Upload batch_3

batch_3/PMC2527490.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527490",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527490\nAUTHORS: Elisabet Selga, Cristina Morales, Véronique Noé, Miguel A Peinado, Carlos J Ciudad\n\nABSTRACT:\nBackgroundMethotrexate is one of the earliest cytotoxic drugs used in cancer therapy, and despite the isolation of multiple other folate antagonists, methotrexate maintains its significant role as a treatment for different types of cancer and other disorders. The usefulness of treatment with methotrexate is limited by the development of drug resistance, which may be acquired through different ways. To get insights into the mechanisms associated with drug resistance and sensitization we performed a functional analysis of genes deregulated in methotrexate resistant cells, either due to its co-amplification with the dhfr gene or as a result of a transcriptome screening using microarrays.MethodsGene expression levels were compared between triplicate samples from either HT29 sensitive cells and resistant to 10-5 M MTX by hybridization to the GeneChip® HG U133 PLUS 2.0 from Affymetrix. After normalization, a list of 3-fold differentially expressed genes with a p-value < 0.05 including multiple testing correction (Benjamini and Hochberg false discovery rate) was generated. RT-Real-time PCR was used to validate the expression levels of selected genes and copy-number was determined by qPCR. Functional validations were performed either by siRNAs or by transfection of an expression plasmid.ResultsGenes adjacent to the dhfr locus and included in the 5q14 amplicon were overexpressed in HT29 MTX-resistant cells. Treatment with siRNAs against those genes caused a slight reduction in cell viability in both HT29 sensitive and resistant cells. On the other hand, microarray analysis of HT29 and HT29 MTX resistant cells unveiled overexpression of caveolin 1, enolase 2 and PKCα genes in resistant cells without concomitant copy number gain. siRNAs against these three genes effectively reduced cell viability and caused a decreased MTX resistance capacity. Moreover, overexpression of E-cadherin, which was found underexpressed in MTX-resistant cells, also sensitized the cells toward the chemotherapeutic agent. Combined treatments targeting siRNA inhibition of caveolin 1 and overexpression of E-cadherin markedly reduced cell viability in both sensitive and MTX-resistant HT29 cells.ConclusionWe provide functional evidences indicating that caveolin 1 and E-cadherin, deregulated in MTX resistant cells, may play a critical role in cell survival and may constitute potential targets for coadjuvant therapy.\n\nBODY:\nBackgroundColorectal cancer is the third most common form of cancer and the second leading cause of cancer-related death in the Western world. Colon cancer causes 655,000 deaths worldwide per year [1]. Therapy is usually through surgery, followed in many cases by chemotherapy, which is used to slow tumor growth, to shrink tumor size and to reduce the likelihood of metastasis development.Chemotherapy effectiveness in cancer cells is compromised by the achievement of drug resistance. Therefore, gaining insight into the mechanisms underlying drug resistance is basic to develop more effective therapeutic approaches. Morales et al. [2] hypothesized that the genetic features related with the progression pathway in colorectal cancer may condition its chemoresistance capability. In fact, it has been described that the tumor's ability to survive, grow and metastasize is conditioned by its genetic and phenotypic heterogeneity [3].Methotrexate (MTX) is an antimetabolite and antifolate drug used in treatment of cancer and autoimmune diseases. MTX competitively and reversibly inhibits dihydrofolate reductase (DHFR), an enzyme that participates in folate metabolism, and essential for DNA synthesis and cell growth [4]. MTX is used for the treatment of lymphoblastic leukemia, lymphoma, osteosarcoma, breast cancer, and head and neck cancer [5]. Treatments combining MTX and other drugs are used in colorectal cancer [6-8]. However, MTX resistance can be easily acquired through different ways, although amplification of the target gene (dhfr) has been shown to be the most important mechanism of resistance in cultured cells [9-11]. Indeed, amplification of 5q12-14 regions, where dhfr is located, has been described in MTX-resistant HT29 cells [2].In the present study, we wanted to identify genes implicated in MTX resistance in HT29 colon cancer cells and to explore their relative contribution to this phenotype. We analyzed the differential gene expression between MTX-resistant and MTX-sensitive HT29 cells using oligonucleotide microarrays containing the full human genome. Changes in the DNA content between both cell lines were also determined. We showed a role for specific differentially expressed genes in MTX resistance. Using siRNAs against caveolin 1, enolase 2 and PKCα or plasmid overexpression for E-cadherin, a clear chemosensitization toward MTX was observed.MethodsCell CultureHuman colon adenocarcinoma cell line HT29 was routinely grown in Ham's F12 medium supplemented with 7% fetal bovine serum (FBS, both from Gibco) at 37°C in a 5% CO2 humidified atmosphere. Cells resistant to 10-5 M MTX, which corresponds to a 1000-fold increase in resistance with respect to the sensitive cells, were previously obtained in the laboratory [12] upon incubation with stepwise concentrations of MTX (Lederle) and were rutinely grown in selective DHFR medium (-GHT medium, GIBCO) lacking glycine, hypoxanthine and thymidine, the final products of DHFR activity. This medium was supplemented with 7% dialyzed fetal bovine serum (GIBCO).MicroarraysGene expression was analyzed by hybridization to the GeneChip® Human Genome U133 PLUS 2.0 from Affymetrix, containing over 47,000 transcripts and variants. Total RNA for oligo arrays was prepared from triplicate samples of both HT29 sensitive and resistant cells using the RNAeasy Mini kit (Qiagen) following the recommendations of the manufacturer. Labeling, hybridization and detection were carried out following the manufacturer's specifications. The data discussed in this publication have been deposited in NCBIs Gene Expression Omnibus [13] and are accessible through GEO Series accession number GSE11440.Microarray data analysisQuantification was carried out with GeneSpring GX software v 7.3.1 (Silicon Genetics), which allows multi-filter comparisons using data from different experiments to perform the normalization, generation of restriction lists and functional classifications of the differentially expressed genes. Normalization was applied in two steps: i) \"per Chip normalization\" by which each measurement was divided by the 50th percentile of all measurements in its array; and ii) \"per Gene normalization\" by which all the samples were normalized against the median of the control samples (HT29 sensitive cells). The expression of each gene was reported as the ratio of the value obtained after each condition relative to the control condition after normalization of the data. Then, data were filtered using the control strength, a control value calculated using the Cross-Gene Error Model on replicates [14] and based on average base/proportional value. Measurements with higher control strength are relatively more precise than measurements with lower control strength. Genes that did not reach this value were discarded. Additional filtering was performed to determine differentially expressed genes. A restriction t-test p-value of less than 0.05 including multiple testing correction (Benjamini and Hochberg false discovery rate) was applied. The output of this analysis was then filtered by fold expression, to specifically select those genes that had a differential expression of at least 3-fold. The 375 transcripts included in this list can be viewed in Additional file 1.RT-Real-Time PCRmRNA levels of the different selected genes were determined by RT-Real-time PCR. Total RNA was extracted from cells (4 × 106) using Ultraspec™ RNA reagent (Biotecx) following the recommendations of the manufacturer. Complementary DNA was synthesized in a total volume of 20 μl from RNA samples by mixing 500 ng of total RNA, 125 ng of random hexamers (Roche), in the presence of 75 mM KCl, 3 mM MgCl2, 10 mM dithiothreitol, 20 units of RNasin (Promega), 0.5 mM dNTPs (AppliChem), 200 units of M-MLV reverse transcriptase (Invitrogen) and 50 mM Tris-HCl buffer, pH 8.3. The reaction mixture was incubated at 37°C for 60 min and the cDNA product was used for subsequent Real-time PCR amplification using SYBR Green. A standard 20 μl reaction contained 25 ng of the cDNA mixture, 0.5 μM of the forward and reverse primers and the SYBR Green Master Mix. Primers used are listed in the Additional file 2.Determination of gene copy numberGenomic DNA from either HT29 sensitive or resistant cells was obtained with the Wizard™ Genomic DNA Purification Kit (Promega) following the manufacturer's recommendations. Five nanograms of DNA were used for Real-Time PCR amplification in a 20 μl reaction containing 0.5 μM of the forward and reverse primers and the SYBR Green Master Mix in an ABI Prism 7000 Sequence Detection System (Applied Biosystems). A list of the primers used is provided as Additional file 3.Functional validationsA) transfection of siRNAs against selected genesHT29 cells (30,000) were plated in 1 ml of -GHT medium and transfection was performed eighteen hours later. For each well, 4 μl of Lipofectamine™ 2000 (Invitrogen) in 100 μl of serum free -GHT medium were mixed in Eppendorf tubes with 100 nM of siRNA in 100 μl of serum free -GHT medium. The mixture was incubated at room temperature for 20 min before addition to the cells. MTX (5 × 10-8 M) was added 48 hours after siRNA treatment and MTT assays were performed [15] after 5 days from the beginning of the treatment. Treatment of HT29 resistant cells was performed following the same protocol using 2 μl of Lipofectamine™ 2000 and 10-5 M MTX. When screening for mRNA levels of the different genes after siRNA treatment, 30,000 cells, either sensitive or resistant, were incubated with increasing amounts of siRNA (1–100 nM) maintaining a 3:1 ratio (μl of Lipofectamine : μg siRNA) and following the procedure previously described. Cells were harvested 48 hours after siRNA treatment for RNA extraction and RT-Real-time PCR. In the combination experiments with siRNAs, 100 nM of each siRNA were diluted in the same eppendorf containing 100 μl of serum free -GHT medium and combined with Lipofectamine™ 2000 as described above. MTX was added as in the single siRNA experiments, mRNA levels were determined and MTT was performed as previously described. In all cases, a non-related siRNA was used as negative control. The treatment was performed as described above and cell viability and mRNA levels for each gene were quantified in parallel. The siRNAs were designed using the software iRNAi. Then, BLAST resources in NCBI were used to assess the degree of specificity of the sequence recognition for these siRNAs. We only selected the siRNAs that reported the target gene as the only mRNA hit. The sequences for the sense strand of all siRNAs used are available in the supplementary material provided (see Additional file 4).B) transfection of an expression plasmid encoding for E-cadherinHT29 cells were seeded into 6-well plates at a density of 3 × 104 cells/well in 1 ml of HAM F12 selective medium. Eighteen hours later, transfections with the expression plasmid for E-cadherin (pBATEM2-CDH) were performed in the presence or in the absence of MTX. The overexpression of E-cadherin was monitored by determining its mRNA levels after 48 h upon transfection. For each well, Lipofectamine™ 2000 was diluted in 100 μl of serum free -GHT medium and was combined with different amounts of the plasmid (500 ng-5 μg) in 100 μl of serum free -GHT medium, always maintaining a 2:1 ratio (μl of Lipofectamine : μg of plasmid). After 20 min at room temperature, the mixture was added to the cells. When combining pBATEM2-CDH transfection and MTX treatment, 5 × 10-8 M MTX was added 48 h after transfection. Cell viability was measured by the MTT assay after 5 days from the beginning of the treatment. Treatment of HT29 resistant cells was performed following the same steps but using 10-5 M MTX.C) co-transfection of siCAV1 and pBATEM2-CDHWhen transfection of siCAV1 and pBATEM2-CDH was performed simultaneously, 100 nM of siRNA and 1 μg of plasmid were diluted together in Eppendorf tubes with 100 μl of serum free -GHT medium and mixed with lipofectamine™ 2000 in 100 μl of serum free -GHT medium (6 μl for the sensitive cells and 3 μl for the MTX-resistant cells). The mixture was incubated at room temperature for 20 min before addition to the cells (3 × 104 cells/well in 1 ml of HAM F12 selective medium, pre-seeded eighteen hours earlier). The mRNA levels after transfection were determined for both genes as previously described and MTT assay was used to determine cell viability.ResultsIdentification of genes deregulated in association with MTX resistanceThe expression profile of the 47,000 transcripts and variants included in the HG U133 PLUS 2.0 microarray from Affymetrix was compared between HT29 sensitive cells and resistant to 10-5 M MTX. GeneSpring GX software v7.3.1 was used to analyze the results. A list of 3-fold differentially expressed genes was generated as described in Methods (Additional file 1). The expression values for genes in this list can be viewed in their corresponding chromosomal position (Figure 1). This overlapping view evidenced a highly overexpressed region in chromosome 5 that covers dhfr and the surrounding loci. The set of upregulated genes in this location included dhfr, zfyve16, msh3, rasgrf2, ssbp2, xrcc4, hapln1 and edil3 (Figure 2), which were selected for further studies. Additional genes that were clearly overexpressed or underexpressed and located in other human chromosomes were also selected according to their function and after literature mining of genes related to drug resistance. The expression levels of most of the selected genes were validated by RT-Real-time PCR (Table 1). The correlation between microarray and qPCR was calculated using the log-transformed values of the fold change obtained for the selected genes, obtaining an r-value of 0.95. To test if changes in the DNA content were responsible for the expression levels of the selected genes in the resistant cells, we determined the copy number for all of them using Real-Time PCR. The results, presented in table 1, showed amplification of all the genes in chromosome 5 flanking dhfr, as well as of mtus1, located in chromosome 8. E-cadherin was the only gene clearly lost among the selected genes.Table 1mRNA levels and copy number determination of differentially expressed genes in HT29 MTX-resistant cells.GenBankGene NameChromosomeCopy Number (Q-PCR)ExpressionGene FunctionMicroarraysValidation (RT-PCR)NM_002961S100A410.85 ± 0.13.7 (p = 5.5e-6)5.68 ± 0.4AngiogenesisBU078629ZFYVE16516.81 ± 2.16.1 (p = 7.7e-6)6.7 ± < 0.1Zinc ion bindingAI144299DHFR516.09 ± 1.47.1 (p = 1.2e-7)11.05 ± 0.5Nucleotide metabolismNM_002439MSH354.97 ± 0.53.9 (p = 5.5e-6)4.23 ± 0.4Missmatch repairAI912976RASGRF2517.76 ± 0.44.6(p = 8.9e-5)6.10 ± 0.5MAPK signalingAF912976SSBP2510.27 ± 0.72.4 (p = 3.4e-3)2.96 ± 0.2ss DNA bindingNM_022406XRCC4517.31 ± 1.17.1 (p = 4.7e-6)8.90 ± 2.3ds break repairU43328HAPLN1511.55 ± < 0.1147 (p = 2.9e-10)1111.9 ± 80.7Cell adhesionAA053711EDIL3514.3 ± 0.7157 (p = 9.1e-8)N/DCell adhesionU17496PSMB860.91 ± < 0.10.1 (p = 0.01)N/DProteasome subunitNM_004666VNN160.84 ± < 0.10.04 (p = 0.01)N/DNitrogen metabolismAU147399CAV171.14 ± < 0.110.9 (p = 1.5e-4)15.00 ± 0.8Integ. plasma membr.BE552421MTUS183.52 ± 0.13.4 (p = 1.8e-6)N/DMitoc. tumor suppressorU05598AKR1C1100.94 ± < 0.14,6 (p = 3.9e-6)6.72 ± 0.7Xenobiotics metabolismNM_001975ENO2120.92 ± < 0.16.0 (p = 4.6e-6)3.90 ± 0.1GlycolisisAK000345DHRS2140.97 ± < 0.10.12 (p = 0.01)N/DOxidoreductaseL08599CDH1160.33 ± < 0.10.19 (p = 0.01)0.15 ± < 0.1Cell adhesionAI471375PRKCA171.05 ± < 0.14.2 (p = 1.7e-5)2.55 ± 0.2Regulation cell cycleBQ003811SLC19A1210.84 ± < 0.10.1 (p = 0.01)N/DCell adhesionNM_001569IRAK1X1.25 ± < 0.10.26 (p = 7.3e-3)N/DIL1 receptor KinaseNM_004135IDH3GX0.85 ± < 0.10.28 (p = 7.3e-8)N/DTCA cycleNM_001183ATP6AP1X0.68 ± < 0.10.3 (p = 0.01)N/DATP biosynthesisTwenty-two genes belonging to the 3-fold differentially expressed list were selected according to their possible relation with drug resistance and/or chromosomal localization. It is shown the GenBank accession number of all genes next to their common name, and their chromosome number. Real-Time PCR was used to determine their copy number, and the expression levels for all them are presented both as the values found in the microarrays (in fold changes relative to the control; t-test p-values included) and as validated mRNA levels (using RT-Real-Time PCR). All experimental results are expressed as fold changes referred to the sensitive cells and values are the mean of triplicate experiments ± SE. The last column indicates the functional categories of the genes. N/D, nondetermined value.Figure 1Chromosomal view of the differentially expressed genes in HT29 MTX-resistant cells. The expression values of genes included in the 3-fold differentially expressed list were viewed in their respective chromosomal location. The names for all the genes studied are depicted on top of their chromosome position. Red is used to color the overexpressed genes and blue is used to highlight the underexpressed genes.Figure 2Localization of dhfr and other genes in chromosome 5 that are overexpressed in HT29 MTX-resistant cells. It is presented a magnification of the region in chromosome 5 where dhfr is located (5q14). The left part is an ideogram of chromosome 5; the right part shows the relative position of all genes studied that are located in this chromosome and that were amplified. The arrows indicate their transcription orientation and the values in parentheses under the names correspond to their respective copy-number validated by Real-time PCR.Effect on MTX sensitivity of siRNAs against genes flanking dhfrTo investigate if the genes that were both overexpressed and co-amplified with dhfr contributed to MTX resistance, their mRNA levels were brought down by means of siRNAs. Effective reduction (≈ 70%) of the respective mRNAs was obtained upon transfection of 100 nM of each single siRNA, both in sensitive and resistant cells. These treatments, though, caused a small reduction in the viability of both cell lines, and addition of MTX to the siRNAs did not sensitize the cells toward the chemotherapeutic agent (data not shown). On the contrary, a siRNA against dhfr mRNA (siDHFR) caused a reduction in cell viability of 30% in HT29 sensitive cells, which was increased up to 90% with the addition of MTX (Figure 3B). mRNA levels after siDHFR treatment were reduced by 70% in this cell line (Figure 3A).Figure 3Determination of mRNA levels and cytotoxicity upon siRNA treatment of HT29 sensitive cells. (A) The mRNA levels of CAV1, ENO2, PKCα and DHFR were determined by RT-Real-time PCR 48 h after treatment of HT29 cells with the indicated concentrations of siCAV1, siENO2, siPKCα, siDHFR and a non-related siRNA. Symbols used for each mRNA are presented as an insert within the figure. (B) Cells were treated with 100 nM of each siRNA as previously described and 5 × 10-8 M MTX was added after 48 h. Cell viability was determined after 5 days from the beginning of the treatment. All results are expressed as percentages referred to untreated cells. Values are the mean of three independent experiments ± SE. A non-related (NR) siRNA was used as negative control.Effect on MTX sensitivity of siRNAs against CAV1, ENO2, PKCα and DHFRAs the knocking down of genes co-amplified with dhfr showed only a slight contribution to MTX sensitivity, we focused in three genes that were clearly overexpressed in the resistant cells and located in different chromosomes, namely caveolin 1(CAV1), enolase 2 (ENO2) and PKCα. We quantified the mRNA levels of these three genes after treatment with different concentrations of the corresponding siRNA using RT-Real-time PCR. The three siRNAs were effective in reducing the mRNA levels of their targets, both in sensitive (Figure 3A) and in MTX-resistant (Figure 4) HT29 cells. One hundred nanomolar was the most effective concentration for all of them, and was used in subsequent experiments. The mRNA levels upon treatment with the siRNA against DHFR are also presented in these series (Figure 3A &4). The mRNA levels of the four genes after treatment with their respective siRNAs in the resistant cells were reduced down almost to the expression levels found for these genes in HT29 sensitive cells (compare Y-axes between Figure 3A &4). A non-related siRNA was used as negative control, and did not produce any significant reduction on the mRNA levels of any of the four genes, either in sensitive or in resistant HT29 cells.Figure 4mRNA levels of CAV1, ENO2, PKCα and DHFR upon siRNA treatment of MTX-resistant cells. Treatments with increasing amounts of each siRNA were performed in MTX-resistant HT29 cells. Forty-eight hours later, mRNA levels for each gene were determined and expressed as percentages of the untreated control (A, CAV1; B, ENO2; C, PKCα and D, DHFR). A non-related (NR) siRNA was used as negative control. Results are depicted taking into account the relative gene expression in the resistant cells (% of the sensitive cells). Values are the mean of three independent experiments ± SE.Viability of HT29 sensitive cells (Figure 3B) was moderately reduced upon treatment with 100 nM siENO2 or siDHFR and treatment with siCAV1 caused a marked reduction of cell viability on its own. No effect on cell viability was observed upon treatment with siPKCα. In all cases, treatment with 100 nM of each single siRNA increased the sensitivity of HT29 cells toward MTX with respect to the control: 80% when using siCAV1; 70% with siENO2; 40% with siPKCα and 90% when siDHFR was used. However, when the same treatments were performed in MTX-resistant cells (data not shown), cell viability was reduced only by 15% when using either siCAV1, siENO2 or siPKCα, and by 25% when siDHFR was used. None of these effects were improved by the combination of siRNAs with MTX. Transfection with 100 nM of a non-related siRNA did not cause any significant reduction on cell viability, either in sensitive or in resistant HT29 cells.Effect of the combination of siRNAs against CAV1, ENO2, PKCα and DHFR on MTX sensitivityAs we had observed a chemosensitization toward MTX in sensitive cells when using individual siRNAs, we performed experiments including the siRNAs against CAV1, ENO2 and PKCα (triple combination) or in combination with siDHFR (quadruple combination) to test if these combinations also increased the sensitivity toward MTX. Treatments combining the three siRNAs (siCAV1, siENO2 and siPKCα) at 100 nM each reduced cell viability by 30% and effectively increased MTX sensitivity by 60% with respect to the control in HT29 sensitive cells (Figure 5A). Addition of 100 nM siDHFR to the previous combination caused a reduction on cell viability of the same degree as the triple combination but increased MTX sensitivity to about 75%. In the case of MTX-resistant HT29 cells, treatments were performed with the same combinations (Figure 5B). The triple combination reduced cell viability by 15% on its own. However, MTX sensitivity was not improved. The quadruple combination did not affect cell viability on its own but caused a reduction of 20% on cell viability when combined with MTX. It was confirmed that the mRNA levels of the four genes were decreased after the siRNA combination treatments in both cell lines (Table 2). Treatments with a non-related siRNA at 400 nM were performed in order to assess the citotoxicity of triple and quadruple combinations and to verify the mRNA levels of all four genes. No effect was observed in any case in either sensitive or resistant cells.Figure 5Effects of combining siRNAs against CAV1, ENO2, PKCα and DHFR on MTX sensitivity. Treatments with combinations of siRNAs at100 nM each were performed both in sensitive (A) and in resistant cells (B). MTX was added after 48 h and cell viability was determined by the MTT assay after 5 days from the beginning of the treatment. The triple combination includes the siRNAs against CAV1, ENO2 and PKCα; the quadruple combination includes the three previous siRNAs plus siDHFR. Results are presented as percentages referred to the untreated cells. Values are the mean of three independent experiments ± SE.Table 2mRNA levels upon treatment with combination of siRNAs against CAV1, ENO2, PKCα and DHFR.ATreatmentCAV1ENO2PKCαDHFRsiCAV1 + siENO2 + siPKCα50.1 ± 4.350.5 ± 4.466.2 ± 2.682.9 ± 2.3siCAV1 + siENO2 + siPKCα + siDHFR37.7 ± 3.759.1 ± 0.847.6 ± 4.747.3 ± 1.2400 nM NR-siRNA96.8 ± 10.498.4 ± 0.596.3 ± 0.3100.6 ± 13.4BTreatmentCAV1ENO2PKCαDHFRsiCAV1 + siENO2 + siPKCα62.7 ± 0.244.20 ± 0.973.2 ± 4.997.7 ± 8.8siCAV1 + siENO2 + siPKCα + siDHFR45.4 ± 0.836.34 ± 4.240.6 ± 2.533.4 ± 3.8400 nM NR-siRNA99.1 ± 9.2100.57 ± 15.9103.2 ± 1.195.5 ± 10.2Treatments combining the siRNAs against CAV1, ENO2 and PKCα at 100 nM each were performed both in sensitive (A) and in resistant (B) HT29 cells. The mRNA levels for all three genes were determined at 48 hours of treatment. DHFR mRNA levels were also determined. Treatments combining 100 nM of the previous three siRNAs plus a siRNA against DHFR were also performed in both cell lines and mRNA levels for the four genes quantified by RT-Real-Time PCR. A non-related (NR) siRNA was used as negative control. Results are expressed as percentages of mRNA referred to untreated cells (mean ± SE) of triplicate experiments.Effect of overexpressing E-cadherin on its mRNA levels, cell viability and MTX sensitivitySince E-cadherin was lost and underexpressed in the resistant cells, it was transiently expressed in HT29 sensitive and resistant cells by means of an expression vector (pBATEM2-CDH). Cells were harvested after 48 hours of treatment. RT-Real-Time PCR was used to quantify E-cadherin mRNA levels in both cell lines (Figure 6A &6B). Transfection of more than 1 μg of the expression vector caused a marked reduction on cell viability in both cell lines. Therefore, 1 μg of plasmid was used in all subsequent experiments. Overexpression of E-cadherin was performed in HT29 sensitive cells in the absence or in the presence of 5 × 10-8 M MTX. This treatment increased by 50% the effect of methotrexate (Figure 6C), thus providing evidence that loss of E-cadherin can confer increased resistance of HT29 cells toward MTX. The same approach was used with HT29 resistant cells, in combination or not with 10-5 M MTX. Overexpression of E-cadherin reduced by 10% cell viability of the resistant cells and only a small improvement was observed when combining E-cadherin overexpression with MTX treatment (Figure 6D).Figure 6Determination of mRNA levels and cytotoxicity upon overexpression of E-cadherin. Cells were treated with increasing amounts of an expression plasmid encoding for E-cadherin (pBATEM2-CDH). Forty-eight hours after the treatment, mRNA levels were determined in HT29 sensitive (A) and MTX-resistant cells (B). Three independent experiments were performed and results are expressed as percentages referred to untreated cells. Values are the mean ± SE. Simultaneous experiments of cells transfected with pBATEM2-CDH were treated with MTX 48 h after transfection, and viability of both sensitive (C) and MTX-resistant cells (D) was assessed after 5 days from the beginning of the treatment by the MTT assay. The mean value ± SE of three independent experiments is depicted. An empty plasmid was used as negative control both for mRNA levels and cytotoxicity determination.Effect of co-transfection of siCAV1 and pBATEM2-CDH on MTX sensitivityE-cadherin has been shown to be an important permissive element in defining the functions of CAV1 [16]. Thus, we performed co-transfection experiments to reduce the mRNA levels of CAV1 and to overexpress E-cadherin simultaneously. mRNA levels after co-transfection were determined in both cell lines (Table 3). As observed in figure 7A, altering the mRNA levels for the two genes reduced the viability of HT29 sensitive cells by almost 40%. Moreover, addition of MTX to the previous combination further reduced cell viability by 90%. Importantly, when performing these co-transfection experiments in HT29 resistant cells, cell viability was reduced by 80%, although in this instance MTX did not improve the effect (Figure 7B).Table 3mRNA levels upon treatment with siCAV1 and pBATEM2-CDH.TreatmentCell lineCAV1E-cadherinsiCAV1 + pBATEM2-CDHHT29 sensitive21.8 ± 1.9252.3 ± 3.5siCAV1 + pBATEM2-CDHHT29 resistant25.6 ± 0.2199.7 ± 16.9Transfection experiments combining the siRNA against caveolin 1 (siCAV1) and the expression plasmid for E-cadherin (pBATEM2-CDH) were performed both in HT29 sensitive and MTX-resistant cells. Forty-eight hours later, mRNA levels for the two genes were determined in both cell lines by RT-Real-Time RCR. Results are expressed as percentages of mRNA referred to untreated cells (mean ± SE) of at least 3 independent experiments.Figure 7Effect of combining the siRNA against CAV1 and the expression plasmid for E-cadherin. One hundred nanomolar siRNA against CAV1 and 1 μg of the expression plasmid for E-cadherin were transfected in both sensitive (A) and resistant (B) HT29 cells. MTX was added 48 hours after transfection and the MTT assay was used to determine cell viability. Results are expressed as percentages referred to untreated cells. Values are the mean of three independent experiments ± SE.DiscussionIn the present study, genes differentially expressed in HT29 colon cancer cells resistant to MTX were identified and their relative contribution to this phenotype evaluated. We observed a cluster of genes flanking the dhfr locus in chromosome 5 that were overexpressed in MTX-resistant HT29 cells. Two of the genes included in this cluster, MSH3 and XRCC4, are known to be involved in DNA repair [17-19]; other two, RASGRF2 and SSBP2, have been related to signaling pathways [20-22]; and EDIL3 has been suggested to prevent apoptosis and to promote cell proliferation [23,24]. Despite the confirmation of the co-amplification of all these genes with dhfr in the resistant cells, we did not observe a clear sensitization toward MTX when reducing their respective mRNA levels by means of iRNA technology. Our observations indicate that the increase in copy-number and the resulting upregulation of the studied genes in 5q14 may be a consequence of dhfr amplification more than an adaptation of the cells to MTX resistance. Indeed, many mammalian species (mouse, rat, bull, cock, dog and chimpanzee) show this set of genes in the same order around dhfr as in human chromosome 5 (using the MapViewer at NCBI), indicating a conserved pattern of gene organization. In keeping with this, its overexpression in the resistant cells could have been useful to improve some cellular processes that might facilitate survival. However, as shown in this work, the increase in copy number of this set of genes does not favor MTX resistance. Thus, we decided to search for other differentially expressed genes (CAV1, E-cadherin, ENO2 and PKCα) that had been previously related with resistance or with colon cancer and to evaluate their relative contribution in our cell system.Enolase 2 (ENO2) is induced by hypoxia, an intrinsic condition of tumors. Moreover, ENO2 is a glycolysis-related gene that has been described to play an important role in tumorogenesis of colorectal cancers [25]. Indeed, ENO2 is upregulated in a variety of cancers [26-28] and alpha-enolase is significantly upregulated in a metastasic colon cancer cell line, suggesting a possible association with the metastasic process in vitro and in vivo [29]. Indeed, we observed a notable contribution of ENO2 to MTX resistance when treating the sensitive cells with siENO2.Both the α-isozyme of PKC and caveolin 1 has been described to be associated with multidrug resistance [30,31], and thus represent good targets to be analyzed. PKCα phosphorylates different proteins, which triggers a wide variety of cellular responses including proliferation, differentiation, membrane transport, gene expression and tumor promotion [32,33]. Chemical inhibitors of PKC activity have been proposed as resistance modulators in MTX chemotherapy [34]. Furthermore, decreasing PKCα mRNA levels attenuates the MDR phenotype in tumor cells [35] and increases the sensitivity to anticancer drugs, both in vitro [36-38] and in vivo [39]. These observations are in accordance with our result showing that the decrease of PKCα mRNA levels by means of iRNA technology causes a sensitization of the cells toward MTX. Caveolin 1 (CAV1), the principal component of caveolae, has been associated with progression of colon and breast carcinomas [40,41] and with enhanced invasiveness in lung adenocarcinoma cells [42]. Although suggested as tumor suppressor gene, and downregulated in some oncogene-transformed and tumor-derived cells [31], overexpression of CAV1 has been found in prostate and esophageal cancer [43-45]. Moreover, re-expression of CAV1 at latter stages of tumor development has been described in human and mouse prostate adenocarcinomas [41], a scenario that could resemble chemotherapy resistance. Indeed, Bender et al. [46] found significantly higher levels of CAV1 in MTX resistant HT29 clones. We have confirmed the implication of CAV1 in MTX resistance in our HT29 cell line.Nevertheless, as Benimetskaya and collaborators observed with PKCα [47], downregulation of a gene alone may be insufficient to completely chemosensitize the cells. Therefore, we considered a combination therapy in order to improve MTX sensitivity. As shown in figure 5, the combination of siRNAs against CAV1, ENO2 and PKCα sensitizes the cells toward MTX, and the effect is improved by the additional downregulation of DHFR. The effects of the triple or the quadruple combinations, however, are not the sum of the effects caused by each single siRNA. This probably reflects the difficulty of transfecting more than one siRNA at 100 nM each. Indeed, the mRNA levels for the four genes after the combination treatment were not as reduced as with the single treatments. In the case of all treatments performed in the resistant cells, probably the overexpression by amplification of the dhfr locus was powerful enough to mask the effects of the siRNAs used.Not only the overexpression of some genes may cause the resistance phenotype. One of the most underexpressed genes that we confirmed to be clearly lost in our HT29 MTX-resistant cells is E-cadherin. In fact, loss of E-cadherin, frequently observed in epithelial tumors, has been associated with tumor progression [48,49] and is considered a crucial event that favors metastasis and invasiveness [50,51]. In addition, the mRNA levels of E-cadherin in adenocarcinoma are 2-fold lower than in normal colon cells [52]. Thus, there is a functional correlation between E-cadherin levels and malignancy. It has been described an event of loss of heterozygosity at the 16.1q chromosome band in most human prostate cancers, where E-cadherin is located [53], which is associated with tumor grade, advanced clinical stage and poor survival [54]. Our experiments show a decrease of 3-fold in E-cadherin levels in resistant cells and also that a mild overexpression of E-cadherin causes a higher sensitivity toward MTX. One has to be cautious, however, about the expression levels of E-cadherin since an increase of more than 3-fold in any of both cell lines caused a reduction in cell viability. This is in accordance with the experiments of Derksen et al. [55] that suggested that loss of E-cadherin could play a causal role in the acquisition of anoikis resistance, as parental mammary cells lacking E-cadherin survived while re-expression of the gene caused apoptosis [55]. Previous works show that loss of E-cadherin in either skin or mammary epithelium does not induce tumor formation [55]. Thus, an overall view of the events occurring in our HT29 cells resistant to methotrexate is needed.It has been shown that activated PKCα translocates from the nucleus to the membrane [56], where it associates with caveolae [57,58], and regulates the function and formation of such biological structures. PKCα has been described to directly interact with CAV1. The union is performed between the caveolin 1 scaffolding domain peptide and PKCα caveolin 1 binding motif [59]. Further, activation of PKCα by phorbol esters dislocates the enzyme from caveolae. All these observations indicate that PKCα interacts functionally with this membrane structures. Moreover, PKCα has been proposed to be involved in the rearrangement of the cytoskeleton. Masur et al. showed that a high level of PKCα expression plus a low E-cadherin level predicts an elevated migratory activity of colon carcinoma cells, which could be derived more easily to metastasis [56]. Lahn et al. speculate that PKCα overexpression may represent an important cellular event leading to enhanced tumor progression, as they concluded that MCF-7 breast cancer cells transfected with PKCα had reduced expression of E-cadherin and β-catenin, resulting in a loss of cell-cell adhesion and thus in a more aggressive tumor phenotype [38].Specific protein-protein interactions between CAV1 and other proteins have been proposed to regulate cell signaling [57,60]. Indeed, CAV1 is known to control cell proliferation and viability by inhibiting expression of survivin, a member of the IAP (inhibitor of apoptosis) family via a transcriptional mechanism involving the β-catenin-Tcf/Lef-1 pathway [16]. One of the possible locations of β-catenin is within a complex with E-cadherin in the adherence junctions, specialized cell-cell adhesion sites that link the cadherin molecules to the actin microfilaments [61]. E-cadherin promotes co-localization and co-imunoprecipitation of CAV1 with β-catenin, as well as inhibition of β-catenin-Tcf/Lef-1 dependent transcription of a wide variety of genes regulated by this pathway, among which survivin is found. However, the ability of CAV1 to regulate survivin expression and cell proliferation is severely impaired in metastasic cancer cells lacking E-cadherin [16]. If E-cadherin is lost, β-catenin is not retained in the plasma membrane and can be then translocated into the nucleus [62], thus activating Tcf/Lef-1 transcription factors-mediated expression of genes implicated in cell proliferation and tumor progression [50]. E-cadherin has been shown to be an important permissive element in defining the functions of CAV1, since several characteristics potentially relevant to CAV1 function as a tumor suppressor are compromised in E-cadherin-deficient HT29 cells [16]. A diagram showing all these relations is presented in Figure 8.Figure 8Scheme for HT29 sensitive and MTX-resistant cells. A diagram is presented showing the compartment localization of CAV1, E-cadherin (E-CDH) and β-catenin (β-CAT) in HT29 sensitive cells (with E-cadherin) and MTX-resistant cells (without E-cadherin, gene loss) and the effects caused by this differential situation. In the sensitive cells, β-catenin is located in the adherence junctions within a complex with E-cadherin. CAV1 co-localizes with β-catenin in these complexes, interfering in β-CAT signaling. If E-cadherin is lost, as in the resistant cells, β-catenin is not retained in the plasma membrane and then can be translocated into the nucleus, thus activating Tcf/Lef-1 transcription factors-mediated expression of genes implicated in cell proliferation and tumor progression.Our results show that HT29 cells can be well sensitized toward MTX by simultaneous treatment with siCAV1 and pBATEM2-CDH. Importantly, we can revert the resistant scenario by reducing the levels of caveolin 1 and by overexpressing E-cadherin simultaneously in the resistant cells, demonstrating the roles that play both genes in MTX resistance.ConclusionWe demonstrate that, aside from dhfr, the contribution of the 5q14 co-amplified genes to MTX resistance is small in HT29 colon cancer cells. On the other hand, we have identified genes deregulated in MTX resistant cells and have demonstrated a role for caveolin 1, E-cadherin, enolase 2 and PKCα in MTX resistance. Very importantly, the concomitant knocking down of CAV1 with overexpression of E-cadherin in HT29 resistant cells markedly reduced cell viability.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsES carried out the siRNA and plasmid transfections, the determination of mRNA levels and associated cytotoxicity after these treatments, and drafted the manuscript. CM carried out the mRNA levels validations, the copy number determinations and helped to draft the manuscript. VN helped with data interpretation and to draft the manuscript, critically revising it. MAP participated in the design of the study and in its coordination, and helped to revise the manuscript. CJC conceived the study, participated in microarray data analyses and drafted the manuscript. All authors read and approved the final manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:Supplementary MaterialAdditional file 1Table of genes differentially expressed by 3 fold. Excel file containing the list of 3-fold differentially expressed genes generated using GeneSpring software v 7.3.1. It includes the GenBank numbers of all genes, their respective common names and the associated description. The chromosomal localization of the 375 entries and the fold change values relative to the control are provided. A final column informing about one of the Gene Ontology categories to which the genes belong according to GeneSpring is included. The differentially expressed transcripts corresponding to open reading frames, transcribed sequences, cDNA clones or hypothetical genes were deleted.Click here for fileAdditional file 2Primers used to validate mRNA levels of selected genes. PDF file with sequences for the primers used to quantify the mRNA levels of genes studied, next to their common name and the number of the chromosome where they are located.Click here for fileAdditional file 3Primers used to determine the copy-number of selected genes. PDF file with common names, chromosome number and the sequences of primers used to amplify their respective DNAs for all selected genes.Click here for fileAdditional file 4Sequences for the sense strand of all siRNAs used. PDF file where the sequences for the sense strand of all the siRNAs used are provided next to the names used to designate all them and the genes they are directed against.Click here for file\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2527492.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527492",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527492\nAUTHORS: YF Rong, WH Lou, DY Jin\n\nABSTRACT:\nA 60-year-old woman presented with vague abdominal pain for one week was referred to pancreatic tail carcinoma accompanied with splenic metastasizes. She came to our hospital for further treatment. Ultrasonography and abdominal computed tomography (CT) revealed a pancreatic tail tumor with splenic metastasizes. There was no history of tuberculosis. Laparotomy was performed because pancreatic tail carcinoma with splenic metastasizes was highly suspected. Indurated mass in the pancreatic tail and sporadic metastasizes in the spleen had been found during the surgery. The pancreatic tail and the spleen were removed and proved to be tuberculosis on histological examination of a frozen section. The patient was given antituberculosis therapy and is now getting well. Tuberculosis should be considered in the differential diagnosis of pancreatic masses. The response to antituberculosis treatment is very favorable.\n\nBODY:\nBackgroundPancreatic tuberculosis (TB) is considered to be extremely rare. However, in the past 20 years, the report of the pancreatic TB has increasing. Most cases of pancreatic TB are diagnosed only after tissue pathology via biopsy or exploratory laparotomy. Because almost all cases of pancreatic TB are effective to antituberculosis management, highlighted awareness of the incident of pancreatic tuberculosis could help to manage this disease. Here we present a case of TB in the pancreatic tail and spleen mimicking pancreatic carcinoma with splenic metastasizes in a 60-year-old woman from China.Case representationThe patient was a 60-year-old woman. She had a past medical history significant for appendectomy, cholecystectomy, left ovariotomy, left breast fibromectomy, left parotectomy and lung nodal with the therapy of steroid. She initially went to the outside hospital on Jan 22, 2007, with complaints of upper abdominal pain. She described her pain as nonconstant and variable in severity. She denied weight loss, fever, chills, nausea, or vomiting. Her physical examination at the time was unremarkable and included normal respiratory, cardiovascular, and abdominal examinations. Her routine laboratory test values were normal, including liver function tests and tumor mark tests. Abdominal ultrasonography scan showed that sporadic metastasizes in the spleen and the common computed tomography (CT) scan showed the malignant tumor in the tail of the pancreas with metastasizes in the spleen.She was referred to our department for surgical treament on Jan 25, 2007. At this time, her symptoms and physical examination results remained unchanged. Hemoglobin: 127 g/dL, white blood cell count: 4,300/mm3, platelets: 147 × 103/mm3, AST: 26 U/L, ALT: 15 U/L, total bilirubin: 14.3 μmol/L, serum creatinine: 76 μmol/L, CEA: 1.29 ng/dL, CA 199:7.1 U/L (normal < 37 U/L). HIV test was negative. A CT scan (performed as a three-dimensional multidetector scan) revealed a low-density mass in the tail of the pancreas with metastasizes in the spleen (Fig 1A and 1B). There was no evidence of ascites, and all of the perihepatic and peripancreatic visceral vessels were not invaded. And the Chest X-ray had no positive founding (Fig 2). Exploratory laparotomy was performed because pancreatic tail carcinoma with splenic metastasizes was highly suspected.Figure 1CT scans of the pancreas. CT scan of pancrease demonstrating a mass in the pancreatic tail () and metastasizes in the spleen (→).Figure 2Chest X-ray. Chest X- ray shows no TB signs.At abdominal exploration on Jan 29, 2007, there were multiple nodules present in the pancreatic tail and spleen (with the diameter of 0.5–2 cm). So we excised the pancreatic tail and the spleen. Once the specimen had been removed, it was submitted to histological examination of a frozen section. The specimen sent to pathology revealed granulomas with caseating necrosis with Langhan's giant cells, suspicious for TB. Given the information from frozen section, the decision was made to close the abdomen without lymph nodes clearance.All pathology specimens were routinely processed (10% formalin), paraffin embedded, and stained with hematoxylin and eosin. Histopathologic examination demonstrated multiple granulomas with caseating necrosis, highly suspicious for TB (Fig 3A and 3B).Figure 3(A and B) Histological examination. Fig 3A low power view of the pancreas demonstrating granulomatous inflammation () and Langhans' giant cells (→) (H & E stain, 10×); Fig 3B high power view of the same field of Fig 3A (H & E stain, 20×).After the operation, the patient was placed in appropriate isolation with strict contact and droplet precautions. The postoperative course was uncomplicated, and the patient was discharged on postoperative day 7 in stable condition. Before her discharge, her case was reported to the Shanghai Health Department and send to the lazaretto for the further therapy of the TB. She has been followed and is currently doing well, having completed her anti-TB therapy.DiscussionTB is caused by the pathogen M. tuberculosis, a rod-shaped aerobic bacterium notable for its acid-fast staining properties which is still a common disease worldwide. The incidence, prevalence, and mortality of the disease vary greatly among different nations, and notably all forms of TB are continuing to increase in all regions [1]. Possible explanations include increased cases of HIV, expanded use of immunosuppressant therapy, globalization of the world's population, and increased transmission in environments such as prisons, homeless shelters, other reasons like evolutionary changes in the biology of the bacterium, drug resistance and so on[2].Although TB often occurs in lung, primary abdominal TB is not uncommon. In 1986, for example, of the 22,768 reported cases of TB in the United States, only 0.58% was found to be located in the peritoneum [3]. However the prevalence of abdominal TB in developing countries has been estimated to be as high as 12% [4]. Although one might expect abdominal TB to be accompanied with active pulmonary TB, only 6–38% had this association [5]. TB does easily disseminate to the gastrointestinal tract, liver, spleen and mesenteric lymph nodes; however the involvement of pancreatic TB is rare. In 1944 Auerbach first reported TB mimicking pancreatic cancer [6]. He described 1656 autopsies of tuberculous patients and identified 297 cases of miliary TB. Only 14 cases had direct pancreatic involvement that may have mimicked neoplasia. Franco-Paredes and colleagues reported two cases of pancreatic TB and reviewed the current literature involving pancreatic TB among nonimmunosuppressed individuals [7]. The authors found between 1980 and 2002 that 50 cases of pancreatic TB had been reported. Thirteen of these cases were categorized as pancreatic masses, mimicking pancreatic carcinoma. While Eric S reported an additional 25 cases of pancreatic TB presenting as discrete pancreatic masses and 42 cases of peripancreatic TB mimicking pancreatic masses in nonimmunosuppressed individuals [2]. In Chinese language literature there were more than 70 cases of pancreatic TB having been described [8-14].The diagnosis of pancreatic TB often proves to be extremely challenging. This is in part because the presentation of abdominal TB is slow and insidious, with nonspecific signs and symptoms. In the present case, for example, vague upper abdominal pain was the only presenting symptoms. Those with immunosuppression status such as HIV positive patients or who live in endemic areas are more likely to develop pancreatic and peripancreatic TB. However in the absence of either of these characteristics, it would be difficult to identify patients with pancreatic TB. In China Feng Xia et al. reviewed literatures and revealed several clinical characteristics of pancreatic TB as follows:1) pancreatic TB is mostly suffered in young people, especially female, while pancreatic tumor is most common in old person; 2) some patients have a history of TB in past, and most often come from areas having high incidence of active tuberculosis; 3) the patients often present with epigastric pain, fever and weight loss; 4) ultrasound and CT scan show pancreatic mass and peripancreatic nodules, some with focal calcification[8].Abdominal TB presents with nonspecific signs and symptoms, making reliance on symptomatology alone for diagnosis insufficient. The diagnostic techniques used for pancreatic TB can be divided into two types: noninvasive and invasive. Noninvasive techniques rely mainly on CT. Sinan et al. reviewed the CT characteristics of pancreatic TB[15]. The most common features were peritoneal involvement and lymphadenopathy. Pancreatic involvement typically appears as an enhancing hypodensity mass, with irregular borders, occasionally mimicking pancreatic carcinoma. In contrast to noninvasive techniques, invasive diagnostic techniques can be used to obtain tissue for pathologic examination and thus are more reliable diagnostic tools. Techniques for biopsy include endoscopic US-guided biopsy, CT/US-guided percutaneous biopsy, and surgical biopsy (open or laparoscopic). Unfortunately, in most cases in the literature, the diagnosis of peripancreatic TB was made only after exploratory laparotomy, as in the present case.For a definitive diagnosis of peripancreatic TB, microbiological and/or histologic confirmation is needed. In the review of Eric S and colleagues suggested that direct smear for the detection of acid-fast organisms is less sensible than combined visual and histopathologic diagnosis [2]. Once the tissue diagnosis has been made, the management of TB rest on the medical treatment. Medical management of pancreatic TB generally consists of isoniazid and rifampin, with pyrazinamide and ethambutol added for severe or resistant cases. In addition to anti-TB treatment, prevention of bacterium spread to other individuals is essential in the management of the disease. While in the hospital, the patient must be quarantined and hospital staff must wear appropriate protective clothing to protect themselves and prevent spread.ConclusionThe present case illustrates pancreatic tail and splenic TB mimicking pancreatic malignant with splenic metastasizes. Once the correct diagnosis was made, our patient was started on a four-drug regimen of anti-TB drugs. However the diagnosis of pancreatic TB often is extremely difficult. This is because patients present with nonspecific symptoms and signs. In addition, CT scan does not reliably distinguish pancreatic TB from pancreatic adenocarcinoma, especially the current case with splenic TB. In present patient we did not pay much attention to the history of taking steroid with an immunosuppressed status. In conclusion, pancreatic TB is rare and the diagnosis is challenging. However if doctors are aware of its clinical features and conduct more diagnostic modalities including CT scan and ultrasound-guided FNA or laparoscopic biopsy, diagnosis of pancreatic tuberculosis without laparotomy is possible and the disease can be effectively treated with antituberculous drugs.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsYFR was a major contributor in writing the manuscript. WHL and DYJ performed the surgery and give consultation to the manuscript. All authors read and approved the final manuscript.Consent sectionWritten informed consent was obtained from the patient for publication of this case report and accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal.\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2527499.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527499",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527499\nAUTHORS: Martial Koenig, Jérôme Morel, Jacqueline Reynaud, Cécile Varvat, Pascal Cathébras\n\nABSTRACT:\nWe report the case of a 39-year old male patient who presented with anaphylactoid shock and diffuse bleeding with prolonged activated partial thromboplastin time at the emergency room. The diagnosis of aggressive mastocytosis was suspected and then confirmed by raised tryptase level and mastocytic infiltration of the bone marrow. The outcome was favorable with supportive measures, antihistamine drugs, and imatinib mesylate.\n\nBODY:\nBackgroundMastocytosis is a rare and heterogeneous disease characterized by the presence of excessive numbers of mast cells in various organs, mainly the skin and the bone marrow [1]. Systemic symptoms are related to mast cell degranulation. Although heparin is one of the mediators released by mast cells, spontaneous bleeding has been very rarely reported in mastocytosis.Case presentationOn December 2006 21th, a 39-year-old male patient of North-African origin, non smoker and who did not drink alcohol, with unremarkable family and personal medical history, at the exception of a history of skin rash after intake of aspirin, presented to the emergency room of a general hospital with abdominal pain, vomiting, and a diffuse skin rash. The patient was alert and afebrile. There was no abdominal tenderness. Initially, blood pressure was 170/90 mm Hg and pulse rate was 70/mn, but hemodynamic parameters rapidly deteriorated despite fluid infusion. Laboratory investigations revealed acute renal failure (creatinine 211 μmol/L), hypokaliemia (2.6 mmol/l), hemoconcentration (protidemia 92 g/L), and clotting tests showed prolonged activated partial thromboplastin time (aPTT) (187\" versus 30\" control value) and prothrombin time (17\"2 versus 11\"8 control value). Blood count showed hyperleucocytosis (18 × 109/L) mainly due to a rise in neutrophil count (16 × 109/L) with a normal platelet and eosinophil count. Haemoglobin level was 12.6 g/dL. C-reactive protein was only slightly elevated (14 mg/L). The patient was transferred to the intensive care unit of the university hospital because of anuria and unexplained abnormalities of clotting tests. On admission, diffuse skin rash was still present. The patient presented signs of shock (blood pressure 80/60 mm Hg, pulse rate 130/mn, anuria and agitation). Orotracheal intubation was thus performed, and mechanical ventilation and continuous hemofiltration were started. Epinephrine infusion corrected hemodynamic status, and the skin rash quickly disappeared. Septic or toxic shock were the first hypotheses investigated, but no infection was documented, and there was no argument for disseminated intravascular coagulation, since the platelet count remained normal. Repeated clotting tests showed however an aPTT up to 200\", a prothrombin time raised up to 90\", and anti-Xa activity was 2.5 UI/mL. Fibrinogen was 1.8 g/L, antithrombin level was 56%. These abnormalities persisted despite the infusion of 10 units of fresh frozen plasma, and haemoglobin level dropped to 8.6 g/dl because of diffuse bleeding at the sites of venous puncture. Total body computed tomography showed no cerebral haemorrhage, no organomegaly, but an hematoma of the duodenal wall was noticed. Gastroscopy revealed non-specific oesophagitis without active bleeding. Medical records revealed that the patient had not received any anticoagulant treatment prior or since he was admitted to the hospital, and thus the abnormalities of clotting tests were attributed to an endogenous heparin-like factor production. This hypothesis, combined with initial symptoms of vasoplegic shock, led the hemostasis specialist to suggest to the clinicians the diagnosis of systemic mastocytosis, which was confirmed by subsequent workup, including a serum tryptase level up to 200 μg/L (normal < 13) and a bone marrow biopsy showing multifocal infiltrates of spindle-shaped mast cells [Figure 1]. The patient was initially treated with fresh frozen plasma and red cell transfusions, and then protamine was infused at a rate of 1200 UI/h combined with IV glucocorticoids, enteral H1 and H2 antihistamines, and imatinib mesylate (400 mg/d). aPTT and prothrombin time were normalized within four days. The patient's status allowed his discharge from intensive care unit after 15 days. On clinical examination in the internal medicine unit, urticaria pigmentosa with Darier's sign (urtication reaction at the site of the papulo-macular lesions when scratched) was demonstrated on the trunk. The patient had not noticed these reddish-brown spots before. Complementary workup revealed long bones involvement on radiographics, diffuse bone abnormality on technetium scintography, diminished bone mineral density (lumbar T-score -1.4; femoral T-score -0.8), and the presence of mastocytic infiltrates in oesophageal wall. No skin biopsy was performed. C-kit mutation D816V was not demonstrated. Serum tryptase level was 13.5 μg/L (N < 13) on day 15. A final diagnosis of aggressive systemic mastocytosis was established, and the patient was discharged on day 30 with ranitidine, cetirizine, glucocorticoids, alendronate, and imatinib mesylate (200 mg/d). On his last follow-up visit in June 2008, he remained asymptomatic under the same treatment at the exception of steroids which had been discontinued.Figure 1Bone marrow aspiration showing infiltration with mast cells (*).ConclusionSystemic mastocytosis is a rare disease characterized by abnormal growth and accumulation of mast cells in various organ [1]. It can follow a benign or indolent course, or it may be associated with invalidating or even life-threatening symptoms such as hypotension, syncope, flushing, urticaria, bronchospasm, peptic ulcer disease, diarrhea, malabsorption, osteoporosis, weigh loss and fatigue [1]. Patients with aggressive systemic mastocytosis usually present with enlarged liver, spleen, and lymph nodes, with or without eosinophilia [1]. The diagnosis of systemic mastocytosis is established by demonstrating mast cell infiltration in an involved tissue, particularly the bone marrow, using special staining techniques or flow cytometry, but the measurement of serum tryptase is a good screening test, since almost all patients with systemic mastocytosis have serum tryptase levels exceeding 20 ng/mL [2]. Clinical pattern depends on mast cells burden in different organs and release of clinically relevant mediators such as histamine, leukotrienes, tryptase and heparin [1,2]. Kinetics of blood clotting may be altered due to fibrinogenolytic and anticoagulant activities of tryptase and heparin respectively [3]. Severe bleeding leading to the death of a patient with systemic mastocytosis due to heparin-like anticoagulant has been recently reported [4], and may represent a difficult diagnosis and a therapeutic challenge in the emergency room. The treatment of systemic mastocytosis is mainly focused on avoidance of triggering factors (e.g. physical stimuli such as heat or cold, alcohol, drugs such as aspirin and other NSAIDS) and symptomatic therapy (H1 and H2 antihistamines, proton pump inhibitors, antileukotrienes, anticholinergics, glucocorticoïds, and epinephrine in case of systemic hypotension). In aggressive forms of systemic mastocytosis, treatments such as interferon alpha, cladribin, and imatinib mesylate should to be considered. Imatinib seems to be more effective in patients without the D816V C-kit mutation [2].AbbreviationsaPTT: activated partial thromboplastin time.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsMK took care of the patient in the internal medicine department and wrote the first version of the manuscript. PC took care of the patient in the internal medicine department, revised and edited the manuscript, and is the attending physician of the patient. JR performed the clotting assays and suggested the diagnosis of mastocytosis to the clinicians. JM and CV took care of the patient in the intensive care unit. All authors read and approved the final manuscript.ConsentWe have obtained written, informed consent from the patient for open access publication of this case report and accompanying image. A copy of the written consent is available for review by the Editor-in-chief of this journal.\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2527501.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527501",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527501\nAUTHORS: Derek Ruths, Luay Nakhleh, Prahlad T Ram\n\nABSTRACT:\nBackgroundIn systems biology the experimentalist is presented with a selection of software for analyzing dynamic properties of signaling networks. These tools either assume that the network is in steady-state or require highly parameterized models of the network of interest. For biologists interested in assessing how signal propagates through a network under specific conditions, the first class of methods does not provide sufficiently detailed results and the second class requires models which may not be easily and accurately constructed. A tool that is able to characterize the dynamics of a signaling network using an unparameterized model of the network would allow biologists to quickly obtain insights into a signaling network's behavior.ResultsWe introduce PathwayOracle, an integrated suite of software tools for computationally inferring and analyzing structural and dynamic properties of a signaling network. The feature which differentiates PathwayOracle from other tools is a method that can predict the response of a signaling network to various experimental conditions and stimuli using only the connectivity of the signaling network. Thus signaling models are relatively easy to build. The method allows for tracking signal flow in a network and comparison of signal flows under different experimental conditions. In addition, PathwayOracle includes tools for the enumeration and visualization of coherent and incoherent signaling paths between proteins, and for experimental analysis – loading and superimposing experimental data, such as microarray intensities, on the network model.ConclusionPathwayOracle provides an integrated environment in which both structural and dynamic analysis of a signaling network can be quickly conducted and visualized along side experimental results. By using the signaling network connectivity, analyses and predictions can be performed quickly using relatively easily constructed signaling network models. The application has been developed in Python and is designed to be easily extensible by groups interested in adding new or extending existing features. PathwayOracle is freely available for download and use.\n\nBODY:\nBackgroundReconstructing cellular signaling networks and understanding how they work are major endeavors in cell biology. The scale and complexity of these networks, however, render their analysis using experimental biology approaches alone very challenging. As a result, computational methods have been developed and combined with experimental biology approaches, producing powerful tools for the analysis of these networks. These tools aid biologists in interpreting existing experimental findings, evaluating hypotheses, enumerating possible biological behaviors, and, ultimately, in quickly designing experiments that maximize the amount of useful information gained. By assisting biologists in maximizing the amount of information obtained from their experiments through improved experimental design and more thorough analysis of results, computational tools increase the pace of scientific discovery.Biological network analysis can generally be classified as either structural or dynamic [1]. Structural analysis provides insights into global properties of the network, among them decomposition of the network into functional modules (e.g., [2]), enumeration of signaling paths connecting arbitrary protein pairs (e.g., [3-5]), and the identification of key pathways that determine the behavior of the network (e.g., [2,6-10]). Dynamic methods, on the other hand, simulate the actual propagation of signals through a network by predicting the changes in the concentration of signaling proteins over time. These predictions will be of varying degrees of resolution and accuracy, depending largely on the accuracy and level of detail of the model from which they are produced.The prevailing methods for dynamic analysis involve systems of ordinary differential equations (ODEs) [11,12]. These approaches require kinetic parameters for the individual biochemical reactions involved in the signaling process. This requirement often poses a significant hurdle for researchers as the numerical values of such parameters are difficult to obtain and may be the object of the researcher's project in the first place. In [13], we presented a novel signaling network simulation method which uses a non-parametric Petri net model of network to predict the signal flow under various experimental conditions. Our simulation method uses a novel technique to approximate the interaction speeds and predicts the qualitative behavior of the signaling network dynamics.The advantage of our method over ODEs is the wide availability of connectivity-based models of signaling networks, and the relative speed with which they can be constructed. Numerous databases exist which catalog known signaling interactions (e.g., [14-16]). Thus, the existence and type (activating or inhibition) of an interaction can often be inferred directly from literature and/or these databases. This presents a stark contrast to the kinetic parameters required by ODEs, the numerical values for many of which must be determined experimentally for each experimental condition and cell line of interest [2].In this paper, we present the software tool PathwayOracle, an integrated environment for connectivity-based structural and dynamic analysis of signaling networks, supporting• visualization of signaling network connectivity;• two versions of the simulation method described in [13] where- the first allows prediction of signal flow through a given network for a specific experimental condition, and- the second predicts the difference in signal flow through a given network induced by two different experimental conditions;• enumeration of the paths connecting arbitrary pairs of nodes in the network; and• visualization of experimental concentration data on the signaling network display.In future releases we plan on expanding capabilities in all three areas of analysis – dynamic, structural, and experimental – with a focus on providing effective ways of integrating results from each together.PathwayOracle has been designed in a modular fashion in order to facilitate extension of existing capabilities and the addition of new features.Since PathwayOracle's most distinctive analytical capability involves the signaling Petri net simulator, a new dynamic analysis technique for signaling networks, we first provide an overview of the signaling Petri net modeling approach. Then in subsequent sections, we focus on PathwayOracle and explain the architecture and core concepts underlying the tool and then examine the individual features, how they can be used, and how they compare to existing tools.The Signaling Petri Net SimulatorPetri nets provide a graphical and executable model of processes in which information or material flows among a series of places or entities [17]. A Petri net consists of places, transitions, and tokens (see Figure 1). Quantities of tokens are assigned to individual places. This assignment is called a marking. As Figure 1 illustrates, the network flow is modeled by the reassignment of tokens to individual places in the Petri net in response to transition firings.Figure 1An example of how tokens move among places. In a Petri net, quantities of tokens are assigned to places. In (a), three tokens are assigned to place pA and zero tokens are assigned to place pB. The two places are connected by a transition, t1. The arcs in and out of t1 indicate the direction in which tokens move. When t1 fires, it moves some number of tokens from pA and puts them in pB. In (b), transition t1 has fired and moved two tokens from pA to pB.A signaling Petri net is an extension of the Petri net formalism to model a signaling network. Places are signaling proteins and transitions implement directed protein interactions; each transition models the effect of a source protein on a target protein. The marking of (number of tokens in) protein p at time t is interpreted as the activity-level of that protein – the number of activated molecules of that type. Figure 2 shows the correspondence between a signaling network and a signaling Petri net model.Figure 2An example signaling network and its corresponding Petri net. An example signaling network (a) and its corresponding Petri net (b). Each signaling protein in the network, A, B, and C, is designated as a place pA, pB, and pC. A signaling interaction becomes a transition node and its input and output arcs. Note that the connectivity for an activating edge differs from that of an inhibitory edge.The signaling Petri net simulator models signal flow as the pattern of token accumulation and dissipation within proteins over time in the Petri net. Through transition firings, the source can influence the marking of (the number of tokens assigned to) the target, modeling the way that signals propagate through protein interactions in cellular signaling networks.In order to overcome the issue of modeling reaction rates in the network, signaling dynamics are simulated by executing the signaling Petri net (SPN) for a set number of steps (called a run) multiple times, each time beginning at the same initial marking. For each run, the individual signaling rates are simulated via generation of random orders of transition firings (interaction occurrences). When the results of a large enough number of runs are averaged together, we find that the change in distribution of tokens in the network correlate with experimentally measured changes in the activity-levels of individual proteins in the underlying signaling network. In essence, the tokenized activity-levels computed by our method should be taken as abstract quantities whose changes over time correlate to changes that occur in the amounts of active proteins present in the cell. It is worth noting that some of the most widely used experimental techniques for protein quantification – western blots and microarrays – also yield results that are treated as indications, but not exact measurements, of protein activity-levels within the cell. Thus in some respects, the predictions returned by our SPN-based simulator can be interpreted like the results of a western blot or microarray experiment looking at changes relative to \"control\".During a simulation run, the simulator imposes a strict ordering on transition firing such that it creates a two-time scale simulation. The smaller time scale is discretized as the firing of a single transition. This unit is referred to as the firing time scale. Firing steps are nested within a larger time scale, called time blocks, within which each transition is fired exactly once. The values returned by the simulator are the averaged token-counts for each protein at each time-block (across all runs).Figure 3 provides a small example of a simulation run whose duration is two time blocks. As mentioned previously, within a given time block, each transition fires exactly once. Thus, in the table (Figure 3(c)), there is one column for each transition in each time block. The ordering of the transitions is shuffled in each time block in order to sample a different set of signaling rates within the networks.Figure 3An example signaling Petri net simulation. (a) is the signaling network being simulated. (b) is the signaling Petri net that models that signaling Petri net. The table in (c) provides the markings for the Petri net over the course of a simulation run whose duration is two time blocks. The proteins are given the initial marking shown in the Initial column. Each subsequent column corresponds to a single time step during which one transition fired, producing a new marking of the network. The bold number in each column indicates which protein's marking was affected by the transition that fired in that time step. The red columns – always the last time step in the block – highlight the markings whose values would be averaged and used as part of the final result. These red columns are the sources of the markings that PathwayOracle reports.In the first time block, transition t2 fires first: it reads 2 tokens out of Grb2 and places 2 additional tokens in Ras. Transition t1 fires second, reading 3 tokens out of Grb2. Transition t3 is evaluated last. The final marking for the network, highlighted as the red column in block 1 is used by the simulator as the marking for that block when averaging across runs.At the conclusion of block 2, compare the values highlighted in red in the Initial column and at the end of both blocks. Note how the distribution of tokens have changed over the course of the simulation. Grb2 has the same number of tokens, implying that its activity-level has remained unchanged – this is consistent with the signaling network since no activating or inhibiting edges affect it in the model. AKTs token-count has risen, consistent with the fact that it is only activated in the signaling network. Ras's token-count has fallen which is one plausible behavior of the system since it is activated by Grb2, but inhibited by AKT.ImplementationPathwayOracle is written in Python [18]. The user experience is oriented around visualization of and interaction with three main types of data: the signaling network, markings, and paths. At any given time, one signaling network is open, which is the basis for all analyses. Any simulation or concentration data is loaded and inspected as markings. Currently all static analyses revolve around paths, which are the third data type. In the following subsections, these individual data types and the user interfaces to them are discussed in more detail.The Signaling Network ModelWhile the implementation of our methods use the signaling Petri net model discussed in an earlier section of this paper, we provide a simpler and more convenient representation of the network to the user which omits the internal topology of the transitions and allows the user to specify interactions simply as either activating or inhibiting. Thus, for the remainder of this paper we use the following definition of the signaling network which is consistent with the experience the user will have when working with PathwayOracle. The signaling network connectivity is a directed graph G = (V, E) where• V is the set of nodes, which are signaling proteins and complexes (hereafter referred to collectively as signaling nodes) and• E is the set of edges, which are signaling interactions. Each edge is of one of two types: u → v for activation and u ⊣ v for inhibition.Within PathwayOracle, each signaling node has a name, unique within the network. A signaling edge has no properties besides its type and is only defined by its source and target.In order to facilitate the rapid construction of such signaling network models, we devised a file format called the Connectivity Format. It is capable of expressing both general networks as well as paths. When representing a network in the format, as shown in the example in Figure 4(b), one signaling interaction is written on a line with the formatFigure 4An example of a Network in the Connectivity Format. (a) A graphical representation of a signaling network's connectivity. (b) The signaling network in (a) written in the Network Connectivity Format.u -> v   or   u - | vwhere u is the name of the source signaling node and v is the name of the target signaling node. Each node is taken to represent the active form of the protein it is named for. Thus, from the example above, the interaction PI-3-K→AKT means that the active form of PI-3-K increases the activity-level of AKT whereas the interaction PTEN⊣AKT means that the active form of PTEN decreases the activity-level of AKT. While these types of unparameterized relationships can be represented in SBML, SBML was designed for encoding much more information than just connectivity [19]. As a result, we deemed it appropriate to design a more concise format for our purposes. However, in a future release, PathwayOracle will support loading and saving in the SBML format.At a given point in time, only one signaling network can be open in PathwayOracle. The main window displays a graphical representation of the network. The layout of the network can be modified by dragging nodes or by shift-clicking on edges to create, remove, or move waypoints. These layouts can be saved with the network and loaded again.Signaling Network MarkingsIn signaling networks, signal flow is measured and quantified as the fluctuation of concentrations of various forms of signaling proteins over time. In PathwayOracle, we model concentrations using the concept of a network marking, which was adapted from Petri nets in which it was first used [9].MarkingsIn PathwayOracle, a marking, μ is an assignment of real values to the nodes of a signaling network such that every signaling node receives a value. Earlier, the concept of a marking was introduced as the assignment of tokens to protein places in the signaling Petri net. In a signaling Petri net, tokens are discrete. In PathwayOracle, a marking is an average of the markings from many independent simulation runs, which gives rise to the real, rather than integral values, assigned by the marking.As discussed earlier, the value of the marking of a signaling node, μ(v), can be interpreted as an estimate of the concentration or change in concentration of the active form of the signaling protein v (we call the amount of the active form of the signaling protein its activity-level). The two different versions of the simulator generate markings with these different meanings. The first simulator predicts the signal flow due to an experimental condition and generates markings whose values are taken to represent the actual activity-level of signaling protein present over the assumed basal levels. The second version of the simulator predicts the difference in signaling due to changing experimental conditions. The values assigned by markings produced by this simulator correspond to the change in the activity-level of the protein induced by the change in experimental condition. This will be discussed further in the Results and Discussion section.Marking SeriesIn order to model signal flow, a single marking is not enough since it only provides a single snapshot of concentrations throughout the network. A marking series is an sequence of markings, (μ1, μ2,..., μT) in which the marking μt is a snapshot of the concentration distribution at time step t. Thus, it is possible to see how the activity-level of protein v changed by plotting the values μ1(v), μ2(v),..., μT(v). PathwayOracle provides the ability to do this.PathwayOracle supports loading a marking series dataset from comma-separated value (.csv) files. As shown in Figure 5(a), the file has a header row which specifies, for each column, the name of the molecule whose concentration values will appear in that column. Each subsequent row contains the value assignments for a marking: the second row contains the marking for time step 1, the third row contains the marking for time step 2, and so on.Figure 5Examples of marking series and group file formats. (a) An example marking series dataset in the comma-separated value file format. The first row specifies the signaling proteins whose concentrations were measured. Each row thereafter specifies the concentration for a given time step: row i specifies the concentrations for each signaling protein at time step i - 1. (b) An example marking group dataset in the comma-separated value file format. The first row specifies the signaling proteins whose concentrations were measured. The first column specifies the names for each marking in the group dataset. The numbers in each row specify the concentration measured for each signaling protein in that marking.Marking GroupsIn many experiments, the activity-level of various proteins are sampled at different time points and under different experimental conditions. Since the marking series is not able to represent changes due to different experimental conditions, we introduced the more general concept of a marking group in which each marking can correspond to an arbitrary activity-level distribution. Each marking is given a descriptive label that can be used to identify the conditions under which the activity-level was sampled.Like the marking series, a marking group is loaded from a .csv file. However, unlike the marking series in which each row corresponds to a time step, in the marking group, each row corresponds to an independent marking (experimental condition). As shown in Figure 5(b), the first row is a header row specifying the molecule names for each column, the first column specifies the names for the individual markings (experimental conditions).The Marking ManagerPathwayOracle includes a specific user-interface, the Marking Manager, designed to manage the three different types of markings. The Marking Manager provides a central interface within which it is possible to view all markings loaded and inspect them in ways that are relevant to their type (marking, marking series, or marking group). The specific ways in which markings can be inspected will be discussed further in the Results section.Signaling PathsThe current structural analysis capabilities available in PathwayOracle allow inspection of signaling paths within the network. A signaling path p is a sequence of nodes, (v1, v2,..., vk) where vi ∈ V ∀1 ≤ i ≤ k, and (vi, vi + 1) ∈ E ∀1 ≤ i <k. In this case, we say that node v1 is the source of path p, and node vk is the target of p. Given a path, a variety of statistics may be of interest to the user. Additionally, it may be useful to view the path within the larger network. PathwayOracle provides these capabilites which will be discussed in the Results and Discussion section.Sets of paths can be saved to a file and loaded back into a session. Like networks, paths are also stored in the Connectivity Format. When representing a set of paths, as shown in Figure 6, the full node names and the edge types are written so that all path information is directly available within the file itself. One line contains one path.Figure 6An example of a Path in the Connectivity Format. (a) A graphical representation of two signaling paths. (b) The signaling paths in (a) represented in the Connectivity Format. Each line corresponds to a single signaling path.ResultsPathwayOracle provides a variety of tools for analyzing the structural and dynamic properties of a signaling network based on its connectivity. While its main differentiating feature is the ability to predict signal flow through a network using only the connectivity of the signaling network, PathwayOracle also provides the ability to visualize the network, analyze its connectivity, and inspect concentration-based experimental data.With the exception of the signaling Petri net simulator, PathwayOracle's features can be found in various combinations in other tools. Figure 7 provides a matrix of the features and capabilities of several tools most commonly-used for signaling network analysis. While other tools support a variety of simulation techniques, PathwayOracle, alone, provides non-parameterized simulation capabilities. It is worth noting that the commercial software package CellIllustrator [20] provides Petri net-based simulation capabilities. The difference between CellIllustrator and PathwayOracle Petri net approaches is the extensive set of kinetic parameters required by CellIllustrator in order to simulate a biological system. In this regard, hybrid functional Petri nets, the underlying technology used by CellIllustrator, are not significantly different from ODEs.Figure 7A comparison of features supported by tools commonly used for signaling network analysis. The table shows the features and analytical capabilities supported by different tools commonly used for the analysis of signaling networks. Tools included in the comparison are: CellDesigner [20], CellIllustrator [24], CellNetAnalyze [25], COPASI [22], Cytoscape [21], the System Biology Toolkit for Matlab [26], and PathwayOracle.Another important distinguishing characteristic of PathwayOracle is the combination of features that it supports. Biological network analysis is a multi-faceted process that may involve structural, dynamic, and data analysis in parallel. Whereas other tools tend to focus on one or two of these general areas of analysis, we considered it important for PathwayOracle to incorporate all three in order to provide the researcher a single environment in which all their analysis could be done. In future releases we plan to increase PathwayOracle's support for all three of these directions of investigation: structural, dynamic, and data analysis.In the remainder of this section, we discuss the features currently available in PathwayOracle.Network VisualizationAs in many other computational analysis tools for signaling networks (e.g., [20,21]), an interactive graphical representation of the signaling network connectivity is at the center of the PathwayOracle interface. The main window provides a visualization of the signaling network connectivity. This visualization interface allows the user to edit the layout of the network by clicking on and dragging nodes and by shift-clicking on edges to create, remove, or move waypoints. Waypoints are points that lie on an edge. Holding down shift will display all edge waypoints. Existing waypoints can be dragged to change the path that an edge follows. Right-clicking on a waypoint will remove it. Left-clicking on a straight segment of the edge will create a new waypoint.The network visualization also provides a view onto which path and experimental data analysis may be mapped. As will be discussed in subsequent sections, selected paths may be highlighted in this view and markings from experiments can set the colorings of individual nodes.Network Signal Flow SimulationThe main feature differentiating PathwayOracle from other tools, such as CellDesigner [20] and COPASI [22], is its ability to simulate signal flow using an unparameterized signaling network model. Simulations can be performed in two different ways. In the first (Single Simulation), the simulator predicts the signal flow through the network for a specific experimental condition. In the second (Differential Simulation), the simulator predicts the difference in signal flow due to two different experimental conditions on the same network. These simulation methods themselves are described in [13]. Here we focus on how simulations are configured, run, and analyzed.Whereas the consensus networks typically represent the connectivity in normal cells, many experiments are conducted on abnormal cells in which oncogenic mutations, gene knockous, and pharmacological inhibitors have altered the behavior of various signaling nodes in the network. In PathwayOracle users can model these cell- and experiment-specific conditions by specifying each signaling node as either High, Low, or Free. The High state models any condition under which a protein's activity-level is held high for the duration of the experiment. This may be due to external stimulation or a known mutation in the protein that makes it constitutively active, for example. Similarly, a Low state models any phenomenon that forces a protein to have a persistently suppressed activity-level. This may be due to mutations that render the protein inactive, gene knockouts, or pharmacological inhibitors that force the activity-level of the protein low. In general, most signaling nodes will be Free, which means that their activity-level is unconstrained throughout the simulation. Only those nodes designated as High or Low will have their activity-level fixed for the duration of the simulation.In order for a protein to be held high during the simulation, it is necessary to indicate the initial activity-level that the protein will be elevated to. This is done by specifying the number of tokens that the protein will receive. Since a protein with a High state cannot be inhibited (even if inhibitory edges target it in the actual network), the protein's activity level will never fall below this initial value. The initial value for a High protein is indicated by placing it in parentheses next to the protein's name, as shown in Figure 8. Two other parameters that must be specified for a simulation are:Figure 8The tokenized simulator user interface. (a) The setup window for the tokenized simulator. The simulation is being configured to have two High nodes, EGF and LKB-auto. EGF will be initialized with a token-count of 10, LKB-auto with a token-count of 3. The token-count of AMPK will be zero for the duration of the simulation. (b) The setup window for the differential simulator. Two different scenarios are being compared through simulation: different token assignments are being tried with EGF and LKB-auto, with and without AMPK being fixed low. (c) The plot window for the marking series generated by a simulation. Observe that the signaling nodes whose activity-levels are plotted correspond to those selected in the checklist directly to the left of the plot.• the number of simulation runs to perform and• the number of time blocksThe number of runs sets the number of independent simulations whose time block markings are averaged together to yield the overall simulation markings. In general, using more runs is a tradeoff between reliability of the results and simulation speed. In practice, the number of runs needed is dependent on the signaling network model and should be selected by observing the reproducability of the simulation results. An appropriate number of iterations will be large enough so that for a given experimental condition, the results are very similar across multiple simulations.The time block, as discussed earlier, is a fundamental unit of time in the simulator. The appropriate number of time blocks for which to simulate will vary depending on the size of the signaling network and the scale of the network behavior of interest. Generally it should be selected by running simulations for a variety of time block values and determining which yields the most biologically reasonable activity-level changes for a known protein. While this is a manual process in the current version of PathwayOracle, we are investigating automated methods for estimating the number of time blocks by training against experimental time series data.In PathwayOracle, the setup window for the Single Simulation (see Figure 8(a)) prompts the user for a single experimental condition. The setup window for the Differential Simulation (see Figure 8(b)) prompts the user for two experimental conditions. Both simulators produce a marking series. The tokenized simulation marking series corresponds to the activity-level time series predicted for the specified experimental condition. The differential simulation marking series corresponds to the change in activity-levels over time produced by switching from experimental condition 2 to experimental condition 1. The marking series produced by a simulation can be accessed through the Marking Manager. Choosing to inspect a marking series will present the user with a blank plot. By selecting signaling nodes, the plot is populated by the marking series values for individual nodes over time, as shown in Figure 8(c).While this plot generation capability exists in many other dynamic simulation tools, the simplicity of the model used for simulation and the speed with which a simulation runs set PathwayOracle apart from other tools which require specification of the numerical values of kinetic parameters for each reaction in the network of interest (e.g., [20,22]). PathwayOracle, because of its novel approach, does not have such requirements. It is worth noting, however, where PathwayOracle provides approximations of signal flow, an ODE generates the actual concentration changes using extremely detailed and accurate models of the underlying biochemistry. The simulators in PathwayOracle provide an attractive, time- and resource-saving alternative this more exhaustively parameterized techniques. In particular, PathwayOracle's features will benefit researchers interested in quickly assessing characteristics of signal flow in their network.For some networks, biologists will have partial knowledge of kinetic parameters or of other biological details which the signaling Petri net model does not, at present, consider. By integrating this knowledge into the simulator, it may be possible to improve the simulator's predictions. We identify this as a direction for future investigation. As the signaling Petri net simulator is extended, these new capabilities will be incorporated in future releases of PathwayOracle.Signaling Path AnalysisThe use of the simulators and plotting tools allows the user to observe trends in the activity-level of individual signaling nodes over time. Since the activity-level of a node is determined by the activity-level of other nodes in the network, the activity-level time series of a node may be explained by changes in the activity-level history of nodes upstream of it. In order to investigate such indirect interactions, it is useful to enumerate all the paths leading from a specific protein to the protein of interest. PathwayOracle provides this capability. Additionally, it provides various statistics on the set of paths linking two signaling nodes as well as a classification of the effect of each path as either coherent or incoherent (e.g. [23]). A coherent path is a directed series of interactions that leads from x to y such that an increase in the activity-level of x causes an increase in the activity of y and a decrease in the activity-level of x causes a decrease in the activity-level of y. An incoherent path is a directed series of interactions leading from x to y such that an increase in the activity-level of x causes a decrease in the activity-level of y and a decrease in the activity-level of x causes a increase in the activity-level of y. It is possible to classify a path p as either coherent or incoherent by counting the number of inhibitory edges along p. A path with an even number of inhibitory edges is coherent; a path with an odd number of inhibitory edges is incoherent [5]. This logic is assumed in PathwayOracle. All simple paths (paths without loops) connecting two specified signaling nodes are enumerated by an exhaustive depth-first search. These paths then are classified as either coherent or incoherent, and presented to the user for further inspection in a window similar to the one shown in Figure 9(a). When a path is selected in the results window, it is highlighted in the main window, allowing the user to evaluate it within the context of the complete network (see Figure 9(b)).Figure 9The path interrogation user interface. (a) The result window enumerating the set of all paths between Ras and mTOR/raptor. (b) The main network view showing the selected path highlighted.Experimental Data AnalysisA model of the connectivity of a signaling network makes it possible to identify components of the model that are inconsistent with experimental data or visa versa. PathwayOracle enables this kind of analysis by allowing users to load experimental concentration data and visualize it both as a heatmap (see Figure 10(a)) or superimposed on the network view (see Figure 10(b)). Several other software tools provide similar capabilities (e.g., [21]). In PathwayOracle, experimental concentration data is loaded as a marking group in which a single marking corresponds to a condition for which concentrations were sampled. Figure 10(a) shows a marking group with 24 conditions (rows). The concentration of seven signaling proteins were sampled for each condition. This is the heatmap view for the marking group. When a specific marking in the group is selected, the colors for that marking are applied to the network view. This is particularly useful when assessing whether the experimental data is consistent with the interactions in the model. In Figure 10, the MDA231-B-DMSO2 marking has been superimposed on the network. We can see that RSK has a relatively low concentration despite the high concentration of MAPK. Given that, in the model, RSK is activated by MAPK, this combination of activity-levels seems unlikely to occur. Such an inconsistency suggests that there may be other signaling interactions contributing to the overall activity-level of RSK. Such an insight can help a researcher quickly identify areas where the model or experimental results need to be re-evaluated or improved.Figure 10The marking group user interface. (a) The heat map visualization of a marking group. The selected marking, MDA231-B-DMSO1, is highlighted in blue. (b) The color distribution for the selected marking in the group is applied to the network view in the main window. Note that signaling nodes for which values were not given are not assigned a color on the valid red to green spectrum.Future DirectionsOur goal is to develop PathwayOracle into an integrated and expansive suite of tools that allow the biologist to extract as much information as possible from models of signaling network connectivity and experimental data relating to those models. We consider future directions for PathwayOracle to fall into several categories: network construction, network augmentation, experimental and computational analysis integration, and architecture.One of the benefits of working with connectivity models of signaling networks is the abundance of databases and other online resources that publish connectivity-level data. Future versions of PathwayOracle will have support for querying such databases for connectivity components and, ultimately, for automated connectivity construction based on a set of signaling nodes specified by the user.Analysis of network connectivity and topology is increasingly relevant to biological research. We intend to expand PathwayOracle's structural analysis features to include the ability to search for and identify motifs in the signaling networks.Network connectivity can also be inferred from experimental data, which provides another direction for research and development. By using experimental results to identify inconsistencies between experimental results and the current network model, it may be possible for PathwayOracle to augment the network with new connectivity based on hints supplied by experimental results. At present only experimental concentration data is supported. However, as experiments produce more information beyond concentration profiles of signaling nodes, we plan to expand the experimental data that PathwayOracle can load, visualize, and use as part of network analyses.Experimental results can also provide computational analysis methods information that can improve their final predictions or decompositions. Taking advantage of the additional, potentially obfuscated, information present in experimental results to improve the results returned by computational tools is a major goal for future versions of PathwayOracle.A longer term direction for PathwayOracle is the integration of transcriptional and metabolic network analysis. In the biological systems of interest, the behavior of any one of these networks is dependent on the characteristics of the other two. As a result, developing a complete understanding of signaling, transcriptional regulation, or metabolism depends in part on integrating knowledge from the others. Finally, an ongoing priority in the design of PathwayOracle is its role as an open platform for the development and deployment of new analytical capabilities by other groups. Currently PathwayOracle employes a modular architecture that facilitates easy integration of new functionality. However, in future releases we plan to expose a plugin interface which will make it easier to developers and researchers to develop and deploy tools within PathwayOracle.ConclusionPathwayOracle is an integrated software environment in which biologists may conduct structural and dynamic analysis of signaling networks of interest. PathwayOracle is distinguished from other tools in the field of systems biology by its ability to predict the signal flow through a network using a simplified, connectivity-based model of the signaling network. Simulations are fast and, based on a published study, predictors of signal propagation. This novel simulation capability, combined with support for structural analysis of connectivity between pairs of proteins and for analysis of certain kinds of experimental data make PathwayOracle a powerful asset in the experimentalist's endeavor to gain a more complete understanding of the cellular signaling network.Availability and requirementsProject name: PathwayOracleProject home page: Operating system(s): Platform independentProgramming language: PythonOther requirements: Python 2.4 or higherLicense: GNU GPLAny restrictions to use by non-academics: NoneAuthors' contributionsDR designed and developed the PathwayOracle application, participated in evaluating features for inclusion, and drafted the manuscript. LN participated in application design and feature selection. PTR contributed biological case studies and data for PathwayOracle feature design. All authors read and approved the final manuscript.\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2527503.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527503",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527503\nAUTHORS: Ahmed Alzaraa, Imran Ghafoor, Andrew Yates, Alhad Dhebri\n\nABSTRACT:\nIntroductionSebaceous gland tumours are rare and their presence should be considered as a marker for Muir-Torre Syndrome, alerting to search for an occult malignancy.Case presentationA 43-year-old Caucasian female patient underwent excision of a sebaceous cyst. Histopathology confirmed a sebaceous carcinoma. Further investigations revealed multiple intra-abdominal malignancies. She has been under regular follow-up in the relevant clinics.ConclusionSebaceous carcinoma should be excised completely and followed-up for the detection of possible metastases. Surgical removal of primary or metastatic cancers may be curative and should be attempted wherever possible. It is very important for clinicians not to miss such skin lesions as they may precede the presentation of internal malignancies.\n\nBODY:\nIntroductionMuir-Torre Syndrome (MTS) is defined by the combination of a sebaceous gland tumour and at least one visceral carcinoma occurring in the same individual in the absence of other precipitating factors such as radiotherapy or AIDS. Typical skin lesions associated with this syndrome include sebaceous adenoma, epithelioma and carcinoma.Case presentationA 43-year-old Caucasian woman had an excision of an infected sebaceous cyst from the skin of her left breast in 2007. Histopathology reported an incompletely excised sebaceous carcinoma, suggestive of MTS. The patient was referred to the breast clinic for further assessment. She had an extensive family history of cancers. Her father died of prostate cancer, her mother had a hysterectomy for uterine cancer, her maternal grandmother had uterine cancer, her sister had bowel cancer and her great paternal aunt had breast cancer. Clinical examination revealed a scar in the lower half of her left breast and two sebaceous cysts over her left nipple and left upper arm. She was also tender in the right iliac fossa and was investigated by a gynaecologist. Her chest X-ray and bilateral mammogram were normal. The patient underwent wider excision of the left breast scar and of skin lesions over the left nipple and the left upper arm. Histology reported no residual neoplasia in the excised scar and the skin lesions were benign epidermoid cysts.A computed tomography (CT) scan of the chest, abdomen and pelvis showed bilateral ovarian cysts and diverticular disease of the colon. The tumour markers were raised: CA125: 149 (0–35); CA19-9: 61.8 (0–35). She later underwent total abdominal hysterectomy, bilateral salpingo-oophorectomy and infracolic partial omentectomy. The histology was Stage IC moderately differentiated endometrioid adenocarcinoma of the right ovary, Stage IC endometrioid endometrial adenocarcinoma, and the omentum was normal. She was referred to a geneticist, oncologist, gastroenterologist and dermatologist. The genetic analysis showed (MSH2s.1578DEL). She had a gastroscopy which showed duodenitis and a colonoscopy which was normal. She had chemotherapy followed by radiotherapy.DiscussionCutaneous lesions associated with hereditary cancer syndromes are known as cancer-associated genodermatoses [1]. One of these is MTS, defined by the combination of a sebaceous gland tumour and at least one visceral carcinoma occurring in the same individual in the absence of other precipitating factors such as radiotherapy or AIDS [2]. It was first described by Muir in 1967 and Torre in 1968, and is recognized as a subtype of Lynch Type II Hereditary Nonpolyposis Colon Cancer (HNPCC) [3,4]. A review by Akhtar et al. (1999) identified a total of 205 reported cases in the world literature [5]. The male/female ratio is 3:2 [6]. It is an autosomal dominant disorder with a high degree of penetrance and variable expression and the children of an affected individual have a 50% risk of inheriting the cancer predisposition. The genetic disorder is an inherited germline mutation in one of the DNA mismatch repair (MMR) genes, most commonly MSH2, which eventually leads to microsatellite instability (MSI) [7].Typical skin lesions associated with this syndrome include sebaceous adenoma, epithelioma and carcinoma (sebaceous hyperplasia and nevus sebaceous of Jadassohn are generally excluded). Keratoacanthomas and basal cell carcinomas with sebaceous differentiation can also occur. Sebaceous gland tumours are rare and their presence should be considered as a marker for MTS, alerting to search for an occult malignancy [8]. Skin lesions may precede the presentation of internal malignancies, but often develop later. Fifty-six percent of skin lesions occur after the diagnosis of the first malignancy, 6% occur concomitantly and 22% occur as the first malignancy of the syndrome [5]. The cutaneous lesion may occur as much as 25 years before or 37 years after the internal malignancy. Multiple primary carcinomas at different sites are characteristic of MTS (Table 1), and up to nine visceral cancers have been reported in one individual [9].Table 1Prevalence of cancers associated with MTSCancerPercentage (%)Colorectal80Stomach11–19Hepatobiliary tract2–7Small intestine1–4Brain or central nervous system1–3Endometrial20–60Ovarian9–12Urinary tract4–5SkinIncreased riskColorectal cancer is the commonest visceral neoplasm to occur in MTS, and the most frequent initial cancer [10,11], though not in our case. In common with other forms of HNPCC, colorectal cancers in MTS are usually proximal in location and tend to have a more indolent course than usual colorectal cancers [11]. Fifty-one percent of MTS patients develop at least one colorectal cancer, and multiple colorectal cancers are common [12]. Colonic polyps are found in more than 25% of MTS patients, and are especially prevalent in patients with colorectal carcinoma [12].The second most common site is the genitourinary tract (including endometrium and ovary), representing approximately one-quarter of visceral cancers. A wide variety of other cancers have been reported involving breast, upper gastrointestinal tract, upper respiratory tract including larynx, salivary gland, and haematological malignancies including lymphoma and leukaemia. Intestinal polyps occur in at least one-quarter of patients [12].In families with proven germline mutation, individuals should be offered regular screening examinations. In those who can be demonstrated not to have inherited the germline mutation, cancer surveillance is not necessary [13]. Screening for malignancy at all possible sites is impractical in MTS given the wide range of associated malignancies, and should probably concentrate on the colorectum, female genital tract and possibly the renal tract. In some families, the occurrence of certain other tumours would be an indication for other screening modalities, for example, upper gastrointestinal endoscopy [14]. Cohen et al. [11] suggested that a search for internal malignancies should be undertaken in those with MTS-associated sebaceous gland tumour, those with MTS, and in family members of an MTS patient. They also suggested a surveillance programme for patients with MTS or MTS-associated sebaceous gland tumours including annual clinical examination, carcinoembryonic antigen (CEA), cervical smear, chest radiography, urine cytology, colonoscopy or barium enema every 3 to 5 years, and for female patients, mammography annually or biennially to age 50 and annually thereafter, and endometrial biopsy every 3 to 5 years. Other authors have suggested that colonoscopy should be more frequent in view of the high frequency of colonic cancer and its proximal predominance, and have advocated annual colonoscopy from the age of 25 years [10]. MTS screening may be extensive and is not always performed. A search for mutations in either MSH1 or MSH2 is expensive and time consuming. One could analyse MMR protein expression as a surrogate for assaying for the respective gene mutations [15]. Benign sebaceous tumours and keratoacanthomas can be conservatively treated with excision or cryotherapy. Sebaceous carcinoma should be excised completely and followed-up for the detection of possible metastases. It has been suggested that, because of their relatively good prognosis and non-aggressive course, surgical removal of primary or metastatic cancers may be curative and should be attempted wherever possible [5,7]. The combination of interferon with retinoids (isotretinoin) seems to be of promise in preventing tumour development in MTS. A dosage of 0.8 mg/kg/day may be effective [10].ConclusionSebaceous gland tumours are rare and their diagnoses should suggest the possibility of MTS and prompt a search for associated malignancies, and for the underlying genetic mutation. Family members should be offered screening to detect early cancers. Genetic studies can help identify the inherited molecular defect that causes MTS.AbbreviationsAIDS: Acquired Immune Deficiency Syndrome; CA125: Cancer Antigen 125; CA19-9: Cancer Antigen 19-9; CEA: carcinoembryonic antigen; CT: computed tomography; DNA: deoxyribonucleic acid; HNPCC: Hereditary Non-Polyposis Colon Cancer; MMR: mismatch repair; MSI: microsatellite instability; MTS: Muir-Torre Syndrome.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsAA searched the literature and drafted the manuscript, IG searched the literature, AY evaluated the histology, and AD operated on the patient and edited the manuscript.Figure 1Sebaceous carcinoma showing multivacuolated cells with clear cytoplasm and indented nuclei as evidence of sebaceous differentiation.\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2527517.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527517",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527517\nAUTHORS: Monika Madan, Salomon Amar\n\nABSTRACT:\nBackgroundAccumulating evidence implicates a fundamental link between the immune system and atherosclerosis. Toll-like receptors are principal sensors of the innate immune system. Here we report an assessment of the role of the TLR2 pathway in atherosclerosis associated with a high-fat diet and/or bacteria in ApoE+/− mice.Methods and ResultsTo explore the role of TLR2 in inflammation- and infection-associated atherosclerosis, 10 week-old ApoE+/−-TLR2+/+, ApoE+/−-TLR2+/− and ApoE+/−-TLR2−/− mice were fed either a high fat diet or a regular chow diet. All mice were inoculated intravenously, once per week for 24 consecutive weeks, with 50 µl live Porphyromonas gingivalis (P.g) (107 CFU) or vehicle (normal saline). Animals were euthanized 24 weeks after the first inoculation. ApoE+/−-TLR2+/+ mice showed a significant increase in atheromatous lesions in proximal aorta and aortic tree compared to ApoE+/−-TLR2+/− and ApoE+/−-TLR2−/− mice for all diet conditions. They also displayed profound changes in plaque composition, as evidenced by increased macrophage infiltration and apoptosis, increased lipid content, and decreased smooth muscle cell mass, all reflecting an unstable plaque phenotype. SAA levels from ApoE+/−-TLR2+/+ mice were significantly higher than from ApoE+/−-TLR2+/− and ApoE+/−-TLR2−/− mice. Serum cytokine analysis revealed increased levels of pro-inflammatory cytokines in ApoE+/−-TLR2+/+ mice compared to ApoE+/−-TLR2+/− and TLR2−/− mice, irrespective of diet or bacterial challenge. ApoE+/−-TLR2+/+ mice injected weekly for 24 weeks with FSL-1 (a TLR2 agonist) also demonstrated significant increases in atherosclerotic lesions, SAA and serum cytokine levels compared to ApoE+/−-TLR2−/− mice under same treatment condition. Finally, mass-spectrometry (MALDI-TOF-MS) of aortic samples analyzed by 2-dimentional gel electrophoresis differential display, identified 6 proteins upregulated greater than 2-fold in ApoE+/−-TLR2+/+ mice fed the high fat diet and inoculated with P.g compared to any other group.ConclusionGenetic deficiency of TLR2 reduces diet- and/or pathogen-associated atherosclerosis in ApoE+/− mice, along with differences in plaque composition suggesting greater structural stability while TLR-2 ligand-specific activation triggers atherosclerosis. The present data offers new insights into the pathophysiological pathways involved in atherosclerosis and paves the way for new pharmacological interventions aimed at reducing atherosclerosis.\n\nBODY:\nIntroductionAtherosclerosis is a multifactorial chronic inflammatory disease characterized by the accumulation of cells of both the innate and acquired immune systems within the intima of the arterial wall [1], [2]. In atherosclerosis, the normal homeostatic functions of the endothelium are altered, promoting an inflammatory response that results in increased expression of adhesion molecules. This in turn leads to recruitment of leukocytes, including monocytes, which penetrate into the intima, predisposing the vessel wall to lipid accretion [1], [3], [4]. Inflammatory mediators enhance uptake of modified lipoprotein particles and formation of lipid-laden macrophages. The adaptive immune response in atherosclerosis is mediated by T cells that enter the intima and secrete cytokines, which subsequently amplify the inflammatory response and promote the migration and proliferation of intimal smooth muscle cells. [2], [5].The innate immune system involves several different cell types, most importantly those of the mononuclear phagocyte lineage [6], [7], [8]. Macrophages and endothelial cells (EC) express receptors that recognize a broad range of molecular patterns foreign to the mammalian organism but commonly found on pathogens. These molecules include lipopolysaccharides and lipoproteins from Gram-negative bacteria, peptidoglycan and lipoteichoic acids from Gram-positive bacteria, lipoproteins from mycoplasma, and zymosan from yeast [9]. These pattern-recognition receptors include various scavenger receptors (ScRs) and Toll-like receptors (TLRs).TLRs are members of a large superfamily containing the interleukin-1 receptors (IL-1R) that share significant homology in their cytoplasmic domain, which is known as the Toll/IL-1R (TIR) domain [10]. Ligation of most TLRs transmits transmembrane signals that activate the NF-κB and mitogen-activated protein kinase (MAPK) pathways [8], [11], [12], [13].Both in vitro and in vivo knockout mouse studies have implicated TLRs in neointima formation and intimal hyperplasia involving modulation of inflammatory responses to exogenous and endogenous stimuli [14]. Although TLRs mediate protection against infection, various studies have demonstrated increased expression of TLR1, 2, and 4 in human atherosclerotic lesion, mechanistically linking TLRs, inflammation and atherosclerosis [6], [15], [16] with downstream signaling of TLR directly regulating inflammatory genes. In vitro stimulation of TLRs in human fibroblasts with a synthetic fibroblast stimulating lipopeptide (FSL-1; Pam2CGDPKHPKSF) leads to activation of NF-κB and the production of inflammatory cytokines in a MyD88-dependent manner [17], [18].Two important observations suggest TLR2 as a novel target to consider for therapeutic intervention in atherosclerosis. One is that TLR2 mediates responses to lipoproteins derived from multiple pathogens. Its unique ability to heterodimerize with TLR1 or TLR6 thus results in a relatively broad range of ligand specificity [19], [20], [21] which may contribute to atherogenesis in the context of exposure to a variety of pathogens.The potential importance of infectious agents of oral/periodontal origin, such as Porphyromonas gingivalis (P.g) in the development of atherosclerosis has recently been described [22], [23], [24], [25], [26], [27]. To our knowledge, the direct role of TLR2 in the bacteria-enhanced atherogenic process had never been addressed and warranted further investigation. Therefore, in the work reported here, we examined the effect of genetic deletion of TLR2 on the progression of atherosclerosis driven by a high fat diet and/or P. g infection in the ApoE+/− murine model. Stimulation by the specific agonist FSL-1 was used to further establish the role of TLR2 in modulating the progression of diet and/or bacteria enhanced atherosclerosis in mice with normal expression of TLR2.ResultsLevels of Glucose, Total Serum Cholesterol, LDL, and HDLMice were monitored for metabolic status by measuring blood glucose and lipids. No significant differences in body weight were observed in mice in connection to genotype or treatment, (e.g. P. g, FSL-1). Total serum cholesterol, LDL, HDL or glucose levels also revealed no significant differences among ApoE+/−-TLR2+/+, ApoE+/−-TLR2+/− and ApoE+/−-TLR2−/− mice that received similar treatments and were maintained on a similar diet. However, we observed a tendency of increased cholesterol, LDL and decreased HDL level in ApoE+/−-TLR2+/+ mice injected with P. g as compared to vehicle injected group. Furthermore, the lipid and glucose profiles did not reveal any differences between mice of the same genotype that were injected with P. g or FSL-1 (supplemental data: Table S1 and Table S2).\nEn face and Histomorphometric Analysis of Atheroma LesionsQuantitative en face analysis revealed statistically significant smaller lesions in ApoE+/−-TLR2+/− mice compared to ApoE+/−-TLR2+/+ mice after 24 week treatments (p<0.05): High fat diet inoculated with P. g (HP) demonstrated 12.5±1.7%, High fat diet injected with vehicle (HS) 5.2±1.8% and Chow diet inoculated with P. g (CP) demonstrated 2.6±0.8% of the aorta occupied by lesion. Interestingly, we did not observe any lesions on the aortic surface in ApoE+/−-TLR2−/− mice, irrespective of the diet or inoculum (Figs 1A–J). We also did not observe any lesions in any of the mice that were maintained on a chow diet and injected with vehicle (CS), regardless of their genetic backgrounds. We previously observed similar results from en face analysis performed after 14 weeks of these inoculation treatments of mice these same three genotypes (data not shown).10.1371/journal.pone.0003204.g001Figure 1\nP. g and/or high fat diet increases aortic atherosclerotic lesions in ApoE+/−-TLR2+/+ mice when compared to ApoE+/−-TLR2+/−, and ApoE+/−-TLR2−/− mice.\nEn face analysis: (1A–1I): Representative en face view of aortic surface lesions in ApoE+/−-TLR2+/+, ApoE+/−-TLR2+/−, and ApoE+/−-TLR2−/− mice after 24 weeks of treatments. (1J): Calculated percentages of aortic surface area covered by lesions after 24 weeks of treatments (bacterial challenge or vehicle control) among mice of three genotypes maintained on standard chow or high fat diets. Values represent means±SD; *p<0.05 for ApoE+/−-TLR2+/+ mice compared to ApoE+/−-TLR2+/− mice and **p<0.05 for ApoE+/−-TLR2+/+ mice compared to ApoE+/−-TLR2−/− mice in the same treatment condition and maintained on the same diet. Abbreviations are as defined in the text.Similarly, FSL-1 treatment for 24 weeks failed to induce any atherosclerotic changes in the aortas of ApoE+/−-TLR2−/− mice maintained on standard lab chow (data not shown). However, in ApoE+/−-TLR2+/+ mice maintained on a high fat diet and treated with FSL-1 for 24 weeks,11.2±0.6% of the aorta was covered by lesion. In contrast, in mice maintained on standard chow, FSL-1 treatment for 24 weeks resulted in significantly smaller atherosclerotic lesions that occupied only 1.94%±0.39% of the aorta (Fig. 2A). No statistically significant differences could be observed in the extent of aortic lesions in ApoE+/−-TLR2+/+ mice injected with P.g or FSL-1, irrespective of the diet (Fig. 2B).10.1371/journal.pone.0003204.g002Figure 2TLR2 activation through FSL-1 demonstrated no significant difference in aortic lesions when compared to P. g in ApoE+/−-TLR2+/+ and ApoE+/−-TLR2−/− mice.(2A) Percentage of aortic surface area covered by lesions in chow-fed groups for mice from two genetic backgrounds (ApoE+/−-TLR2+/+ and ApoE+/−-TLR2−/−) injected with P. g or FSL-1 for 24 weeks. Values represent means±SD; *p<0.05 between ApoE+/−-TLR2+/+ mice and ApoE+/−-TLR2−/− mice injected with P. g; **p<0.05 between ApoE+/−-TLR2+/+ mice and ApoE+/−-TLR2−/− mice injected with FSL-1. No lesions were detected in ApoE+/−-TLR2−/− mice irrespective of the treatment. (2B) Percentage of aortic surface area covered by lesions in ApoE+/−-TLR2+/+ mice maintained on either diet and injected weekly with P. g or FSL-1 for 24 weeks. Values represent means±SD.Histomorphometric analysis revealed significantly smaller lesions in the proximal aortas of ApoE+/−-TLR2+/− mice compared to ApoE+/−-TLR2+/+ mice and ApoE+/-TLR2−/− mice. In mice fed with high fat diet and injected with P.g the proximal aorta occupied by lesion was 46.2±6.6% in ApoE+/−-TLR2+/+, 18.6±2.4 in ApoE+/−-TLR2+/− and 3.04±0.8 in ApoE+/-TLR2−/− mice. In mice fed with high fat diet and injected with saline the proximal aorta occupied by lesion was 22.7±2.9% in ApoE+/−-TLR2+/+, 11.7±2.5 in ApoE+/−-TLR2+/− and 2.2±0.6 in ApoE+/-TLR2−/− mice. In mice fed with chow diet and injected with P.g the proximal aorta occupied by lesion was 16.1±3.2% in ApoE+/−-TLR2+/+, 7.6±2.3 in ApoE+/−-TLR2+/− and 1.5±0.5 in ApoE+/-TLR2−/− mice. (Fig. 3) Similar results were obtained at 14 weeks post inoculation also (data not shown). However, we did not observe any lesions in the chow-fed, vehicle-treated mice (CS), irrespective of their TLR2 genotype. Histomorphometric analysis of the proximal aortas revealed that 37.5±4.6% of the aortic lumen was occupied by lesion in ApoE+/−-TLR2+/+ mice injected with FSL-1 and maintained on a high fat diet, compared with only 10.4±1.3% of ApoE+/−-TLR2+/+ mice kept on a chow diet (Fig 4A). No lesions were observed in ApoE+/−-TLR2−/− mice after 24 weeks of FSL-1 injections and a standard chow diet (data not shown). Furthermore, there was no statistically significant difference in the percentage of aortic lumen occupied by lesions in mice injected with P. g when compared to FSL-1 treatment, irrespective of diet (Fig 4B).10.1371/journal.pone.0003204.g003Figure 3\nP.g and/or high fat diet increases atherosclerotic lesions in proximal aorta of ApoE+/−-TLR2+/+ mice when compared to ApoE+/−-TLR2+/−, and ApoE+/−-TLR2−/− mice.Microscopic cross-sections (10 µm) of the proximal aortic root were stained with Sudan IV and counterstained with hematoxylin to reveal lipid deposition, which was quantified by digital morphometry. (3A–D): results from mice maintained on a standard chow diet and inoculated weekly with P. g (CP). (3E–H): results from mice maintained on a high fat diet and inoculated weekly with vehicle (HS) (normal saline). (3I–L) results from mice maintained on a high fat diet and inoculated weekly with P. g (HP). (3D, H, l): data are presented graphically as percentage of total lumen of the proximal aorta occupied by lesions after 24 weeks of injections. Values represent means±SD; *p<0.05 between ApoE+/−-TLR2+/+ mice and ApoE+/−-TLR2+/− mice; **p<0.05 between ApoE+/−-TLR2+/+ mice and ApoE+/−-TLR2−/− mice in the same condition and maintained on the same diet. Abbreviations are as defined in the text. Photomicrographs shown are representative images obtained at the end of the 24 week treatment period. Original magnifications 20×. Scale bar represents 0.5 mm.10.1371/journal.pone.0003204.g004Figure 4TLR2 activation through FSL-1 demonstrated no significant difference in proximal aortic lesions when compared to P. g in ApoE+/−-TLR2+/+ and ApoE+/−-TLR2−/− mice.Microscopic cross-sections (10 µm) of the proximal aortic root were stained with Sudan IV and counterstained with hematoxylin to reveal lipid deposition, which was quantified by digital morphometry for samples from ApoE+/−-TLR2+/+ and ApoE+/−-TLR2−/− mice. (4A): percentage of total lumen of the proximal aorta occupied by lesions after 24 weeks of treatment in ApoE+/−-TLR2+/+ and ApoE+/−-TLR2−/− mice maintained on a standard chow diet and injected weekly with P. g or FSL-1. Values represent means±SD; *p<0.05 for differences between mice injected with P. g; **p<0.05 for differences between mice injected with FSL-1. No lesions were detected in ApoE+/−-TLR2−/− mice irrespective of the treatment. (4B): percentage of total lumen of the proximal aorta occupied by lesions in ApoE+/−-TLR2+/+ mice maintained on a chow diet or a high fat diet after 24 weeks of injections with P. g or FSL-1. Values represent means±SD.Imunohistochemical Analysis of Proximal AortaAfter establishing the involvement of TLR2 in diet/bacteria induced atherosclerosis, we performed a detailed examination of plaque compositions by immunofluorescence staining. Five sections of the proximal aorta per animal (n = 8) each separated by 80 µm, were selected and staining specific for macrophages, smooth muscle cells, and apoptotic cells using MOMA-2, α-SMA and TUNEL were performed, respectively. Significant differences in the plaque composition were observed between atherosclerotic lesions of all ApoE+/−-TLR2+/+ mice when compared to those from ApoE+/−-TLR2+/− and ApoE+/−-TLR2−/− mice, irrespective of the treatment. The atherosclerotic lesions in ApoE+/−-TLR2+/+ mice exhibited a greater percentage of infiltrating macrophages than smooth muscle cell accumulation. In contrast, in ApoE+/−-TLR2+/− and ApoE+/−-TLR2−/− mice, macrophage content was either less than or equal to the smooth muscle cell accumulation in plaques (Figs. 5A–5G). Furthermore, the marked increase in the inflammatory component of the lesions in ApoE+/−-TLR2+/+mice was associated with a substantial increase in the occurrence of apoptosis within their plaques. In ApoE+/-TLR2+/+ mice we observed: 6.3%, 5.2% and 4.9% inflammatory component in the HP, HS and CP groups, respectively (Figs. 5I–L)10.1371/journal.pone.0003204.g005Figure 5\nP. g and/or high fat diet results in unstable plaque in ApoE+/−-TLR2+/+ mice when compared to ApoE+/−-TLR2+/−, and ApoE+/−-TLR2−/− mice.Representative photomicrographs of atherosclerotic plaques from the aortic sinus of ApoE+/−-TLR2+/+, ApoE+/−-TLR2+/− and ApoE+/-TLR2−/− mice maintained on a high fat diet and inoculated weekly with P. g (HP) for 24 weeks. Stains identify sections of macrophage infiltration (MOMA-2 red staining) (5a, 5b&5c); smooth muscle cells (α-SMA red staining) (5E, 5F&5G); TUNEL positive cells (green spots coinciding with nuclear stain DAPI) (5I, 5J&5K). Quantitative computer-assisted image analysis (as described in Materials and Methods) was used to quantify the percentage of macrophage-positive areas (5D), smooth muscle cell area (5H) and TUNEL/DAPI positive cells (5I) in proximal aortic lesions in ApoE+/−-TLR2+/+, ApoE+/−-TLR2+/− and ApoE+/-TLR2−/− mice of all the groups at the conclusion of the 24 week treatment period. Data represent means±SD; *p<0.05 for ApoE+/−-TLR2+/+ mice compared to ApoE+/−-TLR2+/− mice, and **p<0.05 for ApoE+/−-TLR2+/+mice compared to ApoE+/−-TLR2−/− mice in the same treatment condition and maintained on the same diet. Abbreviations are as defined in text. Original magnifications 100× for macrophages and smooth muscle and 200× for TUNEL/DAPI staining. Scale bar represents 0.5 mm.Serum Amyloid A LevelAt the conclusion of the 24 week treatment period, SAA levels were highest in mice on high fat diets and also inoculated with P.g. High fat mice inoculated only with saline had the next highest SAA levels, followed by mice on a normal diet but challenged with weekly inoculations of P.g. Mice from all genetic backgrounds maintained on standard chow and inoculated with vehicle had the lowest SAA levels, regardless of genotype (Fig 6A). The serum SAA levels were undetectable in ApoE+/−-TLR2−/− mice maintained on a chow diet and injected with FSL-1 (data not shown). Also, there was no significant difference in serum SAA levels between ApoE+/−-TLR2+/+ mice injected with FSL-1 when compared to the P. g injected group, irrespective of the diet (Fig 6B).10.1371/journal.pone.0003204.g006Figure 6TLR2 activation through FSL-1 demonstrated no significant difference in SAA levels when compared to P. g in ApoE+/−-TLR2+/+ and ApoE+/−-TLR2−/− mice.(6A) SAA levels in serum samples obtained after 24 weeks of inoculations, as determined by ELISA. Data represent means±SD; *p<0.05 between ApoE+/−-TLR2+/+mice and ApoE+/−-TLR2+/− mice; **p<0.05 between ApoE+/−-TLR2+/+mice and ApoE+/−-TLR2−/− mice maintained under the same conditions and on the same diet. Abbreviations are as defined in the text. (6B) SAA levels in serum samples obtained at the end of the study, determined by ELISA. Data represent means±SD in ApoE+/−-TLR2+/+ mice maintained on either standard chow or a high fat diet, and injected weekly with P. g or FSL-1.Serum Cytokine LevelsTo further correlate the serum cytokine levels within advanced stage atherosclerotic lesions, serum samples were analyzed for 32 cytokines. Statistical significance was evaluated by ANOVA followed by the post-hoc Scheffe test. A level of p<0.05 was considered significant. Significantly altered cytokine levels were observed as a consequence of the high fat diet and of inoculation with P. g. These are presented in supplement sheet Fig-3. Compared to ApoE+/−-TLR2+/− and ApoE+/−-TLR2−/− mice, ApoE+/−-TLR2+/+ mice displayed profoundly higher levels of most proinflammatory cytokines and chemokines, including IL-1α, IL-1β, IL-6, IL-12p40, IL-12p70, TNF-α, MCP-1, VEGF, M-CSF, and GM-CSF in high fat diet and/or bacterial-challenged animals. No significant differences were observed in the cytokine levels between all 3 genetic background mice fed the chow diet and saline-inoculated. The most significantly elevated cytokine levels were seen in ApoE+/−-TLR2+/+ mice fed the high fat diet and also inoculated with P.g (HP) (Fig. 7).10.1371/journal.pone.0003204.g007Figure 7\nP. g and/or high fat diet results in increased proinflamatory cytokines in ApoE+/−-TLR2+/+ mice when compared to ApoE+/−-TLR2+/−, and ApoE+/−-TLR2−/− mice.Serum cytokine levels (pg/ml) in mice maintained on a standard lab chow diet and inoculated weekly for 24 weeks with P. g. (Fig 7A&7B). Serum cytokine levels (pg/ml) in mice maintained on a high fat diet and injected with saline weekly for 24 weeks (Fig 7C&7D). Serum cytokine levels (pg/ml) in mice maintained on a high fat diet and inoculated weekly for 24 weeks with P. g (Fig 7E&7F).Cytokine profiling in serum samples obtained from ApoE+/−-TLR2+/+ mice treated for 24 weeks with FSL-1 or wild type P.g. were compared. The expression of cytokines was increased with both stimuli in ApoE+/−-TLR2+/+ stimulated with FSL-1 or wild-type P.g. irrespective of the diet (Fig. 3 and 4 Supplemental Data S1). FSL-1 did not stimulate cytokine expression in ApoE+/−-TLR2−/− mice (Fig. S3 and S4 Supplemental Data S1).Two-dimensional Protein Maps of Aortic Tissues from ApoE+/−-TLR2+/+ and ApoE+/−-TLR2−/− MiceProtein extracts from aortic tissues of ApoE+/−-TLR2+/+ and ApoE+/−-TLR2−/− mice maintained on a high fat diet and/or inoculated with P. g were separated using 2-DGE. An example of the overall 2-DGE patterns of aortic protein extracts of ApoE+/−-TLR2+/+ mice fed a high fat diet and inoculated with P. g is shown in Fig 8. A total of 34 different protein spots were detected in response to P.g challenge: 21 protein spots increased least by 2-fold (Fig 8 red); 3 protein spot decreased at least by 2-fold (Fig 8 green) and 10 unmatched protein spots (Fig 8 black). Out of the 21 protein spots with at least 2 fold increase 6 proteins: Vesl-2 protein, Sod-2 protein, fumarate hydratase, myosin light chain polypeptide 3, aconitase, and gelsolin were identified (Table 1). Out of the 10 unmatched spots a protein: Hb was identified (table 1) only in ApoE+/−-TLR2+/+ mice maintained on the high fat diet and inoculated with P. g. Magnified gel regions corresponding to Hb from mice of both genotypes maintained on high fat diets after challenge with P.g or vehicle are compared in Figs 8A–D.10.1371/journal.pone.0003204.g008Figure 8\nP.g and/or high fat diet demonstrated changes in the aortic protein in ApoE+/−-TLR2+/+ mice when compared to ApoE+/−-TLR2−/− mice.Two-dimensional electrophoresis gel image of the proteins extracted from aortas (n = 5) from ApoE+/−-TLR2+/+ mice fed a high fat diet and injected weekly with P. g. Enlarged spots representing Hb were observed in aortas proteins from ApoE+/−-TLR2+/+ mice maintained on a high fat diet and injected with P. g (8B). Corresponding gel regions from aorta proteins from ApoE+/−-TLR2+/+ mice maintained on a high fat diet alone (8A) or ApoE+/−-TLR2−/− mice fed a high fat diet alone (8C) or also injected with P.g. (8D) did not exhibit the spots. Gels were stained by SYPRO RUBY stain. The spot numbers correspond to proteins listed in Table 1.10.1371/journal.pone.0003204.t001Table 11 Protein identified by MALDI TOF.Spot numberProteinExpectationAccession no.ApoE+/− TLR2+/+\nHFD+P.g\n111,108,109Gelsolin5.30E-058606238>2fold increase104,95,80Aconitase1.20E-0718079339>2 fold increase599,597,596,595Hemoglobin8.30E-0631982300Unmatched689SOD-26.05E-0617390379>2 fold increase682Vesl-22.00E-043766297>2 fold increase404Fumarate hydratase1.30E-0533859554>2 fold increase417Myosin, light polypeptide 35.70E-0433563264>2 fold increaseIdentification of proteins differentially expressed in ApoE+/−-TLR2+/+ mice fed a high fat diet and inoculated with P. g as compared to other groups.DiscussionIn the present study we demonstrate that TLR2 plays an important role in the pathogenesis of bacteria-enhanced diet-dependent atherosclerosis in the ApoE+/− murine model, establishing a key link between atherosclerosis and immune defense against foreign pathogens and/or endogenous inflammatory ligands. Both en face and histomorphometric data revealed that a greater percentage of the aorta and aortic lumen was occupied by the atherosclerotic lesions in ApoE+/−-TLR2+/+ mice as compared to either ApoE+/−-TLR2+/− or ApoE+/−-TLR2−/− mice. ApoE+/−-TLR2+/+ mice fed a high fat diet and inoculated with P. g (HP) exhibited larger lesions compared to mice fed a high fat diet and inoculated only with saline vehicle (HS) or mice fed the standard lab chow diet and inoculated with P. g. (CP). Increased cholesterol, LDL and decreased HDL level seen in ApoE+/−-TLR2+/+ mice injected with P. g in current work, corroborates well previous studies showing that periodontal pathogens can influence the systemic lipoprotein profile [33].Our data support previous studies showing that both endogenous (diet) and exogenous (Pam3CSK4) TLR2 ligands play important roles in the modulation of atherosclerosis [14], [28]. Indeed, mice deficient in TLR4, TLR2 and MyD88 all have reduced atherosclerosis which establishes that TLR-dependent pathways contribute to disease development. Although it is likely that total “infectious burden” contributes to atherosclerosis progression, endogenous ligands may also initiate and modulate Toll-like receptor signaling pathways [29], [30], [31].The unstable plaque phenotype is characterized by increased vulnerability to rupture and thrombosis. Histologically, an unstable plaque is identified by its thin fibrous cap, low smooth muscle cell count, high macrophage content, increased apoptosis and large lipid core [32], [33], [34], [35], [36]. The loss of the smooth muscle cells in particular is thought to be detrimental for plaque stability since most of the interstitial collagen fibers, which are important for the tensile strength of the fibrous cap, are produced by these cells [37]. Our detailed immunohistochemical analysis of atherosclerotic lesions for smooth muscle cells, macrophages, and apoptotic regions found that all the signs of plaque instability were consistently observed in ApoE+/−-TLR2+/+ mice, while ApoE+/−-TLR2+/− and ApoE+/−-TLR2−/− mice showed lesions more characteristic of stable plaque, with less macrophage content, less apoptosis, smaller lipid cores, and higher smooth muscle cell mass (Fig. 5). A significant increase of apoptosis in the lesions of ApoE+/−-TLR2+/+ mice, along with upregulation of proinflammatory cytokines which regulate the release of the matrix-degrading proteinases and may favor the unstable plaque phenotype [38].SAA, the mouse counterpart of human C-reactive protein, is an acute phase reactant known as a marker for systemic inflammation. It has been demonstrated that CRP is produced in the liver in response to IL-6, IL-1β, and TNF-α [39]. A strong association has also been shown between circulating levels of SAA and the extent of atherosclerosis in the aorta [23], [26], [27]. Our data demonstrated significantly higher serum levels of SAA, IL-6, IL-1β, and TNF-α in ApoE+/−-TLR2+/+ mice compared with ApoE+/−-TLR2−/− mice, irrespective of the diet or bacterial challenge treatment. Furthermore, SAA levels significantly correlated with the extent of aortic lesions examined after 24 weeks of challenge. Interestingly, significantly lower levels of SAA in ApoE+/−-TLR2−/− mice suggest that TLR2 deficiency may also lead to a lower overall systemic inflammatory status.We performed cytokine profiling in order to further investigate the systemic inflammatory status associated with TLR2 deficiency and the involvement of TLR2 in atherosclerosis. Our data show that TLR2 elicits differential expression of inflammatory cytokines and co-stimulatory molecules upon challenge with atherogenic stimuli (P.g. and/or high fat diet). Maximum induction of a host of proinflammatory cytokines (IL-1α, IL-1β, IL-6, IL-18, IFN-γ, IL-12p40, IL-12p70, TNF-α, MCP-1, VEGF, M-CSF, and GM-CSF) was observed in ApoE+/−-TLR2+/+ mice maintained on a high fat diet and challenged with P. g. Most of these cytokines and chemokines are proinflammatory factors, favoring cell migration, proliferation [40], [41] and chemo-attraction of inflammatory cells, such as monocytes/macrophages and T cells [42], [43], [44]. These results further implicate an aspect of antigen-specific adaptive immunity mostly characteristic of a Th1 response, including cytokines IL-2, IL-18, IFNγ and TNF-α [9]. The differential cytokine induction also implies that P. g and/or a high fat diet can activate different receptors to mediate intracellular signaling. It is known that the formation of heterodimers between TLR2 and other TLRs (TLR1 or TLR6) dictates the specificity of ligand recognition, thereby diversifying the possible outcomes of TLR2 activation [20]. In this context, it is noteworthy that the ApoE+/−-TLR2−/− genotype conferred atheroprotective effects, which may result in part from reduced systemic inflammation as shown by reduced expression of proinflammatory cytokines and chemokines in all treatment groups.To further establish the role of TLR2 in modulating the progression of atherosclerosis, we stimulated mice with the TLR2 agonist known as FSL-1. En face and histomorphometric analysis revealed that systemic exposure to FSL-1 dramatically increased lesion severity in a manner similar to P. g. In contrast, the absence of TLR2 resulted in complete prevention of lesions in mice on the chow diet and injected with FSL-1. Furthermore, the percentage of the lesions observed both in aorta and aortic sinus were comparable to the lesions observed in ApoE+/−-TLR2+/+ mice injected with P. g irrespective of the diet after 24 weeks. FSL-1 stimulation also altered the systemic inflammatory status as monitored by increased serum SAA levels in the ApoE+/−-TLR2+/+ mice when compared to the ApoE+/−-TLR2−/− mice. It is noteworthy that the levels of serum SAA in ApoE+/−-TLR2+/+ mice stimulated with FSL-1 were comparable to the levels obtained when mice were challenged with P. g irrespective of the diet, thus confirming the role of TLR2 in upregulation of systemic inflammation produced by P. g challenge. Cytokine profiling showed that both FSL-1 stimulation and P.g challenge resulted in a relatively similar expression of proinflammatory cytokines.Our expression proteomic approach extends the growing body of literature linking TLR2 and atherogenesis by identifying proteins involved in P. g- and/or diet-induced atherosclerosis in ApoE+/− mice. Using 2D gel electrophoresis (2-DGE) in combination with mass spectrometry (MS), we found that in ApoE+/−-TLR2+/+ mice, P. g stimulation in combination with a high fat diet up-regulated the expression of a set of proteins (Hb, Vesl-2 protein, Sod-2 protein, fumarate hydratase, myosin light chain polypeptide 3, aconitase and gelsolin) compared to high fat diet alone. Some of these proteins were found of interest in improving our understanding of the mechanisms linking atherogenesis to infection, inflammation, and immune response.Hemoglobin (Hb) is known to enhance the biological function of bacterial endotoxins [48] and therefore increased Hb can contribute to aheightened systemic response. Furthermore increased Hb content in the blood leads to increased viscosity, with detrimental effects on blood flow. Moreover, intraplaque hemorrhage seen in advanced lesions also causes the deposition of Hb. In our study, the detection of Hb only in ApoE+/−-TLR2+/+ mice fed a high fat diet and injected with P. g indicates that elevated Hb levels may be useful as a biomarker for an unstable plaque phenotype. Its presumed mechanism of vascular injury would include both oxidative heme toxicity caused by its ineffective clearance and also the subsequent consumption of nitric oxide, an important mediator of vascular homeostasis.Vesl-2 (Homer 2) is a post-synaptic adaptor protein that has been shown to be present in multiple tissues such as brain and heart. It may be linked to atherogenesis through its associations with glutamate receptor complexes and also the actin cytoskeleton. These glutamate receptors are coupled with G-proteins and activate phospholipase C, ultimately activating the IP3 receptor (IP3R) to release intracellular calcium, which can alter the function of ECs. Endothelial dysfunction typically results in platelet aggregation at the damaged site. Elevated intracellular calcium also leads to increased uptake of macromolecules in plasma such as fibrinogen and LDL, eventually forming atherosclerotic plaque. Thus it may be speculated that increased vesl-2 protein in ApoE+/−-TLR2+/+ mice fed a high fat diet and injected with P. g may lead to an increase in intracellular calcium, contributing to the increased atherosclerosis in this group of mice.Our observation of increased SOD-2 along with reduced smooth muscle cell content in ApoE+/−-TLR2+/+ mice fed a high fat diet and injected with P. g agrees well with a recent report in which SOD-2 deficient smooth muscle cells can exhibit a hypertrophic and hyperplastic phenotype. Thus we may speculate that P.g challenge upregulates mitochondrial SOD-2 and affects downstream pathways involving MAP kinases. Thus, increased SOD-2 may play a crucial role in determining plaque phenotype as it directly affects smooth muscle cell phenotype.Gelsolin is an actin-binding protein that is a key regulator of actin filament assembly and disassembly. It is regulated by Ca2+- and polyphosphoinositide 4, 5-bisphosphate (PIP2) and plays an important role in actin remodeling by regulating actin filament severing and capping. It has also been shown to play a role in apoptosis [45] and in sepsis-induced cell injury. Increased gelsolin activity has also been shown in failing human hearts [46] probably in reaction to the cell injury. Thus, increased level of gelsolin seen in ApoE+/−-TLR2+/+ mice fed a high fat diet and injected with P. g may be linked to the increased apoptosis and atherosclerosis observed in this group.Aconitase and fumarate hydratase are Krebs cycle enzymes. These enzymes area lso known to play important roles in the response to oxidant stress, which can inactivate aconitase and other Krebs cycle enzymes [47]. We observed increased aconitase and fumarate hydratase levels in ApoE+/−-TLR2+/+ mice fed a high fat diet and injected with P. g. These proteins probably represent an adaptive response to increased oxidative stress but they are also important predisposing factors in the progression of the atherosclerotic process.Taken together, our results confirm the important role for TLR2 signaling in diet and/or bacteria enhanced atherosclerosis in an ApoE+/− mouse model, providing a link between innate immunity, inflammation and atherosclerosis. Due to TLR2 central role in the disease process, it represent a target of immunomodulatory therapy with the goal of tipping the balance from excessive chronic inflammation towards resolution of inflammation, while not compromising host defenses or atheroprotective immune functions. Therefore manipulation of TLR2 pathways has great therapeutic potential. TLRs inhibitors or their associated signaling molecules hold great promise in the prevention of atherosclerosis.Materials and MethodsPlease refer to the supplemental data and Fig. S1 and S2 for details. Briefly, all animal protocols were approved by the Boston University Medical Campus Institutional Animal Care and Use Committee. To investigate the role of TLR2 in inflammation- and/or infection-associated atherosclerosis, 10 week-old ApoE+/−-TLR2+/+, ApoE+/−-TLR2+/− and ApoE+/−-TLR2−/− mice were fed either a high fat diet or a regular chow diet. All mice were inoculated intravenously, once per week for 24 consecutive weeks, with 50 µl live P. g (107 CFU) or vehicle (normal saline). Animals were euthanized 24 weeks after the first inoculation. Histomorphometric analysis of the aortic lesions and the proximal aorta using Sudan red stain were performed. Immunofluroscent staining for macrophage, smooth muscle cell and apoptosis were performed on the proximal aortic sections. Metabolic profile, serum amyloid A and serum cytokine levels were also performed for all the three genotypes. For the TLR2 agonist study a second set of four week old male ApoE+/−-TLR2+/+ fed either a HFD or a regular chow diet for 6 weeks (n = 10) were used. All mice were inoculated intravenously, once per week for 24 consecutive weeks, with 5 µg FSL-1in 50 µl saline or vehicle (normal saline). Animals were euthanized 24 weeks after the first inoculation. To compare and further establish an absolute effect of TLR2 in bacteria-enhanced atherosclerotic lesions, a third set of experiments involved only ApoE+/−-TLR2−/− mice fed only the standard chow diet. Four week old ApoE+/−-TLR2−/− mice maintained on a regular chow diet for 6 weeks (n = 10), were inoculated with 50 µl saline vehicle or 5 µg FSL-1 in 50 µl saline. All groups were analyzed for atherosclerotic lesions, metabolic profile, serum amyloid A and serum cytokine levels after 24 weeks of inoculations.Statistical AnalysisAll histomorphometric measurements were made by an examiner blinded to the identity of the samples. All quantitative measurements were confirmed by random analysis of one fourth of the specimens by the same examiner (R>0.92) and by another independent examiner (a pathologist) to ensure consistency. The intra-examiner and inter-examiner variation were each <10%. All histomorphometric and serum assay data were analyzed by ANOVA followed by the post-hoc Scheffe test. A level of p<0.05 was considered significant.Supporting InformationData S1Supplemental material document(0.08 MB DOC)Click here for additional data file.Figure S1Animal grouping and experimental time schedule for P.gingivalis expeiments. Four week old male ApoE+/−-TLR2+/+, ApoE+/−-TLR2+/− and ApoE+/−-TLR2−/− mice were fed either a HFD or a regular chow diet for 6 weeks (n = 8), then inoculated once per week for 24 weeks with 50 µl of either vehicle (normal saline) or 107 CFU) P. g while maintained on the chosen diet. Thus, there were 4 groups for each genotype of mice: Group 1 was fed a standard chow diet and inoculated weekly with 50 µl saline vehicle (CS); Group 2 was fed a standard chow diet and inoculated with 50 µl (107 CFU) P. g. (CP); Group 3 was fed a high fat diet and inoculated with 50 µl saline vehicle (HS); Group 4 was fed a high fat diet and inoculated with 50 µl (107 CFU) P. g (HP). In summary, mice (n = 8) in each group received 24 tail vein injections of either vehicle or P. g once weekly.(0.83 MB TIF)Click here for additional data file.Figure S2Animal grouping and experimental time schedule for FSL-1expeiments. Effects of FSL-1 were tested in two sets of experiments. In the first, four week old male ApoE+/−-TLR2+/+ were fed either a HFD or a regular chow diet for 6 weeks (n = 10) then inoculated once per week for 24 weeks with 50 µl of either vehicle (normal saline) or 5 µg FSL-1 while maintained on the chosen diet. The resulting 4 groups were: Group 1a was fed a standard chow diet and inoculated with 50 µl saline vehicle (CS); Group 2a was fed a standard chow diet and inoculated weekly with 50 µl (5 µg) FSL-1; Group 3a was fed a high fat diet and inoculated weekly with 50 µl saline vehicle (HS); Group 4a was fed a high fat diet and inoculated with 50 µl (5 µg) FSL-1. All groups were tested after 24 weeks of their diet and inoculation regimens. For the second set of experiments, four week old ApoE+/−-TLR2−/− mice maintained on only a regular chow diet for 6 weeks (n = 10), then were divided into 2 groups: Group 1b was inoculated weekly with 50 µl vehicle saline (CS); Group 2b was inoculated weekly with 50 µl (5 µg) FSL-1. All groups were tested after 24 weeks of inoculations.(0.53 MB TIF)Click here for additional data file.Figure S3TLR2 activation through FSL-1 demonstrated similar expression of increased proinflamatory cytokines as compared chow fed and injected with P. g in ApoE+/−-TLR2+/+ mice. Serum cytokine levels (pg/ml) in ApoE+/−-TLR2+/+ mice fed a standard chow diet and injected weekly with P. g or FSL-1. Data represent mean+SD.(0.78 MB TIF)Click here for additional data file.Figure S4TLR2 activation through FSL-1 demonstrated similar expression of increased proinflamatory cytokines as compared high fat fed and injected with P. g in ApoE+/−-TLR2+/+ mice. Serum cytokine levels (pg/ml) in ApoE+/−-TLR2+/+ mice fed a high fat diet and injected weekly with P. g or FSL-1. Data represent mean+SD.(0.77 MB TIF)Click here for additional data file.Table S1Metabolic profiles of ApoE+/−-TLR2+/+, ApoE+/−-TLR2+/− and ApoE+/−-TLR2−/− mice maintained on a standard lab chow diet or a high fat diet, and injected weekly with either saline or with P. g, at 24 weeks. *Significance between ApoE+/−-TLR2+/+, ApoE+/−-TLR2+/− and ApoE+/−-TLR2−/− for respective groups. Abbreviations are as defined in the text.(0.07 MB DOC)Click here for additional data file.Table S2Metabolic profile of ApoE+/−-TLR2+/+ and ApoE+/−-TLR2−/− mice fed with either a standard lab chow diet or a high fat diet, and injected weekly with either P. g or FSL-1; measurements were obtained after 24 weeks of treatments.(0.04 MB DOC)Click here for additional data file.\n\nREFERENCES:\n1. RossR\n1999\nAtherosclerosis–an inflammatory disease.\nN Engl J Med\n340\n115\n126\n9887164\n2. LibbyPRidkerPMMaseriA\n2002\nInflammation and atherosclerosis.\nCirculation\n105\n1135\n1143\n11877368\n3. PrescottSMMcIntyreTMZimmermanGAStafforiniDM\n2002\nSol Sherry lecture in thrombosis: molecular events in acute inflammation.\nArterioscler Thromb Vasc Biol\n22\n727\n733\n12006383\n4. RossR\n1999\nAtherosclerosis is an inflammatory disease.\nAm Heart J\n138\nS419\n420\n10539839\n5. LibbyP\n2002\nInflammation in atherosclerosis.\nNature\n420\n868\n874\n12490960\n6. MedzhitovR\n2001\nToll-like receptors and innate immunity.\nNat Rev Immunol\n1\n135\n145\n11905821\n7. MedhitovRHC\n2000\nAdvances in Immunology.\nN Engl J Med\n8. JanewayCAJrMedzhitovR\n2002\nInnate immune recognition.\nAnnu Rev Immunol\n20\n197\n216\n11861602\n9. HanssonGKLibbyPSchonbeckUYanZ-Q\n2002\nInnate and Adaptive Immunity in the Pathogenesis of Atherosclerosis.\nCirc Res\n91\n281\n291\n12193460\n10. TakedaKAkiraS\n2004\nMicrobial recognition by Toll-like receptors.\nJ Dermatol Sci\n34\n73\n82\n15033189\n11. MuzioMNatoliGSaccaniSLevreroMMantovaniA\n1998\nThe Human Toll Signaling Pathway: Divergence of Nuclear Factor kappa B and JNK/SAPK Activation Upstream of Tumor Necrosis Factor Receptor-associated Factor 6 (TRAF6).\nJ Exp Med\n187\n2097\n2101\n9625770\n12. FaureEEquilsOSielingPAThomasLZhangFX\n2000\nBacterial Lipopolysaccharide Activates NF-kappa B through Toll-like Receptor 4 (TLR-4) in Cultured Human Dermal Endothelial Cells. DIFFERENTIAL EXPRESSION OF TLR-4 AND TLR-2 IN ENDOTHELIAL CELLS.\nJ Biol Chem\n275\n11058\n11063\n10753909\n13. GordonS\n2002\nPattern Recognition Receptors: Doubling Up for the Innate Immune Response.\nCell\n111\n927\n930\n12507420\n14. SchoneveldOudeNvanMLamandeK\n2005\nToll-like receptor 2 stimulation induces intimal hyperplasia and atherosclerotic lesion development.\nCardiovascular Research\n66\n162\n169\n15769459\n15. XuXHShahPKFaureEEquilsOThomasL\n2001\nToll-Like Receptor-4 Is Expressed by Macrophages in Murine and Human Lipid-Rich Atherosclerotic Plaques and Upregulated by Oxidized LDL.\nCirculation\n104\n3103\n3108\n11748108\n16. EdfeldtKSwedenborgJHanssonGKYanZ-q\n2002\nExpression of Toll-Like Receptors in Human Atherosclerotic Lesions: A Possible Pathway for Plaque Activation.\nCirculation\n105\n1158\n1161\n11889007\n17. NakamuraJShibataKHasebeAIntoTWatanabeT\n2002\nSignaling pathways induced by lipoproteins derived from Mycoplasma salivarium and a synthetic lipopeptide (FSL-1) in normal human gingival fibroblasts.\nMicrobiol Immunol\n46\n151\n158\n12008923\n18. OkusawaTFujitaMNakamuraJ-iIntoTYasudaM\n2004\nRelationship between Structures and Biological Activities of Mycoplasmal Diacylated Lipopeptides and Their Recognition by Toll-Like Receptors 2 and 6.\nInfect Immun\n72\n1657\n1665\n14977973\n19. TakeuchiOSatoSHoriuchiTHoshinoKTakedaK\n2002\nCutting Edge: Role of Toll-Like Receptor 1 in Mediating Immune Response to Microbial Lipoproteins.\nJ Immunol\n169\n10\n14\n12077222\n20. OzinskyASmithKDHumeDUnderhillDM\n2000\nCo-operative induction of pro-inflammatory signaling by Toll-like receptors.\nJ Endotoxin Res\n6\n393\n396\n11521061\n21. TakeuchiOKawaiTMuhlradtPFMorrMRadolfJD\n2001\nDiscrimination of bacterial lipoproteins by Toll-like receptor 6.\nInt Immunol\n13\n933\n940\n11431423\n22. GibsonFCIIIHongCChouH-HYumotoHChenJ\n2004\nInnate Immune Recognition of Invasive Bacteria Accelerates Atherosclerosis in Apolipoprotein E-Deficient Mice.\nCirculation\n109\n2801\n2806\n15123526\n23. LiLMessasEBatistaELJrLevineRAAmarS\n2002\nPorphyromonas gingivalis Infection Accelerates the Progression of Atherosclerosis in a Heterozygous Apolipoprotein E-Deficient Murine Model.\nCirculation\n105\n861\n867\n11854128\n24. HajishengallisGMartinMSchifferleREGencoRJ\n2002\nCounteracting interactions between lipopolysaccharide molecules with differential activation of toll-like receptors.\nInfect Immun\n70\n6658\n6664\n12438339\n25. Hajishengallis GSARussellMWGencoRJ\n2002\nInteractions of oral pathogens with toll-like receptors: possible role in atherosclerosis.\nAnn Periodontol\n7\n72\n78\n16013219\n26. ChiHMessasELevineRAGravesDTAmarS\n2004\nInterleukin-1 receptor signaling mediates atherosclerosis associated with bacterial exposure and/or a high-fat diet in a murine apolipoprotein E heterozygote model: pharmacotherapeutic implications.\nCirculation\n110\n1678\n1685\n15353494\n27. MadanMBishayiBHogeMMessasEAmarS\n2007\nDoxycycline affects diet- and bacteria-associated atherosclerosis in an ApoE heterozygote murine model: Cytokine profiling implications.\nAtherosclerosis\n190\n62\n72\n16563401\n28. MullickAETobiasPSCurtissLK\n2005\nModulation of atherosclerosis in mice by Toll-like receptor 2.\nJ Clin Invest\n115\n3149\n3156\n16211093\n29. BjorkbackaHKunjathoorVVMooreKJKoehnSOrdijaCM\n2004\nReduced atherosclerosis in MyD88-null mice links elevated serum cholesterol levels to activation of innate immunity signaling pathways.\nNat Med\n10\n416\n421\n15034566\n30. MichelsenKSWongMHShahPKZhangWYanoJ\n2004\nLack of Toll-like receptor 4 or myeloid differentiation factor 88 reduces atherosclerosis and alters plaque phenotype in mice deficient in apolipoprotein E.\nPNAS\n101\n10679\n10684\n15249654\n31. O'NeillLAGreeneC\n1998\nSignal transduction pathways activated by the IL-1 receptor family: ancient signaling machinery in mammals, insects, and plants.\nJ Leukoc Biol\n63\n650\n657\n9620655\n32. ChiuB\n1999\nMultiple infections in carotid atherosclerotic plaques.\nAm Heart J\n138\nS534\n536\n10539867\n33. MadanMBishayiBHogeMAmarS\n2008\nAtheroprotective role of interleukin-6 in diet- and/or pathogen-associated atherosclerosis using an ApoE heterozygote murine model.\nAtherosclerosis\n197\n504\n14\n17412346\n34. MallatZCorbazAScoazecAGraberPAlouaniS\n2001\nInterleukin-18/Interleukin-18 Binding Protein Signaling Modulates Atherosclerotic Lesion Development and Stability.\nCirc Res\n89\n41e\n45\n35. VirmaniRKolodgieFDBurkeAPFarbASchwartzSM\n2000\nLessons From Sudden Coronary Death: A Comprehensive Morphological Classification Scheme for Atherosclerotic Lesions.\nArterioscler Thromb Vasc Biol\n20\n1262\n1275\n10807742\n36. ShiomiMItoTHirouchiYEnomotoM\n2001\nStability of atheromatous plaque affected by lesional composition: study of WHHL rabbits treated with statins.\nAnn N Y Acad Sci\n947\n419\n423\n11795305\n37. KockxMMHermanAG\n2000\nApoptosis in atherosclerosis: beneficial or detrimental?\nCardiovasc Res\n45\n736\n746\n10728396\n38. AliprantisAOYangRBWeissDSGodowskiPZychlinskyA\n2000\nThe apoptotic signaling pathway activated by Toll-like receptor-2.\nEmbo J\n19\n3325\n3336\n10880445\n39. RifaiNRidkerPM\n2001\nHigh-Sensitivity C-Reactive Protein: A Novel and Promising Marker of Coronary Heart Disease.\nClin Chem\n47\n403\n411\n11238289\n40. BarillariGAlboniciLIncerpiSBogettoLPistrittoG\n2001\nInflammatory cytokines stimulate vascular smooth muscle cells locomotion and growth by enhancing [alpha]5[beta]1 integrin expression and function.\nAtherosclerosis\n154\n377\n385\n11166770\n41. YueTLWangXSungCPOlsonBMcKennaPJ\n1994\nInterleukin-8. A mitogen and chemoattractant for vascular smooth muscle cells.\nCirc Res\n75\n1\n7\n8013067\n42. ReapeTJGrootPHE\n1999\nChemokines and atherosclerosis.\nAtherosclerosis\n147\n213\n225\n10559506\n43. ShahPK\n2000\nCirculating Markers of Inflammation for Vascular Risk Prediction: Are they Ready for Prime Time.\nCirculation\n101\n1758\n1759\n10769272\n44. ShinWSSzubaARocksonSG\n2002\nThe role of chemokines in human cardiovascular pathology: enhanced biological insights.\nAtherosclerosis\n160\n91\n102\n11755926\n45. SunHQYamamotoMMejillanoMYinHL\n1999\nGelsolin, a Multifunctional Actin Regulatory Protein.\nJ Biol Chem\n274\n33179\n33182\n10559185\n46. YangJMoravecCSSussmanMADiPaolaNRFuD\n2000\nDecreased SLIM1 Expression and Increased Gelsolin Expression in Failing Human Hearts Measured by High-Density Oligonucleotide Arrays.\nCirculation\n102\n3046\n3052\n11120693\n47. HausladenAFridovichI\n1996\nMeasuring nitric oxide and superoxide: rate constants for aconitase reactivity.\nMethods Enzymol\n269\n37\n41\n8791635\n48. HoweJGaridelPRoessleMRichterWAlexanderC\n2008\nStructural investigations into the interaction of hemoglobin and part structures with bacterial endotoxins.\nInnate Immun\n14\n39\n49\n18387918"
+}

batch_3/PMC2527525.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527525",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527525\nAUTHORS: Babacar Faye, Frederick Arnaud, Eric Peyretaillade, Emilie Brasset, Bernard Dastugue, Chantal Vaury\n\nABSTRACT:\nBackgroundThe retroviral Integrase protein catalyzes the insertion of linear viral DNA into host cell DNA. Although different retroviruses have been shown to target distinctive chromosomal regions, few of them display a site-specific integration. ZAM, a retroelement from Drosophila melanogaster very similar in structure and replication cycle to mammalian retroviruses is highly site-specific. Indeed, ZAM copies target the genomic 5′-CGCGCg-3′ consensus-sequences. To enlighten the determinants of this high integration specificity, we investigated the functional properties of its integrase protein denoted ZAM-IN.Principal FindingsHere we show that ZAM-IN displays the property to nick DNA molecules in vitro. This endonuclease activity targets specific sequences that are present in a 388 bp fragment taken from the white locus and known to be a genomic ZAM integration site in vivo. Furthermore, ZAM-IN displays the unusual property to directly bind specific genomic DNA sequences. Two specific and independent sites are recognized within the 388 bp fragment of the white locus: the CGCGCg sequence and a closely apposed site different in sequence.ConclusionThis study strongly argues that the intrinsic properties of ZAM-IN, ie its binding properties and its endonuclease activity, play an important part in ZAM integration specificity. Its ability to select two binding sites and to nick the DNA molecule reminds the strategy used by some site-specific recombination enzymes and forms the basis for site-specific integration strategies potentially useful in a broad range of genetic engineering applications.\n\nBODY:\nIntroductionIntegration of the retroviral DNA genome into host-cell DNA is an essential step in the retrovirus replication cycle, permitting viral genomes to become permanently fixed as proviruses into the DNA of the host and to use host transcriptional machinery for the production of viral RNA [1]. This integration is performed by an enzyme called integrase encoded by the retrovirus. Although their mechanism of action is not yet clearly elucidated, retroviral integrases have been shown to carry out all the steps known to be required for processing and joining of the viral DNA [2]. Hotspots of integration exist and these preferences appear to be specific to the individual viruses [3]. Several studies indicate that the intrinsic properties of integrases participate in this selection. For instance, in vitro experiments show that integrases from different retroviruses each display a distinct and unique choice of integration sites when given an identical target DNA [4], [5]. Further experiments also indicate that local DNA sequence can influence the choice of the target site [6]. Indeed, some insertions have been associated with palindromic consensus centred on the virus-specific duplicated target site sequence, or as intrinsically bent DNA [7]. By analysing a number of sequences from HIV-1, avian sarcoma-leukosis virus (ASLV) and Murine Leukaemia Virus (MLV) into human cellular DNA, a symmetrical base preference surrounding HIV-1 and ASLV integration sites has been found [8]. Weak palindromic consensus sequences have also been reported to be a common feature at the integration target sites of many retroviruses [9]. Therefore, local DNA structure can affect insertion specificity but several studies also revealed that the chromatin structure imposed by nucleosomes or by other proteins can influence the efficiency of insertion into a particular target. Some of these proteins can be involved in chromatin structure [10]–[12], in transcription activity of nearby genes [13] or be cellular targeting proteins [4], [5]. Several cellular DNA binding proteins have been described that bind integration complexes and/or facilitate integration, including BAF, HMGa1, Ku, and LEDGF [4], [14]. Overall, despite some preferences, a high DNA sequence specificity for retroviral integration has never been described so far.LTR-retrotransposons replication cycle is very similar to the retroviruses one. They encode gag, pol and a subclass of them have an additional env gene. Like retroviruses, pol encodes protease, reverse-transcriptase, and integrase proteins essential for retrotransposition. Various degrees of bias for the integration target sites in vivo have been described for these elements. The yeast Saccharomyces cerevisiae contains several well-studied retrotransposons –Ty1, Ty3 and Ty5- that display interesting patterns of target site selection [15], [16]. For instance, Ty1 targets the upstream sequences of transfer RNA (tRNA) or other PolIII transcribed genes [17]. Ty3 copies are also found in these regions but at a more precise location, 1–4 bp from the transcription start site [18]. This targeting is achieved by the interaction of Ty3 preintegration complex (PIC) with the PolIII transcription factor TFIIIB/TFIIIC [19]. Instead, Ty5 integrase interacts with the transcription silencing protein Sir4p and specifically targets transcriptionally ‘silent’ regions of the yeast genome, such as telomeres or the silent mating loci HM [20]–[22]. Overall, data from retroviruses and LTR-retrotransposons demonstrate a combined involvement of the Integrase, the DNA sequence and cellular host proteins to direct integration at the desired genomic DNA sites.ZAM is an LTR-retrotransposon of 8,435-bp present within the genome of Drosophila melanogaster [23]. On the basis of sequence, structural, and functional similarities, ZAM displays a striking resemblance to vertebrate retroviruses [24]. Its three open reading frames gag, pol, and env are surrounded by two long terminal repeats or LTRs. The ZAM pol gene is subdivided into three regions, which encode typical retrovirus-like enzymes: protease, reverse transcriptase-RnaseH, and integrase (IN) [25], [26]. The latter displays all the characteristics of canonical retroviral IN [27]. In a previous paper, we reported that ZAM is highly sequence specific in its integration, much higher than any other retrovirus described so far. By exhaustive analyses of ZAM insertions, we have shown that the target sequence chosen by nearly every ZAM element is CGCGCg (lowercase “g” indicates a 50% occurrence of that base) [25]. However, the mechanism of this integration process and the reason of its specificity had not been elucidated.In this paper, we investigated the functional properties of ZAM integrase in order to understand the determinant of this specificity. We investigated its endonuclease property and show that ZAM integrase cleaves specifically a genomic site known to be a target of ZAM integration in vivo. Our results further indicate that ZAM-IN recognizes and binds two distinct DNA sites, the CGCGCg sequence corresponding to the ZAM integration site, and a second site located in the vicinity. Our data strongly argue that ZAM-IN is the main actor in the site specificity of ZAM integration.ResultsThe ZAM integrase displays an endonuclease activity on specific DNA fragmentsTwo reactions catalyzed by the integrases encoded by mammalian retroviruses have been well described: 1) the removal of two bases from the 3′ end of each viral DNA strand, and 2) the covalent attachment of leaving recessed 3′ hydroxyl groups at the viral DNA termini to protruding 5′ phosphoryl ends of host cell DNA (for review [2], [28]). Moreover, the ability of retroviral integrases to recognize, cleave and drive retroviral integration into specific DNA targets has not yet been reported although some preferences for certain genomic sites might be explained by intrinsic properties of the integrases.Since ZAM copies are found integrated in a very specific consensus sequence CGCGCg, we investigated if ZAM-IN properties could explain such a targeting of the host DNA. Thus, ZAM-IN was expressed in bacteria as a GST-fusion protein and fixed on glutathione (GSH)-agarose beads (Fig. 1A, “IN”). Then we examined whether ZAM-IN intrinsic properties display a specific endonuclease activity. To this end, its endonuclease activity was assayed by measuring the ability of the purified ZAM integrase to convert supercoiled plasmids into circular and linear molecules. Experiments were conducted with two types of plasmids. One corresponds to the pUC18 cloning vector containing no insert. This plasmid displays two distinct CGCGCG sites present at nucleotide positions 2–7 and 652–657 (Fig. 1B). The second plasmid corresponds to the pUC18 vector containing a 388 bp genomic fragment taken from the upstream region of the white gene. This genomic fragment called w4278 from the genomic position of its 5′ end, comprises a unique consensus sequence CGCGCG (position 4314) previously described as a target for ZAM insertions [25] (Fig. 1B). Both plasmids were called pUC and pUC/white respectively. When the endonuclease activity of ZAM-IN was assayed on the pUC plasmid (see Materials and Methods), a heavy band corresponding to supercoiled molecules, and a very faint band corresponding to open circle molecules were observed for both treated and non-treated plasmids (Fig. 1C). The open circle molecules observed in ZAM-IN treated and non treated samples indicate that this population of circularized molecules resulted from DNA nicks which randomly occurred probably during DNA extraction. These results indicated that ZAM integrase is unable to cleave the pUC18 vector sequence despite the presence of two CGCGCG sites.10.1371/journal.pone.0003185.g001Figure 1The endonuclease activity of ZAM integrase correlates with the presence of a 388 bp fragment from the white locus.A) Schematic representation of the ZAM integrase “IN” and a carboxy-terminal deleted integrase “ΔIN” used in the in vitro DNA binding assay. The three main domains: the zing finger “HHCC”, the catalytic domain “DDE” and a predictive DNA binding domain “BD” are represented. Nucleotide numbers according to ZAM sequence are indicated below. The hatched box indicated the region deleted to generate the ΔIN protein. The full length and the truncated integrases were expressed in bacteria as GST fusion proteins and fixed on agarose beads. IN and ΔIN purified proteins were analysed on SDS-PAGE electrophoresis followed by coomassie staining (right panel). The molecular masses of proteins are indicated in kilodalton. B) Circular representation of the 2686 bp pUC18 plasmid. Palindromic sequences CGCGCG present in pUC18 are indicated. The Drosophila genomic locus known to be the target of ZAM integration and located 3 kb upstream of the white gene is presented above. The white fragment (from positions 4278 to 4666 according to the drosophila sequence) was cloned in the pUC18 plasmid and is represented by the grey box. The black dot at position 4314 indicates the CGCGCG integration site of ZAM. C) In vitro endonuclease activity of ZAM integrase: pUC, pUC/white and pUC/white1mut plasmids were incubated without (−IN) or with (+IN) purified ZAM-IN. Positions of the supercoiled, nicked (circle) and linear (bar) DNAs are indicated.By contrast, an increase of open circles is clearly observed on the gel when pUC/white is incubated with ZAM-IN (Fig. 1C). This increase is easily registered between treated and untreated samples although circularized pUC/white molecules are initially present in the pUC/white DNA sample before the ZAM-IN treatment (see line 3, Fig. 1C). Thus, a nicking property of the integrase protein is registered when the white fragment is added to the pUC vector. In this set of assays, an increase in linear molecules that likely derive from double strand breaks generated by the ZAM-IN is also observed. However, it must be noticed that the amount of linear molecules in this experiment was higher than generally observed in other similar experiments.Since in the same experimental conditions the pUC/white is cleaved unlike the pUC vector, it is very unlikely that the nicking property results from the activity of the purified ZAM-IN and not from the activity of a bacterial enzyme which would have been co-purified with ZAM-IN.Overall, these results bring evidence that ZAM-IN does not nick any DNA fragment but selects and cleaves only some of them. The 388 bp white fragment added to the pUC plasmid carries all the signals required to drive this specific recognition ending by cleavage. Importantly, even if the CGCGCG sequence is the target site for ZAM integration, its presence is not sufficient for cleavage. This is clearly demonstrated by the fact that the pUC plasmid is not cleaved despite the presence of two CGCGCG sites. Furthermore, in an additional series of assays presented Figure 1C, we found that ZAM-IN retains the ability to cleave a plasmid named pUC/white1mut in which the CGCGCG sequence of the white fragment was disrupted by mutagenesis and replaced by AGAGCG. Therefore, the signal required for the endonuclease activity of ZAM-IN is not the sole CGCGCG site of the white fragment identified as the ZAM integration site.ZAM integrase binds two genomic sites within the target locusSince an unidentified signal might exist in the 388 bp white fragment for the integrase to cleave the DNA, we hypothesized that some specific binding sites for ZAM-IN might be such signals.It is well demonstrated that retroviral Integrases have the property to bind each extremity of the viral DNA within their LTR sequence [29], [30]. Thus, in a first series of experiment we verified that ZAM-IN has also the capacity to bind the ZAM LTR. In vitro DNA binding assays were performed on the full length 5′LTR of ZAM and on two shortened LTR fragments. The GST-fusion protein of ZAM-IN fixed on beads and depicted figure 1A was used in these experiments. The LTR fragment denoted “K” corresponds to the ZAM LTR digested by KpnI to delete 21 bp from its 5′ end. The second denoted “H” corresponds to the LTR digested by HindIII to delete 82 bp from its 5′ end (Fig. 2A). As shown figure 2B, ZAM-IN is able to bind the full length LTR (upper left panel, first lane) as well as the K fragment. However, the H fragment deleted for 82 bp of ZAM 5′ end is no more retained by ZAM-IN.10.1371/journal.pone.0003185.g002Figure 2LTR binding property of ZAM integrase.A) ZAM LTR fragments. The grey box represents the full length LTR of ZAM and the solid bars represent the KpnI “K” and HindIII “H” restriction sites at the position 21 and 82, respectively. A full length LTR and truncated PCR product deleted of the first 40 bp of ZAM LTR called “ΔLTR” are presented below. A double stranded oligonucleotide spanning from position 28 to 40 and called “BS” was also used in these experiments. B) In vitro DNA binding assays with ZAM-IN on the LTR fragments. Left panel: The full length “LTR” and the two truncated LTR fragments digested by KpnI “K” or HindIII “H” were tested as indicated above each lane. Middle panel: the full length LTR and a truncated PCR product “ΔLTR” were used in these assays. Right panel:\nIn vitro binding assays with ΔIN on ZAM LTR fragments: the full length “LTR” and the two truncated LTR fragments digested by KpnI “K” or HindIII “H” were tested as indicated above each lane. C) In vitro DNA binding assays performed with a double stranded oligonucleotide from base 28 to 40 according to ZAM sequence. The 13 bp fragment is retained by ZAM-IN. DNA fragments sizes are indicated as “L”. DNA fragments sizes are indicated for each panel. In B and C, bound and unbound fractions are presented in upper and lower panels respectively. The percentage of bound and unbound fractions is presented below each panel in B. In C, 100% of the 274 bp fragment was recovered in the unbound fraction whereas 100% of the 66 and 48 bp fragments were recovered in the bound fraction.To confirm these results and better localize the region recognized by the integrase, two PCR fragments corresponding to the full length or a 40 bp deleted LTR called ΔLTR were amplified and used in the same in vitro DNA binding assays (Fig. 2A). The results indicated that ZAM-IN is unable to retain the ΔLTR (Fig. 2B, middle panel). ZAM-IN contains three domains: a zinc finger amino-terminal motif (HHCC), a core or catalytic domain characterized by the DD35E motif, and a carboxy-terminal part of the protein which displays a high basicity similar to the DNA binding domain of retroviral integrases [25]. In order to test whether this basic domain at the C terminal end of ZAM integrase is important for its LTR DNA binding activity, this integrase deleted for its last 80 (ΔIN) was produced as a GST fusion protein and fixed on agarose beads (Fig. 1A, “ΔIN”). Then, similar in vitro DNA binding assays were performed with this deleted integrase ΔIN. As shown figure 2B (right panel), the ΔIN protein does not bind any of the LTR, K or H fragments of ZAM (upper panel). This result indicates that the C-terminal part of ZAM-IN is required for its binding property on ZAM LTR. It must be noticed that this experiment also confirms that the binding properties observed in the first set of experiments are indeed due to the integrase “IN” expressed in vitro and not to any non-specific binding. To go further, a double strand oligonucleotide labelled with [γ-32P]ATP and spanning nucleotide positions 28 to 40 of the 5′LTR was assayed. As shown figure 2C, this oligonucleotide called BS is retained by the Integrase. Altogether, these results indicate that ZAM-IN is able to bind the LTR of ZAM in a 13 bp site located between nucleotides 28 and 40.In a second set of experiments, we addressed whether specific binding sites for ZAM-IN exist in the 388 bp white fragment which could help to the target site recognition. Through in vitro binding assays, we searched for ZAM-IN binding sites along the 388 bp DNA fragment of white (Fig. 3A).10.1371/journal.pone.0003185.g003Figure 3ZAM Integrase interacts with specific genomic DNA sequences.A) Diagram of the white DNA fragment from nucleotide positions 4278 to 4666. Two PCRs products used in this experiment and called “w4278” and “w4392” are represented underneath. The two AluI restriction sites and the resulting DNA fragment sizes are presented above. The palindromic cleavage site CGCGCG is indicated by a white box. B) In vitro binding assays with ZAM-IN protein performed on the white PCR products “w4278” and “w4392”. C) Assays performed with the white PCR product w4278 digested by AluI. The percentage of bound (upper panels) and unbound (lower panels) fractions is presented below each panel. DNA fragments sizes are indicated for each panel.The GST-IN fusion proteins bound on beads were incubated with a PCR-amplified w4278 fragment (Fig. 1B and 3A). As shown figure 3B (lane w4278), the w4278 fragment is retained by the ZAM-IN (upper panel). This result indicates that ZAM-IN directly binds at least one DNA sequence within the 388 bp of the white locus.Then, we tested whether this binding might occur indifferently along the whole length of w4278 or whether some specific binding sites could be identified. A PCR product called w4392 corresponding to a 274 bp fragment spanning from nucleotide positions 4392 to 4666 was used for further in vitro binding assays (Fig. 3A). As shown in figure 3B (lane w4392), this fragment is not retained by the integrase (upper panel) while it is recovered in the supernatant (lower panel). Thus, ZAM-IN is unable to bind the white sequence between nucleotides 4392 and 4666.To further analyze the DNA fragment comprised between nucleotides 4278 and 4392, the PCR-amplified fragment w4278 was digested by the AluI restriction enzyme to generate three DNA fragments of 48, 66 and 274 bp long (Fig. 3A), and in vitro binding assays were performed. As illustrated figure 3C, the 48 bp and 66 bp fragments are retained by ZAM Integrase (upper panel). By contrast and as expected, the 274 bp fragment is not recognized by ZAM-IN and is only recovered in the supernatant (Fig. 3C, lower panel). These results indicate that ZAM-IN does not bind DNA in a random manner but has the property to recognize specific sequences. Such binding sites are located within the 48 bp and 66 bp fragments of the white fragment analyzed.To go further in their identification, we performed a new set of in vitro binding assays using four oligonucleotides encompassing the full length 48 and 66 bp fragments retained by the Integrase. Their respective positions and sequences are presented figure 4A. These oligonucleotides, called w0, w1, w2 and w3, are 21, 26, 26 and 45 nts long respectively (Fig. 4A). They were firstly annealed to complementary oligonucleotides to form double-stranded DNA molecules (see material and methods) and then used for the DNA binding assays. As illustrated figure 4B, the double-stranded oligonucleotides w1 and w3 are retained by ZAM Integrase whereas w0 and w2 are not.10.1371/journal.pone.0003185.g004Figure 4ZAM integrase binds two specific genomic DNA sites.A) Sequence of the Drosophila white locus from base 4278 to 4326. The oligonucleotides w0, w1, w1mut, w2, w3, Δw3 and w3mut used in the experiments are represented under the sequence. The integration site CGCGCG, the sequence of the mutated integration site of the w1mut oligonucleotide and the nucleotides mutated to generate the w3mut oligonucleotide are indicated by boxes. B) Left panel: In vitro binding assays were performed with ZAM integrase “IN” and the double stranded oligonucleotides w0, w1, w1mut, w2, w3, Δw3 and w3mut. w1 and w3 are the only two oligonucleotides retained by ZAM integrase. Right panel: In vitro endonuclease activity of ZAM integrase: pUC/white and pUC/white3mut plasmids were incubated without (−IN) or with (+IN) purified ZAM-IN. Positions of the supercoiled, nicked (circle) and linear (bar) DNAs are indicated. C) Alignment of a conserved motif detected in the ZAM LTR and w3. The first 60 nucleotides of the LTR sequence are presented as the upper sequence. The binding site of w3 is presented below.Since ZAM insertion site CGCGCg is present within w1, this sequence was likely a binding site of ZAM-IN. To test this possibility, we used an oligonucleotide called w1mut in which the CGCGCG sequence was replaced by an AGAGCG site (Fig. 4A). As shown figure 4B, ZAM integrase is then unable to bind the w1mut oligonucleotide.w3 does not contain any CGCGCG site. Thus, to identify the binding site of ZAM-IN within the w3 fragment, we tested its ability to bind diverse deleted w3 oligonucleotides. We found that when w3 is deleted for 15 bases from its 5′ end (Δw3), ZAM integrase is not able to bind the remaining sequence (Fig. 4B). Thus a second binding site of ZAM-IN is located within the first 15 bp of w3, a region in which little to no sequence similarity with the target site of integration can be detected. When analyzing the sequence of this 15 bp fragment, we detected a palindromic sequence: AGGCCT. Since the target site CGCGCG is also a palindrome, we hypothetized that some specific DNA structures such as hairpin might be recognized by ZAM-IN. A mutated oligonucleotide w3mut in which the palindromic sequence was disrupted was then tested in the same set of in vitro binding assays. As shown in Fig. 4B, ZAM-IN is then unable to bind w3mut. We then performed a new series of experiments similar to experiments presented Fig. 1C and assayed the endonuclease activity of ZAM-IN on a pUC/white3mut plasmid in which the 388 bp fragment of white displays a mutation affecting the second binding site AGGCCTCGTCTATAA converted to AGGCATAGTCTATAA. We found that whereas ZAM-IN is able to convert supercoiled pUC/white molecules to open circles, it is unable to cleave the supercoiled molecules of the pUC/white3mut plasmid. Open circles molecules were not detected after the Integrase treatment (Fig. 4B, right panel). This result contrasts with what has been observed when the CGCGCG motif of the white fragment is mutated (Fig. 1C, lane pUC/white1mut). Indeed, ZAM-IN retains the ability to cleave a plasmid in which the white integration site is mutated whereas this ability is lost when the second binding site is destroyed.\n\nOverall, the above experiments show that the white fragment necessary for the endonuclease activity of ZAM-IN displays two distinct binding sites for ZAM-IN: one of them is the integration site itself CGCGCg and the second displays a different sequence located 40 to 56 bp apart.DiscussionThe retrotransposon ZAM displays an extreme bias in target site selection. Indeed, it integrates in a consensus sequence CGCGCg. On the basis of sequence similarity and gene organization ZAM is a member of a group of retrotransposons that bears a striking resemblance to the vertebrate retroviruses. Its enzymes involved in reverse transcription and integration are similar to retroviruses [25], [26]. Direct binding of retroviral Integrases on their LTR has been well demonstrated and we also showed that ZAM integrase binds its own LTR. However, the specific binding of retroviral IN on the DNA target sites had not been reported yet. So far, only models, in which tethering of integration machinery to host DNA via protein-protein interaction were proposed to be important for integration site selection [31]–[33], and indeed, mechanisms based on tethering strongly explain targeting of some Integrases [8], [20]. Moreover, a clear consensus motif has never been determined despite studies that highlight the influence of the primary DNA sequence in the choice of retroviral integration [7]. Our results indicate that ZAM-IN clearly binds the host DNA, and suggest that the target sites for ZAM integrations are selected through direct interaction between the target DNA and the Integrase. One binding site corresponds to the consensus CGCGCg identified as the integration site, and a second binding site with a different sequence is located in close proximity. Although the DNA characteristics of this second binding site remain to be identified, our data clearly demonstrate that if absent, ZAM-IN is unable to cleave the CGCGCg consensus site. When comparison between these two binding sites identified in the white fragment and the LTR fragment bound by ZAM-IN have been made, some homology was clearly detected between the second binding site of w3 and a motif located between nucleotides 30 and 45 of the LTR (Fig. 4C). Interestingly, the mutation converting the CTC triplet to ATA in w3mut and abrogating ZAM-IN binding is encompassed within this homologous site (see Fig. 4A and C). Thus, although the CGCGCg integration site is a palindromic sequence, we believe that selection of the second site might not necessarily implicate the presence of a palindrome but rather implicate constraints brought by the DNA structure or conformation. A genomic DNA organisation like a strong bending DNA structure could allow ZAM-IN to bind the site.Among retrotransposons, the non-LTR element called R2 encodes a single protein with reverse transcriptase and endonuclease activities. R2 elements specifically insert into 28S rRNA genes of many animal groups. Christensen et al. (2005) have shown that the complete mechanism of integration involves two R2 protein subunits [34]. The first subunit binds upstream of the cleavage site and is responsible for the initial cleavage and reverse transcription step, while the second subunit binds downstream and is responsible for second-strand cleavage. Such properties are also observed for some restriction endonucleases like FokI, MboII or MlyI which bind a specific target sequence and cleave at a conserved distance from this binding site [35]. According to our results, this strategy is thus likely to be the one used by ZAM.ZAM is generally present at a very low copy number in the lines of Drosophila melanogaster so that its mutagenic impact is low. However, we identified a line in which its transposition frequency suddenly increased and is correlated with a high copy number of ZAM. These insertions were found dispersed on the chromosomal arms [36], [37]. In this context, disruption of required cellular genes by these insertions could have meant suicide for both ZAM and its host. Nevertheless, the recent 20 to 30 copies of ZAM in this line have very little effect on the general biology of the host, and no clear sterility or decrease in the life cycle of the line could be detected. The characteristics of ZAM-IN reported here cannot alone explain the selection of target sites that do not compromise the health of its host. This observation suggests that in vivo, host factors might also contribute to the targeting of ZAM Integrase to safety regions of the genome. Experiments are under investigation to identify putatively tethering of ZAM-IN to host proteins having by themselves an additional specific recognition target. Preliminary results through two hybrid experiments have indicated that ZAM integrase interacts with SNR1, a protein of the SWI/SNF chromatin remodelling complex. Might chromatin remodelling complexes participate to the targeting of ZAM copies at specific genomic site? Further analyses are necessary to better understand the influence of such host factors in the specificity of ZAM integration.Retroviral vectors, which integrate the host chromosomes, are the most widely used method of gene transfer in mammals. However, such insertions within the genome come with a cost. Insertions near cellular proto-oncogenes leading to ectopic gene activation have been seen in two patients undergoing retrovirus-based gene therapy [38]. Understanding the molecular mechanism underlying integration site selection of elements related to retroviruses such as ZAM brings the hope that some new strategy will be found to direct integration to innocuous chromosomal sites and avoid problems generated by the little target specificity of vectors currently used.Materials and MethodsGST-Integrase expression and purificationOligonucleotides ZAM5322BamHI (gaatccgatgcaaatcacttc) and ZAM6207BamHICi (ggattcctgttaggttgtact) or ZAM6448BamHICi (ggattcctaggaggttggtgc) were used to clone at the BamHI restriction site, respectively, the full length ZAM integrase or a deleted ZAM integrase peptide called ΔIN in frame with the GST protein in the pGEX-5-X1 vector (Amersham Pharmacia Biotech). BL21 transformant colonies were inoculated in 100 ml of LB/ampicillin medium and incubated over night at 37°C. Expression of both GST-IN and GST-ΔIN fusion proteins in Escherichia coli BL21 was induced for 4 hrs at 30°C with 0.1 mM IPTG (isopropyl-β-D-thiogalactopyranoside). Pellets were resuspended in 10 ml of ice-cold solubilization buffer (50 mM Tris–HCl pH 7.4, 1 mM EDTA, 100 mM NaCl, 10% glycerol, 1% NP-40, 1 mM DTT, 1 nM PMSF, 10 µg/ml aprotinin, 2 µg/ml leupeptin, 2 µg/ml pepstatin, 0.5 mg/ml lysozyme). After sonication, supernatants were incubated for 30 min with 1 ml of 50% glutathione–agarose beads, washed three times in 1 M NaCl, three times in PBS and resuspended in 1 ml of PBS. The GST-IN and GST-ΔIN proteins fixed on agarose beads were used for in vitro DNA binding assays. For in vitro endonuclease experiments, GST-IN fusion proteins were eluted from beads by incubating for 30 min at 4°C in 10 mM glutathione/50 mM Tris–HCl, pH 9.Constructs and Direct mutagenesisThe white fragment from nucleotide 4278 to 4666 (according to the accession number: X02974) was amplified by PCR on a genomic template and cloned in pUC18 giving rise to the pUC/white plasmid. From the pUC/white plasmid, direct mutagenesis (Pfu Turbo DNA Polymerase from Stratagene) of the palindromic site from CGCGCG to AGAGCG was performed with the following sense and reverse oligonucleotides: white1mut:“tttttatgagacaagagcgtgctgtaacct” and white1mutCi “aaaaatactctgttctcacgacattgga”. The resulting plasmid was called pUC/white1mut.\n\n\nIn vitro endonuclease reactions\nIn vitro endonuclease reactions were performed as followed: 0.5 µg of 5′LTR substrate, 10 ng of purified ZAM-IN fusion protein, and 2 µg of target DNA (pUC18, pUC/white, and pUC/white1mut) were incubated in 20 µl of 20 mM Tris (pH 8.0)-0.01% bovine serum albumin-1 mM dithiothreitol -10% dimethyl sulfoxide, 2 mM MnCl2 and 10 mM MgCl2 at 30°C for 2 hour. Analysis of DNA was performed on a 1% agarose gel stained by Ethidium Bromide.\nIn vitro DNA binding assayThe full length 5′LTR of ZAM (473 bp long) called “LTR” and a “ΔLTR” deleted for the first 40 bp were amplified by PCR using forward primers ZAM1: (agttaccgacccatcggtacc) or ZAM40: (taagccaccacgcctacacaa), respectively, and the reverse primer ZAM473CI: (agttacctccggggagtcttg). The full length LTR product was digested with either KpnI or HindIII located at positions 21 and 82, respectively, from the 5′end of the 5′LTR sequence of ZAM. The LTR and ΔLTR PCR products as well as the KpnI and HindIII digested fragments were used for in vitro DNA binding experiments. Moreover, a doubled stranded oligonucleotide called “BS” from base 28 to 40 according to ZAM sequence was labelled with [γ-32P]ATP by a kinase reaction (Invitrogen) and used in this study. The white locus fragments from position 4278 to 4666 and 4392 to 4666 were amplified by PCR and called w4278 and w4392 respectively. PCR products of w4278 were also digested by the AluI restriction enzyme. These two PCR products and the AluI digested fragments were used for in vitro DNA binding experiments. Double stranded white oligonucleotides: w0 (cccaacggatgttttgatacg), w1 (tttttatgagacgcgcgcgtgctgta), w1mut (tttttatgagacaagagcgtgctgtaa), w2 (agctaacgccgacttccgcttgccat), w3 (aggcctcgtctataactcccggccacgcctcctctcctccagct), w3mut (aggcatagtctataactcccggccacgcctcctctcctccagct), Δw3 (ctcccggccacgcctcctctcctccagct) were also used in these experiments. For each reaction, DNA fragments or oligonucleotides were mixed with 20 µl of GST–IN protein fixed on glutathione–agarose beads in the binding buffer [10 mM HEPES, 50 mM KCl, 1 mM DTT, 2.5 mM MgCl2, 20 µg/ml poly(dI–dC), 7.5% glycerol pH 7.9] and incubated at room temperature for 3 h. Beads were pelleted and washed three times in a binding buffer containing 100 mM NaCl to remove all fragments that were not tightly bound. DNA that remained bound to the beads was extracted by phenol/chloroform, precipitated and resuspended in TE before being analyzed on a 1% agarose gel or a 15% polyacrylamide gel stained by Ethidium Bromide.\n\nREFERENCES:\n1. CoffinJMHughesSHVarmusHEVogtPKVogtVM\n1997\nRetrovirus Edited by J. M. Coffin, S. H. Hugues and H. E. Varmus\nCold Spring Harbor, New York\nCold Spring Harbor Laboratory Press\n2. LewinskiMKBushmanFD\n2005\nRetroviral DNA integration–mechanism and consequences.\nAdv Genet\n55\n147\n181\n16291214\n3. MitchellRSBeitzelBFSchroderARShinnPChenH\n2004\nRetroviral DNA integration: ASLV, HIV, and MLV show distinct target site preferences.\nPLoS Biol\n2\nE234. Epub 2004 Aug 2017\n15314653\n4. BushmanFD\n2003\nTargeting survival: integration site selection by retroviruses and LTR-retrotransposons.\nCell\n115\n135\n138\n14567911\n5. WuXBurgessSM\n2004\nIntegration target site selection for retroviruses and transposable elements.\nCell Mol Life Sci\n61\n2588\n2596\n15526164\n6. PryciakPMVarmusHE\n1992\nNucleosomes, DNA-binding proteins, and DNA sequence modulate retroviral integration target site selection.\nCell\n69\n769\n780\n1317268\n7. WuXLiYCriseBBurgessSMMunroeDJ\n2005\nWeak palindromic consensus sequences are a common feature found at the integration target sites of many retroviruses.\nJ Virol\n79\n5211\n5214\n15795304\n8. HolmanAGCoffinJM\n2005\nSymmetrical base preferences surrounding HIV-1, avian sarcoma/leukosis virus, and murine leukemia virus integration sites.\nProc Natl Acad Sci U S A\n102\n6103\n6107. Epub 2005 Mar 6131\n15802467\n9. WuXLiYCriseBBurgessSMMunroeDJ\n2005\nWeak palindromic consensus sequences are a common feature found at the integration target sites of many retroviruses.\nJ Virol\n79\n5211\n5214\n15795304\n10. PrussDBushmanFDWolffeAP\n1994\nHuman immunodeficiency virus integrase directs integration to sites of severe DNA distortion within the nucleosome core.\nProc Natl Acad Sci U S A\n91\n5913\n5917\n8016088\n11. PryciakPMSilAVarmusHE\n1992\nRetroviral integration into minichromosomes in vitro.\nEmbo J\n11\n291\n303\n1310932\n12. TaganovKDCuestaIDanielRCirilloLAKatzRA\n2004\nIntegrase-specific enhancement and suppression of retroviral DNA integration by compacted chromatin structure in vitro.\nJ Virol\n78\n5848\n5855\n15140982\n13. MaxfieldLFFraizeCDCoffinJM\n2005\nRelationship between retroviral DNA-integration-site selection and host cell transcription.\nProc Natl Acad Sci U S A\n102\n1436\n1441. Epub 2005 Jan 1419\n15659548\n14. EngelmanA\n2003\nThe roles of cellular factors in retroviral integration.\nCurr Top Microbiol Immunol\n281\n209\n238\n12932079\n15. BoekeJDDevineSE\n1998\nYeast retrotransposons: finding a nice quiet neighborhood.\nCell\n93\n1087\n1089\n9657139\n16. KimJMVanguriSBoekeJDGabrielAVoytasDF\n1998\nTransposable elements and genome organization: a comprehensive survey of retrotransposons revealed by the complete Saccharomyces cerevisiae genome sequence.\nGenome Res\n8\n464\n478\n9582191\n17. DevineSEBoekeJD\n1996\nIntegration of the yeast retrotransposon Ty1 is targeted to regions upstream of genes transcribed by RNA polymerase III.\nGenes Dev\n10\n620\n633\n8598291\n18. ChalkerDLSandmeyerSB\n1993\nSites of RNA polymerase III transcription initiation and Ty3 integration at the U6 gene are positioned by the TATA box.\nProc Natl Acad Sci U S A\n90\n4927\n4931\n8389458\n19. KirchnerJConnollyCMSandmeyerSB\n1995\nRequirement of RNA polymerase III transcription factors for in vitro position-specific integration of a retroviruslike element.\nScience\n267\n1488\n1491\n7878467\n20. ZhuYZouSWrightDAVoytasDF\n1999\nTagging chromatin with retrotransposons: target specificity of the Saccharomyces Ty5 retrotransposon changes with the chromosomal localization of Sir3p and Sir4p.\nGenes Dev\n13\n2738\n2749\n10541559\n21. XieWGaiXZhuYZappullaDCSternglanzR\n2001\nTargeting of the yeast Ty5 retrotransposon to silent chromatin is mediated by interactions between integrase and Sir4p.\nMol Cell Biol\n21\n6606\n6614\n11533248\n22. DaiJXieWBradyTLGaoJVoytasDF\n2007\nPhosphorylation regulates integration of the yeast Ty5 retrotransposon into heterochromatin.\nMol Cell\n27\n289\n299\n17643377\n23. LeblancPDessetSDastugueBVauryC\n1997\nInvertebrate retroviruses: ZAM a new candidate in D.melanogaster.\nEmbo J\n16\n7521\n7531\n9405380\n24. FinneganDJ\n1994\nRetroviruses and transposons. Wandering retroviruses?\nCurr Biol\n4\n641\n643\n7953544\n25. LeblancPDastugueBVauryC\n1999\nThe integration machinery of ZAM, a retroelement from Drosophila melanogaster, acts as a sequence-specific endonuclease.\nJ Virol\n73\n7061\n7064\n10400810\n26. ArnaudFPeyretailladeEDastugueBVauryC\n2005\nFunctional characteristics of a reverse transcriptase encoded by an endogenous retrovirus from Drosophila melanogaster.\nInsect Biochem Mol Biol\n35\n323\n331\n15763468\n27. GoffSP\n1992\nGenetics of retroviral integration.\nAnnu Rev Genet\n26\n527\n544\n1482125\n28. LewinskiMKYamashitaMEmermanMCiuffiAMarshallH\n2006\nRetroviral DNA integration: viral and cellular determinants of target-site selection.\nPLoS Pathog\n2\ne60\n16789841\n29. EspositoDCraigieR\n1998\nSequence specificity of viral end DNA binding by HIV-1 integrase reveals critical regions for protein-DNA interaction.\nEmbo J\n17\n5832\n5843\n9755183\n30. JohnsonEPBushmanFD\n2001\nPaired DNA three-way junctions as scaffolds for assembling integrase complexes.\nVirology\n286\n304\n316\n11485398\n31. BushmanF\n1995\nTargeting retroviral integration.\nScience\n267\n1443\n1444\n7878462\n32. BradyTLSchmidtCLVoytasDF\n2008\nTargeting integration of the Saccharomyces Ty5 retrotransposon.\nMethods Mol Biol\n435\n153\n163\n18370074\n33. ZhuYDaiJFuerstPGVoytasDF\n2003\nControlling integration specificity of a yeast retrotransposon.\nProc Natl Acad Sci U S A\n100\n5891\n5895\n12730380\n34. ChristensenSMEickbushTH\n2005\nR2 target-primed reverse transcription: ordered cleavage and polymerization steps by protein subunits asymmetrically bound to the target DNA.\nMol Cell Biol\n25\n6617\n6628\n16024797\n35. WahDAHirschJADornerLFSchildkrautIAggarwalAK\n1997\nStructure of the multimodular endonuclease FokI bound to DNA.\nNature\n388\n97\n100\n9214510\n36. BaldrichEDimitriPDessetSLeblancPCodipietroD\n1997\nGenomic distribution of the retrovirus-like element ZAM in Drosophila.\nGenetica\n100\n131\n140\n9440265\n37. DessetSConteCDimitriPCalcoVDastugueB\n1999\nMobilization of two retroelements, ZAM and Idefix, in a novel unstable line of Drosophila melanogaster.\nMol Biol Evol\n16\n54\n66\n10331252\n38. CheckE\n2002\nRegulators split on gene therapy as patient shows signs of cancer.\nNature\n419\n545\n546"
+}

batch_3/PMC2527528.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527528",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527528\nAUTHORS: James J. McDevitt, Donald K. Milton, Stephen N. Rudnick, Melvin W. First\n\nABSTRACT:\nIn the event of a smallpox outbreak due to bioterrorism, delays in vaccination programs may lead to significant secondary transmission. In the early phases of such an outbreak, transmission of smallpox will take place especially in locations where infected persons may congregate, such as hospital emergency rooms. Air disinfection using upper-room 254 nm (UVC) light can lower the airborne concentrations of infective viruses in the lower part of the room, and thereby control the spread of airborne infections among room occupants without exposing occupants to a significant amount of UVC. Using vaccinia virus aerosols as a surrogate for smallpox we report on the effectiveness of air disinfection, via upper-room UVC light, under simulated real world conditions including the effects of convection, mechanical mixing, temperature and relative humidity. In decay experiments, upper-room UVC fixtures used with mixing by a conventional ceiling fan produced decreases in airborne virus concentrations that would require additional ventilation of more than 87 air changes per hour. Under steady state conditions the effective air changes per hour associated with upper-room UVC ranged from 18 to 1000. The surprisingly high end of the observed range resulted from the extreme susceptibility of vaccinia virus to UVC at low relative humidity and use of 4 UVC fixtures in a small room with efficient air mixing. Increasing the number of UVC fixtures or mechanical ventilation rates resulted in greater fractional reduction in virus aerosol and UVC effectiveness was higher in winter compared to summer for each scenario tested. These data demonstrate that upper-room UVC has the potential to greatly reduce exposure to susceptible viral aerosols. The greater survival at baseline and greater UVC susceptibility of vaccinia under winter conditions suggest that while risk from an aerosol attack with smallpox would be greatest in winter, protective measures using UVC may also be most efficient at this time. These data may also be relevant to influenza, which also has improved aerosol survival at low RH and somewhat similar sensitivity to UVC.\n\nBODY:\nIntroductionSmallpox (variola major) is a high priority bioterrorist threat agent, according to the Centers for Disease Control and Prevention and Department of Homeland Security, which can be easily transmitted from person to person, result in high mortality rates, might cause public panic and social disruption, and require special action for public health preparedness (http://www.bt.cdc.gov/agent/agentlist-category.asp). Airborne spread via respiratory droplet nuclei has been identified as a potential contributing mode of transmission for smallpox[1], [2] and prevention of transmission by vaccination will likely be delayed until public health authorities become aware of the outbreak and initiate a vaccination program. In the early phases of such an outbreak, significant secondary transmission of smallpox will take place especially in locations where infected persons may congregate, such as hospital emergency rooms. Therefore, public health measures in addition to vaccination are needed.Hospitals limit aerosol disease transmission in indoor spaces by reducing the concentration of airborne microorganisms through dilution ventilation. However, these measures are largely impractical beyond a limited number of respiratory isolation rooms due to the large amounts of air exchange needed to significantly reduce the threat of infection and are therefore costly in terms of heating and cooling these large amounts of air. The high ventilation rates required for respiratory isolation rooms are not routinely used in emergency departments and waiting areas. With air disinfection, costs are reduced since air does not have to be removed from occupied spaces to remove potential infectious agents. Disinfection using high-efficiency filtration to significantly reduce the threat of airborne infection can be effective but requires more powerful fans beyond what currently exist in the majority of public buildings and also require additional energy consumption. Air disinfection using upper-room 254 nm (UVC) light can lower the airborne concentrations of infective organisms in the lower part of the room, and thereby control the spread of airborne infections among room occupants without exposing occupants to a significant amount of UVC.[3]–[5] Upper-room UVC systems do not require modification to ventilation systems, are low maintenance, and relatively easy to install.[6], [7] The use of upper-room UVC is also economical. For example, the 25-watt lamps used as part of our study would cost just over $40 per year assuming an electrical cost of $0.20 per kilowatt-hour. Some hospitals currently employ upper-room UVC for this purpose in their emergency departments (e.g. Brigham and Women's Hospital, Boston, MA), but its effectiveness against viral aerosols is not well established.Inactivation of microorganisms using UVC is often assumed to follow a first-order decay with a susceptibility parameter Z = ln(1/f) / D, where (f = organism fractional survival and D = UV dose, where dose is the product of UV fluence rate –expressed as power per cross sectional area—and exposure time, for example mJ/cm2. Using a one-pass UVC exposure chamber, however, we have shown that vaccinia virus (a surrogate for variola major) is susceptible to UVC and that the susceptibility varies as a function of dose and relative humidity (RH).[8] In these dose-response experiments the fluence rate and exposure time, and therefore, dose were carefully controlled. Thus, in each experiment, all viruses received the same dose and we determined susceptibility to UVC by varying dose over several experiments. In an actual room using upper-room UVC, the UVC fluence rate varies even within the upper-room, and the time spent in the upper-room varies from particle to particle. Therefore, the dose for each viral particle depends on the path that the particle travels. With perfect mixing, particle doses would be exponentially distributed. In the case of imperfect mixing, computational fluid dynamic (CFD) models should be capable of describing the more complex distribution of doses that would result. Then, using the pattern of UVC susceptibility we previously reported, it should be possible to estimate the net effectiveness of upper-room UVC. However, given the complexity of UVC susceptibility that we previously described combined with the complexity of CFD models, empirical data are needed. We report experiments designed to measure the effectiveness of upper-room UVC under simulated real world conditions including the effects of convection, mechanical mixing, temperature and relative humidity (RH).ResultsDecayThe environmental conditions within the chamber during decay experiment were maintained at 20±3°C and 50±10% RH. The results of chamber decay experiments performed with background decay, without heat boxes, and with heat boxes are shown either without the ceiling fan operating (Figure 1a) or with the ceiling fan operating (Figure 1b). The exponential regression model fits the data reasonably well. The rate constant shown in these equations can be interpreted as the effective air exchange rate for the chamber expressed in units of air changes per hour (ACH). Based on a model for a chamber in which the air is perfectly mixed, the effective air exchange rate is equal to the amount of virus-free dilution air that would be needed to provide the same reduction of virus concentration that was actually measured. The background decay rate reflects the decrease in infective viruses due to the exhaust airflow required to maintain negative pressure within the chamber, as well as any physical and non-UVC-related biological decay of the virus aerosol.10.1371/journal.pone.0003186.g001Figure 1Background decay rates and decay rates for UVC light with and without heat boxes.a) ceiling fan is not operational; b) when ceiling fan is operational.Virus reduction due to upper-room UVC is equal to the effective air exchange rate with the UVC turned on minus the effective air exchange rate with the UVC turned off (i.e. the background decay rate). This difference, which is usually referred to as equivalent air changes per hour (ACHUVC), is summarized in Table 1. Overall, the rate of reduction of vaccinia virus increased over the background as the amount of natural convection increased; mixing by the ceiling fan overwhelmed natural convection effects and markedly increased virus inactivation. When the ceiling fan was not operating, the ACHUVC increased by 7 ACH above background when viruses were dispersed at 37°C (body temperature). When additional convective currents were added to the room by the addition of two heat boxes (equivalent to the heat generated by two people) the ACHUVC increased by 16 ACH. When the ceiling fan was in operation the ACHUVC increased to greater than 87 ACH and there was no discernable effect attributable to the heat boxes.10.1371/journal.pone.0003186.t001Table 1Equivalent Air Changes per Hour Due to UVC (ACHUVC) for Virus Aerosol Decay Tests with and without ceiling fan and heat boxes.Ceiling Fan OperationalNoYes\nWithout Heat Boxes\n792\nWith Heat Boxes\n1687Steady StateThe average concentration of vaccinia aerosols during steady state conditions with UVC off ranged from 1500 to 27000 pfu/m3. One experiment (summer conditions with 2 ACH and 4 UVC fixtures), in which the initial concentration before the UVC was turned on (1500 pfu/m3) was much lower than any of the other experiments was not used in our analysis, because the initial concentration was too low to accurately measure >85% reductions in concentration. The geometric mean vaccinia concentrations without UVC for the experiments used in the analysis were 3400 (95% confidence interval 2600 to 4300) pfu/m3 under summer conditions and 7800 (CI 5800 to 10000) pfu/m3 under winter conditions. With UVC on, the geometric mean concentrations were 570 (CI 430 to 770) pfu/m3 in the summer and 110 (CI 79 to 150) pfu/m3 in the winter experiments. The experiments under summer conditions showed stronger time trends in aerosol concentrations and greater variability between experimental replicates (i.e. steady state was difficult to achieve even without the UVC fixtures). The fraction of infectious virus remaining (ratio of the concentration of virus at steady state with upper-room UVC on to that measured under steady state conditions without UVC) is shown in Table 2 for the various combinations of tested conditions: two ventilation rates (2 and 6 ACH), numbers of UVC fixtures (1 or 4 fixtures) and seasonal conditions (summer and winter). Equivalent air changes due to UVC under steady state conditions for the various test conditions are shown in Figure 2. UVC achieved greater than 85% reduction in virus aerosol concentrations for all test conditions. Increasing the number of UVC fixtures from 1 to 4 resulted in greater fractional reduction in virus aerosol concentration at both 2 and 6 ACH. The fraction of virus surviving UVC treatment was lower under winter conditions compared to summer conditions. The additional effective air changes per hour due to UVC at each ventilation rate were 4 to 19 times greater during winter than summer.10.1371/journal.pone.0003186.g002Figure 2Equivalent air changes due to UVC under steady state conditions with either 2 or 6 ACH, 1 or 4 UVC fixtures, or winter and summer conditions.10.1371/journal.pone.0003186.t002Table 2Effective Air Changes per Hour Associated with Upper-Room UVC for Steady State Virus Aerosols.ConditionReplicate Trials (Air Samples)Ventilation Rate, ACHUVC FixturesFraction of Infective Virus RemainingACHUVC\nEstimate95% Confidence IntervalEstimate95% Confidence IntervalSummer2 (20)210.0870.0620.1201815301 (12)240.0610.0530.0713126363 (36)610.140.1200.1603831463 (36)640.0780.0650.094715886Winter2 (24)210.0170.0140.021110931402 (24)240.0030.0020.0055804108302 (24)610.0380.0320.0461501201802 (24)640.0060.0040.00810007401400DiscussionThese data show that in a ‘real world’ test setup, upper-room UVC is highly effective for reducing the concentration of vaccinia virus aerosols. We demonstrate through aerosol decay experiments that upper-room UVC fixtures used with mixing provided by a conventional ceiling fan and minimal general ventilation produced decreases in airborne virus concentrations that would require additional ventilation of more than 87 ACH. During steady-state experiments the combined effect of upper-room UVC and ventilation had a nonlinear impact on the fraction of remaining virus aerosol.[9] As a result, under winter conditions when vaccinia is most susceptible to UVC inactivation, the effective ACH due to upper-room UVC (ACHUVC) increased approximately five fold with increasing air exchange from ventilation. The equivalent ventilation achieved by UVC ranged from a low of 18 to 1000 ACHUVC, with winter equivalent ventilation rates consistently >100 ACHUVC.The results from our decay experiments confirm the importance of vertical mixing cited for UVC effectiveness in model rooms.[3], [10] Vertical mixing is required to move organisms from the lower room into the upper-room where UVC intensity is the highest. Without vertical mixing, some viruses may get less UVC exposures or not get exposed at all and, as a result, UVC doses may be insufficient to cause deactivation.[9] During our decay experiments the ventilation system was operated so as to provide minimum negative pressure inside of the chamber to facilitate aerosol containment, while providing minimal mixing and dilution. As a result, when the heat boxes were not activated vertical mixing was primarily attributable to convection associated with the nebulizer diffuser which was heated to 37°C (approximately 17°C above the chamber temperature). The effective air exchange rate was a modest 7 ACHUVC above the background. The effective ACH rate more than doubled (16 ACHUVC) when the heat given off by two people present in the room was simulated by the activation of the heat boxes. This value is consistent with the findings of Riley et el[11] for mycobacteria and are greater than what would be achieved by recommended dilution ventilation in hospital isolation rooms.[12] UVC and mixing using a ceiling fan together produced virus aerosol decay rates equivalent to 87 ACHUVC and thus overwhelmed free convection effects. Similarly, First et al were able to show marked reduction in survival when comparing bacterial aerosol decay with and without ceiling fans in operation.[9]\nAlthough these data are strong indicators that UVC would be an effective intervention, it has been recommended that tests of the efficacy of UVC against bioaerosols be based on steady-state measurements rather decay experiments.[10] We performed steady-state experiments under both summer and winter conditions. Consistent with early experiments on virus aerosol stability[13], [14], in the absence of UVC, vaccinia virus appeared to be more stable and higher aerosol concentrations were achieved with low RH (winter conditions) than with high RH (summer conditions). These experimental results also show that upper-room UVC is more effective when the relative humidity is low, even though mixing was reduced by operating the ceiling fans on a low, updraft, winter setting. These results are consistent with our bench-top experiments showing that vaccinia aerosols are more sensitive to UVC when relative humidity is low.[8]\nExamining the fractional reduction of viral aerosol concentrations under various conditions clearly shows that upper-room UVC is capable of greatly reducing exposure. But, fraction reduction measurements do not easily translate into estimates of the actual level of risk achieved or facilitate decision making about how to best deploy upper-room UVC as part of a protection strategy. To estimate the level of risk with, for example, the Wells-Riley equation, we need to convert the fractional survival measurements into equivalent ventilation rates.[10], [15] This is easily done because at steady-state the ratio of total effective sanitary ventilation (QUV+Q) to actual ventilation through air movement is equal to the ratio of virus concentration without UVC to the concentration with UVC (the inverse of the remaining fraction fss, i.e. (QUV+Q)/Q = fss\n−1), where QUV stands for the supply of virus free air due to UV (see Appendix S1) and Q is the ventilation rate with infective-virus-free air (m3/s).[9] If the fraction of infectious virus remaining were constant when the air exchange was tripled, then the total effective sanitary ventilation and effective ventilation due to UVC would also be tripled. However, in these data, when we tripled the air exchange rate from 2 to 6 ACH, the fraction of infectious virus remaining increased. This does not, however, imply that upper-room UVC gives less protection when ventilation is increased. It is true that while increased ventilation reduced the virus aerosol concentration it also reduced the average residence time of viral particles resulting in lower UVC doses to individual particles. But, the increase in f was not great enough to completely offset the more than additive effect of increased air exchanges. When we increased the air exchange rate from 2 to 6 ACH, a factor of 3, the effective ventilation due to UVC increased by a factor of 1.3 to 1.9. Thus, increased ventilation actually increased UVC fixtures effectiveness in terms of ACHUV – the combination of ventilation and upper-room UVC is more than merely additive.With one UVC fixture under summer conditions, when we increased Q by 4 ACH the effective ventilation from UVC increased by 20 ACH. In the winter with one fixture, when we increased the air exchange by 4 ACH, the effective ventilation from UVC increased by 40 ACH, and with 4 fixtures the effective ventilation increased by 420 ACH. The high UVC susceptibility of vaccinia when RH is low, i.e. the very small f observed under winter conditions, and the nonlinear interaction of UVC disinfection with ventilation produced extremely highly effective ventilation when the two were combined – ranging from >100 to 1000 ACH. In our previous bench top, dose-response studies of vaccinia virus, moderate UVC doses (3 J/m2) reduced vaccinia survival by a factor of >10,000 over natural biological decay.[8] For the present study we used an equation developed by Rudnick and First[16], that relates UVC fixture power output to mean fluence rate for the entire room, to estimate the mean UVC dose for the entire room assuming near perfect mixing, one fixture was in use, and a ten minute exposure time (i.e. 6 AC/hr). Under these conditions the UVC dose was estimated to be 17 J/m2. With four fixures in use the dose would be expected to be 4-times higher. Thus, the fraction of virus surviving, especially when they are most susceptible, would be expected to be quite low.Studies by other researchers have made similar measures of UVC light effectiveness under steady-state conditions with bacterial aerosols. Bacteria such as bacillus subtillus and serratia marcescens have been used in full scale tests of upper-room UVC[4], [9] Bacteria are much more resistant to UVC light and have a correspondingly lower UVC susceptibility parameter, referred to as a Z-value. Riley and Kaufman noted decreased susceptibility to UVC for Serratia Marcescens exposed to UVC when RH exceeded 60% RH.[19] Ko et al noted a similar RH trends with S. marcescens and Mycobacterium bovis aerosols exposed to UVC.[20] Our previous studies of Vaccinia virus aerosol showed vaccinia virus susceptibility was highest when relative humidity was low.[8] Thus, the very high UVC susceptibility of vaccinia virus, especially when relative humidity is low [8], most likely accounts for the extremely high effective air changes per hour associated with UVC we found in comparison to studies[3] using bacterial aerosols. First et al. evaluated vaccinia virus in a full-scale chamber under steady-state conditions at 50% RH and reported similar results to our summer conditions.[9]\nThe additional effective ventilation due to upper-room UVC, even in the summer, ranged between 18 and 71 ACH, rates in excess of what is usually achieved in hospital rooms designed for airborne precautions (approximately 12 ACH) [12]. The additional effective ventilation achieved under winter conditions was phenomenal (110 with a single fixture to 1000 ACH with four fixtures). These data demonstrate that upper-room UVC has the potential to greatly reduce exposure to susceptible viral aerosols. The greater survival at baseline and greater UVC susceptibility of vaccinia under winter conditions suggest that while risk from an aerosol attack with smallpox would be greatest in winter, protective measures using UVC may also be most efficient at this time. These data may also be relevant to influenza, which also has improved aerosol survival at low RH. Given current concern about potential for a pandemic in the near future, and the potential that an important fraction of influenza transmission occurs via aerosols, further studies of UVC susceptibility and upper-room UVC effectiveness for influenza are warranted.Materials and MethodsExperimental ChamberThe testing chamber, ante room, aerosol generation, and sampling arrangements have been described previously and are shown in Figure 3.[9] Briefly, virus aerosols were delivered at 1.5 meters above the floor in the center of a climate controlled 4.60 m×2.97 m×3.05 m high room equipped with a ceiling fan and two black boxes containing 100-watt light bulbs (simulate body heat of two people). The boxes were located approximately one meter from the center of the room. UVC light was provided by combinations of 5 wall-mounted Hygeaire UVC fixtures (Model LIND24-EVO; Atlantic Ultraviolet Corp., Bay Shore, NY), each using one 25-W low pressure mercury discharge lamp with a UVC output of 5W. Fixtures were mounted 2.3 m from the floor and experiments with one fixture used a single fixture pointed down the middle of the long axis of the room, while experiments with four fixtures used two fixtures on each end of the long axis mounted one meter from the wall corners.10.1371/journal.pone.0003186.g003Figure 3Schematic diagram of aerosol chamber and equipment.Note: for clarity ceiling fan is not shown, but is located in the center of the main chamber directly above the virus distributor. For decay and single fixture steady state experiments the fixture shown in the center of the wall on the left of the figure was used and for the four fixture steady state experiments the other four fixtures were used.Aerosol generation and samplingVaccinia virus stock, Western Reserve strain, was prepared to a concentration of 107–108 plaque forming units (PFU)/ml as reported previously[8]. Vaccinia stock solution was suspended in phosphate buffered saline (PBS) with 10% fetal bovine serum and 20 µl of Antifoam A (Sigma, St. Louis). Vaccinia virus aerosols were generated using a 6-jet Collison nebulizer (BGI Inc., Waltham, MA) operating at 138 kPa. The nebulizer was located in a class II biological safety cabinet (BSC) in the ante room and attached to a permanently installed pipe leading to the center of the test chamber. An omni-directional diffuser was attached to the end of the pipe at 1.5 meters above the floor. The pipe was heated to (37°C).A port for aerosol sampling was located in front of the exhaust grill and was connected via a pipe to a valve located within the BSC in the control room. Air was drawn through a two-way valve into either a 37 mm gelatin filter (SKC, Inc.Eighty Four, PA) housed in a polyethylene cassette or through a bypass at 28.3 lpm. Bypass or filtered air was then directed through a high efficiency particulate air (HEPA) filter located before the high volume sampling pump. The bypass was used to clear the dead space in the sampling tube prior to sampling (60 sec) and when changing the filters. Filters were dissolved and vaccinia viruses were enumerated by plaque assay on confluent layers of Vero cells as previously described.[8]\nDecay experimentsVaccinia was aerosolized for approximately 30 minutes to achieve sufficiently high concentrations of virus to allow detection after multiple logs of reduction. The generation was stopped and 5-minute samples were taken at 5 to 10 min intervals for up to 90 min. The aerosolization and sampling procedure was repeated with one UVC fixture on (Figure 3) alternating with no UVC fixture decay runs. Each experiment consisted of three pairs of runs with UVC on and off. Decay experiments were carried out without heat boxes or ceiling fan, with heat boxes, and with heat boxes and the ceiling fan.SteadystateUVC inactivation of vaccinia virus was tested under steady state conditions while simulating indoor summer (20°C, 80% RH, ceiling fan directing air downwards) and indoor winter (20°C, 40% RH, ceiling fan directing air upwards) environmental conditions, with either 2 or 6 ACH ventilation rates, and either 1 or 4 UVC light fixtures (Figure 3). We assumed that 3 air changes were sufficient to establish a 95% chamber equilibration. Thus, virus suspension was nebulized for 30 minutes prior to sampling to achieve steady-state at 6 ACH and 1.5 hours for 2 ACH. Triplicate sequential samples were collected with the UV fixtures off followed by activation of the UVC fixtures and 3 air changes to allow equilibration. Then, triplicate sequential samples were collected with the UVC fixtures on. The fixtures were then turned off and the cycle of sampling with fixtures off was repeated. Each sample was assigned a time of collection as the midpoint of the sampling interval relative to the start of virus nebulization.Data AnalysisDecay experiment observations for the number of pfu/m3 for each sample were divide by the pfu/m3 in the initial sample collected after aerosolization was complete (collected approximately from t = 0 and t = 5 min) to obtain an estimate of the fraction of infectious virus remaining at each time point within an experiment. Each estimate of fraction remaining was assigned to the midpoint of the sampling interval. Thus, t = 2.5 min was assigned the preliminary value 1.0 for fraction remaining. An exponential decay curve was fit to these data using following equation:where fd is the fraction of infectious virus surviving and k is a rate (or decay) constant, and t is time. Each estimate of the fraction remaining was then adjusted by dividing all fractions remaining by the y-intercept of the regression. These adjusted fraction remaining estimates from each of the triplicate experiments with the same conditions were combined in a single regression to estimate the exponential decay constant (i.e. the effective air exchange rate). The equivalent air exchange rate due to UVC is the difference between the decay constant when UVC is in use and when it is not.[9]\nFor steady-state experiments, we computed pfu/m3 for each air sample. For each set of experimental conditions, we regressed ln(pfu/m3) on time, time squared, an indicator variable for operation of the UVC fixtures, an indicator for each experiment, and interactions of experiment indicators with the time variables. This allowed us to determine the effect of UVC controlled for experiment specific time trends in nebulizer output and variations in the virus aerosol concentrations achieved in each replicate experiment. The resulting coefficient for the indicator of UVC operation was the log of fss, the ratio of steady state concentration of infectious virus with and without UVC, averaged over the replicate experiments. The ACH due to UVC was then computed as λU = λo(1−f)/fss where λo is the ACH due to ventilation (See Appendix S1 for derivation). Regression analyses and confidence limits for regression coefficients were computed using R statistical software (R-Project, Version 2.6.0) and summarized in Excel (Microsoft Corp, Redmond, WA).Supporting InformationAppendix S1Derivation of Equivalent Air Exchange Rate Due to UVC. Derivation of equation used in data analysis.(0.03 MB DOC)Click here for additional data file.\n\nREFERENCES:\n1. WehrlePFPoschJRichterKHHendersonDA\n1970\nAn airborne outbreak of smallpox in a German hospital and its significance with respect to other recent outbreaks in Europe.\nBull World Health Organ\n43\n669\n79\n5313258\n2. HendersonDAInglesbyTVBartlettJGAscherMSEitzenE\n1999\nSmallpox as a biological weapon: medical and public health management. Working Group on Civilian Biodefense.\nJama\n281\n2127\n37\n10367824\n3. BricknerPWVincentRLFirstMNardellEMurrayM\n2003\nThe application of ultraviolet germicidal irradiation to control transmission of airborne disease: bioterrorism countermeasure.\nPublic Health Rep\n118\n99\n114\n12690064\n4. XuPPecciaJFabianPMartynyJWFennellyKP\n2003\nEfficacy of ultraviolet germicidal irradiation of upper-room air in inactivating airborne bacterial spores and mycobacteria in full-scale studies.\nAtmospheric Env\n37\n405\n419\n5. XuPKujundzicEPecciaJSchaferMPMossG\n2005\nImpact of environmental factors on efficacy of upper-room air ultraviolet germicidal irradiation for inactivating airborne mycobacteria.\nEnviron Sci Technol\n39\n9656\n64\n16475348\n6. FirstMNardellEChaissonWRileyR\n1999\nGuidelines for the Application of Upper-Room Ultraviolet Germicidal Irradiation for Preventing Tranmission of Airborne Contagion-Part II: Design and Operation Guidance.\nASHRAE Transactions\n105\n869\n876\n7. FirstMNardellEChaissonWRileyR\n1999\nGuidelines for the application of upper-room ultraviolet germicidal irradiation for preventing the transmission of airborne contagion-part I: basic principles.\nASHRAE Transactions\n105\n877\n887\n8. McDevittJJLaiKMRudnickSNHousemanEAFirstMW\n2007\nCharacterization of UVC light sensitivity of vaccinia virus.\nAppl Environ Microbiol\n73\n5760\n6\n17644645\n9. FirstMRudnickSNBanahanKFVincentRLBricknerPW\n2007\nFundamental factors affecting upper-room ultraviolet germicidal irradiation - part I. Experimental.\nJ Occup Environ Hyg\n4\n321\n31\n17365506\n10. NicasMMillerSL\n1999\nA multi-zone model evaluation of the efficacy of upper-room air ultraviolet germicidal irradiation.\nAppl Occup Environ Hyg\n14\n317\n28\n10446484\n11. RileyRLKnightMMiddlebrookG\n1976\nUltraviolet susceptibility of BCG and virulent tubercle bacilli.\nAm Rev Respir Dis\n113\n413\n8\n817628\n12. SiegelJDRhinehartEJacksonMChiarelloL\n2007\n2007 Guideline for Isolation Precautions: Preventing Transmission of Infectious Agents in Health Care Settings.\nAm J Infect Control\n35\nS65\n164\n18068815\n13. SattarASIjzazMK\n1987\nSpread of Viral Infections by Aerosols.\nCRC Critical Reviews in Environmental Control\n17\n88\n130\n14. HarperGJ\n1961\nAirborne micro-organisms: survival tests with four viruses.\nJ Hyg (Lond)\n59\n479\n86\n13904777\n15. RudnickSNMiltonDK\n2003\nRisk of indoor airborne infection transmission estimated from carbon dioxide concentration.\nIndoor Air\n13\n237\n45\n12950586\n16. RudnickSNFirstMW\n2007\nFundamental factors affecting upper-room ultraviolet germicidal irradiation - part II. Predicting effectiveness.\nJ Occup Environ Hyg\n4\n352\n62\n17454503\n17. ACGIH\n2008\nLight and Near-Infrared Radiation.\neditor ^editors\nThreshold Limit Values of Chemical Substances and Physical Agents & Biological Exposure Indices. ed\nCincinnati, OH\nACGIH\n144\n155\n18. FirstMWWekerRAYasuiSNardellEA\n2005\nMonitoring human exposures to upper-room germicidal ultraviolet irradiation.\nJ Occup Environ Hyg\n2\n285\n92\n15848970\n19. RileyRLKaufmanJE\n1972\nEffect of relative humidity on the inactivation of airborne Serratia marcescens by ultraviolet radiation.\nAppl Microbiol\n23\n1113\n20\n4557562\n20. KoGFirstMWBurgeHA\n2000\nInfluence of relative humidity on particle size and UV sensitivity of Serratia marcescens and Mycobacterium bovis BCG aerosols.\nTuber Lung Dis\n80\n217\n28\n11052911"
+}

batch_3/PMC2527533.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527533",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527533\nAUTHORS: Olena Morozova, Vyacheslav Morozov, Brad G. Hoffman, Cheryl D. Helgason, Marco A. Marra\n\nABSTRACT:\nBackgroundSerial Analysis of Gene Expression (SAGE) is a DNA sequencing-based method for large-scale gene expression profiling that provides an alternative to microarray analysis. Most analyses of SAGE data aimed at identifying co-expressed genes have been accomplished using various versions of clustering approaches that often result in a number of false positives.Principal FindingsHere we explore the use of seriation, a statistical approach for ordering sets of objects based on their similarity, for large-scale expression pattern discovery in SAGE data. For this specific task we implement a seriation heuristic we term ‘progressive construction of contigs’ that constructs local chains of related elements by sequentially rearranging margins of the correlation matrix. We apply the heuristic to the analysis of simulated and experimental SAGE data and compare our results to those obtained with a clustering algorithm developed specifically for SAGE data. We show using simulations that the performance of seriation compares favorably to that of the clustering algorithm on noisy SAGE data.ConclusionsWe explore the use of a seriation approach for visualization-based pattern discovery in SAGE data. Using both simulations and experimental data, we demonstrate that seriation is able to identify groups of co-expressed genes more accurately than a clustering algorithm developed specifically for SAGE data. Our results suggest that seriation is a useful method for the analysis of gene expression data whose applicability should be further pursued.\n\nBODY:\nIntroductionWith the advent of high throughput technologies, large-scale gene expression studies have become routine in many biological laboratories. Two conceptually different approaches to high throughput gene expression profiling are microarrays [1] and tag sequencing-based methods, such as Serial Analysis of Gene Expression (SAGE) [2]. While both of these gene expression platforms can generate large genome-wide expression data sets, making full use of the data is still an important bioinformatic challenge [3]. A common aim of high throughput gene expression studies is to identify genes with similar expression profiles since such genes may be functionally related and thus may be used to predict functions of unknown genes. This aim has been most often addressed by various versions of clustering analysis that group genes into clusters with correlations among their expression values [4], [5]. Currently available clustering methods show variable success at identifying functionally-relevant gene groupings [6]–[8].While microarray studies assess gene expression levels by measuring hybridization intensities to the relevant probes [1], SAGE studies use portions of cDNA transcripts known as SAGE tags that are concatenated, cloned, and sequenced to provide a quantitative measure of the transcripts levels in the cell [2]. The use of SAGE had been until recently limited by the sequencing cost and laborious steps inherent in the cloning procedure. However, with modern advances in sequencing technologies, SAGE-related methods have become more cost-effective and are gaining popularity owing to some technological advantages they offer over microarrays [9]. In particular, SAGE does not rely on previous knowledge of gene structure. In addition, it has been suggested that SAGE studies are more robust, and require fewer replicates than microarray studies [9], [10]. Generally, SAGE data have been subjected to the same clustering methods as microarray data [11]. However, more appropriate distance measures accounting for the discreet, Poisson-distributed structure of SAGE data have been shown to produce better clustering results than those achieved with conventional Euclidian or Pearson similarity measures routinely used in microarray data clustering [12]. A successful clustering method for SAGE termed PoissonC accounts for the categorical structure of SAGE data by using the Chi-square statistic and the Poisson distribution in a K-means clustering procedure [12].A notorious feature of gene expression datasets affecting the performance of cluster analysis is high data dimensionality whereby the expression of many genes is assayed over a small number of experimental conditions (time points). This leads to failure of common statistical methods to distinguish real correlation patterns from spurious ones [3], [5], and necessitates the development of alternative approaches for identifying co-expressed genes. Here we explore whether a reordering rather than grouping approach can be used for the identification of co-expressed genes in gene expression data, and whether such an approach would yield fewer false positives than achieved by grouping genes into sets with clustering. Seriation is a statistical method for simultaneously ordering rows and columns of a symmetrical distance matrix for the purposes of revealing an underlying one-dimensional structure [13]. An assumption in seriation analysis is that there is an order (or distinct sub orders) in the data that are biologically meaningful. The inherent orders may represent any sequential structure among the data (e.g. their dependence on time or another variable). Seriation in its different flavors has been successfully applied in multiple fields, including archeology, psychology, and operational research; for instance, in archaeology it has been used to uncover the chronological order of archaeological deposits [14], [15]. The application of seriation for the analysis of high throughput biological data has been limited. One application in gene expression analysis is finding an optimal leaf ordering of a hierarchical clusterogram [16], [17]. In these studies global seriation was conducted after hierarchical clustering to aid in finding an optimal solution. In contrast, the present study examines whether the detection of local ordered structures in the data can be used in place of clustering for the identification of co-expressed genes.Since finding the exact solution to seriation is known to be a nondeterministic polynomial time (NP)-hard problem [18], several heuristics have been developed to achieve an acceptable ordering solution [17]. We developed an original seriation heuristic we term ‘progressive construction of contigs’, which is based on step-wise reordering of the correlation matrix to produce chains or ‘contigs’ of related correlation values. We applied the seriation heuristic to analyses of both simulated and experimental SAGE data and showed that our approach can be used to effectively identify groups of co-expressed genes, and the relationships among these groups in a robust manner. We found that seriation performed better than a SAGE-specific clustering method on SAGE data containing spurious expression patterns that would arise due to measurement uncertainties and small number of experimental conditions compared to the large number of genes [3], [5]. Global patterns in the data revealed by seriation are easily detectable by eye from the reordered correlation matrix and can be interpreted biologically.ResultsSeriation using the progressive construction of contigs heuristicMotivated by the opportunity to improve upon current methods for analyzing large scale expression datasets, we set out to explore the use of seriation as a substitute for clustering for identifying co-expression patterns in SAGE data. Seriation seeks the best enumeration order among objects based on their similarity according to a chosen criterion. Since the problem is NP-hard, we developed a novel heuristic specifically for the SAGE data analysis task. The ‘progressive construction of contigs’ heuristic attempts to put the most similar objects side by side without breaking already established chains of closely related elements we term ‘contigs’. Here we use pairwise correlations between expression vectors (normalized tag counts for a particular tag across all libraries) as the criterion for defining similarities between tags; however, in principle, other similarity criteria can be used for this task. The pairwise correlations between tag expression vectors x and y are calculated using the standard correlation coefficient function, \nx̅ and y̅ are the means of expression vectors x and y, and E is the mathematical expectation. The correlation values are subsequently arrayed into a symmetric matrix, which is subjected to the following progressive seriation procedure.In the first step, the tag pair with the highest correlation value is found and marked as the beginning of the first contig. At each subsequent step the tag pair with the next highest correlation value is identified. If one of the members of the tag pair is involved in a previously formed contig, the columns of the matrix are reorganized to place the other member at the nearest edge of the same contig; since the matrix is symmetrical, the rows are reordered accordingly. Importantly, previously reordered elements are kept intact in this process. If it is impossible to add the similarity maximum of the current step to a contig given the restriction on the previously-moved objects or if the tag pair with the correlation maximum does not involve any of the members of the formed contigs, the current similarity maximum is used to start a new contig. The seriation process continues until all elements have been processed. The result is the production of contigs of similar correlation values that can be displayed along the diagonal of the correlation matrix representing internal topologies in the data. Theoretically, in the case of a Robinson data structure, whereby the data are from a unimodal distribution, the contigs are merged into one and the obtained result is the most optimal single seriation solution [14], [17].A key algorithmic difference between the seriation algorithm described above and a procedurally similar hierarchical clustering algorithm (such as the hierarchical clustering method developed in [19] and implemented in [4]) is the treatment of vectors after the highest pairwise correlation value has been identified at each step. In clustering, the vectors are averaged together into a new vector using a linkage rule (for instance, average linkage clustering) and this new vector is represented by a node in the hierarchical clusterogram. In contrast, in the case of seriation, no new vector or node is formed, and the rows and columns of the correlation matrix are merely reordered to reflect underlying patterns in the data as described above. Therefore, no linkage rule is required in seriation in addition to the distance metric used to define similarities.In the current implementation of the seriation algorithm, ordered structures (contigs) are revealed by color-coding the reordered correlation matrix according to the magnitude of the correlation value. In this manner, visual inspection of the matrix allows for the selection of ordered contigs for further inspection. Due to the visualization component, the algorithm is able to analyze up to 4000 genes at a time (tested on 1.7 IBM PC Pentium 4, Z60t laptop) and is suitable for the analysis of pre-selected sets of genes. Importantly, the algorithm produces a robust solution for each seriation run (in other words, equivalent solution is produced upon repeated seriation of the same data set).Performance of seriation on simulated SAGE dataTo test the performance of the seriation heuristic we generated a simulation dataset containing 500 expression vectors of dimension 5 (corresponding to 500 SAGE tags expressed over 5 different time points or conditions). Since expression data for a gene collected under different experimental conditions or at different time points are not completely independent, distinguishing genes with similar expression profiles in which the dynamics of gene expression changes is considered is of biological interest [5]. We designed the expression vectors to represent 10 different expression profiles that might be of potential biological interest (Figure S1).To test the dependence of algorithm performance on the amount of noise in the data, we initially seriated three of these expression profiles with increasing numbers of noise tags. Pattern 2 corresponds to tags whose expression slightly peaks at time point 2 and then at time point 5; pattern 3 includes tags with a single expression peak at time point 2; and pattern 1 corresponds to tags with an expression peak over time points 3 and 4 (Figure S1). To closely simulate actual SAGE data, we added ‘noise’ or singleton tags whose expression profiles do not conform to any of the three patterns. Such expression profiles are common in gene expression datasets, particularly ones with few experimental conditions sampled relative to the number of genes [5]. Since it has been previously shown that SAGE data can be approximated by a Poisson distribution [12], we used Poisson-based rules for our simulations (see methods). Genes with similar expression profiles were modeled by a Poisson distribution with the same λ [12]. In contrast, genes that do not belong to any of the three patterns of interest (i.e. noise) were simulated by constructing expression profiles based on a Poisson distribution with random λ, obtained from a uniform distribution [1, 300]. We tested the performance of seriation as well as the PoissonC clustering algorithm, a successful K-means clustering algorithm previously developed specifically for SAGE data [12] on the simulation data set in three rounds, each time increasing the amount of noise present among the profiles of interest (Table S1). In each round, seriation yielded three clear contigs along the diagonal corresponding to the three patterns of interest (Figure 1A). Importantly, increasing the amount of tags corresponding to noise from 34 (round 1) to 384 (round 3) did not significantly affect the performance of the seriation algorithm (Table 1). We also applied the PoissonC algorithm to the simulation data using the same design. The optimal value of K was determined as described [20] and was set to K = 4 for each round. Values of K = 5 and K = 6 were also tested for round 2 and round 3 simulations, but did not produce significantly different results from those generated with K = 4. Interestingly, the performance of PoissonC declined with increasing amounts of noise (Table 1) illustrating the common problem with clustering analysis of gene expression data sets [3].10.1371/journal.pone.0003205.g001Figure 1Performance of seriation on simulated SAGE data.(A). Seriation results of the three rounds of simulations with increasing amounts of noise from round 1 (34 tags) to round 3 (384 tags). The dark red squares along the diagonal indicate tags with the expression patterns 1–3 that were grouped together by seriation. (B). Seriation of 10 expression profiles with limited amount of noise. The dark red squares along the diagonal indicate tags in each expression profile that were grouped together. The numbers indicate expression patterns from Figure S1 that were grouped into each contig. Note that two contigs in the middle (5 and 1) appear more similar to each other than any other contig pair indicating similarity of the corresponding expression patterns.10.1371/journal.pone.0003205.t001Table 1Effect of the amount of noise in SAGE data on the performance of seriation and PoissonC.Pattern1Pattern2Pattern3TPFPTPFPTPFPRound 1: 34 noise tagsSeriation411384374PoissonC412382375Round 2: 120 noise tagsSeriation416386373PoissonC4113*\n3814*\n3715*\nRound 3: 384 noise tagsSeriation415383373PoissonC4143*\n3899*\n3761*\nSeriation and PoissonC were applied to a simulated SAGE data set containing three expression patterns and increasing amount of noise tags. The dataset is described in more detail in the text and in Table S1. TP (True Positives) include tags that were correctly classified as belonging to the correct expression group (expression pattern 1, 2, or 3 or noise) by assigning them to the cluster (PoissonC) or contig (seriation) containing other members of the expression group. FP (False Positives) include noise tags that have been erroneously assigned to a cluster or contig with tags that conform to the expression pattern 1, 2, or 3.*The false positive rate is significantly higher for the PoissonC algorithm than it is for seriation mostly due to the erroneous assignment of noise tags to an expression pattern (p<0.05).Overall, it can be noted that both algorithms performed well on data with relatively little noise (round 1); however, as the amount of noise in the data increased, seriation appeared more robust than clustering at identifying correct expression groupings. Importantly, both algorithms correctly grouped tags with similar expression profiles together in all three rounds (true positives), and the reduction in performance of PoissonC was due to an increase in false positives being incorporated into co-expression clusters.Having established the excellent performance of seriation on noisy SAGE data containing a few expression profiles of interest, we went on to evaluate the dependence of performance on the number of expression profiles to be identified in the analysis. For this experiment, we used 10 expression profiles (Figure S1) and conducted the analysis as described above. We seriated 50 tags corresponding to each of the 10 expression profiles and 50 tags corresponding to noise. The resulting color-coded correlation matrix is shown in Figure 1B, and the comparative performance of PoissonC and seriation on the data is summarized in Table 2. The 10 expression profiles were grouped into 10 contigs along the diagonal by seriation. In addition to reordering tags according to the correct expression profile, seriation analysis was able to detect similarities among the profiles themselves. For instance, two squares in the middle of the matrix are distinct yet appear more similar to each other than any other pair of consecutive contigs. These contigs correspond to profiles 5 and 1 which are indeed very similar (Figure S1). Such additional information can not be revealed by clustering with PoissonC.10.1371/journal.pone.0003205.t002Table 2Comparative performance of seriation and PoissonC on a simulated SAGE data set with 10 expression patterns.AlgorithmTPFPSeriation5491PoissonC52822*\nSeriation and PoissonC were applied to the analysis of a simulated SAGE data set containing 10 expression patterns each including 50 tags, and 50 noise tags. TP (True Positives) are tags that were correctly classified as belonging to the right expression pattern or noise. FP (False Positives) are tags that were assigned to the wrong pattern or noise tags that were assigned to an expression pattern.*The false positive rate is significantly higher for the PoissonC algorithm than it is for seriation (p<0.05).Performance of seriation on previously-published experimental SAGE dataTo test the performance of seriation on previously analyzed experimental SAGE data, we applied the algorithm to reorder genes expressed in mouse retinal SAGE libraries based on similarity of their expression profiles [20]. The SAGE data were generated from mouse retinal tissues at 10 different developmental stages ranging from E12.5 (Theiler stage 20) to post natal day 10 (P10) and adult; the data were originally analyzed using the PoissonC algorithm with K = 24. We subjected the same dataset to seriation using the progressive construction of contigs procedure. The algorithm produced 10 contigs, including 2 contig groupings called ‘supercontigs’ (Figure 2). Most of the contigs were composed of members of one or several clusters from Blackshaw et al. [20] (Figure 3A; Table 3). The expression profiles of genes in the 24 clusters generated by Blackshaw et al. [20] are provided in Figure S2. It can be noted that as a result of seriation, clusters with similar expression patters were grouped together into contigs. For instance, clusters 8, 22, and 24, which contain genes whose expression peaks at post natal day 10, were grouped into contig8. Similarly, contig9 included clusters 1, 10, 22, and 24, which contain genes that are highly expressed in the adult library (Figure 3). For a full list of contig memberships of genes expressed in the retinal libraries see Figure S3.10.1371/journal.pone.0003205.g002Figure 2Seriation of genes expressed in mouse retinal SAGE libraries.SAGE data from Blackshaw et al. [20] were subjected to seriation analysis as described in the text. The resulting reordered correlation matrix containing correlation coefficients for each tag pair computed to measure the similarity of their retinal expression profiles is color-coded red to blue to represent decreasing correlation values. Ten contigs, including two supercontigs, recognizable as the squares of high (red) correlation values along the diagonal, are evident from the color-coded correlation matrix. The Figure on the right provides a zoomed-in view of the contigs.10.1371/journal.pone.0003205.g003Figure 3Analysis of seriation contigs of genes expressed in mouse retinal SAGE libraries.(A). Comparison of seriation contigs to the original clusters from Blackshaw et al. [20]. Seriation contigs are color-coded and plotted on the x-axis of the 3D graph. The peaks on the z-axis represent the percent cluster members (y-axis) present in the particular contig. Most seriation contigs are composed of one or several predominant clusters (also see Table 3). (B). Expression profiles of genes in seriation contigs. The relative expression levels from 0% to 100% are plotted on the y-axis for each contig while the retinal libraries derived from developmental stages E12.5, E14.5, E16.5, E18.5, P0.5, P2.5, P4.5, P6.5, P10, and adult are on the x-axis. The ordering of contigs is temporal such that genes expressed in earlier developmental stages tend to be in the first contigs, while genes expressed in later stages are in later contigs. This partitioning is particularly evident from the expression patterns of genes in the supercontigs.10.1371/journal.pone.0003205.t003Table 3Comparison of seriation and PoissonC analyses of genes expressed in retinal SAGE libraries.ContigPredominant clusters*\nPercent of predominant cluster members in contigPercent of contig members in predominant clusterTop Gene Ontology annotations enriched in predominant clusters*\nTop Gene Ontology annotations enriched in contigsContig1432.59%58.4%MitochondrialRibonucleoprotein complex, p = 8.68E-03Contig2579.77%74.19%RibosomalCytosolic ribosome (sensu Eukarya), p = 2.52E-09Contig31296.36%42.74%NoneBiosynthesis, p = 1.11E-02Supercontig 1 (contigs 1, 2, 3)4, 5, 1233.48%, 93.64%, 100%17.24%, 37.24%, 12.64%Mitochondrial, RibosomalRibonucleoprotein complex, p = 7.17E-14Contig4631.94%32.86%RNA processingLigase activity, p = 3.28E-02Contig5633.33%68.57%RNA processingN/AContig6N/AN/AN/AN/AStructural molecule activity, p = 3.29E-05Contig71541.07%67.65%RibosomalCytosolic ribosome (sensu Eukarya), p = 4.89E-07Contig88, 22, 24100%, 37.5%, 51.56%15.09%, 2.8%, 46.7%Ribosomal, Vision, VisionMetal ion binding, p = 2.97E-03Contig91, 10, 21, 22100%, 32.2%, 92.86%, 62.5%4.27%, 40.28%, 43.13%, 4.73%Vision, Transporter activity, Vision, VisionVision, p = 4.74E-15Supercontig 2 (contigs 8, 9)1, 8, 10, 21, 22, 24100%, 100%, 41.67%, 100%, 100%, 56.25%2.13%, 7.57%, 26%, 23.17%, 3.78%, 25.53%Vision, Ribosomal, Transporter activity, Vision, VisionPerception of light, p = 7.77E-14Contig102100%15%Lens proteinsStructural constituent of eye lens, p = 1.17E-19Column 2 contains Blackshaw et al. [20] clusters that are predominant in seriation contigs in column 1. GO categories enriched in clusters/contigs were determined using EASE software [34]; p<0.05.*Data from Blackshaw et al. [20].Significantly, Gene Ontology (GO) enrichment analysis indicated that clusters grouped into the same contig contained the same or similar enriched GO categories (Table 3) suggesting that they were somewhat functionally redundant. As an example, all but one of the clusters that fell into contig9 had ‘vision’ as the top enriched GO category. In addition to ordering co-expressed genes to form a contig, seriation provides insight into the relationship among the contigs. Here, it can be noted that the left-most contigs tend to contain genes whose expression peaks at earlier developmental stages, whereas the right-most contigs contain genes whose expression peaks in the late postnatal or adult lilbraries (Figure 3B). This ordering of contigs is temporal and biologically relevant, since the enriched GO categories of neighboring contigs are related. For instance, contigs 1, 2 and 3 that form supercontig1 are all enriched in genes that have a ribosomal function. Similarly, contigs 8 and 9 (supercontig2) are highly enriched in genes that function in vision (Table 3). Therefore, we argue that seriation provides an overview of the global biologically-relevant patterns in the data. Here, the results indicate that the retinal tissues contain two highly represented functional groups of genes, those involved in the ribosome functionality and those related to vision and light perception. These groupings are easily recognized from the color-coded seriated correlation matrix (Figure 2) as the two supercontigs. While it is possible to extract similar information from the clustering results, seriation provides a means to organize it in a relevant easily-interpretable and visualizable manner.As evident from the simulation study, seriation is more discriminative than clustering analysis at grouping co-expressed genes together resulting in more accurate results. On the other hand, clustering analysis forces all the tags to belong to a cluster thereby resulting in more false positives. Here, many genes in the seriation experiment were not captured in the contigs (Figure 2) as they are presumably not sufficiently similar to any of the patterns present in the contigs. It can be noted that all the GO categories that were found to be enriched in Blackshaw et al. [20] clusters were also present in the contigs (Table 3) suggesting that Blackshaw et al. [20] clusters were somewhat redundant and may contain false positives.Performance of seriation on novel experimental SAGE dataWe next applied the seriation algorithm to the analysis of SAGE libraries we generated as part of the Mouse Atlas Project (www.mouseatlas.org). The Mouse Atlas Project aims to produce a collection of SAGE libraries derived from various mouse tissues representing different developmental stages, ranging from embryonic stem cells to post-natal day 84 [21]; currently the resource contains over 200 different libraries. Due to our interest in the transcriptional regulation of pancreatic development we focused on analyzing the expression of transcription factors expressed in six SAGE libraries representing various stages of pancreatic endocrine cell development ranging from Theiler stage 17 (TS17) to post-natal day 70 (P70). Transcription factors are regulatory proteins that are presumed to be responsible for the coordinated expression of functionally-related genes. Transcription factors are at the top of the regulatory hierarchies that drive pancreatic development and enable beta cell maturation [22]. Thus, global analysis of transcription factor expression may provide insight into the mechanisms of pancreatic development and the misregulation of the mechanisms in disease.SAGE expression profiles of 319 transcription factors expressed in six pancreatic libraries were subjected to seriation analysis. The algorithm yielded five contigs of transcription factor SAGE tags with similar expression profiles (Figure 4). For this analysis, we chose contigs as groupings of co-expressed genes (red squares along the diagonal, Figure 4A) with at least 10 members. Contigs of transcription factors expressed in the pancreatic libraries are provided in Figure S4. Annotation analyses of the resulting contigs suggested that they were functionally relevant based on the enrichment for GO category, SwissProt keyword and Kyoto Encyclopedia of Genes and Genomes (KEGG) annotations (Table 4, Figure S5). It was also evident that the contigs contained transcription factors that were expected to be grouped together based on their known membership in the same pathway. For instance, transcription factors implicated in islet cell type specification as part of FoxO signaling, Neurod1 and Foxa2 [23] were grouped together into contig1. In addition, Pax6 which was recently shown to be a target of Neurod1 [24] was placed in the same contig. Similarly, Neurogenin-Neurod cascade members implicated in endocrine development [25] together with a downstream transcription factor Nkx2-2 [26] were grouped into contig2. Smad3 and Smad4, known TGF-beta targets [23] were placed into contig4, which was enriched for TGF-beta signaling pathway annotation. These results suggest that transcription factor groupings produced by seriation are biologically relevant and recapitulate the transcriptional circuitry involved in the control of pancreatic development.10.1371/journal.pone.0003205.g004Figure 4Seriation of transcription factors expressed in Mouse Atlas pancreatic libraries.SAGE data for transcription factors expressed in the pancreatic libraries from the Mouse Atlas project were subjected to seriation analysis as described in the text. The reordered correlation matrix containing correlation coefficients for each tag pair computed to measure the similarity of their pancreatic expression profiles is color-coded red to blue to represent decreasing correlation values. (A). 5 contigs recognizable as red squares along the diagonal are evident. (B). Expression profiles of transcription factors in contigs in (A). The relative expression levels from 0% to 100% are plotted on the y-axis for each contig while the pancreatic libraries derived from stages TS17, TS19, TS20, TS21, TS22, and P70 are on the x-axis.10.1371/journal.pone.0003205.t004Table 4Summary of seriation analysis of transcription factors expressed in pancreas.ContigTags in contigCharacterized transcription factors in contigGene Ontology annotations enriched in contig*\nSwissProt keywords enriched in contig*\nKEGG annotations enriched in contig*\nContig128\nFoxa2, Neurod1, Pax6, Myst1, Ets2, Mxd4, MntAnatomical structure development, p = 1.52E-03; System development, p = 2.98E-03; Organ development, p = 5.91E-03N/AN/AContig255Kin, Pole4, Foxa3, Tox3, Neurod2, Neurog3, Fos, Fev, Yy1, Jun, Nkx2-2\nDefense response, p = 2.21E-02; Receptor activity, p = 2.13E-02N/AN/AContig361\nHoxb7, Dr1, Foxm1, Hmga1, Klf6, Snai1, Hoxb5, Sox18, Hoxb1, Hoxb6, Hoxa10, Hand1, Pax1, Hoxa5, Hoxa7\nMulticellular organismal development, p = 1.45E-06; pattern specification process, p = 4.64E-05; regulation of cell differentiation, p = 9.97E-03Zinc-finger, p = 1.46E-03; Homeobox, p = 2.19E-03Adherens junction, p = 4.25E-02Contig452Hmga1, Foxp1, Mum1, Lin28, Msh6, Dnase2a, Mxd3, Rest, Gata4, Klf4, Cdx2, Smad4, Smad3\nTransmembrane receptor protein serine/threonine kinase signaling pathway, p = 2.96E-02; regulation of signal transduction, p = 2.12E-02; negative regulation of cellular process, p = 2.14E-02N/ATGF-beta signaling pathway, p = 7.66E-04; Wnt signaling pathway, p = 3.65E-02Contig597\nPax4, Dpm1, Sox9, Stat2, Arid2, Terf1, Dpm1, Elf3, Mtf1, Lass2, Arid3a, Stat4Apoptosis, p = 1.23E-03; Programmed cell death, p = 1.43E-03; Regulation of apoptosis, p = 2.19E-03Apoptosis, p = 6.08E-03; Coiled coil, p = 6.28E-03N/ANumber of tags falling into each seriation contig is shown along with the representative contig members and their representative functional annotation using GO categories, SwissProt keywords, and KEGG pathways. For a full list of annotations enriched in the contigs see Figure S5. Known regulators of pancreatic development as well as the transcription factors discussed in the text are bolded.*GO, SwissProt keyword, and KEGG pathway annotations enriched in contigs were computed using web-based FatiGO+ tool [35] and are provided with raw scores.Previous studies have shown that genes with similar expression profiles are functionally related; moreover, co-expression has been reasonably successfully used to predict function of unknown genes [8], [27]. Therefore, the identified contigs of transcription factors can be used to gain insight into the functionality of unknown transcription factors in pancreatic development. For instance, Hand1, a major regulator of heart development which has been also implicated in vascular development [28], was placed into contig3 together with other basic helix-loop-helix (bHLH) transcription factors, Hox and Pax genes. Homeobox (Hox) and paired box (Pax) transcription factors have been presumed to function together to regulate a variety of developmental processes [29]. Our seriation analysis suggested that these transcription factors together with another bHLH family member Hand1 may function together during pancreatic development, a possibility that can be tested experimentally in the future. In addition, a crucial regulator of pancreatic development, Pax4, was placed into contig5 together with members of the Stat family of transcription factors, suggesting a potential interaction between Pax4 and the JAK/STAT pathway.DiscussionClustering analysis has been the approach of choice for most gene expression studies. However, due to high dimensionality of gene expression datasets, many clustering algorithms are prone to producing false positive expression-based interactions [3]. SAGE data have been particularly poorly exploited by statistical analyses owing to the domination of the gene expression field by microarrays that produce continuous data as opposed to discreet count-derived data produced by SAGE. With the advent of next-generation sequencing technologies, sequence tag-based methods have been gaining popularity for gene expression analysis thereby necessitating the development of statistical methods for analyzing discreet expression data. To date, a few clustering algorithms designed to exploit the digital data structure have been developed for SAGE data analysis, and shown to perform favorably compared to conventional microarray clustering algorithms [12]. However, these methods are still subject to the inherent limitations of the clustering approach itself.We explored the use of local seriation for the identification of co-expression patterns in SAGE data. The primary goal of seriation methods is finding an optimal ordering of a set of objects based on a similarity criterion. Since there are n! ways to order a set of n objects, finding the most optimal seriation order becomes computationally expensive with the increasing size of the data set; therefore, heuristics have been developed to achieve an optimal ordering solution [17]. We developed a novel bottom-up heuristic we termed ‘progressive construction of contigs’ specifically designed for seriation of gene expression vectors according to their similarity. The ‘progressive construction of contigs’ heuristic is based on a greedy process that does not question the previous steps, and thus is fast and can, in principle, be implemented with large datasets. We tested the performance of seriation on both simulated and experimental SAGE data, and compared its performance with that of the PoissonC K-means clustering algorithm, a current state-of-the-art method in the field of SAGE data analysis [12]. We demonstrated that seriation was able to identify contigs of co-expressed genes that were related to clusters of co-expressed genes obtained by PoissonC (Table 3). We showed that the co-expression contigs were enriched for genes with similar functions as defined by both Gene Ontology and SwissProt keyword annotations as well as the known memberships in the same pathway. Therefore, we provided an empirical demonstration that the results from the two approaches are related and are complementary to each other. We further showed that in contrast to clustering, seriation could detect relationships among contigs of co-expressed genes, such as their temporal order, whenever such relationships were present in the data. Moreover, based on the simulation results, seriation appeared less sensitive to noisy data than PoissonC, and produced fewer false positives.The major conceptual difference between seriation and clustering underlying the differential performance of the methods on noisy SAGE data stems from the different primary goals of the two methods. The primary goal of seriation is reordering during which inherent patterns in the dataset (e.g., presence of groups of elements that are related to one another) are revealed. On the other hand, the primary goal of clustering is partitioning the dataset into groups of similar elements. A key advantage of ordering over grouping is that ordering allows for the discovery of gradual progressions in the data while such gradual information is lost in grouping analyses. For instance, Robinson properties in the data can be revealed by seriation but not by clustering [14]. Gene expression changes over various experimental conditions are often of a gradual nature rendering seriation a useful tool for the discovery of similar expression profiles. In other words, the identification of groups of related elements is a consequence of seriation while it is the primary goal of clustering. Due to this fact, following clustering analysis of gene expression datasets, all genes are assigned to the most appropriate cluster based on a generic linkage rule. In contrast, following seriation analysis that does not require a linkage rule, contigs of genes with high pairwise correlation coefficients are revealed by reordering. Real versus spurious co-expression interactions can be thus gauged from the color-coded reordered correlation matrix (e.g., Figure 1, Figure 2, Figure 4A) wherein clear tightly-formed red squares along the diagonal reveal groupings of co-expressed genes while the rest of the matrix represents tags that do not belong to any of the contigs.Overall, we showed that seriation is a useful tool for pattern discovery and visualization in SAGE datasets. The method allows one to estimate the number of co-expression patterns present in the dataset (estimated from the number of formed contigs) as well as the amount of ‘noise’ or spurious expression profiles (estimated from the number of tags that do not appear to belong to a contig). We showed that seriation correctly detected groups of simulated expression profiles, correctly identified enriched GO categories obtained by PoissonC, and correctly revealed a number of known transcription factor interactions from pancreas SAGE data. Therefore, we suggest that the application of seriation for the identification of co-expressed genes in tag-based gene expression studies should be explored further.Materials and MethodsSimulation studyThe simulation design was influenced by a previously described design for short-term time series microarray expression data [5]. However, we used the Poisson distribution that has been shown to be suitable for modeling SAGE data [12] instead of the uniform distribution used in [5] to model microarray data. In brief, we generated a simulation dataset containing 500 expression vectors of dimension 5. The expression profiles for each tag (v0, v1, v2, v3, v4) belonging to one of 10 expression patterns defined by setting (m0, m1, m2, m3, m4) to (0, 1, 0, 0, 2) for pattern 1; (0, -1, 1, 1, -1) for pattern 2; (0, 2, 1, 0, 0) for pattern 3; (0, 1, -1, -1, 1) for pattern 4; (0, -1, 0, 0, -2) for pattern 5; (0, -2, -1, 0, 0) for pattern 6; (0, 1, -1, 1, -1) for pattern 7; (-1, 1, 1, -1, 0) for pattern 8; (2, 0, 0, 1, 0) for pattern 9; and (0, 0, 1, 2, 0) for pattern10 in the equation below, were determined as follows:where Z takes on the values of -1, 0 or 1 with equal probability of 1/3 and represents errors in tag count measurements. λ was kept the same for all genes in the same expression pattern group as suggested in [12]. The resulting expression profiles of the 10 groups are displayed in Figure S1. The noise was modeled using the same rule as above but selecting a random λ ∼ Uniform [1, 300] to model the expression profile of each tag. Simulations were conducted over three rounds with increasing amount of noise (Table S1). Seriation algorithm was run several times for each round and produced the same result. PoissonC algorithm was run over 100 iterations and the iteration results were combined into consensus clusters.Experimental SAGE dataThe retinal dataset was obtained from Blackshaw et al. [20] and was processed as described by the authors. The mouse pancreatic SAGE libraries SM161/SM244, SM231, SM243/SM160, SM225/SM249, SM232 and SM017 were obtained from the Mouse Atlas web site (www.mouseatlas.org). The libraries were built as described in [21] and in [30]. All tag processing, including the removal of linker-derived tags, quality filtering (95% sequence quality cutoff was used) and mapping was done in DiscoverySpace 4.0 software as described [31]. The tag counts in each library were normalized to the depth of 100,000.Mouse transcription factorsWe obtained the list of mouse transcription factors by selecting Ensembl genes containing DNA-binding domains from Pfam [32], [33]. The sequences were selected based on the mouse genome NCBI build 32 and are analyzed in O. Morozova and T.R. Hughes. Patterns of transcription factor evolution in vertebrates. Proceedings of the Third Canadian Student Conference on Biomedical Computing (CSCBC), 2008. We found that out of 994 transcription factors expressed in the Mouse Atlas libraries, 319 were present in the pancreatic libraries with a tag count of 4 or higher.GO category, SwissProt keyword and KEGG pathway enrichment analysisGO category analysis of retinal SAGE clusters and contigs was performed using EASE software as described [34]. GO category, SwissProt keyword and KEGG pathway enrichment analysis of transcription factor contigs was performed using the web-based FatiGO+ tool [35]. P-values of less than 0.05 were considered statistically significant for both analyses.Clustering analysisK-means clustering analysis was performed according to the PoissonC algorithm [12]. The within-cluster dispersion was calculated as described [20]. The java implementation of the clustering algorithm and the within-cluster dispersion calculation was kindly provided by Li Cai (Rutgers University, NJ).Seriation algorithm and its implementationSeriation was conducted on simulated, retinal or Mouse Atlas SAGE data using the custom MATLAB implementation. The algorithm was run three times on each experimental dataset to ensure the seriation result was robust. The analysis of simulated SAGE data was done as described above. The implementation of the algorithm can be made available to interested academic users upon request.Supporting InformationFigure S1Composition of the simulation dataset during three rounds of simulations.(0.86 MB TIF)Click here for additional data file.Figure S2Expression profiles of genes in 24 clusters from Blackshaw et al. [20]. The relative expression levels from 0% to 100% are plotted on the y-axis for each cluster while the retinal libraries derived from developmental stages E12.5, E14.5, E16.5, E18.5, P0.5, P2.5, P4.5, P6.5, P10, and adult are on the x-axis.(2.79 MB TIF)Click here for additional data file.Figure S3Contig membership of genes expressed in retinal SAGE libraries.(0.12 MB XLS)Click here for additional data file.Figure S4Contig membership of transcription factors expressed in pancreas.(0.04 MB XLS)Click here for additional data file.Figure S5Annotations enriched in contigs of transcription factors expressed in pancreas.(0.03 MB XLS)Click here for additional data file.Table S1Composition of the simulation dataset during three rounds of simulations. Simulated SAGE datasets were constructed to include three different expression patterns of potential biological interest (depicted in Figure S1, patterns 1, 2, and 3) and modeled as described in Materials and Methods. To simulate actual SAGE data, we included singleton tags that do not strictly conform to any of the three expression patterns (referred to as ‘noise’). The simulation was conducted over three rounds with constant numbers of tags in each expression category (rows 1–3) and increasing numbers of noise tags (row 4). The expression profiles in each category are shown in column 5 and explained in Materials and Methods.(0.03 MB DOC)Click here for additional data file.\n\nREFERENCES:\n1. SchenaMShalonDDavisRWBrownPO\n1995\nQuantitative monitoring of gene expression patterns with a complementary DNA microarray.\nScience\n270(5235)\n467\n470\n7569999\n2. VelculescuVEZhangLVogelsteinBKinzlerKW\n1995\nSerial analysis of gene expression.\nScience\n270(5235)\n484\n487\n7570003\n3. ClarkeRRessomHWWangAXuanJLiuMC\n2008\nThe properties of high-dimensional data spaces: Implications for exploring gene and protein expression data.\nNat Rev Cancer\n8(1)\n37\n49\n18097463\n4. EisenMBSpellmanPTBrownPOBotsteinD\n1998\nCluster analysis and display of genome-wide expression patterns.\nProc Natl Acad Sci U S A\n95(25)\n14863\n14868\n9843981\n5. ErnstJNauGJBar-JosephZ\n2005\nClustering short time series gene expression data.\nBioinformatics\n21\nSuppl 1\ni159\n68\n15961453\n6. ThalamuthuAMukhopadhyayIZhengXTsengGC\n2006\nEvaluation and comparison of gene clustering methods in microarray analysis.\nBioinformatics\n22(19)\n2405\n2412\n16882653\n7. PrelićABleulerSZimmermannPWilleABuhlmannP\n2006\nA systematic comparison and evaluation of biclustering methods for gene expression data.\nBioinformatics\n22(9)\n1122\n1129\n16500941\n8. ZhangWMorrisQDChangRShaiOBakowskiMA\n2004\nThe functional landscape of mouse gene expression.\nJ Biol\n3(5)\n21\n15588312\n9. BrennerSJohnsonMBridghamJGoldaGLloydDH\n2000\nGene expression analysis by massively parallel signature sequencing (MPSS) on microbead arrays.\nNat Biotechnol\n18(6)\n630\n634\n10835600\n10. AudicSClaverieJM\n1997\nThe significance of digital gene expression profiles.\nGenome Res\n7(10)\n986\n995\n9331369\n11. DennisJLOienKA\n2005\nHunting the primary: Novel strategies for defining the origin of tumours.\nJ Pathol\n205(2)\n236\n247\n15641019\n12. CaiLHuangHBlackshawSLiuJSCepkoC\n2004\nClustering analysis of SAGE data using a poisson approach.\nGenome Biol\n5(7)\nR51\n15239836\n13. KendallDG\n1971\nSeriation from abundance matrices.\nHodsonFRKendallDGTautuP\nMathematics in the Archaeological and Historical Sciences\nEdinburgh\nEdinburgh University Press\n215\n232\n14. RobinsonWS\n1951\nA Method for Chronologically Ordering Archaeological Deposits.\nAmerican Antiquity\n16(4)\n293\n301\n15. PetrieWMF\n1899\nSequences in Prehistoric Remains.\nThe Journal of the Royal Anthropological Institute of Great Britain and Ireland\n29\n295\n301\n16. Bar-JosephZGiffordDJaakkolaTS\n2001\nFast optimal leaf ordering for hierarchical clustering.\nBioinformatics\n17\nS22\nS29\n11472989\n17. CarauxGPinlocheS\n2005\nPermutMatrix: A graphical environment to arrange gene expression profiles in optimal linear order.\nBioinformatics\n21(7)\n1280\n1281\n15546938\n18. PuolämkiKForteliusMMannilaH\n2006\nSeriation in paleontological data using markov chain monte carlo methods.\nPLoS Comput Biol\n2(2)\ne6\n16477311\n19. SokalRRMichenerCD\n1958\nA statistical method for evaluating systematic relationships.\nUniv. Kans. Sci. Bull\n28\n1409\n1438\n20. BlackshawSHarpavatSTrimarchiJCaiLHuangH\n2004\nGenomic analysis of mouse retinal development.\nPLoS Biol\n2(9)\nE247\n15226823\n21. SiddiquiASKhattraJDelaneyADZhaoYAstellC\n2005\nA mouse atlas of gene expression: Large-scale digital gene-expression profiles from precisely defined developing C57BL/6J mouse tissues and cells.\nProc Natl Acad Sci U S A\n102(51)\n18485\n18490\n16352711\n22. KimSKMacDonaldRJ\n2002\nSignaling and transcriptional control of pancreatic organogenesis.\nCurr Opin Genet Dev\n12(5)\n540\n547\n12200159\n23. KitamuraTIdo KitamuraY\n2007\nRole of FoxO proteins in pancreatic beta cells.\nEndocr J\n54(4)\n507\n515\n17510498\n24. MarsichEVetereADi PiazzaMTellGPaolettiS\n2003\nThe Pax6 gene is activated by the basic helix-loop-helix transcription factor NeuroD/BETA2.\nBiochem J\n376\n707\n715\n12962539\n25. GasaRMrejenCLynnFCSkewes-CoxPSanchezL\n2007\nInduction of pancreatic islet cell differentiation by the neurogenin-neuroD cascade. [Epub ahead of print]\n26. HabenerJFKempDMThomasMK\n2005\nMinireview: transcriptional regulation in pancreatic development.\nEndocrinology\n146(3)\n1025\n1034\n15604203\n27. GriffithOLPleasanceEDFultonDLOveisiMEsterM\n2005\nAssessment and integration of publicly available SAGE, cDNA microarray, and oligonucleotide microarray expression data for global coexpression analyses.\nGenomics\n86(4)\n476\n488\n16098712\n28. FirulliAB\n2003\nA HANDful of questions: The molecular biology of the heart and neural crest derivatives (HAND)-subclass of basic helix-loop-helix transcription factors.\nGene\n312\n27\n40\n12909338\n29. WestermanBAMurreCOudejansCB\n2003\nThe cellular Pax-Hox-helix connection.\nBiochim Biophys Acta\n1629(1–3)\n1\n7\n14522074\n30. HoffmanBGZavagliaBWitzscheJRuiz de AlgaraTThiessenN\n2008\nIdentification of Transcripts with Enriched Expression in the Developing and Adult Pancreas.\nGenome Biology\n9(6)\nR99\n18554416\n31. RobertsonNOveisi-FordoreiMZuyderduynSDVarholRJFjellC\n2007\nDiscoverySpace: An interactive data analysis application.\nGenome Biol\n8(1)\nR6\n17210078\n32. HubbardTJAkenBLBealKBallesterBCaccamoM\n2007\nEnsembl 2007.\nNucleic Acids Res\n35(Database issue)\nD610\n7\n17148474\n33. BatemanACoinLDurbinRFinnRDHollichV\n2004\nThe pfam protein families database.\nNucleic Acids Res\n32(Database issue)\nD138\n41\n14681378\n34. HosackDADennisGJrShermanBTLaneHCLempickiRA\n2003\nIdentifying biological themes within lists of genes with EASE.\nGenome Biol\n4(10)\nR70\n14519205\n35. Al-ShahrourFMinguezPTarragaJMedinaIAllozaE\n2007\nFatiGO +: A functional profiling tool for genomic data. integration of functional annotation, regulatory motifs and interaction data with microarray experiments.\nNucleic Acids Res\n35(Web Server issue)\nW91\n6\n17478504"
+}

batch_3/PMC2527569.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527569",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527569\nAUTHORS: Junwu Mu, John C Slevin, Dawei Qu, Sarah McCormick, S Lee Adamson\n\nABSTRACT:\nBackgroundNon-invasive micro-ultrasound was evaluated as a method to quantify intrauterine growth phenotypes in mice. Improved methods are required to accelerate research using genetically-altered mice to investigate the interactive roles of genes and environments on embryonic and placental growth. We determined (1) feasible age ranges for measuring specific variables, (2) normative growth curves, (3) accuracy of ultrasound measurements in comparison with light microscopy, and (4) weight prediction equations using regression analysis for CD-1 mice and evaluated their accuracy when applied to other mouse strains.MethodsWe used 30–40 MHz ultrasound to quantify embryonic and placental morphometry in isoflurane-anesthetized pregnant CD-1 mice from embryonic day 7.5 (E7.5) to E18.5 (full-term), and for C57Bl/6J, B6CBAF1, and hIGFBP1 pregnant transgenic mice at E17.5.ResultsGestational sac dimension provided the earliest measure of conceptus size. Sac dimension derived using regression analysis increased from 0.84 mm at E7.5 to 6.44 mm at E11.5 when it was discontinued. The earliest measurement of embryo size was crown-rump length (CRL) which increased from 1.88 mm at E8.5 to 16.22 mm at E16.5 after which it exceeded the field of view. From E10.5 to E18.5 (full term), progressive increases were observed in embryonic biparietal diameter (BPD) (0.79 mm to 7.55 mm at E18.5), abdominal circumference (AC) (4.91 mm to 26.56 mm), and eye lens diameter (0.20 mm to 0.93 mm). Ossified femur length was measureable from E15.5 (1.06 mm) and increased linearly to 2.23 mm at E18.5. In contrast, placental diameter (PD) and placental thickness (PT) increased from E10.5 to E14.5 then remained constant to term in accord with placental weight. Ultrasound and light microscopy measurements agreed with no significant bias and a discrepancy of less than 25%. Regression equations predicting gestational age from individual variables, and embryonic weight (BW) from CRL, BPD, and AC were obtained. The prediction equation BW = -0.757 + 0.0453 (CRL) + 0.0334 (AC) derived from CD-1 data predicted embryonic weights at E17.5 in three other strains of mice with a mean discrepancy of less than 16%.ConclusionMicro-ultrasound provides a feasible tool for in vivo morphometric quantification of embryonic and placental growth parameters in mice and for estimation of embryonic gestational age and/or body weight in utero.\n\nBODY:\nBackgroundGenetically-altered mouse models are proving powerful tools for studying the genetic regulation of embryonic and placental growth and development [1-3], and the interaction between genes and the environment on intrauterine and postnatal growth [4]. Advancing knowledge gained from such models is important given the critical importance of intrauterine growth as a risk factor for perinatal and childhood morbitity and mortality [5] and for diverse adult-onset diseases including diabetes, cancer, and hypertension [6-8]. Factors regulating intrauterine growth are known to differ from those important postnatally, and they remain poorly understood [9,10]. Thus, methods to monitor embryonic and placental growth efficiently and accurately in utero in mice would accelerate progress in this important area.Body weight is the most common parameter used to quantify growth but it provides no information on whether growth is proportionate or preferentially affects length, girth, or other body proportions. Placental growth is often neglected despite the critical role of this organ in supporting embryonic growth and maternal adaptations to pregnancy. Abnormal placental size is now recognized as an early predictor of poor fetal growth and poor pregnancy outcome in human pregnancy [11]. Furthermore, detection of a decelerating rate of intrauterine growth using ultrasound improves the sensitivity of detection of compromised human fetuses [12] suggesting serial measurements of growth would also be of value when phenotyping mouse models with intrauterine growth abnormalities. Thus there is a pressing need for methods to quantify prenatal growth characteristics as a function of gestation in genetically-altered and/or environmentally-challenged mice.Most prior work in mice has evaluated prenatal growth using ex vivo embryonic and/or placental weights as measured variables. In human pregnancy, ultrasound is extensively used to quantify fetal and placental growth, and to estimate fetal gestational age and/or body weight based on morphometric measurements. Measurement parameters include gestational sac dimension, crown-rump length, abdominal circumference, biparietal skull diameter, and femur length [13,14]. Recent work in mice showed these parameters can be measured in embryos in utero using 7.5 to 15 MHz ultrasound [15-18] and can be used to generate prediction equations for gestational age [16,17]. However, information on normal growth trajectories for embryonic parameters is limited and there is no information on placental parameters or on measurement accuracy, and no body weight prediction equations exist for using ultrasound measurements of mouse embryos.There have been major technological advances in small animal imaging [19,20] including the development of micro-ultrasound [21]. A lateral resolution of ~40 μm is achieved using ~40 MHz ultrasound and this represents an approximate 10-fold improvement over more conventional 15 MHz ultrasound [22]. Micro-ultrasound has been used to quantify growth of the lens of the embryonic eye in mouse embryos from E11.5 to term [23] suggesting that this higher resolution instrumentation might permit growth quantification of other parameters in the embryo and placenta to commence earlier in gestation and provide more precise measurements than previously possible.In the current study, we used 40 MHz ultrasound to image the postimplantation mouse conceptus and determined (1) feasible age range for measuring specific variables using on-screen digital calipers, (2) normative growth curves and gestation prediction equations, (3) accuracy of ultrasound measurements in comparison with light microscopy, and (4) body weight prediction equations using regression analysis for CD-1 mice and their accuracy when applied to other mouse strains.MethodsExperiments were approved by the animal care committee of Mount Sinai Hospital and were conducted in accord with guidelines established by the Canadian Council on Animal Care. The normal developmental time-course for growth parameters were obtained in pregnant out-bred mice between 1 and 5 PM (CD-1; Harlan Sprague Dawley, Indianapolis, IN). Mice were on a 12 h light dark cycle, were housed in SPF conditions, and were fed ad-lib (Purina Picolab Rodent Diet 20). Measurements were obtained from transcutaneous, non-invasive ultrasound images obtained from a total of 211 embryos from at least three pregnant mice per gestational day from E7.5 to E18.5. After ultrasound exams, the mouse was killed while still anesthetized, and the embryos and placentas were collected for direct measurement of weight. In some cases, direct measurements of dimensions by light microscopy were made using an eye-piece graticule (Fig. 1) to evaluate the accuracy of the ultrasound measurements. In these cases, the locations of embryos in the abdomen were recorded during the ultrasound exams and the corresponding embryos were identified post mortem (144 embryos from 16 pregnant mice between E11.5 and E18.5 of gestation). These were used for pair-wise comparisons of in utero and ex utero measurements.Figure 1Validation of ultrasound dimension measurements by light microscopy. (A-D) Ultrasound images of an embryo at E14.5 illustrating measurement locations for crown-rump length (CRL), biparietal diameter (BPD), abdominal anteroposterior diameter (APD), abdominal transverse diameter (ATD), placental diameter (PD), and placental thickness (PT) and (E-H) obtained ex vivo by light microscopy.Day 0.5 of pregnancy (E0.5) was defined as morning on the day a vaginal plug was found after overnight mating. Mice were lightly anesthetized with ~1.5% isoflurane in oxygen by face mask. Hair was removed from the abdomen by shaving, followed by a chemical hair remover. Pre-warmed gel was used as an ultrasound coupling medium. A 30 MHz or 40 MHz transducer operating at 30 frames/s was used to transcutaneously image embryos within the maternal abdomen (Model Vevo 660, VisualSonics Inc., Toronto, ON, Canada). Maternal heart rate and rectal temperature were monitored (Model THM100; Indus Instruments, Houston, TX), and heating was adjusted to maintain rectal temperature between 36 and 38°C.The electronic calipers of the ultrasound software were used to measure embryonic and placental dimensions on the ultrasound screen (Fig. 1, 2). The long axis and the largest dimension perpendicular to the long axis were measured and averaged to provide a measurement of the size of the gestational sac (i.e. the fluid-filled structure containing the embryo which is visualized as an anechoic (dark) space bounded by the surrounding echogenic (white) tissue of the parietal yolk sac) (Fig. 2A). Eye lens diameter was the average of the largest dimension and the orthogonal dimension (Fig. 2B). Femur length was measured on a longitudinal view from the outer edges of the ossified bone (Fig. 2C). The CRL was quantified as the maximum distance from the cephalic pole to the caudal pole (Fig. 1A). The BPD was measured from the outer border of the transverse axial view of the head in which the central midline echo and the lateral ventricles were visible (Fig. 1B). Abdominal circumference (AC) was calculated from the abdominal anteroposterior diameter (APD) and abdominal transverse diameter (ATD) measured from a transverse section of the fetal abdomen at the level of the stomach and the umbilical vein (Fig. 1C), where AC = π (ATD + APD)/2. For placental measurements, a transverse image of the placenta was obtained at the insertion site of the umbilical cord and the placental diameter (PD) was measured. Placental thickness (PT) was measured at the centre of the placenta from the chorionic surface to the echogenic calcium deposits in the giant cell layer [24] (Fig. 1D).Figure 2Ultrasound dimension measurements. Ultrasound images illustrating where dimension measurements were obtained (arrows). (A) Gestational sac at E8.5. (B) Embryo eye at E16.5 showing lens and surrounding vitreous humor. (C) Longitudinal view of femur at E17.5.Results are presented as individual embryo values (Fig. 3) or as the values predicted at each gestational age from the regression equations shown in Table 1 and listed in Table 2. A p value of < 0.05 was considered statistically significant. Non-linear regression analysis was used to determine the relationship between the parameter and gestational age. Regression analysis was used to generate equations relating fetal weight to measured ultrasound parameters. Agreement between ultrasound and light microscopic measurements was quantified using Bland-Altman analysis [25] and was expressed as the 95% confidence interval for the percent difference (100 × (ultrasound - light microscopy)/average of two methods). Equations derived from CD-1 mice were applied to estimate body weight from ultrasound parameters in embryos from different strains, and the agreement between measured body weight and predicted body weight was expressed as the mean absolute percent discrepancy (100 × (absolute value of predicted - measured weight)/measured weight).Figure 3Embryonic growth quantified using ultrasound parameters. (A) gestational sac dimension (GS), (B) crown-rump length (CRL), (C) biparietal diameter (BPD), (D) abdominal circumference (AC), (E) femur length (FL), and (F) lens diameter (LD) measured non-invasively by ultrasound in vivo are shown as a function of gestational age. The lines were generated using the regression equation between the parameter and gestational age shown in Table 1. The regression equations were derived using the datapoints shown (each point is the result obtained in one conceptus).Table 1Prediction equations for growth parameters (in mm) as a function of gestational age (in days)Gestational sac (GS):GS = -9.66 + 1.40(GA)R2 = 0.9602Crown-rump length (CRL):CRL = -9.42 + 1.09(GA) + 0.0281(GA)2R2 = 0.9682Biparietal diameter (BPD):BPD = 50.47 - 17.14(GA) + 2.07(GA)2 - 0.103(GA)3 + 0.00186(GA)4R2 = 0.9733Abdominal circumference (AC):AC = -55.75 + 7.52(GA) - 0.166(GA)2R2 = 0.9723Femur length (FL):FL = -5.02 + 0.392(GA)R2 = 0.8215Lens diameter (LD):LD = 2.205 - 0.518(GA) + 0.0401(GA)2 - 0.000856(GA)3R2 = 0.9618Body weight (BW):BW = 0.5488 - 0.01714(GA) - 0.01180(GA)2 + 0.0008279(GA)3R2 = 0.9906Placental diameter (PD):PD = -11.96 + 2.09(GA) - 0.046(GA)2 - 0.0005(GA)3R2 = 0.8941Placental thickness (PT):PT = 4.10 - 1.14(GA) + 0.115(GA)2 - 0.0031(GA)3R2 = 0.7562Placental weight (PW):PW = -0.54180 + 0.07887(GA) - 0.002243(GA)2R2 = 0.8237GA, gestational ageTable 2Predicted fetal measurements at each gestational age using regression equations in Table 1GA (days)BW (g)GS (mm)CRL (mm)BPD (mm)AC (mm)FL (mm)LD (mm)7.50.848.52.241.880.799.53.643.471.3010.50.0265.045.122.094.910.2011.50.0506.446.833.008.780.2512.50.1088.603.9012.310.3213.50.20410.424.7015.520.4114.50.34312.295.3718.390.5215.50.53114.235.9220.931.060.6216.50.77216.226.3923.141.450.7317.51.0726.8925.011.840.8318.51.4357.5526.562.230.93GA, gestational age, BW, body weight, GS, gestational sac; CRL, crown-rump length; BPD, biparietal diameter; AC, abdominal circumference; FL, femur length; LD, lens diameterResults and DiscussionThe gestational sac dimension was the earliest quantitative measure of growth and was consistently measurable from E7.5. It provides a measure of the fluid space surrounding the embryo. Gestational sac dimension increased linearly by 1.40 mm/d from 0.84 mm at E7.5 to 6.44 mm at E11.5 when measurement of this parameter was discontinued (Fig. 3A and Table 1, 2). By this method, gestational sac dimension was measurable two days earlier than in prior work using 15.5 MHz ultrasound [16]. Both methods yielded similar gestational sac dimensions at E9.5 (4.4 mm vs. 3.64 mm in the current study).The crown-rump length of the embryo was measurable from E8.5 to E16.5 when the length of most CD-1 embryos exceeded the field of view so were no longer measurable. Crown-rump length increased non-linearly from 1.88 mm at E8.5 to 16.22 mm at E16.5 (Fig. 3B and Tables 1, 2). At E8.5 the embryonic headfold is at an early stage of development and the embryo has not yet rotated into the embryonic position characteristic of the rest of gestation. Nevertheless, the 'crown-rump length' measured at this gestation was congruent with the relationship between crown-rump length and gestational age of older embryos (Fig. 3B). Crown-rump length was one of the easiest parameters to measure, and regression analysis showed that it was a good predictor of embryonic body weight and of gestational age (Tables 3, 4). Prior work using 15 MHz ultrasound showed that crown-rump length could be measured as early as E10.5 in CD-1 embryos [16] or E12.5 in C57Bl/6J embryos [15,18]. A more recent publication indicates that crown-rump length is measurable from E5.5 to E18.5 in CD-1 and C57Bl/6J embryos using 7.5–10 MHz ultrasound but the measurement accuracy was not reported at any gestational age. Crown-rump length in the current study tended to be larger than values predicted using formulas for CD-1 embryos [16,17] or reported in Tables for C57Bl/6 embryos [15,18] in prior work using 7.5 to 15 MHz ultrasound (Fig. 4). Nevertheless, we found good agreement between crown-rump length measured by light microscopy ex vivo and ultrasound in vivo (Fig. 5A). Overall, there was no significant bias, and the difference between measurements by the two methods was 25% or less (Fig. 6A). Thus, measurements using lower resolution ultrasound may underestimate crown-rump length.Figure 4Comparison of crown-rump length and biparietal diameter measurements with prior work. Crown-rump length (CRL) and biparietal diameter (BPD) are shown as a function of gestational age. Open symbols show the results obtained for each conceptus in the current study. The solid lines were generated using the regression equations shown in Table 1. Solid symbols show results from prior work. Values were calculated using formulas for CD-1 embryos obtained using 7.5–10 MHz (blue squares; [17]) or 15 MHz (green triangles; [16]) ultrasound or are means reported in Tables for C57Bl/6 embryos (red circles [18], purple diamonds [15]).Figure 5Correspondence between ultrasound and light microscopy measurements. Relationship between measurements obtained by ultrasound (UBM) in vivo and by light microscopy (LMM) ex vivo for (A) crown-rump length (CRL), (B) biparietal diameter (BPD), (C) anterioposterior abdominal dimension (APD), and (D) placental diameter (PD). Each point shows the result obtained in one conceptus. The lines show the line of identity (where y = x).Figure 6Statistical evaluation of ultrasound versus light microscopy measurements. Bland-Altman analyses of the relationship between measurements obtained by ultrasound (UBM) in vivo and by light microscopy (LMM) ex vivo for (A) crown-rump length (CRL), (B) biparietal diameter (BPD), (C) anterioposterior abdominal dimension (APD), and (D) placental diameter (PD). The difference between paired measurements are plotted against the mean of the two measurements. Each point shows the paired results obtained in one conceptus. The solid line in each graph shows the bias between the two measurement methods. The bias was not significantly different from zero for all four variables. The dashed lines show the ± 95% confidence intervals.Table 3Prediction equations for gestational age (in days) from measured variables (in mm)Gestational sac (GS):GA (day) = 6.687 + 1.395(GS) - 0.4391(GS)2 + 0.09837(GS)3 - 0.007091(GS)4R2 = 0.9683Crown-rump length (CRL):GA (day) = 7.622 + 0.5264(CRL) + 0.009440(CRL)2 - 0.0005539(CRL)3R2 = 0.9693Biparietal diameter (BPD):GA (day) = 8.195 + 0.8689(BPD) + 0.08056(BPD)2R2 = 0.9648Abdominal circumference (AC):GA (day) = 7.645 + 0.8774(AC) - 0.07917(AC)2 + 0.004024(AC)3 - 6.508e-5(AC)4R2 = 0.9698Femur length (FL):GA (day) = 12.24 + 3.822(FL) - 0.5103(FL)2R2 = 0.8287Lens diameter (LD):GA (day) = 11.96 - 84.88(LD) + 1470(LD)2 - 4625(LD)3R2 = 0.7582GA, gestational ageTable 4Prediction equations for body weight (in g) from measured variables (in mm)From crown-rump length (CRL):BW = -0.696 + 0.0890(CRL)R2 = 0.938From biparietal diameter (BPD):BW = -34.08 + 32.10(BPD) - 11.130(BPD)2 + 1.68(BPD)3-0.093(BPD)4R2 = 0.945From abdominal circumference (AC):BW = 4.20 - 0.76(AC) + 0.045(AC)2 - 0.00078(AC)3R2 = 0.957From crown-rump length (CRL) and abdominal circumference (AC):BW = -0.757 + 0.0453(CRL) + 0.0334(AC)R2 = 0.962BW, body weightAbdominal dimensions were sometimes measurable at E9.5 but were consistently measurable from E10.5 onwards. Abdominal anteroposterior diameter measured by ultrasound in vivo showed good agreement with light microscopic measurement ex vivo (Fig. 5C), with no significant bias and a discrepancy of <21% (Fig. 6C). Abdominal anteroposterior and transverse diameters were used to calculate abdominal circumference. Abdominal circumference provides an indicator of soft tissue growth of abdominal organs, primarily the liver [26,27]. Abdominal circumference increased non-linearly with advancing gestation (Fig. 3D, Table 1, 2). Regression analysis showed that abdominal circumference was a good predictor of embryonic body weight and gestational age (Table 3, 4), which is consistent with prior work in human pregnancy.Biparietal diameter increased from 0.79 mm at E8.5 (when it was measurable in most embryos) to 7.55 mm at E18.5 based on the non-linear regression equation for biparietal diameter as a function of gestational age (Fig. 3C and Table 1). Biparietal diameter was a good predictor of gestational age (R2 = 0.9648; Table 3) and body weight (R2 = 0.945; Table 4). Biparietal diameters in the current study were generally within the range predicted using formulas published previously for CD-1 embryos (obtained using 7.5 – 10 MHz [17] or 15 MHz ultrasound [16]) (Fig. 4). Whether biparietal diameter was measured by ultrasound in vivo or by a light microscope ex vivo (Fig. 5B), bias was not significant and the difference was <20% (Fig. 6B). Biparietal diameter provides prenatal diagnosis of microcephaly in human pregnancy [28] and may reveal asymmetric growth in intrauterine growth restriction. We note that the bias between ultrasound and light microscopy measurements of biparietal diameter tended to be smaller than for soft tissues (i.e. anterioposterior abdominal dimension, placental diameter, and crown-rump length) (+1% versus -6% to -8%) (Fig. 6). Although biases were not statistically significant, this trend may be caused by the less distinct tissue boundaries for soft tissues when viewed by ultrasound.The femur was first detectable within the hind limb at ~E15.5 and increased linearly in length to term at a rate of 0.392 mm/d (Fig. 3E and Table 1, 2). Prior work also found that the femur was first visualized by ultrasound at this gestational age and suggested it may be useful as a marker for this stage of development [16]. Ossification of the femur was first detected between E14.5 and E15.5 in ex vivo specimens and was primarily localized to the middle of the femur with the extremities of the femur composed of cartilage [29]. Ultrasound-detectable femur length in the current study is 41% to 46% of that determined ex vivo at E15.5 to E18.5 respectively [29], likely because only the middle, ossified region of the femur is detectable by ultrasound. Nevertheless the growth of the femur determined in the current study over this interval (+110%) is similar to that of the whole femur assessed ex vivo (+88%) [29] suggesting that it is a useful non-invasive measure of long bone growth.The lens, vitreous humor and retina of the eye were visible by ultrasound from E10.5 onwards (Fig. 2B) as shown previously using similar ultrasound instrumentation and the same mouse strain [23]. Lens diameter increased non-linearly with gestational age from 0.20 mm at E10.5 to 0.93 mm at E18.5 (Fig. 3F, Table 1, 2). We note that non-linearity in our data was primarily due to the earliest age point and thus may reflect slower growth during early differentiation of the eye. When linear regression was applied as in prior work, the linear growth rate was ~90 μm/day which is similar to the 70 μm/day previously reported [23]. Lens diameter increases approximately linearly with gestational age in human fetuses from 15 to 40 weeks gestation [30]. In human fetuses, slow ocular growth is associated with delayed cerebral development [31,32]. Thus a measurement of lens diameter may provide a useful phenotyping marker for eye and, indirectly, brain development in mouse models.Placental diameter and placental thickness were found to increase non-linearly with gestational age (Table 1, Fig. 7). Both measures of placental size increased progressively from E10.5 to ~E14.5 then remained constant to term (Fig. 7). A growth plateau in late gestation is in accord with the plateau observed in placental weight measurements (Fig. 8B) and contrasts with continued late-gestational increases in fetal body weight (Fig. 8A) and umbilical blood flow velocity [33]. The late-gestational plateau in placental growth corresponds to a maturational phase of placental development in which vascularity increases and the thickness of the materno-fetal interhaemal barrier decreases [34] thereby enhancing placental transfer efficiency.Figure 7Placental growth quantified using ultrasound parameters. Relationship between (A) placental diameter (PD) and (B) placental thickness (PT). The lines were generated using the regression equation between the parameter and gestational age shown in Table 1. The regression equations were derived using the datapoints shown (each point is the result obtained in one conceptus).Figure 8Embryonic and placental growth quantified using ex vivo weight. Measured (A) embryo and (B) placental weights are shown as a function of gestational age. Each point shows the result obtained in one conceptus. Lines were generated using the regression equations shown in Table 1.In human pregnancy, ultrasound parameters are routinely used to estimate fetal gestational age and body weight. Thus, we used regression analysis to generate equations to predict gestational age from individual ultrasound parameters (Table 3). We used CD-1 mice, an out-bred strain often used in reproductive research because it is a reliable and prolific breeder. These equations may be useful in future studies on CD-1 mice to estimate embryonic age when the plug date is unknown. We also used the parameters of crown-rump length, abdominal circumference, and biparietal diameter alone and in combination to generate equations to predict embryonic body weight (Table 4). We found that crown-rump length and abdominal circumference provided a good prediction of embryonic body weight (Fig. 9A) and that there was no significant improvement achieved with the inclusion of biparietal diameter (not shown). We also evaluated the ability of this equation to predict embryonic weights in three other strains of mice with embryos of discrepant size. We used C57Bl/6J and B6CBAF1 mice because they are common background strains for genetically-altered mice, and a hIGFBP1 transgenic model [35] as an example of a genetically-altered mouse model with intrauterine growth restriction. The prediction equation BW = -0.757 + 0.0453 (CRL) + 0.0334 (AC) derived from CD-1 data was used to predict embryonic weights at E17.5 in C57Bl/6J, B6CBAF1, and hIGFBP1 transgenic mice (Fig. 9B). The fit tended to diverge from predicted for embryo weights >0.8 g (Fig. 9B). This may be because these weights are largely in the extrapolated range of the equation or, alternatively, because the equation overestimates these weights due to strain differences. Nevertheless, the mean absolute discrepancy for C57Bl/6J, B6CBAF1, and hIGFBP1 transgenic embryos was 12, 16, and 13% respectively (Fig. 9B) which was similar to the value of 14% calculated for CD-1 mice (E12.5 – E16.5; Fig. 9A). A body weight prediction equation using data from all four strains was also derived (Table 5). Again, crown-rump length and abdominal circumference were found to be the best predictors, with no significant improvement afforded by the inclusion of biparietal diameter. We evaluated the fit of this equation (BW = -0.858 + 0.0659(CRL) + 0.0257(AC)) to the measured body weights of the four strains (Fig. 9C). The mean absolute discrepancy using this equation was 15% and thus was similar to that obtained using the equation derived from CD-1 data alone.Figure 9Correspondence between predicted and measured embryonic body weight. (A) Embryo weight predicted using the multiple regression equation based on ultrasound measurement of crown-rump length and abdominal circumference in CD-1 mice versus measured body weight for each CD-1 embryo. (B) Equation derived from data obtained in CD-1 mice applied to three other strains of mice (C57Bl/6J (dark blue squares), B6CBAF1 (pink triangles), and hIGFBP1 transgenics (orange diamonds)). (C) Equation derived using data from all four strains is shown applied to all four strains (CD-1 (black circles), C57Bl/6J (dark blue squares), B6CBAF1 (pink triangles), and hIGFBP1 transgenics (orange diamonds)). Each point shows the result obtained in one conceptus. The lines show the line of identity (where y = x).Table 5Prediction equations for body weight (in g) from measured variables (in mm) using data from four strainsFrom crown-rump length (CRL):BW = -0.778+ 0.0966(CRL)R2 = 0.906From biparietal diameter (BPD):BW = 12.21 - 7.96(BPD) + 1.67(BPD)2 - 0.110(BPD)3R2 = 0.844From abdominal circumference (AC):BW = -0.08 - 0.018(AC) + 0.00247(AC)2R2 = 0.879From crown-rump length (CRL) and abdominal circumference (AC):BW = -0.858 + 0.0659(CRL) + 0.0257(AC)R2 = 0.918BW, body weightThe use of 40 MHz ultrasound for phenotypic analysis of the conceptus also has important limitations including the skill required and the relatively high cost of the equipment. In addition, it is often difficult to achieve the optimal view for morphometric measurements and this is an important source of measurement error. Depending on the number of embryos and their location, some live embryos may not be visible (~10% in our experience [20]) and some may not be in an appropriate orientation for accurate measurement. There is also the possibility that bioeffects associated with anesthesia and/or ultrasound could affect subsequent development of the conceptus. 40 MHz ultrasound under isoflurane anesthesia during organogenesis (E8.5 or E10.5) had no significant effect on birth weight and minimal effects on postnatal growth [36]. However, fetal ultrasound [37] and embryonic exposure to isoflurane [38] can affect biological outcomes so appropriate controls are necessary.ConclusionEmbryonic and placental growth parameters were quantified using 40 MHz ultrasound generating normal growth curves over parameter-specific gestational intervals. Parameters tested exhibited no systematic errors relative to ex vivo measurements by light microscopy, and embryonic body weights estimated using equations derived from CD-1 mice were similarly accurate in three other mouse strains. We found that in vivo quantification of placental size is adequate to detect the normal cessation of placental growth that occurs at ~E14.5. The capacity to quantify placental growth in vivo is important given the crucial role of the placenta in supporting embryonic growth, and our limited understanding of placental growth control. Thus, micro-ultrasound provides a feasible means for obtaining detailed information on prenatal embryonic and placental growth characteristics in genetically-altered and/or environmentally-challenged mouse models, and may also prove useful for estimating gestational age and/or embryonic body weight in utero.Competing interestsSLA was a member of the Scientific Advisory Board of VisualSonics from 2003 to 2006 but otherwise has no financial interests in the company.Authors' contributionsJM conceived the study, JM and SLA participated in study design, JM and JCS performed ultrasound imaging, DQ participated in breeding and study coordination, SM performed statistical analysis and prepared graphs and tables, and JM and SLA drafted the manuscript. All authors read and approved the final manuscript.\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2527677.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527677",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527677\nAUTHORS: Nelson J. R. Fagundes, Ricardo Kanitz, Sandro L. Bonatto\n\nABSTRACT:\nThe Americas were the last continents to be populated by humans, and their colonization represents a very interesting chapter in our species' evolution in which important issues are still contentious or largely unknown. One difficult topic concerns the details of the early peopling of Beringia, such as for how long it was colonized before people moved into the Americas and the demography of this occupation. A recent work using mitochondrial genome (mtDNA) data presented evidence for a so called “three-stage model” consisting of a very early expansion into Beringia followed by ∼20,000 years of population stability before the final entry into the Americas. However, these results are in disagreement with other recent studies using similar data and methods. Here, we reanalyze their data to check the robustness of this model and test the ability of Native American mtDNA to discriminate details of the early colonization of Beringia. We apply the Bayesian Skyline Plot approach to recover the past demographic dynamic underpinning these events using different mtDNA data sets. Our results refute the specific details of the “three-stage model”, since the early stage of expansion into Beringia followed by a long period of stasis could not be reproduced in any mtDNA data set cleaned from non-Native American haplotypes. Nevertheless, they are consistent with a moderate population bottleneck in Beringia associated with the Last Glacial Maximum followed by a strong population growth around 18,000 years ago as suggested by other recent studies. We suggest that this bottleneck erased the signals of ancient demographic history from recent Native American mtDNA pool, and conclude that the proposed early expansion and occupation of Beringia is an artifact caused by the misincorporation of non-Native American haplotypes.\n\nBODY:\nIntroductionThe Americas were the last continents to be settled by modern humans, most probably from northeast Asia through Beringia, the landmass that connected Asia and the Americas during periods of low sea-level [1]. Archeological data suggest that the continent was colonized in the late Pleistocene after the Last Glacial maximum (LGM). The oldest sites for North and South America are about 14.5 ky old [1], possibly suggesting a fast southward movement of the initial settlers. However, the scarceness of late Pleistocene human remains in northeast Asia make it difficult to evaluate the details of the population processes in Beringia that ultimately led to the peopling of the New World.As an alternative to the study of archeological data, molecular data have been extensively used to infer when and how modern humans colonized the world reviewed in [2], [3]. Since the pioneering study by Cann et al. [4], the mitochondrial DNA (mtDNA) has become the most widely used genetic marker to study recent human evolution [5]. In Native Americans, early studies of mtDNA variation found that these populations have five distinct major mtDNA haplogroups (A, B, C, D and X) [6], [7], all of Asian origin. Although most of these studies seemed to converge on a model suggesting a single pre-Clovis migration [8]–[10], no consensus emerged for details such as the timing and pace of the putative occupation event under this scenario. These controversies notwithstanding, the divergence between Native American and Asian sequences for each mtDNA haplogroup led some authors to suggest that the Native American founder population stayed isolated in Beringia from the remaining Asian populations prior to their entry in the Americas [9], the so called “out of Beringia” model. Under this scenario, Beringia played a key-role in the differentiation of the mtDNA haplogroups presently found in Native Americans.The study of complete mtDNA sequences from Native Americans [11]–[14] has allowed investigators to examine mtDNA variation in the New World with much greater resolution. The first systematic survey of coding-region mtDNA sequences including individuals from Native American ancestry was carried by Herrnstadt et al. [11], who studied individuals sampled from an urban population and relied on the screening of an incomplete set of HVS-I markers to identify mtDNA haplotypes as being of Native American origin [15]. Afterwards, a thorough revision of these sequences partially changed the original classification of the Native American genomes [15]. Further studies, using samples obtained mainly from native American populations, revealed new putative founder haplotypes [12], [14], and suggested an average coalescence time for the most common haplogroups of around 13.5 thousand years ago (kya) or 19 kya depending on the estimates being based on only synonymous transitions [16] or on all substitutions [17], respectively. The most extensive work to date [13] showed that all five major haplogroups have a similar pattern of genetic diversity, and that they expanded together towards the end of the last glacial maximum (LGM) around 18 kya.The development of new analytical methods allowed the estimation of the past demography changes using the Bayesian Skyline Plot (BSP) approach, which allows inference of past population size changes without assuming any a priori demographic scenario [18]. Thus, so far, four roughly synchronous studies [13], [19]–[21], using partially overlapping data sets, applied such an analysis using putative Native American mtDNA genomes. Adjusting for the different substitution rates used, all of them showed evidence for quick and strong population growth in the late Pleistocene, likely near the end of the LGM, preceded by a population bottleneck that lasted for a few thousand years e.g., [13].Interestingly, only one of these studies concluded that the BSP of Native American mtDNAs suggested an additional and more ancient period of population growth followed by a long period of population stability [20]. The authors interpreted these results as representing the expansion out of Central Asia into Beringia after divergence from Asians (∼43–36 kya) followed by a long period (∼20,000 years) of population stability in Beringia and finally by the strong population growth stage (∼16 kya) after the LGM associated to the peopling of the Americas. They called this scenario the “three-stage” colonization model for the peopling of the Americas, even though several of the more recent colonization models could likewise be described as having “three stages”: a first stage from Asia into Beringia, a second stage of isolation in Beringia and a third stage with an expansion out of Beringia into the Americas [9], [10], [12]–[14], [15]. These studies highlight the importance of a stage in Beringia prior to the peopling of the Americas, and one of them even provided a rough estimate of the time the Native American founder population spent in Beringia using the number of diagnostic substitutions found in Native American mtDNA sub-haplogroups [13]. However, what differentiates the study of Kitchen et al. [20] from the others is that it was the only one that estimated a detailed demographic history for the first two stages, although it used evidence and methods (BSP) similar to those employed by the other works. Nonetheless, one possible problem with this study is that the data set of mtDNA genomes they used consisted primarily of the original data set of Herrnstadt et al. [11], in which several mtDNA genomes regarded as Native American are most likely of non-Native ancestry [15] and which also includes several sequence errors detected recently but not corrected in the mtDB database that they used [22], [23].In this study, we provide a reanalysis of the BSP results from Kitchen et al. [20] using a rigorous criterion for defining Native American ancestry [12], [14], [15] to determine if the specific three-stage model suggested by these authors is still supported when a more reliable data set is used. We also investigate the likely source of the early expansion detected by that study.Materials and MethodsData setsInitially, the original Kitchen et al. [20] data set (n = 77) was used to replicate their original findings using the evolutionary model specified in that report and another one (see below). Based on information detailed in Bandelt et al. [15] and Achilli et al. [14], we removed seven individuals who are very likely of Asian ancestry. These include three individuals assigned to haplogroup E in Herrnstadt et al. (Figure 2 in [11]), as well as four individuals belonging to the Asian sub-haplogroups B1 or B4c [15]. Another individual (Herrn552; see [11]) was removed, since it actually belongs to the West European haplogroup H. This individual was probably incorporated into the data set by mistake, as individual 532 (from Native American haplogroup A2) is absent from their data set. Finally, another individual (Kiv2870) was removed, since it had all the diagnostic coding-region mutations for West European sub-haplogroup X2b (8393, 13708, 15927), and none of the diagnostic coding-region mutations for Native American sub-haplogroup X2a (8913, 12397, 14502), and thus it is likely of recent European ancestry. That modifications resulted in a new, corrected, data set of 68 sequences of likely Native American origin, and, in a third data set with only the 9 sequences which have been misincorporated in the original data set of 77 sequences analyzed by Kitchen et al. [20]. Finally, to better test whether the anomalous early expansion seen in Kitchen et al. [20] BSP results could be explained by the non-Native American mtDNA genomes, we created a set of 10 alignments with nine genomes each randomly selected from the mtDNA genomes from the macrohaplogroups M and N [24].Bayesian Skyline PlotBSPs [18] have been constructed in the program Beast 1.4.7 (http://beast.bio.ed.ac.uk/). For all analyses, Markov Chain Monte Carlo (MCMC) samples were based on 100,000,000 generations, logging every 2,500 steps, with the first 10,000,000 generations discarded as the burn-in. All analyses were run multiple times to check for convergence. Following Fagundes et al. [13], we used the HKY+Γ evolutionary model, a log-normal relaxed molecular clock with a mean substitution rate of 1.26×10−8 mutations/site/year [17] for the complete coding sequence. The scaled effective population size was converted to the effective female population size Nef, assuming a generation time of 25 years. Importantly, assumptions about the mutation rate and the generation time will only affect the scale of the BSP, but not its shape.ResultsOur analysis of the BSP for the original Kitchen et al. data set [20] reproduced their original results, which was two periods of population growth (Figure 1A) with a long stasis between them. However, the corrected data set provided evidence for a single, post-LGM, population growth (Figure 1C), in close accordance to the other BSP analyses using only mtDNA of Native American origin [13], [19], [21]. That is, there was only a long tail of roughly constant population size between the time for most recent common ancestor (TMRCA) of each Native American haplogroup around the LGM and the sample TMRCA.10.1371/journal.pone.0003157.g001Figure 1Bayesian Skyline Plots using different sequence sets.BSPs estimated with 100 million MCMC iterations sampled every 2,500 steps using log-normal relaxed clock and HKY model plus gamma (eight categories) with the standard substitution rate of 1.26×10−8 sites/yr and a generation time of 25 yr. The y axis represents the female effective population size in a log scale and the x axis shows time in thousands of years ago (kya). The thicker blue lines are the median for population size and the thinner black lines represent the 95% higher posterior density (HPD) intervals. (A) BSP using the original 77 individuals from [20]. (B) BSP for the nine misincorporated non-Native American sequences. (C) BSP for the 68 confirmed Native American haplotypes in [20] in blue and black; and the BSP from [13] in red dashed (median) and gray lines (95% HPD interval).More strikingly, when we considered only those individuals which were removed from the original data set, the signal for the earlier expansion reappeared, despite the very low sample size in this data set (n = 9) (Figure 1B). Using our substitution rate, this expansion began ∼60 kya. These results clearly show that the signal for a population expansion that they detected [20] in Native Americans mtDNAs much before LGM is an artifact caused by the incorporation of non-Native American haplotypes into the analysis. Since these nine haplotypes seem to be of Asian and European origin, from macrohaplogroups M and N, we conjecture that this signal of expansion may be related to the demographic expansion out-of-Africa that gave rise to Eurasians. The BSPs of the ten data sets from genomes selected from these two macrohaplogroups (Figure 2) showed a very similar expansion pattern, corroborating this hypothesis.10.1371/journal.pone.0003157.g002Figure 2Bayesian Skyline Plot for ten replicates of nine random non-African haplotypes.Ten BSPs using random samples of nine non-African individuals from [24] belonging to macrohaplogroups M and N, showing a similar pattern of expansion between ∼80 and ∼40 kya. All BSPs were calculated with 100 million MCMC generations sampling every 2,500 using the same model applied to the BSPs in Figure 1. Axes and lines are as in Figure 1.DiscussionOur results strongly suggest that the demographic expansion putatively associated with the geographical expansion out of Central Asia and the initial peopling of Beringia as well as the estimation of ∼20 ky of occupation of Beringia by human groups before they entered the Americas [20] is merely a database artifact caused by the incorporation of mtDNA genomes of non-Native American ancestry in the analysis. Because the mtDNA haplogroups and sub-haplogroups have a strong and extensively studied geographic association e.g., [5], it is possible to identify almost unambiguously the ancestry of most haplotypes [15]. While this may be also true for the Y-chromosome [25], it is certainly not for most other nuclear markers, which typically display low levels of interpopulation differentiation and extensive haplotype sharing among populations e.g., [26], [27]. Possible applications of the BSP to autosomal or sex-linked haplotypes must carefully select the sampled populations to avoid incorporating into the analysis those recently introduced by admixture. Our analyses suggest that even a relatively small proportion of 12% (9/77) of “admixed” (or misassigned) haplotypes may significantly bias the overall result.The population expansion that began 60–55 kya when non-Native American haplotypes are incorporated into the analysis most likely reflects, at least in part, the early expansion and diversification of macrohaplogroups M and N in Eurasia [28], which is unrelated to the specific process of the peopling of Beringia. As a consequence, Kitchen et al.'s estimation of a period of ∼20 ky of population occupation in Beringia based on the time interval between the “two expansions” is meaningless in the context of the peopling of Beringia or the Americas. In addition, it is important to stress that, because the mtDNA haplogroups currently in America represent derivations of both macrohaplogroup M (C, D) and N (A, B, X) e.g., [14], their TMRCA reflects the TMRCA of macrohaplogroups M and N in Asia (∼60 kya) [28]. Therefore, the >40 ky of constant population size found in the corrected data sets extending from the LGM bottleneck to the past to the TMRCA of all Native American mtDNA haplogroups does not offer any detailed view of the demographic history of Native Americans before the bottleneck. The genetic bottleneck associated with human isolation in Beringia [29], [30] may have erased from the recent non-Beringian Native American mtDNA data most of the details of its pre-Beringian demographic history. In this regard, discerning the population size changes during this period would mostly require acquiring mtDNA information of ancient samples from this time.Interestingly, an almost identical pattern of population size change was found with the Kitchen et al. corrected data set and our previous analyses of mostly Native South American mtDNAs [13]. These results, therefore, strongly corroborate the mtDNA scenario for the peopling of the Americas presented in Fagundes et al. [13] and the integrated model that we suggested elsewhere [31]. This model suggests that the ancestral population colonized Beringia more than five thousand years before the LGM, remained isolated there during LGM, and likely experienced a population reduction and loss of genetic diversity by drift. The strong population expansion shown to have started around the end of LGM (∼18 kya) probably reflects the fast migration south of the Laurentide and Cordilleran ice sheets. Taking into account that the ice-free corridor between the ice sheets had not opened completely by this time interval, and that it could not have supported a viable human population earlier than 14 kya [32], [33], these findings support a coastal route as the major pathway for the peopling of the Americas, in agreement with recent published data from a panel of STR markers [34] and archeological data [35], [36].However, time estimates are dependent on the evolutionary rate used in the analysis. The mtDNA evolutionary rate that we used [17] has been the most extensively used estimate in studies of human evolution e.g., [13], [14], [28], [37]. Nevertheless, other calibrations are available, although they are usually faster than ours [16], [21], [24]. The use of an internal calibration [21] results in a rate similar to that used by Kitchen et al. [20], which pinpointed the post-LGM population growth at ∼16 kya. Another rate, which uses only synonymous substitutions [16], suggests a mean coalescent age of ∼14 ky for the major Native American haplogroups [12]. However, the corroboration of the human occupation of southern Chile ∼14.5 kya [36] strongly suggests that haplogroup coalescences and the expansion out of Beringia should have occurred >16 kya, implying that the faster rates are unlikely to be accurate. In any case, the choice of a rate affects only the absolute numeric estimates, and does not change the shape of the BSP. Thus, our refutation of the Kitchen et al. [20] model for the peopling of Beringia is independent of all the controversies about the correct mtDNA substitution rate e.g., [38]–[40].\n\nREFERENCES:\n1. GoebelTWatersMRO'RourkeDH\n2008\nThe Late Pleistocene Dispersal of Modern Humans in the Americas.\nScience\n319\n1497\n1502\n18339930\n2. SchurrT\n2004\nThe Peopling of the New World: Perspectives from Molecular Anthropology.\nAnnu Rev Anthropol\n33\n551\n583\n3. SalzanoFM\n2007\nThe prehistoric colonization of the Americas.\nCrawfordMH\nAnthropological Genetics: Theory, Methods, and Applications\nCambridge\nCambridge University Press\n433\n455\n4. CannRLStonekingMWilsonAC\n1987\nMitochondrial DNA and human evolution.\nNature\n325\n31\n36\n3025745\n5. TorroniAAchilliAMacaulayVRichardsMBandeltHJ\n2006\nHarvesting the Fruit of the Human MtDNA Tree.\nTrends in Genetics\n22\n339\n345\n16678300\n6. SchurrTGBallingerSWGanYYHodgeJAMerriwetherDA\n1990\nAmerindian mitochondrial DNAs have rare Asian mutations at high frequencies, suggesting they derived from four primary maternal lineages.\nAm J Hum Genet\n46\n613\n623\n1968708\n7. BrownMDHosseiniSHTorroniABandeltHJAllenJC\n1998\nMtDNA Haplogroup X: an Ancient Link between Europe/Western Asia and North America?\nAm J Hum Genet\n63\n1852\n1861\n9837837\n8. ForsterPHardingRTorroniABandeltHJ\n1996\nOrigin and Evolution of Native American MtDNA Variation: a Reappraisal.\nAm J Hum Genet\n59\n935\n945\n8808611\n9. BonattoSLSalzanoFM\n1997\nA Single and Early Migration for the Peopling of the Americas Supported by Mitochondrial DNA Sequence Data.\nProc Natl Acad Sci USA\n94\n1866\n1871\n9050871\n10. BonattoSLSalzanoFM\n1997\nDiversity and age of the four major mtDNA haplogroups, and their implications for the peopling of the New World.\nAm J Hum Genet\n61\n1413\n1423\n9399887\n11. HerrnstadtCElsonJLFahyEPrestonGTurnbullDM\n2002\nReduced-Median-Network Analysis of Complete Mitochondrial DNA Coding-Region Sequences for the Major African, Asian, and European Haplogroups.\nAm J Hum Genet\n70\n1152\n1171\n11938495\n12. TammEKivisildTReidlaMMetspaluMSmithDG\n2007\nBeringian Standstill and Spread of Native American Founders.\nPLoS One\n2\ne829\n17786201\n13. FagundesNJRKanitzREckertRVallsACSBogoMR\n2008\nMitochondrial Population Genomics Supports a Single Pre-Clovis Origin with a Coastal Route for the Peopling of the Americas.\nAm J Hum Genet\n82\n583\n592\n18313026\n14. AchilliAPeregoUABraviCMCobleMDKongQP\n2008\nThe Phylogeny of the Four Pan-American MtDNA Haplogroups: Implications for Evolutionary and Disease Studies.\nPLoS One\n3\ne1764\n18335039\n15. BandeltHJHerrnstadtCYaoYGKongQPRengoC\n2003\nIdentification of Native American Founder mtDNAs Through the Analysis of Complete mtDNA Sequences: Some Caveats.\nAnn Hum Genet\n67\n512\n524\n14641239\n16. KivisildTShenPWallDPDoBSungR\n2006\nThe Role of Selection in the Evolution of Human Mitochondrial Genomes.\nGenetics\n172\n373\n387\n16172508\n17. MishmarDRuiz-PesiniEGolikPMacaulayVClarkAG\n2003\nNatural Selection Shaped Regeional MtDNA Variation in Humans.\nProc Natl Acad Sci USA\n100\n171\n176\n12509511\n18. DrummondAJRambautAShapiroBPybusOG\n2005\nBayesian Coalescent Inference of Past Population Dynamics from Molecular Sequences.\nMol Biol Evol\n22\n1185\n1192\n15703244\n19. KitchenAMiyamotoMMMulliganCJ\n2008\nUtility of DNA Viruses for Studying Human Host History: Case Study of JC Virus.\nMol Phylogen Evol\n46\n673\n682\n20. KitchenAMiyamotoMMMulliganCJ\n2008\nA Three-Stage Colonization Model for the Peopling of the Americas.\nPLoS One\n3\ne1596\n18270583\n21. AtkinsonQDGrayRDDrummondAJ\n2008\nMtDNA Variation Predicts Population Size in Humans and Reveals a Major Southern Asian Chapter in Human Prehistory.\nMol Biol Evol\n25\n468\n474\n18093996\n22. BandeltHJKongQPRichardsMMacaulayV\n2006\nEstimation of Mutation Rates and Coalescence Times: Soma Caveats.\nBandeltHJMacaulayVRichardsM\nHuman Mitochondrial DNA and the Evolution of Homo sapiens\nBerlin\nSpringer\n47\n90\n23. BandeltH-JYaoY-GSalasAKivisildTBraviCM\n2007\nHigh penetrance of sequencing errors and interpretative shortcomings in mtDNA sequence analysis of LHON patients.\nBiochem Biophys Res Commun\n352\n283\n291\n17123466\n24. IngmanMKaessmannHPääboSGyllenstenU\n2000\nMitochondrial Genome Variation and the Origin of Modern Humans.\nNature\n408\n708\n713\n11130070\n25. The Y chromosome Consortium\n2002\nA Nomenclature System for the Tree of Human Y-Chromosomal Binary Haplogroups.\nGenome Res\n12\n339\n348\n11827954\n26. FagundesNJRSalzanoFMBatzerMADeiningerPLBonattoSL\n2005\nWorldwide Genetic Variation at the 3′-UTR Region of the LDLR Gene: Possible Influence of Natural Selection.\nAnn Hum Genet\n69\n389\n400\n15996168\n27. BattilanaJCardoso-SilvaLBarrantesRHillKHurtadoAM\n2007\nMolecular Variability of the 16p13.3 Region in Amerindians and its Anthropological Significance.\nAnn Hum Genet\n71\n64\n76\n17227477\n28. MacaulayVHillCAchilliARengoCClarkeD\n2005\nSingle, Rapid Coastal Settlement of Asia Revealed by Analysis of Complete Mitochondrial Genomes.\nScience\n308\n1034\n1036\n15890885\n29. BattilanaJFagundesNJRHellerAHGoldaniAFreitasLB\n2006\nAlu insertion polymorphisms in Native Americans and Related Asian Populations.\nAnn Hum Biol\n33\n142\n160\n16684689\n30. FagundesNJRRayNBeaumontMNeuenschwanderSSalzanoFM\n2007\nStatistical Evaluation of Alternative Models of Human Evolution.\nProc Natl Acad Sci USA\n104\n17614\n17619\n17978179\n31. Gonzales-JoséRBortoliniMCSantosFRBonattoSL\n2008\nThe Peopling of America: Craniofacial Shape Variation on a Continental Scale and its Interpretation From an Interdisciplinary View.\nAm J Phys Anthropol\nIn press: doi: 1002/ajpa.20854\n32. FladmarkKR\n1979\nRoutes: Alternate Migration Corridors for Early Man in North America.\nAmerican Antiquity\n44\n55\n69\n33. DixonEJ\n1993\nQuest for the Origins of the First Americans\nAlbuquerque\nUniversity of New Mexico Press\n160\n34. WangSLewisCMJakobssonMRamachandranSRayN\n2007\nGenetic Variation and Population Structure in Native Americans.\nPLoS Genet\n3\ne185\n18039031\n35. GilbertMTPJenkinsDLGötherstromANaveranNSanchezJJ\n2008\nDNA from Pre-Clovis Human Coprolites in Oregon, North America.\nScience\n320\n786\n789\n18388261\n36. DillehayTDRamírezCPinoMCollinsMBRossenJ\n2008\nMonte Verde: Seaweed, Food, Medicine, and the Peopling of South America.\nScience\n320\n784\n786\n18467586\n37. AchilliARengoCMagriCBattagliaVOlivieriA\n2004\nThe Molecular Dissection of mtDNA Haplogroup H Confirms That the Franco-Cantabrian Glacial Refuge Was a Major Source for the European Gene Pool.\nAm J Hum Genet\n75\n910\n918\n15382008\n38. HoSYWPhillipsMJCooperADrummondAJ\n2005\nTime Dependency of Molecular Rate Estimates and Systematic Overestimation of Recent Divergence Times.\nMol Biol Evol\n22\n1561\n1568\n15814826\n39. HoSYWShapiroBPhillipsMJCooperADrummondAJ\n2007\nEvidence for Time Dependency of Molecular Rate Estimates.\nSyst Biol\n56\n515\n522\n17562475\n40. BandeltHJParsonW\n2008\nConsistent treatment of length variants in the human mtDNA control region: a reappraisal.\nInt J Legal Med\n122\n11\n21\n17347766"
+}

batch_3/PMC2527838.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2527838",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2527838\nAUTHORS: T L Lash, L Pedersen, D Cronin-Fenton, T P Ahern, C L Rosenberg, K L Lunetta, R A Silliman, S Hamilton-Dutoit, J P Garne, M Ewertz, H T Sørensen\n\nABSTRACT:\nTamoxifen remains an important adjuvant therapy to reduce the rate of breast cancer recurrence among patients with oestrogen-receptor-positive tumours. Cytochrome P-450 2D6 metabolises tamoxifen to metabolites that more readily bind the oestrogen receptor. This enzyme also metabolises selective serotonin reuptake inhibitors (SSRI), so these widely used drugs – when taken concurrently – may reduce tamoxifen's prevention of breast cancer recurrence. We studied citalopram use in 184 cases of breast cancer recurrence and 184 matched controls without recurrence after equivalent follow-up. Cases and controls were nested in a population of female residents of Northern Denmark with stages I–III oestrogen-receptor-positive breast cancer 1985–2001 and who took tamoxifen for 1, 2, or most often for 5 years. We ascertained prescription histories by linking participants' central personal registry numbers to prescription databases from the National Health Service. Seventeen cases (9%) and 21 controls (11%) received at least one prescription for the SSRI citalopram while taking tamoxifen (adjusted conditional odds ratio=0.85, 95% confidence interval=0.42, 1.7). We also observed no reduction of tamoxifen effectiveness among regular citalopram users (⩾30% overlap with tamoxifen use). These results suggest that concurrent use of citalopram does not reduce tamoxifen's prevention of breast cancer recurrence.\n\nBODY:\nTamoxifen is a selective oestrogen receptor modulator (Jordan and Dowse, 1976) that reduces by half the risk of breast cancer recurrence in early-stage patients whose tumour cells express the oestrogen receptor (Early Breast Cancer Trialists' Collaborative Group, 2005). To be pharmacologically active, tamoxifen must be metabolised to secondary metabolites that bind the oestrogen receptor 100-fold more readily than tamoxifen itself (Malet et al, 1988). Four cytochrome P-450 enzymes (CYPs) catalyse this activation (CYP2D6, CYP3A4, CYP3A5, and CYP2C9) (Malet et al, 1988). CYP2D6 catalyses formation of 4-hydroxytamoxifen from tamoxifen (Coller et al, 2002) and formation of 4-hydroxy-N-desmethyltamoxifen from N-desmethyltamoxifen (Stearns et al, 2003). These two secondary metabolites have the highest binding affinity for the oestrogen receptor, and binding affinity correlates with inhibition of cell growth (Coezy et al, 1982). The secondary metabolites are, therefore, the most important modulators of the oestrogen receptor in the tamoxifen pathway (Lim et al, 2005).Breast cancer patients treated with tamoxifen may also take other prescription medications that are metabolised by some of the same enzymes that activate tamoxifen. For example, depression is a common comorbidity in breast cancer patients (Massie, 2004), and many selective serotonin reuptake inhibitors (SSRI), which are widely used medications indicated primarily to treat depression (Hansen et al, 2003), are metabolised by CYP2D6 (Zanger et al, 2004). SSRI competition with tamoxifen and N-desmethyltamoxifen for CYP2D6, or direct inhibition of CYP2D6 by SSRI, could reduce the production of the tamoxifen metabolites with high receptor-binding affinity, and thereby reduce tamoxifen's prevention of breast cancer recurrence. Competition between tamoxifen and the SSRI paroxetine reduced the plasma concentration of endoxifen in a cross-over clinical trial (Stearns et al, 2003). Furthermore, the mean plasma concentration of 4-hydroxy-N-desmethyltamoxifen was more than two-fold greater among women who were taking no CYP2D6 competitor drug than among women who were taking such a drug (Jin et al, 2005). In vivo studies thus demonstrate a compelling biological basis for the hypothesis that concomitant use of SSRI would reduce tamoxifen's prevention of breast cancer recurrence.In the largest study to date of the potential for drug–drug interaction to reduce tamoxifen's protection against breast cancer recurrence, we examined whether Danish breast cancer patients with oestrogen-receptor-positive tumours who were treated with tamoxifen for 1, 2, or most often for 5 years had a higher rate of recurrence if they were concomitantly taking the SSRI citalopram or its S-stereoisomer (‘citalopram’ from here onwards) than if they were not. As described in more detail below, citalopram was the most frequently prescribed SSRI in the study population.Materials and methodsThe study was approved by the Boston University Medical Campus Institutional Review Board and The Regional Committee on Biomedical Research Ethics of Aarhus County.Study populationThe source population included female residents of four Northern Danish counties (Aarhus, North Jutland, Viborg, and Ringkøbing) aged 35–69 at diagnosis of primary International Union Against Cancer stage I, II, or III breast cancer (UICC, 1997) between 1985 and 2001 and who were reported to the Danish Breast Cancer Cooperative Group (DBCG). The DBCG has enrolled nearly all Danish breast cancer patients younger than age 70 at diagnosis into its clinical database since 1977 (Andersen and Mouridsen, 1988; Jensen et al, 2003). More than 90% of Danish breast cancer cases are reported to the DBCG and more than half of the DBCG patients are enrolled in clinical trials (Andersen and Mouridsen, 1988). The same standardized forms are used to follow all patients reported to the DBCG, regardless of whether they enrol in a trial, so the registry provides the data quality advantage of a clinical trial setting with the generalisability advantage of a population-based setting.We divided the source population into three groups: (a) group I women whose tumour expressed the oestrogen receptor protein and who were treated with tamoxifen for at least 1 year; (b) group II women whose tumour did not express the oestrogen receptor protein, were not treated with tamoxifen, and who survived for at least one year; and (c) group III women, comprising all others, who were excluded from this analysis. Group I women were assigned to tamoxifen therapy protocols of 1, 2, or 5 years, depending on the guideline extant in Denmark at the time of their diagnoses. We included group II women to estimate the direct association of citalopram prescription with recurrence rate, if any. We further restricted the source population to women diagnosed with breast cancer after the date that their county of residence began to maintain an electronic prescription database (Aarhus=1996, North Jutland=1989, Ringkøbing=1998, Viborg=1998), which were used to ascertain use of prescription medications, including citalopram. Follow-up time began 1 year after the date of breast cancer diagnosis and continued until the date of the first of breast cancer recurrence, death from any cause, loss to follow-up (e.g., emigration), 10 years of follow-up, or 1 September 2006.Cases were women with local or distant breast cancer recurrence occurring during their follow-up time among the members of groups I and II. We selected one control for each case without replacement from members of the source population who had not had a breast cancer recurrence after the same amount of follow-up time. We matched controls to cases on (a) group membership (group I or II), (b) menopausal status at diagnosis (premenopausal or postmenopausal), (c) date of breast cancer surgery (caliper matched±12 months), (d) county of residence at the time of diagnosis, and (e) UICC stage at diagnosis (stage I, II, or III).Data collectionWe used the Danish Civil Personal Registration (CPR) number assigned to each case and control to link data sets. The CPR is a unique identification number assigned to all Danish residents alive on 1 April 1968, born thereafter, or upon immigration.We collected demographic information (age, menopausal status, and hospital of diagnosis), tumour characteristics (UICC stage, histological grade, and oestrogen-receptor expression), and therapy characteristics (primary surgical tumour management, receipt of radiation therapy, receipt of chemotherapy, and receipt of tamoxifen therapy) from the DBCG database.We collected data on receipt of citalopram prescription and other potential CYP2D6 inhibitors (including other SSRI) by linking the CPR number of cases and controls to the prescription databases maintained by each county (see, for example, the description of North Jutland's database (Gaist et al, 1997)).Analytic variablesRecurrenceWe used the DBCG definition of breast cancer recurrence as any type of breast cancer subsequent to the initial course of therapy (Andersen and Mouridsen, 1988). Given the definition of the source population and follow-up time, all cases of recurrence occurred between 1 and 10 years after the primary breast cancer diagnosis.Prescription statusPrescription medications are coded by the Anatomical Therapeutic Chemical (ATC) classification system (WHO Collaborating Centre for Drug Statistics Methodology, 2007). We defined SSRI antidepressants as all those classified in group N06AB by the ATC. These are the SSRI drugs: zimeldine (N06AB02), fluoxetine (N06AB03), citalopram (N06AB04), paroxetine (N06AB05), sertraline (N06AB06), alaproclate (N06AB07), fluvoxamine (N06AB08), etoperidone (N06AB09), and escitalopram (N06AB10). We defined citalopram exposure as any prescription for citalopram (N06AB04) or its S-stereoisomer escitalopram (N06AB10).We classified cases and controls as those with no record of a citalopram prescription during their follow-up time (never citalopram) and those with any record of prescription for citalopram during their follow-up time (ever citalopram). We used a similar procedure to classify cases and controls as ever or never users of another SSRI or of another prescription medication that is a CYP2D6 inhibitor or substrate, aside from those indicated to treat breast cancer recurrence or its effects. See the Supplementary online material for a complete list of these medications and the frequency of their use in the study population.For group I women who ever had a citalopram prescription, we calculated the percentage of time on tamoxifen when they were simultaneously taking citalopram. We created categories of (a) intermittent citalopram use, defined as citalopram use overlapping tamoxifen use for more than 0% but less than 30% of the time on tamoxifen and (b) regular citalopram use, defined as citalopram use overlapping tamoxifen use for 30% or more of the time on tamoxifen. We chose 30% as the overlap boundary to allow sufficient sample size in the regular citalopram subgroup, while also investigating a substantial period of SSRI and tamoxifen comedication.CovariatesWe defined the following set of covariates: (a) time period of breast cancer diagnosis (1985–1993, 1994–1996, and 1997–2001), (b) age at diagnosis (35–44 years, 45–54 years, 55–64 years, and 65–70 years), (c) menopausal status at diagnosis (premenopausal and postmenopausal), (d) county of residence at diagnosis (Aarhus, North Jutland, Viborg, and Ringkøbing), (e) UICC stage at diagnosis (stages I, II, and III), histological grade (grade I, II, III, and missing), surgery type (breast conserving surgery and mastectomy), and receipt of systemic adjuvant chemotherapy (yes and no), and (f) receipt of a prescription for another medication that is a CYP2D6 inhibitor or substrate, including other SSRI, while taking tamoxifen.Analytic strategyAll analyses were conducted within strata of the two groups (oestrogen-receptor positive and treated with tamoxifen or oestrogen-receptor negative and not treated with tamoxifen). We computed the frequency and proportion of cases and controls within categories of assigned protocol of tamoxifen duration, of citalopram use, of use of other CYP2D6 inhibitors or substrates, and of the covariates. We calculated the number of cases and controls ever receiving citalopram, the number of total prescriptions for citalopram summed over all cases or controls, and the range of the number of prescriptions for citalopram received by each individual case or control.We estimated the rate ratio associating citalopram prescription with breast cancer recurrence as the odds ratio (OR) in a conditional logistic regression including only citalopram use as the exposure variable and conditioned on the matched factors. By design, this ratio adjusts for confounding by the matched factors (Greenland, 2008). We examined whether the effect of citalopram use was modified by duration of tamoxifen therapy in a stratified analysis. Finally, we adjusted for residual confounding by the covariates that were not included in the matching by including them as independent variables in the conditional logistic regression. We retained in the final model any covariate that affected the log OR from the conditional logistic regression model associating citalopram use with breast cancer recurrence rate by more than 10% (Greenland, 1989). All estimates are accompanied by a 95% confidence interval (CI) calculated by the profile likelihood method. All analyses were performed using SAS version 9.ResultsTable 1 shows the frequency and proportion of cases and controls, within strata of group, in the categories of the covariates. About two-thirds of cases and controls in both groups were diagnosed with primary breast cancer during the period 1997–2001, and the majority was resident in Aarhus or North Jutland counties, because the prescription registries began first in these two counties. A large majority had mastectomy as their primary surgical intervention, which is consistent with the clinical practice pattern previously reported in this region during this time period (Ahern et al, 2008). Group I women (positive oestrogen-receptor expression and treated with tamoxifen) were more likely to be post-menopausal (87%) than were group II women (66%; negative oestrogen-receptor expression and not treated with tamoxifen). Group I women were also less likely to receive systemic adjuvant chemotherapy (11 and 13% of cases and controls, respectively) than were group II women (80 and 70% of cases and controls, respectively); reflecting the preference for hormonal therapy over systemic adjuvant chemotherapy in women whose tumours expressed the oestrogen receptor. Between 3 and 11% of cases and controls ever used citalopram while taking tamoxifen (group I) or during their follow-up period (group II).Table 2 depicts the pattern of SSRI prescriptions received by cases and controls. In both groups, SSRI prescriptions were primarily written for citalopram or its S-stereoisomer, escitalopram. For example, 17 of 23 group I cases (74%) ever prescribed an SSRI had at least one prescription for citalopram, accounting for 86% of the total number of prescriptions. Similarly, 22 of 30 group I controls (73%) ever prescribed an SSRI had at least one prescription for citalopram, accounting for 64% of their prescriptions. Sertraline accounted for the majority of the remaining prescriptions (11% of the total for cases and 23% for controls).Group I women who ever used citalopram while taking tamoxifen did not have a higher rate of breast cancer recurrence than women who never used citalopram while taking tamoxifen (Table 3; OR=0.79, 95% CI=0.40, 1.6). This OR was not substantially modified by duration of tamoxifen therapy (P=0.23 for test of homogeneity; data not shown). The approximately null effect persisted with adjustment for age category and ever/never use of another CYP2D6 inhibitor or SSRI (OR=0.85, 95% CI 0.42, 1.7). The effects were likewise approximately null within cumulative citalopram prescription categories (intermittent use OR=0.72, 95% CI 0.30, 1.7; regular use OR=1.1, 95% CI 0.37, 3.3). Citalopram use also had no substantial effect on recurrence in group II women (adjusted OR=0.78, 95% CI 0.17, 3.6), suggesting that citalopram does not directly affect the risk of breast cancer recurrence.DiscussionThe results of this study do not support the hypothesis that citalopram, taken concurrently with tamoxifen, reduces tamoxifen's protective effect against breast cancer recurrence in early-stage patients whose tumour cells express the oestrogen receptor.Our results extend the findings from an earlier study of 28 stage II and III breast cancer patients with recurrence and their matched controls at a single United States oncology centre, which also reported no substantial modification of tamoxifen effectiveness by concomitant use of SSRI inhibitors of CYP2D6 (Lehmann et al, 2004). These results may seem at odds with the strong biological rationale and in vivo evidence that support the hypothesis that CYP2D6 inhibition would reduce tamoxifen's prevention of breast cancer recurrence. It is possible, however, that SSRI medications could reduce the plasma concentration of tamoxifen's secondary metabolites without reducing its anti-tumorigenicity (Ponzone et al, 2004; Ratliff et al, 2004; Stearns et al, 2004). Tamoxifen doses as much as 20-fold lower than the typical US dose of 20 mg day−1 affect biomarkers of cardiovascular, bone, and tumour end points (Decensi et al, 1998, 2003), so the approximately three-fold reduction in the plasma concentration of tamoxifen's secondary metabolites associated with concomitant receipt of the SSRI paroxetine (Jin et al, 2005) may have little consequence.The key mechanistic question may be whether reduced concentrations of active tamoxifen metabolites result in substantially reduced occupancy of the oestrogen receptor. Dowsett and Haynes (2003) estimated that, in postmenopausal women on a daily dose of 20 mg tamoxifen, tamoxifen and its metabolites occupy 9994 of 10 000 oestrogen receptors. Replicating their calculation using the plasma concentrations of tamoxifen and its metabolites in women with no CYP2D6 variant allele (Jin et al, 2005), tamoxifen and its metabolites would occupy 9999 of 10 000 receptors in women not taking any SSRI and 9997 of 10 000 receptors in women taking the strong CYP2D6-inhibiting SSRI paroxetine. Steady-state concentrations of tamoxifen and its metabolites may be sufficient to manifest fully tamoxifen's antitumorigenic effect in postmenopausal women regardless of whether CYP2D6 inhibition reduces the concentration of some tamoxifen metabolites.Nonetheless, our results should be considered with the following limitations in mind. First, the majority of SSRI prescriptions in our study were for citalopram or its S-stereoisomer, both originally manufactured by Lundbeck, a company headquartered in Denmark. Citalopram is a modest inhibitor of CYP2D6 compared with some other SSRI medications (Jeppesen et al, 1996). These more potent inhibitors may reduce tamoxifen's protection against breast cancer recurrence, but their interaction with tamoxifen would not have been well measured by this study.Second, we have not collected genotype data to characterize functional CYP2D6 variants (Hayhurst et al, 2001) that affect the metabolism of tamoxifen (Jin et al, 2005). The combination of genotype and receipt of CYP2D6-inhibiting medications has been related to tamoxifen effectiveness in a previous study (Goetz et al, 2007). We do not, however, expect ever-receipt of citalopram while taking tamoxifen to be related to CYP2D6 genotype, as this genotype would be unknown to the patient and provider at the first citalopram prescription. This study's results therefore pertain to the usual clinical setting. In addition, CYP2D6 genotype is unlikely to cause citalopram prescription, or to share a common causal ancestor, so CYP2D6 genotype does not satisfy the requisite causal structure of a confounder (Greenland et al, 1999). It may be possible that CYP2D6 genotype is related to adherence to citalopram prescription or to long-term maintenance of the prescription, resulting from differences in the occurrence of adverse drug reactions in women with the different alleles. Such a relation could confound the association between breast cancer recurrence and duration of citalopram prescription while taking tamoxifen. Some non-randomized studies suggest such a relation between genotype and SSRI adherence (Rau et al, 2004; Zourková et al, 2007), whereas others suggest no such relation (Stedman et al, 2002; Gerstenberg et al, 2003; Roberts et al, 2004; Hedenmalm et al, 2006; Sugai et al, 2006; Suzuki et al, 2006). In the only randomized trial, CYP2D6 genotype was not related to either the occurrence of adverse events or to adherence to paroxetine prescription (Murphy et al, 2003). Paroxetine is the most potent CYP2D6 inhibitor of tamoxifen metabolism among the SSRI class (Jin et al, 2005). If CYP2D6 genotype does not affect receipt or adherence to SSRI prescription, then it cannot confound the association we have reported.Last, we do not know the indications for which citalopram was prescribed to the study participants, although ordinarily it would be prescribed primarily to treat depression. SSRI may also be prescribed to treat hot flushes (Stearns, 2006), but such prescriptions are rare in Danish breast cancer patients.Weighing against these limitations are the strengths of the data quality. This study relied upon the Danish Breast Cancer Cooperative Group's registry of breast cancer patients, which provides clinical trial quality data in a population-wide setting in the four Northern Danish Counties. For example, the positive predictive value of breast cancer recurrence recorded by the DBCG equaled 99.4% in a validation study (Hansen et al, 1997), showing that there are few false-positive recurrences registered in the DBCG. In addition, of 1888 local and distant recurrences identified by medical record review among 4455 breast cancer patients assigned to a DBCG protocol, 1813 (96%) were correctly registered as recurrences in the DBCG database, 74 (3.9%) were identified as breast cancer deaths, and only 1 (0.05%) was not identified as either a recurrence or breast cancer death.The prescription databases are generated by a computerised pharmacy accounting system that sends data to the Danish National Health Service, which refunds part of the costs associated with prescribed drugs. Given the direct connection between receipt of prescription medications and the pharmacy accounting system of the Danish National Health Service, we expect the prescription records to have excellent validity. The prescriptions from the four counties are merged into a research database at Aarhus University. In Denmark, antidepressants are available only at pharmacies and the patient must have a prescription from a medical doctor. Therefore, the county prescription databases are expected to have high sensitivity and specificity for ascertainment of citalopram prescriptions in the source population. Furthermore, because the prescription records antecede the date of breast cancer recurrence, they are a prospective data source presumably immune to differential classification bias (Rothman et al, 2008).Despite these advantages, the study yielded only 17 cases of breast cancer recurrence among tamoxifen-treated women who had used citalopram while taking tamoxifen. The study was designed with 80% power to detect an OR of 1.6, and ultimately had 90% power to detect an OR of 2.3.The results presented herein are, nonetheless, important and timely. A United States Food and Drug Administration advisory committee recently recommended relabelling tamoxifen with information on gene–drug and drug–drug interactions mediated by CYP2D6 (American Cancer Society, 2007). Furthermore, the current practice guidelines of the United States National Comprehensive Cancer Network note that some SSRI reduce the formation of active tamoxifen metabolites, that citalopram and venlaflaxine appear to have minimal impact on tamoxifen metabolism, and that ‘the clinical impact of these observations is not known’ (National Comprehensive Cancer Network, 2008). Breast cancer patients taking tamoxifen and their physicians may therefore be concerned about SSRI comedication, even when antidepressants are strongly indicated. Our results suggest that citalopram prescription does not reduce tamoxifen's prevention of breast cancer recurrence.\n\nREFERENCES:\n1. Ahern TP, Larsson H, Garne JP, Cronin-Fenton DP, Sørensen HT, Lash TL (2008) Trends in breast-conserving surgery in Denmark, 1982–2002. Eur J Epidemiol\n23: 109–11417987392\n2. American Cancer Society (2007) Tamoxifen: some women don't get full benefit. American Cancer Society News Center http://www.cancer.org/docroot/NWS/content/NWS_1_1x_Tamoxifen_Some_Women_Dont_Get_Full_Benefit.asp. Last accessed 13 June 2007\n3. Andersen KW, Mouridsen HT (1988) Danish Breast Cancer Cooperative Group (DBCG). A description of the register of the nation-wide programme for primary breast cancer. Acta Oncologica\n27: 627–6473064772\n4. Coezy E, Borgna JL, Rochefort H (1982) Tamoxifen and metabolites in MCF-7 cells: correlation between binding to estrogen receptor and inhibition of cell growth. Cancer Res\n42: 317–3237053859\n5. Coller JK, Krebsfaenger N, Klein K, Endrizzi K, Wolbold R, Lang T, Nüssler A, Neuhaus P, Zanger UM, Eichelbaum M, Mürdter TE (2002) The influence of CYP2B6 and CYP2D6 genotypes on the formation of the potent antioestrogen Z-4-hydroxy-tamoxifen in human liver. Br J Clin Pharmacol\n54: 157–16712207635\n6. Decensi A, Bonanni B, Guerrieri-Gonzaga A, Gandini S, Robertson C, Johansson H, Travaglini R, Sandri MT, Tessadrelli A, Farante G, Salinaro F, Bettega D, Barreca A, Boyle P, Costa A, Veronesi U (1998) Biologic activity of tamoxifen at low doses in healthy women. J Natl Cancer Inst\n90: 1461–14679776411\n7. Decensi A, Robertson C, Viale G, Pigatto F, Johansson H, Kisanga ER, Veronesi P, Torrisi R, Cazzaniga M, Mora S, Sandri MT, Pelosi G, Luini A, Goldhirsch A, Lien EA, Veronesi U (2003) A randomized trial of low-dose tamoxifen on breast cancer proliferation and blood estrogenic markers. J Natl Cancer Inst\n95: 779–79012783932\n8. Dowsett M, Haynes BP (2003) Hormonal effects of aromatase inhibitors: focus on premenopausal effects and interaction with tamoxifen. J Steroid Biochem Mol Biol\n86: 255–26314623519\n9. Early Breast Cancer Trialists' Collaborative Group (2005) Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomized trials. Lancet\n365: 1687–171715894097\n10. Gaist D, Sørensen HT, Hallas J (1997) The Danish prescription registries. Dan Med Bull\n44: 445–4489377907\n11. Gerstenberg G, Aoshima T, Fukasawa T, Yoshida K, Takahashi H, Higuchi H, Murata Y, Shimoyama R, Ohkubo T, Shimizu T, Otani K (2003) Relationship between clinical effects of fluvoxamine and the steady-state plasma concentrations of fluvoxamine and its major metabolite fluvoxamino acid in Japanese depressed patients. Psychopharmacol\n167: 443–448\n12. Goetz MP, Knox SK, Suman VJ, Rae JM, Safgren SL, Ames MM, Visscher DW, Reynolds C, Couch FJ, Lingle WL, Weinshilboum RM, Fritcher EG, Nibbe AM, Desta Z, Nguyen A, Flockhart DA, Perez EA, Ingle JN (2007) The impact of cytochrome P-450 2D6 metabolism in women receiving adjuvant tamoxifen. Breast Cancer Res Treat\n101: 113–12117115111\n13. Greenland S (1989) Modeling and variable selection in epidemiologic analysis. Am J Public Health\n79: 340–3492916724\n14. Greenland S (2008) Introduction to regression modeling. In Modern Epidemiology, 3rd edn, Rothman KJ, Greenland S, Lash TL (eds) pp 418–458. Philadelphia: Lippincott Williams & Wilkins\n15. Greenland S, Pearl J, Robins JM (1999) Causal diagrams for epidemiologic research. Epidemiology\n10: 37–489888278\n16. Hansen DG, Søndergaard J, Vach W, Gram LF, Rosholm JU, Kragstrup J (2003) Antidepressant drug use in general practice: inter-practice variation and association with practice characteristics. Eur J Clin Pharmacol\n59: 143–14912721774\n17. Hansen PS, Andersen E, Andersen KW, Mouridsen HT (1997) Quality control of end results in a Danish adjuvant breast cancer multi-center study. Acta Oncologica\n36: 711–7149490088\n18. Hayhurst GP, Harlow J, Chowdry J, Gross E, Hilton E, Lennard MS, Tucker GT, Ellis SW (2001) Influence of phenylalanine-481 subsitutions on the catalytic activity of cytochrome P-450 2D6. Biochem J\n355: 373–37911284724\n19. Hedenmalm K, Güzey C, Dahl ML, Yue QY, Spigset O (2006) Risk factors for extrapyramidal symptoms during treatment with selective serotonin reuptake inhibitors, including cytochrome P-450 enzyme, and serotonin and dopamine transporter and receptor polymorphisms. J Clin Psychopharmacol\n26: 192–19716633151\n20. Jensen AR, Ewertz M, Cold S, Storm HH, Overgaard J (2003) Time trends and regional differences in registration, stage distribution, surgical management, and survival of breast cancer in Denmark. Eur J Cancer\n39: 1783–179312888375\n21. Jeppesen U, Gram LF, Vistisen K, Loft S, Poulsen HE, Brøsen K (1996) Dose-dependent inhibition of CYP1A2, CYP2C19 and CYP2D6 by citalopram, fluoxetine, fluvoxamine and paroxetine. Eur J Clin Pharmacol\n51: 73–788880055\n22. Jin Y, Desta Z, Stearns V, Ward B, Ho H, Lee KH, Skaar T, Storniolo AM, Li L, Araba A, Blanchard R, Nguyen A, Ullmer L, Hayden J, Lemler S, Weinshilboum RM, Rae JM, Hayes DF, Flockhart DA (2005) CYP2D6 genotype, antidepressant use, and tamoxifen metabolism during adjuvant breast cancer treatment. J Natl Cancer Inst\n97: 30–3915632378\n23. Jordan VC, Dowse LJ (1976) Tamoxifen as an anti-tumour agent: effect on oestrogen binding. J Endocrinol\n68: 297–303176295\n24. Lehmann D, Nelsen J, Ramanath V, Newman N, Duggan D, Smith A (2004) Lack of attenuation in the antitumor effect of tamoxifen by chronic CYP isoform inhibition. J Clin Pharmacol\n44: 861–86515286089\n25. Lim YC, Desta Z, Flockhart DA, Skaar TC (2005) Endoxifen (4-hydroxy-N-desmethyl-tamoxifen) has anti-estrogenic effects in breast cancer cells with potency similar to 4-hydroxy-tamoxifen. Cancer Chemother Pharmacol\n55: 471–47815685451\n26. Malet C, Gompel A, Spritzer P, Bricout N, Yaneva H, Mowszowicz I, Kuttenn F, Mauvais-Jarvis P (1988) Tamoxifen and hydroxyl-tamoxifen isomers vs estradiol effects on normal human breast cells in culture. Cancer Res\n48: 7193–71993056610\n27. Massie MJ (2004) Prevalence of depression in patients with cancer. J Natl Cancer Inst Monogr\n32: 57–71\n28. Murphy Jr GM, Kremer C, Rodrigues HE, Schatzberg AF (2003) Pharmacogenetics of antidepressant medication intolerance. Am J Psychiatry\n160: 1830–183514514498\n29. National Comprehensive Cancer Network (2008) NCCN Clinical Practice Guidelines in Oncology: Breast Cancer v.2.2008. pp. 37 http://www.nccn.org/professionals/physician_gls/PDF/breast.pdf\n30. Ponzone R, Biglia N, Sismondi P (2004) Re: Active tamoxifen metabolite plasma concentrations after coadministration of tamoxifen and the selective serotonin reuptake inhibitor paroxetine. Letter. J Natl Cancer Inst\n96: 883–884\n31. Ratliff B, Dietze EC, Bean GR, Moore C, Wanko S, Seewaldt VL (2004) Re: Active tamoxifen metabolite plasma concentrations after coadministration of tamoxifen and the selective serotonin reuptake inhibitor paroxetine. Letter. J Natl Cancer Inst\n96: 883\n32. Rau T, Wohlleben G, Wuttke H, Thuerauf N, Lunkenheimer J, Lanczik M, Eschenhagen T (2004) CYP2D6 genotype: impact on adverse effects and nonresponse during treatment with antidepressants-a pilot study. Clin Pharmacol Ther\n75: 386–39315116051\n33. Roberts RL, Mulder RT, Joyce PR, Luty SE, Kennedy MA (2004) No evidence of increased adverse drug reactions in cytochrome P450 CYP2D6 poor metabolizers treated with fluoxetine or nortriptyline. Hum Psychopharmacol\n19: 17–2314716707\n34. Rothman KJ, Greenland S, Lash TL (2008) Types of epidemiologic studies. In Modern Epidemiology, 3rd edn, Rothman KJ, Greenland S, Lash TL (eds) pp 95–97; Philadelphia: Lippincott, Williams & Wilkins\n35. Stearns V (2006) Serotonergic agents as an alternative to hormonal therapy for the treatment of menopausal vasomotor symptoms. Treat Endocrinol\n5: 83–8716542048\n36. Stearns V, Johnson MD, Rae JM, Morocho A, Novielli A, Bhargava P, Hayes DF, Desta Z, Flockhart DA (2003) Active tamoxifen metabolite plasma concentrations after coadministration of tamoxifen and the selective serotonin reuptake inhibitor paroxetine. J Natl Cancer Inst\n95: 1758–176414652237\n37. Stearns V, Johnson MD, Rae JM, Novielli A, Bhargava P, Hayes DF, Desta A, Flockhart DA (2004) Re: active tamoxifen metabolite plasma concentrations after coadministration of tamoxifen and the selective serotonin reuptake inhibitor paroxetine. Letter. J Natl Cancer Inst\n96: 884–88515173277\n38. Stedman CA, Begg EJ, Kennedy MA, Roberts R, Wilkinson TJ (2002) Cytochrome P450 2D6 genotype does not predict SSRI (fluoxetine or paroxetine) induced hyponatraemia. Hum Psychopharmacol\n17: 187–19012404686\n39. Sugai T, Suzuki Y, Sawamura K, Fukui N, Inoue Y, Someya T (2006) The effect of 5-hydroxytryptamine 3A and 3B receptor genes on nausea induced by paroxetine. Pharmacogenomics J\n6: 351–35616534507\n40. Suzuki Y, Sawamura K, Someya T (2006) Polymorphisms in the 5-hydroxytryptamine 2A receptor and cytochromeP4502D6 genes synergistically predict fluvoxamine-induced side effects in Japanese depressed patients. Neuropsychopharmacology\n31: 825–83116205777\n41. UICC (1997) TNM Classification of Malignant Tumours, 5th edn. Switzerland: Springer\n42. WHO Collaborating Centre for Drug Statistics Methodology (2007) About the centre. http://www.whocc.no/atcddd/. Last accessed 31 May 2007\n43. Zanger UM, Raimundo S, Eichelbaum M (2004) Cytochrome P-450 2D6: overview and update on pharmacology, genetics, biochemistry. Naunyn-Schmiedeberg's Arch Pharmacol\n369: 23–3714618296\n44. Zourková A, Cesková E, Hadasová E, Ravcuková B (2007) Links among paroxetine-induced sexual dysfunctions, gender, and CYP2D6 activity. Sex Marital Ther\n33: 343–355"
+}

batch_3/PMC2528011.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2528011",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2528011\nAUTHORS: Patrick Gérardin, Vanina Guernier, Joëlle Perrau, Adrian Fianu, Karin Le Roux, Philippe Grivard, Alain Michault, Xavier de Lamballerie, Antoine Flahault, François Favier\n\nABSTRACT:\nBackgroundChikungunya virus (CHIKV) caused a major two-wave seventeen-month-long outbreak in La Réunion Island in 2005–2006. The aim of this study was to refine clinical estimates provided by a regional surveillance-system using a two-stage serological assessment as gold standard.MethodsTwo serosurveys were implemented: first, a rapid survey using stored sera of pregnant women, in order to assess the attack rate at the epidemic upsurge (s1, February 2006; n = 888); second, a population-based survey among a random sample of the community, to assess the herd immunity in the post-epidemic era (s2, October 2006; n = 2442). Sera were screened for anti-CHIKV specific antibodies (IgM and IgG in s1, IgG only in s2) using enzyme-linked immunosorbent assays. Seroprevalence rates were compared to clinical estimates of attack rates.ResultsIn s1, 18.2% of the pregnant women were tested positive for CHIKV specific antibodies (13.8% for both IgM and IgG, 4.3% for IgM, 0.1% for IgG only) which provided a congruent estimate with the 16.5% attack rate calculated from the surveillance-system. In s2, the seroprevalence in community was estimated to 38.2% (95% CI, 35.9 to 40.6%). Extrapolations of seroprevalence rates led to estimate, at 143,000 and at 300,000 (95% CI, 283,000 to 320,000), the number of people infected in s1 and in s2, respectively. In comparison, the surveillance-system estimated at 130,000 and 266,000 the number of people infected for the same periods.ConclusionA rapid serosurvey in pregnant women can be helpful to assess the attack rate when large seroprevalence studies cannot be done. On the other hand, a population-based serosurvey is useful to refine the estimate when clinical diagnosis underestimates it. Our findings give valuable insights to assess the herd immunity along the course of epidemics.\n\nBODY:\nBackgroundChikungunya fever is an arbovirosis caused by Chikungunya virus (CHIKV), a mosquito-transmitted alphavirus belonging to the Togaviridae family [1,2]. CHIKV was first isolated in 1952, during a Tanzanian outbreak [3]. It circulated in Africa and Asia, where periodic outbreaks were described in the past 50 years. In some areas, attack rates had reached 80 to 90% [1,2]. Between February 2005 and August 2006, a large Chikungunya fever outbreak swept the Indian Ocean islands [4,5], including La Réunion Island since April 2005, an overseas French department of 787,836 inhabitants (Figure 1). The mosquito specie involved in La Réunion outbreak was Aedes (A.) albopictus [6]. Most CHIKV infections were symptomatic [7] and characterized by a dengue-like illness of sudden onset combining high fever, poly-arthralgia, myalgia, headache, asthenia and rash [8,9].Figure 1Map of La Réunion Island. The territory is divided into four regions: north bounded by orange and red lines, west bounded by orange, light green, dark green and red lines, south bounded by dark green and red lines, east by red lines. For each municipality, the ratio of laboratories which participated to the survey on pregnant woman and the related amount of the sera collected (n = x) are listed in parentheses.In La Réunion, the epidemic pattern was monitored through a regional surveillance-system managed by the Cellule Interrégionale d'Epidémiologie (CIRE) based on \"suspected cases\", defined as subjects with a sudden fever (T > 38.5°C) and incapacitating arthralgia [10,11]. This surveillance-system relied on self-reports, emergency stays, physician declarations, biology laboratories activity, and active case-finding around the cases reported by a sentinel physician network [11]. At the beginning of the outbreak it consisted in an active and retrospective case detection around the cases declared, and then, when the incidence sharply increased (by December 2005), in an estimation of the cases obtained from reports of a sentinel network [12].Before the explosion of the epidemic in mid-January 2006, a herald wave occurred during the previous rainy season; between March and July 2005, and led the CIRE to record about 3,000 suspected cases of Chikungunya [13]. Later on and until December 2005, low case rates were recorded without interruption. An exponential increase of the cases reported was observed in late December 2005, and January 2006 with a peak in February 2006 [11] (Figure 2). On February 15th 2006, the CIRE estimated 157,000 suspected cases of Chikungunya, i.e. a prevalence rate of 20.3%. On July 5th 2006, the CIRE estimated the burden more than 266,000, i.e. a prevalence rate of 34.3% [14].Figure 2Number of weekly incident cases of Chikungunya, La Réunion Island, March 28th, 2005 – April 16th, 2006 (n = 244,000). Reported by the active case-finding system between weeks 9 and 50, 2005 or estimated from the sentinel physician network between week 51 of 2005 and week 15, 2006. Published by Renault P, et al. in Am J Trop Med Hyg, 2007, 77: 727–731 [11], and reprinted with the kind permission of the American Society of Tropical Medicine and Hygiene (Atlanta, USA). \"Survey 1\" corresponds to the rapid serological survey on pregnant women (January 15th 2006 to February 15th 2006); \"Survey 2\" corresponds to the population-based SEROCHIK survey (August 17th to October 20th 2006).The purpose of the study was to refine the estimates of attack rates provided by the surveillance-system for the population of La Réunion Island at two critical times of the 2005–2006 outbreak. That is why we conducted two serosurveys, the first using stored sera of pregnant women during the epidemic upsurge aimed at assessing the extent of the outbreak, the second using a random sample of the population aimed at giving a precise idea of the herd immunity in the post-epidemic era.MethodsRapid survey on pregnant women, epidemic phaseWe gathered sera of pregnant women yet available in outpatient laboratories from a mandatory monthly serological screening for congenital toxoplasmosis. The sera, collected between January 15th and February 15th 2006, were neither directly nor indirectly nominative, and could only be identified by a unique code number. All of the 46 biological labs of La Réunion Island were invited to participate to the survey. Out of these, the 28 participating ones served the entire territory (Figure 1). However, only 19 labs provided valid sera which led to a non-representative amount of 888 valid sera, taken out of the 3888 sera collected during the study period (389 in the north, 305 in the south, 174 in the west, 20 in the east). For this study, designed to inform without delay public health authorities on the extent of the outbreak, a selection bias related to the absence of randomisation of labs was tolerated. Nevertheless, as daily routine sampling of pregnant women was not dedicated to a precise laboratory, it is reasonable to think that the repartition bias was not significant. Statistical analysis was performed using SAS software version 8 (SAS Institute, Cary, NC).Population-based survey, post-epidemic phaseA cross-sectional study, the SEROCHIK survey, was conducted between August 17th and October 20th 2006 from a random sample of the community of Reunion Island [7]. At the first sampling stage, the French National Institute for Statistics and Economical Studies (INSEE) randomly selected 3032 households after stratification on age, gender, the geographical area, municipality size, and type of habitat. The geographical area of habitat was defined according to the regional administrative boundary into four micro-regions (Figure 1). The municipality size was divided in ≤ or > 10,000 inhabitants. The type of habitat was categorized into individual or collective housing (multifamily ≤ 20 or > 20 housings). At the second stage, a Kish method was used to randomly choose one person for each selected household [13].Of the 3032 households randomly selected by INSEE, the sampling plan led to a set of 2442 eligible subjects (after exclusion of absents, persons with invalid address or who refused to participate) which was recovered by INSEE on age, gender, geographical area, and type of habitat.The study was approved by the ethical committee for studies with human subjects (CPP) of Bordeaux and the National Commission for Informatics and Liberty (CNIL). All participants provided their informed consent to answer the questionnaire and for collection of blood on filter paper.Statistical analysis was done by accounting for the sampling design, and was performed using Stata software (College Station, Texas). The population-based seroprevalence was compared to CIRE clinical estimates using a Chi square test. A P-value < 0.05 was considered significant. The population size used for the calculation of incidence was 787,836 inhabitants (INSEE, April 2006).Detection of chikungunya infectionsFor the rapid survey in pregnant women, 100 microliters of stored sera already available in outpatient laboratories were used. In the SEROCHIK survey, for each person consenting to a fingertip prick, a drop of blood was deposited onto Whatman no.1 filter paper [15]. For both studies, anti-CHIKV specific antibodies were screened using the same enzyme-linked immunosorbent assay (ELISA) and a CHIKV antigen produced by the Centre National de Référence pour les Arbovirus (CNR, Lyon, France) [15]. For the rapid survey, the ELISA was performed at the CNR whereas for the SEROCHIK survey, it was done using the Groupe Hospitalier Sud – Réunion (GHSR) laboratory facilities. Both IgM and IgG anti-CHIKV specific antibodies were screened in sera from pregnant women, whereas only IgG anti-CHIKV specific antibodies were screened in the community. In parallel with the SEROCHIK survey, the ability of the prick-method to discriminate the serological status was validated in an independent sample (Fianu et al, unpublished data).ResultsThe Chikungunya serological status (positive/negative) stratified for each study is given in Table 1.Table 1Chikungunya serological status during the epidemic upsurge phase (rapid survey) and the post-epidemic era (population-based survey), Reunion Island outbreak, 2005 – 2006.StudyNegative serologyPositive serologyTotalPregnant women (rapid survey)726(81.8)162(18.2) †888Population (population – based survey)1475(60.4)967(39.6) ‡2442Data are numbers of persons questioned and (row) percentages.† IgM and IgG anti-ChikV specific antibodies tested.‡ IgG anti-ChikV specific antibodies tested.Survey on pregnant women, epidemic phaseDuring the studied period (epidemic phase, Figure 2), 162 pregnant women (out of 888 enrolled, i.e. 18.2%) tested positive for antibodies to CHIKV (IgM and/or IgG). There was serological evidence for a recent Chikungunya infection, as 123 subjects (13.8%) showed both IgM and IgG, and 38 (4.3%) had IgM in the absence of IgG. Isolated positive IgG were detected in only one case (0.1%).Application of the Chikungunya prevalence obtained from the pregnant women group to the community resulted in a rough estimate of 143,386 (787,836 × 0.182) infected cases in La Réunion Island by February 15th 2006 (Figure 3).Figure 3Comparison of monthly suspected, self-reported and confirmed cases of Chikungunya in La Réunion Island between April 2005 and October 2006. The number of suspected cases recorded weekly by the CIRE (left scale) is compared to the number of cases identified in the population-based SEROCHIK survey (right scale). For the serosurvey, both self-reports (all subjects who have declared that they have been infected, without taking into account serology results) and confirmed self-reports (with a positive serology) are noted. We refer to the date of first clinical signs declared by the subjects during the survey conducted between August 17th 2006 and October 20th 2006. \"Suspected cases\" are defined as cases with a sudden onset of fever with temperature > 38.5°C and incapacitating arthralgia.Population-based survey, post-epidemic phaseThe seroprevalence result in the sample questioned is presented in Table 2. The overall seroprevalence for the community after recovery by the sampling plan weight (data not shown) was estimated at 38.2% (95% CI, 35.9 to 40.6%), i.e., about 300,000 people infected (95% CI, 283,000 to 320,000) which was significantly different that the 34.3% of people infected reported by CIRE (Chi square test: 16.4, P < 0.001). The rate of inapparent cases combined to atypical cases (i.e., all subjects who were not aware of their infection but were tested serologically positive to CHIKV) was calculated to be 5.0% (95% CI 3.9 – 6.2%), and the rate of false positive (i.e., subjects who declared that they were infected but had negative serology to CHIKV) was estimated to be 4.5%% (95% CI 3.7 – 5.6%). Out of the 967 serologically positive persons, 162 (16.7%) reported no symptoms.Table 2Chikungunya clinical status × serological status in the community.Population-based survey, Reunion Island outbreak, August – October 2006 (post-epidemic era)Chikungunya declaredNegative serologyPositive serologyTotal\"No\"[82.5]1217[12.0]116[54.5]1333(91.3)(8.7)(100)Yes\"[8.0]118[83.2]805[37.9]923(12.8)(87.2)(100)\"I don't know\"[9.5]140[4.8]46[7.6]186(75.3)(24.7)(100)Total[100]1475[100]967[100]2442(60.4)(39.6)(100)Data are numbers of persons questioned and (row) or [column] percentages.As the CIRE only considered suspected cases, i.e. subjects with fever above 38.5°C with incapacitating arthralgia, we also took into account the symptoms reported in the questionnaire of the SEROCHIK population-based survey.The ten most frequent symptoms self-reported by the subjects who declared a Chikungunya are presented with their positive predictive value (PPV) in order of decreasing frequency in Table 3. People who declared a Chikungunya with sudden fever and incapacitating arthralgia represented 88.5% of all people who declared the disease and 87.0% of all positive serology (sensitivity and PPV for both signs: 88.5% and 87.0%, respectively).Table 3Chikungunya self-reported symptoms × serological status in the community.Population-based survey, Reunion Island outbreak, August – October 2006 (post-epidemic era)SymptomsNegative serologyPositive serology *TotalArthralgia111(12.6)770(87.4)881Fever92(10.9)753(89.1)845Asthenia89(12.8)605(87.2)694Headache69(11.2)545(88.8)614Myalgia68(12.0)497(88.0)565Rash41(7.7)493(92.3)534Pruritus36(7.9)421(92.1)457Vomiting31(14.2)188(85.8)219Diarrhea21(13.5)135(86.5)156Depression19(13.2)125(86.8)144Fever and arthralgia89(13.0)594(87.0)683Total118(100)805(100)923Data are numbers of persons questioned and (row) percentages. Positive Predictive value (PPV) *DiscussionIn the current paper, we demonstrate the usefulness of two different epidemiological methods to assess the burden of a Chikungunya outbreak at two critical times of its evolution.The rapid seroprevalence survey conducted at the peak of La Réunion Island epidemic on sera from pregnant women provided an 18.2% seroprevalence rate (or a rough estimate of 143,000 people infected) by February 15th, 2006. This result was obtained at a very low cost, at a time when the attack rate of the infection in the population and the herd immunity were unknown. Since 99% of the 162 tested sera harbored IgM anti-CHIKV antibodies and only one IgG anti-CHIKV antibodies only, it excluded a previous (recent) significant circulation of the virus in the island and thus suggested that the outbreak emerged into a naïve population. This result is in agreement with the 20% prevalence rate calculated for all parturient women delivered at the GHSR maternity in mid-February 2006 [16], and with the magnitude of 26% reported in pregnant women by April 2006 in Mayotte [17].The rough estimate by the rapid serosurvey in pregnant women is slightly higher but of the same magnitude than the 130,000 cumulative number of suspected cases deducted from the CIRE at the same time (attack rate of 16.5%) [18,19]. It is therefore noteworthy that the rate observed in a targeted population of pregnant woman gives a valuable insight upon the magnitude of the attack rate in the community, and beyond, of the herd immunity. This congruent result with the CIRE data suggests that in February 2006 pregnant women yet behaved like everyone and that their level of exposure to CHIKV was not different to that observed for anybody else. In other words, the message aimed at crystallising the pregnant woman as vulnerable person to the threat of Chikungunya [20] and the measures aimed at reducing her exposure [21], e.g., wear of long clothing, free distribution of insecticide-treated nets and repellents, soon implemented in the maternities of La Réunion, were not effective against A. albopictus bites, a vector which was proven to exercise a diurnal activity [6].The slight difference between pregnant women and community may account for a selection phenomenon, the pregnant women being not representative of the community. However, the prevalence was higher in pregnancy which argues against a significant selection bias, since more than 50% of serum collections came from south and west labs at a time when transmission was still predominant in theses micro-regions, leading to an unexpected geographical adjustment on the transmission level.The population-based survey provided a 38.2% prevalence rate in the post-epidemic era, i.e., about 300,000 people infected, by October 20, 2006 [7]. At the same time, the CIRE data, published on the Institut national de Veille Sanitaire website, estimated at 266,000 (34.3%) the number of people infected [14]. Thus, the seroprevalence in the community was slightly higher than the 34.3% estimated for suspected cases (P < 0.001), but of the same magnitude. This difference might correspond to undeclared cases to the CIRE, i.e., (1) the inapparent cases (5.0%), although these would be compensated by an approximately equal proportion of false positives (4.5%); (2) patients who did not consult and performed auto-medication; (3) patients who did not match the clinical criteria for \"suspected cases\", i.e., sudden fever > 38.5 C° and incapacitating arthralgia. Based on laboratory confirmations of atypical presentations, Renault et al. calculated that approximately 3% of patients did not fulfil this definition [11]. Importantly, the SEROCHIK survey showed that 25.8% of CHIKV-infected subjects did not declare fever combined to arthralgia. This significant discrepancy may result both from an information bias due to the structure of the questionnaire (subjects who were not aware of their infection did not answer to questions about clinical signs), or a memory bias due to the time interval between the SEROCHIK survey and the onset of symptoms (2 to 15 months) that could preclude mild cases to remind their symptoms. However, the declaration-based surveillance system based on suspected cases may have also underestimated the attack rate at the epidemic phase. The difference was particularly notable from April to December 2005 (Figure 3), when the incidence was less than 500 new cases per week (Figure 2) and the surveillance relied upon active and retrospective case detection around the cases declared. Ditto, it was verified from June 2006 (Figure 3), when the incidence dropped dramatically shortly before the epidemic stopped in August (Figure 2). It could be explained by a lower PPV for each symptoms (< 85 to 92%) and for the clinical definition (< 87%), as the transmission was low (Table 3). Indeed, it is well known that when the incidence of a communicable disease is low, its contribution to the clinical forms that can evoke it decreases, with a consequent decline of the PPV for clinical signs to identify the disease [22]. Moreover, the possibility of concomitant circulation of other infections, such as Influenza or Dengue [23], may have challenged the diagnosis of Chikungunya [24], especially between April and December 2005, or from June 2006.Another important aspect of Chikungunya disease disclosed in the course of La Réunion outbreak is the low proportion of inapparent forms (16%), in comparison with those usually observed for other arbovirosis, such as Dengue Fever (> 50%) [25,26] or West-Nile virus infection (> 70%) [27,28].Finally, it is noteworthy that the 38.2% seroprevalence rate observed in the post-epidemic phase [7] gives a valuable insight upon the herd immunity and a clearer picture of susceptible people who could be infected in the future (62.8%). Importantly, the seroprevalence observed in La Réunion Island was far inferior to those reported recently from the Kenyan island of Lamu (75%) [29] and the Grande Comoro Island (63%) [30], two areas where CHIKV emerged before to reach La Réunion [4,5].Several hypotheses may explain this discrepancy in prevalence rates: 1°) Kenyan and Comorian climates are less prone to seasonal variations and therefore more conducive to a sustainable transmission; 2°) A. aegypti, the classical vector of Chikungunya, involved in Kenyan and Comorian outbreaks, keeps a better capability to spread the disease in domestic environment than the less anthropophilic A. albopictus; 3°) for the same reason, the density of susceptible hosts would be less important in peridomestic environment in La Réunion, than in Kenyan and Comorian homes invaded by highly anthropophilic A. aegypti; 4°) in La Réunion, floods brought by the cyclone Diwa drove away larvae of A. albopictus from gullies and hastened the decrease in transmission from early March 2006; 5°) effective vector control measures combining eradication of breeding sites, adulticide and larvicide treatments contributed to limit the density of vectors throughout the Reunion outbreak; 6°) In La Réunion, the herd immunity was gained more rapidly in the littoral plains (where most of the population lives in contact with vectors) which reduced transmission as the entry in the dry austral winter. Indeed, some micro-geographical differences in prevalence rates would have been not detected by the SEROCHIK survey (whose sampling plan was aimed at discriminating between micro-regions but not within), and thereby prevalence rates in littoral plains would have far exceeded the overall 38.2% rate, which would have led to the premature decline of the epidemic due to effective herd immunity.Prevalence rates of 60 to 70% were necessary to delay resurgences to 20 to 30 years in areas where CHIKV had circulated before [31]. According to a recent mathematical model [32], it was concluded that in the best-fitting case (reproductive number of 3.7), the attack rate would have been of 73% which suggests that, with only 38.2% of people infected, a re-emergence in La Réunion Island cannot be excluded for the next years, as long as viral strains circulate in the region. However, the scenario observed in La Réunion seems thwart this mathematical prediction, because so far, no case of Chikungunya has been scientifically confirmed since August 2006. This could be explained at municipality level by heterogeneity of reproductive numbers that would have been very sensitive to local interventions in vector control [33].ConclusionCongruent estimates of Chikungunya attack rate were observed at the upsurge of La Réunion Island outbreak, either using clinical declaration of suspected cases (16.5%), or a rapid serosurvey in pregnant women (18.2%). In contrast, a discrepancy was observed in the post-epidemic era, when clinical diagnosis underestimated the attack rate (34.3%) in comparison to seroprevalence estimate (38.2%; 95% CI, 35.9 to 40.6%). Thus, a rapid serosurvey in a targeted population can be helpful to assess the extent of epidemics at time of emergency when large seroprevalence studies cannot be done. Beyond this indication, our findings suggest that prospective real time surveillance of attack rates in pregnant women would serve as a good model for population monitoring in the event of Chikungunya outbreaks. However, although it may fail to detect micro-geographical differences in prevalence rates at municipality level, a population-based serosurvey can still be useful to refine the clinical estimates and to assess more precisely the herd immunity. Moreover, only a representative survey can bring an overview on risk factors and other conditions facilitating the transmission. Finally, this work speaks to the usefulness of serosurveys for the quantification of epidemics, as to cost-containment in public health. It also provides valuables clues for monitoring epidemics in high and low-incomes countries.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsFF, ANF and XDL conceived and designed the experiments. KLR, PHG and AM performed the experiments including serology assays. JP designed the sampling plan for the SEROCHIK survey. JP, ADF and PAG analysed the data. VG wrote the initial draft and PAG revised the manuscript, which was extensively reviewed and approved by all authors.Pre-publication historyThe pre-publication history for this paper can be accessed here:\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2528014.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2528014",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2528014\nAUTHORS: Mei Tian, Ya-Zhou Cui, Guan-Hua Song, Mei-Juan Zong, Xiao-Yan Zhou, Yu Chen, Jin-Xiang Han\n\nABSTRACT:\nBackgroundThere is an urgent need to discover more sensitive and specific biomarkers to improve early diagnosis and screen high-risk patients for pancreatic ductal adenocarcinoma (PDAC). Pancreatic juice is an ideal specimen for PDAC biomarkers discovery, because it is an exceptionally rich source of proteins released from pancreatic cancer cells.MethodsTo identify novel potential biomarkers for PDAC from pancreatic juice, we carried out difference gel electrophoresis (DIGE) and tandem mass spectrometry (MS/MS) to compare the pancreatic juice profiling from 9 PDAC patients and 9 cancer-free controls. Of the identified differently expressed proteins, three up-regulated proteins in pancreatic cancer juice, matrix metalloproteinase-9 (MMP-9), oncogene DJ1 (DJ-1) and alpha-1B-glycoprotein precursor (A1BG), were selected for validation by Western blot and immunohistochemistry. Serum MMP-9 levels were also detected by enzyme linked immunosorbent assay (ELISA).ResultsFourteen proteins were up-regulated and ten proteins were down-regulated in cancerous pancreatic juice compared with cancer-free controls. Increased MMP-9, DJ-1 and A1BG expression in cancerous pancreatic juice were confirmed by Western blot. Immunohistochemical study showed MMP-9, DJ-1 and A1BG positively expressed in 82.4%, 72.5% and 86.3% of pancreatic cancer tissues, significantly higher than that in normal pancreas tissues. Up-regulation of DJ-1 was associated with better differentiation (p < 0.05). Serum MMP-9 levels were significantly higher in PDAC (255.14 ng/ml) than those in chronic pancreatitis (210.22 ng/ml, p = 0.009) and healthy control (203.77 ng/ml, p = 0.027).ConclusionThe present proteome analysis revealed MMP-9, DJ-1 and A1BG proteins as elevated in pancreatic juice from PDAC, which suggest their further utility in PDAC diagnosis and screening. This is the first time A1BG was identified as a potential biomarker in pancreatic cancer associated samples. The measurement of serum MMP-9 might be clinically useful for PDAC diagnosis.\n\nBODY:\nBackgroundPancreatic ductal adenocarcinoma (PDAC) is the fourth or fifth most common cause of cancer-related mortality worldwide. Because of late presentation and rapid aggressiveness, most PDAC cases are diagnosed at late stage, and its prognosis is accordingly poor. So detection of PDAC at an early disease stage is critical for successful clinical therapy. Carbohydrate antigen (CA) 19-9 is the most commonly used tumor marker for PDAC, but it lacks satisfactory sensitivity and specificity, especially in early diagnosis [1,2]. There is an urgent need to discover more sensitive and specific biomarkers to improve early diagnosis and screen high-risk patients for PDAC [3].The proteomic approach is a powerful tool to identify novel biomarkers or therapeutic targets from cancer-associated samples. Pancreatic juice is an ideal specimen in proteomic studies, because it is an exceptionally rich source of proteins which are released from pancreatic cells in the physiological state and in pathological conditions [4]. It is reasonable that biomarkers identified in pancreatic juice could subsequently be measured in more accessible specimens such as serum. Therefore, mining pancreatic juice proteome might help to identify novel protein markers for pancreatic diseases such as pancreatic cancer.Recently, many efforts have been made to analyze pancreatic juice by proteomic methods. Rosty et al. [5] identified hepatocarcinoma-intestine-pancreas/pancreatitis-associated protein I (HIP/PAP-I) as a biomarker for PDAC by surface enhanced laser desorption ionization (SELDI) techonology. Chen et al. [6] used isotope-code affinity tag (ICAT) technology to compare the pancreatic juice protein profiling from pancreatitis patients and normal controls. Gronborg et al. [3] used one-dimensional electrophoresis combined with liquid chromatography tandem mass spectrometry (1-DE-LC-MS/MS) to identify a total of 170 unique proteins in pancreatic juices from 3 cases of PDAC patients, and confirmed PAP-2 as a novel marker for PDAC.In the current study, we characterized the pancreatic juice protein profiling in PDAC compared with cancer-free controls using difference gel electrophoresis (DIGE) and tandem mass spectrometry (MS/MS), and identified a number of novel protein markers in pancreatic juice which might be a promising target for pancreatic cancer diagnosis and screening.MethodsPatients and samplesThe study was approved by the Ethical   Commitee of Shandong Academy of Medical Science. Fresh pancreatic juice samples were collected with informed consent from 9 PDAC patients and 9 cancer-free controls undergoing endoscopic retrograde cholangiopancreatography (ERCP) in Qilu Hospital and Qianfoshan Hospital of Shandong University. Clinical data of the patients included were summarized in Table 1. After collection, the pancreatic juice samples were centrifuged at 10,000 rpm for 20 min at 4°C and supernatant of each was aliquoted and stored at -80°C until used.Table 1Clinicopathological data of PDAC patients and cancer-free controls in proteomic analysisNo. of samplesSexAgeHistology*Tumor locationMetastasesPDAC1female55Moderately differentiated ductal adenocarcinomaTail of pancreasYes2male65Well differentiated ductal adenocarcinomaHead of pancreasNo3male38Well differentiated ductal adenocarcinomaHead of pancreasYes4male57Moderately differentiated ductal adenocarcinomaBody of pancreasYes5female53Poorly differentiated ductal adenocarcinomaHead of pancreasNo6male59Moderately differentiated ductal adenocarcinomaHead of pancreasYes7female44Moderately differentiated ductal adenocarcinomaTail of pancreasYes8female51Well differentiated ductal adenocarcinomaHead of pancreasNo9male48Moderately differentiated ductal adenocarcinomaTail of pancreasNoCancer-freeClinical diagnosis1male53Chronic pancreatitis2female66benign cystic neoplasm of pancreas3male49Gallstone pancreatitis4male50Chronic pancreatitis5female52benign cystic neoplasm of pancreas6female57benign cystic neoplasm of pancreas7male46Chronic pancreatitis8female47cystic fibrosis of pancreas9male60Chronic pancreatitis* PDAC differentiation was classified according to the WHO standard (2nd edition, 1996).Pancreatic juice protein extractionPancreatic juice samples were first precipitated with acetone. Briefly, 1.2 mL cold acetone (Fluca) was added to 300 μL pancreatic juice and kept at -20°C for 2 hours, then centrifuged at 13,000 rpm for 10 min at 4°C. Supernatant was discarded and pellet was dissolved in 500 μL of lysis buffer containing 30 mM Tris, 8 M urea, 2 M thiourea, 2% 3-[(3-cholamidopropyl) dimethylammonio]-1-propanesulfonate (CHAPS), 18 mM dithiothreitol (DTT) and 0.5% IPG Buffer (GE Healthcare). The mixture was then centrifuged at 12,000 rpm for 10 min at 4°C, and the supernatant was collected and stored at -80°C. Protein concentration was determined using a 2-D Quant Kit according to the manufacturer's instruction (GE Healthcare).Two-dimensional gel electrophoresis (2-DE)Cancerous or control pancreatic juice protein extracts were first pooled for traditional 2-DE. Two hundred micrograms of each pooled protein sample was diluted in 450 μL rehydration buffer (7 M urea, 2 M thiourea, 4% CHAPS, 0.5% IPG buffer, 0.002% bromophenol blue). Isoelectric focusing was performed on Ettan IPGphor (GE Healthcare) with 24 cm IPG strips (pH 3–10 NL, GE Healthcare). The IPG strips were first rehydrated at 30 V for 12 hours, then focused at 500 V for 1 hour, 1,000 V for 1 hour, and maintained at 8,000 V until a total of 65,000 Vhr was arrived. After isoelectric focusing, the strips were equilibrated with 0.375 M Tris-HCl (pH 8.8), 6 M urea, 20% glycerol, 2% sodium dodecyl sulfate (SDS), and 0.2% bromophenol blue. IEF strips were first treated with 130 mM DTT for 10 min, followed by 135 mM iodoacetamide for 10 min with constant shaking. The equilibrated strips were transferred to 12.5% SDS polyacrylamide gel electrophoresis (SDS-PAGE) on Ettan DALT twelve system (GE Healthcare) with constant power (0.2 W/gel, 1 hour; 1.7 W/gel, 4.5 hours) at 20°C. All gels were stained with Coomassie blue R350 (GE Healthcare), and scanned using a PowerLook 2100 XL scanner system (Umax USA).Two-dimensional difference gel electrophoresis (DIGE)Fifty micrograms of each of cancerous and cancer-free control pancreatic juice protein extracts was minimally labeled with 400 pmol Cy3 or 400 pmol Cy5 fluorescent dye (GE Healthcare). An internal standard pool generated by combining equal amounts of extracts from all the samples was labeled with 400 pmol Cy2. The labeling reaction was carried out in the dark on ice for 30 min, and quenched with 10 mM lysine for 10 min. 2-DE was performed as described above, except that low-fluorescent glass plates were used. The Cy2, Cy3, and Cy5-labeled images were acquired on a Typhoon Trio scanner (GE Healthcare) at the excitation/emission of 488/520 nm, 532/580 nm and 633/670 nm, respectively.Trypsin digestion and MALDI-TOF-MS and MS/MSSpots of interest were excised from gels stained by Coomassie Blue R350, and were digested with sequencing grade modified porcine trypsin (Promega) as described previously [7]. Subsequent protein identification was carried out on the ABI 4700 Proteomic Analyzer MALDI-TOF-MS/MS mass spectrometry (Applied Biosystems, USA) on a reflective mode. Peptide mass fingerprint (PMF) was acquired between 800–3500 Da. The strongest five peaks from PMF were selected to obtain MS/MS spectra. PMF and MS/MS data were then searched against a human subset of the Swiss-Prot database using GPS explorer software (Applied Biosystems, USA).Western blotThirty micrograms of pancreatic juice protein extracts from 6 PDAC and 6 cancer-free controls were used for SDS-PAGE. The separated proteins were transferred to nitrocellulose membranes (GE Healthcare). The membranes were blocked for 1 hour and incubated with anti-DJ-1 antibody (1:1000 dilutions, MBL international corporation, MA), anti-MMP-9 antibody (1:1000 dilutions, Santa Cruz Biotechnology) and anti-A1BG antibody (1:1000 dilutions, Aviva System Biology, San Diego, CA) overnight at 4°C. After peroxidase-conjugated secondary antibody was added, proteins were detected using an ECL Plus kit (GE Healthcare).Immunohistochemical analysis (IHC)To study the tissue compartment for DJ-1, MMP-9 and A1BG, a tissue array (Chaoying Biotechnology co. Xian, China) containing 51 cases of PDAC and 8 adjacent normal pancreas tissues was used in immunohistochemical analysis. Clinicopathological data of the tissue array were summarized (Table 2). Paraffin-embedded tissue array slides were processed for antigen retrieval using microwave heating in citrate buffer, and immunohistochemically stained with the same antibodies used in Western blot analysis at 1:100 dilution. Visualization was performed using ABC kit according to the manufacturer's recommendation (Zhongshan Bio co., Beijing, China).Table 2Data of tissue array included in immunohistochemical analysisNo. of samplesAgeSexHistologyGrade164MaleModerately differentiated ductal adenocarcinomaII264MalePoorly differentiated ductal adenocarcinomaIII364MaleModerately differentiated ductal adenocarcinomaII462FemalePoorly differentiated ductal adenocarcinomaIII562FemalePoorly differentiated ductal adenocarcinomaIII662FemalePoorly differentiated ductal adenocarcinomaIII712FemaleNormal pancreatic tissue-812FemaleNormal pancreatic tissue-912FemaleNormal pancreatic tissue-1049MalePoorly differentiated ductal adenocarcinomaIII1149MaleModerately differentiated ductal adenocarcinomaII1249MaleModerately differentiated ductal adenocarcinomaII1369FemaleModerately differentiated carcinomaII1469FemaleModerately differentiated carcinomaII1569FemaleWell differentiated Mucinous adenocarcinomaI1673FemalePoorly differentiated ductal adenocarcinomaIII1773FemalePoorly differentiated ductal adenocarcinomaIII1873FemalePoorly differentiated carcinomaIII1949MaleModerately differentiated ductal adenocarcinomaII2049MaleModerately differentiated ductal adenocarcinomaII2149MaleModerately differentiated ductal adenocarcinomaII2267MalePoorly differentiated ductal adenocarcinomaIII2367MalePoorly differentiated ductal adenocarcinomaIII2467MalePoorly differentiated ductal adenocarcinomaIII2568FemaleModerately differentiated ductal adenocarcinomaII2668FemaleModerately differentiated ductal adenocarcinomaII2768FemaleModerately differentiated ductal adenocarcinomaII2851FemaleModerately differentiated ductal adenocarcinomaII2951FemalePoorly differentiated ductal adenocarcinomaIII3051FemalePoorly differentiated ductal adenocarcinomaIII3145MaleWell differentiated ductal adenocarcinomaI3245MaleWell differentiated ductal adenocarcinomaI3345MaleWell differentiated ductal adenocarcinomaI3460MalePoorly differentiated ductal adenocarcinomaIII3560MalePoorly differentiated ductal adenocarcinomaIII3660MaleModerately differentiated ductal adenocarcinomaII3765MalePoorly differentiated ductal adenocarcinomaIII3865MalePoorly differentiated ductal adenocarcinomaIII3965MaleModerately differentiated ductal adenocarcinomaII4055MalePoorly differentiated ductal adenocarcinomaIII4155MalePoorly differentiated ductal adenocarcinomaIII4255MalePoorly differentiated ductal adenocarcinomaIII4370MaleWell differentiated ductal adenocarcinomaI4470MaleModerately differentiated ductal adenocarcinomaII4570MaleWell differentiated ductal adenocarcinomaI4630MaleNormal pancreatic tissue-4730MaleNormal pancreatic tissue-4851FemaleWell differentiated ductal adenocarcinomaI4951FemaleWell differentiated ductal adenocarcinomaI5051FemaleWell differentiated ductal adenocarcinomaI5162FemaleModerately differentiated papillary carcinomaII5262FemaleModerately differentiated ductal adenocarcinomaII5362FemaleModerately differentiated ductal adenocarcinomaII5450MalePoorly differentiated ductal adenocarcinomaIII5550MalePoorly differentiated ductal adenocarcinomaIII5650MalePoorly differentiated ductal adenocarcinomaIII5733FemaleNormal pancreatic tissue-5814FemaleNormal pancreatic tissue-5938FemaleNormal pancreatic tissue-Enzyme linked immunosorbent assay (ELISA)Serum from 8 PDAC patients, 9 chronic pancreatitis patients and 8 healthy controls not included in the proteomic analysis were collected for MMP-9 ELISA analysis. Serum levels of human MMP-9 were analyzed with a commercially available kit (Biotrak) according to the manufacturer's recommendation. Serum samples were diluted at 1:100; all assays were done in duplicate. The sensitivity limit of MMP-9 ELISA was 0.08 ng/ml.Data analysisDIGE images were analyzed by ImageMaster 6.0 DIGE-enable software (GE Healthcare). The best internal standard image was assigned as the master reference. The protein spots on the remaining internal standard images were matched to the master reference to ensure that the same protein spots were compared between gels. Spot intensity was normalized by dividing each Cy3 or Cy5 spot volume with the corresponding Cy2 (internal standard) spot volume. Statistical analyses were performed using SPSS 13.0 software. The relationships with DJ-1, MMP-9 and A1BG expression and clinicopathological parameters were analyzed using Chi-square or Fisher's exact tests. The differences of MMP-9 levels in serum among the various groups were analyzed using independent-samples t-test. p < 0.05 was considered to be statically significant.ResultsDifferentially expressed Proteins in pancreatic juice from PDAC and cancer-free controlsGiven the limited amount of pancreatic juice sample available, we first made sample pools of all cancerous and cancer-free control pancreatic juice protein extracts separately to identify the differently expressed proteins by traditional 2-DE. Each pool was repeated three times. Coomassie blue R350 staining was applied to visualize the protein spots, because of its compatibility with protein identification by MS. Moreover, we carried out DIGE on each individual sample to verify differential protein expression found by the traditional 2-DE. Most spot changes in DIGE analysis were consistent with those found in Coomassie blue stained 2-DE analysis (Figure 1). Twenty four proteins with more than two-fold expression change between pancreatic cancer juice and cancer-free pancreatic juice were identified (Table 3). Fourteen proteins were significantly up-regulated and ten proteins were significantly down-regulated in pancreatic cancer juice compared with the cancer-free controls (Figure 2). Three up-regulated proteins MMP-9, DJ-1 and A1BG were further confirmed by Western blot and immunohistochemistry.Table 3Differentially expressed proteins in pancreatic juice from PDAC patients and cancer-free controls (with and without pancreatitis)Spot positionProtein NameMW(Da)PIMascot PMF ScoreSequence Coverage (%)Protein Species Score (C.I.%)Fold changePDAC/Cancer-free controlsPancreatitis/No pancreatitis controls1406Alpha-1-antitrypsin precursor467375.3729839.7881003.43-1.721394Serum albumin precursor693675.9249173.4931002.67-1.272106Apolipoprotein A-I precursor307785.5628956.0541004.831.301652Glutathione S-transferase P323115.439068.24999.9983.84-1.361312Alpha-1B-glycoprotein precursor542735.5812433.8631002.39-1.371471Vitamin D-binding protein precursor529645.4030060.6221005.60-2.032193Cationic antimicrobial protein268869.7512738.51299.9972.73-3.902196Superoxide dismutase [Mn], mitochondrial precursor247228.3510936.5241002.542.121543Serotransferrin precursor770506.8151976.6741005.091.031971Ig lambda chain V-IV region Hil115176.0410249.0561002.661.441253matrix metalloproteinase-9784275.6921432.4581003.67-1.031318Hemopexin precursor516766.5521342.2891003.231.062155Oncogene DJ1198916.3311415.8781002.59-1.251739Fibrinogen beta chain precursor559288.5424753.0671004.15-2.562048Trypsin-1 precursor265586.0812135.176100-2.642.801807Carboxypeptidase A2 precursor468285.6820852.449100-13.001.412007Complement C4-B precursor1927936.7312331.488100-2.10-1.502114Chymotrypsinogen B precursor278706.7914636.051100-2.431.302052Elastase-3A precursor294756.4314335.156100-5.71-1.802030Elastase-3B precursor292935.6515742.945100-2.301.121967Fibrinogen-like protein 1 precursor363925.5819657.347100-2.71-1.511857Haptoglobin precursor452056.1316345.013100-2.00-1.722064Trypsin-3 precursor324997.4612027.118100-2.823.202129Complement C3 precursor1871486.0224854.849100-3.784.04Figure 1Representative gel images of proteins extracted from pancreatic cancer juice and cancer-free controls juice. Representative 2-DE gel images of pancreatic juice proteins from PDAC (A) and cancer-free controls (B), and representative DIGE overlay image (C). Labeled spots are significantly up-regulated proteins in pancreatic cancer juice (A), in cancer-free controls juice (B) and a total of 24 differentially expressed protein spots (C).Figure 2Individual differentially expressed protein spots in pancreatic cancer juice and cancer-free controls juice. Selected areas of 14 up-regulated (A) and 10 down-regulated (B) spots and their corresponding DIGE images.Western blot analysis of MMP-9, DJ-1 and A1BGAs seen in Figure 3, compared with cancer-free controls, increased MMP-9, DJ-1 and A1BG expression in cancerous pancreatic juice were detected by Western blot. In addition, cancerous pancreatic juice with MMP-9 expression evinced both 92 kDa and 82 kDa bands, corresponding to the latent and activated forms of MMP-9, respectively.Figure 3MMP-9, DJ-1 and A1BG expression analysis by Western blot. Increased MMP-9 (A), DJ-1 (B) and A1BG (C) were detected in the cancerous juice samples compared with cancer-free pancreatic juice samples; MMP-9 expression evinced both 92 kDa and 82 kDa bands, corresponding to the latent and activated forms of MMP-9, respectively.Immunohistochemical validation of MMP-9, DJ-1 and A1BGThe expressions of MMP-9, DJ-1 and A1BG were confirmed by immunohistochemistry in 51 pancreatic cancer and 8 normal pancreas samples that were not included in the proteomic experiment (Table 4; Figure 4). MMP-9 was detected in the malignant ductal epithelia in 82.4% of PDAC tissues, but was barely detectable in normal pancreas. In addition, 9 cases of PDAC demonstrated strong MMP-9 expression in the stroma.Table 4Summary of MMP-9, DJ-1 and A1BG immunohistochemical study on PDAC tissue arrayPDACNegativePositiveTotalMMP-9DJ-1A1BGMMP-9DJ-1A1BGWell differentiated2 (22.2%)2 (22.2%)1 (11.1%)7 (77.8%)7 (77.8%)8 (88.9%)9Moderately differentiated2 (10.5%)3 (15.8%)2 (10.5%)17 (89.5%)16 (84.2%)17 (89.5%)19Poorly differentiated5 (21.7%)9 (39.1%)4 (17.4%)18 (78.3%)14 (60.9%)19 (82.6%)23Figure 4MMP-9, DJ-1 and A1BG expression analysis by Immunohistochemistry (×200). MMP-9 (A), DJ-1 (C) and A1BG (E) over-expressed in PDAC tissues. MMP-9 (B), DJ-1(D) and A1BG (F) were not detectable in normal pancreas tissues.In normal pancreas, DJ-1 was negatively or weakly expressed in duct epithelium, acinar cells and islet cells. In PDAC, 72.5% demonstrated DJ-1 over-expression in cancer cells. In addition, DJ-1 over-expression was related to the differentiation of PDAC. DJ-1 positive stain was observed in 7/9 well differentiated tumors, 16/19 moderately differentiated tumors, and 14/23 poorly differentiated tumors, respectively. There was a significant difference of DJ-1 over-expression between moderately differentiated PDAC with poorly differentiated PDAC (p < 0.05).No positive A1BG staining was detected in all the normal pancreas tissues. A1BG was over-expressed in the cytoplasma of malignant epithelia in 86.3% of pancreatic cancer tissues, significantly higher than that in normal pancreas tissues (p < 0.01).Serum MMP-9 levels by ELISAAs shown in Figure 5, serum MMP-9 levels were significantly higher in patients with pancreatic cancer (median = 255.14 ng/ml; quartile range, 125.43 ng/ml) than those with chronic pancreatitis (median = 210.22 ng/ml; quartile range, 12.48 ng/ml; p = 0.009) and normal controls (median = 203.77 ng/ml; quartile range, 17.04 ng/ml; p = 0.027).Figure 5Box plot of MMP-9 serum levels for PDAC, chronic pancreatitis and normal controls by ELISA. The serum levels of MMP-9 in PDAC patients were significantly higher than those in chronic pancreatitis patients and in healthy controls (p < 0.05). PDAC: pancreatic ductal adenocarcinoma patients; CP: chronic pancreatitis patients; N: healthy controls.DiscussionPDAC is the most common pathological subgroup of pancreatic cancer. During the development of PDAC, malignant ductal cells preferentially shed into the ductal lumen, making pancreatic juice a rich source of cancer-specific proteins. Therefore, pancreatic juice is an ideal specimen for identifying new tumor markers for PDAC.We adopted a quantitative proteomic technology (DIGE) to compare the protein profiling of pancreatic juice from PDAC and its cancer-free controls. DIGE technology not only provided reliable quantification, but also can minimize the run-to-run reproducibility of conventional 2-DE [8]. In this study, a total of 24 differentially expressed proteins between pancreatic juice from PDAC and cancer-free controls, including 14 up-regulated proteins and 10 down-regulated proteins, were identified. Half of the up-regulated proteins in PDAC, such as DJ-1, MMP-9, apolipoprotein A-I, A1BG, SOD [Mn], serotransferrin and Ig lambda chain V-IV region were firstly identified in cancerous pancreatic juice. Three up-regulated proteins MMP-9, DJ-1 and A1BG in PDAC pancreatic juice were further validated. Western blot analysis demonstrated elevated protein levels of MMP-9, DJ-1 and A1BG in cancerous pancreatic juice compared with cancer-free pancreatic juice, which is consistent with our proteomic findings.For the scarcity of cancer-free pancreatic juice specimens, two types of pancreatic juice control subjects, one with pancreatitis and the other without pancreatitis, were included in our study. In a previous proteomic study by Shen et al. [9], it was demonstrated that there are obvious differences in protein profiling between pancreatitis and healthy control tissues. So the protein profiling of pancreatic juice between pancreatitis patients and control subjects without pancreatitis was also compared (Table 3). Seven pancreatic juice proteins, such as trypsin-1 precursor, were demonstrated to be differently expressed between cancer-free controls with and without pancreatitis. In addition, five down-regulated proteins identified in pancreatic cancer juice (carboxypeptidase A2, chymotrypsinogen B, elastase-3A, elastase-3B and trypsin-1) was also down-regulated in pancreatic cancer tissues compared with pancreatitis and normal pancreas tissues as described by Shen et al. [9].Over-expression of MMP-9 and DJ-1 in pancreatic cancer tissues have been shown in previous studies [10,11]. It is believed that MMP-9 over-expression results in the degradation of the basement membrane and contributes to local invasion or distant metastases during pancreatic carcinogenesis [12,13]. DJ-1 is a novel mitogen-dependent oncogene involved in ras-related signal transduction pathway [14]. Recently, several previous studies have shown that DJ-1 is over-expressed in multiple cancer tissues including pancreatic cancer [15,16]. Our immunohistological study extended the above investigation to larger sample size and revealed that MMP-9 and DJ-1 were expressed in 82.4% and 72.5% of PDAC tissues, significantly higher than that in normal pancreas. Furthermore, DJ-1 over-expression was associated with PDAC differentiation. Moderately differentiated PDAC demonstrated a higher expression level of DJ-1 than poorly differentiated tumor subgroup, which suggests that DJ-1 may be a differentiation related protein.A1BG, a member of the immunoglobulin superfamily, is believed to be a secreted plasma protein, but its function is unknown [17]. In this study, we first identified and confirmed A1BG as an elevated protein in pancreatic juice and cancer tissues of PDAC. Our immunohistochemical study further validated that A1BG protein over-expression was seen in most PDAC tissues, but not detected in normal pancreas tissues. Using a glycoprotein profiling method, Kreunin et al. [18] recently found that A1BG was detected in urinary samples from bladder cancer patients, but in none of the samples obtained from non-tumor-bearing individuals. Yoon et al. [19] also demonstrated A1BG over-expression in liver cancer cell lines and tissues by semi-quantitative RT-PCR. These findings together with our proteomic results suggest that A1BG might be a cancer-associated gene and a novel tumor marker of cancer, and its possible functions in carcinogenesis deserve further investigation.The novel protein biomarkers identified in ERCP-obtained pancreatic juice might have enormous potential for use in the diagnosis of PDAC. It is reasonable to believe that proteins indicative of cancer in pancreatic juice are also elevated in blood. Therefore, we measured serum levels of MMP-9 in PDAC, and found that serum MMP-9 had a higher level in PDAC than in chronic pancreatitis and healthy controls. The findings suggest that serum measurement of MMP-9 or other up-regulated proteins in cancerous pancreatic juice might be helpful in the diagnosis of pancreatic adenocarcinoma and deserve further investigation.ConclusionIn the present study, we carried out a comparative proteomics analysis on pancreatic juice from PDAC and its cancer-free controls. A total of 24 differentially expressed proteins were identified, of which 14 were over-regulated and 10 were under-regulated in pancreatic juice of PDAC. We further confirmed the expression levels of three up-regulated proteins (MMP-9, DJ-1 and A1BG) in pancreatic juice, tissue and serum samples using Western blot, IHC or ELISA. A1BG was firstly identified as a biomarker in cancer-associated samples. Serum MMP-9 measurement might be helpful in discriminating pancreatic adenocarcinoma from chronic pancreatitis and healthy controls. The newly identified proteins in this study might be useful for developing new pancreatic juice-related or serum-related diagnostic markers for PDAC.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsMT and YZC participated in the experimental design, electrophoresis, trypsin digestion, MS identification, Western blot, statistical analysis and the manuscript writing; JXH was responsible for the experimental design, technique and financial support, and manuscript writing; GHS and XYZ were responsible for IHC experiments; MJZ and YC participated in electrophoresis, trypsin digestion. All authors contributed to this work read and approved the final manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2528051.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2528051",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2528051\nAUTHORS: Véronique Pons, Pierre-Philippe Luyet, Etienne Morel, Laurence Abrami, F. Gisou van der Goot, Robert G Parton, Jean Gruenberg\n\nABSTRACT:\nAfter internalization, ubiquitinated signaling receptors are delivered to early endosomes. There, they are sorted and incorporated into the intralumenal invaginations of nascent multivesicular bodies, which function as transport intermediates to late endosomes. Receptor sorting is achieved by Hrs—an adaptor-like protein that binds membrane PtdIns3P via a FYVE motif—and then by ESCRT complexes, which presumably also mediate the invagination process. Eventually, intralumenal vesicles are delivered to lysosomes, leading to the notion that EGF receptor sorting into multivesicular bodies mediates lysosomal targeting. Here, we report that Hrs is essential for lysosomal targeting but dispensable for multivesicular body biogenesis and transport to late endosomes. By contrast, we find that the PtdIns3P-binding protein SNX3 is required for multivesicular body formation, but not for EGF receptor degradation. PtdIns3P thus controls the complementary functions of Hrs and SNX3 in sorting and multivesicular body biogenesis.\n\nBODY:\nIntroductionCell surface lipids and proteins as well as solutes are endocytosed into animal cells through several routes, including clathrin-coated pits and vesicles, caveolae, and pathways that do not depend on caveolae or clathrin [1,2]. Some receptors may be differentially sorted into clathrin-coated pits or caveolae/rafts, depending on their fate: signaling or degradation [3–5]. These endocytic routes all seem to lead to early endosomes eventually; from which some components, including housekeeping receptors, are returned to the cell surface for reutilization while others are transported to the trans-Golgi network and the biosynthetic pathway. By contrast, signaling receptors that need to be down-regulated, including the activated EGF receptor, as well as other endocytosed proteins and lipids, are efficiently sorted away from recycling molecules within early endosomes, and are then routed towards late endosomes and lysosomes, where degradation occurs [6,7].Major progress has been made in understanding the molecular mechanisms responsible for EGF receptor sorting in early endosomes. Sorting signals are provided by the addition of multiple ubiquitin molecules to the receptor cytoplasmic domain [8]. Ubiquitin coupled to the receptor then binds Hrs through its ubiquitin interacting motif, while Hrs also interacts simultaneously with membrane PtdIns3P via a FYVE domain and with clathrin, leading to the concentration of activated receptor molecules into specialized clathrin-coated regions of early endosomes [9–11]. The epidermal growth factor (EGF) receptor then interacts sequentially with the ESCRT-I, ESCRT-II, and ESCRT-III protein complexes, and eventually appears within lumenal vesicles present in multivesicular regions of the early endosomes [12–14]—a process that uncouples activated receptors from cytosolic signaling partners and thus efficiently terminates signaling. This mechanism also appears to contribute to the ubiquitin-independent sorting of the delta opioid G-protein–coupled receptor in mammalian cells [15] and Sna3p in yeast [16–18]. However, sorting of the melanosomal protein Pmel17 does not depend on ubiquitin, Hrs, and ESCRT components [19], and likewise, other proteins, including proteins of the limiting membrane, are presumably transported in an ESCRT-independent manner.Eventually, these multivesicular regions detach—or mature—from early endosomes, and become free multivesicular endosomes or bodies (MVBs), which then serve as transport intermediates (ECVs or endosomal carrier vesicles), in the microtubule-dependent transport towards late endosomes—herein referred as ECV/MVBs. Formation of ECV/MVBs appears to be itself under the control of EGF signaling [20] and depends on annexin family members [21,22], whereas conversion of the small GTPase Rab5 to Rab7 may occur on early endosome [23] or during early-to-late endosome transport [24]. Upon fusion with late endosomes, ECV/MVB lumenal vesicles are delivered to late endosomes, and are eventually packaged within lysosomes, where they are degraded together with their cargo of down-regulated receptors.Evidence shows that the sorting process mediated by Hrs and ESCRTs is coupled somehow to the mechanism of membrane invagination towards the lumen of the endosome, including during the topologically equivalent process of HIV budding at the plasma membrane, which takes advantage of the ESCRT machinery [25–28]. In yeast, mutation of the genes that encode ESCRTs and other components of this pathway leads to the formation of an aberrant endosomal compartment (Class E VPS phenotype). Then, endocytosed membrane proteins that would normally be degraded accumulate in this Class E compartment but also on the membrane of the vacuole (see [12] and reference therein), which is functionally equivalent to lysosomes. Similarly, Hrs knockout in Drosophila [29] or knockdown in mammalian cells [20,30] causes a reduction in the number of lumenal vesicles, presumably because the downstream machinery responsible for membrane invagination fails to assemble in cells lacking Hrs. Knockdown of Tsg101, an ESCRT-I subunit, causes pleiotropic changes in early endosome morphology, including tubulation, reduction in the number of lumenal vesicles [20], and perhaps, the formation of a mammalian equivalent of the yeast Class E compartment [31]. While it is clear that Hrs and ESCRTs play an essential role in the lysosomal targeting of many ubiquitinated proteins, the mechanism controlling membrane invagination itself remains elusive. All proteins that are known to induce membrane deformation act in the topologically opposite direction—towards the cytoplasm—via direct insertion into the bilayer or protein–lipid interactions [32].Here, we have further investigated the mechanisms that control endosomal membrane dynamics, and in particular the role of PtdIns3P and its effectors. Indeed, PtdIns3P is well known to regulate endocytic membrane traffic and protein sorting through numerous effectors that contain FYVE or PX PtdIns3P-binding domains. Moreover, PtdIns3P controls both the sorting of signaling receptors [33] and the biogenesis of lumenal membranes [34–36] at least in part via the FYVE-containing protein Hrs [37]. Searching for other PtdIns3P effectors, we screened proteins of the sorting nexin (SNX) family that all contain the phosphoinositide-binding PX domain [38]. Several SNX proteins play a role in protein trafficking and some contain a BAR domain, involved in sensing and/or inducing membrane curvature [39]. We find that SNX3 plays a direct role in multivesicular body formation, but is not involved in EGF receptor degradation. By contrast, Hrs seems to be essential for lysosomal targeting but dispensable for multivesicular body biogenesis and transport to late endosomes.ResultsWhen ectopically expressed, green fluorescent protein (GFP)-tagged SNX3 showed a characteristic punctate distribution, and colocalized with EEA1, an effector of the small GTPase Rab5, and to some extent with the transferrin receptor, but not with late endocytic markers, including lysobisphosphatidic acid (LBPA) (Figure S1A) or Lamp1 (Figure 1C and 1F). Consistently, GFP-SNX3 cofractionated with Rab5, but not with LBPA (Figure S1B). GFP-SNX3 was thus primarily present on early endosomes, much like the endogenous protein [40], and this association depended on PtdIns3P and required an intact PtdIns3P-binding domain PX (Figure S1C), as expected [40].Figure 1SNX3 in Endosomal Transport(A and B) HeLa cells expressing GFP-SNX3 were incubated with 0.25 μg/ml EGF and 10 μg/ml cycloheximide for the indicated time periods. Cell lysates (100 μg) were analyzed by SDS gel electrophoresis and western blotting with antibodies against EGFR (20-ESO4 against a peptide of the cytoplasmic domain), α-tubulin (α-tub), or GFP. Blots were scanned and the quantification is shown in (B).(C and D) After cell surface binding, biotin-EGF coupled to streptavidin-R-phycoerythrin was internalized for 50 min at 37 °C in HeLa cells expressing GFP-SNX3. Cells were labeled with antibodies against Lamp1 (C) or EEA1 (D) and analyzed by triple channel fluorescence. The various combinations of merged colors for (C) and (D) are shown in Figure S3A and S3B, respectively. Insets in the lower right are a magnification of the regions shown in the boxes.(E) Individual endosomes containing both EGF and EEA1 were counted in (D) and in control cells (Figure S4A). Values are expressed as a percentage of the total number of EGF-containing endosomes (>10 cells per experiment).(F) Rhodamine-dextran was pulsed for 10 min at 37 °C in HeLa cells expressing GFP-SNX3 and then chased for 40 min. Cells were analyzed as in (C and D) using antibodies against Lamp1, and the various combinations of merged colors are in Figure S3C.(G) After cell surface binding, Shiga toxin B-subunit conjugated to Cy3 was internalized for 50 min at 37 °C into HeLa cells expressing GFP-SNX3. Cells were analyzed as in (C and D) using antibodies against Rab6, and the various combinations of merged colors are the same as in Figure S3D.In (B and E), each condition is the mean of three independent experiments; standard errors are indicated.(C, D, F, and G) Scale bar indicates 10 μm.To investigate the possible role of SNX3, we first tested whether overexpression affected transport along the endocytic pathway. SNX3 overexpression had no effect on EGF receptor endocytosis, since internalized EGF colocalized with both GFP-SNX3 and EEA1 in early endosomes within 10 min of incubation at 37 °C (Figure S2A), much like in control cells (unpublished data, see [33]). However, upon longer incubation times at 37 °C, EGF receptor degradation was delayed by SNX3 overexpression (Figure 1A; quantification in Figure 1B), in agreement with previous findings [40]. This delay was caused by defective transport, since the bulk of endocytosed EGF failed to reach late endosomes containing Lamp1 (Figures 1C and S3A) or LBPA (unpublished data) after 50 min in cells overexpressing SNX3. Then, EGF remained in early endosomes containing GFP-SNX3 and EEA1 (Figures 1D and S3B; quantification in Figure 1E), whereas it was mostly transported beyond early endosomes in control cells (Figure S4A). The inhibitory effect of SNX3 was specific, since overexpression of other SNX family members, including SNX1, SNX2, and SNX16, did not interfere with EGF receptor transport (Figure S5), as expected [41,42].Much like with EGF receptor, endocytosis of the bulk phase marker rhodamine-dextran was not affected by SNX3 overexpression, and the tracer accumulated in early endosomes containing GFP-SNX3 and EEA1 within 10 min at 37 °C (Figure S2B). However, after a subsequent 40-min chase at 37 °C, SNX3 overexpression markedly impaired dextran transport to late endosomes. The bulk of the tracer then remained in early endosomes and failed to reach Lamp1-positive late endocytic compartments (Figures 1F and S3C), as was observed with the EGF receptor (Figure 1C and S3A) and in contrast to control cells (Figure S4B). These inhibitory effects of excess SNX3 were specific, since SNX3 overexpression did not affect the transport of endocytosed Shiga toxin B-subunit from early endosome (Figure S2C) to the trans-Golgi network (Figures 1G and S3D; compare with controls in Figure S4C), which requires SNX1 [43]. Altogether these observations indicate that SNX3 overexpression caused the selective retention of signaling receptors and bulk tracers in early endosomes containing Rab5 and its effector EEA1, and thereby inhibited more-distal steps of early-to-late endosome transport.To further investigate the role of SNX3 in early-to-late endosome transport, we made use of vesicular stomatitis virus (VSV), which infects cells from the endocytic pathway. Indeed, endocytosed virions must be transported beyond early endosomes for efficient infection to occur [44,45]. Overexpression of SNX3 had no effect on endocytosis of VSV (Figure 2A), as observed with EGF and dextran (Figure S2A and S2B), but significantly reduced VSV infection, which was monitored by the synthesis of the viral glycoprotein G (Figure 2B). This inhibition did not result from some indirect effects of SNX3 on the G-protein biosynthetic pathway, since replication of the viral genome, quantified by real time-PCR (RT-PCR), was also similarly reduced (Figure 2C), indicating that excess SNX3 prevented efficient release of the viral nucleocapsid into the cytosol.Figure 2SNX3 in VSV Infection and MVB Biogenesis(A) VSV (1 multiplicity of infection [MOI]) was bound on ice to the surface of control HeLa cells (ctrl) or HeLa cells expressing GFP-SNX3. Cells were incubated for 15 min at 37 °C to allow VSV internalization. Then, total endosomes were prepared by fractionation and viral RNA was quantified by RT-PCR. Values are expressed as a percentage of untransfected controls.(B) Experiments were as in (A), except that cells were incubated for 3 h, instead of 15 min, at 37 °C to allow VSV infection to proceed. Cells were analyzed by immunofluorescence with antibodies against VSV-G protein (left panels); stars indicate cells expressing GFP-SNX3 and not VSV-G. The total number of cells expressing the G-protein was counted and is expressed as a percentage of the controls (≈60% of untransfected control cells were infected).(C) Experiments were as in (B), except that 0.1 MOI VSV was used and viral RNA replication was quantified by RT-PCR. Values are expressed as a percentage of untransfected controls.(D) Experiments were as in (A), except that 3 MOI Dil-labeled VSV was used and that cells were incubated for 35 min at 37 °C. Then, viral fusion events were visualized as fluorescent spots (due to Dil dequenching) by fluorescence microscopy, and quantified. Values are expressed as in (B).(E) Control HeLa cells (Figure S6A) or cells expressing GFP-SNX3 (Figure S6B) were fixed and embedded in Epon. Random fields of the different areas in one section were captured at 25.000× magnification (the procedure was repeated 3× for controls and 5× after GFP-SNX3 overexpression). Early endosomes (EE) were identified as vacuole of 200–500 nm containing five or fewer clearly defined internal vesicles, and multivesicular regions as vacuole of 200–500 nm containing five or more clearly defined spherical internal vesicles [66]. The volume of EE and multivesicular regions was calculated by stereological means, and the mean values are expressed as a percentage of the cytoplasm volume ± standard error of the mean (SEM).(F–I) HeLa cells expressing GFP-SNX3 were analyzed by electron microscopy after immunogold labeling of cryosections using antibodies against GFP ([F–H], arrows) and the transferrin receptor (TfR) ([H and I], arrowheads) followed by proteinA-gold, as indicated.In (A–E), each condition is the mean of at least three independent experiments; standard errors are indicated.(B) Scale bar indicates 10 μm; (F–I) scale bar indicates 0.2 ��m.The viral nucleocapsid is released into the cytoplasm after low pH-triggered fusion of the viral envelope with endosomal membranes. To monitor viral fusion events, VSV was labeled with self-quenching amounts of Dil, a fluorescent long-chain dialkylcarbocyanine dye, and bound to the cell surface at 4 °C [44]. After endocytosis at 37 °C, the fusion of individual virions was revealed in the light microscope by the appearance of fluorescent spots in endosomes, due to Dil dequenching [44]. Strikingly, overexpression of SNX3 markedly reduced VSV fusion (Figure 2D). Since VSV fusion normally occurs beyond early endosomes [44,45], these observations indicate that virions, much like EGF receptor and dextran (Figure 1) remained trapped in early endosomes in cells overexpressing SNX3.The effects of SNX3 overexpression on the transport of viral particles, EGF receptor, and fluid phase markers led us to investigate by electron microscopy whether endosome morphology was then affected. SNX3 overexpression caused a dramatic accumulation of multivesicular structures without causing a general expansion of endosome vesicular regions (Figure S6; quantification in Figure 2E), perhaps suggesting that only specialized regions of the early endosome are competent to become MVBs. Multivesicular elements after SNX3 overexpression were frequently clustered in groups of five to ten individuals, as revealed in low (Figure S6) and high (Figure 2F) magnification views, perhaps connected to each other (Figure 2F). Immunogold labeling of cryosections showed that GFP-SNX3 itself was abundant on the limiting membrane of these multivesicular structures. These SNX3-positive structures all exhibited a similar multivesicular and spherical appearance (diameter ≈ 0.4 μm), which closely resembles the morphology of MVBs or ECVs that mediate early-to-late endosome transport [46]—herein referred to as ECV/MVBs. However, in marked contrast to free ECV/MVBs during transport towards late endosomes, these SNX3-positive structures exhibited the characteristic features of early endosomes, including all early endosomal markers that were tested (Figure 1 and see Figure 3). Moreover, the transferrin receptor, which is restricted to early and recycling endosomes, was found closely associated with these SNX3-positive ECV/MVB-like structures (arrowheads in Figure 2H and 2I), consistent with data from us (Figure 1) and others [40]. The transferrin receptor was often present in tubulo-cisternal elements connected to multivesicular elements (Figure 2I), presumably corresponding to forming recycling endosomes. It thus appears that SNX3 overexpression causes the accumulation of multivesicular regions on early endosomal membranes. These ECV/MVB-like structures fail to detach, or to mature, from early endosomes. Such a frustrated process of ECV/MVB formation accounts nicely for our observations that the transport of EGF receptor, VSV and bulk markers beyond early endosomes is then inhibited (see model, Figure S10).Figure 3SNX3 Colocalizes with Hrs, Ubiquitinated Proteins, and Clathrin on Rab5Q79L Enlarged Endosomes.(A–D) HeLa cells co-expressing both GFP-Rab5Q79L and mRFP-SNX3 were processed for immunofluorescence using different antibodies as indicated and analyzed by triple channel fluorescence. Arrowheads point at regions containing GFP-SNX3. Scale bar indicates 10 μm. Insets in the lower right are a magnification of the regions shown in the boxes.In our electron microscopy analysis, SNX3 was often observed on, or close to, membrane regions containing electron-dense materials on the cytoplasmic membrane face (e.g., Figure 2G), which resembled the early endosomal Hrs-clathrin coat that mediates ubiquitinated receptor sorting into lumenal vesicles during ECV/MVB formation on early endosomes [9,10]. To further characterize SNX3 distribution, we thus used the constitutively active mutant Rab5Q79L, which induces the formation of enlarged early endosomes by promoting their homotypic fusion. On these large endosomes, regions containing Hrs, clathrin, and ubiquitinated receptors could be resolved by light microscopy from those containing EEA1, indicating that components of the machinery that sorts down-regulated receptors into ECV/MVB internal vesicles are concentrated in specific early endosomal domains [9]. When GFP-Rab5Q79L was coexpressed with monomeric red fluorescent protein (mRFP)-SNX3, both proteins were present on enlarged early endosomes, as expected. Approximately 90% of endosomes containing SNX3 were also labeled with Rab5Q79L, but some variation in SNX3 association with individual Rab5Q79L-endosomes was observed (Figure 3), presumably reflecting different angles of visualization in the confocal planes. SNX3 was clearly seen to colocalize preferentially with Hrs (Figure 3A), ubiquitinated proteins (Figure 3B), and clathrin (Figure 3C) in regions that seemed devoid of EEA1 (Figure 3D). These observations further demonstrate that ectopically expressed SNX3 accumulates on early endosomal membranes. They also indicate that this accumulation occurs preferentially in multivesicular regions that contain, in addition to SNX3 itself, the protein machinery responsible for sorting into lumenal vesicles. It thus seems that the lumenal invagination process continues in the presence of excess SNX3, leading to the accumulation of multivesicular regions on early endosomes, but that more distal transport events, including ECV/MVB detachment—or maturation—(Figures 1A–1F and 2B–2D), are inhibited, perhaps because excess SNX3 limits the access or binding of downstream machineries.Since overexpression of SNX3 caused an expansion of multivesicular regions on early endosomes, we investigated the impact of SNX3 down-expression on the formation of lumenal vesicles by electron microscopy. To this end, the lumen of ECV/MVBs was labeled with endocytosed horseradish peroxidase (HRP) pulsed for 15 min and then chased for 30 min at 37 °C, after microtubule depolymerization with 10 μM nocodazole [46]. As expected [46], multivesicular structures with the characteristic ECV/MVB morphology were found in controls (Figure 4A, upper panel), accounting for approximately 70% of the total HRP-positive structures (see quantification in Figure 5D). By contrast, after SNX3 knockdown to approximately 20% of the control levels (inset in Figure 4C), approximately 70% of the total HRP-positive profiles did not seem to contain lumenal vesicles, but were otherwise similar to controls (Figure 4B, upper panel). All SNX3 knockdown experiments were repeated with two small interfering RNA (siRNA) target sequences without significant differences (see Figures 6A, 6B, and S8C).Figure 4SNX3 Silencing Inhibits Membrane Formation within MVBs but Does Not Affect EGFR Early-to-Late Endosomal Transport and Degradation(A) The ECV/MVB content of mock-treated HeLa cells was labeled with endocytosed HRP [22,46]. Samples were processed for plastic embedding after HRP cytochemical detection, and analyzed by electron microscopy. The micrographs show representative high-magnification views of individual HRP-containing structures (upper panel). In separate experiments, the ECV/MVB content was labeled with 5-nm proteinA-gold endocytosed for 30 min at 37 °C (lower panel).(B) As in (A), but cells were treated with SNX3 siRNAs.(C) SNX3 siRNA or mock-treated HeLa cells were incubated with 3 MOI Dil-labeled VSV and viral fusion was monitored as in Figure 2D (inset: SNX3 down-regulation in cells treated with SNX3 siRNAs).(D) HeLa cells were treated with SNX3 siRNAs or mock-treated, and then microtubules were depolymerized or not with 10 μM nocodazole. Then, cells were infected with 0.1 MOI VSV, and RNA replication was quantified as in Figure 2C.(E and F) HeLa cells treated with SNX3 siRNAs, Hrs siRNAs, or mock-treated were stimulated with EGF and analyzed as in Figure 1A. Blots were scanned and the quantification is shown in (F).(G) Individual endosomes containing both EGF and EEA1 or Lamp1 were counted in SNX3 siRNA-treated cells (see [H] below) and in control cells (Figure S8A). Values are expressed as a percentage of the total number of EGF-containing endosomes in each category (>10 cells per experiment).(H) EGF-biotin coupled to streptavidin-AlexaFluor 488 was endocytosed as in Figure 1C and 1D in cells treated with SNX3 siRNA. Cells were analyzed by immunofluorescence with the indicated antibodies. Insets in the lower right are a magnification of the regions shown in the boxes.In (C and D) and (F and G), each condition is the mean of at least three independent experiments; standard errors are indicated.(A and B) Scale bar indicates 0.25 μm; (H) scale bar indicates 10 μm.Figure 5SNX3 Rescues the Formation of Internal Vesicles in Hrs siRNA-Treated Cells(A) HeLa cells treated with siRNAs against Hrs or SNX3 or mock-treated were lysed. Lysates were analyzed by SDS gel electrophoresis and western blotting using indicated antibodies. The red box highlights the reduction in SNX3 levels observed after Hrs knockdown.(B) HeLa cells were treated with Hrs siRNAs. Then, the endosomal content was labeled with HRP (left panel) or 5-nm proteinA-gold (right panel) and analyzed by electron microscopy as in Figure 4A.(C) As (B), except that GFP-SNX3 was overexpressed during the last 24 h.(D) The number of MVBs and “empty” MVBs were counted in the experiments shown in Figure 4A (mock-treated cells), Figure 4B (SNX3 siRNAs), Figure 5B (Hrs siRNAs), and Figure 5C (Hrs siRNAs followed by SNX3 overexpression), and are expressed as a percentage of the total number of HRP-labeled endosomes (≈50 individual endosomes counted for each condition). All structures containing one or more intralumenal vesicles were counted as MVBs). To ensure unbiased analysis and quantification, micrographs were taken in Brisbane, and each condition was number coded. Analysis and quantification were then performed blind in Geneva.Figure 6SNX3 Controls the Formation of Intralumenal Vesicles That Incorporate the EGF Receptor(A and B) HeLa cells were mock-treated or treated with siRNAs against SNX3 or Hrs (two different siRNAs in each case) and transfected with GFP-Rab5Q79L (green) during the last 24 h. Alternatively, mRFP-SNX3 was overexpressed in cells treated with each anti-Hrs siRNA. After cell surface binding, EGF was internalized for 15 min at 37 °C. Cells were labeled with anti-EGF-R antibodies (blue) and analyzed by confocal microscopy. In (B), the relative amount of EGF-R in the lumen of endosome was quantified [50] and expressed as the percentage of the total amount of EGF-R. Each condition is the mean of at least three independent experiments; standard errors are indicated. (A) Scale bar indicates 10 μm. Insets in the lower left are a magnification of the regions shown in the boxes.To better visualize the presence or absence of lumenal vesicles, ECV/MVBs were labeled with 5-nm proteinA-gold endocytosed for 30 min at 37 °C with or without nocodazole. In the mock-treated control, the gold particles distributed within ECV/MVBs, which typically contained approximately 20 vesicles per profile, whether microtubules were present (Figure 4A, lower panel) or not (unpublished data). After SNX3 knockdown, however, 5-nm gold particles labeled vesicles of the same diameter as ECV/MVBs, but with only three to five internal vesicles per profile, whether microtubules were intact (Figure 4B, lower panel) or not (unpublished data). The appearance and size of internal vesicles were otherwise undistinguishable from the controls. Altogether, our data thus indicate that SNX3 plays a direct and specific role in the formation of intralumenal membrane invaginations within nascent ECV/MVBs, since multivesicular regions are increased by overexpression (Figure 2F) and decreased by down-expression (Figure 4B).Interestingly, SNX3 knockdown did not affect virus fusion (Figure 4C) and caused only a small, marginal decrease in nucleocapsid release (Figure 4D). Previously, we had found that the VSV envelope undergoes fusion primarily with the membrane of ECV/MVB internal vesicles, thus releasing the capsid into their lumen, where it remains hidden [44]. In late endosomes, back fusion of these vesicles with the endosome-limiting membrane then ensures capsid delivery to the cytoplasm, indicating that VSV fusion and capsid release occur in sequential steps of the pathway. In particular, we found that depolymerization of the microtubules, which reduces early-to-late endosome transport [46], does not affect viral fusion, but efficiently inhibits VSV delivery to late endosomes and capsid release [44] (see Figure 4D). In contrast to controls, capsid release was only marginally affected by microtubule depolymerization in cells treated with SNX3 siRNAs (Figure 4D)—much like in cells treated with PI 3-kinase inhibitors or Hrs siRNAs [44], which both decrease intralumenal membranes in endosomes [30,35] (see also Figure 5B). This was not due to some indirect effects of SNX3 siRNAs, since early-to-late endosome transport remained microtubule dependent after SNX3 knockdown (see below and Figure S7A and S7B). It thus appears that, when ECV/MVBs lack intralumenal vesicles after SNX3 knockdown, VSV fusion can be triggered at the limiting membrane, thus by-passing the need for transport to late endosomes—again much like after PI 3-kinase inhibition or Hrs knockdown [44].The “empty” endosomes in cells lacking SNX3 closely resemble endosomes observed after Hrs knockdown in mammalian cells [30] (see Figure 5B) or mutagenesis in Drosophila [29]. In these studies, Hrs depletion also inhibited EGF receptor degradation, supporting the view that sorting into intralumenal invaginations mediates lysosomal targeting [13,37]. To our surprise, SNX3 knockdown had little effect on EGF receptor degradation (Figure S7C), significantly less than Hrs knockdown (see blot in Figure 5A) in parallel experiments (Figure 4E, quantification of the blots in Figure 4F). Consistently, a wave of fluorescent EGF reached late endocytic compartments containing Lamp1 in cells treated with SNX3 siRNAs (Figure 4H; quantification in Figure 4G) as in mock-treated cells (Figure S8A), and this transport required intact microtubules (Figure S7A and S7B). Then, EGF no longer colocalized with EEA1 in early endosomes (Figure 4H; quantification in Figure 4G) as in mock-treated cells (Figure S8A). This is in contrast to the inhibition observed after SNX3 overexpression (Figure 1C–1E). SNX3 knockdown thus appears to prevent the formation of intralumenal invaginations within endosomes without interfering with EGF receptor transport and degradation, indicating that the lysosomal targeting of signaling receptors is then uncoupled from sorting into ECV/MVBs.Both Hrs and SNX3 seem to play a role in the membrane invagination process, but only Hrs, and not SNX3, appears to be involved in EGF receptor targeting to lysosomes, perhaps suggesting that Hrs acts upstream of SNX3 in receptor sorting and multivesicular body biogenesis. Consistent with this notion, Hrs knockdown selectively reduced the expression of SNX3—without affecting any other protein involved in endosome membrane dynamics that we tested (Figure 5A)—and in particular decreased the membrane-associated pool of SNX3 (Figure S8B). By contrast, SNX3 knockdown had no effect on Hrs expression (Figure 5A and see Figure S8C). We thus wondered whether the known effect of Hrs knockdown could be due, at least in part, to reduced levels of SNX3.As expected [30], approximately 60% of the total HRP-labeled endosomes appeared to contain fewer internal membranes (Figure 5B, left panel; quantification in Figure 5D) in cells treated with Hrs siRNAs, much like endosomes in cells treated with SNX3 siRNAs (Figure 4B)—and in contrast to endosomes in mock-treated cells (Figure 4A). Parallel analysis of cells after proteinA-gold internalization revealed that these endosomes contained five to ten internal vesicles per profile, which were otherwise similar to controls (Figure 4A), whether microtubules were intact (Figure 5B, right panel) or not (unpublished data)—again much like after SNX3 knockdown and in contrast to controls (Figure 4A and 4B). The multivesicular morphology of endosomes in cells treated with Hrs siRNAs could, however, be restored by overexpression of GFP-SNX3 (Figure 5C; quantification in Figure 5D), despite that fact that Hrs expression remained silenced (Figure S8D). Structures labeled with 5-nm proteinA-gold then contained approximately 25–30 lumenal vesicles per profile, similar to endosomes in mock-treated controls. These data thus strongly suggest that the morphological phenotype of endosomes in Hrs-depleted cells is at least in part due to low SNX3 levels, and also demonstrate that SNX3 is a component of the molecular machinery that drives intralumenal membrane invagination.It has been suggested that EGF receptor is trafficked through a subpopulation of multivesicular endosomes in a process that involves annexin A1 [21]. Annexin A1 is structurally, functionally, and biochemically related to annexin A2 [47], and both proteins colocalize on early endosomes [48,49]. Annexin A2 was also proposed to play a role in multivesicular endosome biogenesis, but not in the invagination process [22]. Consistently, endogenous annexin A1 colocalized with annexin A2-GFP and mRFP-SNX3, and endogenous annexin A2 with mRFP-SNX3 (Figure S9A). These observations show that SNX3 is present on endosomes that contain both annexin A1 and annexin A2.Next, we investigated whether SNX3 plays a role in the formation of intralumenal vesicles that mediate EGF receptor sorting into multivesicular endosomes. To this end, we made use of the ability of the active Rab5 mutant Rab5Q79L to form enlarged early endosomes that provide high spatial resolution by light microscopy [9,50], as in Figure 3. When mock-treated cells were challenged with EGF for 15 min at 37 °C, the EGF receptor was endocytosed into these enlarged endosomes, where greater than 50% of the endocytosed receptor accumulated in the lumen (Figure 6A, quantification in Figure 6B), as expected [50]. Similarly, EGF colocalized with the receptor in the lumen of these enlarged endosomes (unpublished data). Knockdown of Hrs with either one of two siRNAs significantly reduced EGF receptor sorting into the lumen of enlarged endosomes (Figure 6A, quantification in Figure 6B), consistent with our electron microscopy analysis (Figure 5B–5D) and in agreement with previous findings [30,50]. Similarly, SNX3 depletion with either one of two siRNAs inhibited EGF receptor incorporation in the lumen of large endosomes to the same extent as Hrs knockdown (Figure 6A, quantification in Figure 6B). Finally, SNX3 re-expression in the Hrs knockdown background restored EGF receptor accumulation in the lumen of enlarged endosomes to the same extent as observed in mock-treated controls (Figure 6A, quantification in Figure 6B). These observations unambiguously demonstrate that SNX3 controls the formation of lumenal membranes that carry the EGF receptor, further confirming the role of SNX3 in the lumenal invagination process.Although reduced levels of SNX3 seemed to account for the invagination defect in Hrs knockdown cells, SNX3 siRNAs did not affect EGF receptor transport to late endosomes (Figure 4G–4H) and degradation (Figure 4E and 4F). Similarly, a wave of endocytosed EGF receptor was exported from EEA1-positive early endosomes and reached Lamp1-positive late endosomes (Figure 7A, quantification in Figure 7B) under our conditions of Hrs knockdown (≈80%, Figure 5A) much like in mock-treated cells (Figure S8A) or in cells treated with SNX3 siRNAs (Figure 4H). Since, under the same Hrs knockdown conditions, the formation of internal vesicles (Figures 5B and 6A) and EGF receptor degradation were inhibited (Figure 4E) and SNX3 levels reduced (Figure 5A), these observations strongly suggest that Hrs is an essential component of the lysosome targeting machinery, which can function independently of receptor sorting into and incorporation within multivesicular endosomes.Figure 7MEK1 Cleavage by Anthrax Toxin Lethal Factor(A and B) Experiments (A) and quantification (B) were as in Figure 4G and 4H, except that cells were treated with Hrs siRNAs. Insets in the lower right are a magnification of the regions shown in the boxes.(C) Hrs or SNX3 was knocked down, or Hrs was knocked down and GFP-SNX3 overexpressed. The trypsin-nicked Protective Antigen (500 ng/ml) and the Lethal Factor (100 ng/ml) of anthrax toxin were then bound to the cell surface on ice [51], and cells were incubated at 37 °C. Lysates (40 μg) were analyzed by western blotting using antibodies against MEK1 N-terminus to detect MEK1 cleavage by Lethal Factor released into the cytosol.In (B), each condition is the mean of at least three independent experiments; standard errors are indicated. (A) Scale bar indicates 10 μm.To further discriminate between Hrs and SNX3 functions, we made use of anthrax toxin, which is translocated across the membrane of ECV/MVB intralumenal vesicles. Like VSV nucleocapsids, the toxin hijacks these vesicles to reach late endosomes, where back fusion with the limiting membrane releases the lethal factor into the cytosol, leading to the cleavage of mitogen-activated protein kinase kinases (MAPKKs), and in particular MEK1 [51]. After addition of anthrax toxin, MEK1 cleavage was slightly retarded in cells treated with Hrs or SNX3 siRNAs (Figure 7C). Presumably, in the absence of intralumenal membranes (Figures 4B, 5B–5D, 6A, and 6B), toxin translocation could then occur across the limiting membrane of these “empty” ECV/MVBs—in good agreement with our observations on VSV capsid release after Hrs [44] or SNX3 (Figure 4D) knockdown. Interestingly, re-expression of GFP-SNX3 in the Hrs knockdown background prevented toxin translocation (Figure 7C). Presumably, the toxin, once released, remained trapped in the lumen of intralumenal vesicles. Similarly, EGF receptor degradation remained inhibited after re-expression of GFP-SNX3 in the Hrs knockdown background (Figure S9B). Indeed, excess SNX3 not only restored intralumenal vesicles (Figures 5C, 6A, and 6B), but also inhibited ECV/MVB detachment (or maturation), and thus transport beyond early endosomes towards late endosomes (Figures 1 and 2).DiscussionIt is generally believed that the lysosomal targeting of signaling receptors is controlled by sorting via Hrs and its downstream ESCRT partners into membrane invaginations on early endosomal regions, which will become multivesicular endosomes. Then, intralumenal vesicles are transported to lysosomes, where they are degraded together with their receptor cargo. Our observations now demonstrate that lysosomal targeting can function independently of this multivesicular endosome sorting event, since the transport of EGF receptor to late endocytic compartments and its degradation are not affected by SNX3 knockdown, in the absence of ECV/MVB lumenal membranes (see model in Figure S10). This agrees with previous observations that PI 3-kinase inhibition with wortmannin inhibits the formation of lumenal vesicles, but not the delivery of the EGF receptor to the lysosomes [35]. A corollary of these observations, however, is that Hrs indeed controls lysosome targeting, as expected [29,30,52], but that this function can be uncoupled from the invagination process in the multivesicular pathway towards late endosomes. Even in the absence of ECV/MVB lumenal membranes after SNX3 knockdown, sorting into putative Hrs platforms [53] is sufficient to ensure, not only receptor transport to late endosomes, but also packaging into lysosomes and degradation.Whereas Hrs is necessary for lysosomal targeting, the protein appears to be dispensable for the membrane invagination process that leads to ECV/MVB biogenesis. By contrast, SNX3, which is dispensable for lysosomal targeting, appears to function as a necessary component of the molecular machinery that drives the formation of membrane invaginations within forming ECV/MVBs. Indeed, SNX3 knockdown inhibits the invagination process, whereas excess SNX3 promotes the formation of intralumenal membranes and the accumulation of multivesicular regions on early endosomes. It thus appears that PtdIns3P signaling regulates differentially lysosomal targeting and the biogenesis of ECV/MVBs, via at least two PtdIns3P effectors, Hrs and SNX3, which act sequentially in the pathway. Although our data demonstrate that SNX3 regulates the membrane invagination process itself, the molecular mechanism is unclear. In yeast, the SNX3 homolog Grd19p is involved in selective retrieval of some membrane proteins from the prevacuolar compartment, presumably late endosomes, to the TGN [54,55]. It is not known whether the function of yeast Grd19p is somehow related to the function we propose for SNX3 in mammalian cells. However, it is possible that Grd19p and SNX3 play different roles, since endosome-to-Golgi transport of Shiga toxin B-subunit seems to require not SNX3 (this study), but SNX1 [43].Several proteins and lipids were reported to play a role in the formation of intralumenal membranes, including Hrs and ESCRTs [13,14], SNX3 (this study), ceramide [50], and perhaps, LBPA and Alix [56]. Future work will be needed to determine the precise relationships that may exist between these components. At present, our observations indicate that SNX3 functions downstream of Hrs, and one may speculate that it recruits other proteins, perhaps ESCRTs, which then drive the invagination process. However, SNX3 is a short member of this protein family that does not contain any structural feature other than the PX domain itself [38,40]. Moreover, we failed to detect ESCRT components in SNX3 pull-downs, and two-hybrid screens showed that Grd19p interacts with proteins of the early secretory pathway, but the significance of these observations is not clear [57]. Alternatively, SNX3 itself may play a direct role in the invagination process by deforming membranes in a direction opposite to BAR-containing proteins [39].Materials and MethodsCells, antibodies, reagents, and constructs.Cell maintenance was described [46], as was the mouse monoclonal anti-LBPA antibody [58]. We are very grateful to Ludger Johannes (Paris. France) for supplying us with Cy3-labeled Shiga toxin B-subunit; to Reinhard Jahn (Göttingen. Germany) and Volker Gerke (Muenster. Germany) for mouse monoclonal antibodies against RAB5 and annexin A2 (HH7), respectively; to Wanjin Hong (Singapore. Singapore), Harald Stenmark (Oslo, Norway), and Marino Zerial (Dresden, Germany) for rabbit polyclonal antibodies against SNX3 or SNX16, Hrs, and EEA1, respectively. We also used mouse monoclonal antibodies against: GFP (Roche Diagnostics); transferrin receptor (Zymed Laboratories); EEA1 and annexin A1 (Transduction Laboratories); human LAMP1 (Pharmacia); EGF receptor (BD Biosciences); α-tubulin (Sigma-Aldrich); ubiquitinated proteins (FK2) (Affinity Research Products); clathrin (X22) (ABR-Affinity BioReagents); and TSG101 (GeneTex). Polyclonal antibodies made in rabbit against Rab6 were from Santa Cruz Biotechnology and made in sheep against the EGF receptor from Fitzergald. HRP-labeled secondary antibodies were from Amersham or Sigma-Aldrich and fluorescently labeled secondary antibodies from Jackson Immunoresearch Laboratories. Wortmannin, nocodazole, EGF, HRP, and o-dianisidine were from Sigma-Aldrich, EGF-biotin, streptavidin R-phycoerythrin conjugate, AlexaFluor 488 EGF complex, 10-kDa rhodamine dextran from Molecular Probes. ProteinA-gold (5 nm) was from Utrecht University. We obtained pGFP-Rab5Q79L from Marino Zerial (Dresden. Germany), GFP-tagged SNX1 and Flag-tagged SNX2 from Gordon N. Gill (University of California, San Diego, California), myc-tagged SNX3 from Carol R. Haft (National Institutes of Health, Bethesda, Maryland), and pDMYC-SNX16 from Wanjin Hong (Singapore, Singapore). SNX1, SNX2, and SNX3 were amplified by PCR and introduced into pEGFP-C2 or fused with mRFP.Viral fusion, RNA replication, and infection.In experiments using VSV, the fusion of the viral envelope with endosomal membranes, the replication of viral RNA, and the appearance of the viral glycoprotein G in the host-cell biosynthetic pathway were analyzed as described [44]. Briefly, VSV fusion was measured using virus labeled with self-quenching amounts of the fluorescent dye Dil (long-chain dialkylcarbocyanine). Labeled virions were bound to the plasma membrane on ice and then cells were incubated for 35 min at 37 °C. Individual VSV fusion events resulted in the appearance of brightly fluorescent spots upon dye dequenching. For quantification, cells containing fused virions were counted. The replication of VSV RNA minus-strand was quantified by TaqMan RT-PCR; total RNA was extracted from infected HeLa cells, precipitated, and used for retrotranscription using specific oligonucleotides of the genomic VSV-G RNA. Finally, G-protein appearance in biosynthetic membranes 3 h postinfection was visualized by immunofluorescence using anti-VSV-G antibodies.Endocytic transport in vivo. Cells were transfected with FuGene (Roche Diagnostics) according to the manufacturer's instructions. In RNA interference (RNAi) experiments, cells were transfected twice at a 24-h interval using OligoFectamin (Invitrogen) with 21-nucleotide RNA duplexes (Qiagen), replated 4 h after the second transfection, and analyzed 36 h later. In HeLa cells, we used an Hrs target sequence siRNA1 that was described [30] and already used in our previous studies [44], as well the target sequence siRNA2 5′-AAGCGGAGGGAAAGGCCACTT-3′; the target SNX3 sequences were: siRNA1, 5′-AAGGGCTGGAGCAGTTTATAA-3′; siRNA2, 5′-AACAAGGGCTGGAGCAGTTTA-3′. To follow fluid phase transport, cells were incubated for 10 min at 37 °C with 2.5 mg/ml rhodamine-dextran (pulse) and further incubated for 40 min without the marker (chase) when indicated. Alternatively, cells were incubated for 1 h at 4 °C with 1 μM Cy3-labeled Shiga toxin B-subunit, or incubated for 1 h at 4 °C with 400 ng/ml EGF-biotin-streptavidin-R-phycoerythrin complex or 500 ng/ml EGF-biotin-streptavidin-AlexaFluor488 complex. Then, the marker was endocytosed for the indicated time periods at 37 °C. To quantify EGF receptor degradation, cells were preincubated in serum-free medium, and incubated with 0.25 μg/ml EGF and 10 μg/ml cycloheximide for the indicated time periods. Microtubules were depolymerized with 10 μM nocodazole for 2 h [46,59]. When indicated, cells were incubated for 1 h at 4 °C with 500 ng/ml trypsin-nicked Protective Antigen (PA) and 100 ng/ml Lethal Factor (LF) of anthrax toxin, and then transferred to 37 °C for different periods of time in a toxin-free medium, as described [51].Microscopy.Immunofluorescence microscopy has been described [60]. For annexin immunolabeling, cells were permeabilized before fixation with 100 mM KCl, 2 mM MgCl2, 1 mM CaCl2, and 1 mM Hepes (pH 6.9) containing 0.1% triton X-100 [61]. In some experiments, 500 ng/ml EGF-biotin-streptavidin-AlexaFluor488 complex was bound for 1 h at 4 °C to the surface of cells overexpressing Rab5Q79L, and then the complex was endocytosed for 15 min at 37 °C. The quantification of EGF receptor distribution in the lumen or on the limiting membrane of enlarged endosomes has been described [50]. Pictures were captured using a Zeiss Axiophot microscope equipped with a 63× Plan-NOEFLUAR objective or a Zeiss LSM 510 META confocal microscope equipped with a 63× Plan-Apochromat objective. To visualize the content of ECV/MVBs by electron microscopy, HRP was endocytosed for 15 min at 37 °C followed by a 30-min chase after microtubule depolymerization [22,46]. Cells were then fixed, HRP was revealed cytochemically with DAB as substrate, and processed for plastic embedding [62]. Alternatively, cells were incubated with 5-nm proteinA-gold (optical density [OD] at 520 nm = 10) for 30 min with or without microtubule depolymerization to label ECV/MVBs, and samples were processed as above. In this analysis, it was not possible to identify unambiguously cells that have been transfected. However, given the extent of knockdown observed on western blots (80%–90%), the corresponding protein (Hrs or SNX3) must be depleted in at least four out of five (or nine out of ten) cells, if depletion were complete in most cells, rather than partial in all cells. The latter view is consistent with observations that 100% of the cells were labeled with fluorescently labeled siRNAs (unpublished data). In any case, to ensure unbiased electron microscopy analysis and quantification of both HRP and proteinA-gold experiments, samples were prepared in Geneva and analyzed by western blotting to ensure that knockdown and re-expression of SNX3 were efficient. Then, samples were shipped to Brisbane. There, each experiment was number coded and analyzed by electron microscopy. The corresponding micrographs in number-coded folders were then transferred to Geneva. Analysis and quantification were then performed blind in Geneva. Electron microscopy after immunogold labeling of cryosections has been described [63].Other methods.LBPA quantification by ELISA [58] was described, as was early and late endosome fractionation by flotation in a sucrose step gradient [59].Supporting InformationFigure S1GFP-SNX3 Is Present on Early Endosomes and This Distribution Depends on PtdInsI3P(A) HeLa cells expressing GFP-SNX3 were processed for immunofluorescence using the indicated antibodies.(B) BHK cells expressing GFP-SNX3 were homogenized and a postnuclear supernatant (PNS) was prepared. From this PNS, early endosomes (EE) were separated from late endosomes (LE) and heavy membranes (HM) by floatation in a sucrose step gradient [59,64]. Fractions were analyzed by SDS gel electrophoresis and western blotting, using antibodies against GFP to detect GFP-SNX3 or against the early endosomal marker Rab5 (upper panel), and by ELISA using antibodies against the late endosomal marker LBPA (lower panel).(C) HeLa cells expressing GFP-SNX3 or the 3-phosphoinositide-binding defective mutant GFP-SNX3R70A were treated or not with 100 nM wortmannin for 30 min at 37 °C and then analyzed by fluorescence microscopy. SNX3R70A was no longer membrane-associated, much like the SNX3 R69RY to AAA triple mutant and Y71A single mutant [40].(A and C) Scale bar indicates 10 μm.(516 KB PDF)Click here for additional data file.Figure S2The Internalization of EGFR, Dextran, and Shiga Toxin B-Subunit Is Not Affected in Cells Expressing GFP-SNX3(A) HeLa cells expressing GFP-SNX3 were incubated with biotin-EGF coupled with streptavidin-R-phycoerythrin for 1 h at 4 °C, chased for 10 min at 37 °C, and analyzed by fluorescence microscopy after labeling with antibodies against EEA1.(B) HeLa cells expressing GFP-SNX3 were incubated with rhodamine-dextran for 10 min at 37 °C, and analyzed as in (A).(C) HeLa cells expressing GFP-SNX3 were incubated with 1 μM Shiga toxin B-subunit conjugated to Cy3 for 1 h at 4 °C, chased for 10 min at 37 °C, and analyzed by fluorescence microscopy after labeling with antibodies against the transferrin receptor.(A–C) Scale bar indicates 10 μm.(703 KB PDF)Click here for additional data file.Figure S3SNX3 in Endosomal Transport(A) After cell surface binding, biotin-EGF coupled to streptavidin-R-phycoerythrin was internalized for 50 min at 37 °C in HeLa cells expressing GFP-SNX3. Cells were labeled with antibodies against Lamp1 and analyzed by triple channel fluorescence; the various combinations of merged colors are shown for the micrographs in Figure 1C.(B) The experiment was as in (A), but cells were labeled with antibodies against EEA1. The various combinations of merged colors are shown for the micrographs in Figure 1D.(C) Rhodamine-dextran was pulsed for 10 min at 37 °C in HeLa cells expressing GFP-SNX3 and then chased for 40 min. Cells were labeled with antibodies against Lamp1. The various combinations of merged colors are shown for the micrographs in Figure 1F.(D) After cell surface binding, Shiga toxin B-subunit conjugated to Cy3 was internalized for 50 min at 37 °C into HeLa cells expressing GFP-SNX3, and cells were analyzed using antibodies against Rab6. The various combinations of merged colors are shown for the micrographs in Figure 1G.(A–D) Scale bar indicates 10 μm(1.76 MB PDF)Click here for additional data file.Figure S4EGFR, Dextran, and Shiga Toxin B-Subunit Transport in Control Cells(A) HeLa cells were incubated with biotin-EGF coupled with streptavidin-R-phycoerythrin for 1 h at 4 °C and then chased for 50 min at 37 °C. Cells were then labeled with antibodies against EEA1 and analyzed by fluorescence microscopy.(B) HeLa cells were incubated with rhodamine-dextran for 10 min at 37 °C, chased for 40 min, and then analyzed by fluorescence microscopy after labeling with antibodies against Lamp1.(C) HeLa cells were incubated with 1 μM Shiga toxin B-subunit conjugated to Cy3 for 1 h at 4 °C and then chased for 50 min at 37 °C. Cells were then labeled with antibodies against Rab6 and analyzed by fluorescence microscopy.(A–C) Scale bar indicates 10 μm.(722 KB PDF)Click here for additional data file.Figure S5Overexpression of SNX1, SNX2, or SNX16 Does Not Affect Transport of the EGF Receptor to Late Endosomes and LysosomesHeLa cells expressing GFP-SNX1 (A), GFP-SNX2 (B), or myc-SNX16 (C) were incubated with biotin-EGF coupled with streptavidin-R-phycoerythrin for 1 h at 4 °C and then chased for 50 min at 37 °C. Cells were then labeled with antibodies against Lamp1 (A–C) or myc (C) and analyzed by fluorescence microscopy.(A–C) Scale bar indicates 10 μm.(535 KB PDF)Click here for additional data file.Figure S6GFP-SNX3 Expression Causes an Expansion of the Multivesicular Regions of Early EndosomesControl HeLa cells (A) or HeLa cells expressing GFP-SNX3 (B) were fixed, embedded in Epon, and then processed for electron microscopy. Arrows point at structures with a characteristic multivesicular appearance. In (B), arrowheads surround a group of seven or eight multivesicular structures—such clusters were frequently observed in cells overexpressing GFP-SNX3 (quantification is shown in Figure 2E). Scale bar indicates 1 μm.(403 KB PDF)Click here for additional data file.Figure S7Early-to-Late Endosomal Transport Is Not Affected in Cells Treated with SNX3 siRNAs, and Depends on an Intact Microtubule Network(A and B) HeLa cells were treated with SNX3 siRNAs, and then microtubules were depolymerized (B) or not (A) with 10 μM nocodazole for 2 h. EGF-biotin coupled to streptavidin-AlexaFluor 488 was endocytosed for 50 min at 37 °C, and cells were analyzed by immunofluorescence using antibodies against Lamp1. Scale bar indicates 10 μm.(C) HeLa cells expressing GFP-SNX3 were incubated with 0.25 μg/ml EGF and 10 μg/ml cycloheximide for the indicated time periods, as in Figures 1A and 4E. Cell lysates (100 μg) were analyzed by SDS gel electrophoresis and western blotting with an antibody raised against the purified recombinant partial cytoplasmic domain of human EGFR, as in Figures 1A and 4E. A blot of the complete gel is shown to illustrate the fact that degradation intermediates were not detected with this antibody. The arrow points at a nonspecific band.(555 KB PDF)Click here for additional data file.Figure S8Treatment with siRNAs(A) HeLa cells treated with mock siRNAs were incubated with 500 ng/ml EGF-biotin coupled with streptavidin-AlexaFluor 488 for 1 h at 4 °C, chased for 50 min at 37 °C, and analyzed by fluorescence microscopy after labeling with antibodies against EEA1 or Lamp1, as indicated. The quantification of these experiments is shown in Figure 4G. Scale bar indicates 10 μm.(B) HeLa cells were mock-treated or treated with siRNAs against Hrs or SNX3 for 72 h. After homogenization, postnuclear supernatants were prepared and fractionated by high-speed centrifugation. Membranes and cytosol were recovered from the high-speed pellet (HSP) and high-speed supernatant (HSS), respectively, and analyzed by SDS gel electrophoresis and western blotting using the indicated antibodies.C) HeLa cells were mock-treated or treated with two different siRNAs against SNX3 (1 and 2) or against Hrs (Hrs2) for 72 h. Cells were harvested and total lysates were analyzed by SDS gel electrophoresis and western blotting using antibodies against Hrs, SNX3, and actin.(D) HeLa cells were mock-treated or treated with siRNAs against Hrs for 72 h. GFP-SNX3 was overexpressed during the last 24 h of the Hrs RNAi treatment. Cells were then harvested and total lysates were analyzed by SDS gel electrophoresis and western blotting using antibodies against α-tubulin, Hrs, and SNX3.(557 KB PDF)Click here for additional data file.Figure S9GFP-SNX3 Is Present on Early Endosomes Containing Both Annexin A1 and Annexin A2(A) HeLa cells expressing mRFP-SNX3 and/or GFP-Anx A2 were fixed [61] and processed for immunofluorescence using the indicated antibodies. Scale bar indicates 10 μm.(B) HeLa cells were mock-treated or treated with siRNAs against Hrs with or without GFP-SNX3 overexpression (as in Figures 5C and 6), and then incubated with 0.25 μg/ml EGF and 10 μg/ml cycloheximide for the indicated time periods. Cell lysates (100 μg) were analyzed by SDS gel electrophoresis and western blotting with antibodies against EGFR (as in Figures 1A and S7C) or actin.(1.10 MB PDF)Click here for additional data file.Figure S10ModelThe current view is that, after endocytosis, the ubiquitinated EGF receptor (EGF-R) is sorted into membrane regions of the early endosome that are coated with clathrin and enriched in PtdInsI3P (sorting platforms). Then, EGF-R interacts sequentially with ESCRT I, II, and III complexes and is incorporated into membrane invaginations (in green), giving rise to a forming ECV/MVB [14,37]. The lumenal vesicles are transported to late endosomes and eventually to lysosomes, where they are degraded together with their receptor cargo. Our previous studies show that VSV envelope fuses with ECV/MVB lumenal membranes, thereby delivering the viral capsid to the lumen of these internal vesicles, where it remains hidden [44]. After delivery to late endosomes, back fusion with the limiting membrane ensures capsid delivery to the cytoplasm. The delivery of anthrax toxin lethal factor to the cytoplasm [51] follows a pathway very similar to that of the nucleocapsid (not indicated here for the sake of simplicity).(A) SNX3 Overexpression. Our observations indicate that SNX3 overexpression causes an accumulation of multivesicular regions on early endosomes and inhibits ECV/MVB detachment from early endosomes (or maturation) and thus interferes with more-distal transport steps. Then, VSV and EGF-R remain in early endosomes and viral RNA release is inhibited. By contrast, Hrs overexpression was reported to mimic its down-expression by RNAi [11,65].(B) SNX3 RNAi. We find that SNX3 knockdown reduces the formation of ECV/MVB lumenal vesicles, but not EGF-R transport to late endosomes and degradation in lysosomes, indicating that EGF-R sorting can be uncoupled from its incorporation into ECV/MVB lumenal vesicles. Presumably, Hrs (which is not affected by SNX3 knockdown) then ensures proper EGFR targeting to the lysosomes. Consistently, the delivery of viral RNA to the cytoplasm after SNX3 knockdown no longer requires transport to late endosomes, presumably because virions then fuse with the limiting membrane of “empty” ECV/MVBs, devoid of internal vesicles.(C) Hrs RNAi. Hrs knockdown reduces the formation of ECV/MVB lumenal vesicles [20,29,30] and causes a concomitant decrease in SNX3, which is likely to be responsible for the reduction in lumenal vesicles. Indeed, these can be restored by SNX3 re-expression after Hrs knockdown. After Hrs knockdown, the delivery of viral RNA to the cytoplasm no longer requires transport to late endosomes, because virions then fuse with the limiting membrane of “empty” ECV/MVBs [44]—much like after SNX3 knockdown. In addition, EGF-R transport to late endosomes is not affected, again much like after SNX3 knockdown. These data together with observations from others that EGF-R can be transported towards lysosomes after Hrs knockdown [20] suggest that Hrs functions are dispensable for early-to-late endosomal transport. Then, EGF-R sorting in early endosomes may be achieved by other components of the sorting machinery. Alternatively, a disruption of this machinery may allow uncontrolled EGF-R incorporation into the degradation pathway, including, for example, nonubiquitinated molecules. In any case, and in marked contrast to SNX3 knockdown, we find that EGF-R degradation in lysosomes is seriously compromised by Hrs knockdown, as observed by others [29,30], confirming the crucial function of Hrs for EGF-R lysosomal targeting.(46 KB PDF)Click here for additional data file.\n\nREFERENCES:\n1. KirkhamMPartonRG\n2005\nClathrin-independent endocytosis: new insights into caveolae and non-caveolar lipid raft carriers\nBiochim Biophys Acta\n1746\n350\n363\n2. ConnerSDSchmidSL\n2003\nRegulated portals of entry into the cell\nNature\n422\n37\n44\n12621426\n3. Di GuglielmoGMLe RoyCGoodfellowAFWranaJL\n2003\nDistinct endocytic pathways regulate TGF-beta receptor signalling and turnover\nNat Cell Biol\n5\n410\n421\n12717440\n4. ChenHDe CamilliP\n2005\nThe association of epsin with ubiquitinated cargo along the endocytic pathway is negatively regulated by its interaction with clathrin\nProc Natl Acad Sci U S A\n102\n2766\n2771\n15701696\n5. SigismundSWoelkTPuriCMasperoETacchettiC\n2005\nClathrin-independent endocytosis of ubiquitinated cargos\nProc Natl Acad Sci U S A\n102\n2760\n2765\n15701692\n6. MaxfieldFRMcGrawTE\n2004\nEndocytic recycling\nNat Rev Mol Cell Biol\n5\n121\n132\n15040445\n7. van der GootFGGruenbergJ\n2006\nIntra-endosomal membrane traffic\nTrends Cell Biol\n16\n514\n521\n16949287\n8. MukhopadhyayDRiezmanH\n2007\nProteasome-independent functions of ubiquitin in endocytosis and signaling\nScience\n315\n201\n205\n17218518\n9. RaiborgCBacheKGGilloolyDJMadshusIHStangE\n2002\nHrs sorts ubiquitinated proteins into clathrin-coated microdomains of early endosomes\nNat Cell Biol\n4\n394\n398\n11988743\n10. SachseMUrbeSOorschotVStrousGJKlumpermanJ\n2002\nBilayered clathrin coats on endosomal vacuoles are involved in protein sorting toward lysosomes\nMol Biol Cell\n13\n1313\n1328\n11950941\n11. UrbeSSachseMRowPEPreisingerCBarrFA\n2003\nThe UIM domain of Hrs couples receptor sorting to vesicle formation\nJ Cell Sci\n116\n4169\n4179\n12953068\n12. BabstM\n2005\nA protein's final ESCRT\nTraffic\n6\n2\n9\n15569240\n13. HurleyJHEmrSD\n2006\nThe ESCRT complexes: structure and mechanism of a membrane-trafficking network\nAnnu Rev Biophys Biomol Struct\n35\n277\n298\n16689637\n14. SlagsvoldTPattniKMalerodLStenmarkH\n2006\nEndosomal and non-endosomal functions of ESCRT proteins\nTrends Cell Biol\n16\n317\n326\n16716591\n15. HislopJNMarleyAVon ZastrowM\n2004\nRole of mammalian vacuolar protein-sorting proteins in endocytic trafficking of a non-ubiquitinated G protein-coupled receptor to lysosomes\nJ Biol Chem\n279\n22522\n22531\n15024011\n16. McNattMWMcKittrickIWestMOdorizziG\n2007\nDirect binding to Rsp5 mediates ubiquitin-independent sorting of Sna3 via the multivesicular body pathway\nMol Biol Cell\n18\n697\n706\n17182850\n17. OestreichAJAboianMLeeJAzmiIPayneJ\n2007\nCharacterization of multiple multivesicular body sorting determinants within Sna3: a role for the ubiquitin ligase Rsp5\nMol Biol Cell\n18\n707\n720\n17182849\n18. WatsonHBonifacinoJS\n2007\nDirect binding to Rsp5p regulates ubiquitination-independent vacuolar transport of Sna3p\nMol Biol Cell\n18\n1781\n1789\n17332499\n19. TheosACTruschelSTTenzaDHurbainIHarperDC\n2006\nA lumenal domain-dependent pathway for sorting to intralumenal vesicles of multivesicular endosomes involved in organelle morphogenesis\nDev Cell\n10\n343\n354\n16516837\n20. RaziMFutterCE\n2006\nDistinct roles for Tsg101 and Hrs in multivesicular body formation and inward vesiculation\nMol Biol Cell\n17\n3469\n3483\n16707569\n21. WhiteIJBaileyLMAghakhaniMRMossSEFutterCE\n2006\nEGF stimulates annexin 1-dependent inward vesiculation in a multivesicular endosome subpopulation\nEMBO J\n25\n1\n12\n16052208\n22. MayranNPartonRGGruenbergJ\n2003\nAnnexin II regulates multivesicular endosome biogenesis in the degradation pathway of animal cells\nEMBO J\n22\n3242\n3253\n12839987\n23. VonderheitAHeleniusA\n2005\nRab7 associates with early endosomes to mediate sorting and transport of Semliki forest virus to late endosomes\nPLoS Biol\n3\ne233\ndoi:10.1371/journal.pbio.0030233\n15954801\n24. RinkJGhigoEKalaidzidisYZerialM\n2005\nRab conversion as a mechanism of progression from early to late endosomes\nCell\n122\n735\n749\n16143105\n25. PornillosOGarrusJESundquistWI\n2002\nMechanisms of enveloped RNA virus budding\nTrends Cell Biol\n12\n569\n579\n12495845\n26. von SchwedlerUKStuchellMMullerBWardDMChungHY\n2003\nThe protein network of HIV budding\nCell\n114\n701\n713\n14505570\n27. Martin-SerranoJ\n2007\nThe role of ubiquitin in retroviral egress\nTraffic\n8\n1297\n1303\n17645437\n28. Martin-SerranoJYarovoyAPerez-CaballeroDBieniaszPD\n2003\nDivergent retroviral late-budding domains recruit vacuolar protein sorting factors by using alternative adaptor proteins\nProc Natl Acad Sci U S A\n100\n12414\n12419\n14519844\n29. LloydTEAtkinsonRWuMNZhouYPennettaG\n2002\nHrs regulates endosome membrane invagination and tyrosine kinase receptor signaling in Drosophila\nCell\n108\n261\n269\n11832215\n30. BacheKGBrechAMehlumAStenmarkH\n2003\nHrs regulates multivesicular body formation via ESCRT recruitment to endosomes\nJ Cell Biol\n162\n435\n442\n12900395\n31. DoyotteARussellMRGHopkinsCRWoodmanPG\n2005\nDepletion of TSG101 forms a mammalian ‘Class E' compartment: a multicisternal early endosome with multiple sorting defects\nJ Cell Sci\n118\n3003\n3017\n16014378\n32. McMahonHTGallopJL\n2005\nMembrane curvature and mechanisms of dynamic cell membrane remodelling\nNature\n438\n590\n596\n16319878\n33. PetiotAFaureJStenmarkHGruenbergJ\n2003\nPI3P signaling regulates receptor sorting but not transport in the endosomal pathway\nJ Cell Biol\n162\n971\n979\n12975344\n34. Fernandez-BorjaMWubboltsRCalafatJJanssenHDivechaN\n1999\nMultivesicular body morphogenesis requires phosphatidyl-inositol 3-kinase activity\nCurr Biol\n9\n55\n58\n9889123\n35. FutterCECollinsonLMBackerJMHopkinsCR\n2001\nHuman VPS34 is required for internal vesicle formation within multivesicular endosomes\nJ Cell Biol\n155\n1251\n1264\n11756475\n36. OdorizziGBabstMEmrSD\n2000\nPhosphoinositide signaling and the regulation of membrane trafficking in yeast\nTrends Biochem Sci\n25\n229\n235\n10782093\n37. GruenbergJStenmarkH\n2004\nThe biogenesis of multivesicular endosomes\nNat Rev Mol Cell Biol\n5\n317\n323\n15071556\n38. WorbyCADixonJE\n2002\nSorting out the cellular functions of sorting nexins\nNat Rev Mol Cell Biol\n3\n919\n931\n12461558\n39. PeterBJKentHMMillsIGVallisYButlerPJ\n2004\nBAR domains as sensors of membrane curvature: the amphiphysin BAR structure\nScience\n303\n495\n499\n14645856\n40. XuYHortsmanHSeetLWongSHHongW\n2001\nSNX3 regulates endosomal function through its PX-domain-mediated interaction with PtdIns(3)P\nNat Cell Biol\n3\n658\n666\n11433298\n41. KurtenRCCadenaDLGillGN\n1996\nEnhanced degradation of EGF receptors by a sorting nexin, SNX1\nScience\n272\n1008\n1010\n8638121\n42. HansonBJHongW\n2003\nEvidence for a role of SNX16 in regulating traffic between the early and later endosomal compartments\nJ Biol Chem\n278\n34617\n34630\n12813048\n43. BujnyMVPopoffVJohannesLCullenPJ\n2007\nThe retromer component sorting nexin-1 is required for efficient retrograde transport of Shiga toxin from early endosome to the trans Golgi network\nJ Cell Sci\n120\n2010\n2021\n17550970\n44. Le BlancILuyetPPPonsVFergusonCEmansN\n2005\nEndosome-to-cytosol transport of viral nucleocapsids\nNat Cell Biol\n7\n653\n664\n15951806\n45. WhitneyJAGomezMSheffDKreisTEMellmanI\n1995\nCytoplasmic coat proteins involved in endosome function\nCell\n83\n703\n713\n8521487\n46. GruenbergJGriffithsGHowellKE\n1989\nCharacterization of the early endosome and putative endocytic carrier vesicles in vivo and with an assay of vesicle fusion in vitro\nJ Cell Biol\n108\n1301\n1316\n2538480\n47. GerkeVMossSE\n2002\nAnnexins: from structure to function\nPhysiol Rev\n82\n331\n371\n11917092\n48. SeemannJWeberKGerkeV\n1997\nAnnexin I targets S100C to early endosomes\nFEBS Lett\n413\n185\n190\n9287141\n49. EmansNGorvelJPWalterCGerkeVKellnerR\n1993\nAnnexin II is a major component of fusogenic endosomal vesicles\nJ Cell Biol\n120\n1357\n1369\n8449982\n50. TrajkovicKHsuCChiantiaSRajendranLWenzelD\n2008\nCeramide triggers budding of exosome vesicles into multivesicular endosomes\nScience\n319\n1244\n1247\n18309083\n51. AbramiLLindsayMPartonRGLepplaSHvan der GootFG\n2004\nMembrane insertion of anthrax protective antigen and cytoplasmic delivery of lethal factor occur at different stages of the endocytic pathway\nJ Cell Biol\n166\n645\n651\n15337774\n52. KomadaMSorianoP\n1999\nHrs, a FYVE finger protein localized to early endosomes, is implicated in vesicular traffic and required for ventral folding morphogenesis\nGenes Dev\n13\n1475\n1485\n10364163\n53. RaiborgCWescheJMalerodLStenmarkH\n2006\nFlat clathrin coats on endosomes mediate degradative protein sorting by scaffolding Hrs in dynamic microdomains\nJ Cell Sci\n119\n2414\n2424\n16720641\n54. HettemaEHLewisMJBlackMWPelhamHR\n2003\nRetromer and the sorting nexins Snx4/41/42 mediate distinct retrieval pathways from yeast endosomes\nEMBO J\n22\n548\n557\n12554655\n55. VoosWStevensTH\n1998\nRetrieval of resident late-Golgi membrane proteins from the prevacuolar compartment of Saccharomyces cerevisiae is dependent on the function of Grd19p\nJ Cell Biol\n140\n577\n590\n9456318\n56. MatsuoHChevallierJMayranNLe BlancIFergusonC\n2004\nRole of LBPA and Alix in multivesicular liposome formation and endosome organization\nScience\n303\n531\n534\n14739459\n57. VollertCSUetzP\n2004\nThe phox homology (PX) domain protein interaction network in yeast\nMol Cell Proteomics\n3\n1053\n1064\n15263065\n58. KobayashiTStangEFangKSde MoerloosePPartonRG\n1998\nA lipid associated with the antiphospholipid syndrome regulates endosome structure and function\nNature\n392\n193\n197\n9515966\n59. AnientoFEmansNGriffithsGGruenbergJ\n1993\nCytoplasmic dynein-dependent vesicular transport from early to late endosomes\nJ Cell Biol\n123\n1373\n1387\n8253838\n60. GuFAnientoFPartonRGGruenbergJ\n1997\nFunctional dissection of COP-I subunits in the biogenesis of multivesicular endosomes\nJ Cell Biol\n139\n1183\n1195\n9382865\n61. MorelEGruenbergJ\n2007\nThe p11/S100A10 light chain of annexin A2 is dispensable for annexin A2 association to endosomes and functions in endosomal transport\nPLoS ONE\n2\ne1118\ndoi:10.1371/journal.pone.0001118\n17971878\n62. PartonRGSchrotzPBucciCGruenbergJ\n1992\nPlasticity of early endosomes\nJ Cell Sci\n103\nPt 2\n335\n348\n1478937\n63. GriffithsGMcDowallABackRDubochetJ\n1984\nOn the preparation of cryosections for immunocytochemistry\nJ Ultrastruct Res\n89\n65\n78\n6544882\n64. GorvelJPChavrierPZerialMGruenbergJ\n1991\nrab5 controls early endosome fusion in vitro\nCell\n64\n915\n925\n1900457\n65. BishopNHormanAWoodmanP\n2002\nMammalian class E vps proteins recognize ubiquitin and act in the removal of endosomal protein-ubiquitin conjugates\nJ Cell Biol\n157\n91\n101\n11916981\n66. MariMBujnyMVZeuschnerDGeertsWJGriffithJ\n2008\nSNX1 defines an early endosomal recycling exit for sortilin and mannose 6-phosphate receptors\nTraffic\n9\n380\n393\n18088323"
+}

batch_3/PMC2528150.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2528150",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2528150\nAUTHORS: A H G Cleven, B G Wouters, B Schutte, A J G Spiertz, M van Engeland, A P de Bruïne\n\nABSTRACT:\nStromal expression of hypoxia inducible factor 2α (HIF-2α) and carbonic anhydrase 9 (CA9) are associated with a poorer prognosis in colorectal cancer (CRC). Tumour cell death, regulated by a hypoxic stromal microenvironment, could be of importance in this respect. Therefore, we correlated apoptosis, TP53 mutational status and BNIP3 promoter hypermethylation of CRC cells with HIF-2α- and CA9-related poor outcome. In a series of 195 CRCs, TP53 mutations in exons 5–8 were analysed by direct sequencing, and promoter hypermethylation of BNIP3 was determined by methylation-specific PCR. Expressions of HIF-2α, CA9, p53, BNIP3 and M30 were analysed immunohistochemically. Poorer survival of HIF-2α and CA9 stromal-positive CRCs was associated with wild-type TP53 (P=0.001 and P=0.0391), but not with BNIP3 methylation. Furthermore, apoptotic levels were independent of the TP53 status, but lower in unmethylated BNIP3 CRCs (P=0.004). It appears that wild-type TP53 in CRC cells favours the progression of tumours expressing markers for hypoxia in their stroma, rather than in the epithelial compartment. Preserved BNIP3 function in CRC cells lowers apoptosis, and may thus be involved in alternative cell death pathways, such as autophagic cell death. However, BNIP3 silencing in tumour cells does not impact on hypoxia-driven poorer prognosis.These results suggest that the biology of CRC cells can be modified by alterations in the tumour microenvironment under conditions of tumour hypoxia.\n\nBODY:\nHypoxia has been reported to influence tumour biology in opposing ways. It can directly induce cell death by activating apoptosis or autophagy, yet hypoxic zones in solid tumours also harbour viable cells resistant to treatment, which contributes to poor patient outcome (Erler et al, 2006). Hypoxia influences the expression of genes involved in cell death and energy homeostasis mainly by stabilisation and activation of the hypoxia inducible factor (HIF) family of transcription factors, influencing angiogenesis (VEGF), glycolysis (GLUT1), pH regulation (CA9), apoptosis (TP53, BNIP3) and autophagy (BNIP3) (Bacon and Harris, 2004; Keith and Simon, 2007).A common hallmark of solid tumours under hypoxic stress is increased ATP requirement, which is supplied by the induction of anaerobic glycolysis. This subsequently leads to a high production of intracellular lactate, requiring regulation of intracellular pH, a process mediated partly by HIF-dependent upregulation of carbonic anhydrase 9 (CA9). Carbonic anhydrase 9 catalyzes the extracellular trapping of acid by hydrating cell-generated CO2 into HCO3− and H+ (Swietach et al, 2007). Consequently, CA9 expression may serve as a marker for metabolic adaptation during hypoxia.Upregulation of p53 and BNIP3 proteins, which stimulate cell death via apoptosis and/or autophagy, appears contradictory to the adverse association between tumour hypoxia and prognosis. However, it was suggested that through induction of cell death, hypoxia selects for cells with defective cell death regulators, such as TP53 (Graeber et al, 1996). In non-selected cells hypoxia can induce expression of p53 and activate p53-mediated G0/G1 arrest or apoptosis, although secondary effects, such as extracellular acidosis and glucose deprivation, are necessary for p53-mediated apoptosis (Graeber et al, 1996; Schmaltz et al, 1998; Pan et al, 2004; Toledo and Wahl, 2006). Over 50% of human tumours contain somatic mutations in the TP53 gene, resulting in a defective apoptotic response (Kato et al, 2003; Soussi and Lozano, 2005). Therefore, TP53 mutations are expected to decrease the susceptibility of tumour cells to hypoxia-induced cell death, as shown in vitro (Graeber et al, 1996; Fei et al, 2004). BNIP3 is a Bcl-2 family member, containing a single BcL-2 homology 3 (BH3) domain and a transmembrane domain localising it to the outer mitochondrial membrane (Vande Velde et al, 2000; Lee and Paik, 2006). It is activated by HIF during hypoxia and initiates programmed cell death through apoptosis or autophagy (Ray et al, 2000; Mellor and Harris, 2007). Epigenetic silencing of BNIP3 by promoter hypermethylation has been reported in several cancer types and contributes to resistance to hypoxia-induced cell death (Okami et al, 2004; Abe et al, 2005; Yan et al, 2006). The role of BNIP3 in colorectal cancer (CRC) is unknown, although 66% tumours show BNIP3 silencing by promoter hypermethylation (Murai et al, 2005; Bacon et al, 2007).Usually, the effects of hypoxia in solid tumours are studied within the tumour cells themselves, neglecting the hypoxic response in tumour-associated stroma. In a previous study, we found that hypoxia within the tumour-associated stroma is indeed correlated with a poorer outcome in patients with CRC who are treated by surgery alone. In a multivariate model, stromal expressions of both HIF-2α and CA9 were independent adverse prognostic factors, whereas HIF-1α was not. Furthermore, expression of hypoxia-related proteins HIF-1α, GLUT1 and CA9 in the tumour cells self was not associated with poorer patient survival (Cleven et al, 2007).Our previous findings indicate that hypoxic conditions may modulate the tumour stroma in such a way that a more aggressive tumour behaviour is facilitated, ultimately leading to decreased patient survival.This study attempts to elucidate whether changes in the epithelial cell compartment of CRC, such as apoptosis and concomitant (epi)genetic changes that are confined to the tumour cells, are related to hypoxia-related changes in the stromal compartment. For this purpose, we correlated alterations of TP53 and BNIP3 in tumour cells with expression of hypoxia-related proteins HIF-2α and CA9 in relation with patient outcome and apoptotic activity in CRCs.Patients and methodsPatient populationPatients were registered for two multicentre prospective clinical trials in The Netherlands between 1979 and 1981. One trial was designed to compare patient survival after treatment of colon cancer by conventional surgery or the no-touch isolation technique (Wiggers et al, 1988). The second trial was conducted to compare survival in rectal cancer patients with or without preoperative radiotherapy. In the current study, we included only the patients who did not undergo preoperative radiotherapy. At the time the trial was conducted, only surgical removal of the tumours was performed, and adjuvant chemotherapy was not yet a standard practice. This study population therefore enables unbiased study of the influence of hypoxic conditions on tumour biology.Tumour tissues were fixed in buffered formalin, sectioned and embedded in paraffin. Experienced pathologists documented the histopathological characteristics of the tumours (Table 1). Follow-up took place every 3 months during the first 3 years and every 6 months between 3 and 5 years after initial diagnosis and surgery. Standard protocols were followed, with routine blood counts and chemical studies (including CEA levels) at each visit, and liver ultrasound, chest X-ray and colonoscopy annually, to evaluate recurrence of disease and disease-related death. After a 5-year follow-up period, only time and cause of death were registered. Follow-up was complete for all patients. Failure was defined as death due to recurrent disease, excluding postoperative mortality within 30 days and non-disease-related death.For immunohistochemical and molecular analyses, tumour tissues from 195 CRC patients were available. The distribution of age, gender, tumour stage, location and type of tumour, frequency of events and mean follow-up time of the patients in this study are representative of the patients in the trial (see Table 1).Genomic DNA isolationGenomic DNA was extracted from CRC tissues using PureGene™ genomic DNA isolation kit (Gentra Systems, Minneapolis, MN, USA) based on the manufacturer's protocol.TP53 sequencingMutation analyses of TP53 exons 5–8 were performed using a semi-nested PCR approach, (see Supplementary Table 1 for primer sequences). Caco2 (exon 6, codon 204 nonsense mutation) was included as a control. Direct sequencing of PCR products was performed using the BigDye® terminator v1.1 cycle sequencing kit (Applied Biosystems, Foster City, CA, USA) and analysed on the ABI 3730 DNA Analyzer (Applied Biosystems). Mutation was detected using Mutation Surveyor DNA Variant Analysis Software v3.0 (SoftGenetics LLC, State College, PA, USA). The results of the mutation analyses are listed in Table 2. Furthermore, we assessed whether TP53 missense mutants were transcriptionally active, on the basis of the IARC prediction models (http://p53.iarc.fr/MutationValidationCriteria.asp). Missense mutations were classified as either transactivation-incompetent or transactivation-competent missense mutations.BNIP3 promoter methylation analysisBNIP3 promoter methylation was determined by sodium bisulphite modification of genomic DNA using the EZ DNA methylation kit (ZYMO Research Co., Orange, CA). Methylation-specific PCR was performed as described in detail elsewhere (Herman et al, 1996; Derks et al, 2004). DNA was first amplified with BNIP3 flanking PCR primers that amplify bisulfite-modified DNA but do not preferentially amplify methylated or unmethylated DNA. The resulting template was used as the template for BNIP3 methylation-specific PCR. For primer sequences see Supplementary Table 1. All PCRs were performed with a control for unmethylated BNIP3 alleles (normal lymphocyte DNA), a positive control for methylated BNIP3 alleles (Sssl methyltransferase (New England Biolabs, Beverly, MA, USA)-treated normal lymphocyte DNA) and a negative control without DNA. Each PCR product was loaded onto a 2% agarose gel, stained with Gelstar® (Cambrex Bioscience Rockland Inc., Rockland, ME, USA) and visualised under UV illumination.ImmunohistochemistrySerial formalin-fixed, paraffin-embedded tissues sections (4 μm) were stained for HIF-2α and CA9, as described previously (Cleven et al, 2007). Brief descriptions are as follows.HIF-2α stainingAntigen retrieval was performed by microwave treatment (750 W for 20 min in 1 mM TE buffer, pH 8.0), followed by cooling in buffer for 30 min. Slides were blocked in 25% normal serum for 10 min. Sections were incubated with primary antibody HIF-2α (1 : 500) for 100 min (anti-HIF-2α monoclonal: ab8365; AbCam, Cambridge, UK).CA9 stainingSlides were blocked in 25% normal serum for 10 min, and then incubated for 45 min with primary CA9 antibody MoAb M75 (1 : 50, anti-human CA9; kindly supplied by Dr S Pastorekova) at room temperature.In addition to the above-mentioned staining procedures, serial sections were stained for p53, BNIP3 and M30, as follows:p53 stainingAntigen retrieval was performed by microwave treatment (750 W for 15 min in Antigen Retrieval (DAKO, Glostrup, Denmark)), followed by cooling in buffer for 30 min. Slides were blocked in 25% normal serum for 10 min. Sections were incubated for 45 min at room temperature with primary antibody p53 (1 : 500, anti-p53 monoclonal (DO-7); M7001 DAKO).BNIP3 stainingAntigen retrieval was performed by microwave treatment (750 W for 15 min in Antigen Retrieval (DAKO)), followed by cooling in buffer for 30 min. Slides were blocked in 25% normal serum for 10 min. Sections were incubated for 180 min at room temperature with primary antibody BNIP3 (1 : 400, anti-BNIP3 monoclonal (Ana40); ab10433, Abcam, Cambridge, UK).M30 stainingAntigen retrieval was performed by microwave treatment (750 W for 10 min in Antigen Retrieval (DAKO)), followed by cooling in buffer for 30 min. Slides were blocked in 25% normal serum for 10 min. Sections were incubated for 45 min at room temperature with primary Cytodeath antibody M30 (1 : 50, mouse monoclonal (CloneM30); Roche Applied Science, Mannheim, Germany).Each staining protocol was started with pre-incubating in 0.6% hydrogen peroxide for 20 min to block endogenous peroxidase activity. Furthermore, as a negative control, TBS buffer was used instead of primary antibody. Visualisation was performed using Dako Envision, Peroxidase, mouse System (K4001; DAKO). Powerenvision poly-HRP (50510–60307; Immunologic, Duiven, The Netherlands) was used for M30 visualisation. The slides were counterstained with haematoxylin.Evaluation of immunohistochemistryEvaluation for HIF-2α and CA9 staining was performed as described previously in detail (Cleven et al, 2007). Briefly, localisation (epithelial or stromal) was scored separately. For the category stromal staining, only the stromal myofibroblasts were taken into account, not the tumour-infiltrating inflammatory cells or the lamina propria of the normal mucosa. If nuclear staining was present in >5% of the tumour epithelial cells or stromal cells, the sample was considered positive for HIF-2α.If membranous staining occurred in >5% of the tumour epithelial cells or stromal cells, samples were considered positive for CA9 (Yoshimura et al, 2004).TP53 and BNIP3 staining were considered positive by the presence of nuclear staining for TP53 and cytoplasmic staining for BNIP3, in >5% of tumour cells.M30 expression was documented as the number of positive M30 cells per square millimetre of tumour cells (counted in 10 high-power fields ( × 100) per tumour) (Figures 1C and D). Apoptosis was categorised as ‘low’ apoptosis when the number of M30-positive cells ⩽10 (mean) and as ‘high’ apoptosis when the number of M30-positive cells >10 (Marijnen et al, 2003; de Bruin et al, 2006).Data analysisCorrelations between HIF-2α, CA9, BNIP3, TP53, M30 and clinicopathological parameters were determined by the Pearson χ2- and Fisher's exact tests, where appropriate. To evaluate the relationship between HIF-2α, BNIP3, TP53 and survival, Kaplan–Meier survival curves were calculated. Differences between groups were determined by using the Log-rank test. The end point for analyses was overall survival starting from the day of surgery. All P-values are two sided and P<0.05 was considered statistically significant. Correction for multiple comparisons was performed using the Bonferroni procedure. Patients with unknown and unspecified scores have been omitted from analyses for that specific variable. SPSS 12.0 software was used for data analyses.ResultsTP53 mutationsTP53 mutation analysis was successful in 155 out of 195 (79%) CRCs. Out of the 155 CRCs, 72 (46%) were classified as having no TP53 mutations, 4 (3%) as having silent mutations and 2 (1%) as having known common polymorphism (exon 6, codon 213, CGA>CGG, R/R, refSNP rs1800372) (Table 2). A total of 37% (57 out of 155) CRCs were classified as transactivation-incompetent missense mutations and 5% (8 out of 155) CRCs as transactivation-competent missense mutations. Correlations between TP53 and other variables did not change with respect to the predicted presence or absence of transcriptional activity of TP53 mutants. Therefore, in further analyses, CRCs were classified as TP53 wild type when no mutations, silent mutations or a known common polymorphism were found, and as mutant TP53 when CRCs had either missense, nonsense or frame-shift mutations. Using this classification, 77 out of 155 (50%) CRCs showed TP53 mutations (Table 1), which is in agreement with data published by others (Cripps et al, 1994; Pricolo et al, 1997; Borresen-Dale et al, 1998; Kressner et al, 1999; Tortola et al, 1999; Garrity et al, 2004; Tang et al, 2004; Conlin et al, 2005; Mollevi et al, 2007; Petitjean et al, 2007).p53 protein expression was only observed in the nucleus of epithelial cells (Figure 1A). There was a significant correlation between the absence of p53 protein expression and wild-type TP53 vs the presence of p53 protein in mutant TP53 (P=0.029; data not shown). No correlation was observed between the TP53 mutation status and clinicopathological data (Table 1).TP53 mutations, patient survival and apoptosisOverall, no significant survival difference was observed between wild-type and mutant TP53 CRCs (data not shown). However, the previously reported association between HIF-2α- or CA9-positive CRCs and poor prognosis was found to exist exclusively in wild-type TP53 CRCs (P=0.001 and P=0.0829, respectively; Figures 2A and B). Furthermore, there was a significant difference in survival between stromal CA9 expression (38%, 5-year survival) and epithelial CA9 expression (71%, 5-year survival) within wild-type TP53 tumours (P=0.0391, data not shown). Overall levels of HIF-2α or CA9 expression were not different between wild-type and mutant TP53 CRCs (data not shown). Survival of mutant TP53 CRCs was not related to HIF-2α or CA9 expression (P=0.9312 and P=0.8456, respectively; Figures 2E and F). These data suggest that wild-type TP53 CRCs are less susceptible to the adverse effects of hypoxia. As TP53 can induce apoptosis during hypoxia, we assessed the extent of apoptosis (M30 staining; Figures 1C and D). Overall, we found no differences in apoptotic levels between wild-type and mutant TP53 CRCs (Table 3), or between HIF-2α- and CA9-positive or -negative CRCs (data not shown). This was found regardless of the TP53 mutation status. These results indicate that the presence or absence of functional p53 protein is not decisive for determining the extent of apoptosis in CRCs.BNIP3 methylationBNIP3 promoter methylation analysis was successful in all of the 195 CRCs (100%). Overall, 53% (103 out of 195) CRCs showed BNIP3 promoter hypermethylation (Table 1), which is in agreement with data published earlier (Murai et al, 2005; Lee and Paik, 2006).The relationship between protein expression of BNIP3 and BNIP3 promoter methylation status was analysed in a randomly selected subset of patients (n=31). BNIP3 protein expression was only observed in the cytoplasm of epithelial cells (Figure 1B). BNIP3 promoter-methylated CRCs less frequently demonstrated BNIP3 protein expression than unmethylated CRCs (25 vs 75%, respectively).BNIP3 methylation, patient survival and apoptosisOverall, there was no significant survival difference between BNIP3 methylated and unmethylated CRCs (data not shown). Although in HIF-2α stromal-negative CRCs, BNIP3 methylation occurred in 61% (33 out of 54) and did not influence prognosis, in the HIF-2α-positive tumours, methylation was observed at almost equal frequency, 52% (68 out of 132), but was associated with poorer patient survival (P=0.006; Figure 2G). Similarly, exclusively stromal (and not epithelial) expression of CA9 was an indicator of a poorer prognosis in both BNIP3 methylated and unmethylated tumours (P=0.0495 and P=0.0725, respectively; Figures 2D and H).This suggests that hypoxic CRCs with stromal expression of HIF-2α and CA9 have a poorer prognosis, independent of BNIP3 methylation.As BNIP3 has been reported to induce apoptosis in response to hypoxia, its methylation and associated downregulation might be expected to result in less apoptosis in the HIF-2α subgroup. However, a low apoptotic activity (low M30 expression) was detected more frequently in BNIP3 unmethylated CRCs 68% (50 out of 73) compared with BNIP3 methylated CRCs 46% (42 out of 91, P=0.004; Table 3). Although tumours with both methylated BNIP3 and stromal HIF-2α expression showed a poorer patient survival when compared with HIF-2α-negative tumours, this was not related to apoptosis (data not shown). Furthermore, we did not detect differences in apoptotic levels between tumours with or without CA9 expression, regardless of the BNIP3 methylation status.DiscussionIn a previous study on the expression of hypoxia-related markers (HIF1α, HIF2α, CA9 and GLUT1) in colorectal adenocarcinomas, we found that in all tumours at least one of these proteins is immunohistochemically expressed. This indicates that hypoxia is an all but ubiquitous phenomenon in colorectal tumours. However, expression of only HIF2α and CA9 in the tumour-associated stroma was correlated with a poorer prognosis, suggesting that tumours with this particular phenotype follow a more aggressive course (Cleven et al, 2007). These results are in contrast to some other studies on colon and rectum tumours that can be summarised by the observation that HIF1α and GLUT1 expression in rectal cancer cells is of prognostic significance (Haber et al, 1998; Yoshimura et al, 2004; Lu et al, 2006; Theodoropoulos et al, 2006). Our finding was relatively new, in the sense that others have mainly reported the relation between tumour prognosis and expression of hypoxia markers in cancer cells, without paying attention to stromal expression. The results in our study did not differ between tumours of the colon and rectum.So far, the biological basis of stromal expression of hypoxia-related markers in CRC is unclear. It could either be a serendipitous finding, or a genuine indication of altered epithelial–mesenchymal interactions within a subset of tumours. In these tumours, hypoxia-driven metabolic and cell biological changes could hypothetically alter the tumour stroma toward an environment that can facilitate cancer progression, along multiple routes, leading to an enhanced proliferation or survival of tumour epithelial cells (Beppu et al, 2008). Under hypoxic conditions, the tumour stroma can select for the propagation of certain subclones of cancer cells that are optimally endowed for tumour progression. A recent study on breast cancer, on how changes in stromal gene expression affect epithelial tumour progression, showed that gene expression profiling of microdissected tumour stroma resulted in a set of stromal genes, which could predict clinical outcome. This set notably included enhanced stromal expression of hypoxia-associated genes (Finak et al, 2008).The current study attempts to pinpoint specific genetic and epigenetic features of CRC cells, which are both associated with the observed more aggressive presentation of CRCs expressing HIF2α and CA9 in their surrounding stroma, and impact on one of the hallmarks of cancer, namely regulation of apoptosis under hypoxic conditions.Tumour hypoxia results in the induction of pro-death signals, mediated partly by TP53 and BNIP3 (Brahimi-Horn and Pouyssegur, 2006). Therefore, it could be envisaged that hypoxia provides a selective environment for outgrowth of cells in which these genes have become mutated or silenced. Loss of pro-death genes, such as TP53 and BNIP3, may result in increased hypoxia tolerance and cross-resistance to other death-inducing stimuli associated with metabolic stress or treatment. Immunohistochemical staining showed that TP53 and BNIP3 expressions are confined to the epithelial cell compartment in CRC, and are not present in the surrounding mesenchymal cells, making these two genes interesting candidates to study the hypothesis of cancer cell selection under hypoxia-driven modification of tumour–stroma interactions.Intriguingly, our results indicate that tumours expressing HIF-2α or CA9 in their stroma have a poorer prognosis in wild-type TP53 tumours compared with mutant tumours. It is unclear as to how wild-type TP53 might benefit this tumour subgroup, but several possibilities exist. Firstly, p53 is involved in a metabolic switch to glycolysis when oxidative phosphorylation is impaired during hypoxia (Lum et al, 2007). Also, other means of adaptation to metabolic stress, such as increased fatty acid β oxidation, have been shown to be present in tumour cells with intact p53. With respect to this, increased apoptosis was reported in p53-deficient HCT-116 CRC cells as compared with wild-type p53 HCT-116 cells, when challenged by metabolic stress (Buzzai et al, 2007). The second possibility is that wild-type TP53 does not act directly, but simply correlates with defects in another pathway, such as the BNIP3 cell death pathway, which substitutes for TP53 loss in a similar fashion during carcinogenesis.With respect to apoptosis- and hypoxia-driven tumour progression, we did not find important effects related to the mutational status of TP53. However, apoptotic levels were lower in BNIP3-expressing tumours, when compared with tumours with epigenetically silenced BNIP3 (P=0.004), which is somewhat surprising given the fact that functional BNIP3 is thought to induce cell death downstream of hypoxia inducible transcription factors. Apparently, things are more complicated. BNIP3 may not be restricted to regulation of apoptosis, but could also regulate other pathways, such as autophagy, in which there is a delicate balance between cell survival and cell death (Papandreou et al, 2005). Conceivably, the lower apoptotic activity in tumours with functional BNIP3 might be due to autophagic rescue of the tumour cells. Furthermore, BNIP3 levels appear to modulate cell death not only via apoptosis or autophagy, but also via necrosis. Also, the net effect of BNIP3 is determined by the level of expression: too high BNIP3 expression will lead to autophagic cell death, whereas lower levels of BNIP3 expression, as in cells where BNIP3 is silenced, will induce necrosis (Tracy et al, 2007). In our study, we used the immunohistochemical marker M30, which exclusively measures apoptotic cell death, and thus were not able to differentiate between other forms of cell death, such as autophagic death and necrosis (Leers et al, 1999; Ueno et al, 2005).Summarising the results from the current study, levels of apoptosis do not play an important role in determining the poorer prognosis of hypoxic CRCs, as defined by stromal expression of HIF-2α and CA9. However, the latter phenotype is correlated with the presence of wild-type TP53 in the tumour cells, and the presence of functional p53 does indeed appear to have an important impact on poorer prognosis. This prognostic effect is not established through regulation of programmed tumour cell death, but may rather be connected to an enhanced capacity for adequate adaptation to metabolic stress. As TP53 mutations occur in a relatively early stage of colorectal carcinogenesis, the potential deleterious effects of hypoxia on CRC biology may already be programmed in a very early phase of tumour development.As opposed to TP53, functionally or epigenetically silenced BNIP3 did not turn out to be of influence in determining tumour prognosis. Furthermore, preservation of BNIP3 function was shown to decrease apoptotic activity, and may thus be involved in enhanced cell survival through autophagic rescue, or could be implicated in alternative cell death pathways, such as autophagic cell death or necrosis, which we were unable to measure in our experimental approach.The findings in this translational study, on the relation between expression patterns of hypoxia-related markers in clinical samples of CRC and the functional status of genetically or epigenetically modified proteins involved in regulation of tumour cell death on the one hand and patient outcome on the other, open up interesting new avenues for more fundamental studies on the mechanisms underlying tumour hypoxia-induced changes in epithelial–mesenchymal interactions.Understanding the mechanisms by which hypoxic tumours can overcome cell death signals and adapt through metabolic changes is critical for our understanding of tumour progression and development of effective therapeutics in CRC patients with adverse prognostic profiles.\n\nREFERENCES:\n1. Abe T, Toyota M, Suzuki H, Murai M, Akino K, Ueno M, Nojima M, Yawata A, Miyakawa H, Suga T, Ito H, Endo T, Tokino T, Hinoda Y, Imai K (2005) Upregulation of BNIP3 by 5-aza-2′-deoxycytidine sensitizes pancreatic cancer cells to hypoxia-mediated cell death. J Gastroenterol\n40: 504–51015942716\n2. Bacon AL, Fox S, Turley H, Harris AL (2007) Selective silencing of the hypoxia-inducible factor 1 target gene BNIP3 by histone deacetylation and methylation in colorectal cancer. Oncogene\n26: 132–14116799636\n3. Bacon AL, Harris AL (2004) Hypoxia-inducible factors and hypoxic cell death in tumour physiology. Ann Med\n36: 530–53915513303\n4. Beppu H, Mwizerwa ON, Beppu Y, Dattwyler MP, Lauwers GY, Bloch KD, Goldstein AM (2008) Stromal inactivation of BMPRII leads to colorectal epithelial overgrowth and polyp formation. Oncogene\n27: 1063–107017700526\n5. Borresen-Dale AL, Lothe RA, Meling GI, Hainaut P, Rognum TO, Skovlund E (1998) TP53 and long-term prognosis in colorectal cancer: mutations in the L3 zinc-binding domain predict poor survival. Clin Cancer Res\n4: 203–2109516972\n6. Brahimi-Horn C, Pouyssegur J (2006) The role of the hypoxia-inducible factor in tumor metabolism growth and invasion. Bull Cancer\n93: E73–E8016935775\n7. Buzzai M, Jones RG, Amaravadi RK, Lum JJ, DeBerardinis RJ, Zhao F, Viollet B, Thompson CB (2007) Systemic treatment with the antidiabetic drug metformin selectively impairs p53-deficient tumor cell growth. Cancer Res\n67: 6745–675217638885\n8. Cleven AH, van Engeland M, Wouters BG, de Bruine AP (2007) Stromal expression of hypoxia regulated proteins is an adverse prognostic factor in colorectal carcinomas. Cell Oncol\n29: 229–24017452775\n9. Conlin A, Smith G, Carey FA, Wolf CR, Steele RJ (2005) The prognostic significance of K-ras, p53, and APC mutations in colorectal carcinoma. Gut\n54: 1283–128615843421\n10. Cripps KJ, Purdie CA, Carder PJ, White S, Komine K, Bird CC, Wyllie AH (1994) A study of stabilisation of p53 protein versus point mutation in colorectal carcinoma. Oncogene\n9: 2739–27438058340\n11. de Bruin EC, van de Velde CJ, van de Pas S, Nagtegaal ID, van Krieken JH, Gosens MJ, Peltenburg LT, Medema JP, Marijnen CA (2006) Prognostic value of apoptosis in rectal cancer patients of the Dutch total mesorectal excision trial: radiotherapy is redundant in intrinsically high-apoptotic tumors. Clin Cancer Res\n12: 6432–643617085656\n12. Derks S, Lentjes MH, Hellebrekers DM, de Bruine AP, Herman JG, van Engeland M (2004) Methylation-specific PCR unraveled. Cell Oncol\n26: 291–29915623939\n13. Erler JT, Bennewith KL, Nicolau M, Dornhofer N, Kong C, Le QT, Chi JT, Jeffrey SS, Giaccia AJ (2006) Lysyl oxidase is essential for hypoxia-induced metastasis. Nature\n440: 1222–122616642001\n14. Fei P, Wang W, Kim SH, Wang S, Burns TF, Sax JK, Buzzai M, Dicker DT, McKenna WG, Bernhard EJ, El-Deiry WS (2004) Bnip3L is induced by p53 under hypoxia, and its knockdown promotes tumor growth. Cancer Cell\n6: 597–60915607964\n15. Finak G, Bertos N, Pepin F, Sadekova S, Souleimanova M, Zhao H, Chen H, Omeroglu G, Meterissian S, Omeroglu A, Hallett M, Park M (2008) Stromal gene expression predicts clinical outcome in breast cancer. Nat Med\n14: 518–52718438415\n16. Garrity MM, Burgart LJ, Mahoney MR, Windschitl HE, Salim M, Wiesenfeld M, Krook JE, Michalak JC, Goldberg RM, O’Connell MJ, Furth AF, Sargent DJ, Murphy LM, Hill E, Riehle DL, Meyers CH, Witzig TE (2004) Prognostic value of proliferation, apoptosis, defective DNA mismatch repair, and p53 overexpression in patients with resected Dukes’ B2 or C colon cancer: a North Central Cancer Treatment Group Study. J Clin Oncol\n22: 1572–158215117979\n17. Graeber TG, Osmanian C, Jacks T, Housman DE, Koch CJ, Lowe SW, Giaccia AJ (1996) Hypoxia-mediated selection of cells with diminished apoptotic potential in solid tumours. Nature\n379: 88–918538748\n18. Haber RS, Rathan A, Weiser KR, Pritsker A, Itzkowitz SH, Bodian C, Slater G, Weiss A, Burstein DE (1998) GLUT1 glucose transporter expression in colorectal carcinoma: a marker for poor prognosis. Cancer\n83: 34–409655290\n19. Herman JG, Graff JR, Myohanen S, Nelkin BD, Baylin SB (1996) Methylation-specific PCR: a novel PCR assay for methylation status of CpG islands. Proc Natl Acad Sci USA\n93: 9821–98268790415\n20. Kato S, Han SY, Liu W, Otsuka K, Shibata H, Kanamaru R, Ishioka C (2003) Understanding the function–structure and function–mutation relationships of p53 tumor suppressor protein by high-resolution missense mutation analysis. Proc Natl Acad Sci USA\n100: 8424–842912826609\n21. Keith B, Simon MC (2007) Hypoxia-inducible factors, stem cells, and cancer. Cell\n129: 465–47217482542\n22. Kressner U, Inganas M, Byding S, Blikstad I, Pahlman L, Glimelius B, Lindmark G (1999) Prognostic value of p53 genetic changes in colorectal cancer. J Clin Oncol\n17: 593–59910080604\n23. Lee H, Paik SG (2006) Regulation of BNIP3 in normal and cancer cells. Mol Cells\n21: 1–616511341\n24. Leers MP, Kolgen W, Bjorklund V, Bergman T, Tribbick G, Persson B, Bjorklund P, Ramaekers FC, Bjorklund B, Nap M, Jornvall H, Schutte B (1999) Immunocytochemical detection and mapping of a cytokeratin 18 neo-epitope exposed during early apoptosis. J Pathol\n187: 567–57210398123\n25. Lu XG, Xing CG, Feng YZ, Chen J, Deng C (2006) Clinical significance of immunohistochemical expression of hypoxia-inducible factor-1α as a prognostic marker in rectal adenocarcinoma. Clin Colorectal Cancer\n5: 350–35316512994\n26. Lum JJ, Bui T, Gruber M, Gordan JD, DeBerardinis RJ, Covello KL, Simon MC, Thompson CB (2007) The transcription factor HIF-1α plays a critical role in the growth factor-dependent regulation of both aerobic and anaerobic glycolysis. Genes Dev\n21: 1037–104917437992\n27. Marijnen CA, Nagtegaal ID, Mulder-Stapel AA, Schrier PI, van de Velde CJ, van Krieken JH, Peltenburg LT (2003) High intrinsic apoptosis, but not radiation-induced apoptosis, predicts better survival in rectal carcinoma patients. Int J Radiat Oncol Biol Phys\n57: 434–44312957255\n28. Mellor HR, Harris AL (2007) The role of the hypoxia-inducible BH3-only proteins BNIP3 and BNIP3L in cancer. Cancer Metastasis Rev\n26: 553–56617805942\n29. Mollevi DG, Serrano T, Ginesta MM, Valls J, Torras J, Navarro M, Ramos E, Germa JR, Jaurrieta E, Moreno V, Figueras J, Capella G, Villanueva A (2007) Mutations in TP53 are a prognostic factor in colorectal hepatic metastases undergoing surgical resection. Carcinogenesis\n28: 1241–124617259658\n30. Murai M, Toyota M, Suzuki H, Satoh A, Sasaki Y, Akino K, Ueno M, Takahashi F, Kusano M, Mita H, Yanagihara K, Endo T, Hinoda Y, Tokino T, Imai K (2005) Aberrant methylation and silencing of the BNIP3 gene in colorectal and gastric cancer. Clin Cancer Res\n11: 1021–102715709167\n31. Okami J, Simeone DM, Logsdon CD (2004) Silencing of the hypoxia-inducible cell death protein BNIP3 in pancreatic cancer. Cancer Res\n64: 5338–534615289340\n32. Pan Y, Oprysko PR, Asham AM, Koch CJ, Simon MC (2004) p53 cannot be induced by hypoxia alone but responds to the hypoxic microenvironment. Oncogene\n23: 4975–498315107830\n33. Papandreou I, Krishna C, Kaper F, Cai D, Giaccia AJ, Denko NC (2005) Anoxia is necessary for tumor cell toxicity caused by a low-oxygen environment. Cancer Res\n65: 3171–317815833847\n34. Petitjean A, Achatz MI, Borresen-Dale AL, Hainaut P, Olivier M (2007) TP53 mutations in human cancers: functional selection and impact on cancer prognosis and outcomes. Oncogene\n26: 2157–216517401424\n35. Pricolo VE, Finkelstein SD, Hansen K, Cole BF, Bland KI (1997) Mutated p53 gene is an independent adverse predictor of survival in colon carcinoma. Arch Surg\n132: 371–374; discussion 374–59108757\n36. Ray R, Chen G, Vande Velde C, Cizeau J, Park JH, Reed JC, Gietz RD, Greenberg AH (2000) BNIP3 heterodimerizes with Bcl-2/Bcl-X(L) and induces cell death independent of a Bcl-2 homology 3 (BH3) domain at both mitochondrial and nonmitochondrial sites. J Biol Chem\n275: 1439–144810625696\n37. Schmaltz C, Hardenbergh PH, Wells A, Fisher DE (1998) Regulation of proliferation-survival decisions during tumor cell hypoxia. Mol Cell Biol\n18: 2845–28549566903\n38. Soussi T, Lozano G (2005) p53 mutation heterogeneity in cancer. Biochem Biophys Res Commun\n331: 834–84215865939\n39. Swietach P, Vaughan-Jones RD, Harris AL (2007) Regulation of tumor pH and the role of carbonic anhydrase 9. Cancer Metastasis Rev\n26: 299–31017415526\n40. Tang R, Wang JY, Fan CW, Tsao KC, Chen HH, Wu CM, Chen JS, Changchien CR, Hsieh LL (2004) p53 is an independent pre-treatment markers for long-term survival in stage II and III colorectal cancers: an analysis of interaction between genetic markers and fluorouracil-based adjuvant therapy. Cancer Lett\n210: 101–10915172127\n41. Theodoropoulos GE, Lazaris AC, Theodoropoulos VE, Papatheodosiou K, Gazouli M, Bramis J, Patsouris E, Panoussopoulos D (2006) Hypoxia, angiogenesis and apoptosis markers in locally advanced rectal cancer. Int J Colorectal Dis\n21: 248–25716052307\n42. Toledo F, Wahl GM (2006) Regulating the p53 pathway: in vitro hypotheses, in vivo veritas. Nat Rev Cancer\n6: 909–92317128209\n43. Tortola S, Marcuello E, Gonzalez I, Reyes G, Arribas R, Aiza G, Sancho FJ, Peinado MA, Capella G (1999) p53 and K-ras gene mutations correlate with tumor aggressiveness but are not of routine prognostic value in colorectal cancer. J Clin Oncol\n17: 1375–138110334521\n44. Tracy K, Dibling BC, Spike BT, Knabb JR, Schumacker P, Macleod KF (2007) BNIP 3 is an RB/E2F target gene required for hypoxia-induced autophagy. Mol Cell Biol\n27(17): 6229–624217576813\n45. Ueno T, Toi M, Linder S (2005) Detection of epithelial cell death in the body by cytokeratin 18 measurement. Biomed Pharmacother\n59(Suppl 2): S359–S36216507409\n46. Vande Velde C, Cizeau J, Dubik D, Alimonti J, Brown T, Israels S, Hakem R, Greenberg AH (2000) BNIP3 and genetic control of necrosis-like cell death through the mitochondrial permeability transition pore. Mol Cell Biol\n20: 5454–546810891486\n47. Wiggers T, Jeekel J, Arends JW, Brinkhorst AP, Kluck HM, Luyk CI, Munting JD, Povel JA, Rutten AP, Volovics A et al (1988) No-touch isolation technique in colon cancer: a controlled prospective trial. Br J Surg\n75: 409–4153292002\n48. Yan J, Yun H, Yang Y, Jing B, Feng C, Song-bin F (2006) Upregulation of BNIP3 promotes apoptosis of lung cancer cells that were induced by p53. Biochem Biophys Res Commun\n346: 501–50716765911\n49. Yoshimura H, Dhar DK, Kohno H, Kubota H, Fujii T, Ueda S, Kinugasa S, Tachibana M, Nagasue N (2004) Prognostic impact of hypoxia-inducible factors 1α and 2α in colorectal cancer patients: correlation with tumor angiogenesis and cyclooxygenase-2 expression. Clin Cancer Res\n10: 8554–856015623639"
+}

batch_3/PMC2528311.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2528311",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2528311\nAUTHORS: R. Samant, I. Alomary, P. Genest, L. Eapen\n\nABSTRACT:\nRecently published studies clearly indicate that there are now several acceptable options for managing stage i testicular seminoma patients after orchiectomy. We therefore decided to survey Canadian radiation oncologists to determine how they currently manage such patients and to compare the results with previous surveys.Our results demonstrate that adjuvant single-agent chemotherapy is being considered as an option by an increasing proportion of radiation oncologists (although it is not considered the preferred option), the routine use of radiotherapy is declining, and surveillance is becoming increasingly popular and is recommended most often.\n\nBODY:\n1. INTRODUCTIONTesticular cancers are the most common—and the most curable—malignancies among young men in North America 1,2. Seminomas account for approximately half of these cancers, and most patients (80%) present with stage i disease 3,4. Treatment is highly successful, with 5-year overall and disease-specific survivals approaching 100% for stage i seminoma 2,5.For decades, the standard treatment for this cancer has been radical inguinal orchiectomy, followed by adjuvant radiotherapy to the para-aortic and ipsilateral pelvic regions 2,5,6. This approach led to a recurrence rate of less than 5%, with salvage chemotherapy being highly effective in the few patients that did relapse 2,3,5,6. However, long-term follow-up data (beyond 10–15 years) now indicate that treatment-related morbidity and mortality (especially from a second malignancy) are significant concerns 3,7,8. Various approaches have therefore been investigated to minimize the toxicity associated with routine use of radiotherapy. One approach has been to minimize toxicity by reducing radiotherapy field sizes and doses 9,10; another approach has been to avoid radiation altogether 3. This change has, in turn, resulted in the evolution of surveillance after orchiectomy as a viable option because of the general availability of computed tomography (ct) imaging for follow-up purposes 11. In addition, the use of single-agent chemotherapy (most commonly 1–2 cycles of carboplatin) has also been recognized as a potential option in place of radiotherapy 12–14.As a result, management of stage i seminoma is currently focused not only on maintaining high rates of cure, but also minimizing both short- and long-term toxicity 3,11,15. Previously, a survey of radiation oncologists conducted by Choo et al. in 2001 regarding management of stage i seminoma revealed considerable variation in practice, especially with regard to treatment 16. Most radiation oncologists at that time routinely discussed surveillance as an option, but thought that only 20% of patients would choose that option. At that time, the authors found that adjuvant radiotherapy was usually the preferred choice. However, data have continued to accumulate regarding the viability of options that do not include the routine use of adjuvant radiotherapy.Review of seminoma management at our own institution 17 showed a steady decline in the use of adjuvant radiotherapy since the 1990s and the increasing use of surveillance for stage i seminoma patients—a trend that has also been reported elsewhere 18,19. Although recent studies suggest that the use of single-agent chemotherapy with carboplatin is very encouraging and the use of adjuvant chemotherapy is becoming increasingly popular in Europe 20, this approach is still not routinely considered at our institution. Our experience has been that chemotherapy tends to be reserved for the salvage of radiotherapy failures or in patients wanting adjuvant treatment who are not deemed suitable for radiation. Based on the increasing options for management of stage i seminoma, we decided to survey Canadian radiation oncologists to see if their management approaches had changed since earlier in the decade.2. METHODSWe developed an electronic survey to assess management of stage i testicular seminoma patients. The survey was specifically designed for radiation oncologists, and the categories evaluated included staging investigations, treatment options, radiotherapy treatment planning details, surveillance protocols, and respondent demographic information. Intended for self-completion, the survey takes approximately 15 minutes to finish. After obtaining approval from the research ethics board of the Ottawa Hospital to proceed with this survey study, we sent the survey by e-mail in 2005 to Canadian radiation oncologists identified as treating genitourinary malignancies. The list of radiation oncologists was formulated from information obtained from the directory of the Canadian Association of Radiation Oncologists at that organization’s Web site and from correspondence with radiation oncology department heads at cancer centres across Canada. Initial non-responders were sent reminder notices, also by e-mail. Remuneration was not offered for completing the survey.The completed surveys were collated and analyzed for this study. The chi-square statistic was used to assess associations between the respondent’s practice history and that person’s responses to survey questions. To compare differences in choice of treatment attributable to age and years of practice, a test of mean differences was applied using analysis of variance. The SPSS software package (SPSS, Chicago, IL, U.S.A.) was used to perform the analyses.3. RESULTSElectronic surveys were sent to 119 Canadian radiation oncologists, and 93 responses were received (78% response rate). Of the respondents, 14 indicated that they did not treat seminoma patients, and 1 declined to complete the survey. The survey completion rate among eligible responders was therefore 74% (78/105). Among respondents completing the survey, 89% were men and 11% were women. Median age was 43 years. Mean length of practice was 13 years, and 80% of the respondents worked in academic centres.Figure 1 shows that the staging investigations most commonly used are ct scans of the abdomen and pelvis (100%) and chest radiographs (84%). Serum tumour markers are also commonly assessed, with alpha-fetoprotein (afp) being measured 95% of the time, and beta–human chorionic gonadotropin (bhcg) levels 100% of the time.Adjuvant radiation and surveillance were considered the most common standard treatment options by 90% and 81% of respondents respectively. However, 30% of respondents also listed adjuvant chemotherapy as a standard treatment option (Figure 2). However, when asked to indicate the management that they felt was most appropriate for most patients, surveillance was chosen by 56%; radiotherapy, by 31%; and chemotherapy, by 1%. The remaining 12% were unsure of the most appropriate management (Figure 3). The most common concerns related to the use of adjuvant radiotherapy were second cancers (84%), infertility (77%), late gastrointestinal toxicity (67%), acute nausea and vomiting (61%), and late renal toxicity (60%).Most respondents (91%) said that they routinely discuss surveillance with patients, but that tumour-related risk factors (size, local extension, and lymphovascular invasion), together with patient age and compliance, influence their recommendations. Nearly all respondents (<95%) started offering surveillance during the last 10 years. Most respondents felt that at least 50% of patients are now choosing surveillance. The most commonly listed reasons, in order of importance, for not offering surveillance were patient fears and anxieties, patient reluctance, increased costs, and the belief that survival was actually better with the use of adjuvant radiotherapy.For patients on surveillance protocols, the investigations commonly used are ct scans of the abdomen and pelvis (93%), chest radiographs (81%), bhcg levels (92%), and afp levels (84%). Surveillance investigations are usually carried out every 3–4 months by 84% of respondents, every 6 months by 15%, and every 12 months by 1%.When treating with radiotherapy, 50% use para-aortic fields only, and 50% use para-aortic and ipsilateral pelvic fields. Planning ct is used for simulation by all respondents (100%); intravenous pyelograms and lymph angiograms are used by only 1%. A dose of 2500 cGy in 15–20 fractions over 3–4 weeks was recommended by more than 90% of respondents, and all respondents use linear accelerators (≥ 6 MV photons) for treatment delivery. Scrotal shielding is routinely used by 49% of respondents to reduce dose to the contralateral testicle, and thermoluminescent dosimeters are used by 47% for verifying dose to the contralateral testicle. Almost all respondents (96%) discussed fertility issues and sperm banking with patients before the start of adjuvant radiotherapy.Prophylactic antiemetics are ordered by 58% of the respondents, with ondansetron, prochlorperazine, dimenhydrinate, and metoclopramide being prescribed by 48%, 25%, 18%, and 8% respectively. Following treatment, 67% recommend that patients take contraceptive measures for at least 3 months (14%), 6 months (38%), 12 months (40%), or 24 months (8%).We observed a trend, although not statistically significant, for older radiation oncologists (> 45 years vs. ≤ 45 years) to choose radiation for their patients (p = 0.07). However, years in practice, type of practice (academic vs. community), and provincial location did not appear to influence management choices.4. DISCUSSIONCancer treatment approaches evolve with time and accumulation of research findings—first to improve and maximize cure rates, and then to minimize toxicity. Stage i seminoma management certainly confirms this paradigm. The use, since the 1960s, of adjuvant radiotherapy after orchiectomy reduced relapse rates to less than 5% and established adjuvant radiation as the standard practice until the 1990s 2,6. In fact, overall 5-year survival rates were in the 98%–99% range, because chemotherapy was highly successful in salvaging the few patients that did relapse post  radiotherapy 20. However, with the accumulation of long-term follow-up data over 25 years or more, it became obvious that second malignancies are a significant problem following radiation 3,7. The prevailing focus therefore moved to reduction of toxicity, because salvage therapies were so effective 3,6.This change in focus led to numerous studies evaluating alternatives to standard radiotherapy, and reductions in radiotherapy treatment volumes and lower radiation doses were both shown to be possible without significantly reducing efficacy 9,10. In prospective and retrospective studies, surveillance approaches have also been shown to be an excellent alternative that do not compromise cure rates, although they require more frequent imaging investigations 11,17–19. More recently, the use of adjuvant single-agent chemotherapy (most commonly with single-agent carboplatin) has been shown to yield results, in terms of relapse rates and overall survival, similar to those seen with the use of adjuvant radiation 14. Therefore, currently, there is evidence that all of the above approaches are effective and reasonable options that have their own unique advantages and limitations. However, long-term follow-up data for the newer approaches are limited, and questions still exist regarding long-term efficacy and toxicity.The effect of recent studies and their influence on actual current practice patterns across North America and Europe has not been fully evaluated. However, it is well known that medical practice patterns often change gradually after research findings are published, as physicians reflect on the available evidence and perhaps discuss them with their colleagues 21–23. Therefore, we believe that it is important to determine whether the accumulating evidence regarding management of stage i testicular seminoma has had an effect at the level of clinical practice in Canada, and that is what we attempted to do in the present study.With a very favourable survey completion rate of 74% 24, we believe that our results generally reflect current opinion among Canadian radiation oncologists regarding stage i seminoma management. Our findings show that staging is fairly consistent and reflects the emergence of ct scans of the abdomen and pelvis as standard, together with chest radiographs 11 and serum tumour markers (afp, bhcg). Radiation treatment planning with ct scans and delivery approaches using linear accelerators, as indicated by the respondents, are also consistent with the published literature, as are the dose and fractionation regimens commonly used 5. However, opinion continues to vary regarding the necessity of ipsilateral pelvic lymph node irradiation.The surveillance protocols among respondents are very similar to published recommendations 15. Although variations in practice still occur, a preponderance of respondents indicate that they routinely discuss both surveillance and adjuvant radiotherapy with their patients, and most believe that surveillance is the preferred option, with at least 50% of patients making the latter choice.Adjuvant chemotherapy is also starting to be recognized as a treatment option by almost one third of radiation oncologists even though our survey was conducted before the results of the Medical Research Council E19 study, published by Oliver et al. 14, indicated that at a median follow-up of 4 years, adjuvant single-agent carboplatin chemotherapy was essentially equivalent in terms of relapse rates and overall survival to adjuvant radiotherapy. Our finding in this regard represents a major shift from the survey results reported in 2002 by Choo et al. 16, when most radiation oncologists thought that adjuvant radiotherapy should be standard, although surveillance was considered an option. Also, chemotherapy was not even considered an option by any respondents at that time, likely reflecting the limited published data to that point.The biggest issue remains the attempt to achieve a balance between minimizing relapse rates and patient fears and anxiety related to recurrences, and avoiding unnecessary treatment for the preponderance of patients that will not relapse and the potential for toxicity that can occur decades later. This balance will likely remain controversial for some time to come, and it is uncertain whether consensus can be achieved in the near future. Individual patient factors, including personal choice, will also be essential in determining the management option that is chosen. However, it is reassuring to confirm that many radiation oncologists have been re-evaluating their approach to management of stage i testicular seminomas over the last 5 years, in parallel with growing evidence of newer management approaches. This re-evaluation may in part be attributable to the growing sub-specialization among many practicing radiation oncologists in terms of site-specific treatment and the resulting familiarity with recently published studies, and also to the Canadian sources of much of the published literature regarding surveillance for management of stage i seminoma 11,17–19. We expect that management approaches will continue to evolve as ongoing studies mature, especially those evaluating the use of single-agent carboplatin chemotherapy.5. CONCLUSIONSCanadian radiation oncologists are routinely discussing surveillance in addition to adjuvant radiotherapy as treatment options for patients with stage i testicular seminoma, but surveillance is usually considered the preferred option. Chemotherapy also appears to be emerging as a viable option among a growing number of radiation oncologists. As a result, the use of radiation is declining.FIGURE 1The frequency of specific staging investigations used for stage i seminoma patients, from a sample of Canadian radiation oncologists. ct = computed tomography; afp = alpha-fetoprotein; β-hcg = beta–human chorionic gonadotropin.FIGURE 2Standard management options used for stage i seminoma patients, from a sample of Canadian radiation oncologists.FIGURE 3The “most appropriate option” for most stage i seminoma patients, from a sample of Canadian radiation oncologists.\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2528552.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2528552",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2528552\nAUTHORS: D. Seely, D. Oneschuk\n\nABSTRACT:\nThe use of complementary and alternative medicine (cam), including the ingestion of natural health products (nhps), is common among cancer patients. Of concern to clinicians and patients alike is the possibility that cam, used concurrently with biomedical therapy, may interact poorly with that therapy, especially chemotherapy and radiotherapy. Proponents of nhps argue that taking such products can help to reduce the side effects of conventional therapy and can provide an additional anticancer effect. However, opponents insist that the potential for harm is too great to warrant the risk of concurrent administration. There are promising examples of specific nhps that may provide patient benefit even when given in close proximity both to chemotherapy and to radiotherapy, but unfortunately, in part because of a rather limited evidence base, caution is warranted when considering the issue of therapeutic interactions. Strategic application of nhps before or after conventional therapy may be considered; however, concurrent application should be avoided as a general principle until further evidence is available regarding specific interactions.\n\nBODY:\nNo Body Content\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2528565.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2528565",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2528565\nAUTHORS: P.R. Fortin\n\nABSTRACT:\nThe Integrating Wellness into Cancer Care conference was held at the University of Toronto, October 4–5, 2007, and was dedicated to the memory of the late Dr. Véronique Benk. This article summarizes the workshops at that conference.The notion of wellness and an integrated approach should be introduced from the outset as part of the cancer patient’s management. Having wellness as part of the treatment sets a standard for taking care of the patient’s emotional, spiritual, physical, and nutritional needs, and for providing information on complementary therapies. A focus on holistic supportive care during treatment and survivorship is important.The whole medical team should support an integrative program. Referral to an education program and one-to-one assessments by a point person such as an advanced nurse practitioner, a social worker, or a psychological counsellor with appropriate special training should be mandatory. The concept of a pathfinder or cancer guide was discussed.\n\nBODY:\nNo Body Content\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2529260.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2529260",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2529260\nAUTHORS: Jeff F Wang, Bo wang, Joshua A Jansen, Eric E Santos, Deba P Sarma\n\nABSTRACT:\nWe are reporting a very rare case of primary bronchogenic squamous cell carcinoma (SCC) with bone metastasis in a 13-year-old boy. A brief review of the English literature on this rare neoplasm in childhood is presented.\n\nBODY:\nCase presentationA 13-year-old boy presented with a two-month history of left shoulder pain. Radiographs followed by MRI demonstrated a destructive lesion of the metaphysis of the proximal third of the left humerus [Fig 1]. The metaphysis of the humerus was replaced by a 5-cm tumor. There was some erosion of the cortex with minimal soft tissue extension by the tumor. The initial impression of the lesion was an osteosarcoma, however, an open biopsy revealed metastatic squamous cell carcinoma. The tumor showed islands and nests of squamous cells with a basaloid appearance at the periphery with maturation and squamous pearl formation in the center [Fig 2]. The cells appeared anaplastic with prominent nucleoli, numerous mitoses and focal necrosis [Fig 3].Figure 1X-ray of left shoulder shows destructive lesion of humerus.Figure 2Invasive islands and nests of squamous cells with basaloid appearance (Hematoxylin and eosin stain, 10×).Figure 3Squamous cells with prominent nucleoli, numerous mitoses and necrosis (Hematoxylin and eosin stain, 20×).Molecular cytogenetic studies on the tumor were positive for rearrangement of the NUT1 region in 54% of the interphase cells. Additional interphase FISH studies were negative for the NUT/BDR4 fusion. The patient underwent a PET-CT that showed multiple foci of increased uptake in the right upper lobe of the lung. Additional abnormal uptake was also noted at T1, T8, T10, and T11, left proximal humerus, right hilar nodes and right acetabulum. A subsequent chest CT confirmed. the malignant lung neoplasm [Fig. 4] Since lung cancer is an extremely rare childhood neoplasm, radon tests are performed in the family home and the results were negative. Upon further questioning, a history of lung cancer in both grandmothers was uncovered. One grandfather also had a history of non-Hodgkin lymphoma and pancreatic cancer. The patient was treated with various chemotherapeutic agents. There was no significant clinical improvement and he died one year later.Figure 4Chest CT confirms the neoplasm in the right lung.DiscussionPrimary lung cancer is the most frequently diagnosed cancer in the USA and the most common cause of cancer mortality worldwide. It occurs most often between the ages of 40 and 70 years, with a peak incidence in the fifties and sixties. Only 2% of all cases appear before the age of 40. Primary lung cancer in childhood is a rare entity and primary bronchogenic squamous cell carcinoma is extremely rare. To our knowledge, only eight cases of primary bronchogenic squamous cell carcinoma in childhood have been reported in English literature.Pulmonary squamous cell carcinoma is most commonly found in men and is closely correlated with a smoking history. Histologically, the tumor is characterized by the presence of keratinization and intercellular bridges. It is graded according to the degree of keratinization, squamous pearl formation, or intercellular bridges. These features are obvious in the well-differentiated tumors but only focally demonstrated in the poorly-differentiated tumors.Cayler et al [1] in 1951 reported 16 cases of primary carcinoma of the lung in children less than 15 years of age Primary bronchogenic squamous cell carcinoma is extremely rare in childhood and adolescence. Eight histologically confirmed cases reported in the English literature [2-9] are summarized along with the present case [Table 1].In 1974, Niitu et al [2] reported one case of squamous cell carcinoma in a boy and reviewed the world literature and found 39 cases of primary lung cancer in children less than 16 years of age. These cases included two cases of bronchogenic squamous cell carcinoma [6,8]. Since then only five additional cases of primary brochogenic SCC have been reported, including one case with substantial family history of cancer [[3-5,7], and [9]]. Most of the patients (eight out of nine) are boys. The clinical presentation of these bronchial cancers varies with the extent of the primary tumor. In our case, the patient presented with bone pain due to metastasis. Four of the reported cases presented with recurrent pneumopathies and hemoptysis. Three of the reported cases were incidentally found by chest x-ray. One case was discovered by routine chest radiograph. There have been no clearly identified risk factors. There is no standard treatment and management essentially depends on the initial findings of the extent of tumor and the presence of metastases. Generally, the prognosis is poor due to metastatic disease.The high frequency of p53 mutations have been seen in all histological types of lung carcinoma. Loss of tumor suppressor gene RB, inactivation of CDK-inhibitor, and overexpression of epidermal growth-factor receptor might contribute to the development of neoplasm. There are no p53 mutations or loss of tumor suppressor gene in our case. However, NUT1, a gene homologous to the major nitrogen regulatory gene [10], is positive for gene rearrangement. It is unknown whether this rearrangement plays any role in the pathogenesis of primay lung bronchogenic squamous cell carcinoma.ConsentWritten consent was obtained from the patient's parents for publication of this case report. A copy of the written consent is available for review by the Editor-in-Chief of this Journal.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsJFW conceived and drafted the manuscript, BW helped with the photomicrographs, EES and JAJ revised and proof-read the manuscript. DPS reviewed the references, made the final corrections and submitted the manuscript. All authors have read and approved the final manuscript.\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2529263.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2529263",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2529263\nAUTHORS: Yi-Wen Chen, Rongye Shi, Nicholas Geraci, Sheela Shrestha, Heather Gordish-Dressman, Lauren M Pachman\n\nABSTRACT:\nBackgroundTo evaluate the impact of the duration of chronic inflammation on gene expression in skeletal muscle biopsies (MBx) from untreated children with juvenile dermatomyositis (JDM) and identify genes and biological processes associated with the disease progression, expression profiling data from 16 girls with active symptoms of JDM greater than or equal to 2 months were compared with 3 girls with active symptoms less than 2 months.ResultsSeventy-nine genes were differentially expressed between the groups with long or short duration of untreated disease. Genes involved in immune responses and vasculature remodelling were expressed at a higher level in muscle biopsies from children with greater or equal to 2 months of symptoms, while genes involved in stress responses and protein turnover were expressed at a lower level. Among the 79 genes, expression of 9 genes showed a significant linear regression relationship with the duration of untreated disease. Five differentially expressed genes – HLA-DQA1, smooth muscle myosin heavy chain, clusterin, plexin D1 and tenomodulin – were verified by quantitative RT-PCR. The chronic inflammation of longer disease duration was also associated with increased DC-LAMP+ and BDCA2+ mature dendritic cells, identified by immunohistochemistry.ConclusionWe conclude that chronic inflammation alters the gene expression patterns in muscle of untreated children with JDM. Symptoms lasting greater or equal to 2 months were associated with dendritic cell maturation and anti-angiogenic vascular remodelling, directly contributing to disease pathophysiology.\n\nBODY:\nBackgroundJuvenile dermatomyositis, a systemic vasculopathy, is the most common of the inflammatory myopathies in children and is characterized by symmetrical proximal muscle weakness, and a pathognomonic rash, which includes a heliotrope discoloration about the eyes, dilated capillaries at the nailbeds and eyelid margins, and thickened erythema over the knuckles (Gottron's papules). The three other diagnostic criteria are: serum elevation of muscle derived enzymes, a muscle biopsy with specific histological features that include mononuclear cell infiltrates as well as perifasicular atrophy with evidence of an occlusive vasculopathy, and a positive electromyogram documenting inflammation [1].There is little information describing the critical variables that influence the development and course of this often devastating illness. Over 3.2 cases/million children/year are diagnosed, with a 2.1 girl to 1 boy ratio [2]. At the time of their first symptom, rash or weakness, the mean age of the patient population is 6.7 years, while 25% of the children are age 4 or younger. It usually takes 4 months or more for the children to be diagnosed with JDM, when the muscle biopsy is obtained [3]. At diagnosis, the extent and severity of the skin and muscle inflammatory response can be assessed using validated disease activity scores (DAS) for skin and muscle involvement [2,4,5].Children with JDM often have a family history of autoimmune disease. The frequencies of the HLA antigens DQA1*0501, DQA1*0301, and DRB3 have been reported to be higher in the JDM population than control populations suggesting a genetic contribution to this disorder, which may be additionally influenced by the TNF-α-308 allelic polymorphisms [6,7].The precise stimulus initiating the inflammatory process is not known. However, there is evidence that newly diagnosed children with JDM have a history of infection, often respiratory or gastrointestinal in nature, within three months prior to the appearance of rash or muscle weakness [5]. In addition, gene expression profile data from MRI directed diagnostic biopsies of muscle from untreated children with active symptoms of JDM identified up-regulation of many type I interferon- (IFNα/β) inducible genes [8]. These findings support the interpretation that the inflammatory milieu in JDM is similar to that seen in anti-microbial responses. In that investigation we also found a marked down-regulation of genes associated with protein synthesis; both observations were subsequently confirmed in studies of muscle from adults with DM [9].Untreated chronic inflammation in children with JDM is associated with the development of pathological calcifications [10]. The phenotype of the children who present in clinic early in their disease course differs from those diagnosed later, both with respect to height and weight, and specific JDM symptoms. Children diagnosed early in the disease course are much weaker than those who have a longer time to diagnosis [10]. However, the extent and severity of skin involvement appears to be stable over time [10].With respect to diagnostic laboratory testing, serum levels of muscle enzymes, generally used to evaluate muscle inflammation, are more likely to be in the normal range when blood is collected 2–4.7 months after the child's first symptom (rash or weakness), making it more difficult to establish a diagnosis of definite JDM [10]. The purpose of the present study was to examine the impact of the duration of untreated chronic inflammation on gene expression in diagnostic muscle biopsies obtained from a large group of untreated girls with clinical symptoms of active JDM in order to identify genes and biological processes associated with disease progression in JDM.ResultsDemographicsAll diagnostic biopsies studied were from untreated girls who had not been given any nonsteroidal or immunosuppressive therapy, and for whom a Disease Activity Score (DAS) was obtained at the time of biopsy (Table 1). The short disease duration group (mean 0.9 ± 0.3 months) was compared with the long duration group (mean 18 ± 27.9 months). Their ages at disease onset date were similar: 5.6 ± 1.0 years for the short duration group compared with 5.2 ± 2.7 years for long duration group. They were similar in disease activity scores. Total DAS for the short duration group was 11.7 ± 4.0 (mean DAS skin 5.0 ± 1.7, mean DAS muscle 6.7 ± 2.5) compared with the Total DAS for the long disease duration group of 12.0 ± 3.9 (mean DAS skin, 6.2 ± 1.4, mean DAS muscle 5.8 ± 3.1). Microarray analysis did not identify any gene significantly differentially expressed among different genotypes (TNF-α-308, DQA1*0301 and DQA1*0501). The second group of girls tested for confirmation of the gene expression changes by immunohistochemistry or RT-PCR, 5 patients each of long and short duration of untreated disease were also similar with respect to age at onset, race, and DAS scores (Table 2).Table 1Demographics of children with JDM in expression profiling study.Patient No.Age at OnsetDAS SkinDAS WeakDAS TotalAge at MBxDuration of Untreated Disease (Months)DQA 0301DQA 0501TNF TypeShort Disease Duration14.74.07.011.04.81.1negposGA26.87.09.016.06.81.1negposGG35.54.04.08.05.500.6negposGAMean5.65.06.711.75.70.9ST. Dev.1.01.72.54.01.10.3Long Disease Duration41.06.05.511.53.530.6negposGA51.86.03.09.02.36.1posnegGG68.39.010.019.08.52.1negnegGG78.85.04.09.09.02.6negposGA83.96.09.015.05.216.1negnegGG91.67.09.016.02.27.7posnegGG107.28.06.014.07.42.2posnegGG118.34.08.012.08.74.9negposAA126.07.03.010.07.720.6posposGG137.76.07.013.08.26.4negposGG142.53.01.04.03.613.6negnegGG152.67.03.010.011.3105.6negnegGA166.17.010.017.06.66.6negnegGA174.56.06.012.05.28.5negposGG185.16.08.014.010.363.5negposAA198.46.01.07.08.63.2negposGGMean5.16.25.812.06.718.8ST. Dev.2.81.43.13.92.927.9DAS: disease activity score; MBx: muscle biopsy; DQA: Major histocompatibility complex, class II, DQ alpha-1; TNF type: Tumor necrosis factor -308 allele.Table 2Demographics of children with JDM in qRT-PCR validation.Patient No.Age at OnsetDAS SkinDAS WeakDAS TotalAge at MBxDuration of Untreated Disease (Months)DQA 0301DQA 0501TNF TypeShort Disease Duration16.85.09.014.06.91.2negposGA26.95.08.013.07.01.7negposGG36.17.09.016.06.31.8negposGG411.65.08.013.011.71.0negposGG58.95.07.012.09.01.5negposGGMean8.15.48.213.68.21.4ST. Dev.2.20.90.81.52.20.4Long Disease Duration68.56.07.013.09.612.9negnegGG75.68.08.016.05.94.0negposGA88.34.08.012.08.74.9negposAA97.76.07.013.08.26.4negposGG1014.75.010.015.015.25.9negposGGMean9.05.88.013.89.56.8ST. Dev.3.41.51.21.63.53.5DAS: disease activity score; MBx: muscle biopsy; DQA: Major histocompatibility complex, class II, DQ alpha-1; TNF type: Tumor necrosis factor -308 allele.Comparison of muscle biopsy gene profiles based on disease durationAs noted above, duration of disease, measured as more than 2 months of symptoms is a critical variable, influencing both clinical and laboratory data [10]. Since changes at the molecular level occur prior to phenotypic changes, we used 2 months (≥ 2 months compared with < 2 months) as the grouping cut-off in the present study. In order to control for gender effects, this study was limited to girls. We identified 79 genes represented by 85 Affymetrix probe sets differentially expressed in the muscle samples. All genes reached statistical significance (p < 0.05) after multiple testing correction with the false positive rate at 5%. In addition, in order to \"bracket\" the optimal cut-off time we compared data from patients with disease duration of 3 months or less with greater than 3 months, and also compared 4 months or less with greater than 4 months. There was no significant difference in expressed genes when the 4 month cut-off was used, and only 3 genes were differentially expressed when the 3 month cut-off was used.For an overview of the changes in relation to the baseline generated using the control samples from age-matched healthy girls undergoing cleft palate repair, we used hierarchical clustering analysis to visualize changes of gene expression patterns associated with the three groups (control, short and long active disease duration) (Figure 1). Six clusters were identified by visualizing the gene tree in Figure 1. Among the 79 genes, 44 genes were expressed at a lower level in the group with active disease greater than 2 months vs. shorter than 2 months (Figure 1, cluster A, B, and E), while 35 genes were expressed at a higher level in the group (Figure 1, cluster C, D, and F).Figure 1Gene tree generated by hierarchical clustering based on gene expression patterns. Genes up-regulated in girls with long (≥ 2 m) vs. short (< 2 m) duration of active disease were clustered into cluster C, D and F, while down-regulated genes were clustered into clusters A, B and E. The color codes represent the ratio between each of the JDM group compared with the age- and sex-matched control samples.Cluster A represents genes up-regulated only in JDM patients with short duration of active disease. Cluster B represents genes significantly down-regulated only in JDM patients with long duration of active JDM. Many genes in these two clusters are involved in protein turn over and stress responses [see Additional file 1]. Most genes in cluster E are either encoding enzymes or proteins with unknown functions. These genes were highly up-regulated in patients with short duration of active disease but not in the other groups suggesting that they are activated at the acute phase of the disease.For genes up-regulated in patients with longer duration of active disease (cluster C, D, and F), genes in cluster C and D were up-regulated in both groups when compared to healthy controls. However, the changes were significantly higher in the patients with long duration of active disease. Interestingly, most of the immune response genes identified in this study were grouped into these two clusters (Table 3). While genes in cluster F were expressed at higher level in the group with long duration, most of the genes in this cluster were down-regulated compared to the baseline (Table 4). In cluster F, most genes are associated with functions of vasculature remodelling. In this report, we will focus on genes in cluster C, D and F (Table 3 and 4), while the complete gene list of each cluster is reported in Additional file 1.Table 3Up-regulation of immune response genes in skeletal muscles of patients with active JDM longer than 2 months (Cluster C and D; genes are in order shown in figure 1)Affymetrix accessionp-valueFold change (long/short)Gene descriptionCluster C203981_s_at6.2E-031.5ATP-binding cassette, sub-family D (ALD), member 4210105_s_at1.5E-031.6FYN oncogene related to SRC, FGR, YES214430_at1.4E-031.6galactosidase, alpha200602_at6.5E-042.0amyloid beta (A4) precursor protein201103_x_at5.7E-031.4hypothetical protein LOC200030209765_at2.3E-031.6a disintegrin and metalloproteinase domain 19216510_x_at9.6E-045.6immunoglobulin heavy constant gamma 1201069_at3.9E-032.0matrix metalloproteinase 2203473_at1.1E-042.2solute carrier organic anion transporter family, member 2B1203742_s_at1.9E-031.5thymine-DNA glycosylase205917_at6.5E-041.6ZNF264212671_s_at2.5E-045.6major histocompatibility complex, class II, DQ alpha 1/2Cluster D218376_s_at4.1E-032.5NEDD9 interacting protein with calponin homology and LIM domains209079_x_at9.1E-041.6protocadherin gamma subfamily38671_at5.9E-041.6plexin D1201279_s_at2.0E-031.6disabled homolog 2, mitogen-responsive phosphoprotein (Drosophila)211066_x_at6.5E-041.6protocadherin gamma subfamily215599_at3.8E-042.1SMA4205717_x_at1.8E-031.4protocadherin gamma subfamily217659_at1.3E-031.6KIAA0261212607_at1.1E-041.3AKT3 (protein kinase B, gamma)211748_x_at4.0E-032.4prostaglandin D2 synthase 21 kDa (brain)215376_at2.4E-041.4CDNA FLJ12295 fis, clone MAMMA1001818202259_s_at1.1E-041.6phosphonoformate immuno-associated protein 5212187_x_at2.1E-031.9prostaglandin D2 synthase 21 kDa (brain)206666_at3.7E-042.5granzyme K (serine protease, granzyme 3; tryptase II)217947_at5.9E-041.9chemokine-like factor super family 6Table 4Genes involved in vasculature remodelling were down-regulated in patients with active JDM shorter than 2 months (Cluster F; genes are in order shown in figure 1).Affymetrix accessionp-valueFold change (long/short)Gene description213290_at4.2E-031.6collagen, type VI, alpha 2201141_at2.1E-063.2glycoprotein (transmembrane) nmb221796_at1.9E-032.0cDNA clone IMAGE:452016207695_s_at1.9E-035.8immunoglobulin superfamily, member 1201369_s_at6.7E-041.6zinc finger protein 36, C3H type-like 2222043_at1.1E-042.9Clusterin201497_x_at1.3E-037.9myosin, heavy polypeptide 11, smooth muscle207961_x_at4.9E-046.4myosin, heavy polypeptide 11, smooth muscle204897_at1.2E-031.9prostaglandin E receptor 4220065_at6.5E-0413.1Tenomodulin205573_s_at5.0E-041.9sorting nexin 7In a previous profiling study, we identified a list of immune response genes differentially expressed in JDM patients, in which many of the genes were induced by type I interferon (IFNα/β) [8]. Although many probe sets on the U133A microarry are different from those on the Affymetrix FL microarray, the changes identified previously were verified using the new microarray. However we found no differential expression of those IFN inducible genes in the short and long disease duration group, suggesting the expression of these inflammatory response genes is independent of disease chronicity.Infiltration of mature dendritic cells in the patients with longer duration of active symptomsTo interpret the changes of genes involved in immune responses in cluster C and D (Table 3), we first searched for genes that were highly expressed in specific immune cell types. Although most of the genes were not specifically expressed by a single cell type, HLA-DR and glycoprotein nmb have been shown to be highly expressed in dendritic cells compared to B cells, monocytes and T cells [11]. This pattern of up-regulation of the genes suggested that there might be a higher number of dendritic cells in the muscle of children with a longer duration of active disease.Using qRT-PCR, we first verified the expression changes HLA-DQA1 which could be either directly involved in antigen presentation or a surrogate marker of an active site. Because supply of diagnostic muscle biopsy tissue was limited, we tested a different group of 5 girls with short duration of disease, less than 2 months and a different group of 3 girls with disease duration greater or equal to 2 months, plus 2 girls from the original long disease duration group, which had been gene profiled in this study (Table 2).The expression level of HLA-DQA1 was 4.28 fold (p < 0.05) up-regulated in patients with longer duration of active disease, which verified our array data (5.6 fold, p < 0.005) (Figure 2). We then performed immunohistochemistry, using an antibody directed against a dendritic cell maturation marker, DC-LAMP to localize the cells in the muscle tissues (Figure 3). Long disease duration JDM patients displayed a greater overall presence of mature dendritic cells (Figure 3A) compared to short disease duration JDM patients (3C). Greater concentrations of mature dendritic cells were found in perivascular regions compared with the endomesium, In addition, many mature dendritic cells were BDCA2 positive suggesting a plasmacytoid origin. No significant differences were found in the distribution of the BDCA2 positive plasmacytoid dendritic cells in JDM of either long or short disease duration (B, D). Muscle from healthy children displayed an absence of mature dendritic cells (E), but a similar display of BDCA2 positive cells (F) compared with JDM.Figure 2Five differentially expressed genes, HLA-DQA1, smooth muscle myosin heavy chain (SMMHC), clusterin, plexin D1, and tenomodulin were verified by quantitative RT-PCR, and their level of expression compared in diagnostic muscle biopsies from 5 girls with untreated symptoms of JDM for a short disease duration < 2 months (open bars) and 5 girls with untreated symptoms of JDM of ≥ 2 months duration (black bars). * p < 0.05, **p < 0.005.Figure 3Comparisons of dendritic cells in skeletal muscle biopsies from JDM, long and short disease duration, and normal patients. DC-LAMP is a membrane bound protein produced by mature activated dendritic cells, while BDCA2 is an antigen produced by immature plasmacytoid dendritic cells (both markers are stained dark brown). Long disease duration JDM patients displayed a greater overall presence of mature dendritic cells (A) compared to short disease duration JDM patients (C). Greater concentrations of mature dendritic cells were found in perivascular and perifasicular regions compared with the endomysium. Many mature dendritic cells co-expressed plasmacytoid markers (A, B). No substantiated differences were found in the distribution of the BDCA2 positive plasmacytoid dendritic cells in JDM of either long or short disease duration (B, D). Normal pediatric muscle displayed an absence of mature dendritic cells with presence of BDCA2 positive cells (E). Images were taken at 10× on a Leica Upright Light Microscope. Scale bars represent 10 μm.Vasculature remodelling in JDM patientsBased on the array data, genes involved in vasculature remodelling, including the structural gene, SMMHC (two probe sets, 7.9 and 6.4 folds, p < 0.005), and regulatory factors, tenomodulin (13.1 fold, p < 0.001), prostaglandin E receptor 4 (1.9 fold, p < 0.005), plexin D1 (1.6 fold, p < 0.001), Akt-3 (1.3 fold, p < 0.001), clusterin (2.9 fold, p < 0.001), and LPGDS (two probe sets, 2.4 and 1.9 folds, p < 0.005) [12-19], were significantly differentially expressed between patients with long and short duration of active disease, suggesting the involvement of vasculature remodelling during this period of time (Table 3 and 4). In addition, matrix metalloproteinase 2 (2.0 fold, p < 0.005) and collagen VIα2 (1.6 fold, p < 0.005) were also up-regulated indicating extracellular matrix remodelling, which is part of angiogenesis [20,21]. Interestingly, most of these genes belonged to cluster F, in which genes were down-regulated more in patients with short duration of active disease when comparing to the baseline (Figure 1, cluster F), suggesting that the expression changes of these specific genes were preferentially misregulated in the earlier stage of the inflammatory disease process.Using qRT-PCR, we confirmed the higher expression of 4 genes related to vascular remodelling in muscle biopsies of girls with longer disease duration compared with disease of short duration. The expression level of smooth muscle heavy chain myosin was elevated 5.9 fold (p < 0.05), clusterin was 4.74 fold increased (p < 0.05); the level of plexin D1 was increased in the long duration group by 2.35 fold (p < 0.005), and tenomodulin was increased by 4.6 fold, (p < 0.05) (Figure 2).A regression analysis was performed of 79 genes and 9 genes (10 probe sets) were identified that had a statistically significant linear relationship between the level of gene expression and duration of active disease (Table 5). Both probe sets representing SMMHC showed linear relationships, suggesting that the gene expression changes are dependent on the duration of the chronic inflammation of active disease.Table 5Nine genes (10 probe sets, SMMCH are represented by two probe sets) for which the expression level correlated with the duration of untreated duration of JDMAffymetrix accessionGene descriptionr squarep-valueCluster207961_x_atmyosin, heavy polypeptide 11, smooth muscle0.470.002F201369_s_atzinc finger protein 36, C3H type-like 20.400.005F203742_s_atthymine-DNA glycosylase0.390.006C214430_atgalactosidase, alpha0.340.011C201497_x_atmyosin, heavy polypeptide 11, smooth muscle0.330.013F204807_attransmembrane protein 50.280.025A214437_s_atserine hydroxymethyltransferase 2 (mitochondrial)0.270.027E215376_atCDNA FLJ12295 fis, clone MAMMA10018180.270.027D202854_athypoxanthine phosphoribosyltransferase 1 (Lesch-Nyhan syndrome)0.250.034E221931_s_atSEH1-like (S. cerevisiae)0.230.042ADiscussionImpact of duration of untreated inflammation on clinical and laboratory findings in children with JDMPrevious studies have confirmed the observation that prolonged duration of untreated symptoms in children with JDM are highly associated with the development of pathologic calcifications, one of the most debilitating complications of JDM [22,23]. Untreated children with a longer duration of chronic symptoms had a higher frequency of loss of nailfold capillary end row loops [24], which persisted after 36 months of therapy [25]. The chronic inflammation associated with nailfold capillary loss was accompanied by impaired absorption of orally administered corticosteroids [26]. In addition, chronic untreated inflammation at diagnosis of JDM was associated with severe bone loss and an abnormal osteoprotegerin: RANKL ratio [10]. A study of 166 previously untreated children with JDM enrolled in a large national NIAMS JDM Research Registry showed that the clinical presentation of the children with shorter duration of active JDM was quite different from those who came to diagnosis later in their disease course. Children diagnosed early in the disease course were more likely to be weaker than later on, and muscle enzymes were more likely to be in the normal range 2–4.7 months after the child's first symptom (rash or weakness) [10] which may be associated with differential utilization of the TRAIL apoptotic pathway [10,27]. In this study we used a 2 month cut-off and identified 79 genes differentially expressed in patients with short (less than 2 months) and long (greater than 2 months or equal) duration of active disease by expression profiling. Identifying these changes provide insight concerning the observed differences in clinical phenotypes at the molecular level over time.Down regulation of genes involved in protein turnover and metabolismDown-regulation of protein turnover and stress response genes in skeletal muscles has been associated with muscular dystrophies and conditions associated with muscle atrophy [28,29]. In this study, genes involved in both protein turnover and stress responses were down-regulated significantly in patients with longer duration of active disease [see Additional file 1]. The genes encoding heat shock proteins (eg. heat shock 70 kDa protein 9B and heat shock 40) and genes involved in proteasome functions (eg. proteasome regulatory particle subunit p44S10, ubiquitination factor E4B, located in cluster A) showed slight upregulation in the short duration group compared to controls, but approached the baseline in the long duration group, reflecting the acute phase of the disease. In contrast, genes involved in protein synthesis (eg. eukaryotic translation initiation factor 2, mitochondrial ribosomal protein S10 and L15) were mildly downregulated in the short duration group and achieve significant down-regulation in the long duration group, suggesting continuous suppression of protein synthesis commonly associated with muscle atrophy. The changes at the molecular level support the observation of ongoing muscle loss in untreated patients with JDM. Gene involved in ER stress response have been previously reported to be differentially expressed in adults with inflammatory myopathies [30]. In our study, we did not observe significant differences between the groups, based on the duration of untreated disease.Dendritic cell infiltrationThe presence of dendritic cells in the muscle of patients with dermatomyositis was previously reported [31,32]. Recent investigation presented data showing that the localization and maturation of resident plasmacytoid dendritic cells in situ in the perivascular areas was essential to the initiation and perpetuation of muscle inflammation in juvenile DM [31]. The present study confirms that the typical pattern of dendritic cell infiltration (with heavy infiltration surrounding the blood vessels) in JDM is a progressive process associated with persistent inflammation without focal infiltration in muscle of patients with short duration of active disease. Many DC-LAMP positive dendritic cells were positive for BDCA2, again confirming that they were of plasmacytoid dendritic cell lineage as recently reported [31]. Dendritic cells are well characterized for efficient antigen presentation to T cells and are speculated to affect both tolerance and T cell activation [33]. Models of autoimmunity through activated dendritic cells in rheumatoid arthritis and in systemic lupus erythematosus (SLE) associated with a Type 1 IFN-α/β initiated response have been proposed by other investigators [34-36]. In addition, increased expression of IFNα/β induced genes have been identified in peripheral blood of children with SLE, systemic onset juvenile rheumatoid arthritis, and in JDM [37-39]. In this study, we found large number of dendritic cells in perivascular areas with dense mononuclear infiltrates in biopsies from children with longer duration of active disease, which suggested that dendritic cells might be actively involved in the progression of the disease by modulating T cells function in the muscle vasculature.Vasculature remodellingVascular smooth muscle cells (SMCs) are highly plastic capable of profound alterations in phenotype in response to changes in local environmental cues [40]. Vascular injury initiates a transition in the phenotype of vascular SMCs whereby \"contractile\" (differentiated) SMCs are capable of undergoing transient modification to a highly \"synthetic\" (dedifferentiated) state. These synthetic vascular SMCs are migratory, highly proliferative, and play a critical role in repair of the vascular injury. Upon resolution of the injury, SMCs reacquire their contractile phenotype and associated markers, which include smooth muscle isoforms of contractile apparatus proteins such as SMMHC [40]. In this study, we found clusterin, tenomodulin and prostaglandin E receptor 4 were down-regulated in JDM patients with short disease duration. In contrast, for children with longer duration of active disease, expression of these genes approached the baseline, and additional genes involved in vasculature remodelling were up-regulated, (plexin D1, Akt-3, LPGDS). These genes are associated with contractile (differentiated) vascular SMCs. These data suggest that early in the course of JDM, a relatively high proportion of vascular SMCs at the site of inflammation are in a synthetic (dedifferentiated or undifferentiated) state; conversely, as time passes and the inflammation becomes chronic, vascular SMCs further differentiate. The linear regression relationship between the duration of active disease and the SMMHC expression suggests that the vasculature remodelling in the muscles is a progressive process associated with disease chronicity and the duration of active disease. This is consistent with the documentation of evidence of cardiovascular compromise in older patients who had JDM in childhood [41]. Among the genes involved in vasculature remodelling, plexin D1 and prostaglandin E receptor 4, are two potent angiogenic factors [12,19]. Plexin D1 is expressed in vascular endothelial cells of developing blood vessels. Signalling between semaphorin 3E and its receptor plexin D1 controls endothelial cell positioning and the patterning of the vasculature during embryonic development, and can negatively regulate angiogenesis as well as activate T cells [42]. Because the hallmark of JDM is occlusion and obliteration of capillaries and arterioles, it is not surprising that expression of tenomodulin is increased, for it is overexpressed in hypovascular connective tissue. The up-regulation of these genes might underlie the molecular mechanisms of the vasculature loss and remodelling we observed in the patients' muscles.In summary, the findings from this study support the clinical observation that identification of the duration of the inflammatory process is a critical component in children with JDM, influencing both the gene expression and the pathophysiology of the immune responses in children with active symptoms. Recognition of this variable, which identifies the chronicity of the process, is important in dissecting out factors involved in immune progression and response.ConclusionWe conclude that the duration of the chronic inflammatory process in untreated JDM alters mRNA expression patterns, including both dendritic cell maturation and vascular remodelling with increased expression of anti-angiogenic factors. We propose that the duration of the inflammatory process must be considered when interpreting gene profiling data as well as clinical and laboratory findings in children with JDM in order to gain insight into potential modes of intervention. We speculate that interventions that diminish the antiangiogenic remodelling present in children with JDM who have a longer duration of untreated disease may be of benefit.MethodsPatient populationAge appropriate informed consent was obtained from a total of 31 girls with definite/probable JDM (IRB# 2002-11762) and the 4 healthy age and sex-matched controls (IRB# 2001-11715) who were enrolled in this study. Ethical approval for this study was obtained from the Children's Memorial Hospital Institutional Review Board, and all procedures were carried out in accordance with the Helsinki Declaration.All the girls with JDM were negative for myositis specific or associated antibodies or for antibodies indicating overlap syndromes at the time of biopsy, and, as part of their diagnostic evaluation, had an MRI directed muscle biopsy. In the first step, 23 muscle biopsies from partially treated (n = 4) patients and untreated (n = 19) JDM patients and 4 control biopsies were expression profiled and used for gene filtering. In the second step, only untreated JDM patients at the time of muscle biopsy (n = 19, 3 short and 16 long duration) were analyzed statistically. In the third step, specific genes identified by the expression profiles were confirmed by testing additional samples from 8 untreated children with JDM. There was enough muscle biopsy material from 2 of the rare group of children with disease duration less than 2 months to be used for qRT-PCR enabling comparison of 5 patients each with long and short disease duration, as well as immunohistochemical studies.The date of recognition of the first symptom (rash or weakness) was defined as the \"disease onset date\". The \"duration of disease\" at the time of the muscle biopsy was defined as the time from disease onset to the date of muscle biopsy. We have reported the results from profile analyses performed using only samples from untreated girls with JDM; their demographics are presented in Table 1.Confirmation of the gene profiles by q-RT PCR utilized a separate set of 5 muscle biopsies from girls with a short duration of disease (1.4 ± 0.4 SD months), matched with 5 muscle biopsies' from girls with a long duration of untreated symptoms (6.8 ± 3.5 SD months); 2 children tested were also part of the gene profile long duration group. The girls were matched for age (8.2 ± 2.2 years compared with 9.5 ± 3.5 years), as well as disease activity, with respect to DAS skin (5.4 ± 0.9, 5.8 ± 1.5 respectively) and DAS muscle (8.2 ± 0.8, 8.0 ± 1.2 respectively). Details are presented in table 2.The four control muscle from non-inflammatory female donors were biopsies, obtained with informed consent, from thoracic muscle from girls undergoing plastic surgery for repair of cleft palate, which do not appear to have evidence of the genetic dysregulation observed in JDM [43]. Their age range was 8–10.3; their mean age was 9.5 years.Disease Activity ScoresOverall measure of severity of the JDM disease activity was assessed at the time of the diagnostic muscle biopsy using the total disease activity score (DAS), a 20 point scale which has two sub-scales, which reflect skin involvement (ranging from 0–9) and muscle inflammation (ranging from 0–11) [4]. The skin component (DAS skin) is based on extent and severity of rash, the presence of telangiectasia (nailfold, palate, eyelids) and Gottron's papules [4] The muscle component (DAS muscle) includes measures of muscle function and the extent of weakness in eight manoeuvres as evaluated by a single physician (LMP) on routine diagnostic physical examination. Both sub scores have been validated for inter-rater reliability.Muscle biopsy samplesA diagnostic muscle biopsy, frequently the vastus lateralis, was obtained from the area of inflammation as defined by an MRI, using a T-2 weighted image with fat suppression. The sample was divided so that one portion was saved for immunohistochemical studies, while the other portion of the sample was used for gene expression profile studies and gene confirmation. Both parts of the muscle biopsy samples were snap frozen and stored in liquid nitrogen (-180°C).Determination of DQA1*0501 and DQA1*0301At the diagnostic visit, immediately prior to the muscle biopsy, peripheral blood mononuclear cells were obtained by Ficoll-Hypaque separation, and frozen in liquid nitrogen at -180°C.Genomic DNA was extracted from whole blood or frozen lymphocytes using the Puregene DNA Purification Kit (Gentra Systems, Minneapolis, MN). DNA from JDM patients was genotyped for HLA-DQA1*0301 and HLA-DQA1*0501 alleles using PCR amplification with sequence-specific primers as previously reported [44,45]. Aliquots of DNA (2 μl) were amplified using TaqBead™ Hot Start Polymerase wax beads (1.25 U/bead; Promega Corporation, Madison, WI) in a reaction volume of 50 μl that contained 5× Green GoTaq™ Reaction Buffer (Promega Corporation, Madison, WI), dNTPs (0.2 mM each), and 0.4 μM allele specific primers for HLA-DQA1*0301 or HLA-DQA*0501.The primers for HLA-DQA1*0301 were: 5'-TTCACTCGTCAGCTGACCAT-3' (Forward) and 5'-CAAATTGCGGGTCAAATCTTCT-3' (Reverse), which amplify a 183 bp product. The primers for HLA-DQA*0501 were: 5'-ACGGTCCCTCTGGCCAGTA-3' (Forward) and 5'-AGTTGGAGCGTTTAATCAGAC-3' (Reverse), which amplify a 186 bp product.Denaturation was performed in each PCR at 94°C for 45 seconds, annealing at 62°C for 45, and extension at 72°C for 2 min (for the final step, extension was for 7 min). Absence or presence of PCR products was visualized by agarose gel electrophoresis.Determination of the TNF-α-308 polymorphismThe TNF-α-308 polymorphism consists of a single base pair G to A substitution. PCR was used to amplify a 107 bp fragment that incorporated the polymorphic site into an NcoI restriction site as previously described [23,45] distinguishes AA from AG from GG.Expression profiles of JDM patient muscle biopsiesTotal RNA was isolated from each biopsy, processed for production of biotinylated cRNA and hybridization to microarrays, as we have previously described [28]. Each sample was then hybridized to Affymetrix U133A microarrays containing approximately 14,500 well-characterized transcripts. Standard operating procedure and quality control was done as previously described [46]. Muscle samples from 23 female JDM patients and 4 healthy age- and sex-matched controls were initially profiled.Generation of hybridization signals (probe set algorithms) of the microarrays was done using Affymetrix MAS (Version 5.0) (Affymetrix, CA), and dCHIP [47]. After the absolute analysis, the gene expression levels were imported into GeneSpring software. The JDM samples were normalized to the mean of the profiles of age- and sex-matched control samples. Data filtering was done by retaining only those probe sets that showed at least two MAS5.0 \"present calls\" across all profiles. This resulted in retention of 67% of probe sets for the U133A microarrays.Welch t-test was used to calculate the probabilities of significant gene expression changes between samples with shorter (< 2 months) and longer (≥ 2 months) disease duration. To reduce false positives, correction for multiple testing was done using Benjamini and Hochberg false discovery rate (5%) [48,49]. In addition, we used all treated and untreated patients to first generate a gene list with genes that showed statistically significant changes to reduce number of genes for multiple testing. We then used only the profiles of untreated patients and performed t-tests on the filtered genes with multiple testing corrections to minimize the false positives. To visualize transcripts showing coordinate regulation as a function of active disease duration, genes sharing temporal patterns were identified by hierarchical clustering using GeneSpring software. Clustering algorithm was based on standard correlation (r = 0.95). For hierarchical clustering, we included all genes with p < 0.05 after multiple testing correction. In order to determine the presence of significant linear relationships between gene expression and duration of untreated disease, linear regression was performed between the gene expression levels of each of the 79 genes and duration of untreated disease. All profiles are publicly accessible via NCBI GEO  (GSE11971).Quantitative Real Time-PCR verification (qRT-PCR)Total cellular RNA was extracted using Trizol Reagent (Invitrogen Corp.) and subsequently DNase treated using DNA-free (Ambion, Austin, TX). Reverse transcription reactions were performed using Superscript III Reverse Transcriptase (Invitrogen Corp., Carlsbad, CA) and random hexamer primers. Relative cDNA quantification of smooth muscle myosin heavy chain (SMMHC), clusterin and an internal reference gene, β-actin, were done using a TaqMan PCR Core Reagent Kit (Applied Biosystems; Roche Molecular Systems, Inc., New Jersey) using fluorescence-based detection method (Applied Biosystems 7500 Fast Real-Time PCR System; Applied Biosystems, Foster City, CA). The PCR reaction was performed using standard methodology as previously described for each gene of interest and the β-actin reference gene was used to normalize input cDNA.The primers and probe for SMMHC are as follows: 5'-CTGGGCAACGTAGTAAAACC-3' (Forward), 5'-TATAGCTCATTGCAGCCTCG-3' (Reverse), and 6FAM-ATAAGCTGGGCGTGGTGGTACACACCT-TAMRA (Probe) [50].The primers and probe for Clusterin are as follows: 5'-GAGCAGCTGAACGAGCAGTTT-3' (Forward) 5'-CTTCGCCTTGCGTGAGGT-3' (Reverse) and 6FAM-ACTGGGTGTCCCGGCTGGCA-TAMRA (Probe) [51]The primers and probe for β-actin are as follows: 5'-TGAGCGCGGCTACAGCTT-3' (Forward) 5'-TCCTTAATGTCACGCACGATTT-3' (Reverse) and 6FAM-ACCACCACGGCCGAGCGG-TAMRA (Probe) [52]For relative cDNA quantification of class II antigen, HLA-DQA1 (#QT00060130), plexin D1 (#QT00036134), tenomodulin (#QT01024590), and an internal reference gene, β-actin (#QT00095431), QuantiTect Primer Assays and QuantiFast Syber Green PCR kits (Qiagen Inc., Valencia, CA) were used (Applied Biosystems 7500 Fast Real-Time PCR System; Applied Biosystems, Foster City, CA). The PCR reaction was performed according to manufacturer's protocol for these genes using β-actin reference gene to normalize input cDNA.Immunohistochemistry assay for mature dendritic cellsThree untreated girls with long duration of active JDM, were age-race matched with three girls with duration of untreated symptoms of less than 2 months and 3 age-matched healthy female muscle donors and studied for presence of mature dendritic cells. Serial 6 um-thick frozen muscle sections were fixed in cold anhydrous acetone. Sections were then blocked for 30 minutes in 10% normal goat or donkey sera and incubated with primary antibody overnight at 4°C. Monoclonal antibodies against DC-LAMP (Beckman coulter, CA), raised in mice, were used at a dilution of 1:10. Polyclonal antibody against BDCA2 (Santa Cruz Biotechnology, CA), raised in goats, were used at a dilution of 1:50. After 3 washes with 1× PBS, slides were incubated for 1 hour at room temperature with biotin conjugated secondary antibodies against mouse and goat respectively (Jackson Immunoresearch, PA). Subsequently, slides were stained with Vectastain Elite (Vector Laboratories, CA), followed by a BioGenex Liquid DAB Substrate Kit (yielding brown coloration at site of positive antibody binding). The slides were then counterstained with haematoxylin and Scott's bluing solution (Ricca Chemical, Texas), dehydrated, and prepared for viewing.Authors' contributionsY-WC oversaw the expression profiling, qRT-PCR and IHC experiments, determined data analysis strategies and performed the analysis, interpreted the profiling, qRT-PCR and IHC data, and prepared the manuscript. RS performed expression profiling and immunohistochemistry. HG-D performed statistical analyses and interpreted results. SS performed the qRT-PCR, supervised the muscle biopsy selection for study and wrote some of the experimental methods. NG performed the immunohistochemistry. LMP created the IRB protocol, oversaw the study, recruited the patients and supervised the biopsy collection, qRT-PCR and immunohistochemistry assays and participated in the literature review and manuscript preparation. All authors have read and approved the final manuscript.Supplementary MaterialAdditional file 1Supplemental table 1. Complete gene list of genes in cluster A-F. All genes are in order shown in figure 1.Click here for file\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2529273.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2529273",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2529273\nAUTHORS: Dmytro Berezhnoy, Maria C Gravielle, Scott Downing, Emmanuel Kostakis, Anthony S Basile, Phil Skolnick, Terrell T Gibbs, David H Farb\n\nABSTRACT:\nBackgroundCompounds targeting the benzodiazepine binding site of the GABAA-R are widely prescribed for the treatment of anxiety disorders, epilepsy, and insomnia as well as for pre-anesthetic sedation and muscle relaxation. It has been hypothesized that these various pharmacological effects are mediated by different GABAA-R subtypes. If this hypothesis is correct, then it may be possible to develop compounds targeting particular GABAA-R subtypes as, for example, selective anxiolytics with a diminished side effect profile. The pyrazolo[1,5-a]-pyrimidine ocinaplon is anxioselective in both preclinical studies and in patients with generalized anxiety disorder, but does not exhibit the selectivity between α1/α2-containing receptors for an anxioselective that is predicted by studies using transgenic mice.ResultsWe hypothesized that the pharmacological properties of ocinaplon in vivo might be influenced by an active biotransformation product with greater selectivity for the α2 subunit relative to α1. One hour after administration of ocinaplon, the plasma concentration of its primary biotransformation product, DOV 315,090, is 38% of the parent compound. The pharmacological properties of DOV 315,090 were assessed using radioligand binding studies and two-electrode voltage clamp electrophysiology. We report that DOV 315,090 possesses modulatory activity at GABAA-Rs, but that its selectivity profile is similar to that of ocinaplon.ConclusionThese findings imply that DOV 315,090 could contribute to the action of ocinaplon in vivo, but that the anxioselective properties of ocinaplon cannot be readily explained by a subtype selective effect/action of DOV 315,090. Further inquiry is required to identify the extent to which different subtypes are involved in the anxiolytic and other pharmacological effects of GABAA-R modulators.\n\nBODY:\nBackgroundGABAA receptors (GABAA-R) are pentameric membrane proteins that belong to the superfamily of cys-loop ligand-gated ion channels (LGIC), which operate as GABA-gated Cl--selective channels. GABAA-R mediate most of the fast inhibitory neurotransmission in the CNS [1-3]. Initially, two subunits of the GABAA-R named α and β were purified [4,5] and subsequently their cDNAs were isolated [6]. Twenty related GABAA-R subunits have been so far identified in mammals (α1–6, β1–4, γ1–3, δ, ε, π, θ, and ρ1–3 [7,8]), yielding a high degree of potential diversity. If all of these subunits could randomly co-assemble, more than one hundred thousand GABAA-R subtypes with distinct subunit composition and arrangement would be formed [9]. The composition of the most abundant GABAA-R type in the CNS is αβγ, and immunohistochemistry studies suggest that receptors containing α1, β2/3 and γ2 subunits are the most widespread GABAA-R subtype in adult mammalian brain and represent about 50% of the total receptor pool [2,10].Typical αβγ GABAA-Rs harbor two agonist (GABA) binding sites located at the two α/β subunit interfaces [2,11]. The function of GABAA-Rs can be modulated by various compounds acting at different allosteric sites located on GABAA-Rs. The benzodiazepine (BZD) site, which is located at an α/γ interface [12,13], is the most frequently targeted site for therapeutic agents, and ligands that enhance GABAA-R function through positive modulation at this site possess anxiolytic, sedative, myorelaxant, anesthetic and amnestic properties [2,3,10,14]. Based on pharmacological studies in transgenic mice, it has been proposed that GABAA-Rs can be classified into the following pharmacological classes according to the effects of BZ site ligands: α1-containing receptors (GABAA1) that mediate sedative effects; α2-containing receptors (GABAA2) that mediate anxiolytic effects; α3-containing receptors (GABAA3) that mediate myorelaxation; and α5-containing receptors (GABAA5) that are involved in learning and memory processes [7,15,16]. This classification is consistent with the sedative/hypnotic profile of GABAA1-preferring compounds such as zolpidem and zaleplon [17], but pharmacological studies in wild-type animals and in man have raised questions regarding the attribution of anxiolytic effects to GABAA2 receptors. In particular, a number of compounds have been identified that exhibit an anxioselective profile in vivo despite lacking the expected GABAA2 selectivity. A series of compounds with mixed preference for α2/α3-containing receptors has been reported to produce robust anxiolysis in animals without noticeable sedation, including one compound that exhibits selectivity for α3-containing receptors [18-21]. Other compounds, such as ocinaplon [22] and DOV 51,892 [23], are anxiolytic in humans and animals without undesired side effects such as sedation and myorelaxation, but do not exhibit strong selectivity among GABAA-Rs sensitive to benzodiazepines (that is, those receptors containing α1–3 and/or α5-subunits)One hypothesis that could explain the anxioselective profile of ocinaplon is the presence of one or more biotransformation products that exhibit selectivity at GABAA2 receptors. To test this hypothesis, we characterized the pharmacological properties of the major biotransformation product of ocinaplon in dogs, rats and man, DOV 315,090 (Fig. 1), using in vitro radioligand binding and two-electrode voltage-clamp electrophysiology. We now report that like its parent compound, DOV 315,090 acts as a positive modulator at GABA receptors, and like its parent, does not exhibit marked selectivity among α1–3 and α5 containing receptors. Thus, while DOV 315,090 may contribute to the pharmacological actions of ocinaplon, the anxioselective profile of ocinaplon cannot be explained on the basis of enhanced subunit selectivity on the part of DOV 315,090.Figure 1Structures of diazepam, ocinaplon and DOV 315,090.MethodsRadioligand Binding AssaysHEK293 cells (CRL 1573, American type Culture Collection, Rockville, MD, USA) were cultured in Dulbecco's modified Eagle's medium (D-MEM, Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (Invitrogen, Carlsbad, CA, USA) and 1% MEM Non-Essential Amino Acids Solution (Invitrogen, Carlsbad, CA, USA). cDNAs encoding rat GABAA-R subunits were in the following vectors: α1 and α5 in pRc/CMV, α2, α3, γ2S and γ3 in pcDNA3 and β2 in pcDNA1. The cells were transiently transfected (5 μg of each cDNA per 100 mm dish) using FuGene™ (Roche Diagnostics Corporation) at a 3:1 FuGene:DNA ratio. Transfection efficiency was 50–80% as measured by co-transfection with green fluorescent protein cDNA (2.5 μg/100 mm dish). Forty-eight hours after transfection, cells were washed with ice-cold PBS, harvested and homogenized. Cell homogenates were centrifuged (100,000 g, 25 min) and washed three times by homogenization in ice-cold PBS buffer followed by centrifugation at 100,000 g for 25 min. The final pellets were stored at -20°C until needed.For competition binding, 100 μg of membrane protein was incubated in 500 μl of PBS buffer with 0.5 nM [3H]Ro15–1788 (78.6 Ci/mmol, PerkinElmer Life Sciences) in the presence of diazepam (1 nM – 10 μM, Sigma-Aldrich), ocinaplon (0.1 – 250 μM, DOV Pharmaceuticals) or DOV 315,090 (0.1 – 50 μM, DOV Pharmaceuticals) for 1 h at 0°C. The samples were then diluted with 5 ml of ice-cold buffer and filtered under vacuum through glass-fiber filters (GF/B Whatman). Filters were washed 3 times with 5 ml of buffer and the radioactivity was quantitated by liquid scintillation counting in 5 ml of Ecolite scintillation fluid (ICN). Non-specific binding determined in the presence of 100 μM Ro 15–1788 (Sigma-Aldrich) was subtracted from total binding to calculate specific binding. Data were analyzed by non-linear regression (Prism, Graph-Pad software).Recording of GABA-Gated Currents from GABAA Receptors Expressed in Xenopus OocytescRNAs encoding GABAA-R α1, α2, α3, α5, β2 and γ2S subunits were injected into oocytes from Xenopus laevis. Forty-eight hours later, measurements of the effects of diazepam, ocinaplon and DOV 315,090 on GABA-gated Cl- currents from oocytes expressing GABAA-Rs were performed using a Warner TEVC amplifier (Warner Instruments, Inc., Foster City, CA) (Park-Chung et al., 1999). GABA (Sigma) was prepared as a 1 M stock solution in ND96. Microelectrodes of 1–3 MΩ when filled with 3 M KCl were used to record from oocytes in a recording chamber continuously perfused with ND-96 buffer solution. During data acquisition, oocytes were clamped at a holding potential of -70 mV. Drugs were applied by perfusion at a rate of approximately 50 μl sec-1 for 20 s followed by a 120 s wash. At the end of each experiment 3 μM of diazepam was applied as a potentiation control. All experiments were performed at room temperature (22–24°C).GABA concentration-response data was obtained for each subunit combination, and the GABA EC10 was determined by nonlinear regression using the logistic equation. This concentration of GABA was used for modulation studies. Peak current measurements were normalized and expressed as a fraction of the peak control current measurements. Control responses to an EC10 concentration of GABA were re-determined after every 2 – 4 applications of modulator + GABA. Percent potentiation is defined as [I(GABA + Drug)/IGABA)-1] × 100, where I(GABA + Drug) is the current response in the presence of diazepam, and IGABA is the control GABA current. Potentiation data from each oocyte was fitted to the equation Potentiation = Emax × [Drug]/([Drug + EC50) by non-linear regression (Prism, Graph-Pad software). Due to a decline in the response at high diazepam concentrations, concentrations of diazepam above 3 μM were excluded from the fit. Some oocytes expressing α1β1γ2 receptors appeared to exhibit a biphasic modulatory response to diazepam, suggesting the possible presence of an additional component of modulation with a sub-nM EC50. For 6 of 8 oocytes, the fit was significantly improved by adding a second, higher-potency component of modulation, but the affinity of this second component was not well resolved in fitting due to its small amplitude. Given the lack of consistency of this possible high affinity effect, we have omitted it in fitting our concentration-effect curves. The choice of fitting to a monophasic or biphasic equation had only a small effect on the EC50 for the major component of modulation. For diazepam, the mean EC50 of the major component was increased from 35 nM to 42 nM when a two-component fit was used for those oocytes in which it produced a significant improvement in the sum of squares.ResultsBiotransformation of ocinaplon into DOV 315,090 in vivoAs shown in Figure 2, DOV 315,090 appears rapidly in plasma following i.v. or oral administration of a behaviorally active dose of ocinaplon (5 mg/kg) to rats. At 1 h, corresponding to the time at which the anticonflict effect of ocinaplon was evaluated [22], the plasma concentration of DOV 315,090 is ~38% of the concentration of parent compound.Figure 2Pharmacokinetics of ocinaplon and DOV 315,090. Blood levels of ocinaplon (●,○) and DOV 315090 (▲,△) were determined at various times after i.v. (●,▲) or oral (○,△) administration of 5 mg/kg ocinaplon to rats. Plotted results do not include one animal that exhibited a low blood level (0.47 μg/ml) of ocinaplon at the initial 10 min time point after oral administration and proportionally lower levels of both compounds throughout the duration of the experiment. This animal may have regurgitated a portion of the dose (of the suspension).Comparison of the binding properties of diazepam, ocinaplon and DOV 315,090Figure 3 and Table 1 document the binding properties of diazepam, ocinaplon and DOV 315,090 in HEK293 cells expressing different GABAA-R subunit combinations. Examination of binding constants shows that ocinaplon and DOV 315,090 have lower affinity than diazepam at all of the receptor subunit combinations tested. The binding profile of DOV 315,090 is similar to that of ocinaplon, with little selectivity among the subunit combinations tested. In contrast to diazepam, which exhibits markedly lower affinity for α1β2γ3 and α2β2γ3 receptors than for α1β2γ2 s and α2β2γ2 s receptors, replacement of γ2S with γ3 had little effect on the affinity of either ocinaplon or DOV 315,090 for any subunit combination (Table 1). Also, whereas diazepam has similar affinity for α1-containing and α2-containing receptors, both ocinaplon and DOV 315,090 have 3–4 fold lower affinity for α2-containing receptors. Specific [3H]Ro15–1788 or [3H]flunitrazepam binding to membrane preparations from cells transfected with α3, β2 and γ3 subunits was not detected, suggesting that these subunits failed to assemble in the HEK293 cells.Figure 3Displacement curves of [3H]Ro 15–1788 binding by diazepam (DZ), ocinaplon and DOV 315,090 in homogenates of HEK293 cells transfected with different subunit combinations. Smooth curves are calculated from the mean parameter values in Table 1.Table 1Binding affinity of diazepam, ocinaplon and DOV 315,090 for GABAA-Rs with different subunit composition.Receptor Typeα1β2γ2Sα1β2γ3α2β2γ2Sα2β2γ3α3β2γ2Sα5β2γ2Sα5β2γ3diazepam (DZ)IC50 (μM)0.030.220.040.210.050.030.09pIC507.54 ± 0.096.67 ± 0.087.50 ± 0.106.80 ± 0.267.32 ± 0.087.57 ± 0.137.09 ± 0.13ocinaplon (OC)IC50 (μM)6.32.324207.79.610pIC505.20 ± 0.145.65 ± 0.014.62 ± 0.144.74 ± 0.155.12 ± 0.065.02 ± 0.035.01 ± 0.18IC50/DZ IC5021810.5759115158355120DOV 315,090IC50 (μM)7.05.524209.32227pIC505.19 ± 0.125.27 ± 0.074.63 ± 0.054.72 ± 0.095.08 ± 0.144.67 ± 0.084.58 ± 0.09IC50/DZ IC5022025760120170790323IC50/OC IC501.022.400.890.981.092.242.59IC50 values were calculated from [3H]Ro15–1788 displacement curves using non-linear regression analysis for each independent experiment. pIC50 values are averages of the negative logarithms of IC50s. Results from each experiment (n = 3) were fitted independently and fitted parameters were averaged to calculate means and SEM. EC50 values were averaged as their negative logarithms (pIC50).Modulation of GABAA-R function by diazepam, ocinaplon and DOV 315,090Consistent with previous studies [22,23], the potency and efficacy of ocinaplon were lower than diazepam at the four receptor subtypes analyzed. The highest efficacy was observed at receptors containing α3 subunits (Table 2). DOV 315,090 also exhibited the highest maximal potentiation at α3-containing receptors; however, its Emax values were similar to those of diazepam at receptors containing α1 or α3 subunits (Table 2).Table 2Properties of diazepam, ocinaplon and DOV315090 determined by two-electrode voltage clamp electrophysiology using Xenopus oocytes injected with cRNA.Receptor Typeα1β2γ2Sα2β2γ2Sα3β2γ2Sα5β2γ2Sdiazepam (DZ)EC50 (μM)0.04 (8)0.03 (10)0.092 (5)0.025 (5)pEC507.46 ± 0.077.60 ± 0.0447.04 ± 0.057.51 ± 0.11Emax, %144 ± 8.0157 ± 14232 ± 31224 ± 24ocinaplon (OC)EC50 (μM)2.93 (4)9.12 (5)8.01 (4)3.5 (4)pEC505.57 ± 0.115.04 ± 0.035.16 ± 0.145.48 ± 0.07EC50/DZ EC507735087139Emax, %132 ± 8150 ± 6181 ± 1884 ± 4Emax/DZ Emax0.910.950.780.37DOV315090 (MET)EC50 (μM)4.87 (4)12.5 (4)10.21 (4)10.14 (4)pEC506.32 ± 0.054.92 ± 0.095.00 ± 0.055.03 ± 0.10EC50/DZ EC50128482111405EC50/OC EC501.661.371.272.92Emax, %192 ± 4139 ± 23 *340 ± 35 *68 ± 8Emax/DZ Emax1.330.881.460.30Emax/OC Emax1.450.921.870.81Drugs were prepared from DMSO stock solution prior to experiment, EC10s of GABA were used, errors are SEM of fitted parameter values from the number of oocytes given in parentheses. Results from each oocyte were fitted independently and fitted parameters were averaged to calculate means and SEM. EC50 values were averaged as their negative logarithms (pEC50) * For these two cases, the extrapolated Emax exceeded the observed maximum observed potentiation by over 25%, but parameter SEM was not substantially increased, indicating that range of concentrations was adequate to project Emax. Higher drug concentrations could not be used due to solubility constraints.DOV 315,090 and ocinaplon exhibited similar efficacies (150% vs. 139% potentiation, respectively) and EC50s (12.5 μM vs. 9.12 μM, respectively, n = 4) at α2β2γ2S receptors (Figure 4, Table 2). In contrast, whereas ocinaplon and DOV 315,090 were approximately equipotent at α3β2γ2S receptors (EC50 = 8.01 μM and 10.21 μM, respectively), the efficacy of DOV 315,090 was almost 1.87 fold greater than that of ocinaplon (340% vs 181% potentiation) (Figure 4, Table 2). Finally, DOV 315,090 was less efficacious and potent than ocinaplon at α5β2γ2S receptors (Figure 4, Table 2). The rank order of potency (EC50) of the pyrazolopyrimidines at enhancing GABA-gated chloride currents in receptors containing different α subunits was: α2≈α3≈α5 < α1 for DOV 315,090, compared to α2≈α3 < α5≈α1 for ocinaplon. Furthermore, DOV 315,090 and ocinaplon had different efficacy (Emax) profiles; the rank order of absolute efficacy was α5 < α2 < α1 < α3 for DOV 315,090, as compared with α5 < α1 < α2 < α3 for ocinaplon.Figure 4Potentiation of GABA-gated currents by diazepam, ocinaplon and DOV 315,090. Rat GABAA-Rs consisting of α1β2γ2S, α2β2γ2S, α3β2γ2S and α5β2γ2S subunits were expressed in Xenopus oocytes. Potentiation was determined using an EC10 concentration of GABA (~10 μM for α1β2γ2S, α2β2γ2S and α3β2γ2S; ~5 μM for the α5β2γ2S). Curves were calculated by normalizing values of relative currents obtained following administration of diazepam (○), ocinaplon (●) or DOV 315,090 (□) in the presence of GABA (from at least four oocytes harvested from at least two batches) to the value obtained following application of GABA. The dose-response curves of diazepam were fitted up to 3 μM. Higher concentrations (in parentheses) were excluded from the fit due to a decline in potentiation at higher concentrations. Smooth curves are calculated based on mean parameter values given in Table 2. Asterisks indicate fits for which the extrapolated Emax is more than 25% greater than the maximum potentiation observed at highest drug concentration.DiscussionIn the CNS, classical 1,4-BZDs such as diazepam, as well as other ligands of the BZD binding site, act on GABAA-Rs that are composed of α, β, and γ subunits. The majority of GABAA receptors contain α1–6, β2/3 and γ2 subunits, whereas the β1 and γ1/3 subunits have very restricted patterns of expression [2]. It has been shown that BZD pharmacology is primarily dependent upon the α subunit subtype present (α1–3 or α5), whereas receptors containing α4 or α6 subunits are insensitive to \"classical\" 1,4-BZDs [7,24,25]. Studies of animals in which genes coding for specific α subunits have been deleted or mutated to eliminate BZD sensitivity (e.g. the α1H101R mutation, which disrupts the BZD binding site) led to the hypothesis that the sedative effects of the BZDs are mediated by α1-subunit containing receptors (designated GABAA1-R), whereas anxiolytic effects are mediated by α2-subunit containing receptors (GABAA2-R) [7,17,26,27]. GABAA-Rs containing α5 subunits are thought to be responsible for the impairment of learning and memory that is induced by BZDs [28]. These finding raised the attractive prospect that BZD-like drugs that specifically target GABAA-Rs that contain a specific α-subunit will be able to produce the intended pharmacological effect (e.g sedation or anxiolysis) with reduced incidence of side effects. Because BZD-like drugs function as allosteric modulators and do not occupy the GABA binding site, specificity may be potentially achieved on the basis of either differences in potency or on differences in modulatory efficacy at specific receptor subtypes.Compounds such as zolpidem and zaleplon, which exhibit higher affinity for α1-containing receptors relative to other subtypes, have been promoted as sedative agents, driven in part by the hypothesis that selectivity for GABAA1-Rs would translate into an improved side-effect profile, particularly with respect to tolerance, withdrawal, and abuse liability. Although these compounds are effective sedative agents, consistent with the identification of GABAA1-Rs as mediating sedation, the selectivity of these compounds for GABAA1-Rs vs. GABAA-Rs containing other α-subunits is generally an order of magnitude or less, and it is unclear to what extent the hypothesized benefits are achieved in clinical practice [17].However, the situation is less clear for compounds possessing anxiolytic properties. Recently published articles describe the pharmacological properties of two novel anxioselective compounds – ocinaplon [22] and DOV 51892 [23]. These compounds do not exhibit a marked selectivity among GABAA-Rs containing different diazepam-sensitive subunits (e.g. α1–3 and α5), yet are reported to be anxioselective, lacking sedative and myorelaxant side effects at anxiolytic doses. In particular, DOV 51892 exhibits higher efficacy than diazepam at GABAA1-Rs.The classic BZD diazepam has been shown to act with high efficacy and similar potency across a broad spectrum of GABAA-Rs [1,10,22] (Table 2). This lack of selectivity with respect to either potency or efficacy among the major GABAA-R types have been hypothesized to account for the side effects associated with the use of diazepam when used as an anxiolytic, which include sedation, myorelaxation, narcosis, and amnesia. However, as has been confirmed by in vivo behavioral studies, such side effects are not observed with ocinaplon (e.g. in motor activity test, inclined screen and rod walking) or for DOV 51892 (e.g. rotarod and grip strength tests), even at doses well in excess of those that enhanced punished responding in the thirsty rat test [22,23]. Further, ocinaplon is an effective anxiolytic in humans at doses that do not produce BZD-like side effects [22]. The present study was designed to test whether the anxioselective profile of ocinaplon is due to metabolism into subtype-selective metabolites. Our pharmacokinetic data demonstrate that in rats, the major metabolite of ocinaplon is a 4'-N' oxide, DOV 315,090. Whereas DOV 315,090 is active as a GABAA-R modulator, and its in vitro binding affinities for recombinant α1β2γ2S, α2β2γ2S, and α3β2γ2S receptors differ only marginally from ocinaplon, its affinity for α5β2γ2S receptors is only slightly lower than that of ocinaplon (~2-fold).Comparison of the pharmacological profile of ocinaplon and DOV 315,090 using two electrode voltage clamp electrophysiology (Table 2) shows that the greatest difference in efficacy occurred at α3β2γ2S receptors. Although a clear maximum was not attained due to solubility limits, the extrapolated maximum potentiation by DOV 315,090 was 1.87-fold greater, followed by a 1.45-fold difference at α1β2γ2S receptors compared to ocinaplon. In contrast, maximum potentiation by DOV 315,090 was lower than that of ocinaplon at the α5β2γ2S receptor subtype. The efficacies of DOV 315,090 and ocinaplon at α2β2γ2S receptors were similar.These results do not support the hypothesis that the anxioselective profile of ocinaplon is attributable to enhanced selectivity of its metabolite DOV 315,090 for α2-containing receptors. Thus, compared to ocinaplon, DOV 315,090 does not exhibit enhanced affinity or potency for α2-containing receptors over α1-containing receptors, whereas the difference in efficacy favors α3-, α5-, or α1-containing receptors over α2-containing receptors. The present experiments examined GABAA-Rs in two different heterologous expression systems (Xenopus oocytes and HEK 293 cells), which may be lacking modulatory proteins or regulatory mechanisms that are only present in neurons. While we cannot exclude the possibility that such interactions somehow confer differences in modulator binding or efficacy, such a hypothesis would require that such interactions modify the structure of the benzodiazepine binding site, which is located in the extracellular domain of the GABAA-R, in such a way as to selectively alter its interactions with different ligands.Recent studies suggest that GABAA3-Rs receptors are also important in mediating anxiolysis [18,20,31-34]. DOV 315,090 has relatively high efficacy at α3β2γ2S, so it is likely that modulation of GABAA3-Rs by DOV 315,090 contributes to the anxioselective profile of ocinaplon; however, adipiplon (NG2-73), an α3-selective positive modulator, has been reported to have sedative/hypnotic activity [35], suggesting that α3 selectivity is not sufficient to confer anxioselectivity.In summary, transgenic mice in which the BZD recognition site of the α2 subunit is disabled exhibit reduced diazepam sensitivity in behavioral tests considered to be predictive of anxiolytic activity, and a similar modification to the α1 subunit reduces sensitivity in tests held to be predictive of sedation [15,26]. These observations have led to optimism that it will be possible to achieve the long-desired goal of developing a nonsedating anxiolytic [36]. And indeed, there has been substantial progress in identifying such compounds [19-22,31,37-40], yet ironically, they do not in general conform to the expected paradigm of favoring α2-containing over α1-containing receptors. This suggests that anxiolysis in humans may prove to be more complex than is suggested by a simple reading of the results from transgenic mice in behavioral models thought to be indicative of anxiety. It remains to be determined whether differences in the design of the behavioral assays [41,42], interspecies differences [43,44], or a combination of these factors account for these discrepancies. Translating such promising results into clinically useful compounds is likely to require an improved understanding of the ways in which BZD-like ligands act at different GABAA-R subtypes and the consequences of these effects upon neural system-mediated behavioral outputs.Conclusion1. DOV315090 is a major metabolite of the anxioselective GABAA-R modulator ocinaplon.2. DOV 315,090 possesses modulatory activity at α1-, α2-, α3-, and α5-containing GABAA-Rs with a selectivity profile similar to that of ocinaplon.3. The anxioselective properties of ocinaplon, demonstrated in both preclinical and clinical studies, are not a consequence of enhanced subtype selectivity by DOV315090.AbbreviationscDNA: complementary deoxyribonucleic acid; cRNA: complementary ribonucleic acid; DOV 51892: (7-(2-chloropyridin-4-yl)pyrazolo- [1,5-a]-pyrimidin-3-yl](pyridin-2-yl)methanone); ocinaplon, (2-pyridinyl [7-(4-pyridinyl)pyrazolo[1,5-a]pyrimidin-3-yl]methanone); DOV 315,090: (7-(1-Oxidopyridin-1-ium-4-yl)pyrazolo [1,5-a]pyrimidin-3-yl)(pyridin-2-yl)methanone, GABA, γ-aminobutyric acid; IGABA: GABA-gated current.Authors' contributionsDB carried out electrophysiological recordings. MCG carried out radioligand binding experiments. EK performed initial electrophysiological experiments. SD developed the data-acquisition hardware and software used in this study. TTG participated in the design of the study, performed the statistical analysis and participated in manuscript preparation. DHF participated in the design of the study and participated in manuscript preparation. PS directed development of ocinaplon at DOV Pharmaceuticals and participated in manuscript preparation. ASB identified major ocinaplon metabolite and participated in manuscript preparation. All authors read and approved the final manuscript.\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2529302.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2529302",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2529302\nAUTHORS: Kieran O'Sullivan, Norelee Kennedy, Emer O'Neill, Una Ni Mhainin\n\nABSTRACT:\nBackgroundLow-dye (LD) taping is commonly used to reduce rearfoot pronation. No studies have previously investigated the effectiveness of LD taping using both plantar pressure distribution (F-Scan) and 3-D (CODA) analysis of rearfoot motion.Methods20 healthy subjects with a navicular drop test exceeding 10 mm participated in the study. T tests were used to determine whether significant (p < 0.05) differences in plantar pressure and rearfoot motion occurred with LD taping.ResultsLD taping resulted in statistically significant increases in peak plantar pressure in the lateral midfoot (p = 0.000), along with significant decreases in pressure in the medial forefoot (p = 0.014), and the medial (p = 0.000) and lateral hindfoot (p = 0.007). No significant changes occurred in the medial midfoot (p = 0.794) or lateral forefoot (p = 0.654). When assessed using motion analysis, taping resulted in a statistically significant decrease in rearfoot pronation (p = 0.006), supination (p = 0.025) and total rearfoot range of motion (p = 0.000). The mean rearfoot position during stance was not significantly different however (p = 0.188).ConclusionLD taping is associated with alterations in peak plantar pressure in the midfoot and forefoot that indicate reduced pronation with LD taping. However, LD taping appears to reduce both pronation and supination in the rearfoot, rather than simply reducing pronation, when assessed using 3D motion analysis. Therefore, it would appear that LD taping does indeed reduce pronation, by restricting rearfoot motion in general, rather than pronation specifically. The degree of change observed with LD taping was however very small, and further research is needed to clarify the clinical significance of these initial findings.\n\nBODY:\nBackgroundPronation is a normal component of the stance phase of gait, however excessive pronation, when the rearfoot remains pronated beyond the midstance phase of gait [1], may cause excessive myofascial and soft tissue stress [2]. Low-dye (LD) taping is commonly used by physiotherapists in the treatment of lower limb symptoms related to altered or excessive pronation [3,4], or to help decide if orthotics may be indicated [5]. LD taping is suggested to limit foot pronation by raising the medial longitudinal arch and controlling the amount of rearfoot pronation occurring [6,7].The effectiveness of LD taping has been examined in many different ways, including static and dynamic measures. Static measures include assessing vertical navicular height (VNH) and the navicular drop test (ND) [6,8]. Using these measures, it appears that LD taping increases VNH and reduces ND in stance [8,9], implying a short-term reduction of pronation with LD taping. Dynamic analysis of the effect of anti-pronation taping on foot motion and alignment has been less commonly used, even though studies have questioned the actual validity of static measures in predicting dynamic foot function [1]. A previous study [7] used two-dimensional (2D) video analysis to measure the pronation angle of the foot with and without tape. Their results did not show any significant difference in dynamic pronation under each of the conditions. In contrast, other authors [10] who also used 2D video analysis, found the arch height ratio to increase, which indicated reduced pronation, in overpronated subjects after the application of LD taping whilst walking. There are no published trials examining the effect of LD taping using three-dimensional (3D) motion analysis. Previous 3D analysis of the effects of soft foot orthotics on overpronation however found significantly reduced overall rearfoot motion when the orthotics are used, and not just reduced pronation [11]. While there has been research into the effect of LD taping on rearfoot pronation, the effect of LD taping on rearfoot supination and overall rearfoot motion is unclear.Numerous studies have used plantar pressure patterns as an indirect measure of foot pronation during walking, as it has been assumed that plantar pressure distribution reflects rearfoot position [12]. The existing evidence suggests that LD taping reduces pronation, as indicated by shifts in midfoot pressure from medial to lateral, as well as changes in forefoot and hindfoot forces [3,12]. Many previous trials examining plantar pressure however did not individually calibrate the sensors for each individual, did not allow subjects wear their usual footwear, and may not have investigated truly normal gait due to subjects having to step onto a sensing platform [3,12].The purpose of this study was therefore to evaluate the immediate effect of LD taping, using 3D motion and plantar pressure distribution analysis. A population of healthy subjects with a navicular drop exceeding 10 mm were chosen to attempt to replicate the patient population who might receive LD taping.MethodsEthical approval was obtained from the University of Limerick Research Ethics committee. Participants gave written informed consent prior to participation.SubjectsA convenience sample of 28 healthy subjects volunteered to participate in this study. An initial screening session determined if subjects had excessive pronation, using the ND test, which is a commonly used method for measuring excessive pronation in healthy individuals, and which has good intra-rater reliability [8,13]. Excessive pronation was defined as navicular drop of > 10 mm, similar to previous research [8,13,14] and all subjects were screened by one investigator. Eight were excluded as they did not have a navicular drop of greater than 10 mm. Tape allergy testing was also performed at the initial screening, for which a piece of zinc oxide tape was applied to the right ankle, and left in situ for at least 24 hours. Subjects with an adverse skin reaction (redness, rash or discomfort) to tape, with a lower limb injury in the past six months, or who were unable to walk painfree were excluded.Study designA repeated measures crossover study design was used. Since the plantar pressure and 3D motion data could not be collected simultaneously due to the practical issues in using both pieces of equipment, the order of testing was structured to minimise the length of time required for testing. Therefore the sequence of testing was always as described in Table 1. This allowed each subject to be analysed using both systems separately with a requirement to be only taped once.Table 1Order of testing for both procedures, and taping condition of each.OrderTest procedureTaped1Plantar pressureNo2Plantar pressureYes33D motion analysisYes43D motion analysisNoTapingLD taping was applied only to the right foot of each subject [3]. A standard LD taping technique using rigid 3.8 cm wide zinc oxide tape (Leukotape) was used, similar to other trials [3,6,12], while palpating subtalar joint neutral position (Figure 1). Feet were washed and dried in advance of taping to optimise tape adherence [5]. To enhance consistency, the same investigator applied all taping and followed a standardised protocol.Figure 1Low-dye taping technique used in the study.Instrumentation: Plantar pressure DataThe F-Scan (Tekscan Inc), a computerised insole sensor system, was used to measure plantar pressure. The sensor consists of a bipedal, thin shoe insole composed of 960 individual pressure-sensing locations, providing a spatial resolution of four sensors/cm2. The insole uses resistance-based technology and the inner surfaces are printed with electrical circuits and in between these circuits is a semiconductive ink whose electrical resistance change inversely proportionally to the pressure applied. Studies have found that the F-Scan has fair to good reliability. Ahroni et al. [15] examined the reliability of the F-Scan in people with diabetes and found moderate ICC values of 0.755 and 0.751. Mueller and Strobe [16] examined the reliability of the F-Scan in ten normal subjects over multiple steps and reported a pearson product moment correlation coefficient of 0.933 between force platform data and F-Scan data. An experimental comparison of the Pedar system and the F-Scan by Hsiao et al. [17] also reported good reliability for both systems provided the limitations of using such measurement devices were identified and reduced where possible. The F-Scan insoles were measured for each individual's right shoe according to manufacturers guidelines. The sensor was then inserted into the subjects shoe and attached to the transducer device that is attached to a computer via a 9.25 m cable. Insole calibration was performed once for each subject as per manufacturers' guidelines. This calibration involved subjects initially walking > 20 steps to allow the insole adjust to conditions in the shoe. The insole was then loaded with total body weight for 1 second by lifting the left foot off the ground, simulating the magnitude and speed of stance phase loading during gait (Figure 2). The same insole was used for each individual for each of his or her walking trials. Because of natural step-to-step variability [18], data from several footfalls was gathered to obtain a representative profile of the subject's foot. Plantar pressure data was collected over 10 metres at a frequency of 50 Hz. Subjects were asked to walk at their normal speed, looking straight ahead. Standardised instructions were given to each subject by the same investigator. Prior to testing, subjects were allowed practice to become comfortable with the procedure. Post-taping, subjects walked around for 2–3 minutes to adapt to the tape. A rest interval between walking trials was offered to all subjects to minimise possible fatigue. Because velocity has been shown to affect plantar pressure values [19], the time taken to complete the walks was also recorded.Figure 2Calibration procedure for the F-scan plantar pressure system.Instrumentation: Motion analysisKinematic data was acquired using a CODA mpx64 (Charnwood Dynamics Ltd., Leicestershire, UK) motion analysis system. This system uses a laboratory-based coordinate system, and calculates joint angles based on skin marker positions without the need to define a 'zero' starting position for the rearfoot. The markers were applied by one investigator in line with both manufacturer guidelines and previous research [20]. Markers were positioned on the lateral aspect of the knee joint line in the median frontal plane, the anterior aspect of the lateral malleolus, the posterior inferior lateral aspect of the heel, and the lateral aspect of the fifth metatarsal head. The markers were fixed to the skin with double-sided adhesive tape. The order of testing required removal and immediate replacement of some of these markers when LD tape was being removed prior to analysis of the 'untaped' condition, however the same investigator did this over a very short time period. During testing subjects walked barefoot across a 10-metre walkway at a comfortable 'normal' walking speed. Subjects were instructed to look at a distant mark to prevent them from looking down at the floor. The subject performed 4 gait cycles with the tape and 4 cycles without the tape, since previous research has recommended the use of at least 3 gait cycles to aid reliability [21]. 3D motion data was collected at 200 Hz for 4 seconds while the subject was performing the walks, similar to previous research [20]. Blinding of the data collector regarding subject condition during the testing procedure was not possible.Data AnalysisPlantar pressure data from the entire stance phase (heel-strike to toe-off) was collected and analysed using Tekscan software. To avoid any acceleration and deceleration associated with the beginning and end of walks, the middle 3 stance phases of each 10 metre walk were analysed. The foot was divided into a grid with 6 distinct areas to display changes in plantar pressure distribution. The same grid was used for taped and untaped data of each subject, but to accommodate different sized feet, different grids had to be developed for each subject. Insole sensor cells occasionally developed \"shorts\" where they appeared to be loaded when they are not, and these were edited prior to analysis as per manufacturers' guidelines. Tekscan software calculated the average peak plantar pressure of the middle 3 stance phases in each of the 6 areas. Peak pressure was defined as the highest value recorded by a cluster of 4 cells over the entire stance phase [22,23]. For kinematic data, the stance phase of gait had to be identified in the absence of a force plate to demarcate stance and swing phases. Therefore heel strike was identified using the lowest vertical component of the heel marker and verified with the stick figure diagram [20]. Kinematic data was calculated and analysed by CODA software, before being extracted and entered into Microsoft Excel and averaged for all subjects. The kinematic data analysed included the following parameters at the subtalar joint during the stance phase of gait;• minimum displacement value, which indicated peak pronation.• maximum displacement value, which indicated peak supination.• total displacement which represented total subtalar joint ROM.• mean displacement value, which indicated mean joint position during stance.These kinematic values are as defined by the manufacturers and other researchers [20].Statistical AnalysisStatistical analysis was undertaken using SPSS 13.0 for Microsoft Windows (Chicago, IL). Data distribution was determined visually using histograms and using the Kolmogornov-Smirnov statistical test. Kinematic data, with the exception of minimum (pronation) values was normally distributed. Plantar pressure data, along with the pronation values from motion analysis, were non-normally distributed. Paired t-tests were carried out on normally distributed data to test for statistically significant differences between taped and untaped conditions. Wilcoxon-Signed Rank tests were carried out on non-normally distributed data to test for significant differences between taped and untaped conditions. The level of significance was set at p < 0.05. The standard error of measurement (SEM) was calculated in line with previous research [24].ResultsDemographic Data20 subjects (6 M, 14 F) met the inclusion criteria. Their mean (+/- SD) age was 22.1 (+/- 5) years.Plantar pressure dataLD taping resulted in statistically significant increases in peak plantar pressure in the lateral midfoot (p = 0.000), along with significant decreases in pressure in the medial forefoot (p = 0.014), and the medial (p = 0.000) and lateral hindfoot (p = 0.007) (Table 2). No significant changes occurred in the medial midfoot (p = 0.794) or lateral forefoot (p = 0.654) (Figure 3). The actual differences in peak plantar pressure values between taped and untaped conditions for all 6 areas of the foot are also detailed for each subject (see additional file 1).Table 2Mean (+/- SD) values for peak plantar pressure in taped and untaped conditions for each region of the foot; medial forefoot (MFF), lateral forefoot (LFF), medial midfoot (MMF), lateral midfoot (LMF), medial hindfoot (MHF) and lateral hindfoot (LHF).UntapedTapedMFF276.85 (+/- 79.66)241.5 (+/- 131.44)*LFF230.65 (+/- 105.43)227.75 (+/- 108.90)MMF57.2 (+/- 15.83)58.7 (+/- 23.17)LMF99.4 (+/- 52.79)149.35 (+/- 65.79)*MHF234.85 (+/- 88.20)192.05 (+/- 43.05)*LHF208 (+/- 63.73)180.2 (+/- 35.49)**indicates the difference was statistically significant (p < 0.05)Figure 3Peak plantar pressure values for taped and untaped conditions for each region of the foot; medial forefoot (MFF), lateral forefoot (LFF), medial midfoot (MMF), lateral midfoot (LMF), medial hindfoot (MHF) and lateral hindfoot (LHF). These differences were statistically significant for the lateral midfoot (p = 0.000), the medial forefoot (p = 0.014), and the medial (p = 0.000) and lateral (p = 0.007) hindfoot.Kinematic dataThe means and standard deviations for pronation, supination, total ROM and joint position under both taped and untaped conditions are displayed in table 3 and figure 4. There was a statistically significant reduction in both pronation (p = 0.006) and supination (p = 0.025) when LD taping was applied. As a result, there was also a significant reduction in total ROM after the application of LD tape (p = 0.000). However the mean rearfoot position was not significantly different between the test conditions (p = 0.188). The actual differences in kinematic values between taped and untaped conditions are also detailed for each subject (see additional file 2).Table 3Mean (+/- SD) values for pronation, supination, total range of motion (ROM) and mean joint position for taped and untaped conditions.TapedUntapedPronation #5.54 (+/- 4.27)4.15 (+/- 3.76)*Supination25.69(+/- 4.06)27.56 (+/- 4.30)*Total ROM20.15(+/- 3.64)23.41 (+/- 3.92)*Mean Position18.05(+/- 3.50)19.16 (+/- 3.48)*indicates the difference was statistically significant (p < 0.05). # lower values for pronation represent increased pronation, and not reduced pronation.Figure 4Kinematic values for pronation, supination, total range of motion (ROM) and mean joint position for taped and untaped conditions. These differences were statistically significant for pronation (p = 0.006), supination (p = 0.025) and total ROM (p = 0.000).Data reliabilityWe did not perform a test-retest reliability study, which significantly limits interpretation of the reliability of the data. Instead, we used the actual study data to calculate values for the SEM, to give an approximate representation of the reliability of the data. Data for plantar pressure could not be used to generate a value for SEM. Kinematic data from each of the four trials was however analysed to obtain values for the SEM of the CODA system (see additional file 3).DiscussionThe findings of this study suggest that LD taping results in reduced rearfoot motion, and changes in plantar pressure patterns, in a small sample of healthy subjects. In agreement with previous trials, LD taping resulted in an immediate short-term reduction in pronation [3,6,8-10,12]. This is the first trial that has shown this reduction in pronation to be present when measured by both plantar pressure and 3D motion analysis. Interestingly rearfoot motion in general, rather than simply pronation, appears to have been reduced with LD taping. This has not been reported previously, but is consistent with similar research demonstrating that addition of foot orthotics resulted in an overall reduction in rearfoot motion, rather than simply reduced pronation [11]. This suggests that it may be inappropriate to refer to LD taping as 'anti-pronation' taping, as its effects are not solely on pronation ROM. In a wider context, this is important because the technique is used by up to 73% of physiotherapists [25], and it is commonly described as an 'anti-pronation' taping technique, with less consideration of it's effect on supination ROM [26].Plantar pressureResults of the current study indicate that taping caused a significant increase in peak plantar pressure in the lateral midfoot, no change in the medial midfoot or lateral forefoot, and significant decreases in the hindfoot (medial and lateral) and the medial forefoot. Since peak plantar pressures are located more medially in excessively pronated feet [27], these results imply that there may be a trend towards reduced pronation in the midfoot and forefoot, but not in the hindfoot. The results are broadly in accordance with the results of previous similar studies [3,12]. Russo and Chipchase [3] found very similar results in the midfoot and forefoot, however they reported contradictory findings in the hindfoot, where peak pressure was increased after taping. Lange et al. [12] also agreed with the results of the current study, showing a significant increase in lateral midfoot pressure and a reduction in hindfoot pressure after LD taping. Vincenzino et al. [5] also demonstrated a significant reduction in hindfoot contact as well as a non-significant increase in lateral midfoot contact after 'augmented' LD taping, similar to the current study. They examined plantar contact area however, rather than plantar peak pressure. The slight inconsistencies between trials may be explained by differences in the pressure-sensor system used, as well as variations in the exact type of LD taping applied. Despite this, the changes observed in the current study are broadly consistent with those described in the literature.KinematicsMaximum pronation was found to decrease significantly (p < 0.05) as a result of LD taping. This finding is in agreement with results found in other studies [5,7-10,26]. The populations studied in these other trials, and the taping techniques used, were similar to those of the current study. Different outcome measures were however used in previous trials, with the majority being related to measures such as ND and VNH [9,10]. This is the first study examining the effect of LD taping on rearfoot motion using more complex 3D analysis, however the findings regarding reduced pronation are in line with previous studies. The findings of a reduction in supination are interesting in that they appear to indicate that LD taping results in a general decrease in mobility of the rearfoot, rather than having a purely 'anti-pronation' effect, as has typically been described in the literature [7,26]. This is further highlighted by the fact that the mean position of the rearfoot during stance did not change significantly between conditions. The observed reduction in overall rearfoot motion has also been described with the use of foot orthotics, albeit using different methods of motion analysis [11,28]. The effects of LD taping and foot orthotics may be similar, however this has not been proven and further research is needed to clarify if the effects seen here with LD taping also occur with foot orthotics. In addition, previous research [29] indicates that ankle taping reduces ankle joint motion in normal subjects. Although the taping technique and joint motion measured differs, their findings are in line with the current study.Mechanism of actionThe main proposed mechanism behind the clinical effectiveness of LD taping has been that it restricts rearfoot pronation [10], thereby reducing medial loading and increasing lateral loading through the foot [7,26]. The findings of this study agree only in part with this proposal. While pronation was reduced, LD taping did not result in increased supination, but rather reduced supination. The motion of the rearfoot as a whole was reduced, and the mean position through stance did not alter, with LD taping. The changes in plantar pressure imply a reduction in pronation, particularly during loading of the midfoot and forefoot. The plantar pressure data does not inform us sufficiently about supination range however. It may be that LD taping acts as a controller of general foot hypermobility rather than having a specific 'anti-pronation' effect. These hypotheses require further research before being proven however.Future researchThe evidence suggests that the effects of LD taping are short lived, although the exact length of time it may be effective for is still unclear [6-9]. This study was limited to the short-term effects of LD taping on non-injured subjects. Obviously further research is required to evaluate if these findings are replicated in a painful population, and how long these effects are maintained. Furthermore, research using foot orthotics suggests that when rearfoot motion is reduced significantly, significant changes may also occur more proximally at the knee joint [11]. Further research into the effects of LD taping on motion in other lower limb joints is warranted. Future use of both kinematic and plantar pressure data in studies examining the effects of LD taping may be warranted as the current study results imply that they inform us of related, but different, aspects of the technique.LimitationsThe main limitation relates to the fact that both 3D motion analysis and plantar pressure systems are known to be linked to variable data output [21,30]. The current study took steps to minimise this variation however, and the degree of variation is similar for both taped and untaped conditions. 3D motion analysis is a relatively new method of analysing the effect of LD taping on rearfoot motion. All surface marking systems carry a certain degree of error when estimating the motion of joints, however the CODA motion analysis system is sufficiently reliable if a number of gait cycles are used, similar to this study [20,21]. It is difficult to compare absolute values of plantar pressure systems across studies, and it is more appropriate to compare plantar pressure distributions under constant conditions, as in this study [31]. Secondly, a strict protocol was followed when using the F-Scan in order to make the procedure reliable. The F-scan system is highly correlated with force platform measures [16] and is sufficiently reliable [15], particularly when a mean of 3 steps is taken as the representative value [16], similar to recommendations for other pressure measurement systems [32]. Thirdly, other factors which could affect validity e.g. walking speed and surface contact [33], were consistent between taped and untaped conditions. The use of footwear was different for each measurement type, but once again this was consistent between taping conditions. Ideally, the measurement of 3D motion and plantar pressure would occur simultaneously to ensure the gait cycle analysed was identical, and the effect of taping could not have changed. The desire to examine in-shoe plantar pressures obviously would not allow visualisation of the skin markers. Therefore, simultaneous data collection was not possible and correlations between changes in kinematics and plantar pressure distribution were neither possible nor appropriate. This potential bias was minimised by gathering multiple cycles for each measurement system, in line with recommendations regarding a suitable number for adequate reliability for each system [16,21]. This resulted in a different number of gait cycles being performed for plantar pressure and motion analysis, however the number of gait cycles did not vary between the taped and untaped conditions. The absence of a force plate also meant the authors had to visually gauge where heel-strike and toe-off occur. This method has, however, been recommended by the manufacturers and been described in previous research [20]. The need to reposition motion analysis skin markers after the removal of LD tape requires that the kinematic results be interpreted with some caution, as there is a small risk that this could have resulted in slight changes in kinematic angles. Similar to some previous LD taping trials [5,7], the reliability of the investigators was not established in the current study, however this is a potential source of error. This is particularly important given the small magnitude of change between conditions and the high variability of the data. The SEM values for kinematic data exceeded the statistically significant difference observed between taped and untaped conditions. Therefore the data should be interpreted with caution, as some of the difference observed between groups could be due to simply measurement error. A clearer indication of the reliability of the study protocol would require a test-retest reliability study to be performed in advance. The sample size is however in line with previous LD taping trials [5,7,9,10,26]. This high level of data variability is commonly noted in studies of plantar pressure, LD taping and lower limb kinematics [5,15,16]. The effect of taping was examined only during the stance phase, due to the fact that maximum pronation has been found to occur during the middle-to-late stance phase of the gait cycle [1], and symptoms are usually related to weight bearing. The size of the change with LD taping was statistically significant, but we cannot say whether this would be clinically significant. We did not examine whether the taping was performed identically for each subject, however one person performed all taping to minimise error and the tape applied did not change between the two measurement techniques. The sample size was small, and a suitable power calculation was not performed due to the exploratory nature of the study, and this limits external validity. Subjects were not randomly selected, but were a sample of convenience. The amount of time subjects were given to become accustomed to the tape varied somewhat between 2 and 3 minutes, which is a potential source of error. Also, there is a very slight risk of a residual effect of taping even after its removal, which could potentially have affected the baseline 'untaped' kinematic data. Finally, neither subjects nor investigators were blinded to taping condition, as this was not feasible.ConclusionWhile this relatively small study does have some limitations, we believe, as it is the first study to combine 3D kinematic and plantar pressure measurement of the effects of LD taping, that its results are noteworthy. The study demonstrated that LD taping reduced both pronation and supination in the rearfoot during the stance phase of gait in healthy subjects with a ND exceeding 10 mm. LD taping also significantly altered the plantar pressure pattern of the foot. Clinically, this may support the use of LD taping in the treatment of symptoms related to increased foot mobility. Despite the description of LD taping as an 'anti-pronation' taping technique, it may work by limiting overall motion at the rearfoot. Further research is needed, particularly in clinical populations and examining the effects of foot orthotics. Further research is also required to establish the effect of reduced rearfoot ROM on other joints of the lower limb and its implications for injured subjects. It is important that future similar studies clarify whether the changes observed are greater than measurement error, which the current study was unable to do. Studies will also need to be conducted to establish the length of time that this effect of LD taping on the rearfoot lasts in a clinical population.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsKOS was involved in conception and design of the study, data analysis and interpretation, as well as drafting and editing the final document for publication. NK was involved in conception and design of the study, data analysis and interpretation, as well as drafting and editing the final document for publication. EON was involved in conception and design of the study, data collection, data analysis, as well as drafting and editing the final document for publication. UNM was involved in conception and design of the study, data collection, data analysis, as well as drafting and editing the final document for publication.FundingNonePre-publication historyThe pre-publication history for this paper can be accessed here:Supplementary MaterialAdditional file 1Peak plantar pressure data for all 20 subjects (averaged) in taped and untaped conditions, for each region of the foot; medial forefoot (MFF), lateral forefoot (LFF), medial midfoot (MMF), lateral midfoot (LMF), medial hindfoot (MHF) and lateral hindfoot (LHF).Click here for fileAdditional file 2Kinematic data for all 20 subjects (averaged) in taped and untaped conditions.Click here for fileAdditional file 3Estimated standard error of measurement (SEM) values (in degrees) for kinematic data, for both taped and untaped conditions.Click here for file\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2529374.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2529374",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2529374\nAUTHORS: Yaoxing Huang, Michael Krasnitz, Raul Rabadan, Daniela M. Witten, Yang Song, Arnold J. Levine, David D. Ho, Harlan Robins\n\nABSTRACT:\nThis manuscript describes a novel strategy to improve HIV DNA vaccine design. Employing a new information theory based bioinformatic algorithm, we identify a set of nucleotide motifs which are common in the coding region of HIV, but are under-represented in genes that are highly expressed in the human genome. We hypothesize that these motifs contribute to the poor protein expression of gag, pol, and env genes from the c-DNAs of HIV clinical isolates. Using this approach and beginning with a codon optimized consensus gag gene, we recode the nucleotide sequence so as to remove these motifs without modifying the amino acid sequence. Transfecting the recoded DNA sequence into a human kidney cell line results in doubling the gag protein expression level compared to the codon optimized version. We then turn both sequences into DNA vaccines and compare induced antibody response in a murine model. Our sequence, which has the motifs removed, induces a five-fold increase in gag antibody response compared to the codon optimized vaccine.\n\nBODY:\nIntroductionIn the effort to create a vaccine for human immunodeficiency virus type 1 (HIV-1), poor immune response to the HIV proteins is a fundamental problem. For a DNA vaccine, the immune response is correlated with protein expression levels, so an increase in expression of these proteins could alleviate a significant road block to the construction of a viable DNA vaccine.[1], [2] Transcription of the HIV DNA copy is an inefficient process that is aided by the addition of an HIV protein (TAT). In addition the large mRNAs from HIV have inherent RNA processing problems and are not efficiently exported from the nucleus in the absence of a helper protein (REV). These m-RNA synthesis and transport problems are presumably due to a set of RNA sequences or structures encoded in HIV DNA and RNAs.[3], [4] We hypothesize that identifying and removing these signals, which cause the poor synthesis and nuclear confinement of HIV RNA, should significantly increase expression levels of these proteins and improve the immune response.The genome of HIV-1 contains nine open reading frames (ORFs), all of which are expressed from a single promoter through alternative splicing. The splice forms for the six ORFs Gag, Pol, Env, Vpu, Vif, and Vpr, along with the full length mRNA, contain Rev response elements (RREs) encoded in their RNA. In the absence of the Rev protein, these six ORFs are poorly expressed. The remaining three ORFs, Tat, Rev, and Nef, are expressed efficiently independently of the Rev protein.[3]\nThe mRNAs which contain RREs likely also contain an as yet unidentified signal or set of signals which prevents normal expression.[4] A primary cause of the poor expression of these ORFs is nuclear confinement.[3] The genome of HIV-1 has an anomalous nucleotide distribution as compared with the set of known coding genes in the human genome. Only 314 of the approximately 25000 genes in the human genome have a higher percentage of adenine (A) than the average clinical isolate of HIV-1. Similar A content can be found in other retrotranscribing viruses (e.g. LINE elements, lentiviruses, spumaviruses); this suggests that retroviruses undergo different selection pressures than ones directing the evolution of the human genome. In the early 90's, the Pavlakis lab showed experimentally that synonymous changes to the Gag ORF, which decrease the A content, significantly increase expression of Gag in human cells.[4] Codon-optimized strains, which are widely used in present experiments and vaccine trials, can increase the protein expression level of Gag transfected into human cells between 500 and 1000 fold.[1], [2] However, the substantial increase in expression due to codon optimization can, at best, indirectly address the problem of poor synthesis and nuclear isolation. We identify multiple nucleotide motifs from a systematic comparison of the HIV-1 genome and the human genome, which we conjecture to play a causative role in poor synthesis and nuclear confinement.In this study, the short motif, AGG, is found to have the maximal differential representation between the coding genes in the human genome and the HIV-1 genome. This identification was made through the use of an information theoretic motif-finding method called the Robins-Krasnitz algorithm, described previously[5].The algorithm identifies dozens of motifs that exhibit substantial differences in representation between the HIV-1 genome and the human coding genes. The study presented here focuses on a single motif in order to isolate its contribution to expression level in a controlled experiment. A codon-optimized version of HIV-1 consensus gag is modified, making synonymous changes to reduce the number of occurrences of AGG. Two plasmids are constructed, one with the original codon-optimized (CO) sequence of gag (AD-gag) and the other with the motif-optimized (MO) sequence with AGG significantly reduced (RK-gag). The (DNA) constructs are transfected into a human epithelial cell line (293 cells) and expression of Gag is shown to be 70% higher for the MO sequence. The two sequences of gag are also tested as DNA vaccines in a murine model for differential immune responses between the two constructs. The mice with the MO version of the vaccine have a 4.5-fold greater anti-Gag antibody response after 4 weeks. With a DNA boost at four weeks and a second readout of anti-Gag antibody titers measured at six weeks, the gap continues to widen between the MO and CO vaccines to 6-fold.ResultsFinding the signal and recoding GagThe Robins-Krasnitz algorithm finds short nucleotide motifs in coding regions of the human genome that are independent of amino acid order and codon usage.[5] Codon usage is defined as the distribution of synonymous codons present in a given gene. The result of the Robins-Krasnitz algorithm is a set of exact nucleotide motifs of length 2–7 bases which are under and over represented in the coding regions of the human genome. The frequency of these motifs in the HIV genome can then be assessed. See Methods for details.Beginning with the set of the 100 most under- and over-represented motifs in the human genome, our study attempts to identify the motif with the largest density difference between the HIV genome and the human genome, after accounting for A content. The motifs are restricted to the set of human genes with A content within 1% of the average HIV A content. The ratios of the densities in the HIV genome are then divided by densities in the human coding regions. If the human density is greater than that of HIV, the quantity is replaced by its reciprocal. We predict that the motif with the largest ratio of densities is responsible for nuclear isolation of HIV mRNAs.The triplet AGG, which is significantly under represented in the coding region of the human genome, is found with a high frequency in HIV when the nucleotide bias of HIV is taken into consideration. We hypothesize that recoding the ORFs of HIV by reducing the frequency of the motif AGG will increase protein expression.For this initial study, our experimental tests focused on the Gag gene. The codon-optimized sequence of gag, referred to as AD-gag, is recoded by systematically removing all AGGs such that the amino acid sequence is not modified and very rare codons are not introduced. The result is the motif optimized RK-gag. Both the AD-gag and RK-gag sequences are found in the supplementary materials.Testing expressionFirst, we determine whether our RK-gag has increased expression as compared to the codon optimized version, AD-gag. Since our version of Gag is undoing part of the codon optimization present in the AD-gag sequence, the protein expression levels should be expected to decrease unless the motif AGG significantly inhibits mRNA synthesis or processing or transport. To compare expression levels, human 293 cells were transfected in vitro with one of the two different versions of Gag (see methods for details). Gag protein expression was measured in the extracts of transfected cells by a quantitative P24 ELISA. RK-gag was 70% higher than the codon optimized AD-gag, with a p-value <0.00001 calculated by a permutation test (see Figure 1).10.1371/journal.pone.0003214.g001Figure 1Gag expression in transiently transfected 293 cells.\nFigure 1 presents the results from four independent transfection experiments. The results are expressed as the mean P24 value (ng/ml, ±SD) of triplicates. The two different gag sequences are the codon optimized version (AD) and the motif optimized version (RK) that we created. Our (RK) version of the Gag gene has approximately two-fold higher expression than the codon optimized version.Humoral immune responseTo test the effect of the almost two-fold gain in expression on the immune response, we created DNA vaccines from each of the sequences. These DNA vaccines were injected into the hind leg muscle of Balb/C mice. The mice were given a booster shot after four weeks. Anti-Gag antibody titers were measured by anti-P24 ELISA at the four week and six week time points (see Methods for details). The results are found in Figure 2. The 70% increase in the expression of the GAG protein in vitro translated into more than a five-fold difference in humoral immune response in a mouse model.10.1371/journal.pone.0003214.g002Figure 2Immunogenicity of Gag DNA vaccines in mouse.The two different versions of Gag were made into DNA vaccines and injected into Balb/C mice with 25 µg/dose, then boosted at four weeks. Anti-Gag antibody levels were measured by ELISA at the four week and six week time points. The results are expressed as the geometric mean antibody titers (±SD) of each group. RK-Gag, induced an immune response that was five times larger than the codon optimized version at four weeks, which increased to a factor of 6 difference after six weeks.Two weeks post-boost, mice were sacrificed and splenocytes were prepared for measuring Gag-specific cell-mediated immune (CMI) responses by an IFN-γ ELISpot assay. Although the difference between these two groups was not statistically significant, there was a trend for RK-gag immunized mice to have stronger CMI responses to both Gag-specific CD4 and CD8 peptides tested (data not shown).DiscussionRecoding the Gag gene in order to reduce the occurrences of a single triplet, AGG, substantially improves immune response to an HIV DNA vaccine in a mouse model. This short sequence motif occurs less frequently in the human coding sequence than in the mouse coding sequence by about twenty percent, so it is possible that these results would be even more dramatic in humans. A set of additional steps would be required to move in the direction of a clinically viable vaccine. These include a recoding of the ENV ORF and a test of its ability to induce an improved immune response. Testing these concepts in primates would be a useful step. The goal of this study was to provide convincing evidence that recoding the HIV ORFs can improve HIV protein expression and the immune response compared to the present codon optimization schemes. It likely that including the other motifs identified by the Robins-Krasnitz algorithm in a systematic way has the potential to improve upon the large gains displayed in this study.Finally this study brings up the question of why the HIV DNA sequence has been selected to express poorly in primate cells, only increasing its levels with the aid of additional proteins that recognize nucleic acid sequences in the genome. Several other retroviruses and retrotransposons have similar sequence complexities. This may reflect an optimal way to regulate these viruses and enhance the viral titers over an extended length of infection. In any event it is becoming clear that nucleotide sequence motifs, in addition to the choice of codons, can have dramatic impacts upon gene expression, RNA processing and transport in a cell.MethodsRobins-Krasnitz algorithmThe first step in the algorithm is the creation of a background sequence to compare with the human genome. This background is a completely randomized version of the coding sequences from the human genome subject to the constraints of amino acid order and codon usage in each gene. We design a Monte Carlo program that randomly permutes the codons for each amino acid within each gene. Figure 3 is an illustrative example.10.1371/journal.pone.0003214.g003Figure 3Example of shuffling procedure.The procedure to get the maximal entropy distribution (MED) involves a set of randomized iterations. The triplets of nucleotides coding for each amino acid are permuted randomly among themselves. This is an illustrative example of a mock short protein with eight amino acids. The shuffling procedure randomly permutes L1, L2, L3, and L4 and separately permutes H1, H2, and H3. Each iteration produces a new sequence. For this example, there are 12 different combinations for the leucines and three combinations for the histidines resulting in 36 unique sequences. They are weighted in the shuffling procedure so that the MED is attained in the limit of a large number of iterations.The shuffling procedure described above yields a set of randomized sequences. From these sequences, we need to extract a probability distribution. As long as the number of occurrences of each motif found in the total set of sequences is reasonably large, we can form a probability distribution, estimating the probability of a given motif by its fraction in the set of all motifs.After the shuffling procedure we can define two distributions, the real distribution found from the actual sequence and the Maximal Entropy Distribution (MED) which we use as the surrogate for the background. We now need a method for choosing under and over-represented motifs. The standard we used is from information theory. The motif that contributes the most bits of information to the difference between the real distribution and the MED is the first motif we chose. Using information theory has the nice feature of putting all results in the same units, number of bits. This allows us to compare motifs of different lengths and motifs that are either over or under-represented. The formula we employ to compute the motif contributing the most bits of information between the two distributions is the Kullback-Leibler distance or the Relative Entropy. We compute the Relative Entropy contribution for each motif and pick out the one with the largest value.Once we have found the most under- or over-represented motif in the sequence, we have to pick out the motif which is the next most under- or over-represented. However, we cannot simply take the motif which has the next largest Relative Entropy. This is because the motifs are overlapping, so under or over representation of a given motif affects the distribution of all the other motifs. The example of CpG illustrates this point. In the human genome, the dinucleotide motif CG will have the largest Relative Entropy. However, all eight trimers which contain CG fall within the top 50 highest Relative Entropy motifs. This is simply an artifact of the selection against CG. We are required to first remove the contribution of CG from the MED before recalculating the Relative Entropy to find the next motif. If we call the motif w, we rescale all motifs that contain w by the same amount so that the rescaled MED had the same distribution for w as the real distribution. This forces the Relative Entropy of w to zero and, at the same time, removes the contribution of w from all other motifs. We can readily show that this choice of rescaling monotonically decreases the overall Relative Entropy between the distributions.The procedure is iterated, so that we remove the contribution of one motif at a time from the Relative Entropy through rescaling of the MED. Then, we choose the next motif. We continue to iterate the procedure, and find additional motifs, until the motif with the largest remaining Relative Entropy is not statistically significant, as determined by comparing shuffled genomes.Experimental protocolsHIV-1 subtype B gag consensus sequence was obtained from the Los Alamos HIV database (www.hiv.lanl.gov). The complete sequence of parental consensus gag was codon optimized to reflect the codon characteristics of eukaryotic expression systems (AD-gag) and assembled in house using overlapping PCR.[4], [6] RK-gag was synthesized by BlueHeron Biotechnology (www.blueheronbio.com). Both constructs have an identical “Kozak signal” located immediately upstream of the initial ATG and were cloned into NotI and XbaI cloning sites of pVAX1 (Invitrogen).Plasmid DNAs were prepared by GenElute Endotoxin-free plasmid purification system (Sigma). For valuation of gag expression in vitro, multiple batches of plasmid DNA were prepared to ensure the reproducibility of each of the independent transfection experiments. Briefly, 0.5 and 1 µg of DNA were transfected into 293 cells using the Lipofectamine reagent in a 24-well plate format according to the manufacturer's specification (Invitrogen). Cell culture supernatants were collected at 48 or 72 hours post transfection. Gag expression was measured by a commercial ELISA kit that detects and quantifies P24 in supernatant (PerkinElmer).For assessing immunogenicity in mouse model, DNA was eluted into saline at the concentration of 0.5 µg/µl. Independent batches of DNA were prepared for immunizations. Six to eight week old female BALB/c mice (Charles River Laboratories) were housed and treated at the Laboratory Animal Research Center of The Rockefeller University in accordance with Institutional Animal Care and Use Committee guidelines. Groups of mice (4 to 5 per group) were immunized with a 25 µg of plasmid DNA vaccine in 50 µl of saline at week 0 and week 4. Serum samples were collected from individual mice at week 4 (4 weeks post first vaccination) and week 6 (2 weeks post second vaccination). Direct ELISA was used to measure serum anti-Gag antibody titers from immunized mice. Briefly, 96-well plates coated with 0.25 µg recombinant P24 protein overnight were blocked for 2 hours with PBS-T containing 5% dry milk and 0.5% BSA. Individual mouse serum samples were added in serial dilutions and incubated for 2 hours. The plates were washed five times with PBS-T and incubated for one hour with AKP-conjugated rat anti-mouse secondary antibodies. The plates were then washed six times with AMPAK washing solution, developed with AMPAK kit (DAKO Corporation). The plates were read on an ELISA reader at 490 nm. The end-point antibody titers were calculated as the reciprocal dilution of the last dilution that was at least 2-fold higher than normal mice sera controls and yields an absorbance of >0.1.\n\nREFERENCES:\n1. LiuMAWahrenBKarlsson HedestamGB\n2006\nDNA vaccines: recent developments and future possibilities.\nHum Gene Ther\n17\n1051\n1061\n17032152\n2. LaddyDJWeinerDB\n2006\nFrom plasmids to protection: a review of DNA vaccines against infectious diseases.\nInt Rev Immunol\n25\n99\n123\n16818367\n3. CullenBR\n2003\nNuclear mRNA export: insights from virology.\nTrends Biochem Sci\n28\n419\n424\n12932730\n4. SchneiderRCampbellMNasioulasGFelberBKPavlakisGN\n1997\nInactivation of the human immunodeficiency virus type 1 inhibitory elements allows Rev-independent expression of Gag and Gag/protease and particle formation.\nJ Virol\n71\n4892\n4903\n9188551\n5. RobinsHKrasnitzMBarakHLevineAJ\n2005\nA relative-entropy algorithm for genomic fingerprinting captures host-phage similarities.\nJ Bacteriol\n187\n8370\n8374\n16321941\n6. KotsopoulouEKimVNKingsmanAJKingsmanSMMitrophanousKA\n2000\nA Rev-independent human immunodeficiency virus type 1 (HIV-1)-based vector that exploits a codon-optimized HIV-1 gag-pol gene.\nJ Virol\n74\n4839\n4852\n10775623"
+}

batch_3/PMC2530858.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2530858",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2530858\nAUTHORS: Deyou Zheng\n\nABSTRACT:\nA systematic analysis of histone modifications between human segmental duplications shows that two seemingly identical genomic copies have distinct epigenomic properties.\n\nBODY:\nBackgroundIt is widely recognized that gene duplications, by providing DNA material for evolutionary innovations, have contributed significantly to the complexity of primate genomes. Characterization of the human genome has highlighted the prevalence of segmental duplications (SDs), defined as continuous blocks of DNA that map to two or more genomic locations [1,2]. Previous studies have identified 25,000-30,000 pairs of SD regions (≥90% sequence identity, ≥1 kb), which occupy 5-6% of the human genome and arise primarily from duplication events that occurred after the divergence of the New World and Old World monkeys [2,3]. Detailed characterization of these SDs indicates that several molecular mechanisms might have been involved in the origin and propagation of SDs; in particular, repetitive sequences (for example, Alu elements) seem to have a major role in many segmental duplications [2].While the contribution of SDs to the architectural complexity of the human genome has been appreciated, the functional and evolutionary consequences of these duplications remain poorly understood. Although studies have begun to define the important roles of SDs in generating novel genes through adaptive evolution, gene fusion or exon exaptation [2,4,5], it remains a mystery how duplicated copies have evolved from an initial state of complete redundancy (immediately after duplications) to a stable state where both copies are maintained by natural selection. On the other hand, recent investigations of duplicated protein coding genes or gene families have provided a glimpse into this important evolutionary process. Those studies have shown that duplicated genes can evolve different expression patterns, leading to increased diversity and complexity of gene regulation, which in turn can facilitate an organism's adaptation to environmental change [6-9]. For example, the expression of yeast duplicated genes appears to have evolved asymmetrically, with one copy changing its expression more rapidly than the other [6].Initiating from these intriguing observations, the current study explores whether the sequence pairs of SDs are subject to different types and levels of molecular regulation, in particular whether the derived sequences are 'less' functional and are more likely to degenerate. As the majority of SDs are not protein coding, whole genome data unbiased towards genic regions is required to address these questions. Furthermore, such data must have sufficiently high resolution but minimal artifacts, which can often be attributed to high sequence similarity (such as cross-hybridization in microarray analysis), in order to reliably identify distinct signals belonging to each of the two individuals in an SD.The human genome is organized into arrays of nucleosomes composed of different histone proteins and higher order chromatin structures. Complex profiles of post-translational modifications (for example, acetylation and methylation) of histone proteins are implicated in regulating gene expression and many other important DNA-based biological functions [10-12]. For example, acetylation and H3K4 methylation are often implicated in gene activation while H3K27 methylation and H3K9 methylation are associated with gene repression. As histone modifications can be viewed, to a great extent, as a characteristic of functional chromatin domains, it will be interesting to know how histone modifications between copies of SDs are different. Furthermore, such a study may shed light on the evolution of SDs since histone modifications can modulate the accessibility of SD regions for DNA transcription, replication, and repair [10,13].This study systematically examined histone modifications in the human SD regions. Using data from a recent chromatin immunoprecipitation and direct sequencing (ChIP-Seq) study [14], the current analysis reveals for the first time that a divergent pattern of modifications exists between the two loci in a pair of SDs, when all SDs are considered collectively. The modifications with an asymmetrical pattern include the methylation of H3K9, H3K27, H3K36, and H3K79. This discovery is very interesting because these modifications have been implicated in a wide range of epigenetic-mediated events, including gene activation, gene repression, and heterochromatin formation [10,14]. Moreover, characterization of SDs emerging after the split of the human and macaque lineages found that the parental copies generally exhibit a higher level of modifications than the derived ones. Intriguingly, parental regions have a greater degree of H3K27me1 and H3K9me1 modifications, but not di- or tri-methylations. Furthermore, the parental loci also differ from the derived loci with respect to gene density, pseudogene density, and the abundance of RNA polymerase II (pol II) association. In short, this study demonstrates that the parental and derived copies of SDs are not functionally identical even though they share ≥90% identity in their primary sequences, suggesting that the descendants in a new genomic environment are more likely the candidates for sequence degeneration or functional innovation.ResultsHistone modification data in segmental duplicationsThe segmental duplications in the human genome were downloaded from the UCSC browser [15,16]. They include 25,914 non-redundant pairs of genomic regions (referred to as SD pairs here) in the released version (hg18) used for this study. The identification of these SDs has been described before [1] and the two sequences in each SD pair have a length of ≥1 kb and share ≥90% sequence identity.Histone modification data were primarily obtained from a recent ChIP-Seq study, which mapped the genome-wide distributions of 20 histone lysine (K) or arginine (R) methylations, as well as H2A.Z, pol II and CTCF (an insulator binding protein) across the human genome [14]. These data are summarized in Table 1, which shows a good number of ChIP-Seq tags (25 nucleotide sequencing reads) from human SDs. Since only tags that can be mapped uniquely to individual SD loci were used, the data in Table 1 indicate that ChIP-Seq can resolve signals from each of the two duplicates in an SD pair. The numbers of tags in SDs, however, decrease as the pairwise similarity within individual SD pairs increases (data not shown). Another set of histone modification data generated by ChIP coupled with paired-end ditags sequencing [17] was also obtained for this study (Table 1). From these two sets of ChIP data, a value measuring the level of a particular nucleosome modification in an SD was derived using a straightforward strategy (Figure 1).Table 1Summary of source dataData typeTotal data points for the human genomePoints within SDsH2AZ7,536,100152,848H2BK5me18,942,880184,251H3K27me110,047,279196,347H3K27me29,070,882180,054H3K27me38,970,141176,060H3K36me18,077,127164,151H3K36me313,572,575313,579H3K4me111,322,526213,535H3K4me25,447,902100,330H3K4me316,845,478361,316H3K79me110,041,806213,775H3K79me22,058,06840,023H3K79me38,114,474240,709H3K9me19,311,627170,633H3K9me29,782,127188,748H3K9me36,348,997147,639H3R2me19,560,224208,646H3R2me26,521,560147,126H4K20me111,015,873205,009H4K20me35,720,089370,598H4R3me27,357,597173,684Pol II4,150,37885,849CTCF2,947,04365,080H3K4me3, ES478,21337,413H3K27me3, ES257,57424,480RefSeq genes18,9573,366Duplicated pseudogenes2,5501,276Processed pseudogenes8,2342,786Other pseudogenes6,8092,412In the analyses of histone modifications and transcription factor binding, a data point is a read (that is, tag) from ChIP sequencing. The third column lists the numbers of ChIP tags (or genes, or pseudogenes) within the human SDs.Figure 1Histone modification ChIP tags in human SDs. A pair of SDs with 91.7% sequence identity was found in chr1:54,212,891-54,214,303 (top) and chr4:83,268,767-83,270,192 (bottom). The top region contained six H3K27me3 and two H3K4me3 ChIP-Seq tags, while the bottom contained two H3K27me3 and seven H3K4me3 tags. Thus, the number of H3K27me3 and H3K4me3 tags per 1 kb are 4.25 and 1.42, respectively, for the top and 1.4 and 4.91 for the bottom region.Asymmetric profiles of histone modifications in the two regions of segmental duplicationTo assess whether two copies of an SD pair exhibit different levels of histone modifications, this study first conducted a paired t-test with the null hypothesis that there is no difference. The Wilcoxon signed rank test was also performed to address a concern that ChIP tag differences between the two loci in SD pairs might not distribute normally. The two statistical tests yielded similar results and, therefore, only t-test data are discussed. After adjusting multiple testing by the Bonferroni method, 7 of the 20 histone marks showed a difference (adjusted p < 0.001; Table 2, all SDs), which include H3K9me2, H3K36me1, H3K79me1, H3R2me1 and the three states of H3K27 methylation. The original ChIP-Seq study also probed the bindings of CTCF and pol II, but the tags for them were distributed between the two loci of SDs without a bias. Similar analysis of the data from human stem cells [17] further indicated that histone modifications are asymmetric between the two copies of SDs (Table 2).Table 2Statistics for ChIP tag differences in the two copies of human SDsAll SDs (n = 25,914)Post-macaque SDs (n = 1,646)FactorsPaired t-test p-valuesWilcoxon signed rank test p-valuesMean of parentalStandard deviation of parentalMean of derivativeStandard deviation of derivativePaired t-test p-valuesMean of differenceWilcoxon signed rank test p-valuesH2AZ3.64E-052.86E-071.3192.3881.1141.9875.51E-030.2051.58E-03H2BK5me12.92E-012.32E-022.6006.5821.2243.2371.49E-151.3772.20E-16H3K27me12.12E-055.92E-032.1472.4151.2501.7862.20E-160.8972.20E-16H3K27me29.71E-101.74E-111.5261.6701.3551.6685.60E-040.1705.08E-05H3K27me32.20E-164.60E-141.4921.7271.4602.0036.09E-010.0312.95E-01H3K36me11.48E-052.28E-101.5331.3921.2421.6388.66E-100.2912.20E-16H3K36me31.29E-021.57E-053.7556.3471.7963.0272.20E-161.9592.20E-16H3K4me13.63E-018.01E-032.7006.8951.1392.7412.20E-161.5622.20E-16H3K4me28.76E-019.66E-061.3542.2900.6511.4582.20E-160.7032.20E-16H3K4me36.68E-017.46E-114.14416.4731.9876.2567.06E-072.1574.44E-16H3K79me16.49E-122.20E-161.9111.6441.4841.7812.20E-160.4272.20E-16H3K79me23.12E-021.26E-020.4760.5200.3560.4991.94E-080.1203.45E-11H3K79me39.62E-042.06E-061.8232.5951.6712.9415.76E-020.1533.76E-05H3K9me12.42E-026.41E-042.2623.8271.0462.1712.20E-161.2152.20E-16H3K9me21.67E-073.73E-141.6181.8651.4891.9361.88E-020.1305.80E-03H3K9me32.40E-033.90E-061.3802.3231.3572.3577.25E-010.0233.61E-01H3R2me12.19E-054.41E-091.8781.7511.4401.8662.20E-160.4382.20E-16H3R2me24.60E-014.41E-011.2921.2631.0791.5001.40E-060.2131.18E-11H4K20me15.72E-046.77E-054.86518.4761.2575.6649.85E-133.6082.20E-16H4K20me33.23E-015.78E-011.6876.4081.9535.6771.26E-01-0.2662.03E-01H4R3me22.16E-015.43E-011.4391.4041.1651.6661.91E-100.2756.66E-16Pol II4.24E-012.45E-041.5384.4760.5070.8126.47E-161.0312.20E-16CTCF9.31E-026.45E-050.7571.5600.5211.2261.41E-060.2362.20E-16H3K4me3, ES0.00080.2945.6739.201.4234.892.30E-064.251.91E-10H3K27me3, ES0.00247.63E-062.2033.2551.9583.3140.710.2450.534The p-values are before adjustment for multiple testing; statistically significant results (by t-test) are in bold.Higher level of histone modifications in the parental versus derivative loci of segmental duplicationsNext, I investigated whether the asymmetry is due to uneven histone modifications between the parental and the derivative regions. Although it has been previously found that two duplicated genes can evolve distinct functions, no systematic study to date has addressed which copy diverges away from its ancestral function. Unfortunately, current SD data do not contain the directionality of duplications, and accurate identification of duplication direction remains a challenge. This study thus adopted a strategy that was recently applied to identify ancestral duplication loci [18]. As illustrated in Figure 2, this approach relies largely on chromosomal synteny (that is, order of sequences on a chromosome) and uses macaque as an outgroup species to assign duplication directions for SDs. It produced more accurate parental-derivative relationships than other methods that were based entirely on mutual best hits established by sequence comparison, because a synteny-based strategy is more appropriate for identifying evolutionarily equivalent sequences in mammalian genomes. Macaque was chosen here because its genome has been sequenced and the average human-macaque sequence identity is approximately 93% [19], which is near the 90% used in identifying SDs. The current approach is not meant to systematically assign SD directions but to select SDs for subsequent analyses, because it can be applied only to SDs that arose after the split of human and macaque lineages. Nevertheless, it was able to determine the parental-derivative relationship for 1,646 SD pairs, referred to here as post-macaque SD pairs.Figure 2A cartoon illustrating the method used here for identifying post-macaque SDs based on chromosomal synteny. Using the liftOver tool [29] from the UCSC genome browser group, a pair of human SDs (A and B) is mapped to the same location (A') in the macaque genome. A and B (large block) are thus considered the product of an SD event that occurred after the split of human from macaque lineages. Then 1 kb sequences (small block) adjacent to A or B were aligned to the macaque genome. If only the sequence next to A was mapped next to A', then A is designated as the parental copy and B as the derivative.A paired t-test for these 1,646 pairs of post-macaque SDs revealed that 14 histone modifications are different between parental sequences and their derivative copies, including H3K36me1, H3K79me1, H3R2me1 and H3K27me1, which also showed asymmetries in the above analysis of all SDs (Table 2). In particular, histones in the parental loci exhibited a higher level of mono-methylation of H3K27 and H3K9 than those in the derivative regions (Table 2), but no difference was detected for di- and tri-methylations. Data from stem cells further supported a difference in H3K4me3 but no difference in H3K27me3. Interestingly, pol II and CTCF were relatively abundant in the parental versus the derivative loci. Noticeably, the analysis of post-macaque SDs yielded a list of histone marks that is quite different from what was obtained for all SDs (Table 2), suggesting that duplication direction is an important factor to include in examining disparate features of duplicated genes.The distribution of ChIP-Seq tags was further examined for human segmental duplications with known duplication directions. Previously, Eichler's research group have determined the duplication directions of nine human SDs by comparative fluorescent in situ hybridization (FISH), using genomic sequences in a human derivative locus as a probe against chromosomes from an outgroup primate species [18]. Four of those nine pairs are depicted in Figure 3. Analysis of ChIP-Seq data found that the levels of histone modifications were in fact quite biased between the two loci of most of these SD pairs. Especially, the parental regions were statistically higher for the following methylations: H2BK5me1, H3K4me2, H3K9me1, H3K27me1, H3K36me3, and H3K79me1. Mono-methylation seems to make up the bulk of the differences. Figure 4 shows the distributions of ChIP-Seq tags for four of these nine SDs.Figure 3Gene and pseudogene annotations in four pairs of human SDs with known duplication directions. The parental locus of each pair is depicted first, followed immediately by its derivative.Figure 4Pattern of histone modifications for the four SD pairs in Figure 3, ordered left to right to match their order from top to bottom in Figure 3. Each point represents the number of ChIP-Seq tags in a 5 kb genomic region, with red for parental and blue for derivative SDs. Horizontal axes are the position relative to the 5' end of a parental locus. Data for a derivative region is ordered with respect to its parent.The paired t-test described above, in principal, compared the sums of ChIP tags in the two copies of an SD pair, but overlooked the intra-SD tag distributions. Thus, a non-statistical method was developed to address this through analyzing ChIP tags in a set of large SDs (>15 kb). Briefly, these SDs were first divided into non-overlapping blocks. Then, for each pair of SDs, one locus was determined to have a higher level of a histone modification if at least two-thirds of its blocks contained more tags of this modification than the corresponding blocks of the other locus. The results not only show that SD loci with a greater degree of modification were three to six times more likely to be parental (Table 3), but also indicated that asymmetry often existed across an SD locus, rather than in one or few narrow sub-regions. Interestingly, all modifications exhibited some degree of asymmetry by this measurement. The second and third examples in Figures 3 and 4 illustrate such a pattern of asymmetrical modifications of histones.Table 3Numbers of large (>15 kb) post-macaque SDs with higher histone modifications in either parental or derivative lociFactorsHigher in parental lociHigher in derivative lociH2AZ6823H2BK5me19215H3K27me19617H3K27me28519H3K27me38529H3K36me19714H3K36me39023H3K4me18314H3K4me26712H3K4me39315H3K79me18719H3K79me2379H3K79me38223H3K9me18416H3K9me28324H3K9me37315H3R2me110319H3R2me29318H4K20me18114H4K20me37227H4R3me28818Pol II5715CTCF5113More parental loci of segmental duplications exhibit 'peak' signals of histone modifications'Peaks' of histone modifications in these large SD pairs were also studied. In agreement with the above observations, the peaks of ChIP-Seq signals were more frequently located within the parental SDs than the derivative SDs, especially for the three marks H3K4me3, H3K9me1, and H2A.Z, which have been previously shown to be enriched in promoters [14]. Data for H3K4me3, H3K27me3, and H3K36me3 are shown in Figure 5 because these methylations are known characteristic marks of promoters and transcribed regions, with H3K4me3 correlating with active genes and H3K27me3 relatively enriched at silent promoters [10,12,14,20]. As shown (Figure 5), SDs with an H3K4me3 peak were 1.5 times more likely to be parental. Such a bias, however, was not detected for H3K27me3. Only approximately 50% of either parental or derivative SDs with H3K4me3 peaks contained genes, suggesting that more functional elements (including novel protein coding and non-coding genes) are yet to be annotated in the human SDs. Interestingly, 9 of the 16 parental SDs versus 4 of the 16 derivative SDs with H3K27me3 peaks contained annotated genes, but these numbers were not statistically significant enough to claim that fewer genes in the derived SDs were repressed in CD4+ T cells. Parental SDs appeared more likely to have H3K36me3 and pol II peaks; however, those peaks did not seem to co-exist in the same SDs as frequently as expected from the correlation previously reported between H3K36me3 and actively transcribed regions [14,20]. This inconsistency needs to be studied in the future. Additionally, it needs to be mentioned that the known correlations between histone methylations and transcription start sites (TSSs) [14] were observed for the TSSs within SDs, and the patterns for parental SDs and derivative SDs were mostly indistinguishable (data not shown).Figure 5The peaks of ChIP-Seq signals in large post-macaque SDs. The numbers of peaks (see Materials and methods) for H3K4me3, H3K27me3, H3K36me3, and pol II are plotted for each of the large SD pairs (from top to bottom), along with the numbers of genes and pseudogenes. The numbers on the left (red) and right (blue) are for parental and derivative SDs, respectively. The H3K4me3 peaks in the first and forth example of Figure 4 are marked by an arrow and labeled with 1 and 4, respectively.In summary, characterization of the pattern of histone modifications by various measurements consistently revealed an asymmetrical pattern of histone modifications, with higher levels biased to the parental regions of SDs, demonstrating that two seemingly 'identical' genomic copies are actually distinct in their epigenomic properties.Parental loci of segmental duplications contain more genes but fewer pseudogenesIt has been reported that SDs are generally enriched with genes [2,3]. This is confirmed by the current survey of genes and pseudogenes in human SDs (Table 1); note that SDs occupy approximately 5% of the human genome. Moreover, Table 1 shows that human SDs are more enriched with pseudogenes than genes, as 36.8% of human pseudogenes and 17.8% of human genes are located in SDs (p << 0.001). Duplicated pseudogenes appear more likely to be associated with SDs than processed pseudogenes, as 50% of human duplicated pseudogenes versus 33.8% of processed pseudogenes are in SDs (p << 0.001). This is consistent with the fact that duplicated pseudogenes are generated by gene duplications whereas processed pseudogenes are from retrotranspositions.A subsequent examination of genes and pseudogenes in the 1,646 post-macaque SDs revealed that 656 parental and 192 derivative loci contain genes (Table 4), while significantly more pseudogenes (all types) are in the derived regions. The numbers of genes and pseudogenes for large SDs are also shown in Figure 5, which clearly illustrates that genes and pseudogenes are enriched in the parental and derived loci, respectively. These data suggest that duplicated sequences in the derived loci are more frequently subject to degeneration and pseudogenization than the parental sequences. It is also possible that duplications yield mostly 'broken' genes in the new locations. However, the combined number of genes and pseudogenes is also higher in the parental SDs. Moreover, when both parental and derived loci were compared to their 'ancestral' locus in the macaque genome (Figure 2), the average sequence identity was 89.8% (±5.9%) and 88.8% (±6.1%) for the parental and derivative, respectively. This difference is statistically significant (p = 3e-10), further suggesting a faster degeneration of derived sequences.Table 4Numbers of post-macaque SD loci with genes or pseudogenesParentalDerivativeRefSeq genes656 (716)192 (213)Duplicated pseudogenes113 (131)251 (279)Processed pseudogenes161 (219)269 (331)Other pseudogenes124 (143)209 (232)For reference, the numbers of genes/pseudogenes are also listed in parentheses as some loci can have more than one gene or pseudogene.Pseudogenization and asymmetry in histone modificationsHow does the asymmetry in histone modifications relate to gene content and gene death in human SDs? The asymmetry of pol II ChIP tags is certainly consistent with the biased distribution of genes because more pol II tags usually indicate higher degrees of transcriptional activity. This correlation is further supported by the observation that most histone modifications enriched at promoters are higher in parental SDs (Tables 2 and 3).The asymmetric distribution of genes, however, cannot fully account for the asymmetric profiles of histone modifications described above. Firstly, the asymmetrical pattern remained present, though consisted of fewer marks, when the above t-test was restricted to 623 post-macaque SD pairs containing neither genes nor pseudogenes in both loci. The significantly different modifications are H3K9me1, H3K27me1, H3K4me1, H3K4me2, H3K79me1, and H3K79me2. Secondly, analysis of SDs without genes also detected a skew for the histone marks H3K9me1, H3K27me1, H3K79me2, H4K20me3, and the three states of H3K4 methylation. All of these modifications occurred more frequently on the parental loci, except H4K20me3, which was previously found to associate with repressive chromatin [21]. Thirdly, an analysis restricted to 419 SD pairs that did not exhibit a difference in pol II between their two copies (defined as difference of pol II <0.3 tag per kb) found several marks with significant asymmetry, including H3K9me1, H3K27me1, H3K27me2, H3K36me1, H3K36me3, and H3K79me2. It is interesting to see that H3K79me2, which was found without a significant preference toward either active or silent genes [14], shows a difference here. In this analysis, the statistics for pol II is a p-value of 0.46.Gene and pseudogene contents, nevertheless, have an influence on the asymmetrical pattern of epigenomic modifications (Figure 5). Not only did fewer marks exhibit a difference in the characterizations of 'gene-depleted' SDs, but also the pattern was less biased to the parental copies. For example, the difference of mean tag densities was 1.215, 0.897, 1.562, 0.703, and 0.427 for H3K9me1, H3K27me1, H3K4me1, H3K4me2, and H3K79me1, respectively (Table 2). These numbers decreased to 0.461, 0.389, 0.741, 0.271, and 0.357, respectively, for the SD pairs without genes or pseudogenes. In addition, a characterization of SD pairs (n = 103) with genes in both of their loci did not find a modification with a significantly asymmetrical pattern, though a difference was observed for H3K36me3 and H4R3me2 (unadjusted p-value < 0.001).Shift in the patterns of differences in histone modification as segmental duplications ageFinally, in order to address the dynamics of the above asymmetries during evolution, the post-macaque SDs were split into four groups based on pairwise nucleotide sequence identity of SD pairs (Table 5). The parental and derivative copies of young SDs (sequence identity ≥0.975) exhibited uneven H3K27me1, H3K36me3, H3K9me1, and H4R3me2 modifications. The first two marks were both enriched downstream of transcription start sites [14]. As SD sequences age, more modifications with an asymmetric pattern emerge and then potentially disappear, but differences in H3K27me1 and H3K9me1 modifications persist. Although a difference in gene content was observed across all age groups, this analysis found that as SDs evolve more genes in the derivative loci have been lost, presumably becoming pseudogenes (Table 5). Pseudogenes (of all three types) were always more abundant in the derivative than the parental loci. This is true even for the oldest SDs, though the difference becomes statistically less significant; for example, the means of duplicated pseudogenes were 0.157 and 0.238 for the parental and derivative regions (p-value = 0.02), respectively.Table 5Features with asymmetric distribution between the parental and derivative loci of post-macaque SDs grouped by sequence identitySequence identity<0.925 (n = 330)0.925-0.95 (n = 444)0.95-0.975 (n = 570)≥0.975 (n = 302)Significantly different modifications (paired t-test, adjusted p-value <0.001)H3K27me1H3K27me1H2BK5me1H3K27me1H3K4me2H3K36me1H3K27me1H3K36me3H3K9me1H3K36me3H3K36me3H3K9me1H3R2me1H3K4me1H3K4me1H4R3me2H3K4me2H3K4me2H3K79me1H3K79me1H3K79me2H3K9me1H3K9me1H3R2me1H3R2me1H4K20me1Genes/pseudogenes (p-value <0.001) RefSeq genes0.375/0.0870.341/0.0870.367/0.1270.410/0.191 Duplicated pseudogenesNone0.089/0.2230.043/0.135None Processed pseudogenesNone0.238/0.4130.104/0.3670.063/0.296 Other pseudogenesNone0.089/0.2530.056/0.2380.055/0.285The values for genes/pseudogenes are the average number of genes (or pseudogenes) per 1 kb for parental/derivative sequences. Only features with statistical significance are listed.DiscussionDuplication of genomic sequence is an important evolutionary process that supplies raw genetic material for architectural as well as functional innovations. Its prevalence has been observed in all three kingdoms of life, with several distinct mechanisms leading to their abundance [2,5,22]. A duplication occurring in a single individual can be fixed or lost in the population, but the most common consequence seems to be the loss of all or part of the newly duplicated sequences through deletion or degeneration. Nonetheless, a novel biochemical function can sometimes arise from the redundant sequences.The asymmetrical distributions of histone modifications, genes, pseudogenes, and transcription (with pol II as the proxy) between parental and derivative loci of human SDs support that degeneration (or pseudogenization) is more common than innovation (or neofunctionalization) after gene duplications. One important discovery here is the depletion of genes and, conversely, the enrichment of pseudogenes in the derivative loci. This implies either that most duplications are incomplete when occurring - that is, only part of a gene is duplicated to the new location, resulting in a pseudogene at birth - or that deletion plays a large role in disabling the descendant sequences. The former is supported by more non-processed pseudogenes in derivative regions, while the latter is probably related to the difference in the sum of genes and duplicated pseudogenes in the two copies (Table 4), though it may be influenced by incomplete gene annotation in SDs as well. The results suggest that the original copy is evolutionarily constrained to maintain its functional status while the descendant is relatively free to mutate and can eventually become a 'non-functional' sequence. It is kind of amazing to see that an organism can achieve this given that the two copies are seemly identical in their primary sequences. The current report of gene difference is also consistent with a recent finding that core duplicons, the common DNA subunits sharing by multiple SDs, are enriched for genes and spliced expressed sequence tags [18]. Unfortunately, due to the limitation of the current strategy for identifying the direction of duplication, not enough SD data were produced to address precisely the different rates of pseudogenization in the parental and derived loci. This issue will be addressed in the future when more primate genomes are sequenced and improved algorithms are developed for reliably identifying SDs of sequence identity <90%.The asymmetry of histone modifications can be a direct consequence of more genes and fewer pseudogenes in the parental loci as histone modification is a process often occurring near genes that can lead to either gene activation (for example, H3K4 methylation) or repression (for example, H3K27 methylation). Such a correlation is apparent for H3K4me3 in large SDs (Figure 5). It is also supported by the analysis of SD pairs containing functional genes in both of their loci, whereas almost no modifications exhibited a significantly unsymmetrical pattern. The small sample size, however, could be an issue for generalizing that result.Alternatively, the current findings may suggest that the chromatins in derivative SDs are looser relative to those in the parental. Under this scenario, the genomic sequences in the derived loci are prone to mutations because of their greater exposure, leading to more pseudogenes in evolution, and the turnover rate of nucleosomes in the derivative regions is higher (that is, exchange faster with free histones), resulting in fewer modified histones being detected experimentally. This can explain why higher levels of various modifications were always seen in the parental SDs. Likewise, loose chromatins are more vulnerable to retrotranspositions; as a result, more processed pseudogenes were inserted into the derived loci of SDs (Tables 4 and 5). Along the same line, it is worth noticing that derived loci containing duplicated pseudogenes often have processed pseudogenes too (Figure 5). Furthermore, this hypothesis is particularly supported by the data from a recent study [23] that mapped nucleosome positions using the Solexa sequencing technique. Analysis of those reads indeed revealed that nucleosomes were relatively depleted (p << 0.001) in the derived SDs.Other biological processes may have also contributed to the asymmetries reported here. First of all, the derived SDs may have ended up in regions of repressive chromatins. The genome distribution of post-macaque SDs showed that centromeres contained slightly more derivative SDs than parental SDs (data not shown). However, characterization of post-macaque SD pairs (n = 1,313) whose two loci were at least 5 Mb away from heterochromatin regions found essentially the same asymmetrical histone modifications that were observed for all post-macaque SD pairs. For example, the study of those restricted SD pairs also showed that mono-methylations of H3K9 and H3K27 were higher in the parental SDs but not di- and tri-methylations of H3K9 and H3K27. Since H3K27 and H3K9 methylations are often associated with chromatin repression [10,14,20] but they did not exhibit an enrichment in the derived SDs, the impact of repressive chromatin on the observed asymmetry of histone modifications is small but warrants further investigations. On the other hand, these results cannot rule out that histone modifications may have been directly involved in the initial regulation of the descendant sequences by keeping the extra genomic copies in a silent chromatin state (for example, by not modifying histones). Conversely, histone modifications may have facilitated degeneration of the descendant sequences by increasing the accessibility of those DNAs for a greater rate of mutations. Both scenarios are very important for understanding SD evolution; however, they cannot be confidently separated in the current analysis.In any case, the characterization of either SDs without genes or SDs without pol II asymmetry shows that asymmetrical distributions of several histone modifications were not entirely entangled with gene/pseudogene asymmetries. It is very difficult to really resolve the interaction between sequence degeneration (or pseudogenization) and epigenomic changes, largely because almost all histone marks that have been characterized were investigated in the context of gene expression (either activation or repression). Analysis of post-macaque SDs in different age groups did not help untangle this issue either. If pseudogenization is the cause of asymmetric histone modifications, we would expect to see more asymmetries emerge as SDs age; conversely, we would see asymmetries fade away if they facilitate pseudogenization. The data in Table 5 provide evidence for both or neither, depending on one's interpretation. Further studies are required to address all these questions, and more generally to fully appreciate the potential importance of epigenetic modifications in the initial regulation and subsequent evolution of duplicated sequences. As shown in Figure 4, not all parental loci of SDs exhibited a higher level of histone modification than their derivative regions. Such SD pairs may contain asymmetrical histone acetylations or phosphorylations, or the derivative loci have newly emerging functional elements. In the future, integrated analysis of different types of modifications is certainly necessary as the effect of an individual histone modification is likely context-dependent and cannot simply be referred to as either activating or repressing a chromatin domain [10,11,13,14].Finally, the asymmetry in histone modifications may be relevant to the established view that divergence in regulatory elements is the first step of functional divergence between duplicated genes. Several previous studies have suggested that duplicated genes often evolve at different rates; for instance, one study found that expression of duplicated genes tends to evolve asymmetrically [6]. The expression of one copy evolves rapidly, likely through changes in its regulation, whereas the other one largely maintains the ancestral expression profile. It will be interesting to see if the changing copy is the parental or the derivative and whether histone modifications are involved in establishing the disparate profile of expression and in facilitating subsequent functional divergence. A separate study has found that retrotransposed genes tend to undergo accelerated evolution relative to their parental genes [24]. The discovery of asymmetrical histone modifications here is consistent with these early results and points to a new direction to explore those early findings.ConclusionThis study is important for understanding both the functional influence and evolutionary fate of SDs because it indicates that derivative sequences of SDs become non-functional more often than the originals, as measured by histone modifications, transcription, and density of genes or pseudogenes. This finding is significant because it pinpoints, for the first time, derived sequences as the main locations of divergent evolution between duplicated genomic regions, suggesting that evolution selects a parental locus to maintain its original biological property but allows its derivative sequence to mutate freely, eventually leading to either degeneration or functional innovation.Materials and methodsThe SD regions were obtained from the UCSC browser [15,16]. The hg18 version contained 51,809 pairs of SDs. After redundant entries were removed, as a pair of segmental duplications was usually listed twice by switching the order of the two regions, 25,914 non-redundant SD pairs were used for this study. RefSeq genes [25] were also downloaded from the UCSC browser and then overlapping transcripts were collapsed into a gene. Human pseudogenes were obtained from the Pseudogene.org database [26]. The identification of these pseudogenes has been described previously [27,28]. Processed and duplicated pseudogenes were separated from the rest, which usually do not contain obvious sequence features of retrotranspositions or exon-intron structures [26,28].Two sets of histone profiling data were used, one for the human resting CD4+ T cells [14] and the other for the human embryonic stem cells [17]. These data (or tags) identified the human genomic regions where modifications of nucleosomes or binding of pol II and CTCF were detected. In both cases, the genomic coordinates of ChIP tags were obtained from the original authors and this study did not re-align ChIP sequencing reads to the human genome. Figure 1 describes the general strategy of counting ChIP tags for individual segmental duplications. For statistical analysis, the number of tags per 1 kb genomic region was used to represent the level of each modification. A similar approach was applied to map genes or pseudogenes into SDs, but a gene or pseudogene was assigned to an SD if it overlaps this SD by at least 1 bp.Figure 2 illustrates the approach for identifying segmental duplications that arose after the split of human and macaque ancestors. Its principle is chromosomal synteny between the human genome and the macaque genome. Using a very strict criterion, this method recognized 2,654 SD events after the divergence; however, it only resolved the direction of duplications for 1,646 SD pairs. This strategy was designed to only extract (post-macaque) SDs with an easily identifiable direction of duplication.In order to characterize the distribution of histone modifications within SDs in detail, a non-statistical method was applied to 185 large (>15 kb) post-macaque SD pairs. Each of these SDs was divided into a set of continuous but disjointed blocks (5 kb), which was in turned represented by a vector describing ChIP tags. Thus, the parental vector was P = [p1, p2, ..., pm] and the derivative vector was D = [d1, d2, ..., dn], where Pi and Di were the numbers of ChIP-Seq tags in the i-th block. 1..n was ordered with respect to 1..m, and m = n for most SDs. Let x = y = 0; and for i in 1..b (b be the smaller of m and n), x increased 1 if Pi > Di but y increased 1 if Pi < Di. Then, for each pair of SDs, its parental locus was considered to have a higher level of a histone modification if x > 2/3 * b, otherwise, the derivative locus was higher if y > 2/3 * b. The result of this analysis is shown in Table 3, and the P and D for four pairs of SDs with their duplication directions known are illustrated in Figure 4.The ChIP-Seq signal 'peaks' in these large SDs were also identified. Many software and algorithms exist for calling peaks from ChIP-Seq reads; however, they were not used here because ChIP-Seq reads in SDs have a distribution quite different from those in non-SD regions (tag density is much lower; Table 1). Instead, a peak here was simply defined as a block (5 kb) with >5 ChIP-Seq reads and the read count was also two standard deviations above the average read in this SD. This method correctly reported the apparent H3K4me3 peaks in the first and forth example of Figure 4. The numbers of such peaks for the 185 large SD pairs are plotted in Figure 5 for H3K4me3, H3K27me3, and H3K36me3 because these three methylations are well characterized in the literature.AbbreviationsChIP-Seq, chromatin immunoprecipitation and direct sequencing; pol II, RNA polymerase II; SD, segmental duplication; TSS, transcription start site.\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2530872.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2530872",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2530872\nAUTHORS: Vu T Chu, Raphael Gottardo, Adrian E Raftery, Roger E Bumgarner, Ka Yee Yeung\n\nABSTRACT:\nMeV+R provides users with point-and-click access to traditionally command-line-driven tools written in R.\n\nBODY:\nRationaleWhile microarray technology has given biologists unprecedented access to gene expression data, reliable and effective data analysis remains a difficult problem. There are many freely or commercially available software packages, but biologists are often faced with trading off power and flexibility for usability and accessibility. In addition to the potentially prohibitive costs, researchers using commercial software tools may find themselves waiting for state-of-the-art algorithms to be implemented with the packages. The Bioconductor project [1,2] is an open source software project that provides a wide range of statistical tools primarily based on the R programming environment and language [3,4]. Taking advantage of R's powerful statistical and graphical capabilities, developers have created and contributed numerous Bioconductor packages to solve a variety of data analysis needs. The use of these packages, however, requires a basic understanding of the R programming/command language and an understanding of the documentation accompanying each package. The primary users of R and the Bioconductor packages have been computational scientists, statisticians and the more computationally oriented biologists. However, in our experience, many biologists find themselves uncomfortable issuing command lines in a terminal. Hence, there is a need for a graphical user interface (GUI) for Bioconductor packages that will allow biologists easy access to data analytical tools without learning the command line syntax. The tcltk package in R adds GUI elements to R by allowing programmers to write GUI-driven modules by embedding Tk commands into the R language [5]. There are also GUIs developed for basic statistical analysis in R, such as the R Commander [6] and windows-based SciViews [7]. However, these GUIs are not designed for microarray analysis. There are Bioconductor packages, such as limmaGUI [8], affylmGUI [9] and OLINgui [10] that are built on the R tcltk package to provide GUIs. LimmaGUI and affylmGUI provide GUIs for the analysis of designed experiments and the assessment of differential expression for two-color spotted microarrays and single-color Affymetrix data, respectively. OLINgui provides a GUI for the visualization, normalization and quality testing of two-channel microarray data. However, no such GUIs are available for the majority of Bioconductor packages. In addition, since each Bioconductor package is often written by a different research group, there is generally no uniformity in the look and feel of the GUIs available for the different packages. Hence, the end user may not be able to easily transfer experience gained with one analysis tool to the use of another.An alternative microarray data analysis tool is the MultiExperiment Viewer (MeV), a component of the TM4 suite of microarray analysis tools [11]. MeV has a user-friendly GUI designed with the biological community in mind. MeV is an open source Java application with a simple to learn, easy to use GUI. It comes with many popular microarray analytical algorithms for clustering, visualization, classification and biological theme discovery, such as hierarchical clustering [12] and Expression Analysis Systematic Explorer (EASE) [13]. MeV was carefully designed to provide an application programming interface (API), thus allowing straightforward contributions by the community. MeV is hosted at SourceForge [14] in a concurrent versions system repository. As such, frequent builds of the source code are made possible, greatly reducing the lag time between version releases.In this paper, we present MeV+R, which is an effort to provide more consistent and well-integrated GUIs for Bioconductor packages by using MeV as a 'wrapper' application for Bioconductor methods. Our work brings the best of both worlds together: providing state-of-the-art statistical algorithms from Bioconductor through the open source and easy to use MeV graphical interface to the biomedical community. MeV+R has many advantages, including platform independence, a well-defined modular API, and a point and click GUI that is easy to learn and use. We demonstrate the successful integration and advantages of three Bioconductor packages (RAMA [15], BRIDGE [16], and iterativeBMA [17]) over existing tools in the MeV environment through case studies. The underlying framework that we used to integrate these Bioconductor packages with MeV is easily extensible to other analysis tools developed in R. The software, documentation and a tutorial are publicly available from our project home page [18].ImplementationOur integration effort is composed of three separate entities (Figure 1). MeV provides the graphical user interface while Rserve serves as the communication layer and R is the language and environment in which the analysis packages run. Rserve is a TCP/IP server that allows various languages to use the facilities of R without the need to initialize R or link against an R library [19]. In other words, we use R as the back end to run Bioconductor packages through the use of Rserve. Rserve is open source, freely available [20], and licensed under GPL.Figure 1Our integration effort is composed of three separate entities: MeV as the GUI, Rserve as the communication layer, and R as the language and environment in which the analysis packages run.As such, Java, Rserve, and R must all be installed on the user's computer, and we provide an automated installer on our project web site. Furthermore, Rserve needs to be running to be used. However, R does not need to be started. Since Rserve works through TCP/IP, it can run on the user's own machine, on an internal network or over the internet. By default, our code assumes Rserve to be running on the local host, but the user can change, add and save additional new hosts using a pull down menu. Once a connection is established, the Java code in MeV converts the user's data from the MeV data structure to the R format and loads it into R. The appropriate R libraries are loaded followed by the R commands that are necessary to initiate the analysis. Upon completion, the returned data from R are explicitly called back into MeV and presented to the user.We have incorporated three Bioconductor packages, RAMA [15], BRIDGE [16], and iterativeBMA [17], into MeV to illustrate the successful MeV+R integration. The Robust Analysis of MicroArray (RAMA) algorithm computes robust estimates of expression intensities from two-color microarray data, which typically consist of a few replicates and potential outliers [15]. RAMA also takes advantage of dye swap experimental designs. Bayesian Robust Inference for Differential Gene Expression (BRIDGE) is a robust algorithm that selects differentially expressed genes under different experimental conditions on both one- and two-color microarray data [16]. Both RAMA and BRIDGE make use of a computationally intensive technique called Markov Chain Monte Carlo for parameter estimation, and it is non-trivial to re-implement these algorithms in Java. Hence, we took advantage of our previous development work by simply using MeV as an interface to the Bioconductor packages. The iterative BMA algorithm is a multivariate gene selection and classification algorithm, which considers multiple genes simultaneously and typically leads to a small number of relevant genes to classify microarray data [17]. The iterativeBMA Bioconductor package implements the iterative BMA algorithm as previously described [17] in R, and its implementation is part of our current integration effort. Both RAMA and BRIDGE are included in the latest release of MeV (version 4.1), and iterativeBMA will be included in future releases. The user interfaces, usage and case studies for RAMA, BRIDGE and iterativeBMA are briefly described below. Detailed documentation is included with the software distribution [21] as well as linked in the MeV application. Help pages are also available as Help Dialogs accessed via buttons on the MeV dialog boxes. Our MeV+R implementation is publicly available and runs on Windows, Mac OS X and Linux.Integrated Bioconductor packages: description and user interfacesRAMA: Robust Analysis of MicroArraysRAMA uses a Bayesian hierarchical model for the robust estimation of cDNA microarray intensities with replicates. This is highly relevant for replicated microarray experiments because even one outlying replicate (such as due to scratches or dust) can have a disastrous effect on the estimated signal intensity. Outliers are modeled explicitly using a t-distribution, which is more robust than the usual Gaussian model. Our model borrows strength from all the genes to decide if a measurement is an outlier, and hence it is better at detecting outliers based on a small number of replicate measurements than other classical robust estimators. Our algorithm uses Markov Chain Monte Carlo for parameter estimation, and addresses classical issues such as design effects, normalization, transformation, and nonconstant variance. Please refer to [15] for a detailed description of the algorithm.User interfaceThe user can start RAMA by clicking 'Adjust Data' - 'Replicate Analysis' - 'RAMA' from the MeV main menu. The RAMA dialog box is then displayed asking the user to label the arrays that were loaded into MeV with their appropriate dye color. At this time, the user is asked to make sure that Rserve is running. On a Win32 system, double clicking Rserve.exe accomplishes this. On a UNIX or Linux or Mac OS X system, the user issues the command 'R CMD Rserve' at a prompt. By default, RAMA will look on the local machine for an Rserve server. However, since Rserve is a TCP/IP server, the Rserve server can be a remote machine. The user is allowed to adjust a few advanced parameters, though suggested values are given as defaults. If an Rserve connection is successfully made, the location of Rserve is written to the user's MeV configuration file and will be available in later sessions. After clicking 'OK', the input data are sent to R. An indeterminate progress bar is displayed while RAMA runs - unfortunately, the architecture of RServe and the R Server do not allow for an accurate indication of the time remaining in an ongoing analysis. Once completed, the user is given a dialog box to save the results. The returned results will then replace the loaded data in a new Multiple Array Viewer (MAV). The old MAV is deleted. The user can then choose to continue using MeV as if the data were loaded through the native loading modules.BRIDGE: Bayesian Robust Inference for Differential Gene ExpressionBRIDGE fits a robust Bayesian hierarchical model to test for differentially expressed genes on microarray data. It can be used with both two-color microarrays and single-channel Affymetrix chips. BRIDGE builds on the previous work of Gottardo et al. [15] by allowing each gene to have a different variance and the detection of differentially expressed genes under multiple (up to three in our current implementation) experimental conditions. Robust inference is accomplished by modeling outliers using a t-distribution, and hence BRIDGE is powerful even with a small number of samples (either biological or technical replicates) under each experimental condition. Parameter estimation is carried out using a novel version of Markov Chain Monte Carlo. The current implementation of BRIDGE does not handle missing values. Please refer to [16] for a detailed description of the model.User interfaceBRIDGE starts when a user clicks the 'BRIDGE' button in the toolbar located on top of the MeV window. The user is once again presented with a dialog box similar to that of RAMA asking for the dye labeling identity of each loaded slide. The user is offered the option to adjust the advanced parameters and to establish an Rserve connection. After clicking OK, the input data are sent to R. An indeterminate progress bar is displayed while BRIDGE runs. The results are presented to the user in three formats: heat maps, expression graphs or tables. In each format, the genes for which there is strong evidence of differential expression are identified as 'Significant Genes', defined by the posterior probability being above 0.5.IterativeBMA: Iterative Bayesian Model AveragingThe iterativeBMA algorithm is a multivariate technique for gene selection and classification of microarray data. Bayesian Model Averaging (BMA) takes model uncertainty into consideration by averaging over the predicted probabilities based on multiple models, weighted by their posterior model probabilities [22]. The most commonly used BMA algorithm is limited to data in which the number of variables is greater than the number of responses, and the algorithm is inefficient for datasets containing more than 30 genes (variables). In the case of classifying samples using microarray data, there are typically thousands or tens of thousands of genes (variables) under a few dozen samples (responses). In the iterative BMA algorithm, we start by ranking the genes using the ratio of between-group to within-group sum of squares (BSS/WSS) [23]. In this initial preprocessing step, genes with large BSS/WSS ratios (that is, genes with relatively large variation between classes and relatively small variation within classes) receive high rankings. We then apply the traditional BMA algorithm to the 30 top ranked genes, and remove genes with low posterior probabilities. Genes from the rank ordered BSS/WSS ratios are then added to the set of genes to replace genes with low probabilities. These steps of gene swaps and iterative applications of BMA are repeated until all genes are subsequently considered. We have previously shown that the iterative BMA algorithm selects small numbers of relevant genes, achieves high prediction accuracy, and produces posterior probabilities for the predictions, selected genes and models [17].The iterativeBMA Bioconductor package implements the iterative BMA algorithm described in Yeung et al. [17] (previously implemented in Splus) when there are two classes. It is part of the original work for this publication. The user documentation (vignette) is included in the package.User interfaceWe have integrated the iterativeBMA Bioconductor package in MeV. IterativeBMA starts after the user clicks on the 'iBMA' icon on top of the MeV window. The current implementation of the iterativeBMA Bioconductor package is limited to only two classes. After loading the data, the user is asked to label the two classes. The default labels for the two classes are 0 and 1, respectively. In the same dialog box, the user is asked to establish an Rserve connection. The user is also given the option of specifying advanced parameters for the analysis. The next dialog box asks the user to assign labels to each of the samples in the data, either by using a pull-down menu or loading an assignment file. At this point, if Rserve is not already running, the user is reminded to start the connection. Then, the data and the parameters are sent to R, and a progress bar is shown warning the user that the computation could take a long time. After the iterativeBMA Bioconductor package finishes running, the following analysis results are displayed: the predicted probability and class for each test sample; the posterior probabilities of the selected genes sorted in descending order; the posterior probabilities of the selected models sorted in descending order; and the heatmaps of the selected genes in both classes.Case studies illustrating the merits of the integrated Bioconductor packagesIn this section, we compare the performance of the integrated Bioconductor packages (RAMA, BRIDGE and iterativeBMA) to existing tools in MeV in order to illustrate the merits of the integrated packages. In addition, we demonstrate that our MeV+R modules can be used together with other MeV modules in the integrated analysis of microarray data, hence, extending the capabilities of MeV.RAMA: Robust Analysis of MicroArraysWe compared the microarray gene intensities estimated using RAMA to that of the log ratios over intensities averaged over all the replicates on two microarray datasets and the results are summarized in Table 1. The first dataset is a subset of the HIV data [24] consisting of the expression levels of 1,028 transcripts, including 13 positive controls and 24 negative controls, in CD4-T-cell lines at time t = 1 hour after infection with HIV virus type 1 hybridized to two-color cDNA arrays. The experimental design consists of four technical replicates and balanced dye swap in which two of the four replicates were hybridized with Cy3 for the control and Cy5 for the treatment and then the dyes were reversed on the other two replicates. The second dataset is a subset of the like and like data [15] consisting of 1,000 genes over four experiments using the same RNA preparation isolated from a HeLa cell line on four different microarray slides. Since the same RNA was used in both channels, no genes from these data should show any differential expression. Both sample datasets are available on our project web site and are included as part of our MeV+R package release.Table 1Comparing the results of RAMA to the averaged log ratios on the HIV data and the like and like dataDataBenchmarkRAMAAveraged log ratioHIV data13 positive controlsAll 13 positive controls have log ratios >1All 13 positive controls have log ratios >124 negative controlsAll 24 negative controls have log ratios <13 negative controls have log ratios >1Like and like dataNo genes expected to be differentially expressedAll log ratios <16 genes with log ratios >1RAMA produced the desired results on both datasets while the averaged log ratio produced three and six false positives, respectively, on these two datasets.Figure 2 shows the log ratios of all genes sorted in descending order after applying RAMA integrated in MeV+R to the HIV data. As shown in Figure 2, the log ratios (to base 2) computed with the robust intensities estimated using RAMA for all 13 positive controls are all greater than one. The log ratios from RAMA for all 24 negative controls are smaller than one (data not shown in Figure 2). On the contrary, computing the log ratios by simply averaging the gene intensities over the four replicates produces log ratios greater than one for three negative controls. Applying RAMA to the like and like data produces no log ratio greater than one as desired since we do not expect any differentially expressed genes. On the contrary, the average log ratio of gene intensities yields six genes with log ratios greater than one. Please refer to the supplementary material [18] for the details of our case studies. To summarize, RAMA produced the desired results on both datasets while the averaged log ratio produced three and six false positives, respectively, on these two datasets.Figure 2The results of applying RAMA to the HIV data. The log ratios computed from RAMA are sorted in descending order, and the top 13 genes with log ratios greater than one are the positive controls.BRIDGE: Bayesian Robust Inference for Differential Gene ExpressionWe compared the differentially expressed genes identified using BRIDGE, t-test and SAM (Significance Analysis of Microarrays) [25] as implemented in MeV on two datasets. Applying BRIDGE to the HIV data described in the previous section identified all 13 positive controls as 'significant' genes (Figure 3). On the other hand, applying the one-sample t-test as implemented in MeV to the same HIV data identified a total of 14 significant genes, including all 13 positive controls and one negative control using a p-value cut-off of 0.01 without any Bonferroni correction. Using a p-value cut-off of 0.05 and standard Bonferroni correction, the one-sample t-test identified only one significant gene (which is one of the 13 positive controls) and incorrectly assigned the remaining 12 positive controls as 'insignificant'. Similarly, using one-sample SAM as implemented in MeV identified 12 out of 13 positive controls using default parameters.Figure 3The significant genes identified by applying BRIDGE to the HIV data.The second dataset we used comprises the Affymetrix U133 spike-in data [26], which consists of three technical replicates of 14 separate hybridizations of 42 spiked transcripts in a complex human background at varying concentrations. Thirty of the spikes are isolated from a human cell line, four spikes are bacterial controls, and eight spikes are artificially engineered sequences believed to be unique in the human genome. The data were preprocessed using GCRMA [27], resulting in a dataset of 22,300 genes across 42 samples. In addition to the original 42 spiked-in genes, we included an additional 20 genes that consistently showed significant differential expression across the array groups and an additional three genes containing probe sequences exactly matching those for the spiked-in genes [28,29]. As a result, our expanded spiked-in gene list contains 65 entries in total. We used a subset of this spiked-in data consisting of 1,059 genes that include all 65 spiked-in genes across two samples in triplicate. In our comparison, only the 65 spiked-in genes should be identified as differentially expressed.BRIDGE identified 45 differentially expressed genes on this data subset. All of these 45 genes identified by BRIDGE are spiked-in genes. On the other hand, the t-test with a p-value cut-off of 0.01 without any correction for multiple comparison identified a total of 33 significant genes, of which 31 were spiked-in genes. Using a p-value cut-off of 0.05 and the standard Bonferroni correction, the t-test identified only four significant genes (which are among the spiked-in genes). SAM identified eight spiked-in genes as differentially expressed.Our comparison results are summarized in Table 2. We have shown that BRIDGE is the only tool that successfully identified all 13 positive controls as 'significant' on the HIV data. In addition, BRIDGE identified the highest number of true positives (spiked-in genes) without any false positives on the Affymetrix spike-in data.Table 2Comparing the results of BRIDGE to t-test and SAM on the HIV data and the Affymetrix spike-in datat-testDatasetBenchmarkBRIDGEp-value cut-off 0.01, no correctionp-value cut-off 0.05, standard Bonferroni correctionSAMHIV data13 positive controls, 24 negative controls DE1314112 TP1313112 FP0100Affymetrix spike-in data65 spike-in genes DE453348 TP453148 FP0200For each dataset and each method, the number of differentially expressed (DE) genes, true positives (TP) and false positives (FP) are shown. For each dataset, the maximum TP and the minimum FP across all methods are shown in bold. BRIDGE produced the best results on both datasets in identifying the highest number of true positives without any false positives.IterativeBMA: Iterative Bayesian Model AveragingWe compared the performance of iterativeBMA (abbreviated as iBMA in our MeV+R implementation) to KNN (k-nearest neighbor) [30] and USC (Uncorrelated Shrunken Centroid) [31] implemented in MeV using the well-studied leukemia data [32]. We used the filtered leukemia dataset, which consists of 3,051 genes, 38 samples in the training data and 34 samples in the test set. The data consist of samples from patients with either acute lymphoblastic leukemia (ALL) or acute myeloid leukemia (AML). On the leukemia data, iterativeBMA produced 2 classification errors using 11 selected genes over 11 models (Figures 4 and 5). On the other hand, KNN does not have a gene selection procedure and produced 2 classification errors using all 3,051 genes. Similarly, USC produced 2 classification errors using 51 selected genes.Figure 4The results of applying iterativeBMA to the leukemia data. A heatmap showing the selected genes from iterativeBMA under the training samples labeled as class 0 and the test samples assigned to class 0 by the algorithm.Figure 5The results of applying iterativeBMA to the leukemia data. A heatmap showing the selected genes from iterativeBMA under the training samples labeled as class 1 and the test samples assigned to class 1 by the algorithm.The second dataset we used is the breast cancer prognosis dataset [33], which consists of 4,919 genes with 76 samples in the training set, and 19 samples in the test set [17]. The patient samples are divided into two categories: the good prognosis group (patients who remained disease free for at least five years) and the poor prognosis group (patients who developed distant metastases within five years). The iterativeBMA algorithm produced three classification errors using four genes averaged over three models. On the other hand, KNN does not have a gene selection procedure and produced five classification errors using all genes. Similarly, USC produced four classification errors using 662 genes.Our results are summarized in Table 3. On the breast cancer prognosis data, iterativeBMA produced higher prediction accuracy using much fewer genes. On the leukaemia data, iterativeBMA produced comparable prediction accuracy using much fewer genes.Table 3Comparing the results of iterativeBMA to KNN and USC on the leukemia data and the breast cancer prognosis dataDataSize of dataiterativeBMAKNNUSCLeukemia data [32]38 training samples11 genes3,051 genes51 genes34 test samples2 errors2 errors2 errorsBreast cancer prognosis data [33]76 training samples4 genes4,919 genes662 genes19 test samples3 errors5 errors4 errorsThe number of selected genes and the number of classification errors are shown for each method. For each dataset, the smallest number of genes and the smallest number of classification errors across all three methods are shown in bold. On the leukemia data, iterativeBMA produced the same number of classification errors using much fewer genes. On the breast cancer prognosis data, iterativeBMA produced fewer errors using much fewer genes.Using other MeV modules in an integrated data analysisThe previous sub-sections showed that our MeV+R modules achieved superior performance when compared to other existing tools implemented in MeV. Here we demonstrate how the R packages that we incorporated into MeV can be used in combination with other existing tools in MeV. This illustrates the fact that the MeV+R framework has extended the capabilities of MeV, and that using these R packages through the MeV GUI adds value to the integrated analysis of microarray data.In this case study, we will follow-up on the results from applying the iterativeBMA algorithm to the leukemia data [32]. The iterativeBMA algorithm is a multivariate gene selection method designed to select a small set of predictive genes for the classification of microarray data. In the case of the leukemia data, the iterativeBMA algorithm selected 11 genes that produced two classification errors on the 34-sample test set. It would be interesting to identify the biological theme in this 11-gene list. Towards this end, we applied EASE [13] as implemented in MeV to determine the over-represented Gene Ontology categories in this gene list relative to all the genes on the microarray. Figure 6 shows the tabular view from the EASE analysis.Figure 6The results of applying EASE to the 11 genes selected by iterativeBMA on the leukemia data.Since iterativeBMA identifies a small set of predictive genes for classification, other genes that exhibit similar expression patterns to the selected genes are likely of biological interest. For example, we would like to explore the gene with the highest posterior probability 'X95735_at' from the iterativeBMA analysis on the leukemia data [32]. We applied PTM (Template Matching) [34] as implemented in MeV to identify genes that are highly correlated with 'X95735_at'. Using a p-value threshold of 0.0001, PTM identified 209 genes that are highly correlated with 'X95735_at'. Our next task was to find the biological theme among these 209 genes, so we applied EASE and TEASE (Tree-EASE). TEASE is a combined analytical tool for hierarchical clustering and EASE. TEASE computes the dendrogram using the hierarchical clustering method and displays the significantly enriched Gene Ontology categories for each subtree in the dendrogram. Please refer to the supplementary materials [18] for the details of our case studies.Incorporating additional R packagesWe have developed a framework with built-in functions for the integration of Bioconductor packages into MeV. Detailed documentation of these built-in functions is provided on our project web site for software developers. Using this framework, we have integrated three Bioconductor packages (RAMA, BRIDGE and iterativeBMA) into MeV as proof of concept. To integrate additional Bioconductor packages into MeV, a software developer can simply call our built-in functions except for complex and non-standard data views.ConclusionMeV+R is a convenient platform to provide biologists with point and click GUI access to Bioconductor packages. We have demonstrated the successful integration of Bioconductor and MeV through three Bioconductor packages, RAMA, BRIDGE and iterativeBM, and that the incorporated Bioconductor packages produced superior results in the analysis of microarray data compared to existing tools in MeV. Additional Bioconductor packages are straightforward to add: the framework for moving data from MeV to R and back is generalized for code re-use, and each new package will merely require the development of a GUI for input and output.AbbreviationsAPI, application programming interface; BMA, Bayesian Model Averaging; BRIDGE, Bayesian Robust Inference for Differential Gene Expression; BSS/WSS, ratio of between-group to within-group sum of squares; EASE, Expression Analysis Systematic Explorer; GUI, graphical user interface; iterativeBMA, iterative Bayesian Model Averaging; KNN, k-nearest neighbor; MAV, Multiple Array Viewer; MeV, MultiExperiment Viewer; PTM, Template Matching; RAMA, Robust Analysis of MicroArray; SAM, Significance Analysis of Microarrays; USC, Uncorrelated Shrunken Centroid.Authors' contributionsVC carried out the software implementation, and drafted part of the initial manuscript. RG and AER designed and wrote the Bioconductor packages RAMA and BRIDGE, and assisted in incorporating these packages into MeV. REB conceived of the study, and designed and coordinated the project. KYY participated in the design and coordination of the study, wrote the iterativeBMA Bioconductor package, carried out the case studies and prepared the manuscript. All authors read and approved the final manuscript.\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2531082.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2531082",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2531082\nAUTHORS: Consuelo Huerta, Saga Johansson, Mari-Ann Wallander, Luis A García Rodríguez\n\nABSTRACT:\nBackgroundVenous thromboembolism (VTE) and thromboembolic arterial diseases are usually considered to be distinct entities, but there is evidence to suggest that these disorders may be linked. The aim of this study was to determine whether a diagnosis of VTE increases the long-term risk of myocardial infarction (MI).MethodsThe incidence rate (IR) and relative risk (RR) of MI in a cohort of patients with a diagnosis of VTE (n = 4890) compared with that of a control cohort without prior VTE (n = 43 382) were evaluated in the UK General Practice Research Database (GPRD). Death during follow-up was also determined. Patients were followed for up to 8 years (mean of 3 years).ResultsThe IR of MI per 1000 person-years was 4.1 (95% CI: 3.1–5.3) for the VTE cohort and 3.5 (95% CI: 3.2–3.8) for the control cohort. The IR of MI was highest in the first year after the VTE episode, but overall differences between the two cohorts were not significant (RR of MI associated with VTE: 1.2; 95% CI: 0.9–1.6). The risk of death was higher in the VTE cohort than the control cohort, even after adjustment for cancer, heart failure and ischaemic heart disease (RR: 2.4; 95% CI: 2.2–2.6), particularly during the first year after VTE (RR: 3.8; 95% CI: 3.4–4.3).ConclusionA VTE episode does not significantly increase the risk of MI, but does increase the risk of death, particularly in the first year following VTE diagnosis.\n\nBODY:\nBackgroundVenous thromboembolism (VTE), usually manifested as deep vein thrombosis (DVT) or pulmonary embolism (PE), is usually considered to be a distinct entity from the thromboembolic arterial diseases, such as myocardial infarction (MI), peripheral artery disease and ischaemic stroke. However, both VTE and thromboembolic arterial diseases involve the formation of clots within blood vessels, and so may be linked. Recent studies have investigated the incidence of arterial events in patients with VTE, and have reported data suggesting that there may be a positive association between VTE and thromboembolic arterial disease [1-3]. However, these studies either lacked a suitable control group or involved relatively small numbers of patients.To determine whether a diagnosis of VTE is associated with an increased risk of arterial disease, the incidence of arterial disease in patients with a prior VTE should be compared with that in matched controls without a history of VTE. In this study we aimed to perform such an analysis by assessing the risk of MI in patients with a VTE diagnosis and control patients without VTE using data recorded prospectively from the UK General Practice Research Database. We also aimed to estimate the overall long-term mortality among patients with VTE during a long-term follow-up period in comparison with that of the control population without VTE.MethodsData sourceThe GPRD contains computerized information entered by primary care physicians (PCPs) in the UK. The vast majority of the UK population is registered with a PCP. About 1500 PCPs participate in the GPRD, covering a population of around 3 million individuals, who are broadly representative of the UK population. The PCPs hold the complete medical record of registered individuals, including demographic data, all medical diagnoses, consultant and hospital referrals, and a record of all prescriptions issued. Prescriptions are generated directly from the PCP's computer and entered into the patient's computerized file. All the information is recorded by PCPs during consultations in a standard fashion and practices regularly anonymize and send these data to the Medicines and Healthcare Products Regulatory Agency (MHRA), which is in charge of quality control and management of the data for use in research projects. Several validation studies have shown the accuracy and completeness of data in the GPRD [4,5]. Previous studies have also confirmed the validity of using the GPRD for epidemiological research in the field of DVT and PE [6-10].Study cohortsThe source population included individuals aged 20–79 years enrolled with a participating PCP for more than 2 years during 1 January 1994 to 31 December 2000, without a previously recorded diagnosis of VTE, as described in a previous cohort study of the natural history of VTE [11]. The resulting source population consisted of 1 856 206 patients, and the first day of meeting these eligibility criteria was used as each individual's start date. Of this source population, 6550 patients had a first recorded diagnosis of VTE from the individual's start date until 31 December 2000. The validation (positive predictive value) of a VTE diagnosis in the GPRD has been described previously: a questionnaire was sent to PCPs for a random sample of 5% of patients with a record of VTE and, after reviewing these questionnaires, the diagnosis of VTE was confirmed in 94% of cases [11]. Moreover, we previously reported that the overall incidence rate of VTE in the study cohort was 74.5 per 100 000 person-years [11]. In other epidemiological studies, the incidence of VTE ranged from 71 to 117 case per 100 000 of the population per year (standardized for age and sex) [12].We classified VTE episodes as idiopathic if they occurred in the absence of the following transient risk factors: fracture, surgery, pregnancy or childbirth, or any hospitalization (all occurring in the 3-month period before VTE); cancer in the year before VTE; or use of hormone replacement therapy or oral contraceptives in the 6 months before the date of the VTE episode. We considered all other cases of VTE to be secondary, in a similar manner to other studies [1,13].A control cohort was also identified. For this, 50 000 individuals without a VTE diagnosis were randomly sampled from the source population and matched by age, sex and calendar year to the VTE cohort.Definition of clinical endpointsMyocardial infarctionPatients with a history of ischaemic heart disease prior to their start date were excluded from both VTE and control cohorts for the analysis of the risk of MI following a VTE, as a history of ischaemic heart disease could mask the influence of VTE on a subsequent MI. Each individual's start date was the date of diagnosis of VTE in the VTE cohort and a random date within the study period for the control cohort. Follow up started a month after the start date in order to exclude patients dying due to the initial episode of VTE. Finally, there were 4890 patients in the VTE cohort and 43 382 patients in the control cohort.Patients in both cohorts were then followed up until one of the following endpoints was reached: a recorded code of MI, age of 80 years, death or the end of the study period (31 December 2002). We manually reviewed the computerized profiles of all patients identified with a code of MI, and all deaths. We considered MI cases to be patients whose diagnosis was confirmed by a letter from a consultant cardiologist or on hospital discharge. We also considered as cases: those who died from coronary heart disease (CHD); patients with post-mortem evidence of a recent MI or a recent coronary artery occlusion; patients with ante-mortem evidence of CHD in the absence of another cause of death; and patients for whom CHD was recorded as the underlying cause of death. We did not contact PCPs for further confirmation of the diagnosis of MI, as our experience from a previous study of MI in the GPRD has shown that case ascertainment after manual review of the computerized information supports our case definition in more than 90% of instances [14].MortalityFor the mortality analysis, individuals with a history of ischaemic heart disease prior to the start date were not excluded, but follow-up was started 1 month after the episode of VTE as before. (Data for patients who died within the first month of the VTE diagnosis have been presented elsewhere [11].) The VTE cohort consisted of 5801 patients and the control cohort consisted of 48 399 patients. Patients were followed until death, age of 80 years or the end of the study period (31 December 2002).Analysis of MI riskRelative risk and Kaplan-Meier survival analysisEstimates of MI occurrence (with 95% CI) were calculated for the VTE and control cohorts. Individuals alive at the end of the study period were regarded as censored from that date, while individuals with their last practice visit before the end of the study were regarded as censored from the date of their last practice visit. The cumulative hazard of MI was calculated using a Kaplan-Meier survival analysis. Cox proportional hazards regression was used to estimate the relative risk (RR) and 95% confidence intervals of MI in the VTE cohort compared with the control cohort (overall and according to type of VTE). Variables included in the multivariate model were the presence of heart failure and hypertension, as well as frequency-matched variables (age, sex, and calendar year).Analysis of mortalityDeaths from any cause during the follow-up period were analysed using Kaplan-Meier life-tables to compare survival between patients with or without VTE. Cox proportional hazards regression was used to estimate the RR and 95% CI of death in the VTE cohort compared with the comparison cohort. Variables included in the multivariate model were the presence of cancer or heart failure as well as the frequency-matched variables (age, sex and calendar year). All statistical analyses were conducted using STATA (version 8.2; Stata Corporation, College Station, Texas, USA).ResultsRisk of myocardial infarctionThe incidence and risk of MI were determined for the cohort of VTE patients (n = 4890) and the control cohort (n = 43 382), in order to provide an estimate of the RR of MI following VTE. The age distribution in the two cohorts was successfully matched, with 12.3% of both cohorts aged 20–39 years, 33.3% aged 40–59 years, 27.1% aged 60–69 years and 27.3% aged 70 years or older. During the mean follow-up period of 3 years (range: 3–8 years; median: 3 years), MI occurred in 55 patients from the VTE cohort and 472 patients from the control cohort. Thus the incidence rate (IR) of MI per 1000 person-years was 4.1 (95% CI: 3.1–5.3) for the VTE cohort and 3.5 (95% CI: 3.2–3.8) for the control cohort. The difference between the two groups was not significant, as shown by the RR of MI (RR: 1.2; 95% CI: 0.9–1.6). The incidence rates of MI were within the range reported in previous population-based studies in the UK (2.73–8.23 and 0.66–2.56 per 1000 person-years in men and women, respectively [15,16]).The IR of MI increased with age in both cohorts (see Figure 1). Although the IR of MI was numerically greater in the VTE cohort than the control cohort for those aged 60–69 years or at least 70 years, determination of RR indicated that these differences were not significant (RR: 1.3; 95% CI: 0.8–2.0 for patients aged 60–69 years, RR: 1.4; 95% CI: 0.9–2.0 for those aged 70 years or more). The cumulative proportion of patients diagnosed with MI over time for the two cohorts is shown in Figure 2. This analysis showed that the cumulative proportion of patients with MI was slightly greater in the VTE cohort than the control cohort in the first year, but this difference narrowed in years 2–4. Indeed, the increased risk of MI in the VTE cohort compared with the control cohort in the first year was of borderline significance (adjusted RR: 1.6; 95% CI: 1.0–2.5). After the first year, the adjusted RR of MI associated with VTE fell to 1.0 (95% CI: 0.7–1.5).Figure 1Incidence rate (IR) and relative risk (RR) of myocardial infarction in the venous thromboembolism (VTE) cohort and control cohort according to age.Figure 2Cumulative proportion of patients diagnosed with myocardial infarction (MI) in the venous thromboembolism (VTE) cohort and control cohort over time (log-rank test > 0.05).Further analysis of the risk of MI in the VTE cohort compared with the control cohort for the first year showed that the excess risk was of borderline significance in patients aged between 60 and 69 years (RR: 2.0; 95% CI: 1.0–4.0) and insignificant in the younger age group (40–59 years of age, RR: 0.70; 95% CI: 0.1–5.3) and older age group (≥70 years, RR: 1.5; 95% CI: 0.8–2.9). The risk of MI in the VTE cohort, however, was similar for patients who had had DVT (n = 33) and for patients who had had PE with or without DVT (n = 22) (Table 1). The risk of MI was also similar regardless of whether the VTE was idiopathic or secondary (Table 1).Table 1Incidence rate and relative risk of myocardial infarction in the venous thromboembolism cohort compared with the control cohort in the first year of follow up.Myocardialinfarction  cases (n = 159)Incidencerate per 1000  person-yearsRelative risk (95% CI)aControl cohort (n = 43 382)1363.4 (2.9–4.0)1Venous thromboembolism cohort (n = 4890)23Deep vein thrombosis155.7 (3.4–9.4)1.7 (1.0–2.9)Pulmonary embolism84.8 (2.4–9.6)1.5 (0.7–3.1)Secondary venous thromboembolism135.2 (3.0–8.9)1.6 (1.0–2.8)Idiopathic venous thromboembolism105.6 (3.0–10.4)1.6 (0.8–3.2)aAdjusted relative risk derived from Cox regression models including sex, age, calendar year, heart failure, hypertension, and smoking.MortalityDuring the total follow-up period of 8 years, 3088 patients died: 2266 of 48 399 in the control cohort and 822 of 5801 in the VTE cohort. Overall mortality was, therefore, higher in the VTE cohort (14.2%; 49.5 per 1000 person-years) than the control cohort (4.7%; 14.5 per 1000 person-years) (Figure 3). After adjustment for the presence of cancer, ischaemic heart disease and heart failure, the RR of death during this 8-year period in the VTE cohort compared with the control cohort was 2.4 (95% CI: 2.2–2.6). When we re-analyzed the mortality findings after excluding patients with a history of ischaemic heart disease to allow comparison with the incidence of MI in these cohorts, the mortality rates dropped to 47.4 and 12.1 per 1000 person-years in the VTE and control cohorts, respectively.Figure 3Kaplan-Meier survival curves for venous thromboembolism (VTE) cohort and control cohort (*log-rank = 1005.25; p < 0.005).The increased risk of death associated with VTE was much greater in the first year after VTE diagnosis (RR: 3.8; 95% CI: 3.4–4.3) than in subsequent years (RR: 1.6; 95% CI: 1.8–1.4) (Table 2). The risk of death in the first year was also greater in patients with a diagnosis of DVT (RR: 4.4; 95% CI: 3.9–5.1) than in those with a diagnosis of PE (RR: 2.9; 95% CI: 2.5–3.5) (Table 2). Compared with the control group, mortality was increased in patients aged 20–59 years (RR: 10.5; 95% CI: 7.3–15.1) and at least 60 years (RR: 3.1; 95% CI: 2.7–3.6).Table 2Mortality and relative risk of MI in the venous thromboembolism cohort compared with the control cohort, according to year of follow up.First yearAfter 1 yearDeaths (n = 1216)Mortality/1000person-years  (95% CI)RR (95% CI)aDeaths (n = 1872)Mortality/1000person-years  (95% CI)RR (95% CI)aControl cohort71616.0 (14.9–17.1)1155013.9 (13.2–14.6)1All VTE cases50097.8 (89.6–106.8)3.8 (3.4–4.3)32228.1 (25.2–31.3)1.6 (1.4–1.8) DVT344113.4 (102.0–126.1)4.4 (3.9–5.1)18627.7 (24.0–32.0)1.6 (1.4–1.9) PE15675.1 (64.2–87.9)2.9 (2.5–3.5)13628.7 (24.2–33.9)1.6 (1.3–1.9)CI, confidence interval; DVT, deep vein thrombosis; PE, pulmonary embolism; RR, relative risk; VTE, venous thromboembolism.aRelative risk estimated by Cox regression adjusted for age, sex, calendar year, consultations in the previous year, cancer, heart failure and ischaemic heart disease.Causes of death for patients dying within the first year of follow up are shown in Table 3. The main cause of death was cancer in both groups, but the percentage of patients dying from cancer was almost two-fold higher in the VTE group (56.0 vs. 29.6%). Conversely, the proportion of patients dying from CHD was approximately two-fold greater in the control cohort than the VTE cohort.Table 3Distribution of causes of death in patients dying during the 11 months of follow up in patients with VTE surviving the first month after VTE diagnosis in comparison with control cohort.Cause of deathVTE cohort (n = 500) n (%)Control cohort (n = 716) n (%)Coronary heart disease (CHD)49 (9.8)153 (21.4)Other cardiovascular and cerebrovascular diseases63 (12.6)107 (14.9)Cancer280 (56.0)212 (29.6)Other non-cardiovascular diseases (respiratory, digestive, urinary, other)47 (9.4)129 (18.0)Unknown61 (12.2)115 (16.1)DiscussionThe results of this study suggest that a first VTE episode does not increase the risk of MI. These results were similar, regardless of VTE type (DVT or PE), or whether VTE was idiopathic or secondary. However, while we did not observe a significant increase in the risk of MI following VTE (RR: 1.2 with a lower 95% CI below 1.0), the upper 95% confidence interval of 1.6 means that an increased risk of MI following VTE cannot be safely excluded on the basis of our results.These results contrast with several, much smaller, studies that have pointed towards an association between VTE and thromboembolic arterial disease. Case-control studies have reported a significantly higher prevalence of carotid plaques in patients with DVT (n = 299) [17] and a significantly higher incidence of coronary artery calcification in patients with idiopathic VTE (n = 89) compared with matched controls lacking VTE [18]. Moreover, Bova and colleagues found a significantly higher risk of arterial events in 151 patients with VTE compared with 151 controls (HR 2.9; 95% CI: 1.1–7.6) [13]. Recently, a cohort study reported the risk of MI to be increased by 60% in the first year after an episode of VTE, with a progressive decline during the subsequent 20 years [19]. As the present study includes nearly 5000 patients with VTE and a control cohort without prior VTE (n = 43 382), our conclusion that VTE is not associated with a major increased risk of subsequent MI is likely to be authoritative.The conclusions of our study contradict those of Bova and colleagues [13], but this may reflect, in part, differences in the criteria used to identify patients with VTE and the arterial events used as endpoints. Moreover the latter study was comparatively small, though to our knowledge it is the only study other than ours and the aforementioned study by Hong et al. in 89 patients with VTE [18] to compare patients with VTE with controls taken from a general population without VTE. Other studies investigating VTE as a risk factor for cardiovascular events did not include a control group without VTE. For example, two studies compared patients with idiopathic VTE with patients diagnosed with secondary VTE [1,2], and one investigated the long-term effects of 6 weeks vs 6 months of anticoagulation treatment on patients with VTE [3]. Given the large number of patients with first VTE (n = 4890) in our study, it seems unlikely that VTE is associated with a subsequent MI for patients without a history of ischaemic heart disease.In contrast to the conflicting results surrounding the association between DVT and cardiovascular disease, it is well recognised that there is an increased mortality after VTE [3]. The present study showed that a first diagnosis of VTE was associated with significantly increased mortality in those who survived the first month following the VTE event, particularly in the subsequent 11 months after VTE diagnosis. This risk was greatest in patients with DVT rather than PE, and in younger patients rather than the elderly. Previous studies have shown that mortality is highest immediately after VTE in patients with PE, and then decreases over the following year [20,21]. However, to our knowledge, few studies have investigated mortality in patients for prolonged periods after VTE. Our study therefore provides important data on long-term mortality following a VTE event in a large number of patients (n = 5801). These findings complement those from a previous study in the same cohort of patients with VTE which reported patient mortality in the first month after VTE [11].The present study showed that, during the first year after VTE, cancer was the most frequent cause of death in VTE patients surviving the first month. Cancer is a well-known risk factor for VTE and death from VTE [3,11,22-24], with the mortality seemingly independent of whether the cancer diagnosis is made before or after VTE diagnosis [24]. As such, cancer (and death because of cancer) is likely to be more common in the VTE cohort. Secondly, the data for this period may be skewed somewhat, as the 1-month period after VTE is omitted. During this 1-month period there were more deaths due to cardiovascular causes than cancer, as has been reported previously [11]. Deaths during the 1-month period after VTE were largely due to PE, rather than DVT, with the 1-month death rate of 1.4% after an episode of DVT and 22.6% after PE with or without DVT [11].The 8-year risk of death in patients surviving the first month after VTE is higher than that in patients without VTE, even after adjustment for cancer, heart failure and ischaemic heart disease. Because of the likely multiple comorbid diseases and risk factors in the predominantly elderly population with VTE in this study, we should be cautious about the reasons for this excess mortality. Thus, although the 8-year risk of death in patients surviving 1-month after VTE is higher than that in patients without VTE, it is difficult to say whether this increased risk arises from VTE itself or other underlying conditions or risk factors.This study has a number of important strengths. Patients were drawn from a large primary care database representative of the UK population and spanning a wide age range, and representing a study population that is an order of magnitude larger than previous studies investigating potential risk factors and complications of VTE [13,17,18,25-29]. Cases of VTE were classified according to whether they were DVT or PE, and also if they were idiopathic or secondary, allowing analysis of possible differences between the types of VTE. VTE cases in a random sample were also identified and validated with a confirmation rate of over 94% [11]. We excluded information bias since information was collected in the same manner for VTE and comparison patients, and information collection in VTE cases was blinded to the later occurrence of MI. Using MI, a major clinical event, as the cardiovascular endpoint was advantageous as previous studies using the GPRD have shown the validity of using codes for MI [14,30]. Limitations of the study include the fact that it only involved a UK population sample and that patients included in the analysis of MI risk may have had subclinical cardiovascular disease prior to the start date of the study. Patients may also have had VTE or MI prior to enrolment in the GPRD, which would not have been systematically recorded. In addition, the limited number of MI cases in the VTE cohort (55) means that the study is not powered to detect modest but potentially clinically important elevations in the incidence rate of MI.ConclusionIn conclusion, our data show that a diagnosis of VTE does not increase the risk of MI in comparison with a control cohort drawn from the general population of the UK. There was some suggestion that the risk of MI may be increased in elderly patients with VTE during the first year following the diagnosis of VTE, but this increase was not statistically significant. While the risk of MI is not increased, patients who survive the first month after a VTE event were significantly more likely than controls to die in the subsequent 8 years. This increased risk is particularly marked over the first year following a VTE. Further studies may thus be required to investigate the causes of death following VTE in more detail, and thus determine how best to reduce mortality in patients after a first VTE event.Competing interestsThis study was funded by a research grant from AstraZeneca R&D Mölndal, Sweden. SJ is an employee of AstraZeneca R&D Mölndal, Sweden, and MAW was an employee of AstraZeneca R&D Mölndal, Sweden at the time of the study. The corresponding author had full access to all the data in the study, and takes responsibility for the integrity of the data and the accuracy of the data analysis.Authors' contributionsCH participated in the design of the study, carried out the statistical analysis, interpreted the data and helped to draft the manuscript. SJ and MAW participated in the design of the study and interpretation of the data. LAGR conceived of the study, participated in its design, analysis and interpretation and helped to draft the manuscript. All authors critically revised the manuscript for important intellectual content, and approved the final manuscript.\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2531130.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2531130",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2531130\nAUTHORS: Cinzia Lavarino, Idoia Garcia, Carlos Mackintosh, Nai-Kong V Cheung, Gema Domenech, José Ríos, Noelia Perez, Eva Rodríguez, Carmen de Torres, William L Gerald, Esperanza Tuset, Sandra Acosta, Helena Beleta, Enrique de Álava, Jaume Mora\n\nABSTRACT:\nBackgroundNeuroblastic tumours (NBTs) represent a heterogeneous spectrum of neoplastic diseases associated with multiple genetic alterations. Structural and numerical chromosomal changes are frequent and are predictive parameters of NBTs outcome. We performed a comparative analysis of the biological entities constituted by NBTs with different ploidy status.MethodsGene expression profiling of 49 diagnostic primary NBTs with ploidy data was performed using oligonucleotide microarray. Further analyses using Quantitative Real-Time Polymerase Chain Reaction (Q-PCR); array-Comparative Genomic Hybridization (aCGH); and Fluorescent in situ Hybridization (FISH) were performed to investigate the correlation between aneuploidy, chromosomal changes and gene expression profiles.ResultsGene expression profiling of 49 primary near-triploid and near-diploid/tetraploid NBTs revealed distinct expression profiles associated with each NBT subgroup. A statistically significant portion of genes mapped to 1p36 (P = 0.01) and 17p13-q21 (P < 0.0001), described as recurrently altered in NBTs. Over 90% of these genes showed higher expression in near-triploid NBTs and the majority are involved in cell differentiation pathways. Specific chromosomal abnormalities observed in NBTs, 1p loss, 17q and whole chromosome 17 gains, were reflected in the gene expression profiles. Comparison between gene copy number and expression levels suggests that differential expression might be only partly dependent on gene copy number. Intratumoural clonal heterogeneity was observed in all NBTs, with marked interclonal variability in near-diploid/tetraploid tumours.ConclusionNBTs with different cellular DNA content display distinct transcriptional profiles with a significant portion of differentially expressed genes mapping to specific chromosomal regions known to be associated with outcome. Furthermore, our results demonstrate that these specific genetic abnormalities are highly heterogeneous in all NBTs, and suggest that NBTs with different ploidy status may result from different mechanisms of aneuploidy driving tumourigenesis.\n\nBODY:\nBackgroundNeuroblastic tumours (NBTs) are one of the most common neoplasms in childhood, accounting for approximately 40% of solid tumours encountered in the first four years of life [1]. NBTs are heterogeneous in terms of their biological, genetic and morphological characteristics and exhibit marked diverse clinical behaviours.The biological bases of these processes are poorly understood. There is an apparent link between NBTs aggressiveness and specific genetic aberrations (i.e., MYCN amplification, chromosome deletions of 1p36, 11q23, 14q32 or 19q13.3; gain of 17q and near-diploid/tetraploid DNA content), indicating that specific genetic alterations are present in individual categories of NBTs and likely contribute to clinical outcome [2-4].Abnormal cellular DNA content is ubiquitous in cancer and has been linked to the rate of cell proliferation, cell differentiation, and prognosis in a variety of tumour cell types. In contrast to most other tumours, hyperploidy confers a favourable prognosis in NBTs [5], acute lymphoblastic leukemia [6], and rhabdomyosarcoma [7]. Non-metastatic loco-regional NBTs (stages 1, 2 and 3) often show modal chromosomal numbers in the near-triploid range (58 to 80 modal chromosome number) and few structural aberrations [5]. On the other hand, karyotypes of metastatic NBTs are commonly near-diploid (44 to 57 chromosomes) or near-tetraploid (81–103 chromosomes) with structural changes [5].The presence of specific and recurrent chromosomal alterations in NBTs suggests that gene copy number abnormalities represent a major biologically relevant event, which contributes to NBT growth and survival. The aim of the current study was to gain further insight into the difference in gene expression of distinct biological entities within NBTs defined by the ploidy status.MethodsPatients and samplesForty-nine diagnostic primary NBT specimens (24 stages 1, 2, and 3; 7 stage 4s; and 18 stage 4) obtained from patients diagnosed and treated at MSKCC were selected for gene expression profiling (Table 1). Risk assessment was defined by the INSS staging classification, the MSKCC biological risk stratification criteria, and the COG clinical staging criteria. NBT stages 1, 2, 3 and 4s were treated without use of cytotoxic therapy, when possible, according to MSKCC protocols. Stage 4 NBTs patients were treated according to N5, N6 or N7 protocols. This study was approved by the MSKCC and HSJD Institutional Review Boards and informed consent was obtained before collection of all samples.Table 1Clinical and Biological characteristics of patients with Neuroblastoma evaluated according to tumour ploidy status.Case numberploidyAgeINSS stageMYCN amplificationDisease StatusSurvival Statusmicroarray analysisvalidation analysis<12m=0; >12m=11,2,3,4s=0; 4=11near-3n10NANPAYY2near-3n10NANPAY.3near-3n10NANPAYY4near-3n00NANPAY.5near-3n00NANPAYY6near-3n00NANPAY.7near-3n00NAPAY.8near-3n00NAPAY.9near-3n00NANPAY.10near-3n00NANPAY.11near-3n10NANPAYY12near-3n00NANPAY.13near-3n11NAPDY.14near-3n10NANPAY.15near-3n00NANPAYY16near-3n00NANPAY.17near-3n00NAPAYY18near-3n11NAPDYY19near-3n10NANPAY.20near-3n10NANPAYY21near-3n00NAPAY.22near-3n10NAPDYY23near-3n00NANPA.Y24near-3n00NANPA.Y25near-3n00NANPA.Y26near-3n00NANPA.Y27near-3n10NANPA.Y28near-3n10NANPA.Y29near-3n10NANPA.Y30near-3n10NANPA.Y31near-3n01NANPA.Y32near-3n00NANPA.Y33near-3n11NAPA.Y34near-3n00NANPA.Y35near-3n00NAPD.Y36near-3n00NANPA.Y37near-3n00NANPA.Y38near-2n11ANPAYY39near-2n11APDYY40near-2n10NAPDY.41near-2n11APDYY42near-2n11APAYY43near-2n11NANPAYY44near-2n11NAPDY.45near-2n11APDY.46near-2n11NANPAY.47near-2n10NAPDY.48near-2n10NAPDY.49near-2n01ANPAY.50near-2n00NANPAYY51near-2n00NAPAY.52near-2n11APDYY53near-2n10ANPAYY54near-2n10NANPAY.55near-2n11NAPDYY56near-2n00NAPAY.57near-2n00NANPAY.58near-2n11NAPDY.59near-2n11NAPDYY60near-2n10NAPDY.61near-2n00NANPA.Y62near-2n11NAPD.Y63near-2n11APD.Y64near-2n11NANPA.Y65near-2n11NAPD.Y66near-2n00ANPA.Y67near-2n11APD.Y68near-2n11NAPA.Y69near-4n10NANPAY.70near-4n01APDYY71near-4n11NANPAYY72near-4n01NAPAY.73near-4n10APD.Y74near-4n11APD.YMYCN amplification status: NA = not amplified, A = amplified. Disease status: NP = no disease progression, P = disease progression. Survival status: A = alive, D = dead. Microarray and validation analyses: Y = cases analyzed.Twenty-one samples (9 stages 1, 2, and 3; 1 stage 4s; and 11 stage 4) of the original MSKCC NBT cohort included in the gene profiling analysis and an independent set of 25 primary NBT specimens (12 stage 1, 2, and 3, 2 stage 4s, and 11 stage 4) obtained at diagnosis from 3 Spanish institutions (HSJD, Barcelona; Hospital La Paz, Madrid; and Department of Pathology, University of Valencia) were available for validation analyses (Table 1). Normal control DNA was obtained from the National DNA Bank of Spain.All tumour-specimens were evaluated by the same pathologists (WG and NP) to assess tumour cell content, only tumours with > 70% were included in the study.DNA content analysisThe modal DNA content was determined by flow cytometry DNA analysis on nuclei isolated from paraffin sections using the method of Hedley modified [8]. DNA index (DI) was expressed as the ratio of tumour DNA content/standard DNA fluorescence; near-diploid DI = 0.90–1.20; near-triploid DI = 1.21–1.75; near-tetraploid DI = 1.76–2.20.Gene expression profilingGene expression profiling was performed of 49 primary NBT samples (22 near-triploid, 23 near-diploid and 4 near-tetraploid) using Affymetrix GeneChip Human Genome U95 Set™ Arrays, as previously reported [9]. Microarray data and sample annotations have been deposited in the caArray database .Differential gene expression analysisGenes with high variability within samples were selected by pair-wise comparison analyses performed by adjusting the type-I error for multiple tests (Step-down permutation (SDP) [10], and False Discovery Rate (FDR) [11]), and with no type-I error adjustment (Raw method). The cut-off Family-wise error applied to select significant genes by means of the T-test for independent data, a univariate screening supervised procedure, was equivalent for all three methods: < 0.1, < 0.05 and < 0.01. Hierarchical clustering analyses were performed for the differentially expressed genes for all the methods of adjustment of Type-I error and cut-off of P-values, using a multivariate unsupervised method, taking into account the relationship between gene expressions. Fisher's exact test and 95% bilateral confidence interval using Wilson method were used to evaluate the proportion with which chromosomes were represented in the selected gene sets in comparison to chromosome representation within the Affymetrix GeneChip U95Av2. Statistical analyses were performed using SAS 9.1 and JMP 5.1 (SAS Institute Inc) for Windows and CIA 2.1.1.Gene Ontology annotation categoriesGene Ontology (GO) annotation categories were analyzed using explore GeneOntology (eGOn v2.0) in Gene Tools web service  to create a biological profile of the differentially expressed genes. Overrepresented GO terms were determined statistically by Fisher's exact test (P < 0.01) and adjusted FDR < 0.01.Quantitative Real-time PCR (Q-PCR)Quantification of transcript levels using Q-PCR was performed of 13 genes located on chromosomes 1 and 17 (see Additional file 1). Concomitant quantification of gene copy number was performed for a set of these genes (see Additional file 1). MYCN gene copy number was analyzed by Q-PCR, and FISH when needed. Validation analyses were performed on 46 primary NBT specimens (see patients and samples).Q-PCR reactions and quantification, using the ΔΔCT relative quantification method, were performed on an ABI Prism 7000 Sequence Detection System using TaqMan® Assay-on-Demand Gene Expression products, according to the manufacturer's protocols (Applied Biosystems, US). All experiments included no template controls and were performed in duplicate and repeated twice independently. Transcript levels were measured relative to 3 normal tissue samples (adrenal gland, lymph node and bone marrow) and normalized to TATA box binding protein (TBP), hypoxantine phosphoribosyltransferase 1 (HPRT1) and succinate dehydrogenase complex, subunit A (SDHA) expression values. Endogenous control genes were chosen on the basis of recent publications regarding accurate normalization of real-time quantitative RT-PCR in primary neuroblastoma [12,13]. These genes are reported within the most stable set of endogenous control genes. Gene copy number quantification was performed as reported previously [14]. Gene copy number was calculated relative to placental DNA using the B-Cell maturation factor (BCMA) as reference gene. The validity of BCMA as reference gene in our cohort of NBTs was determined by copy number ratio: BCMA NB tumour test sample/BCMA placenta calibrator sample. The ratio measured was equal to 1.0016; (tumour DNA 1.0012 ± 0.13 SD)/(placental DNA 0.9996 ± 0.05).Fluorescent in situ hybridization (FISH)FISH was assayed on 4 μm sections of Tissue-Micro-Array (TMA) of formalin-fixed paraffin-embedded NBT samples corresponding to the validation set, and partially matching the MSKCC series described above. Tissue microarrays included only tumour areas showing > 90% of tumour cells. Sections were washed with 2× SSC buffer and fixed in 4% paraformaldehyde in PBS. DNA-probes, CEP 17 Alpha (Ref: 32-112017;Vysis, IL, USA) LSI p53 (Ref:30-190008;Vysis) and/or LSI 1p36 (Ref:30-231004;Vysis), were denatured at 73°C, 5 min., applied to tissue sections and simultaneously denatured using the Hybridizer (DAKO) at 90°C, 4 min. Hybridization was performed for 16 h at 37°C in a humid chamber. Slides were then washed with Buffer post-hybridization (Master Diagnostica, Granada, Spain) and stained with DAPI (6-diamidino-2-phenylindole) and mounted with Vectashield H-1000 medium (Vector). One hundred nuclei were evaluated for each core. Results were recorded as percentage of nuclei present in the sample having each probe signal pattern. Cell populations < 5% of abnormal cells were not scored as significant. Microscope Magnification ×1000.Array comparative genomic hybridization (aCGH)Whole genome BAC-aCGH studies were performed using the Sanger 1 Mb clone set (kindly provided by Dr. K. Szuhai LUMC, The Netherlands). BAC/PAC clones were added to increase resolution for regions of interest: full genomic coverage clones for chromosome 17 (CHORI) and chromosome 11 (BAC/PAC isolated DNA, kindly provided by Dr. J. San Miguel, CIC, Salamanca), and 19q13 enriched medium-coverage set (Invitrogen, CA, USA and kindly provided by Dr. JC Cigudosa, CNIO, Spain). BAC DNA was extracted, amplified by DOP and Aminolinking-PCR and spotted in triplicate onto Codelink slides (Amersham Biosciences, GE, USA).Tumour and reference DNA (an equimolar DNA pool from 40 healthy donors, obtained from the Spanish National DNA Bank) was Cy5/Cy3-dCTP (Amersham, GE) labelled using a non-commercial Random Priming kit composed by Random Octamers dissolved in Eppicentre Exo-Minus Klenow buffer, a dNTPs mix depleted in dCTP and Exo-Minus Klenow enzyme (Eppiocentre). Labelled DNA was purified through Illustra G-50 Microspin Columns, mixed and then precipitated along with Cot DNA (Roche). Hybridization was performed for 48 hours at 42°C and probe excess removed.Imaging acquisition and data analysisLog2 data was acquired using Axon 4000B scanner and GenePix software. Normalization was done with GenePix software using the mean of the median of ratios of all the autosomal features in the array, excluding those removed by the quality flagging scripts. Gpr files were subsequently processed with Bioconductor packages (CRAN) incorporating scripts for removing SD > 0.2 and GenePix flagged spots. DNA copy algorithm and Merge Levels scripts (both implemented in snap CGH package) were applied for segmentation of the data. A graded colour code adjusted to the log2 rank of each individual plot was assigned to define the segments found by the applied algorithm. Universal threshold cut-off values for defining gain/loss were not applied because of subpopulation clonal heterogeneity, ploidy, and percentage of neuroblastic cells, which varied from one sample to another. Due to this, plots were evaluated independently by visual examination and results were depicted using a graded colour code adjusted to the log2 rank of each plot, assigning a colour grade to every segment found by the segmentation algorithm.ResultsDifferential gene expression analysisGene expression analysis was performed on a spectrum of 49 NBTs with varying DNA content (22 near-triploid, 23 near-diploid and 4 near-tetraploid). Owing to reduced number of near-tetraploid cases included in this study and taking into account the reported biological and clinical similarities with near-diploid NBTs [15,16], near-diploid and near-tetraploid NBTs were combined in one group. Pair-wise comparison analyses of near-triploid (n = 22) versus near-diploid/tetraploid (n = 27) NBTs revealed small sets of differentially expressed genes when using a stringent correction for multiple sampling, (6 genes [FDR < 0.01] and 12 genes [SDP < 0.1]) (see Additional file 2). Interestingly, all genes showing a higher expression in the near-triploid group mapped to chromosome 17 (see Additional file 2). Less stringent multiple testing corrections selected a larger set of differentially expressed genes, (51 genes [FDR < 0.05] (Fig. 1) and 254 genes [FDR < 0.1] (see Additional file 2). Again, this resulted in a statistically significant proportion of genes mapping to chromosomes with described recurrent abnormalities in NBTs; chromosome 1 (p = 0.01) and chromosome 17 (p < 0.0001) (Fig. 1). Chromosomal region specificity was observed since the majority of chromosome 1 and 17 differentially expressed genes spread over 1p36-p22.1 and 17p13-17q21 (Fig. 1; see Additional file 2). The majority showed higher expression in near-triploid NBTs; 92% (CI: 78% to 97%) of chromosome 1 genes and 91% (CI: 76% to 96%) of chromosome 17 (see Additional file 2). Only 8% (CI: 2% to 21%) probe sets for genes located on chromosome 1, ENO1 (1p36.2), CCT3 (1q23) and C1orf107 (1q32.2), and 9% (CI: 3% to 23%) for genes on chromosome 17, MAC30 (17q11.2) and NME1 (17q21.3), showed a higher expression within near-diploid/tetraploid NBTs.Figure 1A heatmap illustrating the distinct expression profiles of 49 NB primary tumours with varying ploidy status. Gene expression profiles visualized according to 51 differentially expressed genes [FDR < 0.05]. (Right) Gene dendrogram is divided in 2 main gene clusters. Top cluster: genes displaying higher expression in near-triploid tumours; a statistically significant proportion of genes map to chromosome 1 (p = 0.01) and chromosome 17 (p < 0.0001) (Blue). Bottom cluster: genes with higher expression in near-diploid/tetraploid NBTs. (Bottom) Filled in boxes: Ploidy: black = near-diploid, empty white boxes = near-triploid, grey = near-tetraploid NBTs; MYCN: black = amplified, white = not amplified; Age: black > 12 months, white < 12 months; INSS: black = Stage 4 NBTs, white = stages 1, 2, 3, and 4S.The Gene Ontology biological profile of genes with higher expression in near-diploid/tetraploid NBTs showed enrichment for genes related to protein, macromolecular and nucleic acid biosynthesis, such as, NME1, ATP5I, ATP5C1, NME4, TYMS and GMPS. Whereas, near-triploid tumours included genes involved in vesicle mediated transport, cell communication, signal transduction, nervous system development and regulation of small GTPase mediated signal transduction. A large portion of these genes mapped to chromosomes 1 and 17 (60–100%), among these RERE, CHD5, CLCN6, CDC42BPA, NTRK1, ARHGEF11, PMP22, VAMP2, GARNL4, MAP2K4 and FLOT2.Quantitative Real-time Polymerase Chain Reaction (Q-PCR)Quantification of transcript levels of 13 differentially expressed genes, located mainly on the chromosomal regions 1p36 and 17p13-q21, was performed on two separate groups of NBT specimens: 21 primary NBTs from the original MSKCC cohort as well as on an independent set of 25 NBTs (Table 1). Expression levels identified by Q-PCR confirmed the microarray data in both sets of NBTs (Fig. 2A, B and 2C).Figure 2Quantitative real-time PCR validation of microarray gene expression data. Comparison of gene expression levels of 5 representative genes located on chromosomes 1 and 17. A. Microarray gene expression data in 49 NBT from MSKCC. Gene expression data were log-transformed and normalized to TBP expression levels; B. Q-PCR gene transcript quantification in 21 NBTs from MSKCC; C. Q-PCR gene transcript quantification in 25 NBTs from Spanish institutions. Results were compared by two-tailed independent-sample t test using SPSS v.14.0 for Windows (SPSS, Chicago, IL). Expression data are shown as box plots (SPSS v.14.0).Four genes located on chromosomes 1 and 17 were further analyzed for gene copy number by DNA Q-PCR analysis in 27 cases (Tables 2 and 3; see Additional file 3). Near-triploid NBTs (n = 13) showed, both for chromosome 1 and 17, fold values consistently higher (≥ 1.3-fold) than normal reference gene values, and were considered to represent a minimum trisomic gene copy number. Only case # 2 (Table 2; see Additional file 3) showed 0.8–1.1-fold values reflecting a possible loss of 1p36, subsequently confirmed by FISH and aCGH results. Near-diploid/tetraploid NBTs (n = 13) displayed a wider range of values (0.5–2.7-fold), indicative of losses and gains within a more heterogeneous clonal population, as shown by FISH results. Tumour clonal heterogeneity may often confound analyses performed on the bulk of the tumour specimen and could explain some discrepancies between ploidy and gene copy number.Table 2Results of FISH, aCGH, Q-PCR analyses of chromosome 1, displayed in relation to NBTs ploidy statusCase NumberPloidyMYCNFISH Chromosome 1a CGH Chr. 1Q-PCR Gene copy No. (fold change)Disease StatusSurvival StatusCell % (#DNA probe signals: LSI 1p36: LSI 1q25)pcenqGNB1 (1p36.33)RERE (1p36.1)1near-3nNAn.eGGG1.63.2NPA2near-3nNA50 (2:2), 20 (3:3), 15 (1:3), 15 (2:3)L--0.81.1NPA3near-3nNA60 (2:2), 40 (3:3)---1.52NPA4near-3nNA5 (2:2), 95 (3:3)GGG2.62.4NPA5near-3nNA40 (2:2), 60 (3:3)---1.41.3NPA6near-3nNA50 (2:2), 50 (3:3)GGG1.43NPA7near-2nAn.e.L--0.52.2PD8near-2nNAn.e.GGG1.61.5NPA9near-2nNA95 (2:2), 5 (3:3)n.en.en.e0.70.5NPA10near-2nNA100 (2:2)---0.70.5PD11near-2nNA35 (2:2), 65 (1:3)n.en.en.e12.3PD12near-4nA51 (1:2), 30 (2:2), 19 (1:3)--G0.50.6PD13near-4nA60 (2:2), 30 (3:3), 10 (4:4)---1.32.7PDThirteen representative cases drawn from the HSJD cohort analyzed by FISH, aCGH and Q-PCR of chromosome 1. n.e = not evaluable results. MYCN amplification status: NA = not amplified, A = amplified. Disease status: NP = no disease progression, P = disease progression. Survival status: A = alive, D = dead. FISH: results are displayed as percentage of cells exhibiting the observed number of DNA probe signals, and exact number of signals for the DNA probes used: chromosome 1 (LSI 1p36 and LSI 1q25 DNA probes) and chromosome 17(LSI 17p13.1 and CEP 17 DNA probes). Array CGH: p and q = chromosome arms, cen. = centromeric; G = chromosome gain, L = chromosome loss. Q-PCR: gene copy number fold changes are determined by the ΔΔCT relative quantification method.Table 3Results of FISH, aCGH, Q-PCR analyses of chromosome 17, displayed in relation to NBTs ploidy statusCase NumberPloidyMYCNFISH Chromosome 17a CGH Chr. 17Q-PCR Gene copy No. (fold change)Disease StatusSurvival StatusCell % (# DNA probe signals: LSI 17p13.1: CEP 17)pcenqRUTBC1 (17p13.3)NME1 (17q21)1near-3nNAn.eGGG2.32.5NPA2near-3nNA45 (2:2), 55 (3:3)GGG1.51.5NPA3near-3nNA45 (2:2), 55 (3:3)GGG1.31.4NPA4near-3nNA30 (2:2), 70 (3:3)GGG3.62.2NPA5near-3nNA50 (2:2), 50 (3:3)GGG1.61.3NPA6near-3nNA50 (2:2), 50 (3:3)GGG1.41.5NPA7near-2nNA5 (1:1), 80 (2:2), 10 (3:3), 5 (4:4)n.en.en.e0.71.4NPA8near-2nNA33 (1:1), 66 (2:2)---0.80.7PD9near-2nA7 (1:1), 7 (2:1), 60 (2:2), 20 (1:2), 6 (2:3)--G1.11.4PD10near-2nNA80 (2:2), 15 (2:3), 5 (3:3)--G1.21.5NPA11near-2nNA28 (1:1); 11 (2:1), 56 (2:2), 5 (3:2)n.en.en.e1.10.9PD12near-4nA45 (2:2); 55 (3:3)--G11.4PD13near-4nA45 (2:2), 45 (3:3), 10 (4:4)GGG2.21.9PDThirteen representative cases drawn from the HSJD cohort analyzed by FISH, aCGH and Q-PCR of chromosome 17. n.e = not evaluable results. MYCN amplification status: NA = not amplified, A = amplified. Disease status: NP = no disease progression, P = disease progression. Survival status: A = alive, D = dead. FISH: results are displayed as percentage of cells exhibiting the observed number of DNA probe signals, and exact number of signals for the DNA probes used: chromosome 1 (LSI 1p36 and LSI 1q25 DNA probes) and chromosome 17(LSI 17p13.1 and CEP 17 DNA probes). Array CGH: p and q = chromosome arms, cen. = centromeric; G = chromosome gain, L = chromosome loss. Q-PCR: gene copy number fold changes are determined by the ΔΔCT relative quantification method.Comparison between DNA gene copy number and expression levels (Fig. 3) revealed an overall linear correlation for those analyzed genes that displayed in the microarray analysis higher expression levels in near-triploid NBTs. Conversely, NME1 gene, as from microarray results, showed low expression values, closer to the disomic reference sample expression, in near-triploid NBTs, and high fold increase in mRNA levels in near-diploid and tetraploid cases.Figure 3Comparison between DNA copy number and gene expression levels analyses. Gene expression levels and gene copy number are exhibited as mean values in accordance with NBT ploidy subgroups. Correlation between DNA gene copy number and expression levels was observed in those analyzed genes that displayed in the microarray analysis higher expression levels in near-triploid NBTs.Fluorescent in situ hybridization (FISH)Interphase FISH using the DNA probes LSI 1p36 and LSI 1q25 was performed on 13 primary NBTs drawn from the HSJD cohort; four cases were not evaluable (Table 2). According to chromosome 1 status, near-triploid and near-diploid/tetraploid NBTs were characterized by intratumoural heterogeneous cell population content. Only 1 case showed uniform distribution of probe signals within cells of the tumour specimen (case #10, Table 2). All but one of the near-triploid NBTs were constituted of clonal populations with two LSI 1p36 and LSI 1q25 signals (2:2) and/or three (3:3) DNA probe signal, ranging from 40–60% and 40–100% of the cells, respectively. Case # 2 was the only near-triploid NBT that exhibited a chromosome 1p36 loss in 30% of cells, confirmed by aCGH and Q-PCR. Even higher intratumoural heterogeneity was observed in near-diploid/tetraploid NBTs.Chromosome 17 FISH using centromeric CEP 17 and LSI p53 (17p13.1) DNA probes, was performed on 53 primary NBTs (13 cases from the HSJD cohort, Table 3, and 40 cases from MSKCC, Table 4). Based on chromosome 17 status, near-triploid tumours were constituted of two (2 CEP 17 and 2 LSI p53 signals, 2:2), three (3:3) and four (4:4) chromosome 17 signals clonal populations that ranged from 10–55%, 24–70% and 7–45% of the cells, respectively. Near-diploid/tetraploid NBTs were composed by a more heterogeneous cell population, with a high incidence of chromosomal structural abnormalities. In a large portion of these tumours, alongside with the two (2:2) DNA probe signal clonal populations (6%–100% of cells), the aneuploid cell population counterpart constituted a significant and heterogeneous portion of cell population (Tables 3 and 4).Table 4Chromosome 17 Fluorescence in situ Hybridization results of 40 NBTs obtained from MSKCC, displayed in relation to NBTs ploidy statusCase NumberPloidyMYCNFISH Chromosome 17Disease StatusSurvival StatusCell % (# DNA probe signals: LSI 17p13.1: CEP 17)1near-3nNA23 (2:2), 44 (3:3), 33 (4:4)NPA2near-3nNA50 (2:2), 50 (3:3)NPA3near-3nNA16 (2:2), 41 (3:3), 43 (4:4)NPA4near-3nNA34 (2:2), 42 (3:3), 24 (4:4)NPA5near-3nNA33 (2:2), 50 (3:3), 17 (4:4)PA6near-3nNA25 (2:2), 60 (3:3), 15 (4:4)NPA7near-3nNA31 (2:2), 46 (3:3), 23 (4:4)NPA8near-3nNA35 (2:2), 52 (3:3), 13 (4:4)NPA9near-3nNA23 (2:2), 54 (3:3), 23 (4:4)NPA10near-3nNA13 (3:3), 66 (3:4), 21 (5:5)NPA11near-3nNA16 (2:2), 48 (3:3), 36 (4:4)PA12near-3nNA35 (2:2), 58 (3:3), 7 (4:4)NPA13near-3nNA46 (2:2), 24 (3:3), 30 (4:4)NPA14near-3nNA22 (3:3), 62 (4:4), 16 (4:5)PD15near-3nNA10 (2:2), 29 (3:3), 45 (4:4), 16 (5:5)PD16near-2nNA5 (1:1), 65 (2:2), 5 (1:2), 10 (3:3), 5 (2:3), 5 (4:4), 5 (3:4)PD17near-2nNA100 (2:2)PD18near-2nNA95 (2:2), 5 (3:3)PD19near-2nNA31 (CEP 2), 50 (CEP 3), 18 (CEP 4)NPA20near-2nNA25 (1:1), 75 (2:2)PD21near-2nNA100 (2:2)PD22near-2nNAn.ePD23near-2nNA10 (CEP 1), 40 (CEP 2), 38 (CEP 3), 12 (CEP 4)PA24near-2nNA80 (2:2), 10 (1:2), 5 (3:3), 5 (2:2)PA25near-2nNA100 (2:2)PD26near-2nNA25 (1:1), 75 (2:2)PD27near-2nNA100 (2:2)PD28near-2nNA5 (2:1), 74 (2:2), 5 (1:2), 5 (3:3), 6 (2:3), 5 (4:4)PD29near-2nNA10 (2:1), 70 (2:2), 15 (1:2), 5 (3:3)PD30near-2nNAn.ePD31near-2nA20 (2:2), 32 (3:3), 12 (4:3), 20 (4:4), 16 (3:4)NPA32near-2nA40 (1:1), 60 (2:2)NPA33near-2nAn.eNPA34near-2nA100 (2:2)PA35near-2nA35 (2:2); 5 (3:2), 20 (3:3), 5 (2:3), 30 (4:4), 5 (3:4)PD36near-2nA60 (2:2); 5 (3:2), 25 (3:3), 5 (4:3), 5 (4:4)PD37near-2nA10 (1:1), 90 (2:2)PD38near-4nNA29 (2:2), 6 (3:3), 8 (4:3), 37 (4:4), 20 (3:4)NPA39near-4nNA49 (2:2), 37 (3:3), 9 (2:3), 5 (3:4)NPA40near-4nNA6 (2:2), 20 (3:3), 5 (4:3), 46 (4:4), 18 (5:5), 5 (4:5)PAn.e = not evaluable results. MYCN amplification status: NA = not amplified, A = amplified. Disease status: NP = no disease progression, P = disease progression. Survival status: A = alive, D = dead. FISH: results are displayed as percentage of cells exhibiting the observed number of DNA probe signals, and exact number of signals for the DNA probes used: chromosome 1 (LSI 1p36 and LSI 1q25 DNA probes) and chromosome 17(LSI 17p13.1 and CEP 17 DNA probes). Array CGH: p and q = chromosome arms, cen. = centromeric; G = chromosome gain, L = chromosome loss. Q-PCR: gene copy number fold changes are determined by the ΔΔCT relative quantification method.Intratumoural clonal heterogeneity was observed in all the FISH analyses (Fig. 4).Figure 4FISH analysis. Intra-tumoural cell heterogeneity, cancer cells exhibit different alterations of chromosome 17. FISH analysis using probes for chromosome 17 (red, LSI p53; green, CEP 17) showing different cellular populations within the same NBT in terms of probe signal numbers. In the panels are reported two representative NBT cases; A. near-triploid NBT; B. near-diploid case. Five signal cells in this sample were very rare populations (< 5%) and are not displayed in Table 3.Array comparative genomic hybridization (aCGH)Genome array CGH was performed for 13 cases, drawn from the HSJD validation set of NBTs, with complete FISH and Q-PCR analyses (Tables 2 and 3; Fig. 5). Near-triploid NBTs exhibited the highest incidence of specific chromosomal alterations, with consistent gain or loss of whole chromosomes, being chromosomes 7 and 17 the most frequently gained (83% and 100% cases, respectively), whilst, chromosomes 3, 4, 9, 14, 16 (50% cases), and 19 (67% NBTs) were among the most frequently lost, although the set of cases is not large enough for statistically significant results. Chromosome 1p loss was observed only in one case (case# 2, Table 2), a near-triploid stage 4s tumour.Figure 5Array-Comparative Genomic Hybridization (aCGH) results of 13 NBTs obtained from HSJD. Results are displayed according to tumour ploidy status. Chromosome alterations are visualized as a graded colour code adjusted to the log2 rank of each individual plot assigned to define chromosomal segment alterations. Filled boxes: from orange to pink colour shades represent increasing chromosomal copy number gains, whereas, from light blue to dark blue colour shades indicate chromosome losses. White colour boxes represent no detected chromosome change. Grey colour boxes represent not evaluable results.Specific near-diploid/tetraploid copy number alterations were characterized by a more heterogeneous pattern of chromosomal aberrations than those of near-triploid, being partial chromosomal segment alterations much more frequent than in near-triploid tumours (Fig 5; see Additional file 4). Partial loss of 11q and partial gain of 17q were only observed in near-diploid/tetraploid samples and never in near-triploid NBTs. Chromosome 20 showed a common pattern being one of the most frequent gains both in near-diploid and near-triploid NBTs. MYCN amplification was absent in near-triploid cases and shared by near-diploid/tetraploid cases.Further copy-number alterations that did not reach the maximum log2 values, but were clearly distinguishable in terms of segmentation algorithm, were detected in the array CGH plots and could reflect higher intratumoural clonal heterogeneity (data not shown).DiscussionAneuploidy is ubiquitous in cancer and has been linked to cell proliferation, cell differentiation and prognosis. The karyotypes of most tumours are aneuploid, meaning that chromosomes, which carry thousands of genes, are structurally rearranged, duplicated, broken or entirely missing.Gain of chromosome 17 is one of the most frequent genetic abnormalities observed in NBTs, and may involve either the entire chromosome or partial gain of the distal segment 17q21-qter [17]. Unbalanced translocations, characteristic of near diploid NBTs or tumours with structural rather than numerical chromosome aberrations, are thought to arise from DNA double strand breaks repaired erroneously, suggesting an impaired DNA maintenance or repair pathway [18]. On the other hand, abnormalities in the mitotic segregation of chromosomes are thought to underlie the numerical aberrations characteristic of near-triploid, good prognostic, NBTs. Both mechanisms define the type of aneuploidy behind each of the subgroups of NBTs, determining the kind of genetic aberrations as well as the biological behaviour of each NBT subtype.Gene expression profiling of NBTs with different ploidy status, near-triploid or near-diploid/tetraploid, enabled us to identify distinct expression profiles associated with each subgroup. Interestingly, a statistically significant proportion of genes shown to be differentially expressed mapped to chromosomes described to be recurrently altered in NBTs, chromosomes 1 and 17 [17]. Chromosomal region specificity was also observed for these differentially expressed genes since the majority spread predominantly over the chromosomal regions 1p36-p22.1 and 17p13-17q21. Besides, over 90% of these genes displayed higher expression levels in near-triploid tumours. Only two genes mapping to chromosome 17, MAC30 and NME1, exhibited a higher expression level in near-diploid/tetraploid NBTs. MAC30 gene encodes for a meningioma-associated protein, highly expressed in several types of tumours, but, with unknown clinicopathological and biological significance. The product of the NME1 gene, the nm23A protein, is a nucleoside diphosphate kinase, whose expression has been related to cell proliferative activity [19]. Whereas reduced expression of NME1 is associated with a high potential for metastasis in some tumour types, like breast cancer and melanoma, its expression is increased in aggressive NBTs [20].Genome array CGH, together with FISH and Q-PCR results, confirmed the association of specific chromosomal abnormalities with each of the NBTs subgroups. Therefore, it is not unreasonable to assume that these specific chromosomal alterations are associated with the observed gene expression profiles. The highly significant and strikingly persistent chromosomal localization of the differentially expressed genes made us hypothesize about which transcriptional regulation mechanisms can underlie these gene expression patterns. As a result of aneuploidy, cells possibly produce imbalanced expression of large sets of genes that are amplified or lost. Such gross imbalances would inevitably disrupt critical cellular circuits and destabilize regulatory pathways and cellular structures. It has been assumed that gene dosage effects may play a role in the pathogenesis of malignant diseases. Variations of the transcriptome due to alterations of the gene dosage have been described in vitro [21], in vivo [22] and in human pathologies such as trisomies 13 and 21 [23]. In our hands, when comparing gene expression levels with gene copy number of a set of differentially expressed genes located at chromosomes 1p36 and 17q13-q21, we observed a concordance between copy number and mean expression values in all those analyzed genes that displayed in the microarray analysis higher expression levels in near-triploid NBTs. In contrast, NME1 gene, as from microarray results, showed low expression values, close to the disomic reference sample expression, in near-triploid NBTs, and high fold increase in mRNA levels in near-diploid/tetraploid cases. NME1 gene has been identified as one of the MYCN targets. Correlation between MYCN overexpression and upregulation of NME1 expression has been reported both in NBTs and neuroblastoma cell lines [24]. In our experience, all MYCN amplified NBTs, displaying MYCN overexpression, as well as near-diploid cases with increased copy number of chromosome 17q, showed high NME1 expression levels. However, NME1 overexpression was also observed in 2 near-diploid MYCN single copy cases, with low MYCN expression and no 17q gain. This suggests that in NBTs NME1 gene expression is only partly dependent on gene copy number and MYCN expression, and therefore implies the existence of other mechanisms of NME1 transcriptional regulation.Recently, we reported that clonal ploidy heterogeneity is present in virtually every single loco-regional, near-triploid NBT, and detected the existence of clonal DNA content heterogeneity and evolution [25,26]. In this report our results underscore the clonal heterogeneity of all NBTs, with a marked complexity in the near-diploid/tetraploid tumours. Furthermore, clonal variations reflected in the array CGH plots as copy-number alterations with varying log2 values, could unveil the presence of subpopulations emerged during tumour development. These cellular subpopulations are likely to be the cause of the high cell heterogeneity also observed in the FISH analyses. These findings are important in emphasizing the cellular heterogeneity and karyotypic complexity (aneuploidy) generally associated with malignant tumours, but need a more detailed understanding of their significance.ConclusionWe have found that NBTs with different cellular DNA content display specific transcriptional profiles suggesting that near-diploid/tetraploid and near-triploid NBTs result from two different mechanisms of aneuploidy driving tumourigenesis. A large number of the differentially expressed genes participate in cell differentiation pathways and map to specific chromosomal regions recurrently involved in unbalanced translocations, gains and losses in NBTs. Our results demonstrate that these specific genetic abnormalities are complex, heterogeneous, and translate into a gene expression profile that defines the biological behaviour of each type of NBT.AbbreviationsNBTs: neuroblastic tumours; MIBG: Meta-iodobenzylguanidine; LOH: loss of heterozygosity; MSKCC: Memorial Sloan-Kettering Cancer Center, New York; HSJD: Hospital Sant Joan de Déu, Barcelona; Children's Oncology Group: COG; CT: computed tomography; INSS: International Neuroblastoma Staging System; INPC: International NB pathology committee; CNS: central nervous system; Q-PCR: Quantitative real-time polymerase chain reaction; aCGH: array-Comparative Genomic Hybridization; FISH: Fluorescence in situ hybridization.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsCL and JM are responsible for the initial conception and overall hypothesis of this study. CL, IG and JM are responsible for the design of this manuscript, including the original draft and subsequent revisions and design of this manuscript. CdT assisted with the initial concept and was involved with the draft and revisions of this manuscript; provided guidance for many of the experiments. NKC and WLG are responsible for the procurement and cryopreservation of NBT tissue specimens derived from MSKCC. ER, IG, SA, HB and JM were responsible for the procurement and cryopreservation of NBT tissue specimens derived from the Spanish institutions. WLG and NP evaluated all tumour specimens for tumour staging classification and to assess tumour content. CL, NKC, WLG, and JM are responsible for patient clinico-biological database management and for microarrays studies. NKC and WLG were involved in the drafting and revision of this manuscript. IG and CL are responsible for the quantitative PCR experiments. CM and EdA are responsible for the FISH and aCGH analyses and were also involved with the interpretation of data, draft and revision of this manuscript. ET performed the flow cytometry DNA analysis. CL, GD, JR and IG performed the statistical analysis and interpretation of the data derived from all the samples. HB and SA assisted with valuable technical assistance for experiments associated with this manuscript. All were also involved in the drafting and revisions for this manuscript. All authors read and approved the final manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:Supplementary MaterialAdditional file 1Quantitative Real-time Polymerase Chain Reaction Analysis. List of genes analyzed to determine expression levels and DNA copy number of genes located on chromosomes 1 and 17.Click here for fileAdditional file 2Gene expression profiling of NBTs with different ploidy status. List of differentially expressed genes identified applying different multiple testing corrections. Differentially expressed genes are displayed according to tumour ploidy and chromosomal location. A. List of 6 differentially expressed genes [FDR P < 0,01]; B. List of 12 differentially expressed genes [SDP P < 0,1]; C. List of 51 differentially expressed genes [FDR P < 0,05];.D. List of 254 differentially expressed genes [FDR P < 0,1].Click here for fileAdditional file 3Quantitative Real-time Polymerase Chain Reaction gene copy number analysis and array CGH analysis results. n.e = not evaluable results; n.d. = not done. MYCN amplification status: NA = not amplified, A = amplified. Disease status: NP = no disease progression, P = disease progression. Survival status: A = alive, D = dead. Q-PCR: gene copy number fold changes are determined by the ΔΔCT relative quantification method. Array CGH: p and q = chromosome arms, cen. = centromeric; G = chromosome gain, L = chromosome loss; - = no alteration observed.Click here for fileAdditional file 4Array CGH images of NBT with different DNA content. A. Near-triploid NBT; B. Near-diploid tumour and C. Near-tetraploid NBT.Click here for file\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2531234.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2531234",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2531234\nAUTHORS: Zoe T. Richards, Madeleine J. H. van Oppen, Carden C. Wallace, Bette L. Willis, David J. Miller\n\nABSTRACT:\nBackgroundCoral reefs worldwide face a variety of threats and many coral species are increasingly endangered. It is often assumed that rare coral species face higher risks of extinction because they have very small effective population sizes, a predicted consequence of which is decreased genetic diversity and adaptive potential.Methodology/Principal FindingsHere we show that some Indo-Pacific members of the coral genus Acropora have very small global population sizes and are likely to be unidirectional hybrids. Whether this reflects hybrid origins or secondary hybridization following speciation is unclear.Conclusions/SignificanceThe interspecific gene flow demonstrated here implies increased genetic diversity and adaptive potential in these coral species. Rare Acropora species may therefore be less vulnerable to extinction than has often been assumed because of their propensity for hybridization and introgression, which may increase their adaptive potential.\n\nBODY:\nIntroductionCorals of the genus Acropora are the dominant reef-builders throughout the Indo-Pacific and, although hybridization is thought to have been an important factor in their evolutionary success [1], there are few unambiguous examples of hybrids or hybrid species. In the Caribbean, where only three extant Acropora species are known, A. prolifera is the product of hybridization between the other two Acropora spp. [2], [3]. The low species complexity of the Caribbean coral fauna greatly simplifies unraveling such relationships. By contrast, the extraordinary species-richness of the Indo-Pacific, where over 60 Acropora species may occur in sympatry [Wallace & Muir, unpublished], greatly complicates the identification of hybrids.Allele sharing between species provides evidence for introgressive hybridization [4], [5], but the unknown age of many extant Indo-Pacific species [6] makes it often difficult to distinguish between hybridization and incomplete lineage sorting (i.e. shared ancestral polymorphism) [5], [7]. For the common species on which most work to date has focused, effective (Ne) and census population sizes (N) and coalescence times are unknown but potentially large and long, respectively, therefore incomplete lineage sorting cannot be ruled out. Rare species can provide new insights into the evolution of reef corals due to their intrinsically limited population sizes and therefore very short coalescence times.\nAcropora species typically occupy reef flat, reef crest and upper reef slope habitats (i.e. 2–30 m), however, some rare species occur outside this range, and this suggests an intriguing possibility-that some rare corals may be hybrids that can occupy atypical or non-parental niches, as is the case for the Caribbean hybrid species A. prolifera\n[3]. To address to address the question of whether rare Indo-pacific Acropora species might also be hybrids, we analyzed DNA sequence data from nuclear and mitochondrial loci in a range of rare and common Acropora species from the Indo-Pacific and Caribbean.Materials and MethodsSample collection and census dataSamples (n = 1–3 individuals per species) of 14 rare and 8 common Indo-Pacific species of Acropora (Table 1) were collected from the Great Barrier Reef (Palm Island Group), the Marshall Islands (Rongelap Atoll) and Papua New Guinea (Kimbe Bay). Skeletal and matching tissue samples were collected from all corals sampled (n = 102 corals). Each sample was initially identified by Richards and confirmed by Wallace. All samples used for molecular analyses have matching voucher specimens registered in the World Wide Acropora Collection at the Museum of Tropical Queensland (www.mtq.qld.gov.au). Voucher specimens are available for inspection on request from the museum. For the purpose of this paper, rare species are those which have been recorded at <2.5% of sites for which data are available in the World Wide Acropora Database (which contains >20,000 records for >1,500 sites). Mean (±SE) global census sizes were estimated by >multiplying the mean global reef area available to each species by its mean local abundance per unit area (Supplementary Table S1). Mean global reef area was calculated as the sum of the mean regional reef habitat available for all regions in which each species is known to occur (Supplementary Table S2,). The proportion of reefs and sites occupied by rare species was estimated to be 10–30% of total reef area available. For present purposes, the effective population sizes were assumed to be approximately 11% of the calculated mean global census sizes (Supplementary Methods S1); this relationship is based on a comprehensive meta-analysis of data for 102 species of animals [8].10.1371/journal.pone.0003240.t001Table 1Biological characteristics of species included in this study.SpeciesDistributionRangeEcological nicheCollection location or source\nA. walindii\nRestrictedPNGdeep sandy reef slopesKimbe Bay, PNG\nA. rongelapensis\nRestrictedMicronesia/Indonesiadeep protected sandy slopesRongelap Atoll, RMI\nA. loisetteae\nRestrictedMalaysia, W. Aust, Micronesiaprotected sandy lagoonsRongelap Atoll, RMI\nA. pichoni\nRestrictedPNG, Micronesiadeep submerged shelf reefs, shipwrecksKimbe Bay, PNG\nA. lokani\nRestrictedSE Asiashallow reef flatKimbe Bay, PNG\nA. derawanensis\nRestrictedSE Asiaprotected deep sandy slopesKimbe Bay, PNG\nA. tenella\nRestrictedSE Asiasubtidal protected slopes, shelfsKimbe Bay, PNG\nA. batunai\nRestrictedIndonesia, PNGsubmerged reefs, slopesKimbe Bay, PNG\nA. chesterfieldensis\nRestrictedChesterfield Is., Micronesiasubmerged shallow reefsRongelap Atoll, RMI\nA. kimbeensis\nRestrictedPNG, Micronesiasubmerged reef flatKimbe Bay, PNG\nA. spathulata\nRestrictedGBR, PNGreef flat and slopeOrpheus Island, GBR\nA. kirstyae\nRestrictedIndonesia, GBR, PNG, New Caledoniaprotected interrefal locationsOrpheus Island, GBR\nA. papillare\nRestrictedW. Australia, GBR, Japanultra shallow and exposed reefOrpheus Island, GBR\nA. speciosa\nRestrictedSE Asia, GBR, Central Pacificsubtidal, protected slopes and wallsRongelap Atoll, RMI\nA. jacquelineae\nRestrictedIndonesia, PNGreef slopes and submerged reefsKimbe Bay, PNG\nA. caroliniana\nRestrictedSE Asia-Pacificsubmerged habitatsKimbe Bay, PNG\nA. tortuosa\nRestrictedCentral Pacificsubtidal, protected sandy lagoonsRongelap Atoll, RMI\nA. granulosa\nWidespreadIndo-Pacificreef slopes and wallsRongelap Atoll, RMI\nA. vaughani\nWidespreadIndo-Pacificprotected subtidal habitatsOrpheus Island, GBR\nA. pulchra\nWidespreadIndo-Pacificintertidal or shallow subtidalvan Oppen et al. 2001\nA. aspera\nWidespreadIndo-Pacificintertidal or shallow subtidalvan Oppen et al. 2001\nA. longicyathus\nWidespreadSE Asia-Pacificsubtidal habitatsvan Oppen et al. 2001\nA. loripes\nWidespreadIndo-Pacificsubtidal shallow reef habitatsRongelap Atoll, RMI\nA. gemmifera\nWidespreadIndo-Pacificintertidal or shallow subtidalvan Oppen et al. 2001\nA. microphthalma\nWidespreadIndo-Pacificsubtidal habitatsOrpheus Island, GBR\nA. millepora\nWidespreadIndo-Pacificintertidal or shallow subtidalvan Oppen et al. 2001\nA. digitifera\nWidespreadIndo-Pacificintertidal or shallow subtidalvan Oppen et al. 2001\nA. humilis\nWidespreadIndo-Pacificintertidal or shallow subtidalvan Oppen et al. 2001\nA. austera\nWidespreadIndo-Pacificshallow subtidal habitatsvan Oppen et al. 2001\nA. cerealis\nWidespreadIndo-Pacificshallow subtidal habitatsvan Oppen et al. 2001\nA. nasuta\nWidespreadIndo-Pacificshallow subtidal habitatsvan Oppen et al. 2001\nA. valida\nWidespreadIndo-Pacificshallow subtidal habitatsMagnetic Island, GBR\nA. palmata\nOutgroupAtlantic Oceansubtidal habitatsvan Oppen et al. 2000\nA. prolifera\nOutgroupAtlantic Oceansubtial habitatsvan Oppen et al. 2000\nA. cervicornis\nOutgroupAtlantic Oceansubtidal habitatsvan Oppen et al. 2000\nI. cuneata\nOutgroupIndo-Pacificsubtidal habitatsvan Oppen et al. 2001DNA Extraction, Polymerase Chain Reaction, cloning and sequencingDNA was extracted from ∼1 cm branch fragments of individual corals as previously described [5]. Markers studied were the highly polymorphic single-copy nuclear Pax-C 46/47 intron and the mitochondrial DNA (mtDNA) control region, for which a reference body of data exists from various Acropora species [5], [9]. Details of primers and procedures for PCR, cloning and sequencing are described in [5], [9]. New sequences obtained have been lodged in GenBank under EU918202-EU918288 (mitochondrial data) and EU918771-EU918925 (nuclear intron data).Phylogenetic AnalysisSequences were manually aligned in Sequencher 4.5 against a subset of the existing Acropora Pax-C intron and mitochondrial control region sequences [2], [5], [9] before phylogenetic analysis in a Bayesian statistical framework in Mr Bayes 3.1.2 [10]. The dataset analysed therefore consisted of sequences from 17 rare and 15 common Indo-Pacific species Acropora species, the three Caribbean Acropora species and Isopora cuneata. Genetic distances were calculated as Kimura 2-parameter distances [11]. The optimal model of sequence evolution was identified using hierarchical likelihood ratio tests in MrModeltest 2.2 [12]. The (MCMC) analyses were run for 5 million generations, with burn-in times of 20,000–50,000 (p<0.05). Trees generated from the Pax-C data were rooted using sequences from Isopora cuneata, whereas the mtDNA tree was rooted with A. cervicornis as in this case the degree of divergence of the I. cuneata sequence effectively precluded unambiguous alignment. Analyses were conducted on the full alignments as the exclusion or weighting down of large indels or repeat regions was found not to significantly effect the overall topology (see also [5]).ResultsAllele/haplotype data from nuclear and mitochondrial loci were determined for 17 rare and 15 common Indo-Pacific Acropora species as well as all 3 Caribbean species of Acropora (Table 1) and Isopora cuneata. Only samples from taxonomically unambiguous individuals were included in this study; the morphology of the corals sampled was absolutely consistent with their formal description [6]. To avoid the possibility of sampling clonemates, corals sampled were separated by at least 10 meters. The extreme rarity of several of the species examined limited the number of samples that it was possible to examine. Plots of the number of species distribution records against rank order (Figure 1a) clearly resolve rare species, such as A. pichoni (Figure 1b), from common species, with A. valida and A. nasuta being essentially pandemic throughout the Indo-Pacific.10.1371/journal.pone.0003240.g001Figure 1(a). Global abundance of the Acropora species used in this study.These data are based on numbers of records in the World Wide Acropora Database (n = 1523 sites; [6] and Wallace unpublished). (b). Several rare species, such as A. pichoni shown here, are likely to be unidirectional hybrids and occupy atypical habitats. Photo credit: Maria Beger.Census SizesEffective population sizes in reef corals are expected to be significantly smaller than census sizes for a number of reasons [13]. First, corals are known to undergo extreme variation in census population sizes due to perturbations such as storms and cyclones, bleaching or crown-of-thorns starfish outbreaks, which will substantially reduce effective sizes because it diminishes the proportion of the population involved in reproduction [8]. Second, high variance in fecundity (which is again known in corals [14]) reduces N\ne because neither juveniles nor senescent adults take part in reproduction [15]. Third, some Acropora species reproduce asexually by fragmentation or fission [16], which again reduces N\ne. Here we find mean (±SE) global census population sizes for rare species in this study varied from 32823 (±16412) for A. spathulata to 224 (±117) for A. rongelapensis. Based on the Ne estimate of 11% of the census population size, A. spathulata has a mean effective global population size of 3611 (±1805) and A. rongelapensis, 25 (±13) (Figure 2). Furthermore, it is likely that local population census and effective population sizes are substantially smaller than these conservative global estimates.10.1371/journal.pone.0003240.g002Figure 2Effective population size data for rare Acropora species included in this study.Mean (±SE) global census sizes are shown as black histograms, and predicted effective population sizes as red histograms. Data for A. tortuosa are omitted, as the mean global census size for this species (Supplementary Table S1) is more than two-fold higher than for A. spathulata (of those shown, the species with the largest global census size).Pax-C intron dataResults of phylogenetic analyses of Pax-C intron data (Figure 3) are broadly consistent with previous results, but some details differ due to the selection of taxa. To facilitate comparison with previous analyses, clades are labeled according to published trees [5], [9]. As in previous analyses, the basal clade contains A. longicyathus, and, in the present case, A. austera. In the present tree, a polytomy then gives rise to strongly supported clades corresponding to IIIA, IVB, IIID of previous studies; a major difference is the novel clade V which is composed exclusively of rare species with the exception of a single allele from A. valida. The nuclear tree distinguishes the Caribbean species in the highly supported clade IIID. Within the large terminal clade, two novel subclades (III F+G) were identified, containing predominantly sequences from rare species.10.1371/journal.pone.0003240.g003Figure 3Phylogenetic analysis of PaxC data.The figure shows the majority rule (>50%) consensus tree obtained in a Bayesian analysis of nuclear sequence data for the thirty-five Acropora species included in this study, with Isopora cuneata defined as outgroup. Bayesian analyses used likelihood settings from best-fit model (HKY+G) selected by hLRT in MrModeltest 2.2 [12]: 5 million generations; burn in  = 50,000. Numbers above branches are posterior probability values supporting the topology shown and clades are labelled according to previous [5], [9] analyses. Numbers after species names indicate the coral colonies from which the sequences were obtained; where more than one sequence was obtained per colony, the clone identity is given after an asterisk. Note that in some cases multiple clones (sometimes from different species) had identical sequences.Mitochondrial control region dataPhylogenetic analyses of the mtDNA Control Region (Figure 4) were also broadly consistent with previous results and clades were labeled as in previous publications [5], [9]. The basal clade (IA/IB) again contains A. longicyathus and A. austera, with A. tenuis added. In the present case, clade III is expanded and clade IV contracted relative to published analyses, due to differences in composition of the datasets. Clade IV includes A. aspera, A. humilis and several rare species (e.g. A. kirstyae, A. derawanensis).10.1371/journal.pone.0003240.g004Figure 4Phylogenetic analysis of mitochondrial sequence data.The figure shows the majority rule (>50%) consensus tree obtained in a Bayesian analysis of mitochondrial sequence data for thirty-five Indo-Pacific Acropora species with the Caribbean species Acropora cervicornis defined as outgroup. Bayesian analysis used likelihood settings from best-fit model (HKY+I+G) selected by hLRT in MrModeltest 2.2 [12]: 5 million generations; burn in  = 20,000. Numbers above branches are posterior probability values supporting the topology shown and clades are labelled according to previous [5], [9] analyses. Numbers after species names indicate the coral colonies from which the sequences were obtained.DiscussionIn both the Pax-C and mitochondrial phylogenies many Acropora species are polyphyletic. Previous work [5], [9] provides precedents for this pattern, which has been interpreted as evidence for interspecific hybridization. However, the Indo-Pacific species examined in these previous studies are widespread and locally common, and in these cases lineage sorting will occur slowly. As the fossil record of Acropora is extremely limited, for common and widespread species incomplete lineage sorting cannot be rigorously excluded as an alternative explanation for the observed polyphyletic patterns. However, for the rare species included in the present study, effective population sizes are so small (Fig 2) that lineage sorting will occur on very short time scales, so in contrast to the position with common species, polyphyletic patterns observed for rare species provide unequivocal evidence for hybridization.Comparison of the trees generated from nuclear and mitochondrial data (Figure 5) shows that three of the rare species studied here-A. pichoni, A. kimbeensis and A. papillare-are monophyletic for the mtDNA marker but are polyphyletic and contain highly divergent alleles at the nuclear marker, even within individual corals. The presence of species-specific mitochondrial haplotypes is unusual in Acropora\n[5], [9]. Of the 49 species studied to date, the only other Acropora species that is monophyletic in mtDNA is A. tenuis (Figure 4; however, see also below), which is known to be reproductively isolated through a difference in spawning time [5].10.1371/journal.pone.0003240.g005Figure 5Comparison of nuclear and mitochondrial phylogenies.Asterisks indicate posterior probability values of 100% (black) or >70% (red); for clarity, asterisks are shown only at nodes affecting the positions of sequences from A. papillare, A. pichoni, A. kimbeensis, A. spathulata and A. tortuosa.\nThe mitochondrial phylogeny implies that the three monophyletic rare species have evolved relatively recently, because they fall within derived positions of the large terminal clade that reflects the post-Miocene Indo-Pacific speciation of Acropora (i.e. <5.32 my) [5], [17]. In contrast, sequences from these three species are widely distributed throughout the nuclear tree; for example, alleles from A. papillare occur in both Clades III and V. This pattern in nuclear versus mtDNA loci can be explained by the known faster lineage sorting of mitochondrial haplotypes than alleles at single copy nuclear loci [18]. Unlike their more common relatives, the small effective global population sizes of these three rare species (A. pichoni = 521±125; A. kimbeensis = 1208±707; A. papillare = 284±142) effectively rules out the possibility of incomplete lineage sorting, because of their small population sizes, these rare species have very short coalescence times.There is no evidence that these rare species were historically more common. Moreover, these observed patterns–monophyly with respect to mitochondrial haplotypes accompanied by polyphyly at nuclear loci-cannot be explained as consequences of either recent population crashes or population bottlenecks. Under a population crash scenario one would expect to find divergent mitochondrial haplotypes as well as divergent nuclear alleles, whereas under a population bottleneck scenario (i.e. a crash occurring less recently) low diversity at both nuclear and mitochondrial loci is expected. These alternate possibilities can therefore be ruled out, and the most parsimonious explanation for the observed patterns of allele/haplotype distribution is that A. pichoni, A. kimbeensis and A. papillare are unidirectional hybrids.In the Caribbean, the hybrid species A. prolifera colonizes habitats that are distinct from those of the parental species [2], [3]. Similarly, two of the three rare putative hybrid species from the Indo-Pacific, A. pichoni and A. papillare, occur in atypical habitats. Whereas the vast majority of Acropora spp. occur in relatively shallow reef flat, crest and slope habitats (2–30 m), A. pichoni occurs below 40 m and A papillare, is found in extremely shallow intertidal habitats (<2 m). Specialization in extremely shallow or deep habitats is atypical for Acropora species hence our data provide support for the hypothesis that hybrid species may exploit atypical (or non-parental) niches.Other rare species occurring in small and isolated populations (e.g. A. walindii, A. loisetteae, A. derawanensis and A. jacquelineae) are polyphyletic with respect to both nuclear alleles and mitochrondrial haplotypes. Whilst these patterns are again consistent with hybridization, in these cases alternative explanations, such as recent population crashes, cannot be rigorously excluded.Two species that are geographically restricted but locally common (A. spathulata and A. tortuosa) are also monophyletic at the mitochondrial marker but polyphyletic at the nuclear marker. However, in these latter cases, incomplete lineage sorting cannot be ruled out because of the longer coalescence times for these species resulting from their larger census and predicted effective population sizes.The results presented here imply that a number of rare Indo-Pacific Acropora species are the products of recent hybridization events, and highlight the significance of hybridization in coral diversification. Whether these species have hybrid origins or have evolved and then hybridized in the absence of conspecific gametes remains to be elucidated.In summary, although it has often been assumed that small populations have a decreased potential for adaptation [19], our analyses imply that some rare Acroporid corals may actually have increased adaptive potential as a consequence of introgressive hybridization [20], and therefore may be less vulnerable to extinction than has been assumed.Supporting InformationMethods S1Calculation of mean global census and effective population sizes(0.07 MB DOC)Click here for additional data file.Table S1Estimates of mean global census size for rare species included in this study.(0.12 MB DOC)Click here for additional data file.Table S2Regional estimates of available reef habitat post 2004.(0.09 MB DOC)Click here for additional data file.\n\nREFERENCES:\n1. WillisBLvan OppenMJHMillerDJVollmerSVAyreDJ\n2006\nThe role of hybridization in the evolution of reef corals.\nAnn Rev Ecol Evol Sys\n37\n489\n517\n2. van OppenMJHWillisBLvan VugtHJAMillerDJ\n2000\nExamination of species boundaries in the Acropora cervicornis group (Scleractinia, Cnidarea) using nuclear DNA sequence analyses.\nMol Ecol\n9\n1363\n1373\n10972775\n3. VollmerSVPalumbiSR\n2002\nHybridization and the evolution of coral reef diversity.\nScience\n296\n2023\n2025\n12065836\n4. van OppenMJHWillisBLvan RheedeTMillerDJ\n2002\nSpawning times, reproductive compatibilities and genetic structuring in the Acropora aspera group: evidence for natural hybridization and semi-permeable species boundaries in corals.\nMol Ecol\n11\n1363\n1376\n12144658\n5. van OppenMJHMcDonaldBJWillisBLMillerDJ\n2001\nThe evolutionary history of the coral genus Acropora (Scleractinia, Cnidaria) based on a mitochondrial and a nuclear marker: reticulation, incomplete lineage sorting or morphological convergence?\nMol Biol Evol.\n18\n1315\n1329\n11420370\n6. WallaceCC\n1999\nStaghorn Corals of the World: A revision of the coral genus Acropora (Scleractinia; Astrocoeniina; Acroporidae) worldwide, with emphasis on morphology, phylogeny and biogeography.\nCSIRO Publishing, Melbourne\n7. WolstenholmeJKWallaceCCChenCA\n2003\nSpecies boundaries within the Acropora humilis species group (Cnidaria; Scleractinia): a morphological and molecular interpretation of evolution.\nCoral Reefs\n22\n155\n166\n8. FrankhamR\n1995\nEffective population size/adult population size ratios in wildlife: a review.\nGenetic Resources\n66\n995\n107\n9. MárquezLMvan OppenMJHWillisBLReyesAMillerDJ\n2002\nThe highly cross-fertile coral species, Acropora hyacinthus and A.cytherea, constitute statistically distinguishable lineages.\nMol Ecol\n11\n1339\n1349\n12144656\n10. HuelsenbeckJPRonquistF\n2001\nMrBayes: Bayesian inference of phylogenetic trees.\nBioinformatics\n17\n754\n755\n11524383\n11. KimuraM\n1980\nA simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences.\nJ Mol Evol\n16\n111\n120\n7463489\n12. NylanderJ\n2004\nMrModeltest2.2. Computer software distributed by the University of Uppsala\n13. HughesTPAyreDConnellJH\n1992\nThe evolutionary ecology of corals.\nTrends Ecol Evol\n7\n292\n294\n21236037\n14. WallaceCC\n1985\nReproduction, recruitment and fragmentation in nine sympatric species of the coral genus Acropora.\nMar Biol\n88\n217\n233\n15. CaballeroA\n1994\nDevelopments in the prediction of effective population size.\nHeredity\n73\n6657\n679\n16. AyreDJHughesTP\n2000\nGenotypic diversity and gene flow in brooding and spawning corals along the Great Barrier Reef, Australia.\nEvolution\n54\n1590\n1605\n11108587\n17. WallaceCCRosenBR\n2006\nDiverse Staghorn corals (Acropora) in high-latitude Eocene assemblages: implications for the evolution of modern diversity patterns of reef corals.\nProc. Royal Soc\nB273\n975\n982\n18. TavareS\n1984\nLines of descent and genealogical processes, and their application in population genetics models.\nTheor Pop Biol\n26\n119\n164\n6505980\n19. WilliYBuskirkJVHoffmannAA\n2006\nLimits to the Adaptive potential of small populations.\nAnn Rev Ecol Evol Sys\n37\n433\n458\n20. SeehausenO\n2004\nHybridization and adaptive radiation.\nTrends Ecol Evol\n19\n198\n207\n16701254"
+}

batch_3/PMC2532691.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2532691",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2532691\nAUTHORS: James H Keith, Tresa S Fraser, Malcolm J Fraser\n\nABSTRACT:\nBackgroundThe piggyBac transposable element is a popular tool for germ-line transgenesis of eukaryotes. Despite this, little is known about the mechanism of transposition or the transposase (TPase) itself. A thorough understanding of just how piggyBac works may lead to more effective use of this important mobile element. A PSORTII analysis of the TPase amino acid sequence predicts a bipartite nuclear localization signal (NLS) near the c-terminus, just upstream of a putative ZnF (ZnF).ResultsWe fused the piggyBac TPase upstream of and in-frame with the enhanced yellow fluorescent protein (EYFP) in the Drosophila melanogaster inducible metallothionein protein. Using Drosophila Schneider 2 (S2) cells and the deep red fluorescent nuclear stain Draq5, we were able to track the pattern of piggyBac localization with a scanning confocal microscope 48 hours after induction with copper sulphate.ConclusionThrough n and c-terminal truncations, targeted internal deletions, and specific amino acid mutations of the piggyBac TPase open reading frame, we found that not only is the PSORTII-predicted NLS required for the TPase to enter the nucleus of S2 cells, but there are additional requirements for negatively charged amino acids a short length upstream of this region for nuclear localization.\n\nBODY:\nBackgroundpiggyBac is a short repeat, target-site-specific transposable element originally isolated as mutational insertions within baculovirus genomes that originated from the infected TN-368 cells (Trichoplusia ni: Lepidopteran) [1]. This 2.4 kb transposable element is bounded by an asymmetric repeat configuration consisting of terminal 13 bp and sub-terminal 19 bp inverted repeats separated by either a 5' 3 bp or 3' 31 bp spacer [1]. The single piggyBac open reading frame is 1783 bp long, coding for a protein of 594 amino acids with a predicated mass of 68 kDa [1,2]. TPase catalyzed movement of piggyBac was originally demonstrated by utilizing the baculovirus genome in transfected Spodoptera frugiperda cell cultures as a target for the transposed DNA, and subsequently repeated using simple and rapid tests such plasmid excision assays [2] and interplasmid transposition assays which test for the removal of transposed DNA and its subsequent reinsertion into a different plasmid, respectively. The tests can be carried out in both transfected insect cells and microinjected insect embryos [3].The piggyBac element has several properties that make it an ideal tool for transgenesis, including site-specific integration and excision [2], proven large carrying capacity [4], controllable remobilization [5], and the ability to insert in heterochromatin and euchromatin throughout a genome, in both regulatory and coding regions, greatly facilitating enhancer trapping and random mutagenesis studies [5-7]. This is not to say that all of these properties are shared in all hosts for which they have been assayed. It should be noted that despite the function of piggyBac in the cells of Spodoptera frugiperda [8], attempts to transform the species itself have yet to be successful [9]. Simple plasmid-based mobility assays have also shown piggyBac to be active in human and other primate cells [4,10], in Zea maize cells [11], in Saccharomyces cerevisiae [12], and in the embryos of Aedes triseriatus [13], Heliothis virescens [14], and Danio rerio [10]. Of the species amenable to piggyBac-mediated germ-line or strain transformation, namely, Plasmodium falciparum [15], Mus musculus [4], Tribolium castaneum [5], Anopheles gambiae [16], Ceratitis capita [17], Drosophila melanogaster [18], Bactrocera dorsalis [19], Musca domestica [20], Lucilia cuprina [21], Bicyclus anynana [22], Aedes aegypti [23,24], Anopheles albimanus [25], Anopheles stephensi [26], Bombyx mori [27], Athalia rosae [28], Drosophila willistoni [29], Pectinophora gossypiella [30], Anastrepha suspensa [31], Aedes fluviatilis [32], Harmonia axyridis [33], and the human blood fluke Schistosoma mansoni [34], remobilization assays have only been attempted for Aedes aegypti [35], which was unsuccessful, and Tribolium castaneum [5], and Drosophila melanogaster [6], which both demonstrated remobilization function. In cases of straight transgene introduction, for example foreign protein production by silkworms, or RNAi studies, stable germ-line transformation is preferred. However, others situations, such as enhancer trapping and saturation mutagenesis, remobilization is desired. It is for these reasons piggyBac is proving a valuable tool for functional genomics in D. melanogaster [6] and quickly becoming the transposon of choice for germ line transformation [36].The piggyBac TPase is the archetype of a family of related sequences [37] identified from both computer predictions and EST libraries in a diverse array of organisms such as Takifugu rubripes, Xenopus, Daphnia, and even Homo sapiens [38]. At present, five piggyBac transposable element derived (PGBD) genes, some with multiple isoforms, have been identified among human cDNA clones (Genbank acc#: D88259, CR623168, AK074682, AK094816, and CR597281, respectively) [37]. PGBD3, isolated from human testis cDNA (Genbank acc#: BC034479), overlaps with the excision repair cross-complementing 6 gene (ERCC6), which codes for the cockayne syndrome B protein, CSB [39]. The first 465 residues of PGBD3 and ERCC6 (1061 and 1493 amino acids, respectively) are identical, and occur in the same place on the genome. Cockayne syndrome is a devastating autosomal recessive disease marked by premature physical aging, loss of hair, UV hypersensitivity, and mental retardation. Other signs include severe tooth decay, a high predisposition for a number of cancers, osteoporosis, demyelination of nervous tissue, calcification of the cortex and basal ganglia, and neuronal loss [40].The size of the piggyBac family, its wide utility as a transgene vector, and the patterns of piggyBac related protein expression in human tissues warrant a deeper investigation into the function of this obviously critical family of proteins. Through stepwise mutagenesis we can identify functional and catalytic domains for the TPase, and gain a better understanding of the functional properties of other members of the piggyBac family.TPase catalyzed integration and excision occurs within the eukaryotic nucleus, necessitating either diffusion or transport of the protein across the nuclear envelope through the nuclear pore complexes (NPC). While proteins below a size threshold of 40–60 kDa can passively diffuse [41] through these pores, those of greater mass must be actively transported through pore complexes by nuclear import proteins [42]. Actively transported proteins require one or more nuclear localization signals (NLSs) that facilitate their interaction, either directly or indirectly, with nuclear transport proteins [43]. However, piggyBac may also reside in the nucleus using a nuclear retention signal. In this scenario, piggyBac avoids the requirement for active nuclear transport and could only enter the nucleus during mitosis when the nuclear envelope is not present. While nobody has yet explored the possibility that transposition may only occur during mitosis, and an NLS is not needed, other TPases have already been shown to have NLSs [44-50]. Since the piggyBac TPase has a demonstrated mass of nearly 68 kDa [51], there is no selective pressure for a nuclear retention signal in the absence of active transport as TPase cannot passively diffuse out of the nucleus once entered. We presume that if it is indeed active in the presence of a nuclear envelope, it requires active nuclear transport and therefore may contain a NLS. We therefore find reasonable cause to suspect piggyBac possesses an active NLS as well.The mechanism for nuclear localization is highly conserved among eukaryotes. With the exception of a few specialized NLSs [52], a cell can recognize the NLS of exogenous proteins from highly divergent organisms [43]. Of those NLSs that have been identified, the two most widespread and well characterized are the classic bipartite and monopartite NLS [53,54]. Both of these signals rely on a tract of negatively charged amino acids that are essential for interaction with nuclear importin receptors. The wide host mobility for piggyBac suggests its TPase possesses a conserved NLS that conforms to at least one of the classical types of motifs and can operate in a large variety of cells.Sarkar et al. indicate a PSORTII [55] analysis of the piggyBac TPase predicts a bipartite NLS that falls within a twenty-one amino acid stretch ('PVMKKRTYCTYCPSKIRRKAN') of the C-terminus including residues 551 through 571 [37]. This region of the TPase, in fact, contains four patterns matching characterized NLSs.In this report we define the piggyBac NLS by constructing a series of piggyBac truncations and deletions fused in-frame and upstream of the fluorescent protein EYFP and comparing their nuclear localizing properties to that of a full length TPase-EYFP fusion in transfected Drosophila S2 cells. Using the PSORTII prediction as a starting point, we demonstrate that the regions of the TPase responsible for nuclear localization are located within the carboxy terminal 94 amino acids. Deletion of the PSORTII-predicted bipartite NLS, residues 551–571, eliminates nuclear targeting of the TPase-EYFP fusion protein. However, this sequence does not act as a NLS when placed at the amino-terminus of EYFP. The minimum deletion fragment of the piggyBac TPase required for nuclear localization of the EYFP protein includes the last 94 amino acids (501–594). Additional mutation analyses of piggyBac TPase-EYFP fusions further refine the NLS to within amino acids 501–571.Point mutation analysis identifies at least three individual amino acids located a short distance upstream of the predicted NLS that are essential for nuclear import, but like the predicted NLS, are alone insufficient for nuclear localization. Together these data establish that while the predicted NLS alone is too short to form a recognizable active domain, in its native context within the TPase protein it functions in the translocation of the protein to the nuclear compartment.ResultsFull Length piggyBacA PSORTII analysis of the predicted amino acid sequence for the piggyBac TPase identified NLS patterns between residues 551 and 571 that matched two known consensus signals. The first identified sequence, located at amino acids 554 through 571, is a region that is similar to the bipartite NLS originally defined for Xenopus nucleoplasmin [56] that is composed of 2 basic regions separated by a non-specific 10 residue spacer. This particular region of the TPase is so concentrated with basic amino acids that the bipartite consensus match can begin at either amino acid 554 or 555. In fact, the presence of basic residues in this region is so ubiquitous that, in addition to the bipartite signal, two regions consistent with the requirements for a monopartite NLS can be found in the same stretch: 'PVMKKRT' and 'PSKIRRK' at positions 551–557 and 563–569, respectively. These sequences resemble the monopartite signal exemplified by the SV40 large T antigen, which is defined as a proline followed by a basic region containing either arginine or lysine in 3 out of 4 residues, and within 3 residues of the original proline [54]. This result indicates that piggyBac has up to four possible classic NLS patterns in this short 21 amino acid region.Experimental identification of nuclear localization signalsTo obtain representative examples of what to expect with a nuclear localizing protein, and a diffuse protein, we first imaged the full piggyBac protein fused to EYFP, and the EYFP protein alone. Confocal imaging confirmed nuclear localization of the 96.5 kDa full length piggyBac TPase-EYFP fusion protein, coded by pMT/pBac-EYFP (fig. 1; fig. 2). The nucleus was readily evident in each picture, marked by the red emitting nuclear stain, Draq5. Yellow fluorescence was entirely absent from the cytoplasm and concentrated in the nucleus, which was visible by staining with Draq5. The 96.5 kDa pBac-EYFP fusion protein was well over the molecular weight threshold for passive diffusion of proteins into the nucleus, suggesting active nuclear transport was required. The distribution pattern observed for the pBac-EYFP product was distinctly different from that of the 28 kDa EYFP non-fusion protein control which yielded an evenly dispersed fluorescence in both cytoplasmic and nuclear compartments consistent with passive diffusion into and out of the nucleus (fig. 2). These results confirm an active nuclear localizing capability for the piggyBac TPase that facilitates nuclear import of proteins beyond the passive diffusion limit of 40–60 kDa.Figure 1piggyBac truncations. The piggyBac TPase is shown as an N-terminal fusion to the enhanced yellow fluorescent protein (EYFP). The PSORTII-predicted NLS region is indicated by solid black. The name of each vector is to the left of the visual diagram with the actual changes made listed to the right of the diagram. The observed nuclear localization pattern is indicated in the right column. Sizes and distances are not necessarily to scale. Numbers represent amino acid positions with respect to the piggyBac start codon.Figure 2Confocal microscopy. Confocal microscope images for vectors described in the text. Vector names and their corresponding images are shown. The first column is a transmitted black and white image of the cell. The second column shows EYFP fluorescence pattern observed as a fusion protein with the piggyBac TPase. The third column is the nuclear stain Draq5 while the fourth column is an overlay of the EYFP fluorescence and Draq5 stain. All microscopy work was performed approximately 48 hours post induction. All images are the result of 6 lines averages performed by the imaging software. Each image is zoomed and cropped on the cell or cells of interest but all remain otherwise unenhanced and unaltered.Truncation mutation analysisWe constructed both amino-terminal and carboxy-terminal deletion series for the piggyBac TPase to experimentally verify the location of a functional NLS within the 1782 bp piggyBac TPase open reading frame. We deleted piggyBac from either side in roughly 300 bp increments (fig. 1: pMT/NLS-1 through pMT/NLS-10) in two separate series of deletions. Each of these deletion series were fused upstream and in-frame with EYFP, and positioned for expression within the pMT vector.The compartmentalization pattern for each expressed TPase truncation-EYFP fusion protein from either the N-terminal or C-terminal series was observed following transient expression of transfected S2 cells using confocal microscopy. Cells transfected with vectors expressing fusion proteins that retained the 94 carboxy-terminal amino acids of piggyBac exhibited yellow fluorescence that concentrated within the nucleus, while no significant nuclear localization was evident for EYFP fusions that did not contain these 94 carboxy-terminal amino acids. The smallest contiguous TPase fragment sufficient to yield distinct nuclear localization activity was the c-terminal 94 amino acid sequence expressed in pMT/NLS-5 (Δ1–500; fig. 2), while the largest TPase fusion of the C-terminal deletion series that failed to localize to the nucleus was pMT/NLS-6 (Δ497–594; fig. 2). The difference in localization patterns between the diffuse EYFP-only protein and the larger, nuclear-concentrated pMT/NLS-5 expressed protein was plainly visible. These results demonstrated that the nuclear localization signal must be located downstream of amino acid 500.Analysis of the carboxy-terminusThe N-terminal and C-terminal truncations provided evidence that the carboxy terminal 94 amino acids of the piggyBac open reading frame were both necessary and sufficient to cause the nuclear localization of piggyBac. This sequence included the PSORTII-predicted NLS. We analyzed this region in detail to more accurately define the boundaries and function of the predicted piggyBac NLS. We constructed an in-frame fusion of the NLS-deletion upstream of the EYFP ORF to create pMT/NLS-11 (Δ551–571; fig. 3). Deletion of the entire PSORTII-predicted NLS eliminated expressed fluorescence from the nucleus of S2 cells (fig. 2) which confirmed the necessity of the PSORTII-predicted segment for nuclear localization. Interestingly, this fusion protein appeared to aggregate, forming small but distinct foci in the cytoplasm when viewed under higher magnifications. This aggregation differed significantly from the distribution obtained with the expressed EYFP control protein, which exhibited a diffused, homogenous fluorescence throughout both the nucleus and cytoplasm.Figure 3piggyBac mutation and truncation refinements. Vectors used in the investigation of the nuclear localization pattern of piggyBac in and around the PSORTII-predicted NLS. Deletions are represented by bridged lines. Mutations are specifically indicated. The name of each vector is to the left of the visual diagram with the actual changes made listed to the right of the diagram. The observed nuclear localization pattern is indicated in the right column. Sizes and distances are not necessarily to scale. Numbers represent amino acid positions with respect to the piggyBac start codon.Next, we directly investigated the functionality of solely the PSORTII-predicted piggyBac NLS by fusing this short encoding segment between amino acids 551 and 571, inclusive, to EYFP to yield pMT/NLS- 12 (Δ1–550, Δ572–594; fig. 3). Although the molecular weight of the protein (28 kDa) was below the 40–60 kDa threshold for passive diffusion into the nucleus, the resulting protein was observed in both the nucleus and the cytoplasm (fig. 2), clearly different from pMT/pBac-EYFP. The failure of this fusion protein to concentrate solely in the nucleus indicated an inability of these residues to form a functional NLS domain, suggesting the function of this sequence is context-dependent.Importance of sequences flanking the NLSSince fusion of TPase amino acids 551 through 571 to the N-terminus of EYFP did not allow direct confirmation of a NLS function for the PSORTII-predicted sequences, additional flanking amino acids likely contribute to the activity of this sequence, most likely through facilitation of proper folding. To confirm this hypothesis we constructed two TPase deletion mutations that omitted amino acids either upstream or downstream of the predicted NLS by PCR amplification of the pMT/pBac-EYFP plasmid using inverse-facing primers bounding the area to be deleted. Deletion mutation pMT/NLS-13 (Δ572–594; fig. 3) contained all the amino acids upstream of the predicted NLS. The pattern of fluorescence obtained with this deletion-fusion (fig. 2) was indistinguishable from that of the full length piggyBac-EYFP fusion protein, demonstrating that amino acids downstream of the predicted NLS are dispensable for efficient nuclear trafficking.A second deletion-fusion, pMT/NLS-14 (Δ497–550; fig. 3), removed 54 residues upstream of the predicted NLS. The pMT/NLS-14 fusion protein (fig. 2) remained dispersed in the cytoplasm, demonstrating that the 54 amino acid sequence upstream of the NLS is likely involved in the proper presentation or functioning of the NLS domain.Two additional deletion fusions in this 50 amino acid flanking sequence were also examined for possible contributions to the nuclear localization activity. The specific boundaries of the deletion constructs pMT/NLS-15 and pMT/NLS-16 were chosen based upon the presence of a proline residue at positions 522 and 537, respectively. Deletion fusions pMT/NLS-15 (Δ497–522, Δ572–594; fig. 3) and pMT/NLS-16 (Δ497–536, Δ572–594; fig. 3) were created by deleting portions of the piggyBac open reading frame between amino acid 497 and either proline 522 or proline 537, inclusive, utilizing the deletion plasmid, pMT/NLS-13 as the template. pMT/NLS-15 trafficked efficiently to the nucleus (fig. 2) while the fusion protein lacking the more lengthy segment, pMT/NLS-16, remained confined to the cytoplasm (fig 3). We emphasize that both of these fusion proteins had predicted masses well over the size threshold required for passive diffusion into the nucleus. Taken as a pair, the localization patterns of these two deletion proteins could be interpreted to indicate the NLS is between amino acids 523 and 535. However, pMT/NLS-11 also fails to enter the nucleus, suggesting that both these regions are required for nuclear localization. These results identified the segment of piggyBac required for proper presentation of the predicted NLS as contained somewhere between amino acids proline 522 and glutamic acid 550.Importance of basic amino acids proximal to the predicted NLSThe inability of the isolated TPase PSORTII-predicted NLS motif to cause nuclear localization suggested a role for the adjacent amino acids in the formation of a functional nuclear localization motif. Our deletion plasmids pMT/NLS-15 and pMT/NLS-16 confirmed the requirement for upstream amino acids. Investigation of the area between proline 522 and glutamic acid 550 revealed three basic amino acids K525, R526, and R529. These basic amino acids lie among a stretch of largely neutral residues.Substitution of these residues with neutral amino acids would reveal any specific requirement for these in the nuclear localization of piggyBac. Two plasmids were created: pMT/NLS-17 (Δ497–522, Δ572–594, K525A, R526A, R529A; fig. 3), and pMT/NLS-18 (Δ497–522, Δ572–594, R526A, R529A; fig. 3). Simple replacement of these amino acids with uncharged residues disrupted the nuclear localization activity of fusion proteins that were otherwise trafficked to the nucleus, including those containing the predicted NLS (fig. 2). The altered fluorescence patterns for pMT/NLS-17 and pMT.NLS-18 reveals that while the bipartite signal may contribute the core nuclear localization activity to piggyBac TPase, additional flanking amino acids are somehow involved in the proper function of this signal.DiscussionEukaryotic proteins that bind with or interact with DNA must be capable of entering the nuclear compartment. NLSs have been identified in several eukaryotic TPases including Hermes of Musca domestica [44], mariner of Drosophila mauritania [45], BmTc1 of Bombyx mori [46], Mu [47] and Activator [48] of Zea maize, Tag1 of Arabidopsis Thaliana [49], and the reconstructed salmonid transposon, Sleeping Beauty [50]. Previous studies have demonstrated the nuclear localization capacity of Minos of D. hydei [57]. Analysis of Minos by PSORTII predicts 4 separate amino acid sequences consistent with standard patterns. These are monopartite signals: 'PRDKRQL', 'KKKR', and 'PKRVKCV' at amino acid positions 67, 130, and 325 respectively and a bipartite signal, 'RKRSETYHKDCLKRTTK', at 172. Many prokaryotic recombinases and integrases exhibit enhanced activity in eukaryotic cells when they are linked with eukaryotic nuclear importation signal sequences. For example, recombination activity of the φ C31-integrase is enhanced in eukaryotic cells when the SV40 T-antigen archetypical NLS is fused to the carboxy terminus [58]. Because the piggyBac TPase is larger than the threshold size for passive diffusion it also must be actively targeted to the nucleus to be effective in target site recognition and transposition.A PSORTII examination of the piggyBac TPase sequence predicted multiple mono- and bi-partite NLSs. The classic pat4 monopartite signal pattern is composed of three or four basic residues (K or R) followed by a H or P. Additionally, the monopartite signal can adhere to the pat7 pattern, having a P residue followed closely by a four residue stretch that contains at least three basic amino acids [54]. The bipartite signal follows a somewhat more defined consensus pattern with two basic amino acids followed by a ten residue spacer that connects to at least three out of five basic amino acids [56]. There is considerable variability in the ten residue spacer, allowing for a number of different motifs to be located within the bipartite NLS signal. Our data cannot rule out that either or both of the predicted monopartite signals are the true NLSs of the piggyBac TPase each with a requirement for the upstream basic amino acids for proper function.The NLSs and nucleic acid binding domains of most proteins that exhibit both activities either overlap or are located immediately adjacent to each other [53]. This frequent overlap appears to result from co-evolution of the DNA interacting domain and nuclear localization signal for a given protein [59]. Several examples of overlap or close proximity between the two signals have been characterized [60]. In some cases the NLS may be too short to form an independent functional domain and may have additional requirements for adjacent residues to present a successful secondary structure for nuclear targeting. For example, the bipartite NLS of the human androgen receptor is fully dependent on the presence of the overlapping ZnF, which itself is responsible for DNA binding [61]. Cokol and colleagues (2000) analyzed publicly available protein motif information and concluded that for 90% of proteins in which both the DNA binding domain and NLS are known, these signals overlap. The flexibility of the ten residue spacer in the bipartite signal allows different DNA sequences to be targeted while preserving the underlying NLS pattern and function.In fact, the location of the predicted bipartite NLS and the second predicted monopartite NLS of the piggyBac TPase overlap a ZnF motif 'CTYCPSKIRRKANASCKKCKKVICREHNIDMCQSCF' found at the very C-terminus of piggyBac TPase starting at residue 559. This ZnF is a novel match for the well-known RING-finger motif which, in the case of piggyBac TPase, starts in the spacing region of the bipartite signal and extends downstream to the end of the TPase. ZnFs are classically implicated in the DNA binding, while the RING-finger variant is more apt to be part of a protein-protein domain, a function that piggyBac would require either by itself or through interacting host factors in order to carry out transposition [62]. Previous work by our lab with western blots, co-immunoprecipitation, and the yeast two-hybrid system suggests a multimerization capacity of the TPase (unpublished). For instance, piggyBac has a proven ability to catalyze the transposition of a wide range of load sizes, implying that domains which interact with the piggyBac ITRs are not at a fixed distance relative to each other. Additionally, when used in a cartridge with one upstream ITR and a choice of either a proximal or a distal downstream ITR, piggyBac shows no particular preference for either ITR [51].Deletion of the PSORTII-predicted bipartite NLS, located between amino acids 551 and 571, inclusive, eliminates nuclear targeting of the piggyBac TPase-EYFP fusion protein. However, addition of this same sequence at the amino-terminus of EYFP is insufficient to provide nuclear targeting. Fusion of a series of systematic deletions from both the carboxy and amino termini of the piggyBac TPase upstream of the marker protein EYFP allows us to define the minimum sequence sufficient for nuclear trafficking as the carboxy-terminal 94 residues. In addition, deletion of the last 23 amino acids of the piggyBac open reading frame, encompassing everything downstream of the bipartite NLS, demonstrates this region is unnecessary for nuclear localization. The fact that piggyBac is active in a wide range of host cells and species would indicate that any NLS it possesses is readily recognized by conserved nuclear importing machinery. We find no logical reason to suspect that an NLS displaying such a wide tropism would be any less conserved. We therefore conclude a functional NLS is contained within the 71 amino acids from 501 to 571, and in light of the wide activity of piggyBac, the active NLS is most likely one or more of the 4 common patterns predicted by PSORTII.Our results also demonstrate that a segment of the TPase upstream of the predicted bipartite NLS is also essential for nuclear localization. We therefore attempted to define the involvement of these upstream sequences using point directed mutation analysis and further deletions.The amino acid proline breaks the periodic structure of α-helices and β-sheets, dividing the structure of a protein from one functional domain to the next [63]. If the NLS of the piggyBac TPase lies within a larger conformational domain, then the start of such a domain may be indicated by a proline. Examination of prolines located upstream from the predicted bipartite signal for their possible involvement in delineating regions responsible for the proper presentation of the piggyBac NLS defined a smaller region comprised of amino acids 522 through 571 that is sufficient for nuclear localization. This region includes the predicted bipartite NLS and the 29 amino acids immediately upstream. Nuclear localization was unaffected by deletions upstream of proline-522, but removal of the residues between proline-522 and proline-537 completely abolished nuclear localization. However, these data alone cannot rule out an alternate interpretation that all four PSORTII-predicted NLSs are, in fact, necessary but nonfunctional, and that the upstream flanking basic amino acids constitute the true NLS by interacting in a novel manner with conserved importin machinery.Alteration of the basic amino acids between proline-522 and proline-537 confirmed their importance in nuclear trafficking. The changes K525A;R526A;R529A and R526A;R529A each prevented the EYFP fusion proteins from entering the nucleus. Therefore, these arginines are somehow involved in the formation of a functional nuclear localizing domain within the context of a pBac-EYFP fusion. This requirement for proximal amino acids for the function of an NLS is not without precedent [61].ConclusionWe conclude from these findings that the piggyBac TPase has a functional NLS located between amino acids 551 and 571. However, our deletion and mutation constructs do not allow for a complete examination of the functionality of the monopartite signals either alone or in tandem, separate from the predicted bipartite NLS. Some NLSs function with non-native proteins when they are simply appended to the C-terminus [58], and some require flanking amino acids from their native context to retain nuclear import activity [61]. This short segment of amino acids in the piggyBac TPase is most likely the critical component of the nuclear localization function through its binding, either directly or through an adapter molecule, to a member of the importin family of proteins.We have demonstrated a requirement for the presence of at least two basic amino acids located proximally upstream of the predicted bipartite signal. One conclusion which cannot be ruled out by these data is that these upstream basic amino acids could constitute a novel NLS, with a requirement for the predicted NLS in an auxiliary capacity. To hold true, this interpretation requires all four PSORTII-predicted NLSs to be non-functional and the new putative NLS formed by these amino acids to be conserved across kingdoms and recognized by all cells in which piggyBac functions. The role of NLSs can be influenced by proximal amino acids or tertiary configurations. Therefore, a simpler interpretation of these data is that one or more of the four predicted NLSs is functional and the identified upstream arginines are required for their activity.Finally, sequencing analysis reveals the presence of an overlapping ZnF motif. When taken in the context of previous studies [53] this co-localization suggests the putative ZnF motif may constitute the piggyBac DNA binding domain. This is a logical arrangement when considered in the context of TPase evolution: allowing a TPase to carry out excision and reinsertion in the nucleus even while its sequence recognition sites are changing, giving rise to new family members. There is also the possibility that the ZnF may not function in DNA binding at all, but may be responsible for protein-protein interactions such as dimerization of the piggyBac TPase, binding host auxiliary factors, or heterochromatin interactions. Further investigation into this ZnF will need to be performed to elucidate its exact function, if any, in piggyBac transposition.MethodsPlasmid constructionThe EYFP open reading frame was obtained through PCR amplification of pXL-Bac-EYFP [64] using Pfx high-fidelity polymerase (Invitrogen, Carlsbad, CA) with primers supplying EcoRI (Note: all restriction enzymes obtained from New England Biolabs, Ipswich, MA) and NotI restriction sites at the 5' and 3' ends, respectively (table 1). The resulting PCR product was band isolated from a 9% agarose TAE gel, purified with QIAquick Gel Extraction columns (Qiagen, Valencia, CA) and digested with NotI and EcoRI. The inducible D. melanogaster metallothionein promoter vector pMT/V5-HisA (Invitrogen) was digested with the restriction enzymes NotI and EcoRI and treated with calf intestine alkaline phosphatase. The EYFP open reading frame was subsequently ligated into this vector to obtain pMT/EYFP. Sequencing and restriction analysis of the plasmid verified the presence and integrity of the EYFP open reading frame in pMT/EYFP. Functional fluorescence was tested by transient transfection of S2 cells with Cellfectin (Invitrogen) according to manufacturer's protocol. Expression of EYFP was induced by addition of CuSO4 (Sigma-Aldrich, St Louis, MO) to the medium at a final concentration of 500 μM. Fluorescence was observed with a Nikon Diaphot microscope.Table 1Primers and oligos used in this studyPrimer 1Primer 2pMT/EYFPACTGGAATTCATGGTGAGCAAGGGCGAGGAGCTGCTAGAGTCGCGGCCGCTTTACTTGTApMT/pBac-EYFPTAGAATTCTCGTGACTAATATATAATAAAATGGGTATTAGTGAATTCGAAACAACTTTGGCACATATCpMT/NLS-1AAGAATTCGGGATGGCTTCAAAGTCCACGAGGCGTAGCATTAGTGAATTCGAAACAACTTTGGCACATATCpMT/NLS-2CAGAATTCGTCATGGATCGATCTTTGTCAATGGTGTAATTAGTGAATTCGAAACAACTTTGGCACATATCpMT/NLS-3TGGAATTCAACATGCGTACGAAGTATATGATAAATGGAATTAGTGAATTCGAAACAACTTTGGCACATATCpMT/NLS-4TTGAATTCAACATGGCCCTTACTCTCGTCTCATATAAAATTAGTGAATTCGAAACAACTTTGGCACATATCpMT/NLS-5AGGAATTCAGTATGGAAAAATTTATGAGAAACCTTTACATTAGTGAATTCGAAACAACTTTGGCACATATCpMT/NLS-6TAGAATTCTCGTGACTAATATATAATAAAATGGGTCGGAATTCAACCTTTTCTCCCTTGCTACTGACpMT/NLS-7TAGAATTCTCGTGACTAATATATAATAAAATGGGTAGGAATTCGGGTCCGTCAAAACAAAACATCGpMT/NLS-8TAGAATTCTCGTGACTAATATATAATAAAATGGGTGTGAATTCGTCACACATCATGAGGATTTTTATpMT/NLS-9TAGAATTCTCGTGACTAATATATAATAAAATGGGTAGGAATTCTGTGGACATGTGGTTATCTTTTCTpMT/NLS-10TAGAATTCTCGTGACTAATATATAATAAAATGGGTGTGAATTCTGAAGTTGACCAACAATGTTTATTpMT/NLS-11ATATGGATCCGCATCGTGCAAAAAATGCAAAAAAGTTTTTGGATCCCTCTTCAGTACTGTCATCTGATGTACCpMT/NLS-13TTTGGATCCATTTGCCTTTCGCCTTATTTTAGAGGGGCAAAGGATCCGAAATGGTGAGCAAGGGCGAGGAGCTGpMT/NLS-14CCCGGATCCAACCTTTTCTCCCTTGCTACTGACATTATGGCCCCGGATCCCCAGTAATGAAAAAACGTACTTACTGTACTTACTGCCCCpMT/NLS-15TTTTGAGCTCAACCTTTTCTCCCTTGCTACTGACATTATGGCTTTTGAGCTCCCTACTTTGAAGAGATATTTGCGCGATpMT/NLS-16TTTTGAGCTCAACCTTTTCTCCCTTGCTACTGACATTATGGCTTTTGAGCTCCCAAATGAAGTGCCTGGTACATCAGATGpMT/NLS-17TTTTGAGCTCAACCTTTTCTCCCTTGCTACTGACATTATGGCTTTTGAGCTCCCTACTTTGAAGGCCTATTTGGCCGATAATATCTCTAATATTTTGpMT/NLS-18TTTTGAGCTCAACCTTTTCTCCCTTGCTACTGACATTATGGCTTTTGAGCTCCCTACTTTGGCCGCTTATTTGGCCGATAATATCTCTAATATTTTGpMT/NLS-12AATTCGTAATGGGGCCAGTAATGAAAAAACGTACTTACTGTACTTACTGCCCCTCTAAAATAAGGCGAAAGGCAAATGAATTCATTTGCCTTTCGCCTTATTTTAGAGGGGCAGTAAGTACAGTAAGTACGTTTTTTCATTACTGGCGCCATTACGPrimers used to make each of the vectors described in the text. Vector names are listed on the left with the cooresponding primers used to make the vector given on the right. In the case of pMT/NLS-12, ordered oligos were used as part of the final vector and not for PCR priming, as described in materials and methods.The native piggyBac open reading frame sequence was PCR amplified from p3E1.2 [1] with end specific primers supplying an EcoRI site at either end (table 1). The PCR product was band isolated in 9% agarose TAE gel and digested with EcoRI. The vector pMT/EYFP was linearized with EcoRI, treated with calf intestine alkaline phosphatase and ligated to the piggyBac open reading frame to create a fusion consisting of the full length piggyBac open reading frame joined on its C-terminus to EYFP to form pMT/pBac-EYFP.The vector pMT/EYFP was cut with EcoRI and treated with calf intestine alkaline phosphatase in preparation for the insertion of piggyBac sequences. The C-terminal piggyBac open reading frame truncations pMT/NLS-1 through pMT/NLS-5 (Δ1–100, Δ1–200, Δ1–302, Δ1–400, and Δ1–500, respectively) were all obtained by PCR amplification of p3E1.2 with Pfx high-fidelity polymerase, using a forward primer specific for the start of the piggyBac open reading frame and a reverse primer specific for each truncation as listed in table 1. The N-terminal truncations pMT/NLS-6 through pMT/NLS-10 (Δ497–594, Δ401–594, Δ301–594, Δ197–594, and Δ102–594, respectively) were also PCR amplified using a forward primer specific for each truncation (table 1) and a reverse primer specific for the end of the piggyBac open reading frame minus the stop codon.The PCR product bands were isolated by 9% agarose TAE gel electrophoresis, cut with EcoRI and ligated into the prepared pMT/EYFP vector to obtain a chimeric open reading frame consisting of the piggyBac insertions fused upstream and in-frame with the downstream EYFP. Sequencing and restriction analysis verified the resulting ligations.To obtain the deletion mutations pMT/NLS-11 (Δ551–571), pMT/NLS-13 (Δ572–594), and pMT/NLS-14 (Δ497–550), pMT/pBac-EYFP was PCR amplified using Pfx high-fidelity polymerase with inverted primers as noted in table 1. Briefly, pMT/pBac-EYFP was isolated from the dam methylating bacteria DH10B and subsequently used as a template. The majority of the plasmid except the deleted section was amplified and the resulting PCR reaction ethanol precipitated, washed with 70% ethanol, and resuspended in nuclease free water. Following resuspension, the DNA was cut with BamHI to prepare the product ends for ligation, and DpnI to digest the template. After a second ethanol precipitation, 70% ethanol wash, and resuspension in water, the PCR product was subject to self-ligation to form the respective plasmids. Restriction analysis and sequencing confirmed the integrity of the plasmids.The deletions pMT/NLS-15 (Δ497–522, Δ572–594) and pMT/NLS-16 (Δ497–536, Δ572–594) were created by PCR amplification of dam methylated pMT/NLS-13 with inverted primers containing SacI restriction sites at the 5' ends (table 1). The PCR products were ethanol precipitated, washed with 70% ethanol, and resuspended in nuclease free water. Following resuspension, the DNA was cut with SacI to prepare the product ends for ligation and DpnI to digest the template. After a second ethanol precipitation, 70% ethanol wash, and nuclease free water resuspension, the PCR product was subject to self-ligation to form the respective plasmids. Restriction analysis and sequencing confirmed the integrity of the plasmids.The plasmids containing the amino acid substitutions, pMT/NLS-17 (Δ497–522, Δ572–594, K525A, R526A, R529A) and pMT/NLS-18 (Δ497–522, Δ572–594, R526A, R529A) within the pMT/NLS-15 deletion construct were made by PCR amplification of dam methylated pMT/NLS-13 with inverted primers similar to the construction of the pMT/NLS-15 deletion vector (table 1) Each ligation resulted in a plasmid containing the Δ497–522, Δ572–594 deletion open reading frame with the amino acid substitutions R526A, R529A and K525A, R526A, R529A respectively.The pMT/NLS-12 (Δ1–550, Δ572–594) (fig. 3) fusion vector was constructed by annealing two oligonucleotides (table 1), to form a short double stranded DNA segment corresponding to the upstream and downstream outer boundaries of the PSORTII-predicted nuclear localization signal with EcoRI sticky ends. Briefly, 400 pmol of each oligo were combined in a total volume of 10 μl in a thin-walled PCR tube and heated by floating in 400 mL of boiling water. The water and oligos were then allowed to cool to room temperature undisturbed to facilitate annealing of the two oligos, keeping hairpinning and non-specific binding to a minimum. pMT/EYFP was then cut with EcoRI but not phosphatase treated. Combined with a large molar excess of the oligo mixture and ligated, the resulting vector was designated pMT/NLS-12.Cell culture and transfectionD. melanogaster Schneider 2 (S2) cells were grown in Schneider's medium (Gibco, Carlsbad, CA) supplemented with 10% FBS, 1 mg/mL streptomycin, and 25 μg/mL amphotericin at 28 degrees. Cells were transfected with Cellfectin (Invitrogen) using the recommended manufacturer protocol. Briefly, sterile coverslips were placed in the bottom of 9.4 cm2 wells and used as the surface for cell adherence. 1 ml of cells were seeded in at 6 × 106/ml in S2 medium supplemented with 10% FBS, 1 mg/mL streptomycin, and 25 μg/mL amphotericin (Sigma-Aldrich). The cells were allowed to adhere to the coverslip for 3 hours before undergoing transfection. Adherent cells were washed twice with serum-free S2 medium and resuspended in 800 μl serum-free S2 medium. For each transfection, 3 μl of Cellfectin was hydrated for 15 minutes in 97 μl sterile nuclease free water and added to 5 μg of DNA in 100 μl nuclease free water for a total volume of 200 μl. The Cellfectin-DNA mixture was allowed to incubate at room temperature for 20 minutes and added directly to the cells in a drop-wise manner followed by agitation to mix. The cells were incubated 18 hours at 28 degrees then given fresh S2 medium supplemented with 10% FBS, 1 mg/mL streptomycin, and 25 μg/mL amphotericin, as well as 500 μM CuSO4 (final concentration) to induce metallothionein promoter activity. Initial EYFP fluorescence was detectable at 4 hours post-induction through an EYFP filter (Chroma Technology Corp cat #40128; Excitation: 500 nm; Emission: 535 nm, Rockingham, VT) on a Nikon Diaphot (Nikon, Melville, NY) inverted phase contrast microscope, however the cells were analyzed at 48 h to allow for maximum EYFP signal.Confocal imagingTo prepare cells for confocal imaging, cells were transfected on coverslips as described above. At 48 hours post-induction, the media was aspirated from the coverslip, 200 μl of a 10 μM Draq5 (Biostatus Ltd., Leicestershire, UK) solution in 1× PBS was placed on the coverslip and incubated at room temperature for 10 minutes. The coverslips were then rinsed gently with 1× PBS and a slide was prepared with one drop of ProLong Gold antifade reagent (Invitrogen). The coverslip was sealed to the slide with nail lacquer and imaged with a Leica TCS SP2 True Confocal Scanner (Leica Microsystems, Bannockburn, IL) confocal microscope for EYFP and Draq5 fluorescence. Digital images represent 6 line averages and are cropped but otherwise remain unprocessed in the final images for publication.Authors' contributionsJHK created all plasmids used in this study, performed the confocal imaging, and prepared the manuscript. TSF performed all transfections and slide preparations. MJF conceived of the study and provided guidance. Special thanks to William Archer and Dr. Edward Hinchcliffe for their instruction in the use of the confocal microscope. All authors provided intellectual contributions as the study unfolded and reviewed the manuscript prior to submission.\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2532747.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2532747",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2532747\nAUTHORS: Andrija Tomovic, Edward J. Oakeley\n\nABSTRACT:\nBackgroundWith increasing numbers of crystal structures of protein∶DNA and protein∶protein∶DNA complexes publically available, it is now possible to extract sufficient structural, physical-chemical and thermodynamic parameters to make general observations and predictions about their interactions. In particular, the properties of macromolecular assemblies of multiple proteins bound to DNA have not previously been investigated in detail.Methodology/Principal FindingsWe have performed computational structural analyses on macromolecular assemblies of multiple proteins bound to DNA using a variety of different computational tools: PISA; PROMOTIF; X3DNA; ReadOut; DDNA and DCOMPLEX. Additionally, we have developed and employed an algorithm for approximate collision detection and overlapping volume estimation of two macromolecules. An implementation of this algorithm is available at http://promoterplot.fmi.ch/Collision1/. The results obtained are compared with structural, physical-chemical and thermodynamic parameters from protein∶protein and single protein∶DNA complexes. Many of interface properties of multiple protein∶DNA complexes were found to be very similar to those observed in binary protein∶DNA and protein∶protein complexes. However, the conformational change of the DNA upon protein binding is significantly higher when multiple proteins bind to it than is observed when single proteins bind. The water mediated contacts are less important (found in less quantity) between the interfaces of components in ternary (protein∶protein∶DNA) complexes than in those of binary complexes (protein∶protein and protein∶DNA).The thermodynamic stability of ternary complexes is also higher than in the binary interactions. Greater specificity and affinity of multiple proteins binding to DNA in comparison with binary protein-DNA interactions were observed. However, protein-protein binding affinities are stronger in complexes without the presence of DNA.Conclusions/SignificanceOur results indicate that the interface properties: interface area; number of interface residues/atoms and hydrogen bonds; and the distribution of interface residues, hydrogen bonds, van der Walls contacts and secondary structure motifs are independent of whether or not a protein is in a binary or ternary complex with DNA. However, changes in the shape of the DNA reduce the off-rate of the proteins which greatly enhances the stability and specificity of ternary complexes compared to binary ones.\n\nBODY:\nIntroductionDNA-binding proteins are important for the regulation of many crucial cellular processes (including transcription, recombination, and replication). The number of DNA-binding proteins known is very small compared to the number of regulatory controls they must provide within the nucleus. The problem is solved, at least in part, by the construction of higher-order regulatory complexes composed of multiple proteins. Structural analyses of such complexes may enable us to model the forces driving their assembly and stability which in turn may help us to understand these processes better. Such an understanding may help in predicting DNA-binding specificities. Transcription factors, a large subclass of DNA-binding proteins, are known to act cooperatively in the regulation of gene expression [1]–[7]. Their complexes can include both DNA and non-DNA-binding factors. The DNA-binding factors may be located either remotely (at some distance) or adjacent (with direct contacts) to their promoters [5].Thanks to a large number of recent X-ray and NMR structures of protein∶protein, protein∶DNA, and protein∶RNA complexes, a lot of valuable information about the general features of such complexes has been discovered [8]–[23]. These results indicate that it is very difficult to find universally characteristic rules which can describe all protein-protein, protein-DNA, and protein-RNA interactions. However, some general principles have been deduced. For example, Lys or Arg pair preferentially with any nucleotide in both protein∶DNA and protein∶RNA complexes [16]; two-thirds of all protein-DNA interactions involve van der Waals contacts, compared to about one-sixth involving hydrogen bonds [18]; on average protein-protein interface has approximately the same non-polar character as the protein surface as a whole and carries somewhat fewer charged groups (however, some interfaces are significantly more polar and others more non-polar than the average) [17].The current work comprises a structural analysis of macromolecular assemblies where several proteins are bound to DNA, using data from the Protein Data Bank (PDB) [24]. We analyzed the following chemical and physical properties: the size of interfaces between any two components; the number of residues/atoms involved in contacts between components; residue interface propensities and chemical composition; water-mediated contacts in interfaces; secondary structure motifs in interfaces; and interactions between amino acid side chains either with the DNA or with another protein in the complex. Some of these interface properties for ternary/quaternary complexes (i.e. complexes involving two/three proteins bound to DNA) have been compared with those obtained from binary complexes. One possible hypothesis why the above-mentioned protein-DNA and protein-protein interface properties are expected to depend on the number of proteins in a complex is that when two proteins are free (not bound to DNA) they are more able to find the best patches (on both proteins) to produce the most stable complexes possible, with the highest affinity between components. However, when one protein is bound to DNA then there is a spatial limitation in the movements that are possible in order to find the best interface patches (on both proteins) in order to make stable complexes. This is one possible explanation why protein-protein interface properties can be expected to be different in protein∶protein and in protein∶protein∶DNA complexes. A possible implication is that (if properties are similar or the same) actually two DNA-binding proteins bind first to each other and then bind to DNA together (as a complex). A similar hypothesis can be derived for protein-DNA interfaces in protein∶DNA and in protein∶{protein+}∶DNA complexes. One might suppose that these interfaces can be different, because when one protein binds to DNA there is a higher degree of freedom (rotational, translational) than when one protein should bind to a previously-made protein∶DNA complex. This is useful (from a theoretical point of view) for better understanding protein-DNA interactions which frequently involve complexes of multiple proteins. In addition, this can be useful (from a practical point of view) for the possible modelling of such complexes (their prediction, prediction of order of processes, modelling cis-regulatory modules, etc). In addition the nature of protein-protein interface and protein-DNA interface might be different that there is no any competition between them. This aspect can be also considered with this kind of analysis performed in this paper. In this work we have also calculated and compared, the conformational change of DNA in binary complexes (i.e. single protein-DNA complexes) and ternary/quaternary complexes (protein-protein-DNA/protein-protein-protein-DNA). Next, we analyzed protein-protein and protein-DNA energy binding affinity in protein-protein, single protein-DNA and multiple proteins-DNA complexes using several different tools. In addition, we analyzed and compared the thermodynamic stabilities of these complexes. We have provided an algorithm, and its web-based implementation, for calculating overlapping interface volumes and the number of interface atoms in collision between any two components (macromolecules) from a 3D complex stored in a pdb file.Results and DiscussionWe have performed computational structural analysis and present herewith some general features we have observed about macromolecular assemblies of multiple proteins bound to DNA. The following tools were used in our analysis: PISA [25], [26]; PROMOTIF [27]; X3DNA [28]; ReadOut [29]; DDNA [30] and DCOMPLEX [31]. Additionally, we have developed and used an algorithm for collision detection and overlapping volume of two macromolecules. Web-base implementation of the algorithm is freely available from http://promoterplot.fmi.ch/Collision1/ (see Materials and Methods for details). All data sets, used in this study, are from the PDB database (see Materials and Methods for a definition of data sets used in this study).Physical properties of interfacesDo physical properties of interfaces depend on the number of units in macromolecular assemblies? Are there any differences in physical properties of interfaces among protein∶protein∶DNA, protein∶DNA and protein∶protein complexes? In order to answer these questions, we performed analysis of physical interface properties of different macromolecular assemblies.The number of interfaces in the dataset MutliProteins∶DNA together with their structural characteristics is summarized in Table 1.10.1371/journal.pone.0003243.t001Table 1Descriptive statistics of interfaces.Interface typeNumber of interfacesAverage size of interface (Å2)±SEAverage number of interface residues*±SEAverage number of interface atoms*±SEAverage number of intermolecular H-bonds±SEAverage number of intermolecular salt bridges±SEProtein-protein52929.84±179.449.5±8.4190.9±36.09.36±3.74.08±0.7DNA-protein871002.3±56.552.2±2.9222.2±12.518.0±1.10.0±0.0Descriptive statistics of protein-protein and protein-DNA interfaces of complexes from group-MultiProteins∶DNA.*For both components together in interface.A detailed list of 52 protein-protein and 87 protein-DNA interfaces is given in Table S1. These values represent the sample sizes for the following hypothesis tests between protein-protein and protein-DNA interactions: There was no significant difference in average interface surface sizes (student's t-test, p-value = 0.69); nor the average number of interface residues (student's t-test, p-value = 0.76) nor the average number of atoms (p-value = 0.41). Based on this we can conclude that protein-protein and protein-DNA interfaces have similar average sizes and numbers of residues/atoms involved in their interactions in protein∶protein∶DNA complexes. La Conte et al. [17] found that most protein-protein interface areas are in the range of 1200–2000 Å2. They consider the total area on both components (without dividing by 2 to make the average area) as shown in formula (2). The protein-protein and protein-DNA interface areas for protein∶protein∶DNA complexes are also to this range (Table 1). The average area of protein-protein interfaces of complexes in the group-MultiProteins∶DNA and the average area of protein-protein interfaces of complexes in the group-Protein∶Protein we observe was comparable to those reported by Chakrabarti and Janin [9]. The DNA interface area sizes reported in Table 1 are comparable with those reported in studies considering only single protein-DNA complexes [15], [21]. The number of residues/atoms in protein-protein interfaces in this study was also comparable to previous studies [9], [17]. The situation is similar if we compare protein-DNA interfaces of protein∶protein∶DNA complexes with protein-DNA interfaces of protein∶DNA complexes [15], [21].Based on this we can conclude that average interface size and the average number of interfaces residues/atoms between two macromolecules (DNA, protein) in any kind of complex (protein∶protein, protein∶DNA, protein∶protein∶DNA) are approximately the same. In addition, it appears that these physical properties are not influenced by the number of subunits in the complex.Distribution of hydrogen bonds in interfacesThe purpose of this section was to investigate differences in distributions of hydrogen bonds between interfaces of macromolecular assemblies. There is a statistically significant difference in the average number of intermolecular hydrogen bonds (H-bonds) between protein-protein and DNA-protein interfaces (student's t-test, p-value<0.0001). The number of H-bonds observed in previous protein-protein studies (mean 10.1±0.5) [17] is comparable to those reported in this study for group-MultiProteins∶DNA (Table 1). The situation is similar if we compare protein-protein-DNA verses protein-DNA interfaces [15], [21]. The small observed variations are due to small variations in the interface areas as the number of hydrogen bonds is dependent on this area.In Table S2 we report the numbers of hydrogen bonds observed between the 20 amino acids and the four bases or the backbone of the DNA for the complexes listed in the group-MutliProteins∶DNA. We found that H-bond pairs were significantly different from random (Fisher's test, p<10−6). The most favoured amino acid-DNA base H-bond is ARG-G. In Figure S1 we report the distribution of H-bonds between the DNA bases and the bound proteins in group-MutliProteins∶DNA. 65.69% of all H-bonds where between protein side chains and the DNA backbone (Figure S1). Those H-bonds are not expected to confer specificity of binding but rather assist in complex stability. Most amino acids involved in H-bonds between the proteins and DNA (complex from group-MultiProteins∶DNA) are positively charged, presumably because of the negative charge of DNA (Figure S2). For the H-bonds at the protein-protein interfaces, the situation is different: negative and positively charged amino acids have an approximately equal frequency due to the need to pair charges in electrostatic interactions between donator and acceptor sites in the two proteins. Very similar distributions of H-bonds are found in groups –SingleSameProtein∶DNA and –SubSetMultiProteins∶DNA (Table S3, Table S4, Figure S3, Figure S4).Most H-bonds (53.3%) are made with phosphate groups of the DNA at the protein∶DNA interfaces. Very few H-bonds (12%) are made with deoxyribose (Figure S1). This situation is the same as that reported by Lejeune et al. [16] and Luscombe et al. [18] for protein-DNA interactions. The distribution of H-bonds between the participating amino acids and the DNA is given in Table S2. Entries in Table S2 that diverge from the expected distribution (favoured amino acid-base H-bonds) are also similar to those observed by Luscombe et al. [18].Distributions of interface residuesIn this section we present results about distributions of interface residues. We investigate if distributions of interface residues dependent on the number of units in the complex and if there are any differences in residue distributions between binary and ternary complexes (protein∶protein∶DNA, protein∶DNA, protein∶protein). The amino-acid propensities for the protein-protein and protein-DNA interfaces for complexes from the group-MultiProteins∶DNA are shown in Figure S5. For protein-DNA interfaces, ARG and LYS have the highest propensity values (>1.2), which indicates that they occur greater than 20% higher frequently in the interfaces than in the whole dataset. On other hand, many amino acids (ALA, ASP, CYS, GLN, GLU, ILE, LEU, MET, PHE, PRO, and VAL) are disfavoured in the interactions sites. For protein-protein interfaces, the situation is different and MET is the most favoured residue at interaction sites. In Figure S6 we report the distribution of amino acids involved in protein-protein and protein-DNA interfaces in the complexes from the group-MultiProteins∶DNA. Aliphatic amino acids are dominant in protein-protein interactions, while positively charged amino acids are the most involved in protein-DNA interactions. Those two distributions are significantly different, with a p-value<0.0001 (Chi-square multinomial test). The complexes in group-MutliProteins∶DNA have a number of van der Waals interactions between the amino acids in the proteins and either the DNA bases or backbone that is significantly different from random (Table S5, Fisher's p-value<5×10−6). In order to determine which of the pairings are different from expected, we performed individual Fisher's tests on each pair. The distributions of interface residues for protein-DNA interfaces of the complexes in the groups-SubSetMultiProteins∶DNA and –SingleSameProtein∶DNA are reported in Table S6 and Table S7.Protein-protein interfaces are more hydrophobic than protein-DNA interfaces (they contain significantly more aliphatic amino acids, see Figure S6 for details). Protein-protein interfaces have many more negatively charged amino acids and far fewer positively charged amino acids than protein-DNA interfaces. All these interface parameters give an indication of the overall polar nature of protein-DNA interfaces. Given that the DNA molecule surface is negatively charged, it is perhaps not surprising that it favours positively charged protein surface patches.The frequency distributions of amino acids in protein-DNA interaction sites in this study from the group-MultiProteins∶DNA are similar to those reported by Lejeune [16] (Figure S5 and Figure S6).Distribution of interface structural motifsWe investigated if the distributions of structural motifs in interfaces of components in ternary (protein∶protein∶DNA) complexes are different from those in binary complexes (protein∶protein and protein∶DNA). In order to answer on this question we calculate the propensity values for protein-protein and protein-DNA secondary structure motifs from the group-MultiProteins∶DNA (shown in Figure 1). The most favoured protein-DNA interface motif in is the helix, and the least favoured motifs are γ-turns, β-strands, and β-hairpins. At protein-protein interfaces, the least favoured secondary structure motif is the β-bulge. The distributions of secondary structure motifs between protein-protein and protein-DNA interfaces are significant different (Chi-square multinomial goodness-of-fit test, p-value<0.01). For protein-DNA interfaces, the dominant structural motif is the helix. This result is consistent with the observation that many DNA binding sites on proteins are comprised of helix motifs [32]. The distribution of secondary structure motifs in protein-protein interfaces for the complexes used in this study (group-MultiProteins∶DNA, Figure 1) is similar to that observed by Guharoy and Chakrabarti [33] who observed that the contribution of β-strands is lower than that of helixes and that non-regular structural motifs appear in large numbers.10.1371/journal.pone.0003243.g001Figure 1Secondary structure motif propensities.Secondary structure motif propensities for protein-protein and protein-DNA interfaces. Propensity values which are significantly different from 1 (either above or below), evaluated by the statistical bootstrapping method, are marked with “*”. Significant statistical differences between motif propensities of protein-protein and protein-DNA interfaces are marked with “#”.All previous results (from this and previous subsections) can be summarized in the form:(1)where Xprotein-protein (C) and Xprotein-DNA (C) represent one of the following interface parameters: area, number of residues, number of atoms, number of H-bonds, distribution of residues, distribution of H-bond partners or the distribution of structural interface motifs in either protein-protein or protein-DNA interfaces respectively where complex C is either a protein∶protein, a protein∶DNA or a protein∶protein∶DNA complex. Formula (1) can be easily be expanded to cover quaternary complexes (protein∶protein∶protein∶DNA) as well, but for clarity we have only represented the case for ternary complexes.It is apparent from formula (1) that interface parameters under discussion, for complexes composed of multiple proteins bound to DNA, can be estimated from protein-protein and single protein-DNA complexes alone. A more precise variant of formula (1), for example in the form of a regression equation, would be possible to derive if we had crystal structures of the same protein in all three states: protein∶protein; protein∶DNA and protein∶protein∶DNA.Our results indicate that the physical properties of protein∶protein and protein∶DNA complexes, such as interface area, number of interface residues/atoms and hydrogen bonds and the distribution of interface residues and secondary structure motifs are no different in binary or ternary complexes. Thus, if we have two (or more) proteins which bind together, there will be no influence on these interface parameters of their DNA-binding interface when they bind together as a complex to DNA. This claim is not related to the energy of these interactions and it is expected that the interaction rate constants will not be the same for binary and multiple proteins complexes. If two DNA binding proteins can also bind to each other then this will tether them in the vicinity of the DNA such that when one of the proteins binds to DNA the second will have a faster on-rate because it will have a shorter distance to diffuse to find its binding site thus maintain a higher effective local concentration around the DNA. A detailed analysis of rate constants cannot unfortunately be made from crystal structures which are by definition static snapshots of this dynamic process.Water molecules in protein-protein and protein-DNA interactionsIt has been discussed that water content and water mediated contacts in the protein-DNA interface are important components of protein-DNA interactions [34], [35]. Protein-protein and protein-DNA interfaces contain significant quantities of water [36]. Structural and biochemical data indicate that water-mediated interactions are important for the stability and specificity of recognition, despite the fact that interface solvent molecules exchange rapidly with the bulk solvent [36]. We wanted to evaluate the differences between water mediated contacts at protein-DNA interfaces in protein∶DNA complexes (single proteins bound to DNA) and in protein∶protein∶DNA complexes (multiple proteins bound to DNA). The average number of water mediated contacts between the protein-DNA interfaces of protein∶protein∶DNA complexes is ∼11.82±1.3 (Table S8). This is markedly different from the value of 28 reported for protein∶DNA complexes previously [36]. Similarly, we compared the water mediated contacts in the protein-protein interfaces of protein∶protein and protein∶protein∶DNA complexes. The average number of water molecules for protein-protein interfaces of complexes in the group-MultiProteins∶DNA was ∼4.9±0.83 (Table S8), as compared to ∼22 for protein-protein interactions in binary protein∶protein complexes reported by [36].These results suggest that water mediated contacts in interfaces of components in protein∶protein∶DNA complexes play less important role in the stability and specificity of recognition then in interfaces of components in the binary protein∶protein and protein∶DNA complexes. However, as we discussed later in the text there are other factors which are more important for stability and specificity of component recognition in protein∶protein∶DNA complexes.DNA distortionIn order to check if DNA structural deformation is higher when multiple proteins bind to DNA we performed computational structural analysis of DNA structures. DNA distortion was measured by calculating the root-mean-square deviation (rmsd) when each DNA structure was fitted onto its corresponding canonical A-DNA or B-DNA structure. Distributions of rmsd values for all complexes from the groups MultiProteins∶DNA (black bars) and SingleSameProtein∶DNA (white bars) were calculated (Figure 2). Statistical analysis of these results showed a significant difference in means of rmsd values (student's t-test with equal or unequal variance as appropriate, p-value<0.02) calculated for all complexes from the groups –MultiProteins∶DNA, -SingleProtein∶DNA and –SingleSameProtein∶DNA calculated after fitting each DNA structure onto the corresponding canonical A-DNA and B-DNA structures (Table 2). Further information for each complex is given inTable S9, S10, S11 and S12. The rmsd values for the group-SubMultiProteins∶DNA are the same as those for the group-MultiProteins∶DNA.10.1371/journal.pone.0003243.g002Figure 2Distribution of rmsd values for measuring DNA distortion.Distribution of rmsd values calculated from fitting each DNA structure in the complexes from group-MultiProteins∶DNA (black bars) and group-SingleSameProtein∶DNA (white bars) to a corresponding canonical B-DNA.10.1371/journal.pone.0003243.t002Table 2Measuring DNA distortion.Dataset of complexesAverage rmsd (±SE) from A-DNAAverage rmsd (±SE) from B-DNAGroup-MultiProteins∶DNA8.26±0.44.71±0.5Group-SingleProtein∶DNA5.94±0.2(p<0.001)3.44±0.2 (p = 0.007)#\nGroup-SingleSameProtein∶DNA6.66±0.6 (p = 0.02)2.87±0.4 (p = 0.004)#\nAverage rmsd values calculated from fitting each DNA structure in the complexes from group –MultiProteins∶DNA, -SingleProtein∶DNA, and –SingleSameProtein∶DNA to a corresponding canonical A-DNA and B-DNA.p-values are calculated in comparison with Group A and obtained using the one-tailed Student's t-test.#unequal variance.The rmsd values of the group SubSetMultiProteins∶DNA, including comparisons with the group SingleSameProtein∶DNA, are given in Table S13. DNA distortion, however, is significantly higher when multiple proteins are bound to the DNA (Figure 2, Table 2, Table S13). It has been reported that when a single protein binds to DNA it results in a higher rmsd (conformational change) than that seen in the unbound DNA structure [15]. Here we reported that there are also further conformational changes to the structure of DNA which are induced when multiple proteins bind to it.Energetic properties of interfacesThe energetic properties of cooperatives are useful for understanding of how the essential macromolecular machines of cellular function are assembled and how they work [37]. We analyzed energetic and thermodynamic properties of different mulitcomponent complexes (protein∶protein∶DNA, protein∶DNA, protein∶protein). In Table 3 we report the free energy of dissociation (ΔGdiss) and the free energy of solvation (ΔG\nint) in kJ/mol for complexes from the four groups –MultiProteins∶DNA, -SubMultiProteins∶DNA, -SingleProtein∶DNA, and –SingleSameProtein∶DNA. In Table 4 we also report energy Z-score values for direct and indirect readouts for the three groups –MultiProteins∶DNA, -SubMultiProteins∶DNA and –SingleProtein∶DNA. The p-values in Table 3 were obtained by comparing the means of ΔG\nint, ΔGdiss and the Z-scores for the direct and indirect readouts using the student's t-test (with equal or unequal variance as appropriate). We could not calculate energy Z-scores for the indirect readouts of the group SubMultiProteins∶DNA because the DNA structure is the same for each complex, so the calculated Z-scores would also be the same. Detailed lists of the ΔG\nint, ΔGdiss and Z-scores for both the direct and indirect readouts of each complex and each group are available in Table S14, S15, S16, S17, S18, S19, S20, S21, S22 and S23.10.1371/journal.pone.0003243.t003Table 3Complex energies.Dataset of complexesAverage (±SE) solvation energy ΔG\nint (kJ/mol)Average (±SE) ΔGdiss (kJ/mol)Average (±SE) energy Z-score for direct readoutAverage (±SE)energy Z-score for indirect readoutGroup-MultiProteins∶DNA−234.61.03±18.450.41±6.0−2.81±0.2−2.36±0.1Group-SubMultiProteins∶DNA−123.21±9.8 (p<0.001)#\n47.19±4.9 (p = 0.34)−1.71±0.2 (p<0.001)—Group-SingleProtein∶DNA−114.49±8.6 (p<0.001)#\n48.52±5.3 (p = 0.41)−1.84±0.3 (p = 0.005)#\n−2.14±0.1 (p = 0.13)Group-SingleSameProtein∶DNA−99.79±15.0 (p<0.001)#\n31.06±6.5 (p = 0.03)−1.34±0.3 (p<0.001)#\n−1.48±0.3 (p = 0.007)Average solvation energy (kJ/mol), free energy barrier of assembly dissociation (kJ/mol), and energy Z-scores for direct and indirect readouts for groups –MultiProteins∶DNA, -SubMultiProteins∶DNA, -SingleProtein∶DNA and –SingleSameProtein∶DNA.p-values are calculated in comparison with Group-MultiProteins∶DNA and obtained using the one-tailed Student's t-test.#unequal variance.10.1371/journal.pone.0003243.t004Table 4Affinity of components.Dataset of complexesAverage (±SE) protein-DNA energy binding affinity (kJ/mol)Average (±SE) protein-DNA overlapping volume (Å3)Average (±SE) number of atoms in collision in protein-DNA interfacesGroup-MultiProteins∶DNA−39.05±0.94.26±0.832.06±4.1Group-SubMultiProteins∶DNA−30.93±0.5 (p<0.001)#\n2.04±0.3 (p = 0.007)#\n15.44±1.9 (p<0.001)#\nGroup-SingleProtein∶DNA−33.20±0.6 (p<0.001)3.17±0.56 (p = 0.13)20.45±1.8 (p = 0.006)#\nGroup-SingleSameProtein∶DNA−32.79±0.9(p<0.001)#\n2.313±0.8 (p = 0.04)#\n15.5±3.3 (p = 0.001)#\nAverage protein-DNA energy binding affinity (kJ/mol), interface overlapping volume (Å3) and average number of interface collision atoms for groups –MultiProteins∶DNA, -SubMultiProteins∶DNA, -SingleProtein∶DNA and –SingleSameProtein∶DNA.p-values are calculated in comparison with Group-MultiProteins∶DNA and obtained using the one-tailed Student's t-test.#unequal variance.\nTable 4 shows the average protein-DNA energy binding affinity in kJ/mol for the MultiProteins∶DNA, SubMultiProteins∶DNA, SingleProtein∶DNA and SingleSameProtein∶DNA groups; the average protein-DNA overlapping volume (in Å3) and the number of atoms in collision at the protein-DNA interfaces. All values were compared against the MultiProteins∶DNA group and a student's t-test was used to calculate the p-values. Further information on these parameters can be found in Table S24, S25, S26, S27 and S28.The average protein-protein binding energy for complexes from the MultiProteins∶DNA group (which are bound to DNA) is significantly smaller (student's t-test, p-value = 0.05) than that of complexes from group-Protein∶Protein (Table 5). The average solvation energy (ΔG\nint) and free energy barrier of assembly dissociation (ΔGdiss) for protein-protein complexes from group–MultiProteins∶DNA is, respectively, smaller and larger (student's t-test, p-value<0.001) than that found for complexes from group-Protein∶Protein (Table 5). A list of protein-protein binding affinities for every complex in the MultiProteins∶DNA and Protein∶Protein groups may be found in Table S29–S30.10.1371/journal.pone.0003243.t005Table 5Protein-protein interfaces energies.Dataset of complexesAverage (±SE) protein-protein binding free energy (kJ/mol)Average (±SE) solvation energy ΔG\nint (kJ/mol)Average (±SE) ΔGdiss (kJ/mol)Group-MultiProteins∶DNA−56.27±6.3−234.61.03±18.4*\n50.41±6.0*\nGroup-Protein∶Protein−67.20±2.3 (p = 0.05)#\n−81.937±10.1 (p<0.001)#\n8.22±2.9 (p<0.001)#\nAverage protein-protein binding free energy (kJ/mol), average solvation energy (kJ/mol) and average free energy barrier of assembly dissociation (kJ/mol) for protein-protein complexes from group –MultiProteins∶DNA and –Protein∶Protein.p-values are calculated in comparison with Group-MultiProteins∶DNA and obtained using the one-tailed Student's t-test.#unequal variance.*calculated for the whole complex (the same values as in Table 3).The energetic properties of protein-DNA interfaces of the complexes in group-SubSetMultiProteins∶DNA, including their comparisons with corresponding values from group-SingleSameProtein∶DNA, are given in Tables S31 and S32.The free energy barrier of assembly dissociation (ΔGdiss, Table 3) is higher for complexes involving multiple proteins bound to DNA (MultiProteins∶DNA) than those involving only single protein-DNA complexes (SubMultiProteins∶DNA, SingleProtein∶DNA and SingleSameProtein). The SingleSameProtein∶DNA and the SubMultiProteins∶DNA groups both contain proteins which are also components of the complexes found in the MultiProteins∶DNA group, but the SubMultiProteins∶DNA group was formed by manually removing the extra protein units from the complexes of group-MultiProteins∶DNA in order to get single protein-DNA complexes. We see that in comparison with the SingleSameProtein∶DNA group, complexes in the MultiProteins∶DNA group have significantly (p = 0.03, student's t-test) higher free energy barriers of assembly dissociation (ΔGdiss). This means that multiple proteins-DNA complexes are more thermodynamically stable than single protein-DNA complexes. Comparing the MultiProteins∶DNA group to the three other groups (SubMultiProteins∶DNA, SingleProtein∶DNA, and SingleSameProtein∶DNA), we find a significantly smaller free energy (student's test, p-value<0.001, Table 3) of solvation gain upon complex formation (ΔG\nint). The same result was found when comparing the MutliProteins∶DNA group to the SubSetMultiProteins∶DNA group (Table S31).The energy Z-scores for direct and indirect readouts (conformational energy) have more negative values for complexes with multiple proteins bound to DNA (Table 3 and Table S31). More negative Z-scores mean that the target DNA sequence fits into a given protein structure better [29]. Therefore, DNA-binding proteins fit their targets better when they form a ternary complex with DNA. The Z-score also indicates that ternary complexes may be more stable than binary ones. The binding energy affinity, overlapping volume and number of atoms in collision (Table 4) is significantly higher in protein-protein-DNA complexes than in protein-DNA complexes. Differences in overlapping volume and number of atoms in collision are due not only to the bigger interface area (twice protein∶DNA), but also to the higher affinity of multiple proteins binding (interface area sizes for the SingleProteins∶DNA, SingleSameProteins∶DNA and –SubMultiProteins∶DNA groups are similar, butthe SingleProtein∶DNA and SingleSameProtein∶DNA groups have higher protein-DNA binding affinities, overlapping volumes and numbers of atoms in collision than those in the SubMultiProteins∶DNA group, Table 4 and Table S32). Cis-modules that contain transcription factor binding sites (cis-motifs) of transcription factors which make direct physical contact with each other have higher DNA-binding affinities than cis-modules that contain transcription factor binding sites (cis-motifs) of factors without direct mutual contacts. This information may be used for the prediction of cis-regulatory motifs/modules in the following way: if we say that the value of a scoring function for binding sites which are close to one another (where there might be the physical contact between corresponding transcription factors) may have a lower threshold value than a threshold which should be used for scoring function for binding sites that are further away (where there might not be the physical contact between corresponding transcription factors). Modelling DNA∶protein∶protein∶DNA interactions caused by the bending of DNA would also be a possible explanation for introducing a similar strategy; however, there is still not enough information for computational modelling of DNA-bending (i.e. there are not yet any computational strategies which can predict when two transcription factors which are bound to DNA with a long distance between them would have direct physical contact as a consequence of DNA bending). In addition to that, another important implication for the prediction of CRM or cis-motifs is the overlap between transcription factors which have binding sites close to each other. Based on our collision detection results, we realized that sometimes when transcription factors bind to the different grooves of DNA (major and minor) their binding sites can overlap a lot, but from a 3D point of view there is no physical overlap between factors. On the other hand, if two transcription factors bind to the same groove (usually major) then there can be a large overlap between them from a 3D point of view if there is a large overlap between their binding sites (i.e. this situation is not possible). In other words, if care is taken about the structural classification of transcription factors (i.e. if they bind to the major or minor groove) this information can also be used for CRM or cis-motif predictions.It is interesting to note that protein-protein affinities are higher when proteins are not bound to DNA (Table 5). Interfaces between proteins that are part of a multi-complex (with DNA) can be weaker than those found in binary ones. Binding to DNA may decrease protein-protein affinities, while increasing the overall stability of the complex (significantly higher stability, student's test, p<0.001, Table 5). When two proteins bind freely in solution they are largely unhindered in their rotational movement so they can align themselves using the most energetically favourable orientation which gives them the optimal protein-protein binding energy. When DNA is added to the complex, the three components must arrange themselves to form a global energy minima. However the requirement of binding to DNA introduces a restriction on the possible arrangement of the components such that the protein-protein binding may be weakened by this extra strain but the additional synergistic stability of the three way complex more than compensates for this effect (Table 5).ConclusionIt is very difficult to determine the rules governing the assembly of complexes by data-mining alone [38]. Universal conclusions for the types of complexes used are unreliable because of the limited number of available structures (44). However, many general descriptive features can be elucidated even with a modest data collection. As further structures become available, the confidence in the results presented here can be further constrained. The precedent for such studies, using similar or even smaller number of structures is well documented (e.g. [10], [15], [19], [23]).In this paper, we conclude that protein-protein and protein-DNA interface parameters, such as interface area, number of interface residues/atoms and hydrogen bonds, and distribution of interface residues, hydrogen bonds, van der Walls contacts and secondary structure motifs in complexes where multiple proteins are bound to DNA are no different in protein-protein, single protein-DNA or multiple proteins-DNA complexes. Thus, if we have two (or more) proteins which bind together, there will be no influence on these interface parameters. Also, if we have one protein bound to DNA, then that binding will have no influence (in terms of the interface parameters mentioned) on the types of interface interactions that can occur with subsequent protein-protein complex expansion. The water mediated contacts in interfaces of components in protein∶protein∶DNA complexes play less important role (found in less quantity) in the stability and specificity of recognition then in interfaces of components in the binary protein∶protein and protein∶DNA complexes. Distortion is significantly higher when multiple proteins bind to DNA. This distortion is required to accommodate multiple protein binding events. The combinatorial assembly of transcription factors has been known for a long time to play an important role in stabilizing regulatory complexes. A deeper understanding of structural considerations may be helpful when predicting the assembly of transcription factor complexes. The formation of multiple protein interactions with DNA results in a decrease in protein-protein affinity and an increase in protein-DNA affinity with a net gain in overall stability for a protein-protein-DNA complex. Such effects are clearly important for modelling transcription factor cooperativity.Materials and MethodsDefinition of data setsWe selected 75 crystal complexes from the PDB database which contained two or more proteins bound to DNA with a resolution of 3.25 Å or less. We discarded all homologous complexes with less than 30% protein sequence for all protein components using the PISCES server [39], [40]. Our final dataset contained 46 complexes (Table S33). We determined the UniProt ID of each protein component using the tool [41]. This dataset was called group-MultiProteins∶DNA. Most of the complexes from group-MultiProteins∶DNA are ternary (two proteins bound to DNA), but a few of them are quaternary (three proteins bound to DNA). A very few of them contain one protein which does not make contact with DNA but is bound to another protein which does have a direct contact with DNA. We created a second dataset (group-SubMultiProteins∶DNA) from group-MultiProteins∶DNA which consisted of 91 structures (this number is smaller than 92, because some of the proteins do not have direct contact with DNA), each of which was a sub-structure containing only one protein unit plus DNA. In addition, we analysed a set (group-SingleProtien∶DNA, Table S34) of single protein-DNA complexes (102 structures), which was a subset derived from a previous study [16]. We found 17 PDB structures (group-SingleSameProtein∶DNA, Table S35) which contained single proteins and DNA, but the proteins were all components of complexes in group-MultiProteins∶DNA. Corresponding subgroup of group-MultiProteins∶DNA which contains complexes for each where there is a partner in the SingleSameProtein∶DNA group we call this group-SubSetMultiProteins∶DNA (Table S36). The group-Protein∶Protein (Table S37), which contained 70 protein-protein complexes, came from a previous study [9].Physical and chemical analysis of interfacesWe used the PISA service from the European Bioinformatics Institute [25], [26] to calculate interface areas and compositions. There are two possibilities for defining the interface between two macromolecular components: the first approach defines the interface as the protein surface area which becomes inaccessible to solvents when two chains come into contact; the second method defines the interface as the set of atoms, where the atom centers from different proteins lie within a distance of 1–5 Å. Both approaches are widely used in macromolecular complex analysis and produce roughly equivalent results. The PISA service uses the first approach. The interface area between macromolecular components M1 and M2 is calculated as the difference in total accessible surface areas of isolated and interfacing structures divided by two, i.e.:(2)where ASA(M1) and ASA(M2) are the accessible surface areas of macromolecular components M1 and M2 respectively, and ASA(M1M2) is the accessible surface area of the complex of M1 and M2.We also used the PISA service to calculate hydrogen bonds, salt bridges, disulphide bonds and interface residues. However, PISA provides no information about van der Waals contacts between atoms (residues) because they may be in contact with several other residues. This is the principal difference between the outputs for van der Waals and hydrogen bonds, where inter-atomic links are well determined. However, in order to produce results comparable with previous studies, we have calculated van der Waals contacts in the following way: all atoms not involved in hydrogen bonds but separated by 3.9 Å or less are considered to be interacting through van der Waals contacts [18]. We also analyzed the statistical distribution of amino acid-amino acid and amino acid-nucleotide pairs (“interaction matrices”) for hydrogen bonds and van der Waal contacts. For all amino acid-amino acid and amino acid-nucleotide pairs we calculated contingency tables. The expected values for these tables are based on an assumption of random interactions. We evaluated the contingency tables using Fisher's exact test for count data with simulated p-values based on 200000 repetitions (GNU R). The p-value obtained by Fisher's exact test indicates whether rows and columns in contingency tables are independent or not. However, this does not provide information about which of the pairings are different from expected. To calculate this we performed individual Fisher's tests (GNU R) for each pair.In order to determine the chemical characteristics of the interfaces, we classified the interface residues using Eisenberg's hydrophobicity scale [42] in a similar way to Lejeune et al. [16]: amino acids are assigned to groups which contain those that are positively charged (Arg and Lys), negatively charged (Asp and Glu), polar (Asn, Gln, His, Ser, and Thr), aliphatic (Ala, Ile, Leu, Met and Val), aromatic (Phe, Trp, and Tyr), and particular (Cys, Gly, and Pro). Multinomial distributions obtained in this study were compared using the Chi-square multinomial goodness-of-fit test.In addition, a general indication of the hydrophobicity of the interfaces can be estimated using the residue interface propensities. The residue interface propensities give a measure of the relative importance of different amino acid (nucleic acid) residues in all the interfaces of complexes. The propensity values can be calculated using the accessible surface area of residues, as was done by Ellis et al. [10], or using the frequencies of residues, as was done by Lejeune et al. [16]. Both approaches have the same goal, to determine the relative importance of the different residues. Because of its simplicity, we have used the approach described in [16]. Following that, the propensity Px for the interface residues x (x and y are amino acid or DNA structures) can be calculated by:(3)where Ix is the total number of residues x in the interface area, Tx is the total number of residues in the whole dataset and similar for Ty and Iy. If Px>1 it indicates that the residue x is “favoured” and occurs more frequently at interfaces than in the dataset as a whole. If Px<1 then residue x is “disfavoured” at interaction sites; in all other cases we can say that residue x is neither over- nor under-represented in the interface region in the complexes. In order to evaluate whether a particular propensity value was significantly different from 1 (either above or below), a statistical bootstrapping method was implemented similar to [10].Structural analysis of interfacesWe analyzed the types of secondary structures present within protein-protein and protein-DNA interfaces using the PROMOTIF program [27]. PROMOTIF defines 11 different secondary structure motifs: β-turns, γ-turns, β-bulges, α-helices, 310-helices, β-strands, β-sheets, βαβ units, ψ-loop, β-hairpins, and disulphide bridges. For each structural motif we calculated propensities in the same way as we did for residue propensities (formula (3)).Analysis of DNA distortionDNA distortions were estimated by calculating the root-mean-square deviation (rmsd) when each DNA structure from a complex was fitted onto the corresponding canonical A-DNA and B-DNA structures as in [15], using the whole DNA from crystal strucutres and without normalization to the length of the DNA used. (Regions which are not in interactions do not have significant deformation therefore their contributions to RMSD is not big.) Canonical A-DNA and B-DNA for the nucleotide sequence (with the same length) from the complex were constructed using X3DNA [28]. The fitting was performed with the McLachlan algorithm [43] as implemented in the program ProFit [44].Analysis of water molecules in protein-protein and protein-DNA interactionsWater molecules are defined as interface water molecules if they are less than 3.5 Å from the atoms of the two components of a complex, as in [21]. This analysis was restricted to those structures with 2.4 Å or better resolution as the identification of water in the electron density map may be ambiguous at lower resolutions [21].Analysis of energetic properties of interfacesThe chemical stability of complexes was analysed by calculating the free energy barrier of assembly dissociation (ΔGdiss) and the solvation free energy gain upon formation of the assembly (ΔG\nint) in kJ/mol using PISA. Assemblies with higher positive values of ΔGdiss are more thermodynamically stable, and that value indicates that an external driving force is required to dissociate the assembly. For the calculation of ΔG\nint and ΔGdiss we used structures from all six groups (-MultiProteins∶DNA, -SubMultiProteins∶DNA, -SingleProtein∶DNA, -SingleSameProtein∶DNA, -SubSetMultiProteins∶DNA and –Protein∶Protein).We calculated Z-scores for intermolecular and intramolecular readouts using a ReadOut server [29]. Direct readouts (direct contacts between amino acids and base pairs) and water-mediated contacts are intramolecular energies, whereas indirect energies quantify sequence-dependent DNA conformational energies. The specificity of the complex is given by the Z-score, and larger negative values correspond to higher specificities [45]. For the calculation of the Z-score, we used the data from groups –MultiProteins∶DNA, -SubMultiProteins∶DNA, -SingleProteins∶DNA, -SingleSameProtein, -SubSetMultiProteins∶DNA.We calculated binding energy affinities (protein-DNA) for each structure in groups –MultiProteins∶DNA, -SubMultiProteins∶DNA, -SingleProtein∶DNA, -SingleSameProtein∶DNA, and –SubSetMultiProteins∶DNA using the DFIRE energy function [30].We compared the mean of ΔG\nint, ΔGdiss, the Z-score for direct and indirect readouts, and the binding energy affinities between group-MultiProteins∶DNA and each of the other three groups (-SubMultiProteins∶DNA, -SingleProtein∶DNA and –SingleSameProtein∶DNA) using student's t-test (one-tailed). Differences in the variances of corresponding values between groups were calculated using Bartlett's test. In those cases where we had significant differences in variance between groups, we used student's t-test with unequal variance.For protein-protein complexes (group-Protein∶Protein) we calculated ΔG\nint and ΔGdiss using the PISA server. We have calculated protein-protein binding energy affinities for complexes from group-Protein∶Protein and protein-protein subcomplexes from group-MultiProteins∶DNA using DCOMPLEX [31]. We also compared the average protein-protein binding affinities, average values of ΔG\nint and ΔGdiss between groups –MultiProteins∶DNA and –Protein∶Protein.Collision detections and overlapping volume of two macromoleculesWe calculated the number of atoms in collision and the volume of the overlapping region for protein-protein and protein-DNA interfaces from groups –MutliProteins∶DNA, -SubMultiProteins∶DNA, -SingleProtein∶DNA and –SingleSameProtein∶DNA. Collision detection between two macromolecules is actually collision detection between complex objects, where these objects are composed of collections of spheres. The most straightforward algorithm for modelling this problem (in the case of two objects: A1 and A2) is checking each sphere from object A1 against each sphere from object A2, and we know that objects A1 and A2 intersect only if one or more of these pairs intersect. For two objects with M and N spheres this algorithm requires O(MN) time to complete. There are several geometric algorithms with better speed for collision detection between objects in 3D space such as those based on bounding-volume (BV) hierarchies [46], [47], algorithms based on axis-aligned bounding boxes AABB [48], [49], algorithms based on oriented bounding boxes [50], and spatial hashing [51], [52]. In this study we used an algorithm for collision detection based on spatial hashing [51] and axis-aligned bounding boxes AABB [48], [49]. To perform this, we executed the following steps (Figure S7):Make an AABB around each macromolecule.Check if any pair of AABBs overlaps. In order for two AABBs to overlap they must overlap on all three special axes. If there is no overlap then they cannot be in collision. Otherwise they may be in collision.Perform a special hashing on the overlapping region of each pair of AABBs that contain macromolecules that may be in collision.The overlapping region (a rectangular prism) is divided into a three dimensional grid of cells. Each cell in the grid is a cube with side lengths equal to the diameter of the largest sphere (atom) in the macromolecule. This is a uniform spatial subdivision. Each sphere (atom) in the macromolecule can be assigned to the cell in which it lies using a hash function as follows: First it is necessary to make an AABB for each sphere. Then the (x,y,z) coordinates of the six side centers are assigned to their corresponding cells using the hash function (Figure 3).10.1371/journal.pone.0003243.g003Figure 3Assignment of hash values to the atoms of a macromolecule.Hash values are computed for all the grid cells covered by the AABB of the sphere (atom) from a macromolecule. In this case, sphere S falls into four cells and they are mapped onto a hash table.The hash function we used is given in formula (4) [52]:(4)where p1, p2, and p3 are large prime numbers (in our case 73856093, 19349663 and 83492791 respectively). The size of a cell is defined as 1, the hash table has a size “n”. The function “trunc(x)” rounds the real number “x” down to the next integer. The function “xor” is a Boolean exclusive-or operation.To test whether a sphere “S” from another macromolecule intersects with the first macromolecule, it suffices to find out if that sphere intersects any of the spheres of another macromolecule that share a cell with “S”. The time complexity of this algorithm is linear “O(n)”, where “n” is the number of sphere-atoms found in the overlapping region between two macromolecules AABBs.We extended the collision detection algorithm so that it is able to calculate the number of atoms which are in collision and their overlapping volume. Instead of stopping the analysis as soon as two atoms are found to be in collision, the algorithm is continued until all of the atoms from the different macromolecules have been counted. From this it is a simple matter to estimate the overlapping volume from the colliding spheres.Web-base implementation of the algorithm is freely available from http://promoterplot.fmi.ch/Collision1/. The user submits pdb files and then specifies which chains to test for collision. The output lists the number of atoms from each protein which are in collision and the volume of overlapping region. In addition, with this tool user may display 3D complex from PDB files as interactive web pages using the Corotna VRML Client plug-in or any other VRML plug-in.Supporting InformationFigure S1Distribution of H-bonds according to the nucleotide part (group-MultiProteins∶DNA).(0.91 MB TIF)Click here for additional data file.Figure S2Distribution of amino acids involved in H-bonds in protein-protein and protein-DNA interfaces (group-MultiProteins∶DNA).(0.93 MB TIF)Click here for additional data file.Figure S3Distribution of H-bonds according to the nucleotide part (group-SingleSameProtein∶DNA).(0.91 MB TIF)Click here for additional data file.Figure S4Distribution of H-bonds according to the nucleotide part (group-SubSetMultiProteins∶DNA).(0.91 MB TIF)Click here for additional data file.Figure S5Amino acid propensities for protein-protein and DNA-protein interfaces (group MultiProteins∶DNA). Propensity values which are significantly different from 1 (either above or below), as evaluated using the statistical bootstrapping method, are marked with “*”.(1.08 MB TIF)Click here for additional data file.Figure S6Distribution of amino acids involved in interaction sites of protein-protein and DNA-protein (group-MultiProteins∶DNA).(1.07 MB TIF)Click here for additional data file.Figure S7Visualization of first several steps of the collision detection algorithm. Situation (A) represents scenario when there is on overlapping between two macromolecules and corresponding axis-aligned bounding boxes either; situation (B) represents scenario when there is no overlapping between two macromolecules but with overlapping between corresponding axis-aligned bounding boxes; situation (C) represents scenario when there is overlapping between two macromolecules and corresponding axis-aligned bounding boxes.(3.00 MB TIF)Click here for additional data file.Table S1Detailed list of interface parameters for each complex from group-MultiProteins∶DNA(0.09 MB PDF)Click here for additional data file.Table S2The number of observed hydrogen bonds between amino acid and nucleotide moieties in protein-DNA interfaces (group-MultiProteins∶DNA)(0.07 MB DOC)Click here for additional data file.Table S3The number of observed hydrogen bonds between amino acid and nucleotide moieties in protein-DNA interfaces (group-SingleSameProtein∶DNA)(0.07 MB DOC)Click here for additional data file.Table S4The number of observed hydrogen bonds between amino acid and nucleotide moieties in protein-DNA interfaces (group-SubSetMultiProteins∶DNA).(0.06 MB DOC)Click here for additional data file.Table S5Number of observed van der Waals contacts between amino acid and nucleotide moieties in protein-DNA interfaces (group-MultiProteins∶DNA).(0.06 MB DOC)Click here for additional data file.Table S6Number of observed van der Waals contacts between amino acid and nucleotide moieties in protein-DNA interfaces (group-SingleSameProtein∶DNA).(0.07 MB DOC)Click here for additional data file.Table S7Number of observed van der Waals contacts between amino acid and nucleotide moieties in protein-DNA interfaces (group-SubSetMultiProteins∶DNA).(0.06 MB DOC)Click here for additional data file.Table S8The number of water-mediated contacts in protein-protein and protein-DNA intrerfaces of selected complexes in group-MultipleProteins∶DNA(0.04 MB PDF)Click here for additional data file.Table S9Detailed list of rmsd values calculated from fitting each DNA structure in the complexes from group-MultiProteins∶DNA to a corresponding canonical A-DNA and B-DNA.(0.04 MB PDF)Click here for additional data file.Table S10Detailed list of rmsd values calculated from fitting each DNA structure in the complexes from group-SingleProtein∶DNA to a corresponding canonical A-DNA and B-DNA.(0.04 MB PDF)Click here for additional data file.Table S11Detailed list of rmsd values calculated from fitting each DNA structure in the complexes from group-SingleSameProtein∶DNA to a corresponding canonical A-DNA and B-DNA.(0.03 MB PDF)Click here for additional data file.Table S12Detailed list of rmsd values calculated from fitting each DNA structure in the complexes from group-SubSetMutliProteins∶DNA to a corresponding canonical A-DNA and B-DNA.(0.04 MB PDF)Click here for additional data file.Table S13Average rmsd values calculated from fitting each DNA structure in the complexes from group -SubSetMultiProteins∶DNA and -SingleSameProtein∶DNA to a corresponding canonical A-DNA and B-DNA.(0.03 MB DOC)Click here for additional data file.Table S14Detailed list of energies for each complex in group-MultiProteins∶DNA(0.04 MB PDF)Click here for additional data file.Table S15Detailed list of energies for each complex in group-SubMultiProteins∶DNA(0.04 MB PDF)Click here for additional data file.Table S16Detailed list of energies for each complex in group-SingleProtein∶DNA(0.04 MB PDF)Click here for additional data file.Table S17Detailed list of energies for each complex in group-SingleSameProtein∶DNA(0.04 MB PDF)Click here for additional data file.Table S18Detailed list of energies for each complex in group-SubSetMultiProteins∶DNA(0.04 MB PDF)Click here for additional data file.Table S19Detailed list of energies Z-scores (direct and indirect readouts) for each complex in group-MultiProteins∶DNA(0.04 MB PDF)Click here for additional data file.Table S20Detailed list of energies Z-scores (direct and indirect readouts) for each complex in group-SubMultiProteins∶DNA(0.04 MB PDF)Click here for additional data file.Table S21Detailed list of energies Z-scores (direct and indirect readouts) for each complex in group-SingleProtein∶DNA(0.04 MB PDF)Click here for additional data file.Table S22Detailed list of energy Z-scores (direct and indirect readouts) for each complex in group-SingleSameProtein∶DNA(0.04 MB PDF)Click here for additional data file.Table S23Detailed list of energy Z-scores (direct and indirect readouts) for each complex in group-SubSetMultiProteins∶DNA(0.04 MB PDF)Click here for additional data file.Table S24Detailed list of protein-DNA energy binding affinity, overlapping volume and number of atoms in collision for each complex in group-MultiProteins∶DNA(0.04 MB PDF)Click here for additional data file.Table S25Detailed list of protein-DNA energy binding affinity, overlapping volume and number of atoms in collision for each complex in group-SubMultiProteins∶DNA(0.05 MB PDF)Click here for additional data file.Table S26Detailed list of protein-DNA energy binding affinity, overlapping volume and number of atoms in collision for each complex in group-SingleProtein∶DNA(0.05 MB PDF)Click here for additional data file.Table S27Detailed list of protein-DNA energy binding affinity, overlapping volume and number of atoms in collision for each complex in group-SingleSameProtein∶DNA(0.04 MB PDF)Click here for additional data file.Table S28Detailed list of protein-DNA energy binding affinity, overlapping volume and number of atoms in collision for each complex in group-SubSetMultiProteins∶DNA(0.04 MB PDF)Click here for additional data file.Table S29Detailed list of protein-protein binding free energy for each protein-proteincomplex in group-MultiProteins∶DNA(0.04 MB PDF)Click here for additional data file.Table S30Detailed list of protein-protein binding free energy for each protein-proteincomplex in group-Protein∶Protein(0.06 MB PDF)Click here for additional data file.Table S31Average solvation energy (kJ/mol), free energy barrier of assembly dissociation (kJ/mol), and energy Z-scores for direct and indirect readouts for groups -SubSetMultiProteins∶DNA, -SingleSameProtein∶DNA(0.03 MB DOC)Click here for additional data file.Table S32Average protein-DNA energy binding affinity (kJ/mol), interface overlapping volume (Å3) and average number of interface collision atoms for groups -SubSetMultiProteins∶DNA, -SingleSameProtein∶DNA(0.03 MB DOC)Click here for additional data file.Table S33List of PDB IDs used in the study (group-MultiProteins∶DNA), with description of component (including Swiss Prot ID) and biological process of components.(0.08 MB DOC)Click here for additional data file.Table S34The list of PDB codes of complexes from group-SingleProtein∶DNA(0.03 MB DOC)Click here for additional data file.Table S35The list of PDB codes of complexes from group-SingleSameProtein∶DNA(0.03 MB DOC)Click here for additional data file.Table S36The list of PDB codes of complexes from group-SubSetMultiProteins∶DNA(0.03 MB DOC)Click here for additional data file.Table S37The list of PDB codes of complexes from group-Protein∶Protein(0.03 MB PDF)Click here for additional data file.\n\nREFERENCES:\n1. SinhaSAdlerASFieldYChangHYSegalE\n2008\nSystematic functional characterization of cis-regulatory motifs in human core promoters.\nGenome Res\n18\n477\n488\n18256240\n2. ZhaoGSchrieferLAStormoGD\n2007\nIdentification of muscle-specific regulatory modules in Caenorhabditis elegans.\nGenome Res\n17\n348\n357\n17284674\n3. YuXLinJZackDJQianJ\n2006\nComputational analysis of tissue-specific combinatorial gene regulation: predicting interaction between transcription factors in human tissues.\nNucleic Acids Res\n34\n4925\n4936\n16982645\n4. MoormanCSunLVWangJde WitETalhoutW\n2006\nHotspots of transcription factor colocalization in the genome of Drosophila melanogaster.\nProc Natl Acad Sci U S A\n103\n12027\n12032\n16880385\n5. BanerjeeNZhangMQ\n2003\nIdentifying cooperativity among transcription factors controlling the cell cycle in yeast.\nNucleic Acids Res\n31\n7024\n7031\n14627835\n6. BermanBPNibuYPfeifferBDTomancakPCelnikerSE\n2002\nExploiting transcription factor binding site clustering to identify cis-regulatory modules involved in pattern formation in the Drosophila genome.\nProc Natl Acad Sci U S A\n99\n757\n762\n11805330\n7. BermanBPPfeifferBDLavertyTRSalzbergSLRubinGM\n2004\nComputational identification of developmental enhancers: conservation and function of transcription factor binding-site clusters in Drosophila melanogaster and Drosophila pseudoobscura.\nGenome Biol\n5\nR61\n15345045\n8. ChoKILeeKLeeKHKimDLeeD\n2006\nSpecificity of molecular interactions in transient protein-protein interaction interfaces.\nProteins\n65\n593\n606\n16948160\n9. ChakrabartiPJaninJ\n2002\nDissecting protein-protein recognition sites.\nProteins\n47\n334\n343\n11948787\n10. EllisJJBroomMJonesS\n2006\nProtein-RNA interactions: Structural analysis and functional classes.\nProteins\n66\n903\n911\n11. JonesSDaleyDTLuscombeNMBermanHMThorntonJM\n2001\nProtein-RNA interactions: a structural analysis.\nNucleic Acids Res\n29\n943\n954\n11160927\n12. JonesSThorntonJM\n1995\nProtein-protein interactions: a review of protein dimer structures.\nProg Biophys Mol Biol\n63\n31\n65\n7746868\n13. JonesSThorntonJM\n1996\nPrinciples of protein-protein interactions.\nProc Natl Acad Sci U S A\n93\n13\n20\n8552589\n14. JonesSThorntonJM\n1997\nAnalysis of protein-protein interaction sites using surface patches.\nJ Mol Biol\n272\n121\n132\n9299342\n15. JonesSvan HeyningenPBermanHMThorntonJM\n1999\nProtein-DNA interactions: A structural analysis.\nJ Mol Biol\n287\n877\n896\n10222198\n16. LejeuneDDelsauxNCharloteauxBThomasABrasseurR\n2005\nProtein-nucleic acid recognition: statistical analysis of atomic interactions and influence of DNA structure.\nProteins\n61\n258\n271\n16121397\n17. Lo ConteLChothiaCJaninJ\n1999\nThe atomic structure of protein-protein recognition sites.\nJ Mol Biol\n285\n2177\n2198\n9925793\n18. LuscombeNMLaskowskiRAThorntonJM\n2001\nAmino acid-base interactions: a three-dimensional analysis of protein-DNA interactions at an atomic level.\nNucleic Acids Res\n29\n2860\n2874\n11433033\n19. Mandel-GutfreundYSchuelerOMargalitH\n1995\nComprehensive analysis of hydrogen bonds in regulatory protein DNA-complexes: in search of common principles.\nJ Mol Biol\n253\n370\n382\n7563096\n20. MirnyLAGelfandMS\n2002\nStructural analysis of conserved base pairs in protein-DNA complexes.\nNucleic Acids Res\n30\n1704\n1711\n11917033\n21. NadassyKWodakSJJaninJ\n1999\nStructural features of protein-nucleic acid recognition sites.\nBiochemistry\n38\n1999\n2017\n10026283\n22. PaboCONekludovaL\n2000\nGeometric analysis and comparison of protein-DNA interfaces: why is there no simple code for recognition?\nJ Mol Biol\n301\n597\n624\n10966773\n23. TregerMWesthofE\n2001\nStatistical analysis of atomic contacts at RNA-protein interfaces.\nJ Mol Recognit\n14\n199\n214\n11500966\n24. BermanHMWestbrookJFengZGillilandGBhatTN\n2000\nThe Protein Data Bank.\nNucleic Acids Res\n28\n235\n242\n10592235\n25. KrissinelEHenrickK\n2005\nDetection of Protein Assemblies in Crystals.\nBerholdMRea\nComputational Life Sciences\nHeidelberg\nSpringer Berlin\n163\n174\n26. KrissinelEHenrickK\n2007\nInference of macromolecular assemblies from crystalline state.\nJournal of Molecular Biology\ndoi: 10.1016/j.jmb.2007.05.022\n27. HutchinsonEGThorntonJM\n1996\nPROMOTIF–a program to identify and analyze structural motifs in proteins.\nProtein Sci\n5\n212\n220\n8745398\n28. LuXJOlsonWK\n2003\n3DNA: a software package for the analysis, rebuilding and visualization of three-dimensional nucleic acid structures.\nNucleic Acids Res\n31\n5108\n5121\n12930962\n29. AhmadSKonoHArauzo-BravoMJSaraiA\n2006\nReadOut: structure-based calculation of direct and indirect readout energies and specificities for protein-DNA recognition.\nNucleic Acids Res\n34\nW124\n127\n16844974\n30. ZhangCLiuSZhuQZhouY\n2005\nA knowledge-based energy function for protein-ligand, protein-protein, and protein-DNA complexes.\nJ Med Chem\n48\n2325\n2335\n15801826\n31. LiuSZhangCZhouHZhouY\n2004\nA physical reference state unifies the structure-derived potential of mean force for protein folding and binding.\nProteins\n56\n93\n101\n15162489\n32. LuscombeNMAustinSEBermanHMThorntonJM\n2000\nAn overview of the structures of protein-DNA complexes.\nGenome Biol\n1\nREVIEWS001\n11104519\n33. GuharoyMChakrabartiP\n2007\nSecondary structure based analysis and classification of biological interfaces: identification of binding motifs in protein-protein interactions.\nBioinformatics\n34. JayaramBJainT\n2004\nThe role of water in protein-DNA recognition.\nAnnu Rev Biophys Biomol Struct\n33\n343\n361\n15139817\n35. ReddyCKDasAJayaramB\n2001\nDo water molecules mediate protein-DNA recognition?\nJ Mol Biol\n314\n619\n632\n11846571\n36. JaninJ\n1999\nWet and dry interfaces: the role of solvent in protein-protein and protein-DNA recognition.\nStructure\n7\nR277\n279\n10647173\n37. WilliamsonJR\n2008\nCooperativity in macromolecular assembly.\nNat Chem Biol\n4\n458\n465\n18641626\n38. SaraiAKonoH\n2005\nProtein-DNA recognition patterns and predictions.\nAnnu Rev Biophys Biomol Struct\n34\n379\n398\n15869395\n39. WangGDunbrackRLJr\n2003\nPISCES: a protein sequence culling server.\nBioinformatics\n19\n1589\n1591\n12912846\n40. WangGDunbrackRLJr\n2005\nPISCES: recent improvements to a PDB sequence culling server.\nNucleic Acids Res\n33\nW94\n98\n15980589\n41. MartinAC\n2005\nMapping PDB chains to UniProtKB entries.\nBioinformatics\n21\n4297\n4301\n16188924\n42. EisenbergDWeissRMTerwilligerTC\n1982\nThe helical hydrophobic moment: a measure of the amphiphilicity of a helix.\nNature\n299\n371\n374\n7110359\n43. McLachlanAD\n1982\nRapid comparison of protein structres.\nActa Crystallographica\n38\n871\n873\n44. Martin ACR http://www.bioinf.org.uk/software/profit/\n45. Michael GromihaMSiebersJGSelvarajSKonoHSaraiA\n2004\nIntermolecular and intramolecular readout mechanisms in protein-DNA recognition.\nJ Mol Biol\n337\n285\n294\n15003447\n46. BarequetGChazelleBGuibasLMitchellJTalA\n1996\nBOXTREE: A Hierarchical representation for Surface in 3D\n47. HubbardP\n1996\nApproximation Polyhedra with Spheres for Time-critical Collision Detection.\nACM trans Computer Graphics\n15\n179\n210\n48. BergenG\n1997\nEfficient collision detection of complex deformable models using AABB trees.\nJournal of Graphics Tools\n2\n1\n13\n49. HughesMDiMattiaCLinMManochaD\n1996\nEfficient and accurate interference detection for polynomial deformation and soft object animation\n50. GottschalkSLinMManochaD\n1996\nOBB-tree: A hierarchical structure for rapid interference detection\n51. TurkG\n1989\nInteractive Collision Detection for Molecular Graphics\nChapel Hill\nThe University of North Carolina\n52. TeschnerMHeidelbergerBMuellerMRomeranetsDGrossD\n2003\nOptimized Spatial Hashing for Collision Detection of Deformable Objects.\nMunich, Germany"
+}

batch_3/PMC2532748.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2532748",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2532748\nAUTHORS: Christian B. Gocke, Hongtao Yu\n\nABSTRACT:\nHistone modifications in chromatin regulate gene expression. A transcriptional co-repressor complex containing LSD1–CoREST–HDAC1 (termed LCH hereafter for simplicity) represses transcription by coordinately removing histone modifications associated with transcriptional activation. RE1-silencing transcription factor (REST) recruits LCH to the promoters of neuron-specific genes, thereby silencing their transcription in non-neuronal tissues. ZNF198 is a member of a family of MYM-type zinc finger proteins that associate with LCH. Here, we show that ZNF198-like proteins are required for the repression of E-cadherin (a gene known to be repressed by LSD1), but not REST-responsive genes. ZNF198 binds preferentially to the intact LCH ternary complex, but not its individual subunits. ZNF198- and REST-binding to the LCH complex are mutually exclusive. ZNF198 associates with chromatin independently of LCH. Furthermore, modification of HDAC1 by small ubiquitin-like modifier (SUMO) in vitro weakens its interaction with CoREST whereas sumoylation of HDAC1 stimulates its binding to ZNF198. Finally, we mapped the LCH- and HDAC1–SUMO-binding domains of ZNF198 to tandem repeats of MYM-type zinc fingers. Therefore, our results suggest that ZNF198, through its multiple protein-protein interaction interfaces, helps to maintain the intact LCH complex on specific, non-REST-responsive promoters and may also prevent SUMO-dependent dissociation of HDAC1.\n\nBODY:\nIntroductionThe ordered assembly of genomic DNA into a proteinacious substance—chromatin—allows for high-order regulation of DNA-templated processes, such as transcription, replication, and DNA repair. Chromatin contains repeating units of nucleosomes, which consists of one histone H3/H4 tetramer and two H2A/H2B dimers wrapped around by double-stranded DNA [1]–[3]. Polymers of nucleosomes flanked by various lengths of linker DNA can fold into compacted high-order structures that are subject to dynamic regulation [4].Post-translational modifications on the flexible tails of histones can directly or indirectly affect chromatin structure [5]. Histone acetylation is generally associated with transcriptional activation and is dynamically regulated by histone acetyltransferases (HATs) and histone deacetylases (HDACs) [5]. The effect of histone lysine methylation, catalyzed by methyltransferases, depends on the specific residue and degree of modification (mono-, di-, or trimethylation) [6]. Histone H3 lysine 4 di- and trimethylation (H3K4me2/3) is associated with active promoters [7], [8], but H3K9me2/3 is mostly associated with transcriptional repression [9]. Lysine-specific demethylase 1 (LSD1; also known as BHC110 or AOF2) is a flavin adenine dinucleotide (FAD)-dependent amine oxidase that demethylates histone H3K4me1/2, but not H3K4me3 [10], [11]. Although LSD1 alone can demethylate bulk histones or peptide substrates, it requires a co-factor, REST co-repressor (CoREST), for efficient binding to nucleosomes and demethylation of nucleosomal substrates [12]–[14]. The LSD1–CoREST interaction also stabilizes the LSD1 protein in the cell [14]. A fraction of the abundant class I HDACs, HDAC1 and HDAC2, associate with LSD1–CoREST, forming an LSD1–CoREST–HDAC1/2 (LCH) core ternary complex [15]–[18]. Formation of this complex on chromatin enables HDAC1/2 and LSD1 to stimulate each other's activity through CoREST [19].The LCH complex can be targeted to specific promoters through binding to sequence-specific transcriptional factors, either directly or indirectly. For example, RE1-silencing transcription factor (REST), which is a Krüppel-like zinc finger-containing protein, binds directly to CoREST and recruits the LCH complex to neuron-specific gene promoters that contain RE1 elements, thus repressing the expression of neuron-specific genes in non-neuronal tissues [20], [21]. In addition, LCH can be incorporated into a larger co-repressor complex that also contains CtBP1/2 and the G9a histone H3K9 methyltransferase [16]. CtBP1/2 in turn binds to Krüppel-like zinc finger-containing sequence-specific repressors ZEB1/2, which recruit this complex to chromatin [22]. Finally, LSD1 is targeted to androgen- and estrogen-responsive promoters through interactions with androgen receptor (AR) and possibly estrogen receptor (ER). In this context, LSD1 activates transcription through promoting the demethylation of H3K9me1/2 at these promoters [23], [24]. Whether the entire LCH complex is targeted to AR- or ER-dependent promoters is unclear.Several human proteins, including ZNF198, ZNF237, ZNF261, ZNF262, and ZNF258, contain a stretch of unique tandem zinc fingers called MYM (myeloproliferative and mental retardation) domains [25] (Figure S1). The MYM-domains of ZNF198 are frequently fused to FGF receptor kinase in myeloproliferative syndromes [26]–[28]. Disruptions near the ZNF261 gene have been linked to X-linked mental retardation [29]. Among human MYM-domain proteins, ZNF198, ZNF261, and ZNF262 share a similar domain architecture and possibly perform similar functions (Figure S1). A Drosophila homolog of these proteins, without children (dWoc), is essential for viability, associates with chromatin, and prevents telomere fusions [30]–[32]. Interestingly, ZNF198 and ZNF261 are present in transcriptional corepressor complexes that also contain LCH [13], [14], [18], although their functions in transcriptional regulation have not been explored.Many transcription factors and cofactors are modified by small ubiquitin-like modifier (SUMO). Sumoylation of these factors generally leads to transcriptional repression. For unknown reasons, multiple subunits within a given chromatin-associated complex are often targeted by sumoylation [33], [34]. For example, ZNF198, ZNF262, HDAC1, and LSD1 are known SUMO substrates [33], [35]–[37]. Sumoylation of HDAC1 has been shown to be required for its function. Recent reports have also identified ZNF198 as a non-covalent binding partner for SUMO [38], [39].In this study, we characterize the function and mechanism of ZNF198-like proteins in regulating the LCH complex. We show that depletion of ZNF198, ZNF261, and ZNF262 by RNA interference (RNAi) in HeLa cells causes derepression of E-cadherin, a known target of LSD1. By contrast, ZNF198-like proteins are not required for the transcriptional repression of several REST-responsive genes that are repressed by LSD1. Consistent with this finding, ZNF198 selectively binds to the LSD1–CoREST–HDAC1 ternary complex and binding of ZNF198 to LCH prevents its interaction with REST. Similar to dWoc, ZNF198 associates with chromatin. Depletion of ZNF198-like proteins weakens the association of LCH with chromatin. Furthermore, sumoylation of HDAC1 decreases its affinity toward CoREST, but enhances its binding to ZNF198. Finally, the tandem repeats of MYM-type zinc fingers of ZNF198 mediate its binding to both LCH and sumoylated HDAC1. Collectively, our results suggest that, unlike the Krüppel-like zinc fingers which bind to DNA, the MYM-type zinc fingers of ZNF198-like proteins mediate multiple protein-protein interactions, maintains the integrity of the LCH complex at non-REST-responsive promoters, and may antagonize SUMO-dependent disassembly of the LCH complex.ResultsZNF198 associates with LSD1, CoREST, and HDACs in human cellsThe MYM-type zinc fingers have the CX2CX19–24[F/Y]CX3CX3[F/Y] (X is any residue) consensus motif [28]. Five proteins in the human proteome contain tandem repeats of the MYM-type zinc fingers, including ZNF198, ZNF261, ZNF262, ZNF237, and ZNF258 (Figure S1). Some of the MYM zinc fingers in ZNF237 and ZNF258 lack key conserved cysteines (Figure S1), whereas the zinc fingers in ZNF198, ZNF261, and ZNF262 all appear to be intact. ZNF198, ZNF261, and ZNF262 have additional features that differentiate them from ZNF237 and ZNF258. They contain a proline/valine-rich (P/V-rich) domain downstream of the MYM domain. They also contain a domain at their C-terminal region that is predicted by 3D-Jury [40] to have a fold similar to DNA breaking-rejoining enzymes, such as Cre recombinase. The Cre-like domain is also found in several proteins that do not contain MYM zinc fingers (Figure S1).ZNF198 and ZNF261 have been shown to be present in several LSD1-containing transcriptional corepressor complexes in sub-stoichiometric amounts [13], [14], [18]. To identify the major ZNF198-interacting proteins in human cells, we immunoprecipitated the endogenous ZNF198 protein from HEK293 and HeLa cells (Figure 1A and data not shown). The ZNF198-binding proteins were detected by Colloidal blue staining followed by mass spectrometry. LSD1, CoREST, and HDAC1/2 were present at near stoichiometric levels. ZNF262 was also present at sub-stoichiometric amounts. Several abundant proteins, including tubulin, Hsp70, and dynein, were also identified in the anti-ZNF198 IP, although they might not be specific ZNF198 interactors. This result indicates that LSD1, CoREST, and HDAC1/2 are major binding proteins of ZNF198 in human cells and confirms earlier findings that have demonstrated the interactions between the LCH complex and proteins containing MYM zinc fingers.10.1371/journal.pone.0003255.g001Figure 1ZNF198-like proteins are not required for the repression of REST-responsive genes.(A) LSD1, CoREST, and HDAC1/2 are major binding partners of ZNF198 in 293 cells. IP of ZNF198 from HEK293 whole cell lysates was separated on SDS-PAGE and stained with Colloidal Blue. Bands were excised and identified by mass spectrometry. (B) HeLa cells were transfected with siRNAs against luciferase (siLuc) or ZNF198, ZNF261, and ZNF262 (siMYM). Cell lysates were blotted with the indicated antibodies. (C & D) U2OS cells were transfected with the indicated siRNAs for three days, followed by quantitative RT-PCR using the indicated primer sets. Cycling-time values were normalized to the housekeeping gene cyclophilin B. Each PCR reaction was performed in triplicate, and error bars indicate the standard deviation of three separate experiments.ZNF198-like proteins are not required for the repression of REST-responsive genesThrough its binding to REST, the LCH complex is recruited to neuron-specific genes and represses their transcription in non-neuronal tissues. We first tested whether ZNF198-like proteins were required for the repression of REST-responsive genes. Because ZNF198, ZNF261, and ZNF262 have all been shown to be associated with LSD1-containing corepressor complexes, we depleted from U2OS and other human cells the three ZNF198-like MYM proteins using RNA intereference (RNAi). As shown in Figure 1B, RNAi against ZNF198, ZNF261, and ZNF262 effectively knocked down the levels of ZNF198 without affecting the levels of LSD1, CoREST, or HDAC1. We did not have antibodies against ZNF261 and ZNF262. However, quantitative RT-PCR analysis (QPCR) confirmed that the siRNAs against these two genes effectively reduced their mRNA levels (Figure 1C). As a comparison, we also depleted LSD1 from human cells using RNAi. Cells transfected with siRNA against luciferase were used as a control. As expected, QPCR analysis revealed that LSD1 RNAi caused an up-regulation of mRNA levels of the known LSD1 target genes, SCN3A and NCAM2 [23] (Figure 1D). By contrast, depletion of ZNF198-like proteins did not significantly alter the mRNA levels of SCN3A and NCAM2 (Figure 1D), suggesting that these proteins were not required for the repression of these putative REST-responsive neuronal genes in non-neuronal tissues.To identify additional genes that were repressed by LSD1, we performed microarray analysis of RNA samples from HeLa cells transfected with siRNAs against luciferase or LSD1 (data not shown). Among the genes that were up-regulated by LSD1 RNAi, we confirmed that keratin 17 (KRT17) was a REST-responsive gene, because REST directly bound to the promoter of KRT17 as demonstrated by chromatin immunoprecipitation (ChIP) (Figure 2A). Using QPCR, we confirmed that LSD1 RNAi indeed increased the mRNA levels of KRT17. Depletion of ZNF198-like MYM proteins again had no effect on KRT17 expression (Figure 2B). Therefore, these ZNF198-like proteins do not appear to be required for the repression of REST-responsive genes.10.1371/journal.pone.0003255.g002Figure 2ZNF198-like proteins are required for the repression of E-cadherin.(A) ChIP analysis reveals that REST binds to the promoter of keratin 17 (KRT17). (B) LSD1 is required for the repression of KRT17. U2OS cells were transfected with the indicated siRNAs for three days, followed by quantitative RT-PCR using KRT17 primers. (C) U2OS cells were transfected with the indicated siRNAs for three days, followed by quantitative RT-PCR using the E-cadherin primer set. Cycling-time values were normalized to the housekeeping gene cyclophilin B. Each PCR reaction was performed in triplicate, and error bars indicate the standard deviation of three separate experiments.ZNF198-like proteins are required for the repression of E-cadherinCorepressor complexes containing LSD1, CoREST, and HDAC1 can be recruited to promoters in REST-independent ways. E-cadherin is a well-characterized gene that is repressed by the CtBP corepressor complex containing LSD1, but E-cadherin is not known to be regulated by REST [16]. We confirmed that E-cadherin was indeed repressed by LSD1 (Figure 2C). RNAi against ZNF198-like MYM proteins also significantly elevated the mRNA level of E-cadherin. Therefore, ZNF198-like proteins are required for the repression of at least some LSD1-repressed genes that are not regulated by REST.ZNF198 binds selectively to the LSD1–CoREST–HDAC1 (LCH) ternary complexWe next sought to understand the mechanisms by which ZNF198-like MYM-domain proteins regulated the LCH complex. First, we tested whether ZNF198 bound directly to LSD1, CoREST, or HDAC1. To do this, we purified recombinant His6-ZNF198, GST-CoREST, His6-LSD1, and HDAC1-FLAG (Figure 3A and 3C). Surprisingly, in GST pull-down assays, ZNF198 did not bind efficiently to CoREST alone, LSD1–CoREST, or HDAC1–CoREST (Figure 3B, lanes 6–8 and Figure 3C, lanes 2 and 3). ZNF198, however, bound efficiently to the intact LCH ternary complex (Figure 3B, lane 2; Figure 3C, lane 1). Inhibition of HDAC1 and LSD1 activities using trichostatin A [41] (TSA) and tranylcypromine [42] (TCP), respectively, had no effect on the binding between ZNF198 and LCH (Figure 3B, lanes 3–5). Furthermore, binding of ZNF198 to LCH greatly reduced the binding of REST to LCH (Figure 3D, lanes 5 and 6). This suggests that ZNF198-binding and REST-binding to LCH are mutually exclusive, consistent with our finding that ZNF198 does not appear to be required for the repression of REST-responsive genes.10.1371/journal.pone.0003255.g003Figure 3ZNF198 binds preferentially to the intact LSD1–CoREST–HDAC1 (LCH) ternary complex.(A) Recombinant His6-ZNF198 purified from Sf9 cells was treated with TEV protease and analyzed by SDS-PAGE followed by Coomassie staining. TEV protease digestion removed the His6-tag and caused the protein to migrate faster, thus confirming the identity of the band as His6-ZNF198. (B) Recombinant His6-ZNF198 (6 µg) was added to glutathione-agarose beads that had been preincubated with 1 µg GST-CoREST, 2 µg HDAC1-FLAG, or 3 µg His-LSD1 as indicated. When indicated, TCP or TSA were present for the entire procedure. After washing, bound proteins were detected by western blotting with the indicated antibodies. (C) GST pull-downs assays were performed as in (B), except that bound proteins were stained with Coomassie blue. The bands belonging to His6-ZNF198, His6-LSD1, GST-CoREST, and HDAC1-FLAG are labeled. (D) ZNF198 competes with REST for binding to the LCH complex. HDAC1-FLAG (1 µg) and the His6-LSD1–His6-CoREST binary complex (3 µg) were preincubated with anti-FLAG M2 agarose in the indicated combinations. After washing, 35S-REST was added to each binding reaction in the presence or absence of His6-ZNF198 (10 µg). Bound REST was detected using a phosphoimager (upper panel). The intensities of REST bands in each reaction were quantified and normalized to lane 5. The values were averages of two experiments. Proteins bound to the anti-FLAG beads were also analyzed by SDS-PAGE and stained with Coomassie blue (lower panel). Note that His6-CoREST and HDAC1-FLAG migrate at the same position on the gel.ZNF198-like proteins stabilize the LCH complex on chromatinBinding of ZNF198 to LCH prevents the binding of REST. One possibility is that ZNF198 recruits LCH to specific promoters in a manner similar to REST. However, we were unable to detect high-affinity binding of ZNF198 to DNA in vitro (data not shown). Prompted by the finding that the Drosophila homolog of ZNF198, Woc, associated with chromatin, we tested whether ZNF198 also bound to chromatin in human cells. We transfected HeLa cells with plasmids encoding Myc-ZNF198 or its fragments. The cells were stained with anti-Myc either with or without extraction before fixation (Figure 4A and 4B). To mark the nuclei of transfected cells after extraction, cells were also co-transfected with a plasmid encoding a known chromatin-bound protein GFP-MCM7. As expected, GFP-MCM7 remained in the nucleus after such extraction, indicating that it was bound to chromatin [43] (Figure 4A). Both the endogenous ZNF198 (data not shown) and the full-length Myc-ZNF198 (Figure 4A) were diffusely nuclear localized and this staining was resistant to extraction, consistent with ZNF198 being bound to chromatin. Analysis of the ZNF198 fragments showed that those fragments lacking the P/V-rich domain were not detected in the nuclei after extraction, even though these nuclei contained GFP-MCM7 and had thus been transfected (Figure 4A and 4B). These data indicate that the P/V-rich domain is required for the association of ZNF198 with chromatin. The P/V-rich region alone, however, bound to chromatin much weaker than the full-length Myc-ZNF198 and its larger fragments (Figure 4B). Therefore, multiple regions in ZNF198 contribute to its association with chromatin.10.1371/journal.pone.0003255.g004Figure 4ZNF198 binds to chromatin through its P/V-rich domain.(A) HeLa Tet-on cells were transfected with the indicated Myc-tagged ZNF198 constructs along with GFP-MCM7. Cells were either fixed directly (–Extraction) or extracted prior to fixation (+Extraction) and stained with anti-Myc antibody (red) and DAPI (blue). GFP is shown in green. (B) Summary of the staining data described in (A). The ZNF198 fragments are shown on the left while the percentages of GFP-positive cells that were also Myc-positive after extraction are shown on the right. More than 30 GFP-positive cells from 10 random fields were counted for each fragment. The boundaries of ZNF198 fragments are indicated by triangles.Formation of an intact LSD1–CoREST–HDAC1/2 complex on chromatin is important for optimal co-repressor activity [19]. Consistent with our in vitro finding that ZNF198 interacted specifically with the LCH ternary complex, depletion of LSD1 by RNAi also dramatically reduced the amounts of CoREST and HDAC1 in the α-ZNF198 IPs (Figure 5A, top panel, compare lanes 1 and 3). Through selective binding to the LCH ternary complex, ZNF198 would be expected to stabilize this complex. On the other hand, depletion of ZNF198-like MYM-domain proteins only slightly decreased the association of LSD1 with CoREST and HDAC1 (Figure 5A, middle panel, compare lanes 1 and 2). Thus, ZNF198-like proteins have a role in maintaining the LCH complex, but are not essential for its stability.10.1371/journal.pone.0003255.g005Figure 5ZNF198-like proteins regulate the chromatin association of LSD1.(A) HeLa Tet-on cells were transfected with the indicated siRNAs: Luc (firefly luciferase), MYM (ZNF198, ZNF261, and ZNF262), or LSD1. Lysates from these RNAi cells were immunoprecipitated with either anti-ZNF198 (top panel) or anti-LSD1 (middle panel). The IPs and the lysates (bottom panel) were blotted with the indicated antibodies. (B) After RNAi of the indicated proteins, nuclear pellets were generated by subjecting HeLa Tet-on cells to hypotonic lysis followed by centrifugation. Normalized samples from each step were subjected to SDS-PAGE and blotted with the indicated antibodies. (C) Nuclear pellets in lanes 7–9 in (B) were subjected to extraction with a high-salt buffer. Normalized samples from each step were blotted with the indicated antibodies. (D) HeLa cells transfected with siLuc were subjected to fractionation as in B. The nuclear pellet in lane 7 in (B) was digested with micrococcal nuclease (MNase) followed by extraction with 2 mM EDTA. Supernatants (S) and pellets (P) were blotted with the indicated antibodies.We next tested whether depletion of ZNF198-like proteins affected the chromatin binding of LSD1 by subcellular fractionation [44]. The majority of ZNF198 was found in insoluble chromatin fractions (Figure 5B and 5C, compare supernatant and pellets). LSD1 RNAi did not affect chromatin association of ZNF198, suggesting that ZNF198 bound to chromatin independently of LCH (Figure 5B and 5C). By contrast, only a small fraction of LSD1 was associated with chromatin (Figure 5B). RNAi of ZNF198-like proteins reduced the levels of LSD1 in the chromatin fraction by about 2-fold (Figure 5B, compare lanes 7 and 8). This effect was more pronounced after a high-salt extraction of the chromatin fraction (Figure 5C, compare lanes 4 and 5). Importantly, an unrelated chromatin-binding protein MCM7 was unaffected by RNAi against ZNF198-like proteins [45]. Therefore, ZNF198-like proteins are required for efficient chromatin-association of LSD1 and possibly the LCH complex. HDAC1 levels in chromatin fractions, however, were not affected by RNAi against ZNF198-like proteins, presumably because only a fraction of cellular HDAC1 associated with LSD1 and ZNF198 (data not shown).To further characterize the nature of chromatin association of ZNF198 and LSD1, we digested the chromatin fraction with micrococcal nuclease (MNase) followed by EDTA extraction [44] (Figure 5D). Most of MCM7 and histone H3 were released into the supernatant by MNase digestion (Figure 5D). Although about 50% of HDAC1 and LSD1 and a small fraction of ZNF198 were also released, significant fractions of ZNF198, LSD1, and HDAC1 remained in the pellet. Therefore, at least a portion of the LCH complex associates with nuclease-resistant chromatin materials or nuclear matrix or both.Sumoylation of HDAC1 weakens its binding to CoREST but enhances its binding to ZNF198Both HDAC1 and LSD1 can be sumoylated. Although the function of LSD1 sumoylation has not been established, sumoylation of HDAC1 at its C-terminal region is required for its ability to repress transcription [37]. We were interested in how sumoylation of HDAC1 may affect its interaction with CoREST. Since sumoylation of HDAC1 has been shown to be important for its repressor activity, we hypothesized that SUMO-HDAC1 might show increased affinity for CoREST. We tested this hypothesis by comparing binding of sumoylated and free HDAC1-FLAG to GST-CoREST using GST pull-down assays (Figure 6A). Surprisingly, free HDAC1-FLAG, but not SUMO2-HDAC1-FLAG, bound to GST-CoREST. Thus, sumoylation of HDAC1 inhibits its interaction with CoREST.10.1371/journal.pone.0003255.g006Figure 6Sumoylation of HDAC1 weakens its interaction with CoREST, but enhances its binding to ZNF198.(A) HDAC1-FLAG (40 ng) was incubated with SUMO2 and sumoylation enzymes with or without ATP. The reaction mixtures were then added to glutathione-agarose beads that had been preincubated with buffer or GST-CoREST (1 µg). After washing, bound (top panel) and unbound (bottom panel) proteins were blotted with anti-FLAG. The sumoylated and un-sumoylated HDAC1 bands are labeled. (B) The indicated 35S-labeled in vitro translated proteins were incubated with glutathione-agarose beads bound with 10 µg of GST, GST-SUMO1, or GST-SUMO2. The bound proteins were separated by SDS-PAGE, stained with Coomassie blue (bottom panel, a representative image), and analyzed using a phosphoimager (top panel). (C) HDAC1-FLAG (1 µg) bound to anti-FLAG M2 agarose beads was incubated with SUMO2 and sumoylation enzymes with or without ATP. After washing, either His6-ZNF198 (6 µg) or GST-CoREST (2 µg) was added to the beads. The bound proteins were blotted with the indicated antibodies.Recently, it has been shown that ZNF198 and LSD1 could bind to SUMO non-covalently [38], [39]. To confirm these reports, we tested whether ZNF198, LSD1, CoREST, HDAC1, and PIASxβ bound SUMO non-covalently in GST pull-down assays (Figure 6B). Consistent with previous reports [46], PIASxβ bound to both GST-SUMO1 and GST-SUMO2. By contrast, ZNF198 bound preferentially to GST-SUMO2. We could not detect binding of LSD1, CoREST, or HDAC1 to either GST-SUMO1 or GST-SUMO2. Therefore, our results confirm that ZNF198 binds to SUMO2 non-covalently and suggest that the reported non-covalent binding between LSD1 and SUMO2 in human cell lysates is likely indirect and mediated through ZNF198.We next tested how sumoylation of HDAC1 might influence its interaction with ZNF198. We performed FLAG IPs with either sumoylated or free HDAC1-FLAG as bait (Figure 6C). ZNF198 exhibited minimal binding to HDAC1 alone. Strikingly, ZNF198 binding to HDAC1 was greatly enhanced by HDAC1 sumoylation, even though a minor fraction of HDAC1 was sumoylated in this particular assay. As a control, this minimal sumoylation of HDAC1 did not perturb its binding to CoREST, because the majority of HDAC1 remained unsumoylated and retained the ability to bind to CoREST (Figure 6C). Thus, sumoylation of HDAC1 weakens its binding to CoREST, but enhances its binding to ZNF198 in vitro.Many non-covalent binding partners of SUMO are efficiently sumoylated in vitro\n[46], [47]. Consistent with a previous report [35], ZNF198 was efficiently sumoylated in vitro (Figure S2A). Many efficient SUMO substrates with SUMO binding capacity, such as RanBP2 and PIASxβ, are also SUMO ligases [46], [47]. ZNF198, however, failed to stimulate the sumoylation of either LSD1 or HDAC1 in vitro (Figure S2B). Therefore, we do not have evidence to suggest that ZNF198 functions as a SUMO ligase.MYM-type zinc fingers of ZNF198 mediate its interactions with the LCH complex and sumoylated HDAC1We next mapped the regions of ZNF198 required for binding to the ternary LCH complex or HDAC1-SUMO2. We used recombinant FLAG-tagged HDAC1-SUMO2, HDAC1, or LSD1–CoREST–HDAC1 as baits. As expected, full-length ZNF198 bound to both SUMO2-HDAC1 and LSD1–CoREST–HDAC1, but not to HDAC1 alone (Figure 7A). Analysis of a series of truncation mutants of ZNF198 revealed that the central region of ZNF198 containing the 10 tandem MYM-type zinc fingers was necessary and sufficient for binding to both SUMO2-HDAC1 and LSD1–CoREST–HDAC1 (Figure 7A). Further mapping revealed that a small ZNF198 fragment containing two zinc fingers, MYM8-9, was sufficient for binding to the LCH complex in vitro (Figure 7A). MYM8-9 could also be co-immunoprecipitated with LSD1 from HeLa cell lysates (Figure 7B), although it was unclear whether it bound LSD1 as efficiently as the wild-type ZNF198. By contrast, efficient binding of ZNF198 to SUMO2-HDAC1 required larger fragments of ZNF198 containing most of its MYM-type zinc fingers (Figure 7A). Notably, ZNF198 has three putative SUMO interaction motifs (SIMs) [38]. Deletion of the N-terminal region containing SIM1 and SIM2 or mutation of SIM3 did not affect the binding of ZNF198 to HDAC1-SUMO2 (Figure 7A), indicating that none of these putative SIM motifs of ZNF198 are important for HDAC1-SUMO2 binding. Thus, zinc fingers 8 and 9 of ZNF198 mediate its binding to LCH, whereas additional zinc fingers are required for the binding of ZNF198 to HDAC1-SUMO2.10.1371/journal.pone.0003255.g007Figure 7MYM-type zinc fingers of ZNF198 mediate its binding to the LCH complex and HDAC1-SUMO2.(A) 35S-labeled ZNF198 fragments and mutants were incubated with anti-FLAG beads that had been preincubated with HDAC1-SUMO2, HDAC1, or the LSD1–CoREST–HDAC1 complex. After washing, the bound proteins were visualized by autoradiography and Coomassie blue staining (bottom panel). Three putative SUMO-interacting motifs (SIMs) are indicated. The SIM3 mutant contains V483A, L484A, and V485A mutations. (B) HeLa Tet-on cells were transfected with the indicated constructs for 24 hours. Lysates and the Myc IPs of the transfected cells were blotted with the indicated antibodies. (C) Ribbon drawings of three Krüppel-like zinc fingers bound to DNA (PDB ID: 1ZAA; left) and a MYM-type zinc finger from ZNF237 (PDB ID: 2DAS; right). Zinc ions are shown as spheres while zinc-binding ligands are shown in sticks.DiscussionFunctions of ZNF198-like proteins in transcriptional repressionConsistent with a role of ZNF198 in transcriptional regulation, several well established transcriptional repressors, including LSD1, CoREST and HDAC1/2, are major binding proteins of ZNF198 in vivo. Interestingly, ZNF198 only interacts with the intact LSD1–CoREST–HDAC1 (LCH) ternary complex. Moreover, ZNF198- and REST-binding to LCH are mutually exclusive. Expectedly, ZNF198-like proteins are dispensable for the LSD1-mediated repression of REST-responsive neuron-specific genes in non-neuronal cell lines, such as HeLa and U2OS. This created a serious challenge for our studies on ZNF198, because most known LSD1-repressed genes are REST-responsive. Despite this difficulty, we managed to show that ZNF198-like proteins are required for the transcriptional repression of E-cadherin, a well-established LSD1-repressed gene that is not REST-responsive. Identification of additional genes whose repression requires ZNF198-like proteins is needed to fully understand their biological functions.We have further shown that ZNF198 binds directly to chromatin in part through its proline/valine-rich domain. Depletion of ZNF198 reduces the amount of chromatin-bound LSD1 as revealed by subcellular fractionation experiments. Thus, one possible mechanism by which ZNF198-like proteins facilitate the functions of the LSD1–CoREST–HDAC1-containing corepressor complexes is to recruit or stabilize the LSD1–CoREST–HDAC1 ternary complex on chromatin. Because we have so far failed to detect ZNF198 at specific promoters using ChIP, it remains to be determined whether ZNF198-like proteins are required for the stable association of the LCH complex at specific promoters. It will also be interesting to test whether, in addition to tethering the LCH complex to chromatin, ZNF198 directly facilitates the demethylation or deacetylation or both of nucleosomes by the LCH complex.ZNF198 and sumoylation of HDAC1Sumoylation of HDAC1 is required for its function in transcriptional repression [37], [48]. For example, cells stably expressing wild-type HDAC1, but not its sumoylation-deficient mutant, show cell cycle defects [37]. The sumoylation-deficient mutant of HDAC1 also shows lower deacetylase activity [37], [48], suggesting that sumoylation is required for the full activation of the catalytic activity of HDAC1. We show that HDAC1 sumoylation inhibits its binding to CoREST. Our finding is consistent with previous findings that the C-terminus of HDAC1 mediates its interactions with cofactors [49] and mutation of the HDAC1 sumoylation sites does not disrupt cofactor binding [37].HDAC1 and LSD1 exhibit positive cooperativity in deacetylating and demethylating nucleosomes. CoREST bridges the interaction between HDAC1 and LSD1. Disruption of the HDAC1–CoREST interaction by HDAC1 sumoylation is thus expected to abolish the positive cooperativity between HDAC1 and LSD1. Paradoxically, sumoylation of HDAC1 is required for its function and possibly activity. Our finding that sumoylation of HDAC1 enhances its binding toward ZNF198 provides one possible way to resolve the paradox of HDAC1 sumoylation. Through its abilities to bind multiple components of the LCH complex and to sumoylated HDAC1, ZNF198 may antagonize the disruption of the HDAC1–CoREST interaction by HDAC1 sumoylation and preserve the integrity of the LCH complex on chromatin. The physiological significance of these in vitro findings remains to be established.MYM-type zinc fingers as protein-protein interaction modulesZNF198 has been proposed to recruit transcriptional corepressors to specific promoters through sequence-specific DNA-binding, a function that is analogous to REST. This hypothesis largely stems from the fact that both ZNF198 and REST contain tandem zinc fingers and bind to the LCH complex [18], [20], [21]. Furthermore, ZNF198 competes with REST for CoREST-binding. However, the tandem zinc fingers of REST are Krüppel-like zinc fingers that mediate sequence-specific DNA-binding to RE1-elements in promoters [50], [51]. Structure comparison reveals that the MYM-type zinc fingers in ZNF198 have a fold that is distinct from that of the Krüppel-like zinc fingers (Figure 7C\n). In particular, the long α-helix that mediates DNA binding in Krüppel-like zinc fingers [52] is much shorter in the MYM-type zinc fingers. Therefore, the MYM-type zinc fingers are unlikely to bind to DNA in a manner similar to the Krüppel-like zinc fingers. In fact, a search using the Dali server (http://www.ebi.ac.uk/dali) revealed that the fold of MYM-type zinc fingers is most related to that of the LIM (Lin11, Isl-1 & Mec-3) domain. LIM domains are also often found in tandem repeats and mediate protein-protein interactions [53]. The structural similarity between the MYM and LIM domains further supports a role for MYM-type zinc fingers in mediating protein-protein interactions.In conclusion, we have revealed a functional requirement of ZNF198-like proteins in transcriptional repression of LSD1-repressed genes that are not REST-responsive. Our results further suggest the following model to explain the mechanism by which ZNF198 promotes the functions of LSD1 (Figure 8). In this model, ZNF198 binds to chromatin and recruits the LSD1–CoREST–HDAC1 ternary complex to chromatin. The MYM-type zinc fingers of ZNF198 mediate its interactions with the LCH complex and with sumoylated HDAC1. Through its ability to engage in multiple protein-protein interactions, ZNF198 stabilizes the LCH complex on chromatin and possibly prevents the dissociation of HDAC1 from the complex that is triggered by HDAC1 sumoylation.10.1371/journal.pone.0003255.g008Figure 8Proposed mechanisms by which ZNF198 regulates the LSD1–CoREST–HDAC1 complex.See DISCUSSION for details.Materials and MethodsProtein expression and purificationSUMO-related constructs and protein purification were described previously [33]. The coding regions of CoREST and LSD1 were amplified from human fetal thymus cDNA library (BD Biosciences) and ZNF198 was amplified from a purchased cDNA plasmid (Open Biosystems) by PCR. The PCR products were digested and ligated into appropriate expression vectors. Similar methods were used to construct plasmids encoding various ZNF198 fragments. The pSC-β-REST construct was a gift from Jenny Hsieh.The full-length His6-LSD1, His6-LSD1–His6-CoREST, or His6-ZNF198 were expressed in Sf9 insect cells and purified using a combination of Ni2+-Sepharose (Amersham) affinity chromatography and ion exchange chromatography (Resource Q, Amersham). HDAC1-FLAG was purified with M2 agarose beads (Sigma) from Sf9 cell lysates and eluted with the FLAG peptide. GST-CoREST, GST-SUMO1, and GST-SUMO2 were purified from bacterial lysates using glutathione-agarose resin (Amersham). Proteins were stored in buffers containing 50 mM Tris-HCl, pH 8.1, 50–200 mM KCl, 10% glycerol, and 1 mM DTT.Cell culture and transfectionsHeLa Tet-on (BD Biosciences) and U2OS cells were grown in DMEM (Invitrogen) supplemented with 10% fetal bovine serum, 2 mM L-glutamine, and 100 µg/ml penicillin and streptomycin at 37°C and 5% CO2. DNA and siRNA transfections were performed with the effectene reagent (Qiagen) and Lipofectamine RNAiMax reagent (Invitrogen), respectively, according to manufacturer's protocols. The siRNA sequences are: 5′- GGCCUAGACAUUAAACUGA-3′ (LSD1), 5′-GGGCCAGACAGCUUAUCAA-3′ (ZNF198), 5′-GACCCUGUGUAAGAACUUU-3′ (ZNF261), 5′-CACCACCACUAGUAAAGAU-3′ (ZNF262).Cell fractionationLysates of HeLa Tet-on cells (2×10-cm plates) transfected with siRNAs for 48 hrs were fractionated as described [44] with one significant modification. We included an additional high-salt extraction step using buffer C (10 mM HEPES, pH 7.9, 10 mM KCl, 300 mM NaCl, 1.5 mM MgCl2, 25% glycerol, 0.1% Triton X-100, 1 mM DTT, 10 µg/ml protease inhibitor cocktail, and 0.4 mM PMSF).ImmunofluorescenceHeLa Tet-on cells transfected with various plasmids were either extracted as previously described [43] and then fixed with 4% paraformaldehyde or directly fixed. All samples were then permeabilized with 0.1% Triton-X100 in PBS, and incubated with 1 µg/ml of anti-Myc (9E10, Roche). After washing, fluorescent secondary antibodies (Molecular Probes) were added at 1∶500 dilutions. The cells were again washed three times with PBS, counter-stained with DAPI, and viewed using a 63× objective on a Zeiss Axiovert 200 M microscope. Images were acquired using the Intelligent Imaging software, and pseudo-colored in Adobe Photoshop.Antibodies, immunoprecipitation, and immunoblottingRabbit polyclonal antibodies against ZNF198 and LSD1 were generated using a ZNF198 fragment (residues 923–1377) and an LSD1 fragment (residues 171–852) as the antigens at Zymed and Yenzym, respectively. The following antibodies were purchased from Upstate: α-HDAC1 (05-614), α-CoREST (07-455). Large-scale immuno-purification of ZNF198-containing protein complexes was performed as described [54]. For IP and western experiments, HeLa Tet-on cells from a 10-cm dish were washed and harvested in cold PBS 2 days after transfection and lysed in 1 ml of buffer C supplemented with 0.5 µM okadaic acid. Antibodies immobilized on Affi-prep protein A beads were incubated with the lysates for 2 hrs at 4°C. After washing, the beads were dissolved in SDS sample buffer and analyzed by SDS-PAGE following by immunoblotting. For immunoblotting, crude sera were used at 1∶1000 dilution while purified antibodies were used at a final concentration of 1 µg/ml.\nIn vitro binding and sumoylation assays\nIn vitro transcription and translation and in vitro sumoylation assays were performed as previously described [33]. For binding assays, HDAC1-FLAG, GST-CoREST, or GST-SUMO1/2 proteins together with other proteins were incubated with 5–10 µl M2 agarose (Sigma) or glutathione-sepharose 4B (Amersham) beads in 50 µl binding solution (TBS supplemented with 0.05% Tween-20 and 1 mM DTT) for 1 hr. After washing, the beads were then incubated in 50 µl blocking solution (TBS supplemented with 0.05% Tween-20, 5% dry milk, 1 mM DTT) for 1 hr at room temperature. The appropriate recombinant proteins or 5 µl 35S-labeled in vitro translated proteins were incubated with the beads for 1 hr at room temperature. Beads were then washed four times with the binding solution, boiled in SDS sample buffer, and subjected to SDS-PAGE followed by Coomassie Blue staining and autoradiography. For binding reactions containing ZNF198, 100 µM ZnCl2 was included in all buffers.Reverse transcription and quantitative PCRRNA from U2OS cells grown on 6-well plates and transfected with siRNAs was extracted using TriZOL reagent (Invitrogen) followed by RNAeasy RNA purification kit (Qiagen). RNA was then subjected to DNase digestion and inactivation followed by reverse transcription using random hexamers as primers. 2.5 µl of this cDNA was then used for quantitative PCR in 20 µl reactions using a 2× SYBR Green mix (Bio-Rad). The primers used were: SCN3A-Fwd (5′-ATGCTGGGCTTTGTTATGCT-3′), SCN3A-Rev (5′-TGGCTTGGCTTCAGTTTTCT-3); Cyclophilin B-Fwd: (5′-GGAGATGGCACAGGAGGAA-3′), Cyclophilin B-Rev (5′-GCCCGTAGTGCTTCAGTTT-3′); E-cadherin-Fwd (5′-GGATGACACAGCGTGAGAGA-3′), E-cadherin-Rev (5′-ACAGGATGGCTGAAGGTGAC-3′), NCAM2-Fwd (5′-CACGTTCACTGAAGGCGATA-3′), NCAM2-Rev (5′-GCTGCCCTTTGACTTCGATA-3′). KRT17-Fwd (5′-ATGCAGGCCTTGGAGATAGA-3′), KRT17-Rev (5′-AGGGATGCTTTCATGCTGAG-3′). All primers were validated as described [55].Chromatin immunoprecipitation (ChIP)ChIP experiments were performed as described [56]. About 1×107 HeLa Tet-on cells was used for each IP. Quantitative PCR was performed with 2.5 µl of eluted DNA, using the following primers: KRT17 ChIP-Fwd (5′-GGATAGGCTCTCGGTCTCCT-3′), KRT17 ChIP-Rev (5′-GTCTTTCACCCCACACTGCT-3′), GAPDH ChIP-Fwd (5′-TGTGCCCAAGACCTCTTTTC-3′), GAPDH ChIP-Rev (5′-TATTGAGGGCAGGGTGAGTC-3′).Supporting InformationFigure S1Domain architecture of MYM domain-containing proteins. The following color schemes from this illustration are used throughout the manuscript: MYM-type zinc fingers (MYM, red); proline/valine-rich domain (P/V-rich, green); Cre-like domain (CLD, gold); glutamine-rich domain (Q-rich, gray); potassium-tetramerization domain (K-tetra, black); and transposase-like domain (teal). Non-cysteine residues at zinc-coordinating positions in certain MYM domains are indicated by asterisks. Scale bar indicates 100 amino acids. The Cre-like domain of KCTD1 was used for 3D-Jury analysis (http://Bioinfo.Pl/Meta).(0.50 MB TIF)Click here for additional data file.Figure S2ZNF198 does not stimulate the sumoylation of HDAC1 or LSD1. (A) 35S-labeled in vitro translated ZNF198 was incubated with sumoylation enzymes (E1 and E2) and ATP in the presence or absence of SUMO2. The reaction mixtures were separated by SDS-PAGE and analyzed using a phosphoimager. The bands of unmodified and sumoylated ZNF198 are labeled. (B) Mixtures of His-LSD1 (300 ng), HDAC1-FLAG (300 ng), and GST-CoREST (100 ng) were subjected to in vitro sumoylation reactions in the presence or absence of His-ZNF198 (1–2 µg). The reaction mixtures were blotted with anti-FLAG (top panel) or anti-LSD1 (bottom panel).(0.33 MB TIF)Click here for additional data file.\n\nREFERENCES:\n1. LugerK\n2003\nStructure and dynamic behavior of nucleosomes.\nCurr Opin Genet Dev\n13\n127\n135\n12672489\n2. DaveyCASargentDFLugerKMaederAWRichmondTJ\n2002\nSolvent mediated interactions in the structure of the nucleosome core particle at 1.9 Å resolution.\nJ Mol Biol\n319\n1097\n1113\n12079350\n3. LugerKMaderAWRichmondRKSargentDFRichmondTJ\n1997\nCrystal structure of the nucleosome core particle at 2.8 Å resolution.\nNature\n389\n251\n260\n9305837\n4. LugerKHansenJC\n2005\nNucleosome and chromatin fiber dynamics.\nCurr Opin Struct Biol\n15\n188\n196\n15837178\n5. JenuweinTAllisCD\n2001\nTranslating the histone code.\nScience\n293\n1074\n1080\n11498575\n6. MartinCZhangY\n2005\nThe diverse functions of histone lysine methylation.\nNat Rev Mol Cell Biol\n6\n838\n849\n16261189\n7. BernsteinBEKamalMLindblad-TohKBekiranovSBaileyDK\n2005\nGenomic maps and comparative analysis of histone modifications in human and mouse.\nCell\n120\n169\n181\n15680324\n8. Santos-RosaHSchneiderRBannisterAJSherriffJBernsteinBE\n2002\nActive genes are tri-methylated at K4 of histone H3.\nNature\n419\n407\n411\n12353038\n9. BarskiACuddapahSCuiKRohTYSchonesDE\n2007\nHigh-resolution profiling of histone methylations in the human genome.\nCell\n129\n823\n837\n17512414\n10. ShiYLanFMatsonCMulliganPWhetstineJR\n2004\nHistone demethylation mediated by the nuclear amine oxidase homolog LSD1.\nCell\n119\n941\n953\n15620353\n11. FornerisFBindaCVanoniMAMatteviABattaglioliE\n2005\nHistone demethylation catalysed by LSD1 is a flavin-dependent oxidative process.\nFEBS Lett\n579\n2203\n2207\n15811342\n12. YangMGockeCBLuoXBorekDTomchickDR\n2006\nStructural basis for CoREST-dependent demethylation of nucleosomes by the human LSD1 histone demethylase.\nMol Cell\n23\n377\n387\n16885027\n13. LeeMGWynderCCoochNShiekhattarR\n2005\nAn essential role for CoREST in nucleosomal histone 3 lysine 4 demethylation.\nNature\n437\n432\n435\n16079794\n14. ShiYJMatsonCLanFIwaseSBabaT\n2005\nRegulation of LSD1 histone demethylase activity by its associated factors.\nMol Cell\n19\n857\n864\n16140033\n15. HumphreyGWWangYRussanovaVRHiraiTQinJ\n2001\nStable histone deacetylase complexes distinguished by the presence of SANT domain proteins CoREST/kiaa0071 and Mta-L1.\nJ Biol Chem\n276\n6817\n6824\n11102443\n16. ShiYSawadaJSuiGAffar elBWhetstineJR\n2003\nCoordinated histone modifications mediated by a CtBP co-repressor complex.\nNature\n422\n735\n738\n12700765\n17. YouATongJKGrozingerCMSchreiberSL\n2001\nCoREST is an integral component of the CoREST- human histone deacetylase complex.\nProc Natl Acad Sci U S A\n98\n1454\n1458\n11171972\n18. HakimiMADongYLaneWSSpeicherDWShiekhattarR\n2003\nA candidate X-linked mental retardation gene is a component of a new family of histone deacetylase-containing complexes.\nJ Biol Chem\n278\n7234\n7239\n12493763\n19. LeeMGWynderCBocharDAHakimiMACoochN\n2006\nFunctional interplay between histone demethylase and deacetylase enzymes.\nMol Cell Biol\n26\n6395\n6402\n16914725\n20. AndresMEBurgerCPeral-RubioMJBattaglioliEAndersonME\n1999\nCoREST: a functional corepressor required for regulation of neural-specific gene expression.\nProc Natl Acad Sci U S A\n96\n9873\n9878\n10449787\n21. BallasNBattaglioliEAtoufFAndresMEChenowethJ\n2001\nRegulation of neuronal traits by a novel transcriptional complex.\nNeuron\n31\n353\n365\n11516394\n22. WangJScullyKZhuXCaiLZhangJ\n2007\nOpposing LSD1 complexes function in developmental gene activation and repression programmes.\nNature\n446\n882\n887\n17392792\n23. Garcia-BassetsIKwonYSTeleseFPrefontaineGGHuttKR\n2007\nHistone methylation-dependent mechanisms impose ligand dependency for gene activation by nuclear receptors.\nCell\n128\n505\n518\n17289570\n24. MetzgerEWissmannMYinNMullerJMSchneiderR\n2005\nLSD1 demethylates repressive histone marks to promote androgen-receptor-dependent transcription.\nNature\n437\n436\n439\n16079795\n25. SmedleyDHamoudiRLuYJCooperCShipleyJ\n1999\nCloning and mapping of members of the MYM family.\nGenomics\n60\n244\n247\n10486218\n26. XiaoSNalaboluSRAsterJCMaJAbruzzoL\n1998\nFGFR1 is fused with a novel zinc-finger gene, ZNF198, in the t(8;13) leukaemia/lymphoma syndrome.\nNat Genet\n18\n84\n87\n9425908\n27. SmedleyDHamoudiRClarkJWarrenWAbdul-RaufM\n1998\nThe t(8;13)(p11;q11–12) rearrangement associated with an atypical myeloproliferative disorder fuses the fibroblast growth factor receptor 1 gene to a novel gene RAMP.\nHum Mol Genet\n7\n637\n642\n9499416\n28. ReiterASohalJKulkarniSChaseAMacdonaldDH\n1998\nConsistent fusion of ZNF198 to the fibroblast growth factor receptor-1 in the t(8;13)(p11;q12) myeloproliferative syndrome.\nBlood\n92\n1735\n1742\n9716603\n29. van der MaarelSMScholtenIHHuberIPhilippeCSuijkerbuijkRF\n1996\nCloning and characterization of DXS6673E, a candidate gene for X-linked mental retardation in Xq13.1.\nHum Mol Genet\n5\n887\n897\n8817323\n30. RaffaGDCenciGSiriacoGGoldbergMLGattiM\n2005\nThe putative Drosophila transcription factor woc is required to prevent telomeric fusions.\nMol Cell\n20\n821\n831\n16364909\n31. WarrenJTWismarJSubrahmanyamBGilbertLI\n2001\nWoc (without children) gene control of ecdysone biosynthesis in Drosophila melanogaster.\nMol Cell Endocrinol\n181\n1\n14\n11476936\n32. WismarJHabtemichaelNWarrenJTDaiJDGilbertLI\n2000\nThe mutation without children(rgl) causes ecdysteroid deficiency in third-instar larvae of Drosophila melanogaster.\nDev Biol\n226\n1\n17\n10993670\n33. GockeCYuHKangJ\n2004\nSystematic identification and analysis of mammalian small ubiquitin-like modifier substrates.\nJ Biol Chem\n280\n5004\n5012\n15561718\n34. WohlschlegelJAJohnsonESReedSIYatesJR3rd\n2004\nGlobal analysis of protein sumoylation in Saccharomyces cerevisiae.\nJ Biol Chem\n279\n45662\n45668\n15326169\n35. KunapuliPKasyapaCSChinSFCaldasCCowellJK\n2006\nZNF198, a zinc finger protein rearranged in myeloproliferative disease, localizes to the PML nuclear bodies and interacts with SUMO-1 and PML.\nExp Cell Res\n312\n3739\n3751\n17027752\n36. ZhaoYKwonSWAnselmoAKaurKWhiteMA\n2004\nBroad spectrum identification of cellular small ubiquitin-related modifier (SUMO) substrate proteins.\nJ Biol Chem\n279\n20999\n21002\n15016812\n37. DavidGNeptuneMADePinhoRA\n2002\nSUMO-1 modification of histone deacetylase 1 (HDAC1) modulates its biological activities.\nJ Biol Chem\n277\n23658\n23663\n11960997\n38. HeckerCMRabillerMHaglundKBayerPDikicI\n2006\nSpecification of SUMO1- and SUMO2-interacting motifs.\nJ Biol Chem\n281\n16117\n16127\n16524884\n39. RosendorffASakakibaraSLuSKieffEXuanY\n2006\nNXP-2 association with SUMO-2 depends on lysines required for transcriptional repression.\nProc Natl Acad Sci U S A\n103\n5308\n5313\n16567619\n40. GinalskiKElofssonAFischerDRychlewskiL\n2003\n3D-Jury: a simple approach to improve protein structure predictions.\nBioinformatics\n19\n1015\n1018\n12761065\n41. LahertyCDYangWMSunJMDavieJRSetoE\n1997\nHistone deacetylases associated with the mSin3 corepressor mediate mad transcriptional repression.\nCell\n89\n349\n356\n9150134\n42. LeeMGWynderCSchmidtDMMcCaffertyDGShiekhattarR\n2006\nHistone H3 lysine 4 demethylation is a target of nonselective antidepressive medications.\nChem Biol\n13\n563\n567\n16793513\n43. TodorovITAttaranAKearseySE\n1995\nBM28, a human member of the MCM2-3-5 family, is displaced from chromatin during DNA replication.\nJ Cell Biol\n129\n1433\n1445\n7790346\n44. MendezJStillmanB\n2000\nChromatin association of human origin recognition complex, cdc6, and minichromosome maintenance proteins during the cell cycle: assembly of prereplication complexes in late mitosis.\nMol Cell Biol\n20\n8602\n8612\n11046155\n45. RomanowskiPMadineMALaskeyRA\n1996\nXMCM7, a novel member of the Xenopus MCM family, interacts with XMCM3 and colocalizes with it throughout replication.\nProc Natl Acad Sci U S A\n93\n10189\n10194\n8816774\n46. KotajaNKarvonenUJanneOAPalvimoJJ\n2002\nPIAS proteins modulate transcription factors by functioning as SUMO-1 ligases.\nMol Cell Biol\n22\n5222\n5234\n12077349\n47. PichlerAGastASeelerJSDejeanAMelchiorF\n2002\nThe nucleoporin RanBP2 has SUMO1 E3 ligase activity.\nCell\n108\n109\n120\n11792325\n48. ChengJWangDWangZYehET\n2004\nSENP1 enhances androgen receptor-dependent transcription through desumoylation of histone deacetylase 1.\nMol Cell Biol\n24\n6021\n6028\n15199155\n49. PflumMKTongJKLaneWSSchreiberSL\n2001\nHistone deacetylase 1 phosphorylation promotes enzymatic activity and complex formation.\nJ Biol Chem\n276\n47733\n47741\n11602581\n50. SchoenherrCJAndersonDJ\n1995\nThe neuron-restrictive silencer factor (NRSF): a coordinate repressor of multiple neuron-specific genes.\nScience\n267\n1360\n1363\n7871435\n51. ChongJATapia-RamirezJKimSToledo-AralJJZhengY\n1995\nREST: a mammalian silencer protein that restricts sodium channel gene expression to neurons.\nCell\n80\n949\n957\n7697725\n52. PavletichNPPaboCO\n1991\nZinc finger-DNA recognition: crystal structure of a Zif268-DNA complex at 2.1 A.\nScience\n252\n809\n817\n2028256\n53. KadrmasJLBeckerleMC\n2004\nThe LIM domain: from the cytoskeleton to the nucleus.\nNat Rev Mol Cell Biol\n5\n920\n931\n15520811\n54. TangZShuHQiWMahmoodNAMumbyMC\n2006\nPP2A is required for centromeric localization of Sgo1 and proper chromosome segregation.\nDev Cell\n10\n575\n585\n16580887\n55. BookoutALMangelsdorfDJ\n2003\nQuantitative real-time PCR protocol for analysis of nuclear receptor signaling pathways.\nNucl Recept Signal\n1\ne012\n16604184\n56. PottsPRPorteusMHYuH\n2006\nHuman SMC5/6 complex promotes sister chromatid homologous recombination by recruiting the SMC1/3 cohesin complex to double-strand breaks.\nEMBO J\n25\n3377\n3388\n16810316"
+}

batch_3/PMC2532750.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2532750",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2532750\nAUTHORS: Genc Basha, Gregory Lizée, Anna T. Reinicke, Robyn P. Seipp, Kyla D. Omilusik, Wilfred A. Jefferies\n\nABSTRACT:\nBackgroundCross-presentation by dendritic cells (DCs) is a crucial prerequisite for effective priming of cytotoxic T-cell responses against bacterial, viral and tumor antigens; however, this antigen presentation pathway remains poorly defined.Methodology/Principal FindingsIn order to develop a comprehensive understanding of this process, we tested the hypothesis that the internalization of MHC class I molecules (MHC-I) from the cell surface is directly involved in cross-presentation pathway and the loading of antigenic peptides. Here we provide the first examination of the internalization of MHC-I in DCs and we demonstrate that the cytoplasmic domain of MHC-I appears to act as an addressin domain to route MHC-I to both endosomal and lysosomal compartments of DCs, where it is demonstrated that loading of peptides derived from exogenously-derived proteins occurs. Furthermore, by chasing MHC-I from the cell surface of normal and transgenic DCs expressing mutant forms of MHC-I, we observe that a tyrosine-based endocytic trafficking motif is required for the constitutive internalization of MHC-I molecules from the cell surface into early endosomes and subsequently deep into lysosomal peptide-loading compartments. Finally, our data support the concept that multiple pathways of peptide loading of cross-presented antigens may exist depending on the chemical nature and size of the antigen requiring processing.Conclusions/SignificanceWe conclude that DCs have ‘hijacked’ and adapted a common vacuolar/endocytic intracellular trafficking pathway to facilitate MHC I access to the endosomal and lysosomal compartments where antigen processing and loading and antigen cross-presentation takes place.\n\nBODY:\nIntroductionIn order to generate appropriate CD8+ T cell-mediated immune responses to viral, bacterial, self, or tumor-associated protein antigens, professional antigen-presenting cells (pAPCs) must acquire these antigens from the extracellular milieu, process them into antigenic peptides, load them onto MHC class I (MHC-I) molecules, and present them at the cell surface [1]–[4]. This process of “cross-presentation” is known to occur most efficiently in dendritic cells (DCs), but several mechanistic details remain unclear.Three non-mutually exclusive pathways have been proposed to explain cross-presentation. In the vacuolar pathway, DCs internalize exogenous antigens and transport them through the endocytic pathway, where the internalized proteins are processed by cathepsins and other proteases into antigenic peptides [5], [6]. In this model, loading of peptides onto MHC-I molecules occurs directly within early and late endosomal/lysosomal compartments, often in a transporter associated with antigen processing-independent manner [7], [8]. In the endosome-to-cytosol pathway, internalized protein antigens are transported out of endosomes and into the cytosol by a mechanism that may involve ER-associated retrotranslocation complexes [9]–[12]. Subsequently, protein antigens follow the classical pathway of direct MHC-I presentation involving proteasome-mediated protein degradation and TAP transport of antigenic peptides into the endoplasmic reticulum (ER) for loading onto nascent MHC-I molecules [13]. A third proposed model of cross-presentation describes a unique intracellular compartment of exogenous antigen loading, termed the ergosome, which involves a possible fusion of ER with phagosomes containing internalized antigenic cargo [14], [15]. Although the biogenesis of ergosomes is a matter of some dispute [16], the concept of an ER-phagosome ‘mix’ compartment remains an attractive model for explaining where peptide-receptive MHC-I molecules could intersect with a relatively high concentration of exogenous antigens, presumably a crucial prerequisite for efficient cross-presentation.MHC-I molecules have been reported to reside within endosomes and lysosomes of DCs [17], [18], but their source and intracellular trafficking routes have not yet been clearly defined. Although recycling between the cell membrane and endosomal compartments has been demonstrated for MHC-I in both in T cells and macrophages 19, 20, little is known about MHC-I endocytic trafficking in DCs or how this relates to their function in cross-presenation. We recently demonstrated a crucial role for a conserved, exon 6-encoded MHC-I cytoplasmic tyrosine in DC cross-presentation of exogenous antigens, and in the generation of in vivo cytolytic T lymphocyte responses against viruses [21]. Although we showed that mutation of the cytoplasmic tyrosine drastically reduced localization of MHC-I within endolysosomal compartments of DCs, the dynamic contribution of surface MHC-I to these compartments was not addressed. In this study, we define the mechanism whereby MHC-I molecules gain direct access to the intracellular compartment in DCs where peptide loading takes place. This mechanism underlies the unique ability of DCs to cross-present antigens and promote and initiate primary adaptive immune responses.Materials and MethodsMHC class I internalization in dendritic cellsAll experiments involving mice have been approved and performed in accordance with Canadian Council on Animal Care requirements. Splenic DCs were isolated as described (21),cultured overnight at 37°C in complete RPMI and the next day washed in PBS 3 times, aliquoted and labeled with Fc blocker 2.4G2 Fcγ III/II (BD PharMingen) for 30 min at 4°C to exclude binding of antibodies to Fc receptor, followed by labeling with AF6-88.5 (H-2Kb) specific monoclonal antibody conjugated to fluorescein isothiocyanate (FITC) for 30 min at 0°C in 96-well plates. Next, they were incubated at 37°C and at different time points as indicated, they were chilled to 4°C on ice before fixing in 2% paraformaldehyde. Recovered DCs were pipetted onto coverslips, mounted on slides and examined with a Nikon multiphoton immunofluorescent confocal microscope (ICM).In order to assess the kinetics of internalization of MHC-I molecules, DCs were isolated as described, cultured at 37°C and the next day they were washed and stained with Fc blocker followed by AF6-88.5 (H-2Kb) and 36-7-5 (H-2Kk) antibodies (BD Biosciences, Mississauga, ON, Canada) conjugated to FITC and phycoerythrin (PE) respectively, at 4°C for 30 min in 96-well plates. Next, sample cells were placed at 37°C, whereas control cells were placed at 0°C and after different time points DCs were washed 3 times, fixed in ethanol to preserve FITC fluorescence and resuspended in 2% FCS PBS to reach an equal pH of 7.4 and prevent further quenching. DCs with fluorescently labeled H-2Kb molecules were examined using FACSCalibur™ (Becton Dickinson). Data were analyzed using FlowJo software to examine the H-2Kb and H-2Kk molecules following incubation at 37°C as an indication of MHC-I internalization.Intracellular DC colocalization and image quantificationSpleen-derived DCs from transgenic mice were aliquoted in 96-well plates and stained with H-2Kb-FITC antibody after blocking the Fc receptor. Next, DCs were resuspended in 200 µL of 37°C pre-warmed completed RPMI, mounted onto Poly-D-lysine pre-coated coverslips and incubated at 37°C, allowing antibody-bound H-2Kb molecules to internalize. At different time points, as indicated, internalization was stopped by soaking the coverslips in cold PBS. At the end of the last time point, all coverslips containing DCs were treated with 2% BSA in PBS followed by fixation with 2% paraformaldehyde. Spleen-derived DCs were then permeabilized with 0.1% saponin in 2% BSA PBS followed by incubation with goat anti-mouse EEA1 or LAMP-1 primary antibodies (Santa Cruz Biotechnology, Santa Cruz, CA, USA). Secondary Alexa-568-conjugated rabbit anti-goat antibody (Molecular Probes, OR, USA) was used as a detection reagent. Isotype control antibodies were used in all confocal microscopy experiments to confirm the specificity of antibody staining. Images were acquired using a Nikon-C1, TE2000-U ICM and the EZ-C1 software. Data were analyzed using ImageJ.1 to select single slices and Adobe Photoshop 9.0 to merge images obtained from Red and Green channels.To visualize the acquisition of exogenous OVA peptide by internalized H-2Kb molecules, DCs surface-labeled with H-2Kb-FITC antibody as described above were incubated at 37°C in complete RPMI containing either 5 mg/mL ovalbumin protein (Worthington, NJ) or bovine serum albumin control protein. Next, DCs were mounted onto coverslips and incubated at 37°C for 6 hours, allowing ovalbumin protein uptake, processing and loading onto H-2Kb molecules. Incubation was halted by chilling the coverslips in ice-cold PBS. DCs were then fixed and permeabilized as described and following Fc blocking for 30 min, they were co-stained with goat anti-mouse EEA1, LAMP-1 or rat anti-mouse Giantin (Golgi marker) and with anti-H-2Kb/OVA257–264 antibody. Goat anti mouse or Rabbit anti-goat coupled to Alexa-568 (Molecular Probes) were used to detect the endosomes-EEA1 or lysosomes-LAMP-1 and goat anti-Rat coupled to Alexa-568 was used to detect Giantin-Golgi. Goat anti-mouse Alexa-647 labeled was used to visualize the H-2Kb/OVA257–264 complexes. Dendritic cell surface-derived MHC-I were detected by locating intracellular fluorescent-green punctate dots present. Fluorescence was visualized by ICM using the 488-nm (green) 568-nm (red) and 633-nm (blue) laser lines for excitation of the appropriate fluorochromes. Data were analyzed using ImageJ.1 to select single slices and Adobe Photoshop 9.0 to merge images obtained upon excitation of fluorochromes obtained by red, green and blue channels. Co-localization of three different molecules was evaluated by the presence, intensity and distribution of the white color resulting from the overlapping of green, red and blue.For quantification of colocalization, a total of approximately 50 DCs were examined at 60× magnification. Quantitative confocal image analysis was done by single cell identification using Openlab software and the relative fluorescent intensity of green, red, blue, yellow, purple, light blue and white pixels was assessed. The relative fluorescent intensity of all individual colors was then expressed as percent of total fluorescence intensity.\nIn vitro cross-presentation and T cell proliferationPrimary DCs were isolated from bone marrow precursors of KbWT, Δ7, and ΔY transgenic mice, by culturing in vitro in X63-Ag8-plasmacytoma-derived GM-CSF (gift from David Gray, University of Edinburgh, UK) at 20 µL/10 mL of RPMI complete media. On day 8, bone marrow derived DCs (bmDCs) were stained with antibodies against H-2Kb (AF6.88), I-Ab (AF6.120.1) and CD11c (HL3) (Pharmingen) to test their purity then incubated for 6 hours with 10 mg/mL OVA or with bovine serum albumin or synthetic immunodominant peptide OVA257–264 as controls. Next, they were washed in cold PBS and fixed in 0.0005% glutaraldehide to preserve surface Kb/OVA257–264 complexes. B3Z T cell hybridoma (kind gift from Nilabh Shastri at UC Berkeley) were labeled with 10 µM CFSE (Molecular Probes) at 37°C for 30 min and 105 cells were co-cultured with OVA-pulsed transgenic mice DCs in 48-well plates at 50∶50 ratio for 24 and 48 hrs. Next, the mixed cells were washed and stained with anti-CD3-PE labeled antibody (BD Pharmingen) to detect the T cells and analyzed by flow cytometry. Data were acquired using FACSCalibur™ and the CFSE fluorescence (FL1) and CD3 (FL2) positive populations of B3Z-T cells were analyzed with FlowJo software to assess their proliferation.Statistical AnalysisStudent's T test was used to compare the numbers of fluorescently labeled H-2Kb in the cytoplasm of transgenic mouse DCs assessed by ICM and the percent down regulation of H-2Kb molecules in transgenic mice DCs at 0°C compared to upon incubation at 37°C as measured by FACS. The difference between two populations was considered statistically significant if P<0.05 (two-tailed, two sample equal variance).ResultsConstitutive internalization of MHC-I in dendritic cells is differentially affected by cytoplasmic tail mutationsTo elucidate the role of distinct cytoplasmic domain motifs in MHC-I molecular trafficking and function in vivo, we generated transgenic mice, described previously [21], expressing either wild type H-2Kb (KbWT), Kb containing a point mutation of the highly conserved exon 6-encoded tyrosine (ΔY), or Kb containing a deletion of 13 amino acids encoded by exon 7, including at least one highly conserved serine phosphorylation site (Δ7, Figure 1A). Several distinct founder lines from each of the three strains were obtained by backcrossing the transgenic mice onto a C3H background (haplotype H-2k).10.1371/journal.pone.0003247.g001Figure 1Constitutive MHC-I internalization in DCs is differentially controlled by cytoplasmic tyrosine- and exon 7-dependent mechanisms.(A) Amino acid sequences of the cytoplasmic domains of wild-type H-2Kb and the two cytoplasmic tail mutants Δ7 and ΔY. The asterisk denotes a known conserved serine phosphorylation site. The Δ7 mutant contains a deletion of the 13 amino acids comprising exon 7, indicated as dashed lines. Highlighted amino acids indicate conserved tyrosine and serine residues. TM, transmembrane domain. (B) Splenic dendritic cells isolated from KbWT, and Δ7 and ΔY transgenic mice were labeled with FITC-conjugated H-2Kb-specific mAb, washed, and incubated at 37°C for the indicated time points. DCs were then imaged using confocal fluorescence microscopy to visualize internalized MHC-I-containing vesicles. Data are representative of at least 3 images captured from 2 independent experiments.To determine whether DCs constitutively internalize surface MHC-I, as has been reported for activated T lymphocytes (22), we performed MHC-I endocytosis experiments on spleen-derived DCs. Kb-specific, FITC-labeled antibodies were used to stain DCs and internalization of labeled Kb molecules was evaluated over time. Fluorescently-labeled intracellular vesicles appeared after 30 min in the KbWT transgenic mice DCs and were observed in the majority of DCs examined (Figure 1B). In contrast, DCs from both Δ7 and ΔY transgenic mice showed an almost complete absence of these fluorescent vesicles up to 90 minutes of chase, indicating that spontaneous internalization of Kb molecules was abrogated.In order to more quantitatively measure MHC-I internalization in DCs, we used flow cytometric analysis following staining with labeled antibodies specific for Kb and Kk. As shown in Figure 2A, 2B, and 2C both Kb and Kk were internalized very rapidly in DCs derived from KbWT mice, with both molecules demonstrating close to 50% internalization at 30 minutes of chase. By contrast, only 4% of labeled Kb molecules were internalized from the surface of ΔY-derived DCs over the same time period (Figure 2A and B). As an internal control for MHC-I internalization, these ΔY-derived DCs efficiently internalized endogenously-expressed wild-type Kk molecules (33% internalization at 30 minutes of chase), albeit not as rapidly as observed for KbWT DCs (Figure 2C). DCs from Δ7 mice showed an intermediate phenotype, with ∼20% of surface Kb molecules being internalized at 30 minutes of chase time (Figure 2A and B). These results demonstrate that, while both the Δ7 and ΔY cytoplasmic tail alterations led to impaired MHC-I endocytosis in DCs, point mutation of the single conserved tyrosine residue resulted in a much more pronounced internalization defect compared to complete deletion of the 13 amino acids comprising exon 7.10.1371/journal.pone.0003247.g002Figure 2Quantification of MHC-I internalization in DCs.The dynamics of MHC Class I internalization is assessed by the reduced mean fluorescence units of FITC-labeled H-2Kb surface expression. Following labeling with (A and C) FITC-conjugated H-2Kb- or (B and D) PE-conjugated H-2Kk-specific antibodies and internalization at 37°C for the indicated time points, flow cytometric analysis was conducted to assess internalization of Kb and Kk molecules, as measured by the reduction in FITC and PE mean fluorescence intensities over time, respectively. Data are representative of 3 different experiments performed in triplicate.Cytoplasmic tail motifs differentially regulate trafficking of internalized MHC-I molecules through endocytic compartments of DCsIn order to investigate which intracellular compartments Kb molecules traffic through following internalization from the DC surface, spleen-derived DCs from all transgenic mice were initially surface-labeled at 4°C with Kb-specific mAbs. Following washing and incubation at 37°C to induce MHC-I internalization, DCs were fixed and permeabilized at different time points and stained with antibodies specific for EEA-1, a marker of early endosomes, or LAMP-1, a marker of late endosomes and lysosomes. Both wild-type and exon 7-mutated Kb molecules showed significant colocalization with early endosomal markers at 1.5 and 3 hours of chase. By contrast, ΔY molecules showed minimal or no overlap with early endosomes at those time points (Figures 3A to 3C). Notably, surface ΔY molecules were also largely excluded from LAMP-1-positive late endosomes and lysosomes, even at 5 hours of chase (Figures 4A to 4D). By contrast, co-localization of surface-derived wild-type Kb molecules with LAMP-1 became evident after only 90 minutes of incubation at 37°C (Figure 4B). Interestingly, although little or no colocalization of Δ7 molecules with LAMP-1 could be observed at earlier time points, significant overlap could be detected at five hours of chase (Figure 4B to 4D).10.1371/journal.pone.0003247.g003Figure 3MHC-I cell surface-to-endosome trafficking in DCs is differentially abrogated by mutations in cytoplasmic tyrosine or exon 7-encoded determinants.(A to C) Splenic DCs isolated from KbWT, and Δ7 and ΔY transgenic mice were mounted on coverslips, labeled with FITC-conjugated H-2Kb-specific mAb, washed, then incubated at 37°C for the indicated times. DCs were then fixed, permeabilized and counterstained for EEA-1. Images were acquired using a multiphoton fluorescence confocal microscope. Yellow color indicates co-localization of surface-derived H-2Kb (green) with EEA-1 (red). Data are representative of at least 3 images captured from 2 independent experiments.10.1371/journal.pone.0003247.g004Figure 4MHC-I cell surface-to-lysosome trafficking in DCs is impaired by mutation in the cytoplasmic tyrosine residue.(A to D) Splenic DCs isolated from KbWT, and Δ7 and ΔY transgenic mice were mounted on coverslips, labeled with FITC-conjugated H-2Kb-specific mAb, washed, and incubated at 37°C for the indicated times. DCs were then fixed, permeabilized, and counterstained for LAMP-1. Yellow color indicates co-localization of surface-derived H-2Kb (green) with LAMP-1 (red). Data are representative of at least 3 images captured from 2 independent experiments.Taken together, these data indicate that while Kb molecules lacking exon 7 show significantly delayed internalization kinetics compared with wild-type Kb molecules, they nonetheless can traffic from the cell surface into both early and late endosomal/lysosomal compartments of DCs. In contrast, tyrosine point-mutated Kb molecules were almost completely abrogated in their ability to traffic from the cell surface to endosomal/lysosomal compartments containing either EEA1 or LAMP-1.Cytoplasmic tail mutations reveal distinct pathways of MHC-I trafficking through exogenous peptide-loading compartments of DCsPrior to assessing the entry of surface-derived Kb molecules into endocytic compartments, a competition assay using flow cytometry was designed. No change in H-2Kb binding was observed with sequential staining of Kb molecules preceded or followed by staining of H-2Kb/OVAp complexes (refer to Supporting Information: Text S1 and Figure S1). The same was observed for the H-2Kb/OVAp surface expression indicating that there was no competition of these antibodies for binding to specific sites. In order to assess how entry of surface-derived Kb molecules into endocytic compartments coincides with exogenous antigen loading, Kb surface-labeled DCs were pulsed with soluble OVA protein for 6 hours to allow simultaneous Kb-FITC internalization and OVA uptake and antigen processing. DCs were then fixed, permeabilized and stained with fluorescently-labeled antibodies specific for either EEA1, LAMP-1, or Giantin, a Golgi marker (red) in addition to 25.D1, an antibody that specifically stains Kb/OVA257–264 complexes (blue). Stained DCs were then examined and imaged by confocal microscopy to assess colocalization of the three fluorophores. In order to obtain a more precise quantification of this colocalization, we used image analysis software to obtain pixel counts for the seven colors present in each overlaid confocal image [green, red, blue, yellow (g+r), light blue (g+b), pink (r+b), and white (g+r+b)]. While Figure 5A, B and C shows confocal DC images representative of each mouse strain, quantification of fluorescence intensity was performed on images from a further 30 to 50 individual DCs from each mouse strain in order to generate the quantitative data summarized in Figure 5 (D to F).10.1371/journal.pone.0003247.g005Figure 5Cytoplasmic tail mutations significantly reduce the contribution of surface MHC-I molecules to endosomal and lysosomal peptide-loading compartments.(A to C) Splenic DCs isolated from KbWT, Δ7, and ΔY transgenic mice were labeled with FITC-conjugated H-2Kb-specific mAb, washed and incubated for 6 hr at 37°C in 5 mg/mL ovalbumin protein. DCs were then labeled with mAbs specific for (A) early endosomal antigen, EEA-1, (B) lysosomal marker LAMP-1 and (C) Golgi marker Giantin. All DCs were simultaneously co-stained with purified 25.D1.16 (anti-H-2Kb/OVA257–264) antibody. Cellular markers EEA-1, LAMP-1 and Giantin were visualized by staining with secondary antibodies coupled to Alexa-568 (red) whereas H-2Kb/OVA257–264 complexes were visualized by staining with secondary antibody conjugated to Alexa-647 (blue). Three-color fluorescence was detected by laser scanning confocal microscopy of 488-nm (green), 568-nm (red), and 633-nm (blue) wavelengths. Photographs depict three-color image overlays to assess colocalization of the three markers. White color indicates a triple overlap of all three markers (green+red+blue), whereas yellow (green+red), pink (red+blue), and light blue (green+blue) indicate overlap of two of the three markers. D to F shows a quantitative assessment of internalized MHC-I and MHC-I/peptide complexes within intracellular compartments of DCs. Three-color confocal overlay images of DCs, as shown in Figure 5, were analyzed for relative fluorescent color (pixel) intensity in order to obtain a quantitative measure of fluorophore colocalization. For each data set, 30 to 50 individual DCs derived from each of the indicated mouse strains were analyzed. The green color indicates surface-labeled H-2Kb, the blue color indicates Kb-OVA257–264 peptide complexes, and red color indicates either (A) early endosomal antigen (EEA-1), (B) LAMP-1, or (C) Giantin. White pixels indicate triple overlap of all three markers (green+red+blue), whereas yellow (green+red), pink (red+blue), and light blue (green+blue) indicate overlap of two of the three markers. Graph depicts individual color pixel mean percentages and standard deviations, as calculated by dividing the number of pixels of a given color by the total number of colored pixels counted. Data are representative of at least 3 images captured from 4 independent experiments.When DCs from all transgenic mice strains were incubated at 4°C, no loading of OVA peptide antigen (blue) was detectable and Kb molecules remained at the cell surface, as expected (not shown). Upon incubation at 37°C for 6 hrs, both surface Kb and Kb-OVA complexes showed abundant colocalization (white) with EEA1-positive intracellular compartments in KbWT-derived DCs (Figure 5A). Notably, overlapping of EEA-1 and Kb-OVA complexes without surface Kb (indicated by pink color) was significantly higher (p = 0.002) in Δ7-derived DCs (Figure 5D). By contrast, early endosomes from ΔY-derived DCs contained no detectable surface-derived Kb or Kb-OVA complexes (Figure 5A).Similarly, KbWT-derived DCs demonstrated extensive triple overlap of surface-derived Kb and OVA-loaded Kb molecules with LAMP-1-positive compartments (Figure 5A). By comparison, DCs derived from Δ7 and ΔY mice showed 2-fold (p = 0.004) and 6-fold (p = 0.0001) less triple colocalization, respectively (Figures 5A and 5E). Furthermore, in KbWT-derived DCs the majority (>80%) of LAMP-1-positive compartments also contained surface-derived Kb, with ΔY-derived DCs showing approximately 5-fold less (p = 0.0001) overlap of these two markers and Δ7-derived DCs demonstrating an intermediate phenotype (Figure 5E). Conversely, ΔY-derived DCs showed a strikingly higher (p = 0.006) percentage of LAMP-1 alone (red pixels, not colocalized with the other two markers) compared with KbWT-derived DCs.Co-staining with the Golgi marker Giantin showed very little colocalization of surface Kb or Kb-OVA complexes with Golgi in DCs from KbWT and Δ7 mice. Interestingly, in ΔY-derived DCs, Giantin and surface Kb demonstrated a significant (p = 0.0004) degree of overlap (yellow) and some triple colocalization (white) (p = 0.007) was also apparent (Figures 5C and 5F).In summary, while surface KbWT molecules appear to form an abundant source of MHC-I molecules for early and late endosomal peptide-loading, surface ΔY molecules do not gain access to these peptide-loading sites due to aberrant intracellular trafficking. Surface Kb molecules that lack exon 7, by contrast, appear to traffic readily into early endosomal compartments, but are significantly delayed in their ability to traffic into late endosomes or lysosomes. Taken together, these data reveal distinct intracellular trafficking patterns for KbWT, Δ7 and ΔY molecules in DCs that may provide a plausible mechanism to explain their different abilities to cross-present exogenous antigens.Cross-presentation and T cell activation is impaired in ΔY-derived DCs following 6 hour incubation with OVA antigenTo correlate the impaired intracellular trafficking of MHC Class I molecules in our transgenic mouse DCs with functional impairment of OVA cross-presentation and T cell activation during the 6 hour time period, T cell proliferation was monitored for 48 hours following a 6 hour incubation of transgenic mouse DCs with OVA protein. Proliferation of CFSE-labeled T cells was detectable following 24 hours of co-incubation with KbWT transgenic OVA-pulsed DCs (Figure 6A). However, fewer T cells proliferated following incubation with Δ7-derived DCs while this was barely measurable following co-culture with ΔY-DCs. After 48 hours, more then half of the T cells incubated with KbWT-DCs had proliferated and although less proliferation was detected still a considerable fraction of T cells incubated with Δ7-DCs were dividing. However, when T cells were incubated with ΔY-DCs, over 90% of them did not proliferate indicating that cross-presentation of OVA antigen and T cell activation was severely impaired presumably due to defective internalization and intracellular trafficking of surface ΔY-Kbs. No difference in T cell activation was observed, between transgenic mice DCs following their incubation with OVA257–264.10.1371/journal.pone.0003247.g006Figure 6T cell activation is severely compromised in protein-pulsed DCs of transgenic mice containing the cytoplasmic tail mutation.bmDCs from KbWT, Δ7 and ΔY transgenic mice were isolated, incubated for 6 hours with 10 mg/mL OVA and co-cultured for 24 and 48 hours with B3Z-T hybridoma cells previously labeled with 1 µM of CFSE. Mixed cultures were then stained with anti-CD3-PE antibody and flow cytometry was conducted to examine the CFSE/CD3+ T cell population and proliferation. Histograms depict proliferating T cells following incubation with OVA-pulsed DCs (including one representative of OVA257–264) of transgenic mice for 24 (A) and 48 (B) hours. Data are representative of 1 experiment performed in triplicate.DiscussionAlthough cross-presentation by DCs is a crucial prerequisite for effective priming of cytotoxic T-cell responses against bacterial, viral, and tumor antigens in vivo, several mechanistic aspects of these pathways remain poorly defined. The results presented here delineate constitutive pathways of MHC-I intracellular transport in DCs, and illuminate the nature of the intracellular compartments of DCs where peptides from exogenously-derived antigens are loaded for cross-presentation. Importantly, the internalization data demonstrate that MHC Class I from the cell surface can enter a vesicular compartment that contains also complexes of MHC Class I and cross-presented antigen.Our microscopic analysis of spleen-derived DCs reveals for the first time, that wild-type (WT) surface MHC-I molecules are internalized and transported to early and late endosomal/lysosomal compartments, where peptides from exogenously-derived antigens can be loaded [13], [18], [21], [22]. While efficient MHC Class I loading with exogenous antigen has been shown to occur in early endosomes (13), some endosome-to-lysosome trafficking or exchange of MHC Class I containing antigenic peptide, can not be excluded particularly at later time points (6 hr) that could be occurring during fusion of endocytic compartments while processing the exogenous antigen. However, in this model, it is difficult to assess the relevance of MHC Class I that contain antigenic peptide in late endosomes for priming T cells. Our functional data indicate that peptide-loaded Kb molecules, can reach the cell surface within 6 hours, and are sufficient to induce T cell proliferation. While microscopy could somewhat visualize the peptide-loaded Kbs on the cell surface, this was not convincingly quantifiable and therefore the assessment of T cell activation complemented the previous observations and revealed the relevance of their presence on the cell surface within the time frame assessed. Also, our model including the 6 hour time frame, tends to validate the vacuolar pathway of antigen presentation since it allows little time for the antigen to access the cytosol, although as shown by microscopy and some minimal T cell activation in ΔY-DCs, this could occur even at early time points. Consequently, it is reasonable to assume that this pathway may account for the first wave of T cells that are generated following antigen exposure during the immune response. Importantly in the present study we can clearly demonstrate that this pathway is dependent on the conserved cytoplasmic tyrosine residue encoded by exon 6, confirming its role as part of a tyrosine-based endocytic-sorting motif (YXXA) that controls endocytic trafficking of surface MHC-I molecules [23]. It is thus likely that the previously reported cross-presentation deficiency in ΔY-derived DCs resulted from the inability of these tyrosine-mutated surface Kb molecules to traffic through endocytic loading compartments (ELCs) [21].Our data also demonstrates a distinct role for exon 7-encoded amino acids in controlling MHC-I intracellular trafficking in DCs. The existence of exon 7-deleted MHC-I isoforms that lack conserved serine phosphorylation sites has been reported in several species including mouse, arising in different cell types to varying degrees as a result of differential RNA splicing [24]. As has previously been reported for lymphoblastoid cell lines, exon 7-deleted (Δ7) MHC-I molecules in DCs show significantly delayed kinetics of internalization from the cell surface [25]. However, unlike ΔY, surface Δ7 molecules do possess the ability to traffic though both early and late endosomal compartments, albeit more slowly than KbWT molecules. Delayed cell surface internalization may actually provide an advantage by prolonging antigen presentation to augment CTL priming, a notion consistent with our previous observation that Δ7 mice consistently generated more vigorous antiviral CTL responses compared to KbWT mice [21].The distinct intracellular localizations of newly-formed Kb-OVA complexes observed in KbWT, Δ7, and ΔY-derived DCs support the emerging concept of multiple pathways of peptide loading of cross-presented antigens [26], [27]. In our study, the majority of KbWT-OVA complexes were found within early and late endosomal/lysosomal compartments of DCs after exposure to OVA. Like KbWT, the majority of Δ7-OVAp complexes were also found within early and late endosomes. By contrast, ΔY-OVAp complexes were comparatively less abundant, and those observed were almost exclusively in non-endolysosomal compartments, with some appearing in the Golgi and others colocalizing with the ER marker Grp78 (unpublished).Our data suggests that the vacuolar pathway of exogenous antigen loading plays a principal role in cross-presentation, having observed the majority of KbWT-OVA complexes within early and late endosomes/lysosomes of DCs following addition of soluble ovalbumin. However, the existence of Kb-OVA complexes within non-endolysosomal compartments, as seen in both Δ7 and ΔY DCs, suggests that a second compartment of exogenous antigen loading also exists. These observations could support the endosome-to-cytosol pathway, in which exogenous antigens are extruded into the cytosolic pool of protein antigens and are processed by proteosomes before being transported into the ER for loading onto nascent MHC-I molecules. However, co-localization analysis of Kb-OVA complexes with the ER marker Grp78 showed that only a minor fraction of Grp78-positive compartments contained such complexes, and these compartments also colocalized with surface-derived Kb molecules (unpublished data). This could be interpreted as evidence of “ergosomal” peptide loading, in which the relevant loading sites consist of mixed compartments containing markers of both ER and endolysosomes [28]. While it remains disputed whether phagosomal membranes are ER- or plasma membrane-derived [16], it is possible that cross presentation-competent, ergosomal organelles may be formed through another mechanism of vesicular fusion, following endocytosis. Further studies utilizing different ER specific markers will need to be performed before definitive conclusions can be reached.Collectively, our results support the model of DC MHC-I trafficking and exogenously-derived peptide loading depicted in Figure 7. In this model, KbWT, Δ7, and ΔY molecules initially appear in the ER as nascent MHC-I heavy chains that bind to β2-microglobulin and are loaded primarily with endogenously-derived peptides generated by the proteosome and transported by TAP. After secretory transport through the Golgi and to the cell surface, wild-type Kb molecules are rapidly and efficiently internalized and transported to ELCs, where peptide exchange and loading of peptides from exogenously-derived antigens generated by endolysosomal proteases and cathepsins may occur. These newly-loaded MHC-I complexes are then transported by an undefined mechanism back to the cell surface for cross-presentation to T cells. By contrast, Kb molecules lacking exon 7 are internalized and transported to ELCs, but significantly more slowly than wild-type molecules. Interestingly, this delayed kinetics does not seem to significantly reduce the level of Δ7 molecules residing within ELCs. This raises the possibility that MHC-I molecules might gain access to endolysosomal compartments by means other than internalization from the cell surface. The relative abundance of Δ7-OVA complexes in ELCs that are not colocalized with surface-derived Δ7 molecules (Figures 5A and 5B, pink color) supports this notion. This may indicate loading of newly-synthesized MHC-I molecules through the classical pathway and subsequent transport to endosomes, or alternatively may represent loading of nascent Kb molecules within ELCs. MHC-I molecules may conceivably traffic to ELCs directly from the ER to create ergosomal mix/fusion compartments, but this pathway is not currently well understood [29]. Another potential trafficking route may involve sorting of MHC-I molecules from the trans-Golgi into endolysosomes, a pathway that has been well-described for other transmembrane proteins [30]. It is tempting to speculate that the ΔY molecules we observed in Golgi compartments of DCs may be trapped there as a result of their inability to interact with the appropriate tyrosine motif-binding adaptin proteins to be sorted into clathrin-coated vesicles destined for ELCs [31].10.1371/journal.pone.0003247.g007Figure 7Model of dendritic cell MHC-I trafficking and cross-presentation.Schematic representation of trafficking routes for KbWT, Δ7, and ΔY molecules in dendritic cells, depicting proposed intracellular sites of antigen acquisition. In the direct presentation pathway (bottom), endogenously-synthesized proteins (green) are degraded by cytosolic proteosome complexes into antigenic peptides, which are transported by the transporter associated with antigen processing (TAP) into the endoplasmic reticulum (ER) for binding to nascent class Iα chain/β2-microglobulin dimers. KbWT, Δ7, and ΔY molecules initially loaded in the ER with endogenously-derived peptides (green) are then transported through the cis- and trans-Golgi (TGN) and to the cell surface via the secretory pathway. Alternatively, a subset of KbWT and Δ7, but not ΔY, molecules may be re-routed from the secretory pathway directly into endolysosomal compartments, although this pathway remains largely uncharacterized. In cross-presentation, exogenous protein antigens (orange, top) are internalized into endocytic vesicles, where they can be transported into the cytosol via ER-associated retrotranslocation (asterisk) and subsequently enter the direct presentation pathway. Exogenous antigens may also be transported into endolysosomes, where they are degraded by resident proteases and Cathepsin S into antigenic peptides (orange). Surface MHC-I molecules are constitutively internalized and transported through the endocytic pathway by a mechanism that requires the MHC-I cytoplasmic tyrosine (Y) residue. MHC-I molecules lacking Exon 7 (Ex7), despite abundant colocalization within endosomes and lysosomes, are significantly delayed in their endocytic transport. MHC-I transport through early endosomes and late endosomes/lysosomes seems to be required for acquisition and cross-presentation of exogenously-derived peptides, suggesting that such peptides are bound to recycling MHC-I molecules directly within endocytic loading compartments (ELCs).While the results presented here provide an important first step for delineating intracellular routes of MHC-I trafficking in DCs and defining the molecular mechanisms that contribute to cross-presentation, many questions remain unresolved. For example, it will be interesting to examine whether serine phosphorylation is the major mechanism that modulates MHC-I internalization and recycling, and whether tyrosine phosphorylation also plays a role [32]. DC maturation stimuli (ie. TLR ligands) that induce changes in MHC-I molecular trafficking and remodeling of intracellular compartments will also shed light on how specialized exogenous antigen loading can occur with maximum efficiency [33]. Finally, there are unresolved questions regarding the importance of antigen entry route on cross-presentation. Soluble ovalbumin has recently been shown to be dependent upon the mannose receptor for its uptake into DCs (13). It will be interesting to determine whether the MHC-I cytoplasmic domain plays a similar role in the cross-presentation of pinocytosed or phagocytosed antigens as well as the MHC Class I targeting in the organelle where loading takes place following distinct routes of uptake'. With growing evidence that cross-presentation is crucial for the generation of appropriate CD8+ immune responses, future work should focus on uncovering the relevant mechanisms and better characterizing the dynamic nature of cross-presentation compartments. Understanding these issues will ultimately lead to the development of more potent dendritic cell-based vaccines for cancer and other diseases, as well as novel strategies to target endogenous DCs in vivo to generate optimal CD8+ T-cell priming against defined antigens.Supporting InformationText S1(0.02 MB DOC)Click here for additional data file.Figure S1DC2.4 dendritic cells were incubated with 1 µM OVA257–264 or PBS and labeled sequentially with anti-H-2Kb-FITC followed by anti H-2Kb/OVA257–264 antibodies and vice versa. Flow cytometry was conducted to assess the H-2Kb and H-2Kb/OVA257–264 complexes. Data represents one experiment.(0.06 MB TIF)Click here for additional data file.\n\nREFERENCES:\n1. YewdellJWNorburyCCBenninkJR\n1999\nMechanisms of exogenous antigen presentation by MHC class I molecules in vitro and in vivo: implications for generating CD8+ T cell responses to infectious agents, tumors, transplants, and vaccines.\nAdv Immunol\n73\n1\n77\n10399005\n2. HeathWRCarboneFR\n2001\nCross-presentation, dendritic cells, tolerance and immunity.\nAnnu Rev Immunol\n19\n47\n64\n11244030\n3. CresswellPAckermanALGiodiniAPeaperDRWearschPA\n2005\nMechanisms of MHC class I-restricted antigen processing and cross-presentation.\nImmunol Rev\n207\n145\n157\n16181333\n4. ShenLRockKL\n2006\nPriming of T cells by exogenous antigen cross-presented on MHC class I molecules.\nCurr Opin Immunol\n18\n85\n91\n16326087\n5. PfeiferJDWickMJRobertsRLFindlayKNormarkSJ\n1993\nPhagocytic processing of bacterial antigens for class I MHC presentation to T cells.\nNature\n361\n359\n362\n7678924\n6. ShenLSigalLJBoesMRockKL\n2004\nImportant role of cathepsin S in generating peptides for TAP-independent MHC class I crosspresentation in vivo.\nImmunity\n21\n155\n165\n15308097\n7. BachmannMFHengartnerHZinkernagelRM\n1995\nT helper cell-independent neutralizing B cell response against vesicular stomatitis virus: role of antigen patterns in B cell induction?\nEur J Immunol\n25\n3445\n3451\n8566036\n8. LiuTChambersBDiehlADVan KaerLJondalM\n1997\nTAP peptide transporter-independent presentation of heat-killed Sendai virus antigen on MHC class I molecules by splenic antigen-presenting cells.\nJ Immunol\n159\n5364\n5371\n9548476\n9. Kovacsovics-BankowskiMRockKL\n1995\nA phagosome-to-cytosol pathway for exogenous antigens presented on MHC class I molecules.\nScience\n267\n243\n246\n7809629\n10. DelamarreLHolcombeHMellmanI\n2003\nPresentation of exogenous antigens on major histocompatibility complex (MHC) class I and MHC class II molecules is differentially regulated during dendritic cell maturation.\nJ Exp Med\n198\n111\n122\n12835477\n11. ImaiJHasegawaHMaruyaMKoyasuSYaharaI\n2005\nExogenous antigens are processed through the endoplasmic reticulum-associated degradation (ERAD) in cross-presentation by dendritic cells.\nInt Immunol\n17\n45\n53\n15546887\n12. AckermanALGiodiniACresswellP\n2006\nA role for the endoplasmic reticulum protein retrotranslocation machinery during crosspresentation by dendritic cells.\nImmunity\n25\n607\n617\n17027300\n13. BurgdorfSKautzABohnertVKnollePAKurtsC\n2007\nDistinct pathways of antigen uptake and intracellular routing in CD4 and CD8 T cell activation.\nScience\n316\n612\n616\n17463291\n14. HoudeMBertholetSGagnonEBrunetSGoyetteG\n2003\nPhagosomes are competent organelles for antigen cross-presentation.\nNature\n425\n402\n406\n14508490\n15. GuermonprezPSaveanuLKleijmeerMDavoustJVan EndertP\n2003\nER-phagosome fusion defines an MHC class I cross-presentation compartment in dendritic cells.\nNature\n425\n397\n402\n14508489\n16. TouretNParoutisPTerebiznikMHarrisonRETrombettaS\n2005\nQuantitative and dynamic assessment of the contribution of the ER to phagosome formation.\nCell\n123\n157\n170\n16213220\n17. MacAryPALindsayMScottMACraigJILuzioJP\n2001\nMobilization of MHC class I molecules from late endosomes to the cell surface following activation of CD34-derived human Langerhans cells.\nProc Natl Acad Sci U S A\n98\n3982\n3987\n11274420\n18. KleijmeerMJEscolaJMUytdeHaagFGJakobsonEGriffithJM\n2001\nAntigen loading of MHC class I molecules in the endocytic tract.\nTraffic\n2\n124\n137\n11247303\n19. DasguptaJDWatkinsSSlayterHYunisEJ\n1988\nReceptor-like nature of class I HLA: endocytosis via coated pits.\nJ Immunol\n141\n2577\n2580\n2902138\n20. ReidPAWattsC\n1990\nCycling of cell-surface MHC glycoproteins through primaquine-sensitive intracellular compartments.\nNature\n346\n655\n657\n2166918\n21. LizeeGBashaGTiongJJulienJPTianM\n2003\nControl of dendritic cell cross-presentation by the major histocompatibility complex class I cytoplasmic domain.\nNat Immunol\n4\n1065\n1073\n14566337\n22. GrommeMUytdehaagFGJanssenHCalafatJvan BinnendijkRS\n1999\nRecycling MHC class I molecules and endosomal peptide loading.\nProc Natl Acad Sci U S A\n96\n10326\n10331\n10468607\n23. LizeeGBashaGJefferiesWA\n2005\nTails of wonder: endocytic-sorting motifs key for exogenous antigen presentation.\nTrends Immunol\n26\n141\n149\n15745856\n24. McCluskeyJBoydLFMaloyWLColiganJEMarguliesDH\n1986\nAlternative processing of H-2Dd pre-mRNAs results in membrane expression of differentially phosphorylated protein products.\nEmbo J\n5\n2477\n2483\n3640710\n25. VegaMAStromingerJL\n1989\nConstitutive endocytosis of HLA class I antigens requires a specific portion of the intracytoplasmic tail that shares structural features with other endocytosed molecules.\nProc Natl Acad Sci U S A\n86\n2688\n2692\n2495533\n26. PalliserDGuillenEJuMEisenHN\n2005\nMultiple intracellular routes in the cross-presentation of a soluble protein by murine dendritic cells.\nJ Immunol\n174\n1879\n1887\n15699114\n27. RamirezMCSigalLJ\n2004\nThe multiple routes of MHC-I cross-presentation.\nTrends Microbiol\n12\n204\n207\n15120138\n28. GuermonprezPAmigorenaS\n2005\nPathways for antigen cross presentation.\nSpringer Semin Immunopathol\n26\n257\n271\n15592842\n29. SugitaMBrennerMB\n1995\nAssociation of the invariant chain with major histocompatibility complex class I molecules directs trafficking to endocytic compartments.\nJ Biol Chem\n270\n1443\n1448\n7836413\n30. BardFMalhotraV\n2006\nThe formation of TGN-to-plasma-membrane transport carriers.\nAnnu Rev Cell Dev Biol\n22\n439\n455\n16824007\n31. OwenDJEvansPR\n1998\nA structural explanation for the recognition of tyrosine-based endocytotic signals.\nScience\n282\n1327\n1332\n9812899\n32. SantosSGPowisSJArosaFA\n2004\nMisfolding of major histocompatibility complex class I molecules in activated T cells allows cis-interactions with receptors and signaling molecules and is associated with tyrosine phosphorylation.\nJ Biol Chem\n279\n53062\n53070\n15471856\n33. BlanderJMMedzhitovR\n2006\nToll-dependent selection of microbial antigens for presentation by dendritic cells.\nNature"
+}

batch_3/PMC2533015.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533015",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533015\nAUTHORS: Masae Miyatani, Hiroshi Kawano, Kei Masani, Yuko Gando, Kenta Yamamoto, Michiya Tanimoto, Taewoong Oh, Chiyoko Usui, Kiyoshi Sanada, Mitsuru Higuchi, Izumi Tabata, Motohiko Miyachi\n\nABSTRACT:\nBackgroundSince it is essential to maintain a high level of cardiorespiratory fitness to prevent life-style related disease, the Ministry of Health, Labour and Welfare of Japan in 2006 proposed to determine the maximal oxygen uptake (Vo2max: mL·kg-1·min-1) reference values to prevent life-style related diseases (LSRD). Since muscle mass is one of the determinant factors of Vo2max, it could be used as the reference parameter for preventing LSRD. The aim of this study was to determine and quantify the muscle mass required to maintain the Vo2max reference values in Japanese women.MethodsA total of 403 Japanese women aged 20–69 years were randomly allocated to either a validation or a cross-validation group. In the validation group, a multiple regression equation, which used a set of age and the percentage of muscle mass (%MM, percentage of appendicular lean soft tissue mass to body weight), as independent variables, was derived to estimate the Vo2max. After the equation was cross-validated, data from the two groups were pooled together to establish the final equation. The required %MM for each subject was recalculated by substituting the Vo2max reference values and her age in the final equation.ResultsThe mean value of required %MM was identified as (28.5 ± 0.35%). Thus, the present study proposed the required muscle mass (28.5% per body weight) in Japanese women to maintain the Vo2max reference values determined by the Japanese Ministry of Health Labour and Welfare.ConclusionThe estimated required %MM (28.5% per body weight) can be used as one of the reference parameters of fitness level in Japanese women.\n\nBODY:\nBackgroundPrevious epidemiologic and clinical evidence indicate that a poor cardiorespiratory fitness is a major risk factor for life-style related diseases (LSRD) such as obesity, hypertension, hypercholesterolaemia, arteriosclerosis and diabetes [1-4]. Moreover, low cardiorespiratory fitness has been found to be a predictor of cardiovascular disease (CVD) mortality, and all-cause mortality [5-8]. Thus, it is essential to maintain a high level of cardiorespiratory fitness to prevent LSRD.Cardiovascular fitness is usually evaluated as the maximal oxygen uptake per body mass (Vo2max, mL·kg-1·min-1). The Japanese Ministry of Health Labour and Welfare in 2006 proposed Vo2max reference values for each age group to prevent LSRD [9]. These Vo2max reference values were determined by the \"Committee for the Determination of the Recommended Exercise Allowance and Exercise Guide\" established in August 2005, and were referenced in the \"Exercise and Physical Activity Reference Quantity for Health Promotion 2006 (EPAR2006)\". Originally, the \"Recommended Quantity of Exercise for Health Promotion (1989)\" had been formulated to mainly target the prevention of coronary artery disease. With the passage of more than 15 years following the establishment of this standard, the morbidity pattern of people has worsened and LSRD have increased in prevalence. In order to face this situation, the EPAR2006 was made based on the latest scientific evidence, and was designed to maintain and promote the health of people and prevent LSRD by improving their capacity for physical activity and exercise. These Vo2max reference values proposed in the EPAR2006 were determined by experts through the systematic review of literature regarding the relationship between Vo2max and LSRD such as obesity, hypertension, hypercholesterolemia, diabetes, cerebrovascular disease, CVD mortality and all-cause mortality.It is well known that Vo2max decreases with age [10-20]. It has been suggested that the age-related decline in Vo2max is a consequence of attenuation of central and peripheral functions such as stroke volume, heart rate max (HRmax), peripheral O2 extraction, and lean body mass (LBM) or muscle mass [19,21-25]. Among these determinants, reductions in HRmax and LBM or muscle mass have been suggested to be primary factors [26,27]. While many studies on cardiovascular fitness have focused on cardiac measurements, it should be emphasized that muscle mass is one of the critical determinants of Vo2max [13,14,19,24,26,28-30] since the amount of tissue available to extract oxygen during maximal exercise, i.e., muscle, can directly contribute to the value of Vo2max. For example, Sanada et al. reported the MRI-measured lower body skeletal muscle mass was closely associated to the absolute Vo2max during running [28,30]. Additionally, the age-related decrement in Vo2max can be related to the age-associated muscle loss [24,19]. Further, it is important to notice that LBM or muscle mass can be maintained to some degree by exercise training, while such training cannot prevent age-related declines in HRmax, [26,27].Therefore, we hypothesized that a certain level of muscle mass required to maintain sufficient cardiovascular fitness is present and that it could be a limiting factor of age-related Vo2max attenuation. Based on this hypothesis, it is advantageous to Japanese women's health to propose such muscle mass required to maintain sufficient Vo2max. Thus, the purpose of this study was to determine a required value of muscle mass to maintain the Vo2max reference value determined by the Japanese Ministry of Health Labour and Welfare in 2006 (Ministry of Health, Labour and Welfare of Japan 2006).MethodsSubjectsA group of 403 Japanese women aged 20 to 69 years were randomly allocated to either a validation group (V-group, n = 201) or a cross-validation group (CV-group, n = 202). The subjects were recruited from the community around the National Institute of Health and Nutrition. All subjects were active and free of overt CVD assessed using a medical history questionnaire. All assessments were conducted at the National Institute of Health and Nutrition between February 2004 and October 2006. The study was approved by the Ethics Committee of the National Institute of Health and Nutrition, and written consent was obtained from all participants.Percentage of muscle massThe lean soft tissue mass of legs and arms were measured with a whole-body Dual Energy X-ray Absorptiomettry (DXA) scanner (Hologic QDR-4500, Hologic INC., Waltham, MA, USA). The body regions were delineated according to specific anatomical landmarks using manual DXA analysis software (version11.2.3). The appendicular lean soft tissue mass was calculated as a sum of the lean soft tissue mass of the legs and the arms. The lean soft tissue mass of extremities assessed using DXA was assumed to represent appendicular skeletal muscle mass along with a small and relatively constant amount of skin and underlying connective tissues. The percentage of muscle mass (%MM) was calculated as follows;%MM (%) = (Appendicular lean soft tissue mass)/Body weight × 100.Vo2maxWe assessed peak oxygen uptake (Vo2peak: mL·kg-1·min-1) instead of Vo2max as an index of cardiorespiratory fitness, which is defined as the highest level of oxygen uptake that is determined by the protocol of a graded exercise load. The Vo2peak was measured using the incremental cycle exercise. An initial work intensity of 30 W or 60 W was selected for each patient based on the patient's fitness level. The work intensity was increased thereafter by a step of 15 W/min, until the subject was not able to maintain the required pedaling frequency of 60 rpm. The heart rate and rating of perceived exertion (RPE) were monitored throughout the exercise. The O2 consumption and the minute ventilation were monitored during each 1-min exercise stage (two 30 sec samplings for each stage), after RPE reached 18. The expired air was collected using Douglas bags. Expired O2 and CO2 gas concentrations were measured using a mass spectrometer (ARCO-1000A, ARCO SYSTEM, Chiba, Japan), and gas volume was measured using a dry gas meter (DC-5C Shinagawa Seiki, Tokyo, Japan). If the subject became exhausted and was not able to keep the pedaling frequency at 60 rpm, it was decided that the maximum effort had been achieved and the test was terminated. The highest value of Vo2 during the exercise test was designated as Vo2peak. Note that the oxygen uptake obtained in this procedure is referred to as Vo2peak, to discriminate this from Vo2max in the strict definition. However, we equate the obtained Vo2peak to Vo2max in the present study since the Vo2max reference value was determined using both Vo2max and Vo2peak as mentioned in the next section.Vo2maxk reference valuesThe Japanese Ministry of Health Labour and Welfare proposed Vo2max reference values to prevent life-style related illness for women [9]. The Vo2max reference values are provided for each age group. The procedure to determine Vo2max reference values was described in the EPARQ2006 [9]. In brief, these Vo2max reference values were determined by experts through a systematic review of literature. The target age was 6 years and older. The target LSRD were obesity, hypertension, hyperlipemia, diabetes mellitus, cerebrovascular disorders, death due to circulatory diseases, osteoporosis, ADL and total mortality. By means of this systematic review, the threshold values of the Vo2max or Vo2peak at which the morbidity of LSRD statistically increases in each age group were collected from the literature. The average values of these threshold values for each age group were then calculated and designated as the Vo2max reference values for preventing LSRD. The identified Vo2max reference values (mL·kg-1·min-1) were 33 (20–29 yr), 32 (30–39 yr), 31 (40–49 yr), 29 (50–59 yr), and 28 (60–69 yr).AnalysesFirst, a single regression analysis was used to test the correlation between age and Vo2max, and between %MM and Vo2max in V-group. Then, a multiple regression analysis was performed using Vo2max as a dependent variable, and age and %MM as the independent variables. This analysis was based on the hypothesis that Vo2max can be accounted for by age and %MM. In this hypothesis, we assumed that the age factor included Vo2max determinant factors related to aging except for muscle mass, such as HRmax, maximal stroke volume, and peripheral O2 extraction [21-23,25,27]. The validity of the prediction by the obtained regression equation was tested by applying the obtained regression equation to the CV-group. After the equation was cross-validated, the data from the two groups were pooled together to obtain the final prediction equation and in the subsequent analysis.The purpose of the final prediction equation was to obtain the required %MM to maintain the reference Vo2max value in each age group. Thus, the required %MM for each subject was recalculated by assigning the Vo2max reference values and age in the final prediction equation. If the difference of the required %MM among the age groups was very small, the mean value of the required %MM was calculated to be used in the following analysis. To test the validity of the required %MM, the correlation between the sufficiency of Vo2max, i.e., individual's Vo2max as the percentage of the Vo2max reference values (% Vo2max reference values), and the sufficiency of the required %MM, i.e., individual's %MM as the percentage of the required %MM (%required-%MM), were tested.All data are reported as means ± standard deviations (SD). P < 0.05 was used as a level of significance for all comparisons.ResultsPhysiological characteristicsThe physiological characteristics for each group are shown in Table 1. There were no significant physiological differences between V-group and CV-group.Table 1Characteristics of validation and cross-validation groupV-groupCV-groupn202201Age (yr)41.4 ± 16.741.6 ± 16.9Height (cm)158.5 ± 6.4157.9 ± 6.1Body weight (kg)54.4 ± 7.453.9 ± 7.3Body mass index (kg/m2)21.6 ± 2.721.7 ± 2.9Appendicular muscle mass (kg)16.4 ± 2.416.1 ± 2.3% MMI (%)30.3 ± 3.230.0 ± 3.4Vo2max (ml·kg-1·min-1)33.5 ± 7.932.7 ± 7.7mean ± SD, V-group, Validation group; CV-group, Cross-validation group; %MM, percentage of muscle massRelationship between age and Vo2max in V-groupVo2max in V-group was from 16.4 to 56.9 ml.kg-1min-1 (mean 33.5 ± 7.9) (Table 1). As expected, a strong negative linear correlation was found between Vo2max and age (Figure 1). The decrement was 2.58 ml.kg-1min-1 per decade. The Vo2max reference values for each age group in the EPAR2006 were superimposed in Figure 1. With increasing age, the proportion of subjects with Vo2max values below the reference Vo2max values increased.Figure 1The relationship between age and Vo2max in the V-group. The Vo2max reference values by the Japanease Ministry of Health Labour and Welfare were shown for reference.Relationship between Vo2max and %MM in V-group%MM in V-group was from 18.7 to 37.3% (mean 30.3 ± 3.2%) (Table 1). There was also a strong correlation between Vo2max and %MM, while the correlation was positive (Figure 2).Figure 2Relationship between percentage of muscle mass (%MM) and Vo2max in the V-group.Multiple-regression analysis in V-groupMultiple regression analysis in V-group revealed that age (R2 = 0.286) and %MM (R2 = 0.540) were significant (p < 0.0001) contributors to the prediction of the measured Vo2max. The multiple regression equation obtained in the V-group was the following: Vo2max = -0.135 × Age + 1.315 × %MM -0.799. In this equation, R2 and SEE were 0.522 and 5.4 mL·kg-1·min-1, respectively.Cross-validation of the multiple regression equationThe multiple regression equation derived from the V-group was used to predict Vo2max in the CV-group. Figure 3 shows the residual plot. There was not statistically significant correlation between the predicted Vo2max and residual error (p > 0.05). Thus, the residual plot indicates that there was no bias in the prediction of Vo2max of the CV-group using the multiple regression obtained in the V-group.Figure 3Relationship between estimated Vo2max by the multiple regression equation and the residuals for the CV-group.Final prediction equationData from the two groups were pooled to generate the final equations:(1)Vo2max = -0.131 × Age + 1.344 × %MM - 2.035.In the final equation, analysis revealed that age (R2 = 0.282) and %MM (R2 = 0.570) were significant (p < 0.0001) independent contributors to the prediction of the measured Vo2max. Figure 4 shows the residual plot of the multiple-regression. There was no statistically significant correlation between the predicted Vo2max and residual error (p > 0.05). Thus, the residual plot indicates that there was no bias in the prediction of Vo2max.Figure 4Relationship between estimated Vo2max by the multiple regression equation and the residuals for both the V-group and the CV-group.Estimation of the required %MMThe equation (1) was rearranged to predict required %MM as follow;(2)%MM = (0.131 × Age + 2.035 + Vo2max)/1.344.The required %MM was calculated by assigning the Vo2max reference values, and age in the equation (2). The calculated required %MM was shown in Table 2. The mean value and standard deviation of required %MM was 28.5 ± 0.35%. Figure 5 shows the relationship between the measured %MM and age with the required %MM superimposed on the plot. The older people tended to have a %MM lower than the required. With increasing age, the proportion of subjects with %MM below the required %MM increased.Table 2Required %MM for Vo2max reference values of each age groupAge group20 Y30 Y40 Y50 Y60 YTotaln14348557384403Required MMI (%)28.3 ± 0.2628.6 ± 0.2928.9 ± 0.2728.4 ± 0.2528.6 ± 0.3028.5 ± 0.35Mean ± SD; %MMI, percentage of muscle massFigure 5The relationship between age and the percentage of muscle mass (%MM) in the V-group and the CV-group. Required %MM is shown for reference.The validity of the required %MMFigure 6 shows the relationship between %Vo2max reference values and %required-%MM. The %Vo2max reference values positively correlated with %required-%MM (r = 0.651, p < 0.05).Figure 6The relationship between the sufficiency of Vo2max (%Vo2max reference values) and the sufficiency of the required %MM (% required %MM) in both the V-group and the CV-group. Solid line: regression line, dashed line: lines of 100% of Required %MM and 100% of Vo2max reference values.DiscussionThe primary finding of the present study is that appendicular muscle mass of 28.5% of body weight is needed to maintain the Vo2max reference values determined by the Japanese Ministry of Health Labour and Welfare in Japanese women. By use of the multiple-regression analysis, the regression equation of Vo2max from age and %MM was obtained in the V-group at first. Then the validity of the regression equation was confirmed in the CV-group (Figure 3). The required %MM to maintain the Vo2max reference values was obtained using the final regression equation using the data of V- and CV-groups (equation (2)) and the Vo2max reference values for each age group (Table 2). There was strong correlation between percentages of the required %MM and Vo2max reference values (Figure 6).Required muscle massWe propose the required %MM in Japanese women as a reference value of muscle mass for the usage of maintaining the reference value of Vo2max proposed by the Ministry of Health Labour and Welfare of Japan. Interestingly, the calculated required %MM was not different among age groups (Table 2). Thus, we proposed the averaged required muscle mass (28.5%) as the general value for all age groups. A large portion of the subjects (68%) satisfied the required muscle mass, while with increasing age, the proportion of subjects with %MM below the required %MM increased (Figure 5). This tendency was similar to Vo2max, i.e., with increasing age, the proportion of subjects with Vo2max values below the reference Vo2max values increased (Figure 1). Additionally, there was strong positive relation between percentages of Vo2max reference values and required %MM (Figure 6). The results indicate that subjects with total muscle mass lower than 100% of the required %MM also tended to have lower Vo2max when compared to levels of Vo2max reference values. Thus, our result suggests that one of the reasons for insufficient Vo2max may be insufficient %MM. Women who have %MM less than the required %MM are encouraged to increase their %MM above the required %MM to achieve the Vo2max reference values. The required %MM can be used as an additional parameter for preventing LSRD together with the Vo2max reference values. The required %MM obtained in this study is practical and appropriate for most Japanese women, because it is slightly less than the average %MM of the total number of subjects. Thus, the value is an achievable goal for most of Japanese women. Although strength training is not typically included in exercise programs targeting prevention of the age-related decline in Vo2max or to increase Vo2max, it would be advisable to recommend some form of strength training as well as aerobic training especially for individuals who do not achieve the required %MM.Several prior studies demonstrated the significance of fat free mass, muscle mass, and/or muscle function to morbidity and mortality, although there are few researches targeting women [31-33]. The Japanese Ministry of Health Labour and Welfare also has admitted the importance of muscle mass and muscle function to prevent LSRD and/or mortality in EPAR2006. However practical target values have not been offered in the statement due to the lack of evidences compared to Vo2max. In this present study we determined the target value of muscle mass through the Vo2max reference values, which already has strong evidences. Although we have not confirmed the direct relation between muscle mass and LSRD morbidity and/or mortality, we believe Japanese women could aim to achieve the required %MM as one of targets for their health. Whether an increase of skeletal muscle mass would result in an improvement of exercise capacity and or reduce morbidity and mortality needs to be confirmed by future studies.It should note that some individuals may have a large muscle mass, yet be at a high mortality risk. For example, it is well known that central obesity is one of risk factor of LSRD morbidity. Thus, it is important to remember that muscle mass is not the only important parameter but also, other risk factor should be monitored and considered together.Prediction of Vo2max from age and muscle massThe residuals of the multiple regression might be due to the approximation that all age-related determinant factors were included in age in the multiple regression. In the present model, we hypothesized that determinants such as HRmax, maximal stroke volume, and peripheral O2 extraction were age-related, and therefore their effects were included in the factor of age. It was suggested that HRmax [14,22,26,29,34-39] and peripheral O2 extraction [21,34] do decline with age, and are not influenced by exercise training. However, although maximum stroke volume was also suggested to decline with age in sedentary individuals [23], it was suggested that age-related decline of maximum stroke volume was prevented by exercise [21,34]. Thus, the simplification must be the error factor, and it is likely in future to improve the multiple regression equations using these age-related Vo2max determinants, and to improve the estimation of the required MMI.We studied only a statistical relationship between Vo2max and muscle mass. Therefore, the results do not necessarily suggest a cause-effect relationship. It is possible that muscle mass and Vo2max are physiologically unrelated but indirectly correlated, i.e., people with a high Vo2max may be more physically active and perform activities that increase muscle mass. However, muscle mass is highly likely physiologically important determinant of Vo2max because the amount of tissue available to extract oxygen during maximal exercise directly contribute to the value of Vo2max.Study limitationsThe current study has limitations that require caution when interpreting and generalizing the findings reported herein. This study included only the cross-sectional design, and it did not investigate the relationship between the required %MM and the morbidity of LSRD or mortality by using a prospective design. Thus, it has not been clarified how the required %MM reflects these risks in this present study. Further investigation is required to validate the required %MM through a prospective study with the morbidity and/or mortality as an endpoint. Additionally, the potential difference between methods using %MM or absolute muscle mass (kg) as the indicator of health should be also investigated. Another limitation of this study is the results of this study are applicable to only Japanese women. The decided %MM in this study may not be able be applicable to men and/or other racial group since they may have different characteristics of the relationship between muscle mass and Vo2max.ConclusionIn conclusion, the present study proposed the required muscle mass (28.5% per body weight) in Japanese women to maintain the Vo2max reference values determined by the Japanese Ministry of Health Labour and Welfare. This required muscle mass can be used as one of the reference parameters of fitness level in Japanese women.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsMM performed analysis and data interpretation as well as drafted and revised the manuscript. KM participated in the conception of this study, interpretation of the analysis and critically reviewed this manuscript, and provided comment as Statistical expertise. HK, YG, KY, MT, TO, CU and SK performed data analysis and interpretation, and provided comment and review of the manuscript. MH and IT designed the project, assisted with data interpretation and provided comment and revisions for the manuscript. MM designed the project, participated in the conception of this study, interpretation of the analysis and critically reviewed this manuscript. All authors read and give final approval of the final manuscript for publication.Pre-publication historyThe pre-publication history for this paper can be accessed here:\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2533020.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533020",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533020\nAUTHORS: María Pilar  García Guerreiro, Blanca E Chávez-Sandoval, Joan Balanyà, Lluís Serra, Antonio Fontdevila\n\nABSTRACT:\nBackgroundTransposable elements (TEs) constitute a substantial amount of all eukaryotic genomes. They induce an important proportion of deleterious mutations by insertion into genes or gene regulatory regions. However, their mutational capabilities are not always adverse but can contribute to the genetic diversity and evolution of organisms. Knowledge of their distribution and activity in the genomes of populations under different environmental and demographic regimes, is important to understand their role in species evolution. In this work we study the chromosomal distribution of two TEs, gypsy and bilbo, in original and colonizing populations of Drosophila subobscura to reveal the putative effect of colonization on their insertion profile.ResultsChromosomal frequency distribution of two TEs in one original and three colonizing populations of D. subobscura, is different. Whereas the original population shows a low insertion frequency in most TE sites, colonizing populations have a mixture of high (frequency ≥ 10%) and low insertion sites for both TEs. Most highly occupied sites are coincident among colonizing populations and some of them are correlated to chromosomal arrangements. Comparisons of TE copy number between the X chromosome and autosomes show that gypsy occupancy seems to be controlled by negative selection, but bilbo one does not.ConclusionThese results are in accordance that TEs in Drosophila subobscura colonizing populations are submitted to a founder effect followed by genetic drift as a consequence of colonization. This would explain the high insertion frequencies of bilbo and gypsy in coincident sites of colonizing populations. High occupancy sites would represent insertion events prior to colonization. Sites of low frequency would be insertions that occurred after colonization and/or copies from the original population whose frequency is decreasing in colonizing populations. This work is a pioneer attempt to explain the chromosomal distribution of TEs in a colonizing species with high inversion polymorphism to reveal the putative effect of arrangements in TE insertion profiles. In general no associations between arrangements and TE have been found, except in a few cases where the association is very strong. Alternatively, founder drift effects, seem to play a leading role in TE genome distribution in colonizing populations.\n\nBODY:\nBackgroundTEs are widely distributed in eukaryotes, representing 50% of the human genome [1], 15% of the Drosophila genome, and up to 70% in Zea mays [2]. Because of their capacity of transposition they are able to invade the genome and promote insertional mutations and chromosomal rearrangements. Recurrent mobility allows them to persist in spite of their harmful effects in the host [3]. Most of the proposed models in population dynamic studies [4-8] suggest that TEs are able to invade the genome if their transposition rate is enough to balance out opposing forces as excision and selection against deleterious insertions and chromosomal arrangements. Yet, these models, often too general, do not consider that each element behaves depending on both its own characteristics and the history of the population to which it belongs. This challenge to standard reasoning is most relevant in colonizing populations [9]. Several authors have suggested that bursts of transposition could be induced in colonization by the foreign, often stressful, environment faced by the founders of colonizing populations [10,11]. Moreover, colonizing populations are subjected to well documented founder, drift effects [12]. Both processes generate population instabilities that may incorporate new variables to the interpretation of TE occupancy profiles in colonizing populations. These considerations qualify the study of TEs in colonization as of prime interest to understanding their invasive dynamics and putative evolutionary role in populations.Colonization effects on TEs were studied in Drosophila species [9,11,13] showing that this process plays an important role in the TE chromosomal distribution. In particular studies in colonizing populations of D. buzzatii showed a TE bimodal distribution with sites either highly occupied, in a few cases, or showing low insertion occupancy, in most cases. Molecular studies of TE copies from high and low occupied sites [14] strongly indicated that the most reliable explanation of the observed bimodal distribution is that a founder effect followed by genetic drift occurred during the colonization process. These results notwithstanding, valid for D. buzzatii, cannot be generalized to other colonizing Drosophila species, with different genomic characteristics, and subjected to different environmental pressures.D. subobscura, a Paleartic species belonging to the obscura group [15] and characterized by a rich inversion polymorphism [16], has colonized North and South America almost 30 years ago [17,18]. It was found for the first time in Puerto Montt (Chile) in 1978 [19] and later near Port Townsend in Washington (USA) in 1982 [20]. Thereafter this species showed a rapid spread and adaptation to the new colonized environment in form of latitudinal clines for chromosomal polymorphism and body size that paralleled the Paleartic clines [17,18,21]. Main after-colonization population effects were the presence of allelic lethal genes in different populations [22], the low genetic variability of mtDNA [23,24] and the reduction of microsatellite allele numbers [25] compared to original founder populations. These are expected outcomes of the founder drift effect of colonization. However nothing is known of the impact of colonization on the TE chromosomal distribution in this species.Here we present the study of the distribution of two TEs, gypsy and bilbo, in original and colonizing populations of D. subobscura. Results show that TE frequency distribution differ between original and colonizing populations in a way that colonization, chromosomal inversion polymorphism and particular characteristics inherent to each element can provide a sufficient likely explanation. In this paper we particularly emphasize the importance of population structure and history to explain TE distribution in natural populations.ResultsChromosomal distribution of bilbo and gypsyWe analyzed the distribution of bilbo and gypsy in polytene chromosomes of D. subobscura. Fig. 1 shows two examples of chromosomal distribution: bilbo in chromosome O and gypsy in chromosome U. A different distribution pattern is observed, in general, when we compare colonizing and original populations. Colonizing populations present insertion frequencies of bilbo and gypsy higher than those of the original population. In general the same distribution pattern is observed for the rest of chromosomes. Ten sites (7A, 16A, 20A, 45C, 58D, 74D 82A, 83C, 85A, 89C) show a bilbo insertion polymorphism greater than 32% in at least one colonizing population. Gypsy insertion frequencies are lower than those of bilbo with an occupancy of more than 10% in eight chromosomal sites (39D, 41C, 43D, 49D, 52D, 63C, 71B, 74D). Differences in occupancy profiles between original and colonizing populations are represented in Table 1 that shows the distribution of the number of times that each site is occupied in the studied sample. Thus, in bilbo the occupancy frequency ranges from 1 to 51 times in colonizing populations and only from 1 to 19 in the original population. Although gypsy shows a low occupancy profile compared to bilbo (colonizing populations range: 1–15; original population range: 1–5), the occupancy rate of both TEs in colonizing populations is greater than in the original population. The highest bilbo insertion frequencies are observed, in decreasing order, in Bellingham, Maipú and Davis. In the original population of Bordils, the highest insertion frequency corresponds to one site observed 16 times.Table 1Occupancy profiles of euchromatic sites in original and colonizing populationsTEPopulationsOccupancy profiles1234567891011121314151617–51bilboColonizingDA76941702215110107BE353242121200412210MA884742252323212112OriginalBO321921116251201010311gypsyColonizingDA844012220010110BE1151222100000001MA9106321100000200OriginalBO10125210000000000Population origin: Davis (DA), Bellingham (BE), Maipú (MA), Bordils (BO).Occupancy profile: number of times that each site is occupied in populations.Figure 1Distribution of bilbo and gypsy in chromosomes O and U, respectively, from colonizing (DA: Davis, BE: Bellingham, MA: Maipú) and original populations (BO: Bordils) of D. subobscura. Number of haploid genomes analyzed are given in parenthesis.Table 2 lists the means and variances of copy number for bilbo and gypsy per chromosome and haploid genome. The mean copy number of both TEs for the whole genome (HG) is always higher in colonizer populations than in the original one. The bilbo mean copy number differs greatly among chromosomes ranging from 2.58 copies in chromosome O from Bellingham to 0.55 in chromosome J from Davis. In fact chromosome J hosts the lowest number of bilbo in all populations. A different scenario is found for gypsy in which the A (X) chromosome contains the lowest number of insertions, in colonizer and original populations alike. However, among the autosomes J is the least occupied in all populations. Deviation from Poisson distribution was tested by chi-square goodness of fit tests (for details, see additional files 1 and 2) pooling adjacent classes with low expected numbers. In colonizing populations bilbo distribution in each chromosome fits a Poisson distribution. Gypsy deviates from a Poisson distribution in E chromosome from Davis and Bellingham and in U chromosome from Maipú. When the whole genome is considered both TEs follow a Poisson distribution in the original population and deviate in all colonizing ones, except for bilbo in Bellingham and gypsy in Maipú. For this element the general trend in colonizing populations is a lower than expected number of genomes with a single copy and an excess of genomes with three or more copies (see Table 1). An alternative test was performed using dispersion coefficients (DC), which measure the ratio between the variance (Vn) and the mean (m) (DC = Vn/m, see table 2). DC of 1 indicates that TE distribution is Poisson, and DC > 1 or DC < 1 indicates contagious or repulsive distributions, respectively. When the haploid genome is considered, there is a general tendency towards DCs > 1 for both elements in all populations except for gypsy in Maipú (these results are due to the greater effect of some chromosomes in the final result of the test).Table 2Tests of the Poisson distribution of bilbo and gypsy per chromosome and haploid genome.PopulationsDA (76)BE (88)TECh.mVnDCχ2dfmVnDCχ2dfA0.840.720.860.6121.070.770.732.682J0.550.490.890.1110.760.670.881.612U1.010.870.865.7431.111.000.901.463E1.381.731.254.7040.820.841.030.693O1.711.670.983.4752.581.800.707.064bilboHG5.509.401.71**32.56**116.347.791.2318.6111MA (81)BO (81)mVnDCχ2dfmVnDCχ2dfA1.751.690.961.5451.060.710.678.112J0.910.931.020.2530.750.811.0813.65*4U1.541.280.834.1540.840.690.821.142E1.801.660.921.7041.091.701.57**29.79**5O1.791.991.116.5350.961.281.349.583HG7.8019.512.50**254.34**184.705.761.2211.8810DA (70)BE (84)mVnDCχ2dfmVnDCχ2dfA00---0.020.021.00--J00---0.020.021.00--U0.981.261.285.5330.430.491.141.763E0.580.971.66**44.10**40.330.561.69**38.103O0.040.040.97--0.090.111.171.381gypsyHG1.612.441.51*14.8950.901.601.77**33.18**4MA (80)BO (80)mVnDCχ2dfmVnDCχ2dfA0.120.161.294.4610.020.021.00--J0.140.120.87--0.100.121.161.261U0.420.270.648.49*10.110.100.90--E0.390.320.821.4310.300.341.132.271O0.270.281.010.1210.240.230.990.001HG1.351.270.9411.4840.770.881.143.442TE: Transposable elements; Ch: chromosome; HG: haploid genome; m: mean copy number; Vn: variance of copy number; Numbers of haploid genomes are in parenthesis; DC: dispersion coefficient (Vn/m); *P < 0.05; **P < 0.01. Bonferroni's correction was applied. See Table 1 for population originBecause in some cases TE sites seem to be distributed in a contagious way (DC > 1), linkage disequilibrium was computed for each pair of sites by way of 2 × 2 contingency tables [26]. Linkage disequilibrium between TE sites could be responsible of the non-random distribution detected in some cases. The observed distribution of correlation coefficients between all paired sites was compared to the expected distribution in absence of linkage disequilibrium using Fisher's hypergeometric formula [27]. Figure 2 depicts, as an example, the correlation coefficient distributions (pooled in intervals of 0.1) of bilbo in chromome E from Bordils and in chromosome O from Maipú; and of gypsy in chromosome E from Davis and Bellingham. Tests were significant in most cases where a deviation from Poisson distribution was observed. Moreover we also found significant results in some cases where departures from Poisson distribution were not detected (e.g bilbo on chromosome O of Maipú). The general trend is a defect of class (-0.09–0.00) and an excess of some positive correlation classes. This indicates that some sites tend to stay together, as indicated by a DC > 1. This tendency was observed in all cases where deviations from Poisson distribution were observed, except for gypsy in chromosome U from Maipú where there is an overabundance in class (-0.09–0.00) and the DC is lower than 1.Figure 2Observed and expected frequency distributions of correlation coefficients between all pairs of sites in natural populations: A) bilbo in chromosome E from Bordils and O from Maipú. B)gypsy in chromosome E from Davis and Bellingham.Copy number comparisons among chromosomesMontgomery et al [28] proposed that selection against TE insertions would lead to a lower number of TE copies in chromosome X than in autosomes due to the stronger deleterious effect of recessive insertional mutations in the X chromosome of hemizygous males. In order to test this hypothesis we compared the copy number in the A (X) chromosome with that in autosomes. To estimate the expected number of insertions we multiply the relative proportion of chromatin of each chromosome by the number of total insertions in the population. The relative proportion of chromatin is that reported by Stumm-Zollinger and Goldschmidt [29] corrected by eliminating the dot chromosome, not included in our analyses. If TEs are randomly distributed, we expect a TE copy number per chromosome proportional to the amount of chromatin.Observed and expected proportions were compared by a G test [30] among all chromosomes (Ga), between the A (X) chromosome and autosomes (Gb), and among autosomes (Gc), as indicated in Table 3. Gb values were significant for gypsy in all populations, and for bilbo only in Maipú and Bordils. Because some differences may be due to high insertion sites, additional analyses were done after eliminating these sites. After elimination the significance was maintained for gypsy in all populations except in Maipú, and removed for bilbo. In general gypsy shows a low copy number in the A (X) chromosome compared to autosomes. However, this is not the rule for bilbo where Maipú and Bordils show a high copy number in A (X). Interestingly, those populations that display gypsy copy number differences between A (X) and autosomes, show also significant differences among autosomes (Gc), specially in colonizer populations where chromosomes E and O show a higher copy number than expected.Table 3Comparison of the proportion of gypsy and bilbo sites among chromosomes, autosomes and between chromosome A and autosomesTEgypsybilboCh.P. chromatDABEMABODABEMABOA(X)0.160.000.030.090.030.150.170.220.23J0.200.010.030.100.130.100.120.120.16U0.190.610.480.310.140.180.170.200.18E0.200.360.370.290.390.250.130.230.23O0.250.030.100.200.310.310.400.230.20DfGa4197.23** (115.82**)71.18** (61.31**)22.26** (5.08)21.40** (21.40**)36.98** (0.08)87.16** (16.98**)41.82** (4.80)15.67 (8.66)Gb140.48** (26.15**)15.24** (17.20**)4.63* (1.14)11.06**0.36 (4.8)0.08 (3.47)15.56** (2.9)9.69** (0.52)Gc3156.75** (89.66**)55.93** (44.11**)17.63** 3.9410.34*36.61** (3.47)87.08** (13.50*)26.26** (1.87)5.97 (8.10*)GaTotal12290.67**(182.21**)165.96**(30.06**)Pooled4230.15*(110.30**)102.01**(1.8)H860.52**(71.90**)63.95**(28.23**)GbTotal360.36**(44.50**)16.00**(11.20**)Pooled144.97**(23.78**)5.77*(1.09)H215.38**(20.71**)10.24*(10.10*)GcTotal9230.32**(137.71**)149.95**(18.85)Pooled3185.18**(86.52**)96.25**(0.72)H645.13**(51.19**)53.71**(18.12*)P. chromat: Proportion of chromatin. Ga: Comparison of the proportion of TEs among chromosomes. Gb: Comparison of the proportion of TEs between chromosome A (X) and autosomes. Gc: Comparison of the proportion of TEs among autosomes; H: Heterogeneity test between colonizing populations; Df: Degrees of freedom; Pooled: Only colonizing populations; *P < 0.05; **P < 0.001. Bonferroni's correction was applied. Test values excluding high insertion frequency sites are in parenthesis.In general, copy number tend to be higher for bilbo in chromosome O and for gypsy in chromosome U in all colonizing populations, (except bilbo in Maipú), whereas in the original population the E chromosome hosts the highest proportion of gypsy and bilbo. In order to determine if chromosomal differences have the same tendency in colonizing populations, heterogeneity (H) tests were performed for comparisons among chromosomes, between A (X) and autosomes and among autosomes. Table 3 shows that all cases were heterogeneous for both TEs. However when Maipú is excluded from the analyses and high insertion frequency sites are eliminated, Bellingham and Davis become homogeneous for bilbo (data not shown).Correlation studies between high frequency sites and chromosomal arrangementsAll five pairs of acrocentric chromosomes of D. subobscura are polymorphic for inversions. Frequencies of chromosomal arrangements show clinal variation correlated with latitude in Paleartic populations [31,32] and clines that follow the same latitudinal gradient evolved in recent colonizing populations in both hemispheres of the Americas [17,18]. These parallel observations across continents provided a natural experiment that supports the adaptative role of the chromosomal inversion polymorphism.Frequencies of chromosomal arrangements in the analyzed populations of this work are summarized in Table 4. Each arrangement is conventionally designed by the letter of the chromosome in which it occurs, followed by a combination of digits that identify the set of inversions included in it [16]. Arrangement frequencies are of the same order of magnitude as those previously reported, including the North-South latitudinal variation of most arrangements [21,31,32]. However it is interesting to note that Maipú presents a higher OSt frequency than expected according to its latitude.Table 4Frequencies of chromosomal arrangements in natural populationsArrangementsPopulationsDABEMABOAst0.570.670.550.53A1---0.17A20.430.330.450.30Jst0.410.450.260.40J10.590.550.740.60Ust0.320.420.500.17U1---0.03U1+20.390.400.200.72U1+2+80.290.180.300.07U1+2+3---0.01Est0.540.740.590.53E8---0.01E1+20.070.05-0.30E1+2+90.120.010.080.03E1+2+9+120.050.150.230.12E1+2+9+30.220.050.100.01Ost0.080.230.220.22O2---0.01O50.010.14--O7--0.010.04O3+40.200.110.300.32O3+4+70.170.010.210.04O3+4+20.340.190.220.06O3+4+80.200.320.040.30O3+4+23+2---0.01-: Arrangement absentSome authors consider recombination as the main factor determining the chromosomal distribution of TEs [33,34], but see [34]. The model of ectopic exchange, predicts a negative correlation between recombination rate and TE copy number if ectopic exchange is reduced in parallel with regular meiotic recombination rate [35,36]. Under this model, TEs are expected to be more abundant in regions of low recombination as inversions or inversion break-points. In these regions the probability of induction of deleterious rearrangements produced by unequal recombination between TEs, is low because most of the time inversions will be found in heterozygous state (recombination is suppressed inside). Experimental evidences [37-39] suggest that TEs are responsible of chromosomal inversions in natural populations of Diptera and are particularly abundant inside and near inversion break-points [6,7,40].In order to know whether an association between high insertion sites and arrangements exist, we computed the product-moment correlation coefficient (r) for high-frequency sites (Table 5). We observed two bilbo sites of particular interest (67A and 89C) that show the highest correlation coefficients. The 67A site is located inside the breakpoint of arrangement E12 and is significantly associated with E1+2+9+12 in Davis (r = 0.64) and Maipú (r = 0.85). The 89C site is located near the break-point of O8 arrangement and is significantly correlated with arrangement O3+4+8 in Davis (r = 0.58) and Bellingham (r = 0.34), and only marginally (r = 0.26) in Maipú. Other instances of significant associations are not so easily explained because sites are external to inversion breakpoints. Thus, highly occupied 74D bilbo site is located outside of chromosomal inversions, yet, it is also significantly associated with E1+2+9+12 in Davis (r = 0,33) and Maipú (r = 0,24). This site is also highly occupied by gypsy but in this case associations are not significant. In other cases we observe associations of sites inside highly frequent inversions where the crossing-over is not reduced. This is the case of 11B bilbo site, for example, negatively associated to A2 arrangement in all populations except Bellingham but located inside it.Table 5Correlation coefficients between chromosomal arrangements and high insertion frequency (HF) sitesPopulationsDABEMABOHF sites of bilboArrang.rq-valuerq-valuerq-valuerq-value11BA2-0.13(0.42)0.35**(10-7)0.28*0.02-0.07(0.99)20AJ10.24(0.11)---0.06(0.39)-0.04(0.99)43BUst-0.08(0.46)0.040.440.23*(0.04)0.24(0.99)45CUst0.11(0.56)0.20(0.08)0.40**(3.10-7)0.05(0.99)U1+2-0.13(0.59)0.22*(0.04)-0.190.080.02(0.99)45DUst0.15(0.42)0.49**(3.10-7)--------U1+2-0.02(0.86)-0.34**(2.10-3)--------53AUst-0.20(0.15)-0.35**(3.10-7)-0.22*(0.04)-0.05(0.99)U1+20.26(0.11)0.050.440.18(0.08)0.07(0.99)57DE1+2+9--0.25(0.12)-0.10(0.44)0.57(0.41)E1+2+9+3-0.06(0.72)0.17(0.15)0.28*(0.04)-0.02(0.99)59CEst0.36**(5.10-3)-----0.02(0.80)67AEst-0.15(0.42)-0.25*(0.04)-0.69**(4.10-8)-0.03(0.99)E1+2+9+120.64**(1.10-4)--0.85**(4.10-8)0.22(0.51)74DE1+2+9+120.33*(0.04)--0.24*(0.04)0.13(0.99)82AOst0.12(0.63)-0.10(0.28)0.29*(0.02)-0.16(0.99)83CO3+4+70.00(0.72)0.09(0.08)-0.14(0.17)0.56(0.11)85AO3+4-0.14(0.46)0.27*(0.02)0.10(0.17)0.03(0.99)O3+4+7-0.17(0.41)0.080.53-0.24*(0.04)-0.04(0.99)89CO3+4-0.160.42-0.25*(0.03)-0.21(0.06)-0.09(0.99)O3+4+2-0.36**(5.10-3)-0.31**(7.10-3)0.09(0.22)0.07(0.99)O3+4+80.58**(1.10-4)0.34**(4.10-3)0.26(0.05)0.02(0.99)91BOst0.28(0.16)0.02(0.12)0.32*(0.03)-0.06(0.99)O5-0.03(0.86)0.28*(0.02)--------98DO3+4+20.05(0.720.12(0.18)0.32*(0.03)0.29(0.66)HF sites of gypsy41CU1+20.35*0.52**0.25*--52DU1+20.32*0.28*-0.12--63CEst-----0.46**--E1+2+9+12----0.45**--74DE1+2+9+30.31-0.090.34-0.04Only high insertion sites showing correlation coefficient values either significant or higher than 0.20 at least in one population, are considered. Arrang: arrangement; r: Correlation coefficient; -: indicates cases where correlations cannot be computed because of low ETs copy number; --: indicates the lack of a site or an inversion in the population). See Table 1 for population origin. *P < 0.05; **P < 0.01. Q-value and Bonferroni corrections were applied to bilbo and gypsy respectivelyDiscussionBilbo and gypsy distributions are different in original and colonizer populationsResults show a clear differential TE distribution in original and colonizing populations. While in the original population most sites have low insertion frequencies, colonizing populations present some highly occupied sites, with frequencies higher than 50% for bilbo and close to 20% for gypsy. Interestingly, most of them are common to all populations. Mean copy number of both elements is higher in colonizing populations than in the original one due to the presence of these highly occupied sites.Low occupied sites would represent insertions occurred after colonization and/or copies from the original population whose frequency is decreasing in colonizing populations. An argument in favour of the former hypothesis is the existence of unique sites that would correspond to new transpositions (i.e. site 48D of gypsy), while the latter hypothesis explains the existence of low-occupancy original sites common to different populations (i.e. site 41A of gypsy or 85B of bilbo).High insertion frequency sites are most likely due to a founder event during the colonization process (the founder hypothesis), as previously reported in other Drosophila species [9,11,13]. In D. buzzatii this hypothesis was also verified by molecular studies showing identical Osvaldo retrotransposon structures and flanking genomic sequences in high insertion frequency sites from different colonizing populations [14].In this study two lines of evidence support the founder hypothesis. First, the two studied TEs belong to different subclasses, yet they show a similar population behaviour. Second, most highly occupied sites are located in colonizing population chromosomes, although some exceptions occur for bilbo whose insertion frequency exceeds 10% in 9 sites in the original population of Bordils. All these sites correspond also to high insertion sites in colonizing populations, except 90C site and 21A, which are, respectively, free of insertions or occupied at low frequency in America.The presence of high frequency sites in the original population could be a consequence of the transposition mechanism of bilbo, a LINE element. It has been shown that LINE elements (L1) make 5' truncated copies during their transposition mechanism indicating that 5' sequences are not absolutely necessary to insertion [41-43]. In fact, the majority of the L1 copies present in mammalian genomes are 5' truncated with a length of not more than 1 kb [1,44]. We can think that selection against truncated, \"dead-on-arrival\" (DOA) copies should be weak because they are not transcribed, potentially immobile and shorter than full copies. Thus, deleted copies could persist in some genomic regions without being completely eliminated by natural selection. In fact, some Drosophila TE families (most of them LINE like elements) seem to be only marginally affected by purifying selection, reaching high insertion frequencies in euchromatin [45].On the other hand, some of bilbo high frequency sites from Bordils could be explained by the dragging effect from the rich inversion polymorphism of D. subobscura. For example the 67A site located in the break-point of E12 arrangement presents highly significant correlations with this arrangement in 2 out of 3 colonizing populations. In Bordils, this correlation is not significant because of the lower frequency of this arrangement in this population. Arrangements of chromosome E cover approximately 75% of its length and it is not rare to find this kind of associations. In this chromosome another high insertion frequency site (74D) shows association with the same E1+2+9+12 chromosomal arrangement. This site corresponds to a heterochromatic telomeric site where it is not rare to find an accumulation of TE insertions. In fact gypsy is inserted also in this chromosomal site at occupation rates that range from 1.3 to 11.4%. Accumulation of TEs in heterochromatin is well documented in D. melanogaster where a significant excess of insertions were reported in heterochromatin, dot and Y chromosomes alike [46-49].Seasonal fluctuations in population frequency of chromosomal rearrangements can modify recombination rates and associations between arrangements and genes. In D. subobscura no seasonal fluctuations were reported in some works [50,51], but fluctuations and seasonal changes of associations between chromosomal inversions and allozymes were reported in others, specially in the O chromosome from original populations [52,53]. In the present case, we observe no associations between insertions and specific chromosomal arrangements in the original population, but we do detect this kind of associations in colonizing populations (where fluctuations were not studied). However, changes in associations between chromosomal arrangements and chromosomal sites do not follow a definite trend. As an example, the UST arrangement, whose frequency has increased in all colonizing populations, shows a positive association to 43B and 45C sites but a negative one to 53A site in Maipu. This is a rather odd outcome since increase of rearrangement frequency is always expected to break down associations due to an increase of recombination rate. So, the likely explanation would be that fluctuations do not affect associations or at least not in the same way for every studied rearrangement polymorphism.On the other hand, we favor the general idea that the positive correlation between arrangements and TE copies is not due to an inversion effect but, most probably, to the founder event [19,25,54,55]. This could explain why arrangement E1+2+9+12and the 74D site, which is located outside of the inversion, show a positive association and also why an excess of classes including positive correlation coefficients between chromosomal sites was observed in some chromosomes like E. Genetic estimates suggest that the number of founders ranged from 10 to 150 [25,56]. If some founders carried together this site and this arrangement, both will appear together in all populations because they are identical by descent. The founder hypothesis is favored by the fact that all correlations between sites and arrangements are significant only in colonizing populations. In the original population in spite of having correlation coefficients of 0.57 (in 57D) and 0.56 (in 83C) with E1+2+9 and O3+4+7 respectively, these are not significant. In fact, these two arrangements are currently decreasing in frequency in the Mediterranean populations and perhaps these combinations descend also from a few individuals. All these considerations suggest that most of the associations detected are due to a founder effect.The general rule, as reported in D. melanogaster [45,57], is that TEs are spread and have low insertion frequencies in euchromatin. In some cases, however, accumulations of TEs in some chromosomal sites have been reported, as in the 42B [58], 87C [59] and 38 [60] regions, of D. melanogaster, and the 85D region of D. subobscura [61,62], and even fixation has occurred, as in the 42C site in natural populations of D. simulans [63]. Preferential insertion sites (hotspots) have been suggested for some Drosophila elements [64-66] and we cannot completely discard the possibility of an activation of transposition to specific hotspots during the colonization process. This hypothesis could be verified if a process affecting equally the two TEs studied occurred, as shown in D. melanogaster. In this species some proteins are involved in RNA-silencing mechanisms for retrotransposable elements repression [67-69]. We cannot discard the existence of a similar mechanism in D. subobscura that was de-repressed as a consequence of the colonization process contributing simultaneously to an increase of transposition of different transposable elements.Factors affecting TEs distribution in D. subobscuraIn Drosophila, TEs seem to be maintained in populations as the result of a balance between transposition and opposing forces that reduce their copy number. In this way selection can act either directly against deleterious insertions or indirectly against deleterious chromosomal rearrangements produced by ectopic recombination between TEs [4,5,36,70]. In this work a test of selection against deleterious insertions was done by comparing copy numbers between X and autosomes, selection being more effective in the former than in the latter.For gypsy we observe a clear tendency to follow a selection model, except in Maipú. This result is in concordance with that observed in a natural population of D. melanogaster with this element [71]. For bilbo the data do not fit a selection model against deleterious insertions; even in those cases where the test is significant, a higher copy number on A (X), compared to autosomes, is observed. A possible explanation of this result is that bilbo could have a differential transposition rate between X and autosomes. Some examples of transposition restricted to female or male D. melanogaster germ line have been reported [72,73] and they should be taken into account when X and autosomes are compared. On the other hand, the discrepancies observed between the two elements may be accounted for by the different factors that control copy numbers in each of these elements. In D. melanogaster gypsy is a retrovirus [74] submitted probably to a strong selection effect, its transposition depending on the presence of permissive alleles most likely segregating in natural populations. In D. subobscura this retrotransposon seems to be non infectious because current available copies have an apparently inactive env region [75], but this does not discard the putative presence of alleles that control its transposition. On the other hand bilbo is a LINE element and could be submitted to a soft selection pressure due to its DOA transposition mechanism. Most of the copies are probably deleted and its deleterious capability by transposition is diminished. The model of selection against deleterious insertions has been questioned by some authors [28,48] because neither all ETs nor all populations had a lower insertion frequency on X chromosomes compared to autosomes. However in a later work [76], where the authors reanalyze the data including more results from other species, selection against insertions is considered as the major mechanism of TE copy number control. On the other hand, values of selection coefficients against deleterious mutations could not be comparable to mutations associated to TE insertions. Moreover, deleterious effects of TEs can be species specific and populations may also sometimes suffer TE mobilizations that mask selection effects on TE distribution.In this work each element presents a different behavior probably due to their distinct transposition mechanisms. Moreover we should not forget that elements which are stable in some genome conditions could be unstable in others. Recently mobilized TEs and/or colonization events, in populations, could lead to a differential copy distribution between chromosomes, rendering the selection undetected. This could be the case of Maipú, a new colonizing Argentinian population, which shows a distribution pattern for gypsy and bilbo quite different from the other colonizing populations. In particular, some high insertion frequency sites are more represented, or even exclusive, in this population. It is possible that Maipú was established through a bottleneck of founder flies from Chile as a consequence of a secondary colonization. In this case, we cannot discard the existence of new transposition events in founders induced by the new environmental conditions encountered as previously proposed by other authors [10,11]. If this colonization occurred recently, as indicated by collecting records, selection has not had enough time to act, explaining the discrepancies in this population when comparing A (X) and autosome copy numbers in Table 3 or when this population is included in heterogeneity tests. In addition if TEs are not at equilibrium, departures from random distribution across chromosomes could reflect the insertion pattern rather than the effect of natural selection.Another model proposed to explain the TE dynamics is the selection against deleterious arrangements produced by ectopic recombination between TEs. In D. subobscura accurate measures of recombination rate are not available and it is not possible to calculate a correlation between TE copies and recombination rates. This species has a rich inversion polymorphism in all chromosomes and recombination is reduced in heterokaryotypes. Under this model we expect accumulation of TEs in inverted segments, and in inversion break points or near them. In some cases arrangements include overlapping inverted fragments, often reaching frequencies higher than the standard arrangements, but in other cases, of low frequency arrangements, TE copy number is too low to allow statistical tests. Also recombination between non-overlapping inversions or inversion complexes may also be prevented [77].We looked for accumulations of bilbo and gypsy in breakpoints of inversions but only one high insertion frequency bilbo site, 67A, coincides with an inversion breakpoint (E1+2+9+12). In another case the 89C high frequency site of bilbo is located near the inversion O8 and shows a significant correlation with O3+4+8 arrangement. This is in concordance with several unsuccessful attempts to localize in situ hybridization middle repeated sequences in D. subobscura inversions breakpoints [61,78]. These data notwithstanding, we cannot discard that other elements may be responsible of chromosomal inversion induction as reported in other Drosophila species [37,38].ConclusionWe conclude that the differential distribution of bilbo and gypsy between original and colonizing D. subobscura populations, is mainly due to a founder effect occurred during the colonization process of this species. We have shown that both founder effect and inversion polymorphism contribute notably to an excess of positive correlations between site pairs. Moreover the two transposable elements show a different pattern of distribution in populations that might be due to their differences in transposition and copy number regulatory mechanisms. This paper is also an attempt to emphasize the importance of population structure and history to explain the TE chromosomal distribution. We highlight the fact that comparisons in TE copy number between X and autosomes have to be interpreted cautiously. Sometimes TEs mobilizations can mask the effect of selection on TE distribution.MethodsDrosophila strainsThe control strain chcu carries the recessive markers cherry eyes and curled wings and is homokaryotypic for chromosomal arrangements Ast, Jst, Ust, Est and O3+4. It is kept by mass-culturing to maintain its viability. In situ hybridization for insertions of bilbo and gypsy displayed high stability over generations in 19C, 46A, 46C, 73A, 81D, 84A, 96A for bilbo and in 7C and 52A for gypsy.The original population was sampled in Spring 2005 in Bordils (42.30°N, Girona, Spain). The colonizing populations were sampled in Spring 2004 in Davis (38.33°N, California, USA) and Bellingham (48.45°N, Washington, USA), and in Spring 2005 in Maipú (36.52°S, Argentina).Mating system (prior to \"in situ\" hybridization)Individual males of natural populations were crossed with virgin females of the control line chcu. Insertion profiles were analyzed in F1 female larval progeny to include the X chromosome. The TE insertion profile of each male was deduced by subtracting the TE insertion profile of the control line from that of the F1 larva.In situ hybridization and DNA probesPolytene chromosome [16] squashes from salivary glands of third-instar larvae, prepared as described in [79], were hybridized with digoxigenin labelled probes of bilbo and gypsy. The probes consisted of PCR fragments (2.6 and 2.8 kb long) which included the reverse transcriptase region. Prehybridization solutions and posthybridization washes were done following a protocol by Roche [80]. PCR reactions were carried out in a final volume of 25 μl, including 1× activity buffer (Ecogen), 1.6 mM MgCl2, 0.2 mM of each dNTP (Roche), 0.4 μM primer (Roche), 10–20 ng of genomic template DNA, and 0.04 units per μl of Taq polymerase (Ecotaq from Ecogen). Amplifications were run in a MJ Research Inc. thermocycler programmed as follows: 5 min preliminary denaturation at 94°, 30 cycles of 45 s at 94° (denaturation), 45 s at specific PCR annealing temperatures, 1.5 min at 72° (extension) and a final extension for 10 min at 72°. PCR products were gel purified with a Geneclean kit (BIO 101) and labelled using the random primer method. After hybridization signal development was done using an anti-digoxigenin antibody conjugated with alkaline phosphatase (Roche).In situ hybridization is the more suitable method used in localization of TEs on chromosomal arms. However, the power of resolution of this technique allow us neither discriminate between closely neighbouring sites, nor between elements that diverge below 10%.Statistical analysesStatistical analyses were performed excluding centromeric and pericentromeric TEs insertions. The statistical software SPSS version 14.0 was used for most of the statistical data analyses.In cases of multiple testing, corrections were achieved measuring the significance of False Discovery Rates [81] through q values. To get the q-value we used the software QVALUE [82] on the p values obtained from the multiple test. When this test could not be applied, Bonferroni's correction was performed [83].Authors' contributionsMPGG participated in the design, the chromosomal slides, some statistical analyses and the writing of the manuscript. BECS collected the Bordils population, performed most of the technical work, the reading of slides and some statistical analyses. JB collected Davis and Bellingham populations, supervised all arrangement readings and performed the data set analyses. LS collected the Bordils population, contributed to the design and thoroughly revised the manuscript AF participated in the design, directed the project, coordinated the data analyses, contributed to the writing of the manuscript and collected the Maipú and Bordils populations. All authors read and approved the final manuscriptSupplementary MaterialAdditional file 1Poisson distribution: raw data of bilbo copy number per chromosome and population, P values and chi tests. A table of detailed tests of Poisson distribution of bilbo per chromosome and haploid genome.Click here for fileAdditional file 2Poisson distribution: raw data of gypsy copy number per chromosome and population, P values and chi tests. A table of detailed tests of Poisson distribution of gypsy per chromosome and haploid genome.Click here for file\n\nREFERENCES:\n1. IHGSCInitial sequencing and analysis of the human genomeNature200140986092110.1038/3505706211237011\n2. DeiningerPLRoy-EngelAMCraig NL, Craigie R, Gellert M, Lambowitz AMobile elements in animals and plant genomesMobile DNA II2002Washington D.C. , American Society for Microbiology10741092\n3. Le RouzicACapyPPopulation genetics models of competition between trasposable elements subfamiliesGenetics200617478579310.1534/genetics.105.05224116888345\n4. CharlesworthBCharlesworthDThe population dynamics of transposable elementsGenet Res198342127\n5. CharlesworthBLangleyCHThe population genetics of Drosophila transposable elementsAnnu Rev Genet19892325128710.1146/annurev.ge.23.120189.0013432559652\n6. CharlesworthBLapidACanadaDThe distribution of transposable elements within and between chromosomes in a population of Drosophila melanogaster. I. Element frequencies and distributionGenet Res1992601031141334899\n7. CharlesworthBLapidACanadaDThe distribution of transposable elements within and between chromosomes in a population of Drosophila melanogaster. II. Inferences on the nature of selection against elementsGenet Res1992601151301334900\n8. DecelièreGCharlesSBiémontCThe dynamics of transposable elements in structured populationsGenetics200516946747410.1534/genetics.104.03224315466430\n9. García GuerreiroMPFontdevilaAChromosomal distribution of the transposable elements Osvaldo and blanco in original and colonizer populations of Drosophila buzzatiiGenet Res20017722723810.1017/S001667230100503111486506\n10. VieiraCNardonCArpinCLepetitDBiémontCEvolution of the genome size in Drosophila. Is the invader's genome being invaded by transposable elements?Mol Biol Evol2002191154116112082134\n11. WisotzkeyRGFelgerIHuntJABiogeographic analysis of the Uhu and LOA elements in the Hawaiian Drosophila Chromosoma199710646547710.1007/s0041200502689391219\n12. FontdevilaAHewitt GM, Johnston AWB, Young JPWColonizing Species of DrosophilaMolecular Techniques in Taxonomy199157HHeidelberg , Springer-Verlag 249269\n13. LabradorMSelemeMDCFontdevilaAThe evolutionary history of Drosophila buzzatii. XXXIV. The distribution of the retrotransposon Osvaldo in original and colonizing populationsMol Biol Evol1998151532154712572617\n14. García GuerreiroMPFontdevilaAThe evolutionary history of Drosophila buzzatii. XXXVI. Molecular structural analysis of Osvaldo retrotransposon insertions in colonizing populations unveils drift effects in founder events.Genetics200717530131010.1534/genetics.106.06437817151248\n15. SturtevantAHThe classification of the genus Drosophila with the description of nine new speciesUniv Texas Publ194242135\n16. KrimbasCBDrosophila subobscura.1993Hamburg , Verlag Dr. Kovac\n17. AyalaFJSerraLPrevostiAA grand experiment in evolution: the Drosophila subobscura colonization of the Americas.Genome198931246255\n18. PrevostiARibóGSerraLAguadéMBalañáJMonclúsMMestresFColonization of America by Drosophila subobscura: experiment in natural populations that supports the adaptative role of chromosomal inversion polymorphism.Proc Natl Acad Sci USA1988855597560010.1073/pnas.85.15.559716593967\n19. BrncicDPrevostiABudnikMMonclúsMOcañaJColonization of Drosophila subobscura in Chile I. First population and cytogenetic studies.Genetica1981563910.1007/BF00126923\n20. BeckenbachATPrevostiASerraLColonization of North America by the European species Drosophila subobscura and D. ambiguaAm Midl Nat1986115101810.2307/2425832\n21. MenozziPKrimbasCBThe inversion polymorphism of D. subobscura revisited: Synthetic maps of gene arrangement frequencies and their interpretationJ Evol Biol1992562564110.1046/j.1420-9101.1992.5040625.x\n22. MestresFBalañáJSegarraCPrevostiASerraLColonization of America by Drosophila subobscura: Analysis of the O5 inversions from Europe and America and their implications for the colonizing process.Evolution1992461564156810.2307/2409960\n23. LatorreAMoyaAAyalaFJEvolution of mitochondrial DNA in Drosophila subobscuraProc Natl Acad Sci U S A1986838649865310.1073/pnas.83.22.864916578796\n24. RozasJHernándezMCabreraVMPrevostiAColonization of America by Drosophila subobscura: Effect of the founder event on the mitochondrial DNA polymorphism.Mol Biol Evol199071031092299979\n25. PascualMAquadroCFSotoVSerraLMicrosatellite variation in colonizing and Palearctic populations of Drosophila subobscura.Mol Biol Evol20011873174011319257\n26. CharlesworthBLapidAA study of ten families of transposable elements on X chromosome from a population of Drosophila melanogasterGenet Res1989541131252558961\n27. FisherRAStatistical methods for research workers195813rdNew York , Hafner\n28. MontgomeryEACharlesworthBLangleyCHA test for the role of a natural selection in the stabilization of transposable element copy number in a population of Drosophila melanogasterGenet Res19874931413032743\n29. Stumm-ZollingerEGoldschmidtEGeographical differentiation of inversion systems in Drosophila subobscura. Evolution195913899810.2307/2405946\n30. SokalRRRohlfFJBiometry19953rdNew York , Freeman\n31. BalanyàJSerraLGilchristGWHueyRBPascualMMestresFSoléEEvolutionary pace of chromosomal polymorphism in colonizing populations of Drosophila subobscura: An evolutionary time series.Evolution2003571837184514503625\n32. BalanyàJSoléEOllerJMSperlichDSerraLLong-term changes in the chromosomal inversion polymorphism of Drosophila subobscura. II. European populationsJ Zoolog Syst Evol Res20044219120110.1111/j.1439-0469.2004.00274.x\n33. BartoloméCMasideXCharlesworthBOn the abundance and distribution of transposable elements in the genome of Drosophila melanogasterMol Biol Evol20021992693712032249\n34. RizzonCMaraisGGouyMBiémontCRecombination rate and the distribution of transposable elements in the Drosophila melanogaster genomeGenome Res20021240040711875027\n35. GoldmanASHLichtenMThe efficiency of meiotic recombination between dispersed sequences in Saccharomyces cerevisiae depends upon their chromosomal location.Genetics199614443558878672\n36. Langley CHMontgomery EAHudson RKaplan NCharlesworth BOn the role of unequal exchange in the containment of transposable element copy numberGenet Res1988522232352854088\n37. CáceresMRanzJMBarbadillaALongMRuizAGeneration of a Widespread Drosophila inversion by a Transposable ElementScience199928541541810.1126/science.285.5426.41510411506\n38. LyttleTWHaymerDSThe role of the transposable element hobo in the origin of the endemic inversions in wild populations of Drosophila melanogasterGenetical Research199286113126\n39. MathiopoulosKDDella TorreAPredazziVPetrarcaVColuzziMCloning of inversion breakpoints in the Anopheles gambiae complex traces a transposable element at the inversion junctionProc Natl Acad Sci USA199895124441244910.1073/pnas.95.21.124449770505\n40. SniegowskyPDCharlesworthBTransposable element numbers in cosmopolitan inversions from a natural population of Drosophila melanogasterGenetics19941378158278088526\n41. BelancioVPWheltonMDeiningerPRequirements for polyadenilation at the 3' end of LINE-1 elementsGene20073909810710.1016/j.gene.2006.07.02917023124\n42. BusseauIPélissonABuchetonACharacterization of 5' truncated transposed copies of the I factor in Drosophila melanogasterNucleic Acids Research1989176939694510.1093/nar/17.17.69392550903\n43. FanningTGSingerMFLINE I: A mammalian transposable elementBiochem Biophys Acta19879102032122445384\n44. MedstrandPvan de Lagemaat L NMagerDLRetroelement distributions in the human genome: Variations associated with age and proximity to genes.Genome Res2002121483149510.1101/gr.38890212368240\n45. PetrovDAminetzachYTDavisJCBensassonDHirshAESize matters: Non-LTR retrotransposable elements and ectopic recombination in DrosophilaMol Biol Evol20032088089210.1093/molbev/msg10212716993\n46. BartoloméCMasideXThe lack of recombination drives the fixation of transposable elements on the fourth chromosome of Drosophila melanogasterGenetical Research2004839110010.1017/S001667230400675515219154\n47. CarmenaMGonzálezCTransposable elements map in a conserved pattern of distribution extending from beta-heterochromatin to centromeres in Drosophila melanogasterChromosoma199510367668410.1007/BF003442287664614\n48. CharlesworthBSniegowskyPDStephanWThe evolutionary dynamics of repetitive DNA in eukaryotesNature199437121522010.1038/371215a08078581\n49. DimitriPJunakovicNArcaBColonization of herochomatic genes by transposable elements in Drosophila.Mol Biol Evol20032050351210.1093/molbev/msg04812654931\n50. KrimbasCBThe genetics of Drosophila subobscura populations. III. Inversion polymorphism and climatic factorsMol Gen Genet19679913315010.1007/BF004261585584166\n51. PrevostiAChromosomal polymorphism in Drosophila subobscura populations from Barcelona (Spain)Genet Res196452738\n52. FontdevilaAZapataCAlvarezCSánchezLMéndezJEnriquezIGenetic coadaptation in the chromosomal polymorphism of Drosophila subobscura. I. Seasonal changes of gametic disequilibrium in a natural populationGenetics198310593595517246183\n53. Rodríguez-Trelles FSeasonal cycles of allozyme-by-chromosomal-inversion gemetic disequilibrium in Drosophila subobscuraEvolution200357483984812778553\n54. BalanyàJSegarraCPrevostiASerraLColonization of America by Drosophila subobscura: The founder event and a rapid expansion.J Hered1994854274327995925\n55. PrevostiASerraLRibóGAguadéMSagarraEBalanyàJMonclúsMGarcíaMPThe colonization of Drosophila subobscura in Chile. II. Clines in the chromosomal arrangements.Evolution19853983884410.2307/2408683\n56. MestresFPeguerolesGPrevostiASerraLColonization of America by Drosophila subobscura: lethal genes and the problem of the O5 inversion.Evolution1990441823183610.2307/2409511\n57. Biémont CPopulation genetics of transposable DNA elements. A Drosophila point of view.Genetica199286678410.1007/BF001337121334919\n58. Biémont CGautier CInteractions between transposable elements for insertion in the Drosophila melanogaster genomeHeredity19896312513310.1038/hdy.1989.832548982\n59. MasideXBartoloméCCharlesworthBS-element insertions are associated with the evolution of the Hsp70 genes in Drosophila melanogaster.Current Biol2002121686169110.1016/S0960-9822(02)01181-8\n60. BergmanCMQuesnevilleHAnxolabéhèreDAshburnerMRecurrent insertion and duplication generate networks of transposable element sequences in the Drosophila melanogaster genomeGenome Biol2006711\n61. FelgerISperlichDCytological localization and organization of dispersed middle repetitive DNA sequences of Drosophila subobscuraChromosoma19899834235010.1007/BF002923872612293\n62. VivasMVGarcía-PlanellsJRuízCMarfanyGParicioNGonzález-DuarteRde FrutosRGEM, a cluster of repetitive sequences in the Drosophila subobscura genomeGenet Res19992294757\n63. VieiraCBiémontCGeographical variation in insertion site number of retrotransposon 412 in Drosophila simulans.J Mol Evol19964244345110.1007/BF024986388642613\n64. BelyaevaESAnanievEVGvozdevVADistribution of mobile dispersed genes (mdg-1 and mdg-3) in the chromosomes of Drosophila melanogaster .Chromosoma198490161910.1007/BF00352273\n65. GvozdevVABelyaevaESIlyinIVAmosovaISKaidanovLZSelection and transposition of mobile dispersed genes in Drosophila melanogaster. Cold Spring Harbor Symp Quant Biol1980456736856266759\n66. PasyukovaEBelyaevaESKoganGLKaidanovLZGvozdevVAConcerted transpositions of mobile genetic elements coupled with fitness changes in Drosophila melanogaster.Mol Biol Evol198632993122832692\n67. Mével-NinioMPélissonKinderJCamposARBuchetonAThe flamenco locus controls the gypsy and ZAM retroviruses and is required for Drosophila oogenesisGenetics20071751615162410.1534/genetics.106.06810617277359\n68. PélissonASarotEPayen-GroschêneGBuchetonAA novel repeat-associated small interfering RNA-mediated silencing Pathway downregulates complementary sense gypsy transcripts in somatic cells of the Drosophila ovaryJournal of Virology20078141951196010.1128/JVI.01980-0617135323\n69. VaginVVSigovaALiCSeitzHGvozdevVZamorePDA distinct small RNA pathway silences selfish genetic elements in the germlineScience200631332032410.1126/science.112933316809489\n70. KaplanNLBrookfieldJFYTransposable elements in mendelian populations. III. Statistical results.Genetics19831044859517246144\n71. BiémontCLemeunierFGarcía GuerreiroMPBrookfieldJFGautierCAulardSPasyukovaEGPopulation dynamics of the copia, mdg1, mdg3, gypsy and P transposable elements in a natural population of Drosophila melanogasterGenet Res1994631972128082837\n72. BuchetonABusseauITeningesACraig NL, Craigie R, Gellert M, Lambowitz AMI elements in Drosophila melanogasterMobile DNA II2002Washington DC , ASM press796813\n73. PasyukovaENuzhdinWLFlavellAJGerm line transposition of the copia retrotransposon in Drosophila melanogaster is restricted to males by tissue-specific control of copia RNA levelsMol Gen Genet199725511512410.1007/s0043800504799230904\n74. KimATerzianCSantamaríaPPélissonAPrud´hommeNBuchetonARetroviruses in invertebrates: The gypsy retrotransposon is apparently an infectious retrovirus of Drosophila melanogasterProc Nat Acad Sci USA1994911285128910.1073/pnas.91.4.12858108403\n75. AlberolaTMde FrutosRMolecular structure of a gypsy element of Drosophila subobscura (gypsy Ds) constituting a degenerate form of insect retrovirusesNucleic Acids Res19962491492310.1093/nar/24.5.9148600460\n76. BiémontCVieiraCHooglandCCizeronGLoevenbruckCArnaultCCaranteJPMaintenance of transposable element copy number in natural populations of Drosophila melanogaster and D. simulansGenetica199710016116610.1023/A:10183295283149440269\n77. MestresFSanzJSerraLChromosomal structure and recombination between inversions in Drosophila subobscuraHereditas199812810511310.1111/j.1601-5223.1998.00105.x9652229\n78. CireraSMartín-CamposJMSegarraCAguadéMMolecular Chracterization of the breakpoints of an inversion fixed between Drosophila melanogaster and D. subobscuraGenetics19951393213267705632\n79. LabradorMNaveiraHFontdevilaAGenetic mapping of the Adh locus in the repleta group of Drosophila by in situ hybridizationJ Hered19908183862185305\n80. SmidtERGrünewald-Janho S, Keesey J, Leous M, van Miltenburg R, Schroeder CA simplified and efficient protocol for non-radioactive in situ hybridization to polytene chromosomnes with a DIG-labeled DNA probeNon-radioactive In situ hybridization Aplication Manual1996Mannheim , Roche97107\n81. BenjaminiYHochbergYControlling the false discovery rate. A practical and powerful approach to multiple testing.J R Stat Soc Ser B Stat Meth199557289300\n82. StoreyJDA direct approach to false discovery ratesJ R Stat Soc Ser B Stat Meth20026447949810.1111/1467-9868.00346\n83. HolmSA simple sequencially rejective multiple test procedureScand J Stat197966570"
+}

batch_3/PMC2533120.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533120",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533120\nAUTHORS: Ran Elkon, Reuven Agami\n\nABSTRACT:\nElucidation of regulatory roles played by microRNAs (miRs) in various biological networks is one of the greatest challenges of present molecular and computational biology. The integrated analysis of gene expression data and 3′-UTR sequences holds great promise for being an effective means to systematically delineate active miRs in different biological processes. Applying such an integrated analysis, we uncovered a striking relationship between 3′-UTR AU content and gene response in numerous microarray datasets. We show that this relationship is secondary to a general bias that links gene response and probe AU content and reflects the fact that in the majority of current arrays probes are selected from target transcript 3′-UTRs. Therefore, removal of this bias, which is in order in any analysis of microarray datasets, is of crucial importance when integrating expression data and 3′-UTR sequences to identify regulatory elements embedded in this region. We developed visualization and normalization schemes for the detection and removal of such AU biases and demonstrate that their application to microarray data significantly enhances the computational identification of active miRs. Our results substantiate that, after removal of AU biases, mRNA expression profiles contain ample information which allows in silico detection of miRs that are active in physiological conditions.\n\nBODY:\nIntroductionMicroRNAs (miRs) are a growing class of non-coding RNAs that is now recognized as a major tier of gene control, predicted to target more than 30% of all human protein-coding genes [1],[2]. miRs suppress gene expression via binding to regulatory sites usually embedded in the 3′-UTRs of their target mRNAs, leading to the repression of translation occasionally associated with mRNA degradation. Target recognition involves complementary base pairing of the target site with the miR's seed region (positions 2–8 at the miR's 5′ end), although the exact extent of seed complementarity is not precisely determined, and can be modified by 3′ pairing [2]–[4]. Despite intensive efforts in recent years, biological functions carried out by miRs have been characterized for only a minority of these genes, and therefore, elucidating regulatory roles played by miRs in various biological networks constitutes one of the major challenges facing biology today. Bioinformatics analyses can significantly contribute to elucidation of miR functions; in particular, the integrated analysis of gene expression data and 3′-UTR sequences that holds promise for systematic dissection of regulatory networks controlled by miRs and of cis-regulatory elements embedded in 3′-UTRs.Similar bioinformatics approaches that integrates gene expression data and promoter sequences proved highly effective in delineating transcriptional regulatory networks in a multitude of organisms ranging from yeast to human [5]–[7]. Microarray measurements reflect the total effect of all regulatory mechanisms that control gene expression, including both transcriptional and post-transcriptional mechanisms; thus, genome-wide expression profiles should yield ample information not only on transcriptional networks, but also on regulatory networks regulated by miRs and RNA binding proteins (RBPs) that modulate mRNA stability, and that usually act via regulatory elements in 3′-UTR of their target genes [8]. Although mRNA degradation seems to be a secondary mode of miRs' action (with inhibition of translation being the primary one), since each miR is predicted to directly affect the expression level of dozens of target genes, such an orchestrated effect should be discernable by statistical analysis of wide-scale mRNA expression data, even if the effect on each target is only a subtle one. This orchestrated effect could serve as a molecular fingerprint for miRs activity under given biological conditions. Indeed, several pioneering studies provided strong evidence of the ability to computationally decipher miR-mediated regulatory networks from mRNA expression data alone or in correlation with miR expression profiles [9]–[14].In this study, we applied an integrated analysis of gene expression data and 3′-UTR sequences aimed at identifying miRs that are active in a given biological process. Applying such analysis we discovered in numerous microarray datasets a major bias that resulted in a striking relationship between 3′-UTR AU content and gene response. We show that this surprising link between gene's response and 3′-UTR base composition is secondary to a more basic relationship between gene's response and base composition of its probes on the chip. We demonstrate that this bias causes many false positive calls in computational searches for active miRs from mRNA expression data. Therefore, removal of this bias, which is in order in any analysis of microarray datasets, is of crucial importance when integrating expression data and 3′-UTR sequences to identify regulatory elements embedded in this region. Our results substantiate that computational analysis of mRNA expression data, after appropriate removal of AU biases, can accurately detect active miRs that control various biological processes under physiological conditions.ResultsWe set out to demonstrate that integrated computational analysis of mRNA expression data and 3′-UTR sequences can accurately uncover miRs that participate in the regulation of a given biological process. As the role of miRs in different branches of hematopoiesis is well characterized [15]–[18], we first analyzed a dataset that recorded global gene expression profiles for multi-potent hematopoietic progenitor cells (HPCs) undergoing multi-lineage differentiation [19]. Since miRs often induce degradation of their target mRNAs, we expected the 3′-UTR of genes whose expression is induced during differentiation to be enriched for seed signatures of miRs that become inactive in this process, and vice versa—that the 3′-UTR of genes whose expression is repressed would be enriched for seed signatures of miRs that become active during the process.Before employing statistical tests to identify over-represented seed sequences among up- or down-regulated genes, we examined whether a more global trend in base composition could be detected in the 3′-UTR sequences of the responding genes. For example, if the 3′-UTRs of the up-regulated genes are generally more AU-rich compared to the 3′-UTRs of the non-responding genes, then any statistical search for over-represented seed signatures among the up-regulated genes is expected to yield false positive calls for miRs whose seed signature is AU-rich. One effective means for detecting such false positive calls is to repeat the over-representation tests with randomly permuted miR seeds (which preserve the seed's base composition). If an enrichment of a certain miR seed is accounted for merely by base composition, then it is expected to be non-specific and detected also for randomly permuted seeds derived from the original one.Therefore, as a first step in the analysis of the HPC dataset, we checked whether a global 3′-UTR base composition trend is associated with the multi-lineage differentiation. We detected a very strong correlation between 3′-UTR base composition and gene response at several time points in this dataset. For example, there was an exceptionally strong relationship between AU content and gene response at the 16 hr time point after induction of HPC differentiation into megakaryocytes: 3′-UTRs of down-regulated genes were significantly more AU-rich than those of up-regulated ones (Figure 1). (The mean 3′-UTR AU content of the 5% most down-regulated and most up-regulated genes were 60.6% and 52.7%, respectively, p<10−99, Wilcoxon test.) The other three lineages in this dataset displayed similarly strong trends (Figure S1).10.1371/journal.pcbi.1000189.g001Figure 1Relationship between 3′-UTR AU content and gene response during HPC differentiation.Expression profiles were measured at several time points after stimulation of HPC differentiation into megakaryocytes. To visualize the relationships between 3′-UTR AU content and gene response, the genes were sorted for each time point according to their fold of repression/induction relative to the expression level at t0, and the mean 3′-UTR AU content was calculated in a sliding window that encompassed in each step 5% of the genes included in the analysis. (At each step the sliding window was moved to the right by 5% of its size.) Each plot corresponds to the time point indicated above it. Genes are sorted on the X-axis according to their response, from the most repressed genes at the left to the most induced genes at the right. The Y-axis represents the mean 3′-UTR AU content calculated on each sliding window. The p value above each plot is for the comparison (Wilcoxon test) between the 3′-UTR AU content of the top 5% (most strongly up-regulated) and bottom 5% (most strongly down-regulated) genes at the corresponding time point. Note the striking relationship between 3′-UTR AU content and gene response at the 16 hr time point.The strength of the relationship between 3′-UTR AU content and gene response in the HPC dataset prompted us to search for such trends in other datasets. Surprisingly, we found such relationships, with similarly high statistical significance, in numerous microarray datasets (data not shown). Still more suspicious, we observed the relationship even when we compared different control samples within a dataset. This led us to question whether the relationship observed between 3′-UTR AU content and gene response reflects any true biological regulatory mechanism, or is rather a result of some technical artifact in microarray measurements. We found a definitive answer to this question by analyzing a technical dataset published by van Ruissen et al. [20]. This dataset profiled a universal reference RNA pool in two independent oligonucleotide chips (Affymetrix HGU133A). Comparing the data from these two arrays, which measure identical and artificial RNA pools, we again found a striking relationship between 3′-UTR AU content and difference in gene expression level (Figure 2), pointing to a major AU bias in microarray measurements. This AU response bias is not specific to a particular data preprocessing method, as it existed in data under different preprocessing and normalization schemes; namely, rma, gcrma, and mas5 (Figure S2). In this technical dataset, we detected no preference for A or U in the bias, and no major 3′-UTR length bias (Figure S3).10.1371/journal.pcbi.1000189.g002Figure 2Strong relationship between 3′-UTR AU content and gene response detected in a comparison between technical replicates.The figure shows the relationship between 3′-UTR AU content and gene fold-change in a comparison between two chips hybridized with identical universal reference RNA pools. The plot was generated as described in the legend to Figure 1. A highly significant relationship between 3′-UTR AU content and gene response was detected in this technical comparison (p value = 8.1*10−84 for the comparison between the bottom and top 5% ‘responding’ genes), pointing to a major AU bias in microarray measurements.Next, we sought to elucidate the sources of the AU response bias. A well-documented bias in microarray measurements is the one between probe intensity and response [21], which is routinely visualized using M-A plots. We first suspected that the observed association between 3′-UTR AU content and gene response is a mere reflection of the intensity-response bias. However, there was no intensity-bias in the above technical dataset, which points that the 3′-UTR AU response bias is distinct from the intensity-response bias (see Figure 3A and 3B; in the latter, adopting the concept of M-A plots, we introduced the M-AU plot to visualize the AU response bias). The AU response bias exists over a large range of intensities (Figure S4), and, furthermore, the gcrma method which takes into account the correlation between probe's AU content and intensity did not cancel it.10.1371/journal.pcbi.1000189.g003Figure 3M-A and M-AU plots.(A). M-A plot shows that there is no intensity-response bias in the comparison between the two chips hybridized with identical universal reference RNA pools. The Y axis (denoted as M) represents the log2 fold-change and the X-axis (denoted as A) represents the average log2 intensity. Each dot in the plot corresponds to a gene in the dataset. (B). Adopting the M-A plot concept, we introduced the M-AU plot, in which the Y axis represents the log2 fold-change (as in the M-A plots), and the X axis represents the 3′-UTR AU content of a gene. The M-AU plot shows a major AU bias in this technical dataset. The red line is the lowess smoothing line calculated for the scatter plot.In the vast majority of present chips, probes are selected from the 3′-end of target transcripts. This is also the case for the technical dataset that we have analyzed, which used the Affymetrix HGU133A chip. Therefore, as expected, we observed in this dataset also a strong relationship between probeset AU content and response (similar to the one observed between gene's 3′-UTR AU content and response) (Figure S5). To test whether the AU artifact origins either from base-composition properties of 3′-UTR of target transcripts or of that of the chip probes, the sequence of probes and target 3′-UTRs need to be uncoupled. The new generation Affymetrix chips break this coupling as their probes are selected from all regions of target transcripts. We therefore analyzed a second technical dataset, recently published by Pradervand et al. [22] which used the new Affymetrix Human Gene 1.0 ST Array. In this dataset too, we detected a strong AU response bias. That is, we observed a significant relationship between probeset AU content and response in a comparison between duplicate control chips. Importantly, carrying out a probe-level analysis, we found that probes located at 5′-UTR and CDS regions show a similar AU bias as probes located at 3′-UTRs (Figure 4). This finding indicates that the link between gene's response and 3′-UTR base composition is secondary to a more basic bias in microarray measurements which links gene response with base composition of its probes.10.1371/journal.pcbi.1000189.g004Figure 4The AU response bias is related to probe base composition regardless probe location along the target transcript.Probe-level M-AU plot for the comparison between two chips hybridized with a common human brain reference sample. This dataset used the new generation Affymetrix Human Gene 1.0 ST Array, in which probes are located throughout the target transcripts. We generated plots which either included all probes, or included separately only those mapped to the 5′-UTR, CDS, or 3′-UTR of the targets. (As the length of each probe is 25 bases, probe's AU content (X axis) gets only discrete values in the 0–100% range with jumps of 4%). Probes mapped to the different transcript regions exhibited similar level of AU response bias.We next evaluated the effect of the AU bias on computational identification of active miRs from microarray data. Searching for miRs that are active in biological conditions examined in a dataset, we utilized miR target prediction generated by TargetScanS [2], and applied the following statistical test: for each miR family and for each condition in a dataset, we tested whether the set of predicted miR target genes is significantly induced or repressed compared to a background set consisting of all the non-target genes (see Methods). The technical dataset which profiled the universal reference RNA pool served us as a negative test case in which no real biological signal exists. Applying the statistical tests to this dataset, we identified nine miR families whose target sets showed statistically significant response (Table 1). Of course, in this negative test case, all calls are false positive ones; and, as expected, all the falsely identified miR families had an AU-rich seed (the seed of eight out of the nine calls contained at least 5 A or U bases, while the prevalence of miRs with such seed among all the miRs tested was less than 25%; Table 1). Next, for each miR family identified as significant, we repeated the statistical tests, but this time with randomly permuted miR seeds. In all cases, permuted seeds showed similar statistical significance to the original seeds (Table 1), demonstrating the utility of such permutation tests in detecting non-specific results caused by correlation between base composition of miR-seeds and 3′-UTRs of the responding genes.10.1371/journal.pcbi.1000189.t001Table 1Active miRs falsely identified in the negative test case.Without AU NormalizationmiR ID\np ValuemiR SeedBest Permuted p Valuea\nmiR.1862.41*10−9\n\nAAAGAAU\n1.60*10−11\nmiR.5431.72*10−7\n\nAACAUUC\n2.01*10−7\nmiR.4962.24*10−7\n\nUUACAUG\n3.14*10−7\nmiR.200b.4293.07*10−7\n\nAAUACUG\n9.84*10−10\nmiR.3811.41*10−5\n\nAUACAAG\n2.48*10−10\nmiR.261.62*10−5\n\nUCAAGUA\n4.07*10−5\nmiR.203.11.79*10−5\n\nGAAAUGU\n6.34*10−6\nmiR.132.2120.00017\nAACAGUC\n0.0019miR.1810.00029\nACAUUCA\n3.45*10−10\nAfter AU Normalizationb\nmiR ID\np ValueBest Permuted p Valuea\nmiR.1860.00610.0060miR.5000.00730.0025aBest p-value obtained for 20 randomly permuted seeds derived from the original miR seed.bAfter applying AU normalization to the dataset none of the miRs passed the statistical significance threshold (0.0003, which corresponds to 0.05 after Bonferroni correction for multiple testing). In order to compare the results with the original data (without AU normalization), we listed the top two miRs even though they did not pass the threshold.As shown, the AU response bias causes many false positive calls in computational search for active miRs from expression data, and therefore its removal is crucial when carrying out integrated bioinformatics analysis of mRNA expression data and 3′-UTR sequences. To remove this bias, we adopted the lowess normalization method which is routinely used to remove intensity biases from microarray data [21], and adjusted it to cancel AU biases (Figure 5) (see Methods). Applying AU normalization did not distort the normalization at the M-A plane (Figure S6). Importantly, after applying AU normalization to the negative control dataset, no miR family passed the statistical significance threshold (0.0003, which corresponds to 0.05 after Bonferroni correction for multiple testing) (Table 1).10.1371/journal.pcbi.1000189.g005Figure 5AU normalization.M-AU plots without (A) and after (B) applying an AU normalization scheme to the technical dataset which profiled the universal reference RNA pool.We next searched for an expression dataset that would serve as a positive test case; that is, a dataset that contains known miR signals. We preferred physiologically relevant datasets over ones that over-expressed miRs, which often give expression levels that are far above physiological ones. (Statistical searches for active miRs applied to several datasets that profiled cells over-expressing specific miRs readily detected the correct signals both without and after AU normalization (data not shown).) A recent study that compared expression profiles between stimulated T-cells derived from miR-155 deficient and control mice met this requirement [23]. As in many other datasets, we observed a strong AU bias in this dataset too, and removed it using the AU normalization (Figure 6). Without AU normalization, the statistical tests identified eleven significant miR families; the true hit (miR-155) was the third most significant one (Table 2). (Note that five out of the six most significant miRs falsely identified on the negative dataset were detected also in this positive dataset (compare Tables 1 and 2)). Here too, permutation tests found, in most cases, random seeds whose significance scores were similar to the ones obtained by the original seeds (Table 2). In sharp contrast, after AU normalization, only the true miR (miR-155) was detected and its statistical significance was substantially improved (Table 2). Importantly, none of the permuted seeds derived from the seed of miR-155 obtained a statistically significant score.10.1371/journal.pcbi.1000189.g006Figure 6AU bias in the miR-155 dataset.Relationship between 3′-UTR AU content and gene response in the dataset that compared gene expression profiles between miR-155-deficient and control Th2 cells. (A) Without AU normalization. (B) After applying AU normalization to the dataset. Plots were generated as described in the legend to Figure 1.10.1371/journal.pcbi.1000189.t002Table 2Active miRs identified in the miR-155 dataset.Without AU NormalizationmiR ID\np Valuea\nBest Permuted p Valueb\nmiR.496−1.56*10−10\n−1.13*10−7\nmiR.186−1.43*10−08\n−2.29*10−9\n\nmiR.155\n\n7.07*10−08\n−3.98*10−6\nmiR.26−2.15*10−06\n−1.41*10−7\nmiR.543−2.33*10−06\n−6.04*10−6\nmiR.25.32.92.363.367−6.54*10−06\n−3.71*10−7\nmiR.381−1.09*10−05\n−9.99*10−8\nmiR.329−1.98*10−05\n−1.31*10−3\nmiR.3312.48*10−05\n2.19*10−1\nmiR.493.5p−3.98*10−05\n−1.46*10−10\nmiR.495−7.41*10−05\n−9.60*10−8\nAfter AU Normalizationa\nmiR ID\np ValueBest Permuted p Valueb\n\nmiR.155\n\n1.20*10−12\n−0.023miR.142.5p0.00083−0.024aThe sign of the p-value marks the direction of the response of the miR target set: positive and negative p-values correspond to miRs whose target sets are significantly up- and down-regulated in miR-deficient Th2 cells, respectively, compared to wild type Th2 cells. The results obtained for the true signal in this dataset—miR-155—are emphasized in bold-italic font, and are in the expected direction: that is, the set of miR-155 predicted target genes is up-regulated in miR-155 deficient Th2 cells compared to control Th2 cells.bBest p-value obtained for 20 randomly permuted seeds derived from the original one.For a more challenging test case we used a dataset that monitored gene expression profiles in five distinct human T cells sub-populations representing five phases of T cell differentiation [24]: intrathymic T progenitor (ITTP) cells, double positive (DP) thymocytes, CD4 single positive (SP4), naïve CD4 T cells from cord blood (CB4), and naïve CD4 T cells from adult blood (AB4). To obtain fold-change measures, we divided the expression level at each development phase by the one measured in the mature AB4 T cells. Without AU normalization, the statistical tests identified six significant miR families: the target sets of three were down-regulated in ITTP cells, and the target sets of the other three were up-regulated in the SP4 cells (Table 3). After applying the AU normalization to the data, only the three miR-families whose target sets were repressed in ITTP (miR-17.5p, miR-19 and miR-181 families) remained significant (Table 3), suggesting that members of these three miR families are active in early phases of T cell development and become inactive as T cells mature. There is evidence that all three miR families detected by the statistical analysis play a role in thymocyte maturation and therefore are true hits. Li et al. recently [25] showed that miR-181a is highly expressed in immature T cells and that its expression level goes down as T cells proceed through differentiation. That study further showed that miR-181a plays a critical role in augmenting T cell sensitivity, a propensity that is vital to the elimination of self-reacting T cells early during maturation. Regarding miR-17.5p and miR-19 families, Landais et al. recently reported that the miR-106-363 cluster is over-expressed in 46% of human T-cell leukemias tested [26]. The miR-106-363 cluster is homolog to the miR-17-92 cluster, and miR-19 is contained in both clusters but carries a seed which is different from the one of the other miRs in these two clusters. It is possible that up-regulation of members of the miR-106-363 and miR-17-92 clusters in T-cell leukemia endows these cells with propensities normal to immature T-cells, most probably enhanced proliferation capacity. The identification of true hits on this dataset further demonstrates that computational analysis can accurately dissect active miRs from gene expression data probing cells under physiological conditions. Our statistical analysis utilizes target prediction based on miR seed signatures and therefore cannot distinguish between miRs sharing seed sequences. Empirical biological testing is required to pinpoint which members of the miR-17-92 and miR-106-363 clusters that carry a common seed sequence are actually active during T cell maturation.10.1371/journal.pcbi.1000189.t003Table 3Active miRs identified in the thymocyte maturation dataset.Original DatasetmiR IDITTPDP4SP4CB4miR.17.5p.20.93.mr.106.519.d−1.45*10−9\n−0.0055−0.12−0.75miR.19−7.04*10−8\n−0.0085−0.52−0.040miR.1010.760.00591.27*10−6\n−0.97miR.1440.840.00481.48*10−6\n0.48miR.3810.400.00643.81*10−5\n0.77miR.181−5.30*10−5\n−0.6490.12−0.35After Applying AU NormalizationmiR IDITTPa\n#\nDP4SP4CB4miR.17.5p.20.93.mr.106.519.d−1.04*10−9 (−4.27*10−3)−0.0016−0.0530.99miR.19−2.72*10−8 (−1.24*10−2)−0.0010−0.20−0.08miR.181−3.19*10−5 (−7.07*10−4)−0.200.43−0.67aIn parentheses, the best p-value in 20 random seed permutations.DiscussionIn the course of this study we observed in many gene expression datasets a striking association between gene response and 3′-UTR base composition. The high prevalence of such a relationship in microarray datasets, its exceptional statistical strength, and its detection in technical comparisons between replicate arrays, point unequivocally to a major bias in microarray measurements that was heretofore missed. Such a major AU bias in microarray measurements might have gone undetected because gene expression data are commonly analyzed in association with promoter, rather than 3′-UTR sequences, in attempts to unravel cis-regulatory promoter elements that control gene transcription. Only recently, with the emergence of miRs and RNA-binding proteins as key post-transcriptional regulators of gene expression, has gene expression analysis been coupled with analysis of 3′-UTR sequences. Indeed, it was the search for active miRs that motivated us to integrate gene expression and 3′-UTR sequence data, and led us to the detection of the AU response bias in microarray data.We demonstrated that this bias is distinct from the well-documented intensity-response and AU intensity biases, and that it originates from a systematic association between probe base composition and response. Using the new generation Affymetrix chips that contain probes selected throughout the transcripts, we uncoupled the sequences of probes and target 3-UTRs. We show that probes exhibit similar AU response bias irrespective of their location in the target transcripts. Therefore, the major link between gene response and 3′-UTR base composition that we observed in vast microarray datasets, is secondary to the general probe AU response bias, and simply reflects the fact that chip probes were selected from 3′-UTRs. A reasonable explanation to the AU response bias is that there are subtle differences in hybridization conditions for different arrays in a dataset, and that the effect of such differences is dependent on probe base composition. Further technical examinations are required to test this point.Bioinformatics analysis that integrates gene expression data and 3′-UTR sequences holds promise for systematic dissection of regulatory networks controlled by miRs. However, we demonstrated that the AU response bias causes many false positive calls in such analysis. Permutation tests were highly effective in revealing such false positive hits. Removal of this bias is of crucial importance when aiming to uncover miR-signatures as well as other cis-regulatory elements embedded in 3′-UTRs from mRNA expression profiles. We therefore developed visualization and normalization schemes for the detection and removal of AU biases, and demonstrated that their application to microarray data significantly enhances the computational identification of active miRs. In the case of Affymetrix chips, the normalization scheme that we implemented works at the probe-set or transcript level, and corrects the AU bias in a post-processing step (i.e., ran after probe intensity levels were calculated). A normalization scheme that takes into account the AU response bias at the phase of probe intensity calculation (similar to gcrma, which cancels AU intensity biases) is still required.Our results further substantiate that mRNA expression data contain ample information that allows, after proper removal of AU biases, in silico detection of active miRs. Importantly, this is also true when mRNA profiles were measured under physiological conditions. In view of the importance of elucidating regulatory roles played by miRs in various biological networks, we anticipate that the methods introduced in this study for detection, visualization and removal of the AU response bias from microarray data will be in wide use by the research community.MethodsAll statistical analyses were performed and plots were generated using the R package (http://www.r-project.org/).Data Analysis of Gene Expression DatasetsIn this study, we analyzed four microarray datasets which used 3′-UTR Affymetrix oligonucleotide chips (that is, chips in which probes are selected from targets' 3-UTRs), and one dataset that used the new generation Affymetrix Human Gene 1.0 ST Array, in which probes are located throughout the target transcripts. Raw data files (CEL files) were downloaded from GEO (http://www.ncbi.nlm.nih.gov/geo/) or ArrayExpress (http://www.ebi.ac.uk/microarray-as/aer/#ae-main0) DBs, or obtained directly from the authors of the data.Analysis of datasets that used 3′-UTR Affymetrix chipsThe dataset that profiled HPC multi-lineage differentiation [19] used Affymetrix MGU74Av2 mouse chips. Expression levels were recorded in triplicates at 0, 4, 8, 16, 24, 48, 72, and 168 hrs of differentiation into four lineages: megakaryocytes, neutrophils, erythrocytes and macrophages. The dataset that profiled Stratagene's universal human reference RNA pool in two independent chips ([20], GSE1158) used Affymetrix HGU133A human chips. The dataset that profiled expression levels in miR155-deficient and control T cells ([23], E-TABM-232), used Affymetrix MG-430.2 mouse chips. Expression levels were measured in 5 replicates in miR155-deficient and wild-type Th1 and Th2 cells stimulated for 24 hrs with LPS and IL4. The results reported in our study were derived from the Th2 dataset. The dataset that profiled expression level during T cell maturation ([24], GSE1460), used Affymetrix HGU133A-B human chips. Expression levels were recorded in triplicates in 5 phases during differentiation (intrathymic T progenitor (ITTP) cells, double positive (DP) thymocytes, CD4 single positive (SP4), naïve CD4 T cells from cord blood (CB4), and naïve CD4 T cells from adult blood (AB4).All these four datasets were processed by a similar scheme: First, probeset expression levels were calculated using the rma, gcrma, and mas5 methods implemented in the affy [27] and gcrma packages of the BioConductor project [28]. Unless otherwise stated, results reported in this paper are the ones obtained using the rma method. Similar results were obtained for data processed by the mas5 and gcrma methods. Second, probeset presence flags were calculated using the mas5calls function implemented in the affy package, and probesets that got more ‘Absent’ calls than a certain threshold were removed from subsequent analysis. (Thresholds for the number of ‘Absent’ calls were: 18 (out of 30 chips) in the HPC differentiation into megakaryocytes dataset; one (out of 2 chips) in the universal RNA pool dataset; 3 (out of 10 chips) in the miR-155 dataset; and 10 (out of 18 chips) in the T cell maturation dataset.) Next, probesets were mapped to their corresponding genes using annotation files provided by Affymetrix, and in cases where a gene was represented by several probesets, we used the measurements of the probeset with the highest median intensity level. Intensity levels over replicate chips were averaged.Analysis of the dataset that used the Affymetrix Human Gene 1.0 ST ArrayCEL files of this dataset ([22], GSE9819) were downloaded from GEO, and probe-set expression values were calculated using rma. In this dataset, we detected significant 3′-UTR bias in a comparison between two chips hybridized with a common Ambion Human Brain Reference RNA pool (sample ids GSM247680 and GSM247680). Probe-level intensities were extracted using the pm function implemented by the affy package. Probes' sequences and genome coordinates were obtained from chip annotation files provided by Affymetrix. Genome coordination of 5′-UTR, CDS and 3′-UTR regions of all annotated human transcripts were extracted from Ensembl using BioMart utilities [29]. Mapping of probes to 5′-UTR, CDS, and 3′-UTR regions was done by a Perl script written for this purpose. Before generating the probe-level M-AU plot, we performed the following preprocessing steps: a floor cut-off signal, which was set to the first quartile signal, was applied to each chip; probe expression levels were quantile- normalized; and probes whose signal was above median level were flagged as ‘Present’. Log2 of fold-change and AU content were calculated for each probe. To reduce noise, M-AU plot included only probes that were ‘Present’ in at least one of the chips hybridized with the brain reference sample.3′-UTR Sequences and miR Target Prediction3′-UTR sequences and miR target prediction for human and mouse were downloaded from TargetScanS (http://www.targetscan.org/; version 4.0; July 2007). TargetScanS predicts gene targets of miRNAs by searching 3′-UTRs for the presence of conserved 8-mer and 7-mer sites that match the seed region of each miRNA family [2]. In case a gene has several annotated 3′-UTRs, the longest one is considered.Target prediction for randomly permuted miR seedsFor each conserved miR family, as defined by TargetscanS, we generated 20 randomly permuted seeds derived from the original seed. Targets of these random seeds were predicted by the same program used by TargetScanS for prediction of targets of the original miRs (the program is available at TargetScanS website).AU normalizationAdopting the concepts of MA-plots and intensity-dependent normalization that were introduced by Yang et al. [21] in order to remove intensity biases from microarray data, we used the robust scatter plot smoother ‘lowess’, implemented in R (with default parameters), to remove the AU bias:where I1 and I2 are the intensity signals measured for a gene in chip1 and chip2, and c(AU) is the lowess fit to the M-AU plot (in which the X-axis represents either transcript 3′-UTR, probe-set, or probe's AU content). Applying 3′-UTR-based or probe-set-based AU normalization to the 3′-UTR Affymetrix datasets yielded similar results, as expected, because of the coupling between transcript 3′-UTR and probe-set sequences in these chips.Statistical search for candidate active miRs in mRNA expression datasetSearching for miRs that are active in a microarray dataset, we utilized miR target prediction produced by TargetScanS, and applied the following statistical test: for each miR family and for each condition in a dataset, we tested whether the set of predicted miR target genes is significantly more induced/repressed than the background set consisting of all the non-target genes (for which 3′-UTR sequence and expression data are available). Target and background sets were compared using the non-parametric Wilcoxon test, and a miR family was putatively considered ‘active’ in a certain condition if the p-value obtained for its target set was below 0.05 after applying Bonferroni correction for multiple testing (∼150 miR families were tested).Supporting InformationFigure S1Relationship between 3′-UTR AU content and gene response during HPC differentiation. The plot was generated as described in the legend to Figure 1 and shows the relationship between 3′-UTR AU content and gene response at three time points (4, 8, and 16 h) during HPC differentiation into three lineages (erythrocytes (E), monocytes (M), and neutrophils (N)).(0.13 MB TIF)Click here for additional data file.Figure S2AU bias in microarray data is not specific to a particular preprocessing method. The major AU bias in the dataset that profiled the universal reference RNA pool is not specific to a particular preprocessing method as it existed in data derived using different preprocessing and normalization schemes: rma, gcrma, and mas5.(0.12 MB TIF)Click here for additional data file.Figure S3No preference for A or U in the AU bias. The figure shows the relationship between gene fold-change in the technical dataset and: 3′-UTR AU content, 3′-UTR length, and 3′-UTR single base contents. The figure was generated as described in the legend to Figure 1 (p values indicated above each plot are for the comparison between the top 5% and bottom 5% genes). In this dataset, there is no preference for A or U in the relationship between 3′-UTR AU content and gene response. No major relationship between 3′-UTR length and gene response was observed here.(0.13 MB TIF)Click here for additional data file.Figure S4The AU response bias exists over large range of intensities. To test whether the AU-response bias is confined to probes with low intensities (which are inherently noisier), we redrew the M-A plot in Figure 3A, and colored each point according to the AU content of the corresponding probe (probes were divided into three groups: High, Medium and Low AU content probes; each group contained one third of the probes included in the analysis). The AU response bias is not associated with low intensity but exists over a large range of intensities.(0.33 MB TIF)Click here for additional data file.Figure S5AU bias using probe-set AU content. M-AU plot in which the X-axis represents probe-set AU content (in contrast to transcript 3′-UTR AU content shown in Figure 3B).(0.13 MB TIF)Click here for additional data file.Figure S6AU normalization does not distort the normalization at the M-A plane. This figure presents the M-A plot after applying AU normalization. While this normalization cancels the major bias detected at the M-AU plane, it has only subtle effect on the M-A plane.(0.18 MB TIF)Click here for additional data file.\n\nREFERENCES:\n1. KimVNNamJW\n2006\nGenomics of microRNA.\nTrends Genet\n22\n165\n173\n16446010\n2. LewisBPBurgeCBBartelDP\n2005\nConserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets.\nCell\n120\n15\n20\n15652477\n3. BrenneckeJStarkARussellRBCohenSM\n2005\nPrinciples of microRNA-target recognition.\nPLoS Biol\n3\ne85\ndoi:10.1371/journal.pbio.0030085\n15723116\n4. RajewskyN\n2006\nMicroRNA target predictions in animals.\nNat Genet\n38\nS8\nS13\n16736023\n5. ElkonRLinhartCSharanRShamirRShilohY\n2003\nGenome-wide in silico identification of transcriptional regulators controlling the cell cycle in human cells.\nGenome Res\n13\n773\n780\n12727897\n6. PilpelYSudarsanamPChurchGM\n2001\nIdentifying regulatory networks by combinatorial analysis of promoter elements.\nNat Genet\n29\n153\n159\n11547334\n7. SegalEYelenskyRKollerD\n2003\nGenome-wide discovery of transcriptional modules from DNA sequence and gene expression.\nBioinformatics\n19\ni273\ni282\n12855470\n8. CheadleCFanJCho-ChungYSWernerTRayJ\n2005\nStability regulation of mRNA and the control of gene expression.\nAnn N Y Acad Sci\n1058\n196\n204\n16394137\n9. FarhKKGrimsonAJanCLewisBPJohnstonWK\n2005\nThe widespread impact of mammalian microRNAs on mRNA repression and evolution.\nScience\n310\n1817\n1821\n16308420\n10. LimLPLauNCGarrett-EngelePGrimsonASchelterJM\n2005\nMicroarray analysis shows that some microRNAs downregulate large numbers of target mRNAs.\nNature\n433\n769\n773\n15685193\n11. LiuTPapagiannakopoulosTPuskarKQiSSantiagoF\n2007\nDetection of a microRNA signal in an in vivo expression set of mRNAs.\nPLoS ONE\n2\ne804\ndoi:10.1371/journal.pone.0000804\n17726534\n12. TsangJZhuJvan OudenaardenA\n2007\nMicroRNA-mediated feedback and feedforward loops are recurrent network motifs in mammals.\nMol Cell\n26\n753\n767\n17560377\n13. HuangJCBabakTCorsonTWChuaGKhanS\n2007\nUsing expression profiling data to identify human microRNA targets.\nNat Methods\n4\n1045\n1049\n18026111\n14. SoodPKrekAZavolanMMacinoGRajewskyN\n2006\nCell-type-specific signatures of microRNAs on target mRNA expression.\nProc Natl Acad Sci U S A\n103\n2746\n2751\n16477010\n15. ChenCZLiLLodishHFBartelDP\n2004\nMicroRNAs modulate hematopoietic lineage differentiation.\nScience\n303\n83\n86\n14657504\n16. FelliNFontanaLPelosiEBottaRBonciD\n2005\nMicroRNAs 221 and 222 inhibit normal erythropoiesis and erythroleukemic cell growth via kit receptor down-modulation.\nProc Natl Acad Sci U S A\n102\n18081\n18086\n16330772\n17. GeorgantasRW3rdHildrethRMorisotSAlderJLiuCG\n2007\nCD34+ hematopoietic stem-progenitor cell microRNA expression and function: a circuit diagram of differentiation control.\nProc Natl Acad Sci U S A\n104\n2750\n2755\n17293455\n18. FaziFRosaAFaticaAGelmettiVDe MarchisML\n2005\nA minicircuitry comprised of microRNA-223 and transcription factors NFI-A and C/EBPalpha regulates human granulopoiesis.\nCell\n123\n819\n831\n16325577\n19. BrunoLHoffmannRMcBlaneFBrownJGuptaR\n2004\nMolecular signatures of self-renewal, differentiation, and lineage choice in multipotential hemopoietic progenitor cells in vitro.\nMol Cell Biol\n24\n741\n756\n14701746\n20. van RuissenFRuijterJMSchaafGJAsgharnegadLZwijnenburgDA\n2005\nEvaluation of the similarity of gene expression data estimated with SAGE and Affymetrix GeneChips.\nBMC Genomics\n6\n91\n15955238\n21. YangYHDudoitSLuuPLinDMPengV\n2002\nNormalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation.\nNucleic Acids Res\n30\ne15\n11842121\n22. PradervandSPaillussonAThomasJWeberJWirapatiP\n2008\nAffymetrix Whole-Transcript Human Gene 1.0 ST array is highly concordant with standard 3′ expression arrays.\nBiotechniques\n44\n759\n762\n18476829\n23. RodriguezAVigoritoEClareSWarrenMVCouttetP\n2007\nRequirement of bic/microRNA-155 for normal immune function.\nScience\n316\n608\n611\n17463290\n24. LeeMSHanspersKBarkerCSKornAPMcCuneJM\n2004\nGene expression profiles during human CD4+ T cell differentiation.\nInt Immunol\n16\n1109\n1124\n15210650\n25. LiQJChauJEbertPJSylvesterGMinH\n2007\nmiR-181a is an intrinsic modulator of T cell sensitivity and selection.\nCell\n129\n147\n161\n17382377\n26. LandaisSLandrySLegaultPRassartE\n2007\nOncogenic potential of the miR-106-363 cluster and its implication in human T-cell leukemia.\nCancer Res\n67\n5699\n5707\n17575136\n27. IrizarryRABolstadBMCollinFCopeLMHobbsB\n2003\nSummaries of Affymetrix GeneChip probe level data.\nNucleic Acids Res\n31\ne15\n12582260\n28. ReimersMCareyVJ\n2006\nBioconductor: an open source framework for bioinformatics and computational biology.\nMethods Enzymol\n411\n119\n134\n16939789\n29. FlicekPAkenBLBealKBallesterBCaccamoM\n2008\nEnsembl 2008.\nNucleic Acids Res\n36\nD707\nD714\n18000006"
+}

batch_3/PMC2533295.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533295",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533295\nAUTHORS: Valorie A Crooks, Allison Williams\n\nABSTRACT:\nBackgroundThe goal of Canada's Compassionate Care Benefit (CCB) is to enable family members and other loved ones who are employed to take a temporary secured leave to care for a terminally ill individual at end of life. Successful applicants of the CCB can receive up to 55% of their average insured earnings, up to a maximum of CDN$435 per week, over a six week period to provide care for a gravely ill family member at risk of death within a six month period, as evidenced by a medical certificate. The goal of this study is to evaluate the CCB from the perspective of family caregivers providing care to individuals at end of life. There are three specific research objectives. Meeting these objectives will address our study purpose which is to make policy-relevant recommendations informed by the needs of Canadian family caregivers and input from other key stakeholders who shape program uptake. Being the first study that will capture family caregivers' experiences and perceptions of the CCB and gather contextual data with front-line palliative care practitioners, employers, and human resources personnel, we will be in a unique position to provide policy solutions/recommendations that will address concerns raised by numerous individuals and organizations.MethodsWe will achieve the research goal and objectives through employing utilization-focused evaluation as our methodology, in-depth interviews and focus groups as our techniques of data collection, and constant comparative as our technique of data analysis. Three respondent groups will participate: (1) family caregivers who are providing or who have provided end of life care via phone interview; (2) front-line palliative care practitioners via phone interview; and (3) human resources personnel and employers via focus group. Each of these three groups has a stake in the successful administration of the CCB. A watching brief of policy documents, grey literature, media reports, and other relevant items will also be managed throughout data collection.DiscussionWe propose to conduct this study over a three year period beginning in October, 2006 and ending in October, 2009.\n\nBODY:\nBackgroundThe goal of Canada's Compassionate Care Benefit (CCB) is to enable family members and other loved ones who are employed to take a temporary secured leave to care for a terminally ill individual at end of life. It came into effect in January of 2004 through changes to the Employment Insurance Act and Canadian Labour Code. Its establishment was based on recommendations from a recent health care commission [1] as well as years of advocacy from the palliative care and caregiving communities. The CCB is administered through the federal Employment Insurance (EI) program as a 'special benefit'.Successful applicants of the CCB program can receive up to 55% of their average insured earnings, up to a maximum of CDN$435 per week, over a six week period to provide palliative/end-of-life (P/EoL) care for a family member or other loved one at risk of death within a six month period. In order to qualify, applicants must have worked a minimum of 600 insurable employment hours over the previous 52 week period. Applicants must also meet the designation of 'family member'1 and have access to a medical certificate from the gravely ill or dying individual's doctor, indicating that death is imminent (i.e., within a six month period), in order to be successful.The six weeks of income assistance afforded by the CCB can be taken at once, broken down into one week periods and spread out over six months, and/or be shared between family members. Successful applicants must first go through a two week unpaid waiting period before starting payments. Also, a successful applicant must determine on his or her own when to request that the payments begin, with the first payment to be made within 28 days of beginning the claim.The goal of this study is to evaluate the CCB from the perspective of family caregivers providing care to a terminally ill individual at end of life. Based on a successful pilot evaluation undertaken in 2005 involving interviews with 25 family caregivers [2-4], this study will use Patton's [5] comprehensive utilization-focused approach to evaluation. To this end, there are three specific research objectives:(1) to examine the usefulness of the CCB for family caregivers providing P/EoL care and determine those elements of the program that can be changed/refined to better support their needs;(2) to explore front-line palliative care practitioners' perceptions of the CCB, including the barriers and facilitators to use, and how they determine whether or not to recommend the CCB to family caregivers on a case-by-case basis; and(3) to investigate barriers and facilitators inherent in the organization of specific workplaces and within the labour market in general that shape uptake of the CCB from the perspective of employers and human resources personnel.Meeting these objectives will address our study purpose, which is to make policy-relevant recommendations informed by the needs of Canadian family caregivers and input from other key stakeholders who shape program uptake.ContextThe primary goal of P/EoL care is to improve the quality of life and quality of death for dying people and their families through the provision of excellent care. Confirmed by a growing body of research, family units are assuming the majority of costs and responsibilities associated with P/EoL caregiving in an increasingly rationalized Canadian health care system [6]. The responsibilities associated with P/EoL caregiving are often more considerable than what family members can manage, resulting in compromised emotional, mental, social, financial and physical well-being [6-9]. Although many family caregivers want to provide care for their loved ones at the P/EoL stage, work interference can result in significant stress and burden [10] and many are faced with no provisions for paid leave and a lack of job security when returning to work. Furthermore, MacBride-King [11] reports that 48% of Canadian family caregivers find it difficult to balance caregiving and workplace responsibilities and 42% experience a great deal of stress in trying to meet the dual demands. The burden placed on family members has been shown to be of concern to P/EoL patients. For example, those patients surveyed in the study of Singer et al. [12] identified relieving caregiver burden as one of five elements of quality P/EoL care. Cohen and Leis [13] have also identified the burden placed on family caregivers to be a primary determinant of the quality of life of patients receiving P/EoL care. At the same time, contemporary shifts in the provision of P/EoL care from institutional settings to those in the community are resulting in increased family caregiver burden [14-16] which may outpace caregivers' individual capacities to cope [9,17,18].While family caregivers ignoring their physical and mental health in order to provide P/EoL care is common [19], financial and workplace obligations are more difficult to disregard. For example, in addition to needing to pay for existing personal financial responsibilities during the caregiving period, often by maintaining involvement in paid labour, family members providing P/EoL care contribute, on average, CDN$6000 in unpaid caregiving during the final four weeks of life [20]. Grunfeld et al. [10] note that family caregivers caring for terminally ill patients not only experience depression and anxiety but also adverse work impacts, such as missing scheduled shifts, and typically financial burden, such as the cost of purchasing prescription drugs out-of-pocket. Thus, gaining access to financial support is a particular need of family caregivers providing P/EoL care [21] as such support minimizes financial stressors.Canada's response to the growing demand to provide job and income support to family caregivers providing P/EoL care has been to develop the CCB. The CCB is a health-related social program that falls under the purview of Human Resource & Skills Development Canada. The CCB program and the legislative changes that shape its administration have come about in an era of neo-liberally informed social policy creation in Canada. An important ideology that underscores the creation of social policies in such a political climate is that families and voluntary agencies, rather than local states, should bear the onus of responsibility for assisting persons in need [22,23]. Another element of policymaking in the current neo-liberal climate has been the focus on providing care in the community as opposed to in institutional settings [24]. A significant outcome of this has been an increased reliance on the voluntary sector and unpaid labour in meeting such care needs [25]. It is these types of changes in the role of the state in Canadian society that have informed the development and implementation of the CCB – a program that facilitates care being provided in community settings by family members and other loved ones.According to the Health Council of Canada, the CCB is a P/EoL initiative of international excellence [26]. At the same time, there has been a great deal of national criticism focused on the CCB program. Picard's [27] opinion piece in the national newspaper The Globe and Mail, for example, surmised the following:A social program that provides some modicum of financial relief is entirely appropriate, and much needed. But the current program is not passing muster. It is unduly bureaucratic, inflexible and heartless. In short, the compassionate care program is utterly lacking in compassion.Critics of the program have focused on a number of issues as reflecting this 'lack of compassion'; these include: the two week waiting period for payments [27] and the labour market participation requirements that cannot be met by family caregivers who have taken time off from work to provide long-term care [28], The Canadian Federation of Independent Businesses [29] has also identified concern about the amendment of the Canada Labour Code, expressing that regulating such leaves may negatively affect small businesses in particular due to their smaller workforces. Other concerns have been raised about the design and implementation of the CCB. A recent review of the CCB by The Health Council of Canada [26] has noted that one of the CCB's most significant issues is its problematically low uptake. Further, the Canadian Women's Health Network [30] has pointed out the gender-based disadvantage inherent in the program in that women are more likely to be ineligible for CCB income support because they are more likely to be stay-at-home parents and part-time workers who do not meet the CCB's eligibility criteria. Being the first study that will capture family caregivers' experiences and perceptions of the CCB and gather contextual data with front-line palliative care practitioners, employers, and human resources personnel, we will be in a unique position to provide policy solutions/recommendations that will address these and other concerns.Methods/designAs we are looking to gather lived experiences from both family caregivers and stakeholders who inform caregivers' experiences of the CCB, we propose to conduct an inductive study [31] using a qualitative approach. The method and techniques of data collection and analysis that we propose here are directly informed by the research objectives stated earlier. More specifically, we will achieve this goal through employing: Patton's utilization-focused evaluation as our methodology, in-depth interviews and focus groups as our techniques of data collection, and constant comparative as our technique of data analysis.The objective of utilization-focused evaluation is to inform program and policy improvement using research findings. According to Patton [[5] p.20]:Utilization-Focused Evaluation begins with the premise that evaluations should be judged by their utility and actual use; therefore, evaluators should facilitate the evaluation process and design any evaluation with careful consideration of how everything that is done, from beginning to end, will affect use. Nor is use an abstraction. Use concerns how real people in the real world apply evaluation findings and experience the evaluation process. Therefore, the focus in utilization-focused evaluation is on intended use by intended users. (emphasis in original)This method is appropriate for our study as its purpose is to create policy recommendations which will directly affect family caregivers based on input from family caregivers and those who inform their uptake of the CCB, this being a focus on 'intended use by intended users.' There are twelve broad tasks that shape the method of utilization-focused evaluation; they are to: (1) determine readiness for assessment; (2) assess the readiness of the evaluators; (3) recruit an evaluation taskforce (ETF); (4) consider the evaluation context; (5) identify intended users; (6) determine the evaluation focus; (7) design the evaluation techniques; (8) test data collection techniques; (9) collect data; (10) analyze data; (11) facilitate the use of the findings; and (12) assess the evaluation process. Having already created an ETF that has worked with the research team to interpret the findings of the pilot study [2-4] (i.e., interviews conducted with 25 family caregivers regarding the CCB) and inform the direction of this full evaluation, we have already completed steps one through eight and are ready to move ahead with completing a full evaluation of the CCB from the perspective of family caregivers.A foundational principle of the utilization-focused approach to evaluation is to have the research inform program improvement, not only through collecting relevant data but by increasing the ETF's commitment to employ the data for program improvement. The ETF members, all of whom were engaged in the pilot study, will continue in this role. As in the pilot evaluation, members of the ETF will work with the investigators to finalize data collection instruments, identify and recruit participants, interpret findings, and identify venues for knowledge transfer and translation. The work of the ETF will primarily be done via regular teleconferences, although communication via e-mail will take place between these meetings. The principal responsibility of the ETF will be to ensure that the evaluation produces policy-relevant findings that will: (1) have implications for family caregivers and their use of the CCB, and (2) be of use to key members of the P/EoL care policy community.Upon completion of the pilot evaluation it was determined by the investigators and ETF that data collection will need to take place with three specific respondent groups in this full evaluation: (1) family caregivers who are providing or who have provided P/EoL care; (2) front-line palliative care practitioners; and (3) human resources personnel and employers. Each of these three groups has a stake in the successful administration of the CCB. Further, it is our contention that collecting data with all of these groups is essential in order to gain the fullest contextual understanding of the barriers and facilitators that shape family caregivers' uptake of the CCB while best informing how to better meet their needs. Data collection with each of these groups will be undertaken in five provinces: British Columbia, Manitoba, Newfoundland, Ontario, and Quebec. These provinces were carefully selected by the team upon completion of the pilot evaluation to represent Canada's regional diversity.Respondent Group #1: Family caregivers who are providing or who have provided P/EoL careIn-depth interviews are known to yield rich, nuanced data [32]. Further, in-depth interviews conducted by phone are known to be cost effective and produce reliable data [33,34]. We propose to conduct in-depth phone interviews with three groups of family caregivers: (1) successful CCB applicants; (2) unsuccessful applicants; and (3) non-applicants (i.e., retired, self-employed, or unemployed). Upon completion of the pilot study it was determined that collecting data from each of these three groups was relevant to the overall goal of evaluating the CCB from the perspective of family caregivers. We plan on accessing family caregivers from a wide variety of populations, including those providing P/EoL care for loved ones with cancer, Alzheimer's disease, and end-stage cardiopulmonary disease, in order to capture the diversity of caregiving experiences.We will conduct interviews with five family caregivers from each group in each of the five provinces resulting in a total of 75 interviews (25 in each category, 15 from each province). The interviews will address: (a) to what extent family caregivers are satisfied with the CCB; (b) perceived strengths of the CCB; (c) recommendations for improving the CCB; (d) family caregivers' experiences of employers' responses to taking the CCB leave; and (e) the logistical elements of applying for and/or receiving the CCB. Interview guides were tested and refined during the pilot evaluation. We will also administer a demographic questionnaire which was used and refined during the pilot evaluation to capture standard information about the caregiving experience and the personal characteristics of the caregiver and care recipient.Respondent Group #2: Front-line palliative care practitionersThe pilot evaluation revealed that 20 of the 25 family caregiver interviewees had some degree of awareness of the CCB prior to participating in the study [2]. They had first learned of the CCB from a variety of sources, primarily from the media. We also found that there is a significant difference between being aware of the CCB's existence and having a working knowledge of both how the CCB is administered and its eligibility requirements. Many participants lacked this kind of knowledge. An important group of professionals who have the capacity to share this type of information with potential applicants are front-line palliative care practitioners. For the purpose of this study, we define this group as including clinicians (e.g., nurses, nurse practitioners, family doctors, palliative care specialists), social workers, bereavement councillors, and P/EoL coordinators/program managers. Upon completion of the pilot evaluation it was determined that in the full evaluation, data collection would need to take place with these key informants as they provide important contextual data for the evaluation of the CCB from the perspective of family caregivers. We also believe it is important to consult with these key informants around barriers and facilitators of program uptake as they were some of the strongest advocates for developing the CCB and getting its supporting legislation enacted; thus, they have a useful perspective to contribute.We propose to conduct phone interviews with 50 key informants, ten in each of the five provinces. The interviews will address the following: (a) perceptions of the CCB's usefulness and barriers/facilitators to access; (b) experiences of recommending the CCB to a client/client's family; (c) working knowledge of the CCB's administration and eligibility requirements; and (d) suggestions for improvement. The investigators will work together with members of the ETF in the first year of the project to develop an instrument for data collection that addresses these and related issues. The instrument will also be informed by preliminary findings with the family caregiver respondent group.Respondent Group #3: Human resources personnel and employersFocus groups are known to have many benefits [35,36], including that participants are given the opportunity to engage in discussion with others about a topic of mutual interest. We propose to conduct five focus groups with human resources personnel and employers, one in each of the five study provinces. For the purpose of this study, we consider human resources personnel to be those individuals who take care of payroll, labour management, and/or administering benefits within the company with which they are employed or who work for a human resources management firm. We consider employers to be those individuals who have the ultimate responsibility for managing employees, including hiring/firing and negotiating leaves, within a company they own, direct, and/or manage. Upon completion of the pilot evaluation, the research team and ETF determined that this was another important stakeholder group to target. More specifically, this group will provide us with important contextual information about the logistics of having an employee take leave through the CCB, as well as offering insights into the CCB's usefulness from a labour market perspective.The focus groups will be run with 7–10 participants in each. Thus, we expect to collect data with anywhere from 35 (7 respondents/focus group) to 50 (10 respondents/focus group) employers and human resources personnel. Topics to be covered in the focus groups include: (a) perceptions of the CCB's usefulness and barriers/facilitators to access; (b) experience with having an employee take the CCB; (c) working knowledge of the CCB's administration and eligibility requirements; (d) strategies for supporting employees who are providing P/EoL care while involved in paid labour; and (e) suggestions for improvement. The research team will work together with members of the ETF in the first year of the project to develop a focus group guide that addresses these topics. The instrument will also be informed by preliminary findings with the family caregiver respondent group.Watching briefA watching brief of policy documents, grey literature, media reports, and other relevant items will be managed throughout the period of data collection. These sources will be accessed through conducting frequent searches for updated sources in media and publication search engines. Furthermore, members of the ETF will contribute relevant documents such as newsletters and policy briefings generated by their respective offices and organizations. The purpose of conducting the watching brief is to keep up-to-date on issues of relevance to the CCB including legal appeals and policy changes. These secondary data will assist in tracking any changes to the CCB and, in so doing, shape the policy context and augment our analyses of the three primary datasets.Recruitment: Telephone interviewsWe will identify family caregiver participants in each of the three categories through recruitment strategies that were shown to have success in the pilot evaluation by engaging in both purposeful and snowball sampling. Our first step will be to disseminate calls for participants using the collective resources of the research team, ETF, and the Canadian Institutes of Health Research funded New Emerging Team in Family Caregiving for People at End of Life (NET). The targeted group will be individual caregivers who meet sampling criteria and to those service providers and organizations that have contact with our population of interest. In addition to recruiting through the NET website via a posted advertisement, strategic internet searches will also be undertaken to identify community-based organizations in each of the five provinces that provide services for our target population (e.g., local support groups, family caregiver networks). French and English advertisements will be circulated to these groups via e-mail. We will then place advertisements in provincial newspapers. We will also snowball out from other participants by asking them if they know of anyone else who might be interested in participating in the study. Our last tested strategy will be to send letters to the offices of Members of Parliament in the target provinces to make them aware of the study and ask them to post a recruitment advertisement in their local offices and to share information about the project with any constituents they know who meet our sampling criteria. We found this to be useful in the pilot project as it assisted in identifying those who had applied for the CCB, as some had shared their experiences and even complaints with the constituency office.We will identify members of the key informant group (i.e., front-line palliative care practitioners) using the extensive networks which exist in the research team, ETF, and the NET. As with the family caregiver group, we will circulate advertisements in French and English through our collective networks and to newsletters and listservs of relevant organizations. We will also contact directly key informants with whom there is already a working relationship established.Recruitment: Focus groupsThe identification of employers and human resources personnel to participate in the focus groups will be done through established linkages with relevant professional associations. Advertisements about the focus groups will be circulated in French and English through these associations. To minimize logistical arrangements in meeting with active members, we aim to conduct the five focus groups at the provincial human resources association meetings, likely before or after conference sessions on a day agreed upon by all participants. We will rent a room at the conference location in which to host the focus group. We will run four English-language focus groups and one French-language focus group.Informed ConsentBecause the interviews with family caregivers and front-line palliative care practitioners will take place by phone, verbal consent will be sought. After scheduling an interview, these participants will be mailed or e-mailed a detailed information letter that contains information on their rights as participants. At the start of the phone interview the interviewer will review these details and read a consent script in order to obtain verbal consent. At this point the interviewer will sign a consent form indicating that the script was read and verbal consent has been granted. The respondent will have received a copy of this form in the letter of information package to keep for his/her own records. Human resource personnel will similarly be mailed or e-mailed a detailed letter of information with details of the study, participant rights, and the focus group information. Because these groups will happen face-to-face, a signed consent form will be used. A confidentiality script will be read aloud at the start of the group reminding participants that discussion that happens in the group is to remain confidential. These procedures have been reviewed and approved by the Office of Research Ethics at Simon Fraser University (certificate #37980) and Research Ethics Board at McMaster University (certificate #2006172).Data Management & AnalysisAll interviews and focus groups will be audio taped. Data analysis will proceed with the verbatim transcription of all interviews and focus groups which will be imported into the qualitative data management program NVivo. NVivo has been selected for data management because it allows for collaboration between researchers at multiple sites as in the case of our research team. All researchers and trainees involved in the project will have some form of involvement in data analysis, whether to analyze the findings of a particular respondent group or to collaborate on a particular element of the analysis (e.g. interview themes/codes) to ensure investigator triangulation. Investigator triangulation of this nature will also assist us in enhancing the reliability of the findings [37].Analysis of all focus groups and interviews with participants and key informants will be guided by the constant comparative technique. While this technique was originally developed to be used with the grounded theory method as a way to engage in analysis while data collection is ongoing [38,39], it has usefully been adopted in other types of qualitative approaches [39]. In our study we will use this technique to analyze completed datasets (i.e., data collection and analysis will not be concurrent). It is an appropriate technique as it provides a way to move beyond description of qualitative data and toward explanation, in that comparing findings between groups and explaining differences will allow us to shape the most relevant policy recommendations. This analytic technique \"involves taking one piece of data (one interview, one statement, one theme) and comparing it with all others that may be similar or different in order to develop conceptualisations of the possible relations between various pieces of data\" (p. 69) [39]. Our constant comparative analysis will take place at three levels: (1) the intra-group, (2) the inter-group, and (3) the inter-topic. Undertaking multiple types of comparisons within the same project is a common component of this analytic technique [38].DiscussionWe propose to conduct this study over a three year period beginning in October, 2006 and ending in October, 2009. During the first year (October 2006–07) we will undertake data collection with family caregivers (n = 75) while working to identify potential participants from the other groups (key informants, human resource personnel, and employers). Upon completing collection of these data, we will undertake analysis of the dataset. In the first year (October 2006–07) we will also develop the instrument for collection of data with the key informants and the probes for the focus groups with human resources personnel and employers, both of which will be informed by the preliminary findings of the family caregiver dataset. From October, 2007 through to February, 2008 we will conduct data collection with the key informants, namely front-line palliative care practitioners (n = 50). Full analyses of these data will begin once all interviews have been conducted. We expect to collect focus group data from human resource personnel and employers during the second year (October 2007–08) by attending provincial association conferences that take place during this period (n = 5 focus groups). All data collection will be completed by October 2008 and all analysis will be completed by March 2009.In the final year of the study (October 2008–09), we will work collectively to interpret the data from which policy directions arise. Appropriate venues for dissemination will also be identified. We will also work at this time to assess the evaluation process which is, as described in the study details section, an important element of Patton's [5] utilization-focused evaluation, during the final year. Throughout the three years the watching brief will be updated regularly and will be used to inform the analysis and identification of significant findings. In addition to widely disseminating a full and summary report during the final year, findings will be presented at scholarly and policy conferences and manuscripts will be submitted to peer-reviewed national and international journals. Final research reports and summaries will be made available in both English and French and posted on the NET website . Further, members of the ETF will assist with disseminating research products, distributing them to their membership and other key stakeholders. The Canadian Hospice Palliative Care Association, of which all provincial palliative care associations are members, will play a particularly central role in this regard, advocating for changes to federal government officials in Ottawa. Numerous other organizations have been identified for report dissemination.AbbreviationsP/EoL: palliative/end-of-life; ETF: evaluation taskforce; NET: New Emerging Team in Family Caregiving for People at End of Life; CCB: Compassionate Care Benefit.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsVC and AW contributed equally to the design and writing of this protocol.Appendix 1Endnotes1 At the time this research protocol was written and submitted for funding there was a limited definition of 'family member' that included only immediate relatives (e.g., parents and children) and not siblings, aunts, uncles, cousins, and other relatives. Since funding was obtained the range of eligible caregivers was broadened extensively to include not only all family members, including foster parents and in-laws, but also any loved one deemed as family by the dying individual being cared for. We retain our use of the term 'family caregiver' here because all caregiver respondents for that part of the study were indeed members of either the nuclear or extended family of the dying individual.Pre-publication historyThe pre-publication history for this paper can be accessed here:Supplementary MaterialAdditional file 1Peer review reports. This file contains the external review reports collated by CIHR for this protocol.Click here for file\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2533318.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533318",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533318\nAUTHORS: Gabriella Ferrandina, Vanda Salutari, Marco Petrillo, Arnaldo Carbone, Giovanni Scambia\n\nABSTRACT:\nBackgroundVery little data about the conservative treatment of early stage glassy cell cervical cancer have been reported.Case presentationA 30-year old patient, nulligravida was admitted to the Gynecologic Oncology Unit of the Catholic University of Campobasso for irregular post-coital vaginal bleeding. The patients was staged as having FIGO stage IB1 (tumor diameter = 2 cm) squamous cervical cancer. After extensive counseling of the patient and her family, laparoscopic pelvic lymphadenectomy and cold knife conization were performed. The final diagnosis was FIGO Stage IB1 glassy cell carcinoma. Currently, after a follow-up of 38 months, she has no evidence of disease.ConclusionWe reported a case of early stage glassy cell cancer patient, who was conservatively treated by conization and laparoscopic pelvic lymphadenectomy.\n\nBODY:\nBackgroundOver the past decade, the treatment of cervical cancer has evolved registering a gradual abandonment of radical surgery in favor of more conservative approaches: this becomes even more relevant considering that approximately 15% of all cervical cancers, and 45% of surgically treated stage IB cervical cancers occur in women < 40 years of age [1]. These figures are expected to increase due to the widespread use of cervical cancer screening which results in overall younger age and an earlier stage of disease at diagnosis. In addition, more and more frequently women defer childbearing, so that an increasing number of women would be diagnosed cervical cancer before having started or completed their reproductive program. Among the uterus preserving techniques, radical vaginal trachelectomy (RVT) with laparoscopic pelvic lymphadenectomy [2] has gained acceptance over the years by the gynecologic oncology community due to the favorable results in terms of oncological and obstetrical outcome [3].Among the strict criteria employed in the selection of cases who can potentially be offered uterus preserving approaches, tumor histology per se seems not to be a relevant factor [4], with the exception of rare histological types such as adenosquamous, neuroendocrine tumors or glassy cell carcinomas which have been generally associated with a higher risk of recurrence [5,6], and considered a contraindication to conservative treatment [7,8]. In particular, glassy cell carcinomas first described by Glücksmann and Cherry [9] in the uterine cervix, are typically composed of malignant cells showing a moderate amount of cytoplasm with \"ground glass\" appearance, distinct cell membranes stained with eosin or periodic acid-Schiff, and large nuclei with prominent nucleoli. These tumors have been considered since the beginning as an uncommon variant of poorly differentiated adenosquamous carcinoma [9], endowed with resistance to radiation therapy and unfavorable prognosis [10].To our knowledge, only three cases of glassy cell carcinomas undergoing conservative treatment by laparoscopic pelvic lymphadenectomy and radical vaginal trachelectomy have been reported [11].Here, we report the case of a stage IB1 cervical glassy cell carcinoma patient, who was safely treated with cold knife conization plus laparoscopic pelvic lymphadenectomy.Case presentationA 30-year old patient, nulligravida was admitted in March 2005, to the Gynecologic Oncology Unit of the Catholic University of Campobasso, for irregular post-coital vaginal bleeding. Her medical history was unremarkable. Her gynecological history was negative with menarche at the age of 12 years, and regular menses until 6 months before the occurrence of the symptoms.Gynaecological examination revealed a normal size uterus, and no adnexal masses. A circumscribed, ulcerated lesion (maximum diameter = 2 cm) was documented in the posterior esocervix. Parametria and vagina appeared uninvolved. Colposcopy-guided biopsy and curettage of endocervical canal were performed revealing an invasive squamous cell cervical carcinoma with areas of poor differentiation. Transabdominal and transvaginal ultrasound examination documented the presence of a normal size uterus showing normal echogenicity with the exception of a vascularized hypoechogenic area (18 × 14 × 11 mm) located in the cervix.Staging evaluation including chest X-ray, total body CT scan, and pelvic magnetic resonance imaging (MRI) documented the presence of a tumor mass (maximum diameter = 2 cm) located in the uterine cervix, and no enlarged lymph nodes. Examination under anesthesia revealed an ulcerated lesion of maximum diameter of 2 cm, without vaginal and parametrial involvement. Squamous cell carcinoma antigen levels were negative. The patient was staged as having FIGO stage IB1 cervical cancer.After extensive counseling of the patient and her family, she opted for a conservative approach. Open laparoscopy was carried out: peritoneal washing and a careful inspection of the adnexae and intra abdominal organs was performed. Systematic pelvic lymphadenectomy was performed up to internal iliac lymph nodes, and they returned as negative at frozen section examination. Several biopsies of the vaginal walls were obtained; these were negative for disease on frozen section. A cold knife conization was performed, and frozen section analysis showed that the lateral and deep margins of the tissue specimen were uninvolved. The biopsy of the endocervical canal also resulted negative at frozen section.At definitive pathological examination, a nodular lesion of maximum diameter of 2.0 cm (width extension) located in the cone (height = 2 cm, width = 3 cm), was detected. Microscopic examination revealed a tumor composed of nests of large cells with large eosinophilic cytoplasm presenting a ground-glass appearance (Figure 1). Cell membranes were easily recognizable, and tumor nuclei appeared large, presenting prominent nucleoli, and also areas of abundant eosinophil infiltration were present. The tumor showed a stromal invasion of 8 mm out of 1.7 stromal thickness. The lateral and deep margins of the cone were uninvolved for at least 9 mm. All peritoneal biopsies, as well as pelvic lymph nodes (n = 18) were negative. No lymphovascular space involvement was observed. The final diagnosis was FIGO Stage IB1 poorly differentiated carcinoma with > 90% of the tumor represented by neoplastic cells with glassy cell features. A second pathologist, blinded to the first's impression confirmed the diagnosis. Given the rarity of this histological type and its prognostic features, therapeutic options including radical trachelectomy, hysterectomy, or adjuvant treatment were carefully discussed with the patient, who nevertheless decided to undergo only strict follow-up procedures. The patient was then followed with gynecological examination, pap smear, and colposcopy every 3 months for the first 2 years, and every 6 months thereafter, and was also requested to perform chest x-ray and pelvic MRI every year. Currently, after a follow up of 38 months, she has no evidence of disease.Figure 1Glassy cell carcinoma of the cervix: the undifferentiated, glassy cells display large nuclei with prominent nucleoli and granular cytoplasm. Areas of abundant eosinophils infiltration are present. (Hematoxylin & Eosin, magnification: 200×).Cervical stenosis was documented after 21 months since surgery, and was easily managed by cannulation of the cervical canal under anesthesia.DiscussionWe report a case of early stage glassy cell cancer in a patient, who was conservatively treated by conization and laparoscopic pelvic lymphadenectomy. Indeed, among the fertility preservation approaches to early stage cervical carcinoma, RVT has gained much attention because of the recognized oncologic efficacy and safety. Intra- and postoperative complications have been reported to be approximately 4% and 12% of cases, respectively [8], and even less radical procedures such as conization plus laparoscopic pelvic lymphadenectomy have been investigated in selected cases of stage IB1 squamous cell carcinoma < 2 cm diameter [7]. While the fertility preserving procedures are widely accepted for tumors with squamous histological type, and also adenocarcinomas, which per se should not be considered a contraindication to conservative treatment, some concerns have been raised for rare histological types such as adenosquamous, neuroendocrine or glassy cell carcinomas. In particular, conservatively treated neuroendocrine and adenosquamous tumors have been reported to carry out a very unfavorable prognosis [5,6]. On the other hand, very few data about early stage glassy cell cervical cancer have been reported: of 3 cases treated with laparoscopic pelvic lymphadenectomy and RVT, all were reported as having no evidence of disease at time of publication [11]. No case of early stage glassy cell carcinoma treated with conization plus laparoscopic pelvic lymphadenectomy has been reported until now.Despite the extensive counseling about the possibility to perform trachelectomy or adjuvant treatment after final diagnosis, our patient decided only to undergo strict follow-up procedures, and is currently without evidence of disease after 38 months since initial diagnosis.ConclusionWe report a case of an early stage glassy cell cervical carcinoma patient, who was successfully treated with conization and laparoscopic pelvic lymphadenectomy. Given the rarity of this tumor histological type, and the paucity of data about its natural history, which has been reported to be similar to other histological types only with the employment of multimodal treatment strategies [12], caution should be taken to i) carefully evaluate the patients' fertility potential; ii) extensively counsel the patients about the risk/benefit of a conservative treatment; iii) investigate the patients' compliance to undergo strict follow-up procedures.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsGF conceived of the study, participated in its design and drafting. VS participated in the design of the study and collected the clinical data. MP participated in the design of the study and collected the clinical data. AC carried out the histopathological evaluation. GS conceived of the study, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.ConsentWritten informed consent was obtained from the patient for publication of this case report and any accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal.\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2533335.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533335",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533335\nAUTHORS: John Yeh, Beom Su Kim, Larry Gaines, Jennifer Peresie, Carly Page, Armando Arroyo\n\nABSTRACT:\nBackgroundAim of this study was to test the hypothesis that levels of hyperpolarization activated cyclic nucleotide gated channels 1 to 4 (HCN1-4) are linked to the reproductive age of the ovary.MethodsYoung, adult, and reproductively aged ovaries were collected from Sprague-Dawley rats. RT-PCR and western blot analysis of ovaries was performed to investigate the presence of mRNA and total protein for HCN1-4. Immunohistochemistry with semiquantitative H score analysis was performed using whole ovarian histologic sections.ResultsRT-PCR analysis showed the presence of mRNA for HCN1-4. Western blot analysis revealed HCN1-3 proteins in all ages of ovarian tissues. Immunohistochemistry with H score analysis demonstrated distinct age-related changes in patterns of HCN1-3 in the oocytes, granulosa cells, theca cells, and corpora lutea. HCN4 was present only in the oocytes, with declining levels during the reproduction lifespan.ConclusionThe evidence presented here demonstrates cell-type and developmental age patterns of HCN1-4 channel expression in rat ovaries. Based on this, we hypothesize that HCN channels have functional significance in rat ovaries and may have changing roles in reproductive aging.\n\nBODY:\nBackgroundMolecular studies of ovarian granulosa cells have determined that the granulosa cells of various species express potassium, calcium, sodium, and chloride channels. These channels have electrical activity and generate action potentials. Porcine granulosa cells express a potassium current (IA), a delayed rectifier K+ current (IK) and Ca2+ currents [1,2]. Ion channels such as Kv1.1, Kv1.2, Kv1.3, Kv1.4, Kv1.5, Kv1.6, KCNQ1, KCNE1 have been identified in porcine granulosa cells [2]. Kir6.1 and Kv4.2 are expressed in human granulosa cells [3,4]. Ca2+ subunits Cav1.2 and Cav3.2 are expressed in human granulosa cells and calcium type currents are also found in human granulosa cells [5,6]. Human granulosa cells express a Ca2+ activated K+ current (BKCa), a transient outward K+ current and an ATP-sensitive potassium channel [3,4,7]. In hen granulosa cells, chloride channels are activated by cAMP during LH-stimulated progesterone production [8].During aging, potassium, calcium and sodium channels activities and levels are altered. For the potassium and calcium channels, the channels in the cells in the brain, heart, liver, and pancreas all change during the process of aging [9-12]. Cumulatively, these changes include a decrease in the total number of ion channels present and alterations in the distribution and activity of the channels. For the sodium channels, the changes associated with developmental aging in retinal ganglion cells, myocardium and in kidney epithelium cells include shifts in the number and alterations in conduction activity [13-15]. These reports suggest that there are specific age-related patterns in the expression and physiological activity of ion channels.Hyperpolarization activated cyclic nucleotide gated (HCN) channels generate a pacemaker current (Ih) that controls spontaneous pacemaker activity in the heart and brain [16-19]. There are four members of the HCN gene family and they belong to the voltage-gated K+ superfamily. The four forms of HCN genes (HCN1-4) have highly conserved core transmembrane and cyclic nucleotide binding regions, with each of the four proteins having a six transmembrane region. The four HCN genes have different distributions in the heart and brain, suggesting that they have different functions. HCN channels have been in found in neurosecretory neurons of the hypothalamus, retinal rod photoreceptors, hair cells of the auditory system, olfactory neurons, spinal cord dorsal root ganglion neurons, and the enteric nervous system [16-25]. The wide distribution of the HCN channels suggests that they have roles in a number of different physiological conditions. In addition to the wide distribution of these channels, it has been previously reported that HCN4 expression in the hippocampus is related to developmental age, suggesting that these channels also have aging-related changes [23,24].To our knowledge, no prior studies have investigated the HCN channels in the ovary. Given the important roles of HCN in other organs and given the aging-related changes found in potassium, calcium, sodium and HCN channels, it was hypothesized that HCN channels play vital roles in the ovary and that alterations of their expression would be found during reproductive aging. In this study, we analyzed the expression and localization of HCN1-4 in the rat ovary to assess this postulate.MethodsAnimals and treatmentSprague-Dawley rats (Harlan, Indianapolis, IN) of three age groups were studied: 1.) \"young\", 26 days old, immature control females; 2.) \"adult\", 65–75 day old, adult control females and; 3.) \"reproductive aging\", 8–9 month old retired breeders, experimental females with declining fertility [26]. The animals were maintained under standard housing conditions with a 12 h:12 h light cycle. They were provided access to standard rat chow (Harlan, Indianapolis, IN) and water ad libitum. The animals were euthanized by an overdose of carbon dioxide. Subsequently, both ovaries were dissected out from each animal; one ovary was snap frozen and stored at -80°C while the other one was fixed in 10% formalin and stored at 4°C for paraffin sectioning. All procedures were approved by the Institutional Animal Care and Use Committee of the University at Buffalo (GYN07042N).RNA isolation and RT-PCRRat ovarian total RNA was isolated using Trizol (GibcoBRL, Life Technologies, Grand Island, NY). RT-PCR was performed as previously described by our laboratory [25], using a Promega Access RT-PCR kit (Access RT-PCR System, Promega, Madison, WI). PCR primers were designed to amplify rat HCN1-4 mRNA (Table 1) and were slightly modified from mouse primers used previously [25]. Positive control for HCN1-4 was brain RNA and the negative control was running the PCR reaction without the cDNA template. PCR conditions were as follows: 45°C for 45 min, 94°C for 2 min, and then 40 cycles of 94°C for 30 sec, 60°C for 1 min, 68°C for 2 min, and a final extension of one cycle at 68°C for 7 min. The analysis of the RT-PCR reaction products was by agarose gel electrophoresis.Table 1Primers for RT-PCR reactionsGeneGenBank Accession No.Forward primer (5'-3')Reverse primer (3'-5')Size of product (bp)HCN1NM_053375TTCATGCAGAGGCAGTTCACCACGGTGTTGTTGTTTGCTC248HCN2NM_053684CCATGCTGACAAAGCTCAAACGAGCTGAGATCATGCTGAA377HCN3NM_053685TCGGACACTTTCTTCCTGCTGGTTGAAGATGCGAACCACT364HCN4NM_021658GGGCTTCTCCTGTAGCCTTTTGAGCTTCAGGTCCTGTGTG219Western blottingTotal protein was isolated by procedures used previously [25,26]. In brief, rat tissues were lysed in RIPA buffer, containing 50 mM Tris-HCl, 150 mM NaCl, 0.1% sodium dodecyl sulfate (SDS), 1% NP-40, 1 mM phenylmethanesulfonyl fluoride (PMSF), and 0.5% N, N'-dicyclohexylcabodiimide (DCC) [25,26], along with protease inhibitors (Sigma, St. Louis, MO, 1:100). The protein concentrations were determined by the Bradford method (Bio-Rad, Hercules, CA). Fifty micrograms of protein from rat tissues under reducing conditions were loaded onto a 10% (HCN2 and HCN3) or an 8% (HCN1 and HCN4) Tris-Glycine SDS-polyacrylamide gel (Invitrogen, Carlsbad, CA). After electrophoresis, the proteins were electrically transferred to a nitrocellulose membrane (VWR International, West Chester, PA), blocked with 5% skim milk in TTBS (TBS with 0.1% Tween 20), and then incubated overnight at 4°C with rabbit polyclonal antibodies against HCN1-4 (anti-HCN 1, 2, 3 or 4; product # APC-056, APC-030, APC-057, APC-052, respectively; Alomone Labs Ltd., Jerusalem, Israel) [26] at a dilution of 1:200. Horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG (1:2500; Amersham Pharmacia, Piscataway, NJ) was used to identify the protein bands and they were amplified using SuperSignal West Pico Chemiluminescent Substrate Kit (Pierce Biotechnology, Rockford, IL). Visualization of the protein bands was by CL-Xposure film (Pierce Biotechnology).Immunohistochemistry and H score semiquantitative analysisAfter fixation, ovaries were embedded in paraffin and cut at 4 μm thick sections that were placed on Starfrost Adhesive positively charged microscope slides (Mercedes Medical, Germany) and the procedures used were as previously described [27]. Sections were deparaffinized using xylene and rehydrated using graded alcohol series. Sections were rinsed in distilled water then incubated for 30 minutes in 4N HCl at 37°C for antigen retrieval. Slides were cooled to room temperature, and then washed in PBS for 5 minutes. The slides were then placed in sodium borohydride diluted in PBS at a concentration of 1 mg/ml. The slides were then rinsed three times in PBS. The tissue sections were then blocked for 1 hour at room temperature using 5% goat serum and 5% BSA. The slides were transferred to a humidified chamber, and a rabbit polyclonal primary antibody against one of the HCN channels described in Western blotting. The HCN2 and HCN3 anitibodies were applied at dilutions of 1:200, and HCN1 and HCN4 were at dilutions of 1:50. As negative control, the primary antibodies were omitted. To confirm specificity of immunostaining, an additional negative control was performed for each of the channels. The anti-HCN1-4 antibodies were pre-incubated with the appropriate antibody control antigen as follows: HCN 1 6–24 Peptide, HCN 2 147–161 Peptide, HCN 3 727–744 Peptide, and HCN4-GST fusion protein (provided by Alomone labs). For the HCN1-3 preadsorption control solution, 1 μg of peptide was incubated with 1 μg of antibody and for HCN 4 preadsorption control solution, 3 μg of fusion protein was incubated with 1 μg of antibody for one hour at 37°C, centrifuged at 12,000 rpm, and then applied to the tissue sections. The slides were incubated with primary antibodies overnight at 4°C. The following morning, slides were washed 3 times in PBS for 5 minutes each. To visualize the primary antibody, Alexa Fluor goat anti-rabbit 594 (Molecular Probes, Eugene, OR) were applied to the tissue sections for 30 minutes at room temperature in the dark. To visualize HCN1, 2, and 3, the Alexa Fluor was applied at a dilution of 1:1000. To visualize HCN4, the secondary antibody was applied at a dilution of 1:500. Slides were air dried for 1 hour and cover-slipped using ProLong Antifade (Molecular Probes). The sections were then viewed on a Nikon Eclipse E400 fluorescent Microscope (Micro Video Instruments, Avon, MA) using the appropriate filters.A modified H score system was used to analyze the HCN channels 1 to 4 staining [27,28]. This scoring system was based on two criteria: the distribution of the staining, and the intensity of the staining. The following scale was used to determine the distribution of the stain in a structure: less than 50% of the structure stained was scored as 1, greater than or equal to 50% of the structure stained was scored as 2. To determine the intensity of the stain the following scale was used: no stain = 0, weak staining = 1, moderate staining = 2, and intense staining = 3. The numbers obtained for the distribution and intensity were then multiplied together for a combined score. A total score of 6 was the maximum for any structure. Two independent observers scored the sections for H score analysis. The results from each reviewer were compared and any discrepancy greater than 10% resulted in a reevaluation with both reviewers. Follicles were divided into classes based on the criteria of Oktay et al [29]. Primordial, primary, preantral, and antral follicles were included in the H score analysis, as well as thecal cells, oocytes, and corpora lutea.Statistical analysisThree separate rats in each age group were used for the RT-PCR and western blot experiments. Data for the ovarian immunofluorescence H score analysis are presented as mean +/- standard error and represent results from experiments repeated in triplicate. Statistical analyses of the ovarian H scores for HCN1-4 were performed using a one-way ANOVA followed by a linear contrast (SPSS version 11). P < 0.05 was considered statistically significant.ResultsGene expression and western blot analysis of HCN1-4HCN1-4 mRNA expression in the rat ovary was determined by RT-PCR (Figure 1A). For all four HCN channels, RT-PCR demonstrated the presence of these mRNA in the ovaries in all reproductive stages studied. Western blot analysis showed protein bands for HCN1-3 in rat ovaries of the three reproductive stages studied (Figure 1B). For HCN4, no protein was detected in the rat ovarian tissue by our whole ovarian extract western blot analysis. However, a 150 kDa protein band was found for rat brain and heart, two positive controls.Figure 1(A) RT-PCR analysis of ovarian HCN1-4 gene expression. (a) A 248 bp HCN1 RT-PCR band in all three developmental ages studied. (b) A 377 bp HCN2 RT-PCR band was evident in all three developmental ages studied. (c) A 364 bp HCN3 RT-PCR band was expressed in all three developmental stages studied. (d) A 219 HCN4 RT-PCR band was present in all three developmental stages studied. n = 3 animals studied per gene per reproductive stage. M = marker; B = brain; Y = young; A = adult; RA = reproductive aging; N = negative; H = heart. B. Western blot analysis of ovarian lysates for HCN1-4 protein expression of young, adult and reproductively aged rats. (a) Western blot analysis for HCN1 in ovaries. (b) Western blot analysis for HCN2 in ovaries. (c) Western blot analysis for HCN3 in ovaries. (d) Western blot analysis for HCN4 in ovaries. n = 3 animals studied per protein per reproductive stage.Immunohistochemistry of ovarian HCN1-4Figure 2 depicts ovarian follicles and the staining patterns found using immunofluorescence to analyze for the localization of HCN1-4 in young, adult and reproductively aged rat ovaries. By H score analysis (Figure 2), differences were detected in the spatial and temporal localization of HCN1-4 in the rat ovary. All four HCN proteins were detected in ovaries of all three reproductive ages. However, there were specific spatial protein differences in the distribution of HCN1-4. For HCN1, HCN2, and HCN3, the proteins were found in oocytes and in the granulosa cells of primary, preantral, and antral follicles. In addition, all three proteins were found in thecal cells. Furthermore, all three proteins were localized to corpora lutea. For HCN4 experimental tissue sections, HCN4 protein expression was localized only to the oocytes. In the HCN1-4 negative control experiments, both types of negative control experiments, the preabsorbed control experiments and the omission of the primary antibody experiments, were appropriately negative.Figure 2Immunofluorescence localization of HCN1-4 in young, adult and reproductively aged rat ovaries. Experimental and negative control studies are presented from consecutive ovarian sections for each HCN protein for each reproductive state. (A) Localization of HCN1-4 channels in the young ovary. (B) Localization of HCN1-4 channels in the adult ovary. (C) Localization of HCN1-4 channels in the reproductive aging ovary. n = 3–9 animals studied per protein per reproductive stage.In addition to the spatial findings above, there were age-related findings related to the reproductive age of the ovaries for the specific ovarian structures studied (Figure 3). For the following structures, there were differences in the H scores for the HCN proteins for the three reproductive ages studied: 1.) oocytes: HCN1 (decline in H score with increasing reproductive age; p < 0.05), HCN3 (decline in H score with increasing reproductive age; p < 0.01) and HCN4 (decline in H score with increasing reproductive age; p < 0.01); 2.) preantral follicle granulosa cells: HCN3 (decline in H score with increasing reproductive age; p < 0.05); 3.) primary follicle granulosa cells: HCN3 (decline in H score with increasing reproductive age; p < 0.01); 4.) thecal cells: HCN3 (decline in H score with increasing reproductive age; p < 0.01).Figure 3H Score analysis of HCN1-4 channels in the different ovarian structures throughout the aging process. (A) The H scores of HCN1 (a), HCN2 (b), HCN3 (c), and HCN4 (d) channels in the oocytes during the reproductive aging process. (B) H score analysis of granulosa cells in different follicle classes. (a-d) H scores of HCN1-4 channels in primary follicles. (e-h) Results of H score analysis of HCN1-4 channels in preantral follicle granulosa cells. (i-l) H scores for HCN1-4 channels in the granulosa cells of antral follicles. (C) H score analysis of HCN1 (a), HCN2 (b), HCN3 (c), and HCN4 (d) expression in theca cells. (D) H score analysis of HCN1-4 channels (a-d respectively) in the corpora lutea. N/A – not applicable, as the young rats do not yet have corpora lutea. n = 3–9 animals studied per protein per reproductive stage.DiscussionAll four types of HCN channels are expressed in the ovary as evidenced by RT-PCR, western blot, and immunohistochemical results presented in this report. To our knowledge, this is the first description of the changes in distribution of the HCN channels in ovarian structures in the reproductive life-cycle. These channels have specific patterns in the different ovarian cell types. HCN channels 1–3 are expressed in oocytes, granulosa cells of primary, preantral, and antral follicles, the thecal cells, and in luteal cells, while HCN4 is only expressed in oocytes. This suggests that different ovarian structures use different combinations of HCN channels for normal physiological function. Furthermore, HCN4 appears to be oocyte specific and, thus, this protein may be useful to define the physiological status of an oocyte.Ion channels are involved in ovarian steroidogenesis. Potassium channels mediate gonadotropin regulated progesterone secretion in human granulosa cells [3,4,7,30]. L- and T-type Ca2+ channels mediate hCG stimulated progesterone secretion in human granulosa cells [5]. Sodium channels down regulate progesterone production in primate granulosa cells [31]. Potassium channels and cAMP are involved in FSH-stimulated progesterone production in pig granulosa cells [32]. Given that HCN channels are located in all the cell types which are involved in steroidogenesis, the granulosa, theca and corpora lutea cells, it would not be unreasonable to hypothesize that the HCN channels also participate in this important ovarian activity. HCN channels have been identified in secretory cells including GnRH neurons, pancreatic β-cells, and pituitary lactotrophs [16-19,25]. Several studies have described membrane hyperpolarization in granulosa cells. Activation of BKCa channels resulting in membrane hyperpolarization is required for steroidogenesis in human luteinized granulosa cells [7]. Thus, membrane hyperpolarization may be a mechanism controlling steroid production in the granulosa cells of the ovary. In contrast to most voltage gated channels, HCN channels are activated by membrane hyperpolarization [16-19]. In granulosa cells it is possible that hyperpolarization of the cell membrane could activate HCN channels, thereby resulting in membrane depolarization. Depolarization could activate calcium and cAMP signaling, thus resulting in activation of steroidogenic enzymes and thereby increasing steroid production. This hypothesis would be supported by demonstration of functional HCN channels in granulosa cells.There are age related changes in potassium, calcium and sodium channel expression [9-15]. In the data presented in this report, HCN1 and HCN2 had minimal variation through the aging process, with HCN1 only exhibiting declining levels in the oocytes during reproductive aging. This suggests that the expression of these two channels remains relatively constant throughout granulosa and thecal cell reproductive aging and may have minor or unchanging roles in ovarian physiologic functions such as steroidogenesis or peptide hormone production. HCN3 has different expression patterns in the granulosa cells and theca cells during the aging process, indicating that there is an age-dependent expression of HCN3 in ovarian structures, suggesting that the changes in steroidogenesis during aging might be modulated through this protein. HCN3 expression decreases during aging in oocytes, granulosa cells of preantral follicles, and in theca cells, suggesting a possible function in the decrease of ovarian function in advancing reproductive age. The mechanism of this decline is not yet known and understanding of the mechanism could lead to further insights into the overall aging-related reduction of ovarian function. To date, the physiologic processes in reproductive aging are not yet fully understood. HCN channels may play a role in ovarian aging and, in addition, they could serve as immunohistochemical biomarkers for reproductive aging.HCN4 is an oocyte specific channel and may be an indicator of oocyte quality. There is a linear decrease in the expression of HCN4 throughout the reproductive aging process in the female rat. Brewster et al. and Surges et al. showed a steady decrease in HCN4 expression throughout the maturation process of the rat hippocampus [23,24]. In the ovary, growth differentiation factor 9 (GDF-9) has been found to be oocyte specific [33-35]. GDF-9 has been demonstrated to be essential to the growth and differentiation of early ovarian follicles. In cultured bovine granulosa cells, GDF-9 stimulated the proliferation of granulosa cells from small and large antral follicles and can disrupt the production of progesterone and estradiol [36]. The functions of HCN4 in the oocyte have not been determined to date, but it is possible that the channel may also be necessary for growth and differentiation of ovarian follicles or for steroid production.ConclusionIn conclusion, HCN1-4 channels are expressed in the ovary and there are differential expression patterns for the channels. HCN1-3 are expressed in ovarian structures including oocytes, granulosa and thecal cells, and in luteal cells, while HCN 4 is only expressed in the oocytes. There are decreases in the expression of HCN1, 3, and 4 in the oocytes during reproductive aging, with the decrease in HCN4 being the most pronounced. HCN3 expression are also decreased in the granulosa cells of preantral follicles and theca cells in reproductive aging. Future studies need to be conducted to determine the specific roles of HCN channels in the ovaries and the physiological reasons for the changes in the expression of the channels through the aging process.AbbreviationsHCN: hyperpolarization activated cyclic nucleotide gated channel; RT-PCR: reverse transcriptase polymerase chain reaction; mRNA: messenger ribonucleic acid; Kv1.1 – 1.6: potassium voltage-gated channel, shaker-related subfamily, members 1–6; KCNQ1: potassium voltage-gated channel, subfamily Q, member 1; KCNE1: potassium voltage-gated channel, Isk-related family, member 1; Kir6.1: potassium inwardly-rectifying channel, subfamily J, member 8; Kv4.2: potassium voltage-gated channel, Shal-related family, member 2; IA: potassium current; IK: delayed rectifier potassium current; Cav1.2: calcium channel, voltage-dependent, L type, alpha 1C subunit; Cav3.2: calcium channel, voltage-dependent, T type, alpha 1 H subunit; Ca2+: calcium; K+: potassium; BKca: calcium activated potassium current; ATP: adenosine tri-phosphate; cAMP:  cyclic adenosine mono-phosphate; LH: luteinizig hormone; Ih : pacemaker current, hyperpolarization-activated current, hyperpolarization-activated cation current; RNA: ribonucleic acid; cDNA: complimentary deoxyribonucleic acid; SDS: sodium dodecyl sulfate; PMSF: phenylmethylsulfonyl fluoride; DCC: dicyclohexlcabodiimide; TTBS: tris buffered saline with 0.1% Tween 20; HRP: horseradish peroxidase; HCl: hydrochloric acid; PBS: phosphate buffered saline; hCG: human chorionic gonadotropin; FSH: follicle stimulating hormone; GnRH: gonadotropin releasing hormone; β: betaCompeting interestsThe authors declare that they have no competing interests.Authors' contributionsJY conceived of the study along with AA, and participated in its design and coordination and helped to draft the manuscript. BSK is responsible for the Western blots and RT-PCR, data and statistical analysis, and manuscript preparation. LG carried out the immunohistochemistry and baseline research articles for the initial research. JP carried out the immunohistochemistry, baseline research articles, data and statistical analysis, and manuscript preparation. CP carried out the immunohistochemistry as well as data analysis. AA conceived of the study along with JY, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2533340.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533340",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533340\nAUTHORS: Ali H Zarzour, Mohie Selim, Alaa A Abd-Elsayed, Diaa A Hameed, Mohammad A AbdelAziz\n\nABSTRACT:\nBackgroundIn Egypt, where bilharziasis is endemic, bladder cancer is the commonest cancer in males and the 2nd in females; squamous cell carcinoma (SCC) is the commonest type found, with a peculiar mode of presentation. The aim of this study is to identify and rank the risk factors of muscle invasive bladder cancer (MIBC) in Upper Egypt and describe its specific criteria of presentation and histopathology.MethodsThis is an analytical, hospital based, case controlled study conducted in south Egypt cancer institute through comparing MIBC cases (n = 130) with age, sex and residence matched controls (n = 260) for the presence of risk factors of MIBC. Data was collected by personal interview using a well designed questionnaire. Patients' records were reviewed for histopathology and Radiologic findings.ResultsThe risk factors of MIBC were positive family history [Adjusted odds ratio (AOR) = 7.7], exposure to pesticides [AOR = 6.2], bladder stones [AOR = 5], consanguinity [AOR = 3.9], recurrent cystitis [AOR = 3.1], bilharziasis [odds ratio (OR) = 5.8] and smoking [OR = 5.3]. SCC represented 67.6% of cases with burning micturition being the presenting symptom in 73.8%.ConclusionMIBC in Upper Egypt is usually of the SCC type (although its percentage is decreasing), occurs at a younger age and presents with burning micturition rather than hematuria. Unlike the common belief, positive family history, parents' consanguinity, exposure to pesticides and chronic cystitis seem to play now more important roles than bilharziasis and smoking in the development of this disease in this area.\n\nBODY:\nBackgroundBladder cancer represents a global health problem. It ranks ninth in worldwide cancer incidence. It is the 4th commonest cancer in men and the 12th in women in the USA. It is estimated that about 67,160 Americans were diagnosed with bladder cancer in 2007 and 13,750 died of the disease [1].In Egypt, carcinoma of the bladder is the most prevalent cancer, accounting for as many as 31% of all cancer cases [2]. Currently, it ranks first in males representing 16.2% of male cancer [3]. The estimated incidence in males in rural areas in Egypt is about 32 per 100.000 [4].The exact etiology of bladder cancer is still unknown. Several risk factors have been accused as being involved in its pathogenesis such as cigarette smoking [5], synthetic nitrogen fertilizers [6], organophosphate-based pesticides [7], aromatic amines [8], pelvic irradiation, A cyclophosphamide, chronic cystitis, schistosomiasis [5], human papilloma virus [9], genetic predisposition, and some occupations [5]. The relative importance of such risk factors in the pathogenesis of the disease differs in different populations.The aim of this study is to identify and rank the risk factors of muscle invasive bladder cancer (MIBC) in Upper Egypt and to describe the peculiarities of the disease presentation and histopathology in this specific population.MethodsThis study is an analytical, hospital based, case control study comparing MIBC cases with matched control group in age, sex and residence for the presence of risk factors of bladder cancer.The study was carried out in Upper Egypt which is a narrow strip of land that extends from the cataract boundaries of modern-day Aswan to the area between El-Aiyat and Zawyet Dahshur, south of modern-day Cairo, Figure 1.Figure 1Map of Egypt.The study group were residents of upper Egypt with newly diagnosed histologically proven MIBC admitted to South Egypt Cancer Institute in 2005 (n = 130).Controls were chosen from the healthy visitors of the institute. They were matched to cases for age, sex and residence and were compared to them as regards risk factors. Two controls were matched with each case (n = 260). Full medical history, clinical examination, urinalysis and abdominal ultrasonography were done for controls to exclude the presence of any bladder lesion.Personal interview was conducted to collect socio-demographic data (age, sex, occupation and residence), history suggestive of risk factors (bilharziasis, smoking, chronic cystitis, bladder stones, family history of cancer, parents consanguinity, exposure to chemicals, pelvic radiation and cyclophosphamide chemotherapy), and mode of presentation. History of bilharziasis is defined as finding of ovae on previous urinalysis and history of medical treatment for it. Histopathological pattern of the tumor after cystectomy was recorded.SPSS program (version 13) was used for data analysis, which included descriptive analysis and logistic regression for calculation of risk factors.Approval was obtained from the ethical committee of Faculty of Medicine, Assiut University. An informed written consent was obtained from all the participants, security and confidentiality of all the information obtained was guaranteed.ResultsNone of the controls had any suspicious symptom or sign of MIBC, also no suspicious lesions were found during investigating the controls.The mean age of our patients was 58.34 ± 12.13. Males constituted 83.8% of the patients and 83.1% of controls. Residents of the rural areas were 93.8% of both patients and controls.There was a highly significant statistical difference between cases and controls as regards the exposure to bilharziasis, fertilizer, pesticides, recurrent cystitis, bladder stones, smoking, and positive family history of bladder cancer (p < 0.001). Yet, the type of fertilizer, mode of exposure to it, the type of pesticide, mode of exposure to it, the type of smoking and the degree of relative with bladder cancer were not statistically significantly different between cases and controls.Patients' characteristics are shown in table 1. 73.8% of cases were farmers, 86.2% were married and 91.5% were illiterates.Table 1Patients' characteristics:Patients' CharacteristicsOccupation • Farmer96 (73.8%) • Trader4 (3.1%) • Employee4(3.1%) • House wife22 (16.9%) • Manual worker4 (3.1%)Marital status • Single2 (1.5%) • Married112 (86.2%) • Divorced1 (0.8%) • Widowed15 (11.5%)Education • Illiterate119 (91.5%) • Read and write5 (3.8%) • Primary school graduate2 (1.6%) • Preparatory school graduate1 (0.8%) • Secondary school graduate1 (0.8%) • Intermediate education graduate2(1.5%) • University graduate0 (0%)Risk factors of MIBC as calculated by risk estimate analysis are shown in table 2. The adjusted odds ratio (AOR) as estimated by stepwise logistic regression is shown in table 3. The most important risk factor was the positive family history of the disease (AOR 7.7, CI = 2.1–28.4, p < 0.01) followed by exposure to pesticides (AOR 6.2, CI = 3.5–11.3, p < 0.001).Table 2Risk factors of MIBC as calculated by risk estimate analysis.p-valueOR(95%CI)Positive family history of bladder cancerp < 0.00113.8(4–47.7)Exposure to pesticidesp < 0.0019.4(5.6–15.8)Exposure to fertilizersp < 0.0017.5(4.4–12.8)Bladder stonesp < 0.0017.0(3.5–14.2)Parents' consanguinityp < 0.0016.3(3.9–10.4)Recurrent cystitisp < 0.0016.1(3.4–11.1)Bilharziasisp < 0.0015.8(3.3–10.4)Smokingp < 0.0015.3(3.2–8.7)OR = odds ratio, CI = confidence interval.Table 3Risk factors of MIBC calculated by stepwise logistic regression:P-valueAOR(95% CI)Positive family history of bladder cancerP < 0.017.7(2.1–28.4)Exposure to pesticidesP < 0.0016.2(3.5–11.3)Bladder stonesP < 0.0015.0(2.2–11.4)Parents' consanguinityP < 0.0013.9(2.2–6.7)Chronic cystitisP < 0.013.1(1.5–6.1)AOR = adjusted odds ratio, CI = confidence interval.The results of imaging (ultrasound, IVU and CT) and cystoscopy are shown in table 4; 77% of the patients had an obstructed kidney, 100% of them had a filling defect in IVU. CT showed single bladder lesion in 90% and multiple lesions in 10% of cases.Table 4Patients' clinical, radiological, cystoscopic, and histopathological criteria.(n = 130)First complaint Burning micturition73.8% Haematuria20.8% Loin pain3.8% Frequency1.6%Ultrasound findings Visible growth99.1 Bladder stones8.9% Obstructed kidney77%IVU findings Visible filling defect100% Obstructed kidney77.3% Normal contrast secretion82.7%CT Liver cirrhosisBack pressure on kidneys20% Back pressure on kidneys80%Bladder Single lesion90% Multiple lesion10%Cystoscopy findings Involved urethra0% Involved bladder neck6.6% Involved ureteral orifice Tumor configuration13.1%  Solid91.5%  Papillary8.5%Pathological type Well differentiated SCC43.8% Moderately differentiated SCC20% Transitional cell carcinoma15.4% Anaplastic carcinoma8.5% Spindle cell carcinoma5.4% Poorly differentiated SCC3.8% Adenocarcinoma3.1%Stage of the tumor T2a8.6% T2b47.7% T3a7.8% T3b3.1% T4a28.1% T4b4.7%Lymph node affection Positive13.3% Negative86.7%Table 4 also shows some important clinical and histopathological criteria of our patients. Burning micturition was the first complaint in 73.8% of cases while hematuria was the presenting symptom in only 20.8%. Digital rectal examination revealed a palpable mass in 95.4% of the cases. Squamous cell carcinoma (SCC) constituted 67.6% of cases; 47.7% of cases had a T2b tumor at the time of first presentation. Positive lymph node affection was found in 13.3% of the patients.DiscussionThe mean age of cases in this study was 58.34 ± 12.13 years which agrees with another recent report from Egypt that found that the mean age of bladder cancer cases was 56.24 ± 11 [10]. This age is less than reported in the literature for other parts of the world; Lynch and Cohen, (1995) reported that the median ages at diagnosis for urothelial carcinoma is 69 years in males and 71 years in females [11].The male to female ratio in this study was 5.5: 1. Residents of rural areas constituted 93.8% of cases while only 6.2% of cases lived in urban areas. This difference in male to female ratio than the international ratio of 3:1 might be explained by the fact that women in Upper Egypt aren't equally involved in farming activities with men, hence they are less exposed to the risk factors of the disease that are linked to this occupation (pesticides, fertilizers and bilharziasis) [12].Positive family history of bladder cancer was confirmed in 13.8% of cases. Many studies reported that family history plays a major rule in developing bladder cancer and familial clusters of bladder cancer have been reported in transitional cell carcinoma (TCC) [13]. Moreover, we found that parents' consanguinity is an important risk factor for the development of bladder cancer in offspring, history of consanguinity between parents was found in 50.8% of cases (AOR = 3.9, 95%CI = 2.2–6.9, P < 0.001).Among our patients 73.8% were farmers. In Upper Egypt, cancer risk in this occupational group is considered an important public health problem. Farmers are exposed to several hazardous substances such as fertilizers and pesticides. Moreover, the prevalence and severity of schistosomiasis tend to rise sharply with opportunities for exposure. In Egypt, the disease prevalence increased dramatically after installation of the High Dam, which created perennial irrigation instead of the basin one with subsequent higher exposure to bilharzial infestations [14]. A positive past history of bilharzial infestations was obtained from 87.7% of our cases (OR = 5.8, 95% CI = 3.3–10.4, P < 0.001). Due to the nature of the patients' work as farmers (frequently in contact with water), it is practically very difficult, if not impossible, to define the number of episodes of infestation or the time lag before each treatment course is given, also, the number of treatment courses given doesn't frequently reflect the real number of infestation episodes. Moreover, the heaviness of infestation which is believed to be an important factor in development of bladder cancer can never be practically measured. So an absolute history of exposure to bilharzial infestation was used in this study.There is a plethora of literature incriminating Schistosoma haematobium infestation as a risk factor for bladder cancer, but explanation for this association remains speculative [15]. Evidence that supports the association between schistosomiasis and bladder cancer includes the geographical correlation between the 2 conditions, the distinctive patterns of sex and age at diagnosis, the clinicopathological identity of schistosome-associated bladder cancer, and extensive evidence in experimentally infected animals [16]. The relatively high frequency of bladder cancer in Egypt supports the etiological relationship to urinary schistosomiasis. Despite the marked decrease in prevalence of endemic schistosomiasis over the last 2 decades (decreased from 35% in 1983 to 1.7% in 2003, with complete eradication in certain districts.), Egypt is still paying the toll of the previously high prevalence of the disease. Comparison of the frequency of active urinary schistosomiasis previously reported during the era of high prevalence of the disease and the age-specific incidence rate indicates a strong cohort effect, figure 2. It could be anticipated that in the near future, there will be a marked decrease in bilharziasis associated bladder cancer in Egypt as a sequel to schistosomiasis control. The potential risk is the rise in incidence of bladder cancer related to other risk factors [17].Figure 2Bladder Cancer: Prevalence of Active Schistosomiasis (AS) by Age and Age Distribution of Bladder Cancer (BC) in Egypt. Ibrahim and Khaled (2006) 17.History of exposure to pesticides was obtained in 82.3% of our patients (AOR 6.2, CI = 3.5–11.3, p < 0.001). It is well known that persons exposed to pesticides are at greater risk of developing bladder cancer than persons with no exposure to them [13].El-Mawla et al (2001) stated that urinary bladder stones and chronic cystitis increase the risk of developing bladder cancer and particularly SCC [12]. History of bladder stones was found in 25.4% of our cases (AOR = 5, 95% CI = 2.2–11.4, P < 0.001), while 33.1% of cases had history of recurrent cystitis (AOR = 3.1, 95% CI = 1.5–6.1, P < 0.01).Some authors have claimed that bladder carcinogenesis is related to bacterial infections, which are commonly associated with bilharzial infestation, rather than the parasite itself. Urinary bacteria have a double action: (i) the production of carcinogenic nitrosamines from their precursors in urine, e.g., nitrates and secondary amines, and (ii) the secretion of the enzyme β-glucuronidase, which may clear conjugated carcinogens, yielding free carcinogenic products [18].Among the patients of this study, 80% were current or ex smokers (OR = 5.3, 95% CI = 3.2–8.7, P < 0.001). Cigarette smokers are reported to have up to a fourfold higher incidence of bladder cancer than do people who have never smoked [19]. Radosavljevic et al., (2003) stated that although smoking is still recognized as a major risk factor of cancers including bladder cancer, the increasing incidence of bladder cancer despite the reduction in smoking in the United States suggests that other environmental factors may be playing an increasing role in the development of bladder cancer [20]. This is in accordance with the results of this study where smoking failed to be an independent risk factor for MIBC when adjusted to other environmental factors (pesticides, fertilizers and bilharziasis).The present data contradicts the common belief in Egypt about the major role bilharziasis plays in bladder cancer development. A larger role for the exposure to pesticides and fertilizers is evident in this study. Also the role of family history and consanguinity between parents seems to be higher than ever recognized. The risk factor profile of Egyptian bladder cancer has changed over the last 26 years as exposure to chemical carcinogens play a major role for the development of bladder cancer in Egypt [21].Histopathological examination showed that 67.6% of cases had SCC, 15.4% TCC, 8.5% anaplastic carcinoma and 8.5% had other pathological types (table 4). This seams like a shift from what was previously reported regarding the percentage of SCC which exceeded 75% in Bilharzial bladders [22]. According to many Egyptian authors, the pattern of histopathology of bladder cancer showed a marked change over the previous years, where SCC constituted less than 60% of bladder cancer [12,21,23]. This change might be explained by the introduction of mass treatment of bilharziasis in recent years in contrast to the increased exposure to pesticides and fertilizers.As regards the mode of presentation, globally, the most common presenting symptom of bladder cancer is painless hematuria, which occurs in about 90% of cases [24]. In this study the main presenting complaint was burning micturition, which was the presenting symptom in 73.8% of cases while hematuria was the 1st complaint in only 20.8%. This different mode of presentation might be due to the different tumor configuration which was found to be solid in 91.5% of patients. Unlike the papillary configuration which can bleed easily on shedding of the tumor cells, the solid tumor configuration delays the occurrence of hematuria. This mode of presentation complicates the picture of bladder cancer in Egypt as the patients who are usually accustomed to some kind of burning micturition due to bilharziasis, bladder stones and cystitis don't ask for medical advice until their tumors are already invasive [25]. This was reflected on the result of digital rectal examination where bladder mass was detected in 95.4% of cases.Lymph node affection was found in only 13.3% of the patients. This might be explained by the fibrosis that affects the lymphatics in bilharzial patients. This is in agreement with the findings of other reports from Egypt [26].ConclusionMIBC in Upper Egypt is peculiar in that it is usually of the SCC type (although its percentage is decreasing), occurs at a younger age and presents with burning micturition rather than hematuria.Unlike the common belief, risk factors such as positive family history, parents' consanguinity, exposure to pesticides and chronic cystitis seem to play now more important roles than bilharziasis and smoking in the development of this disease in this area, yet reports on larger numbers of patients are needed to support this conclusion.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsAHZ supervised the work and gave important suggestions, MS supervised the work, participated in data analysis and manuscript writing and directed the clinical assessment, AAA-E carried out the field work, interviewing cases and controls, clinical assessment of cases and controls, data management and analysis and writing the final manuscript, DAH participated in writing the drafts, writing the final manuscript and gave very important clinical suggestions, MAA supervised the work, participated in the clinical assessment of cases and controls, directed the clinical investigations, gave very important suggestions and participated in the manuscript writing.Pre-publication historyThe pre-publication history for this paper can be accessed here:\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2533393.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533393",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533393\nAUTHORS: Tiziana Musso, Sara Scutera, William Vermi, Roberta Daniele, Michele Fornaro, Carlotta Castagnoli, Daniela Alotto, Maria Ravanini, Irene Cambieri, Laura Salogni, Angela Rita Elia, Mirella Giovarelli, Fabio Facchetti, Giampiero Girolomoni, Silvano Sozzani\n\nABSTRACT:\nLangerhans cells (LC) represent a well characterized subset of dendritic cells located in the epidermis of skin and mucosae. In vivo, they originate from resident and blood-borne precursors in the presence of keratinocyte-derived TGFβ. Ιn vitro, LC can be generated from monocytes in the presence of GM-CSF, IL-4 and TGFβ. However, the signals that induce LC during an inflammatory reaction are not fully investigated. Here we report that Activin A, a TGFβ family member induced by pro-inflammatory cytokines and involved in skin morphogenesis and wound healing, induces the differentiation of human monocytes into LC in the absence of TGFβ. Activin A-induced LC are Langerin+, Birbeck granules+, E-cadherin+, CLA+ and CCR6+ and possess typical APC functions. In human skin explants, intradermal injection of Activin A increased the number of CD1a+ and Langerin+ cells in both the epidermis and dermis by promoting the differentiation of resident precursor cells. High levels of Activin A were present in the upper epidermal layers and in the dermis of Lichen Planus biopsies in association with a marked infiltration of CD1a+ and Langerin+ cells. This study reports that Activin A induces the differentiation of circulating CD14+ cells into LC. Since Activin A is abundantly produced during inflammatory conditions which are also characterized by increased numbers of LC, we propose that this cytokine represents a new pathway, alternative to TGFβ, responsible for LC differentiation during inflammatory/autoimmune conditions.\n\nBODY:\nIntroductionLangerhans cells (LC) are specialized dendritic cells (DC) normally found in the epidermis and mucosal stratified epithelia [1]–[5]. Contrary to myeloid and plasmacytoid DC, LC express the C-type lectin CD207 (Langerin), the major constituent of Birbeck granules, which represent the hallmark of LC [6]–[8]. LC also express a characteristic set of cell-surface molecules, such as cutaneous lymphocyte-associated antigen (CLA), E-cadherin and the CC chemokine receptor 6 (CCR6) [9]–[13]. As immature cells, their primary function is to sense the environment for danger signals and capture antigens; then LC undergo a process of functional and phenotypic maturation and migrate to the regional lymph nodes [5]; [10]; [14]; [15].Several evidences suggest that LC are of myeloid origin and can differentiate from monocytes or CD34+ precursors [16]–[20]. While the generation of DC from monocytes requires GM-CSF and IL-4 only, the additional presence of TGFβ1 in the cytokine milieu appears to be essential for the development of LC [17]; [18]. Accordingly, TGFβ1-deficient mice display a severe defect in LC, but not in DC, development [21]. IL-15 is the only other cytokine known until now to skew monocyte differentiation toward LC-type DC; though, IL-15-derived LC are Langerin+ but lack Birbeck granules [22].According to the current model of LC differentiation, in steady-state conditions LC are maintained locally by a stable renewable population present in the skin [23]–[25]. The existence of skin-resident LC precursors was postulated by Lareggina et al who described dermal CD14+ cells that express Langerin and CCR6 and are able to acquire LC features when cultured in the presence of TGFβ1 [26]. When the skin is exposed to inflammatory stimuli (UV rays, infections, allergens) LC increase their expression of class II MHC and costimulatory molecules and migrate to regional lymph nodes. In this situation of accelerated turnover, LC are replaced by blood-borne precursors such as inflammatory Gr-1+ monocytes recruited through a CCR2-dependent mechanism [23]; [27]; [28].The epidermal environment can contribute to the attraction of precursors and to their differentiation into LC. Keratinocyte production of MIP-3α/CCL20 and TGF may direct CCR6+ LC precursors to the epidermis and induce their entry into the LC pathway, respectively [9]; [18]; [21]. However, LC differentiation could also depend on the dermal cytokine environment once LC precursors have entered the skin to colonize the dermis, or are trafficking through the dermis to the overlying epidermis. Indeed, emerging evidence support the concept that in certain skin pathological conditions a conspicuous expansion of the LC population occurs within the dermis [29]–[32] suggesting, that LC differentiation factors might also be produced within the dermal compartment.Activin A is a member of the TGFβ1 family initially identified for its ability to control the secretion of follicle-stimulating hormone. Activin A is presently also known for its activity on growth and differentiation of various cell types during organogenesis, and for its role in wound healing, inflammation and tumor progression [33]–[36]. Activin A binds to specific transmembrane serine/threonine kinase receptors (ActRIB and ActRII) and to follistatin, a secreted protein that inhibits protein functions by sequestration [37]–[39]. Activin A is strongly induced after skin injury, probably by serum growth factors released upon haemorrhage and by macrophage-derived pro-inflammatory cytokines [40]; [41]. Transgenic mice overexpressing Activin A in the epidermis show strong hyperthickened epidermis, accelerated wound healing and enhanced scarring [42]; [43]. Conversely, in transgenic mice overexpressing the antagonist follistatin, skin wound closure is delayed and scar formation reduced [44]. Curiously, LC are strongly reduced in the skin of follistatin transgenic mice, suggesting a role of Activin A in LC biology [45].This study shows that Activin A induces the differentiation of LC in vitro and ex-vivo and candidates Activin A as a new differentiation pathway that might be relevant in conditions characterized by the local production of Activin A and accumulation of LC.ResultsActivin A induces the differentiation of circulating monocytes into DC with phenotypic and ultrastructural features of LCHighly purified monocytes were cultured for 6 days with Activin A in the presence of GM-CSF and IL-4, two cytokines that were shown to cooperate with TGFβ1 in the differentiation of monocytes to LC [16]; [18]. These cells (thereafter called Act A-LC to differentiate them from TGFβ1-LC) were CD14− and expressed typical LC markers, such as CD1a, Langerin, E-caderin, CLA and CCR6. At day 6, Act A-LC presented an immature phenotype with a modest expression of CD80, CCR7 and CD83 (Fig. 1A). The presence of Birbeck granules, the hallmark of epidermal LC [7] was assessed by transmission electron microscopy on ultrathin-sections of Act A-LC cells (Fig. 1B). These cells displayed dendritic morphology with slightly off centred indented nuclei. The cytoplasm contained Birbeck granules, cytoplasmic organelles with rod like profile, electron-opaque central lamella and rounded ends. More open-ended tennis-racket shaped granules were also observed. Taken together this set of data indicates that Activin A, in the presence of GM-CSF and IL-4, induces the differentiation of circulating monocytes into LC. The possibility that the effect of Activin A could be due to the secondary induction of TGFβ, was investigated by real-time PCR. As shown in Figure 1C, TGFβ1 mRNA was barely detectable in Act A-LC cultures and similar results were obtained for TGFβ2 and TGFβ3 (data not shown); TGFβ1 was also weakly induced in TGFβ1-LC. On the contrary, Activin A mRNA was strongly upregulated by both Activin A and TGFβ1, suggesting the existence of an amplificatory loop. In agreement with mRNA levels, TGFβ1 concentration was below the detection limits (sensitivity 4.61 pg/ml) in Act A-LC supernatants, whereas Activin A was induced at 1.05 ng/106 cells (n = 5, n = 6) in TGFβ1-LC. Furthermore, Act A-LC generation was not blocked by the addition of anti-TGFβ1 (data not shown), supporting a TGFβ1-independent differentiation of Act A-LC. Finally, it was tested whether bone morphogenetic protein (BMP-6), another TGF family member protein, could also induce LC differentiation. Figure 1D shows that BMP-6 did not sustain phenotypic LC differentiation, indicating that the ability to induce LC differentiation is not a general feature shared by all TGFβ family member proteins.10.1371/journal.pone.0003271.g001Figure 1Activin A promotes Langerhans cell differentiation from human CD14+ monocytes.(A) Phenotypic analysis of monocytes cultured for 6 days with GM-CSF and IL-4 in the presence of Activin A (Act A-LC) or TGFβ1 (TGFβ1-LC). Cells were stained with the indicated moAbs (filled histograms) or isotype-matched negative control moAbs (open histograms). Percentages of positive cells are shown in the upper right corner of each histogram. The figure shows one experiment representative of at least five independent cultures. (B) Electron microscopy analysis of Act A-LC. Act A-LC exhibited abundant dendritic membrane protrusions and lobulated or indented nuclei (left panel, 3,000X, bar 40 µm). Cytoplasm presented a rough endoplasmic reticulum, many multilamellar organelles and numerous electron-dense structures reminiscent of Birbeck granules (right panel, 12,000X, bar 1 µm). The inset shows rod-shaped Birbeck granules (200,000X, bar 20 µm). (C) TGFβ1 and Activin A mRNA expression in Act A-LC and TGFβ1-LC cultures. Monocytes were cultured in the presence of Act A or TGFβ1 for the indicated time and the expression of TGFβ1 and Activin A mRNA was determined by real-time PCR, relative to GAPDH mRNA used as internal control. The expression level in freshly isolated monocytes was assumed as the 1.0 value. Similar results were obtained in three different donors. (D) Effects of different TGF family members on LC differentiation. Monocytes were cultured for 6 days with GM-CSF in the presence of 10 ng/ml TGFβ1, 100 ng/ml Activin A, or 100 ng/ml BMP6 and analyzed for Langerin, E-caderin and CCR6 expression by flow cytometry analysis. Data are representative of at least four independent cultures.Phenotypical and functional characterization of Act A-LCMembrane phenotype and functions were tested in immature and CD40L-mature Act A-LC. As shown in Figure 2A CD40L-activated Act A-LC showed increased surface expression of CD80 and CD83 as well as the expression of CXCR4 and CCR7, two maturation-associated chemokine receptors; these data are consistent with the acquisition of a mature phenotype. Immature Act A-LC efficiently stimulated T cell proliferation and their allostimulatory capacity was further enhanced upon maturation (Fig. 2B). In agreement with the expression of CCR6 (Fig. 1A and data not shown), immature, but not mature Act A-LC migrated in response to CCL20 (Fig. 2C). On the contrary, CD40L-mature cells migrated in response to CCL19, one of the CCR7 ligands (data not shown). Finally, CD40L-activated Act A-DC secreted IL-12p70, TNF, CCL22 and CCL20 in a similar manner to TGFβ1-LC, with the exception of IL-12 which was consistently produced at higher levels by Act A-LC (Fig. 2D). Therefore, LC generated in the presence of Activin A have the capacity to undergo a full maturation process based on membrane phenotype, migration and functional properties.10.1371/journal.pone.0003271.g002Figure 2Phenotypical and functional characterization of CD40L-activated Act A-LC(A) Expression of maturation markers by Act A-LC. Act A-LC were incubated with CD40L-transfected fibroblasts for 40 hrs and stained with anti-CD80, CD83, CCR7 and CXCR4 moAbs (filled histograms) or isotype-matched negative control Abs (open histograms). Results obtained with TGFβ1-LC are also shown for comparison. The percentage of positive cells is reported in each panel. Data shown are representative of three independent experiments. (B) Allostimulatory capacity of Act A-LC. Irradiated immature or CD40L-matured Act A-LC (or TGFβ1-LC) were cultured with 2×105 allogeneic purified T cells. Proliferation was assayed as uptake of [H3]thymidine added in the last 16 hrs of a 6-day culture assay. Results are expressed as mean counts per minute (cpm)±SD of one representative experiment performed in triplicate. Values are at the net of T cell proliferation in the absence of DC (3250±250 cpm). (C) Act A-LC migrate in response to CCL20. Immature or CD40L-mature Act A-LC or TGFβ1-LC were applied to the upper wells of the chemotaxis chamber. CCL20 was added to the lower level of the chamber. The number of cells migrated to the lower chamber was counted. Each assay was performed in triplicate and the results are expressed as the mean±SD number of migrated cells (representative of three experiments). (D) Cytokine release by Act A-LC. Immature or CD40L-mature Act A-LC or TGFβ1-LC were assessed for their ability to release the indicated cytokines by ELISA. Results are the average determination (±SD) of four independent experiments.Activin A induces the generation of Langerin+ cells ex-vivo in human skin explantsTo investigate the ability of Activin A to induce LC differentiation within the skin milieu, skin explants were intradermally injected with Activin A and subsequently cultured in six-well culture plates at the air-medium interface with the epidermis side up [46]. Explants were subsequently removed and examined by immunohistochemistry. As expected, fresh, untreated skin explants revealed several Langerin+ cells within the epidermis (Fig. 3) and a similar picture was observed following the injection of medium. Instead, the inoculation of 100 ng Activin A led to a profound increase in the number of Langerin+ cells both in the epidermis and in the dermal layer, with a maximal induction observed 72 hrs after the injection. The increase of the number of Langerin+ cells was of about 2-fold and 10-fold (n = 6) in the epidermis and dermis, respectively, and was statistically significant with respect to control skin (p<0.05, by Student's t-test; Fig. 3). The number of CD1a+ cells also increased in parallel to Langerin expression (data not shown). These results show that Activin A is able to induce the differentiation of LC precursors resident within normal skin.10.1371/journal.pone.0003271.g003Figure 3Intradermal injection of Activin A induces the differentiation of dermal and epidermal Langerhans cells in human skin explants.Langerin expression was evaluated in the epidermis and dermis (full thickness skin explants) of skin explants, untreated and 72 hrs after i.d. injection of medium or 100 ng Activin A (magnification 100X, inset 400X) The number of Langerin+ cells were quantified in skin explants by evaluating six different skin sections (0.05 mm2/field; means±SD). * p<0.05 by Student's t test vs. medium (lower right panel).Activin A promotes LC differentiation from precursors cells present in the dermal layerTo exclude that the increased number of LC observed in the dermis following Activin A injection could be due to migration of LC from the epidermis, the dermis was separated by dispase digestion and thereafter injected with Activin A. Also under these experimental conditions, Activin A inoculation strongly increased the number of Langerin+ cells (Fig. 4). To better address the potential of dermal precursors to differentiate into LC under the influence of Activin A, skin migratory cells were recovered from dermal layers and subsequently cultured in the presence of Activin A. At day 6 of culture, a consistent number of cells, 15% and 20% were positive for CD1a and Langerin, respectively. Conversely, no Langerin+ cells could be detected in the absence of Activin A. Altogether, these data show that skin precursors, present within the dermis, can be induced to differentiate into LC by Activin A.10.1371/journal.pone.0003271.g004Figure 4Activin A induces Langerhans cells differentiation in epidermis-depleted skin explants.Langerin expression was evaluated in the dermal layer, separated from skin explants by dispase digestion and subsequently treated for 72 hrs after i.d. injection with 100 ng Activin A (magnification 100X, inset 400X.).Dermal accumulation of Langerhans cells in lichen planus is associated with abundant production of Activin AAlthough LC are predominantly confined to the epidermis, we and others have recently documented that in certain pathological conditions, such as lichen planus, LC are also abundantly found in the stromal compartment [29]; [30]; [32]. As shown in Fig. 5 (panel a), in normal skin and mucosa LC are regularly distributed within the epithelium in the basal and suprabasal layers and are easily recognized based on their dendritic morphology and expression of Langerin [7]. On the contrary, in the large majority of the lichen planus cases investigated (32/34) variable numbers of Langerin+ dendritic cells were identifiable in the stromal compartment (Fig. 5, panel b), distributed as sparse cells or clusters within the mononuclear infiltrate (Fig. 5, panel c). Of interest, many of these Langerin+ cells were detected in close proximity to blood vessels, as shown by double immunofluorescence for Langerin and Factor VIII-related antigen (Fig. 5, panel c, inset). In normal skin a weak reactivity for Activin A was detected in epidermal keratinocytes and rare spindle cells (likely representing dermal macrophages or dendritic cells); a stronger positivity was also observed in scattered mast cell [47] (Fig. 5, panel d). Compared to normal skin, lesional skin and mucosa from lichen planus biopsies showed strong induction of Activin A in the upper layers of the epidermis and, particularly, in the stromal compartment (Figure 5, panel panel e). In the latter, Activin A was mostly produced by non-lymphoid mononuclear cells, endothelial cells and mast cells (Figure 5, panel f)10.1371/journal.pone.0003271.g005Figure 5Dermal accumulation of Langerhans cells in lichen planus is associated to abundant production of Activin A.Sections from normal skin (NS) (a and d) and lichen planus (LP) (b, c, e, f) biopsies were stained for Langerin (a–c) and Activin A (d–f). In normal skin, Langerin+ cells are regularly distributed in basal and suprabasal layers and show multiple fine dendrites; no positive cells are detectable in the dermis (panel a). In LP biopsies, in addition to intraepidermal LC, accumulation of Langerin+ cells is observed in the dermis within the dense monuclear cell infiltrate (panel b). At high power view, Langerin+ cells show an ovoidal/dendritic shape (panel c) and are found surrounding Factor VIII+ dermal blood vessels (arrow head, inset in c). Serial sections from the same tissue blocks were stained for Activin A. Normal skin (panel d) showed weak intraepithelial reactivity (red arrow head); in the dermis, mast cells and occasional spindle cells were positive for Activin A (black arrow heads). In LP, Activin A was strongly induced in the superficial layers of epidermis; in the dermis, a diffuse reactivity can be observed in numerous cells within the inflammatory infiltrate (panel e). This cell population includes endothelial cells and a mixture of non-lymphoid mononuclear cells (panel f). Magnification 100x (a, b, d, e; scale bar 200 micron) and 400x (c, f; scale bar 50 micron).DiscussionThis study reports that Activin A, a protein abundantly produced in the skin during normal and pathological wound healing [40]; [47] and inflammatory/autoimmune diseases (this study), induces the differentiation of human CD14+ monocytes in Langerin+, Birbeck granules+, E-cadherin+, CLA+ and CCR6+ cells. LC originate in vitro from CD34+ bone marrow precursors [16]; [17]; [19]; [20]. Monocytes also represent LC blood precursor cells in vitro and in vivo [18]; [22]; [27]; [28]. Indeed, CD14+ cells can be induced to differentiate into LC by a cytokine combination including GM-CSF, IL-4 and TGFβ1 [18]. The crucial role of TGFβ in LC differentiation has been clearly documented by the observation that TGFβ1 null mice are devoid of LC [21]. More recently, it was shown that the cytokine milieu present at the inflammatory site may favour LC differentiation through an alternative pathway [22]; [28]. Indeed, in vitro experiments have shown that IL-15, in cooperation with GM-CSF, induces the differentiation of monocytes into cells that express LC markers, such as E-cadherin, CCR6 and Langerin but lacking the expression of conventional Birbeck granules [22]. The present study adds Activin A to the limited list of cytokines that possess the potential to promote LC differentiation. Activin A is a member of the TGFβ family and shares some of the intra-cellular signalling pathways with this cytokine [48]. However, the ability to induce LC differentiation is not a general feature shared by all TGFβ family member proteins, as documented by the lack of activity of BMP-6 in our assay conditions. The selective action of Activin A versus BMP-6 is likely to be due to the usage of specific transmembrane receptors and the activation of different signalling pathways [49]; [50].LC form a cellular network in the epidermis that constitutes the first immunological barrier against pathogens and dangerous insults. Following antigen capture, LC leave the epidermis by a mechanism that depends on the expression of chemotactic receptors, adhesion molecules and proteases [10]; [14]; [51]. Emigrating skin cells enter lymphatic vessels located in the superficial dermis to finally reach draining lymph nodes where they present processed antigens to naïve T cells [52]. During this migratory process, LC acquire a mature phenotype that is associated with the expression of homing receptors, co-stimulatory molecules and the ability to release several cytokines [10]; [52]; [53]. The results presented in this study show that LC generated in the presence of Activin A are fully competent to undergo a maturation process, as evaluated by the expression of CCR7 and the downregulation of CCR6, the expression of CD80 and CD83, the ability to induce T cell proliferation and to secrete high levels of chemokines (i.e. CCL20 and CCL22) and cytokines (TNF-α, IL-12p70).LC are normally confined to the basal and suprabasal layer of the epidermis and stratified epithelia of mucosal surfaces. These cells are clearly distinct from dermal/interstitial DC which lack Birbeck granules and Langerin expression, but express DC-SIGN, Factor XIIIa and more rarely CD1a [54]–[58]. The current view proposes that under steady-state conditions, dermal-resident CD14+ precursor cells have the potential to migrate to the epidermis in response to CCL20 and there, in the presence of keratinocyte-derived TGFβ, differentiate into immature resident LC characterized by a weak T cell stimulatory activity. In the presence of the cytokine rich milieu that characterizes many pathological conditions, migratory CD14+ cells further differentiate into more mature LC which possess a higher antigen presenting activity [25]; [26]; [28]. In order to evaluate the potential of skin resident precursor cells to differentiate into LC in response to Activin A, we performed experiments in which skin biopsies were inoculated ex-vivo with Activin A and further incubated in vitro [46]. The immunohistochemical evaluation of these skin explants clearly show that the injection of Activin A induced a strong increase in the number of Langerin+/CD1a+ cells in both the epidermal and dermal compartments in a time-dependent manner. Due to the experimental conditions employed, Langerin+ cells must have originated from skin-resident precursor cells. Further we show that Langerin+ cells can be induced to differentiate by Activin A within the dermis in the absence of epidermis. In agreement with these results, cells emigrated from skin explants could also be induced to differentiate into Langerin+/CD1a+ cells by the presence of Activin A in vitro. These findings are compatible with the description of CD14+, Langerin+ LC precursors located in the superficial and deep dermis, predominantly in perivascular areas [24]; [26]. Although the precise characterization of Activin A-responsive dermal LC precursors is beyond the aim of the present study, these results clearly document that Activin A can induce the local differentiation of dermal LC precursors. In this contest it is interesting to note that Stoitzner et al. reported that in mice overexpressing in the follistatin, the natural Activin A antagonist, the number of LC is reduced [45].Although Activin A is very weakly expressed in normal skin, its expression was dramatically increased in lichen planus biopsies. In this condition, Activin A was expressed both in the superficial epidermis and in the dermis by stromal cells, infiltrating leukocytes, including mast cells and some blood vessels. As expected on the basis of previous work [29]–[32], lichen planus biopsies show a prominent increase in the number of LC which were present in the deep and superficial derma and in the epidermis. Of note, LC were often present in clusters localized around Factor VIII+ blood vessels, suggesting the involvement of newly vascular-recruited precursor cells. Although these data generated in a human disease do not allow formal conclusions, it is tempting to speculate that during certain pathological conditions characterized by the local expression of Activin A and inflammatory cytokines (such as GM-CSF and IL-4), dermal LC precursors, or newly recruited blood elements, can be induced to differentiate to LC within the stromal compartment. This model may help to explain the origin of LC localized in the deep dermal layers, away from the epidermis and from superficial lymphatic vessels.Lichen planus is an autoimmune disease characterized by a prominent cellular infiltrate mainly composed of DC, LC, cytotoxic T lymphocytes and NK cells, localized in close proximity of apoptotic keratinocytes [29]–[32]. The involvement of the TGFβ family members in this pathology is suggested by several observations. First, BMP-4 is upregulated in the epithelium of lichen planus [59]. Second, several lines of evidence suggest that the TGFβ activation and/or signal transduction pathway might be defective in this disease. Indeed, lichen planus is associated with epithelial hyperproliferation, a situation that is usually negatively controlled by TGFβ, and this is consistent with the identification of TGFβ positive T cells in the sub-epithelial lymphocytic infiltrate but not within the epithelium itself. It is therefore of interest to note that this defective TGFβ pathway is associated with a high expression of Activin A (this study). In this context it is likely that Activin A may have a prominent role in LC differentiation in lichen planus. Furthermore, Activin A may contribute to the pathogenesis of lichen planus by favouring epithelial hyperplasia.In summary, this study presents a new model in which Activin A induces the differentiation of circulating CD14+ cells into LC. Since Activin A is abundantly produced during certain inflammatory conditions, we propose that this cytokine represents a new pathway, alternative to TGFβ, responsible for LC differentiation during inflammatory/autoimmune conditions.Materials and MethodsThe study was conducted in accordance with a protocol approved by the Spedali Civili of Brescia Institutional Ethical Board (Brescia, Italy) and the Board of the CTO Hospital (Turin, Italy); written informed consent was obtained from all patients.Cell culturesCD14+ monocytes were isolated from buffy coats (Centro Trasfusionale Brescia, Italy) by positive magnetic separation using CD14 immunomagnetic beads (Miltenyi Biotec, Auburn, CA) [60]. To generate LC, monocytes were cultured for 6 days in 6 wells tissue culture plates (Costar, Corning, Cambridge, MA) in 10% heat-inactivated FCS RPMI 1640 (Sigma, St. Louis, MO) supplemented with 100 U/mL penicillin, 100 µg/mL streptomycin, 2 mM L-glutamine. GM-CSF 100ng/ml and IL-4 10 ng/ml were added at day 0. 10ng/ml TGFβ1 or 100 ng/ml Activin A were also added at day 0. All cytokines were from Peprotech, Rocky Hill, NJ. Half the culture medium was replaced with fresh medium containing cytokines on day 2 and 4. LC maturation was induced by incubation with CD40L-transfected J558 cells (1∶4 ratio) for 24 hrs.Flow cytometric analysisSurface phenotype analysis was performed using the following antibodies: phycoerythrin (PE)-conjugated anti-CD1a (anti-CD1a-PE), anti-Langerin (CD207)-PE, anti-E-cadherin, and anti-CD83-PE (Immunotech, Marseille, France); anti-human leukocyte antigen (HLA)-ABC-PE and FITC-conjugated anti-CD1a (anti-CD1a-FITC; Dako, Glostrup, Denmark); anti-HLA-DR-PE, anti-CD80-PE, (BD PharMingen, San Diego, CA); anti-CD86-PE, anti-CD14-PE, anti-CC chemokine receptor 7 (CCR7)-PE, and anti-CCR6-FITC (R&D Systems). Anti-CLA-FITC rat mAb (BD PharMingen) was also used. Mouse immunoglobulin G1 (IgG1)-PE, mouse IgG1-FITC, rat anti-mouse IgG1-FITC, and rat IgG2a-PE were from BD Pharmingen. Purified mouse IgG1 (R&D Systems), rat IgM-FITC, mouse IgG2b-PE, mouse IgG2a-PE (BD PharMingen), or mouse IgG2b-FITC (Beckman Coulter, Hialeah, FL) were used as an isotype control. Cells were analyzed with a FACScan flow cytometer (Becton Dickinson, San Jose, CA) using CellQuest software.Transmission electron microscopy106 cells were fixed with 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) and postfixed with 1% osmium tetroxide in 0.1 M cacodylate buffer (pH 7.4). Next, cells were dehydrated through a graded series of ethanol and embedded in araldite (Poersch, Frankfurt, Germany). Ultrathin sections were counterstained with uranyl acetate and lead citrate and were examined with a Zeiss electron microscope (EM 906, Zeiss, Oberkochen, Germany).Evaluation of DC functionsIrradiated immature and CD40L-stimulated LC were added in triplicate in graded doses to 2×105 purified allogeneic T cells in 96-well round-bottom plates. [3H]Thymidine incorporation was measured on day 5 after a 16-h pulse (5 Ci/µmol; Amersham Biosciences, Freiburg, Germany). Human IL-12p70, Activin A, TGFβ1, MDC/CCL22, TARC/CCL17, MIP-3α/CCL20 protein levels in the culture supernatants were measured by sandwich ELISA (R&D Systems, Minneapolis, MN). DC migration was evaluated using a 48-well microchemotaxis chamber (Neuroprobe, Pleasanton, CA) with 5-µm pore size polyvinylpyrrolidone polycarbonate filters (Neuroprobe) as previously described [61].RNA purification and real time RT-PCR analysisRNA samples, extracted by using TRIzol (Invitrogen, Carlsbad, CA), were treated with DNase (Invitrogen) and single-stranded complementary DNA (cDNA) was synthesized by reverse transcription of 2 µg total RNA using random hexamers and the Superscript First-Strand Synthesis System for RT-PCR (Invitrogen). The cDNAs were then amplified in duplicate by real-time PCR using the Platinum SYBR Green (Invitrogen) in a final volume of 25 µl and quantitative analysis was carried out as previously described [47]. The sequences of primers were as follows: Activin A (sense 5′-GCA GAA ATG AAT GAA CTT ATG GA-3′; antisense 5′-GTC TTC CTG GCT GTT CCT GAC T-3′), TGFβ1 (sense 5′-GCG TGC TAA TGG TGG AAA-3′; antisense 5′-CGG TGA CAT CAA AGA TAA CCA C-3′), β-actin (sense, 5′-GTT GCT ATC CAG GCT GTG-3′; antisense, 5′-TGT CCA CGT CAC ACT TCA-3′).Skin explant culturesSkin specimens were obtained from patients undergoing corrective breast or abdominal plastic surgery. Cytokines were injected into the dermis with a MicroFine insulin syringe (29 gauge needle) in the indicated amounts and in a total volume of 20 µl. At the site of injection, a ∼5-mm wheal appeared and a 6 mm punch biopsy was taken. For immunohistochemistry, skin biopsies were cultured at air-medium interface with the epidermis side up in a six-well culture plate (Costar) on sterilized stainless steel grids covered with a filter (Millipore, Bedford, MA; 45 µm), at 37° C in 5% CO2 humidified air [46]; [62]. At the indicated times, the explants were harvested, snap frozen, and stored in liquid nitrogen. To study phenotypic development of epidermal and dermal DC separately, the epidermal and dermal layers were separated by dispase digestion for 1 h at 37°C (Dispase grade II, 50mg/ml; Roche). To obtain emigrated skin cells, the epidermal and dermal layers were placed directly in 1 ml culture medium (floating with epidermis side up) in a 48-well culture plate (Costar) with 100 ng/ml Activin A. After 7 h, migratory cells were collected, counted in hemocytometers using trypan blue exclusion and cultured in the presence of GM-CSF and Activin A for 6 days.Tissue specimens and staining proceduresParaffin embedded tissue blocks were taken from the archive of the Department of Pathology of the University of Brescia/Spedali Civili di Brescia; tissues included normal skin, oral (13 cases) and cutaneous (21 cases) lichen planus and granulosa cell tumors (3 cases). Four micron tissue sections were used for immunohistochemical staining using primary Abs to the following antigens: human Activin A (clone E4, dilution 1∶50, Serotec, Oxford, UK), human Langerin (clone 12D6, 1∶200, Vector Laboratories, Burlingame, CA), human Factor VIII-related antigen (rabbit polyclonal, 1∶100, Neomarkers, Westnghouse, CA). Upon antigen retrieval with microwave treatment (3 cycles of 5 min×750 W) or thermostatic bath (40′ at 89°) in EDTA buffer solution, reactivity was revealed using Real EnVision Mouse/Rabbit-HRP (DakoCytomation, Glostrup, Denmark) or SuperSensitive IHC Detection System (BioGenex, San Ramon, CA). For double-immunofluorescence staining anti-human Langerin and Factor VIII-related antigen were revealed respectively using a goat anti-mouse IgG2b (1∶75, Southern Biotek, Birmingham, AL) followed by Streptavidin Texas Red (1∶75, Southern Biotek) and a FITC-conjugated swine anti-rabbit (1∶30, Dako Cytomation). For immunohistochemical staining of Activin A, positive tissue control was represented by granulosa cell tumors that strongly expressed this protein [63]; no reactivity was observed when omission of primary Ab and irrelevant isotype matched primary antibody were used. Immunostained sections were independently examined by two pathologists (WV and FF); digital images taken using the Olympus BX60 microscope and a DP-70 Olympus digital camera were processed using Analysis Image Processing software.\n\nREFERENCES:\n1. BanchereauJBriereFCauxCDavoustJLebecqueS\n2000\nImmunobiology of dendritic cells.\nAnnu Rev Immunol\n18\n767\n811\n10837075\n2. BellDYoungJWBanchereauJ\n1999\nDendritic cells.\nAdv Immunol\n72\n255\n324\n10361578\n3. SteinmanRMBanchereauJ\n2007\nTaking dendritic cells into medicine.\nNature\n449\n419\n426\n17898760\n4. StinglGKatzSIShevachEMRosenthalASGreenI\n1978\nAnalogous functions of macrophages and Langerhans cells in the initiation in the immune response.\nJ Invest Dermatol\n71\n59\n64\n355568\n5. StroblHRiedlEBello-FernandezCKnappW\n1998\nEpidermal Langerhans cell development and differentiation.\nImmunobiology\n198\n588\n605\n9561375\n6. KashiharaMUedaMHoriguchiYFurukawaFHanaokaM\n1986\nA monoclonal antibody specifically reactive to human Langerhans cells.\nJ Invest Dermatol\n87\n602\n607\n3534103\n7. Mc DermottRZiylanUSpehnerDBausingerHLipskerD\n2002\nBirbeck granules are subdomains of endosomal recycling compartment in human epidermal Langerhans cells, which form where Langerin accumulates.\nMol Biol Cell\n13\n317\n335\n11809842\n8. ValladeauJRavelODezutter-DambuyantCMooreKKleijmeerM\n2000\nLangerin, a novel C-type lectin specific to Langerhans cells, is an endocytic receptor that induces the formation of Birbeck granules.\nImmunity\n12\n71\n81\n10661407\n9. PowerCAChurchDJMeyerAAlouaniSProudfootAE\n1997\nCloning and characterization of a specific receptor for the novel CC chemokine MIP-3alpha from lung dendritic cells.\nJ Exp Med\n186\n825\n835\n9294137\n10. SozzaniS\n2005\nDendritic cell trafficking: more than just chemokines.\nCytokine Growth Factor Rev\n16\n581\n592\n15963754\n11. TangAAmagaiMGrangerLGStanleyJRUdeyMC\n1993\nAdhesion of epidermal Langerhans cells to keratinocytes mediated by E-cadherin.\nNature\n361\n82\n85\n8421498\n12. YangDHowardOMChenQOppenheimJJ\n1999\nCutting edge: immature dendritic cells generated from monocytes in the presence of TGF-beta 1 express functional C-C chemokine receptor 6.\nJ Immunol\n163\n1737\n1741\n10438902\n13. YasakaNFurueMTamakiK\n1996\nExpression of cutaneous lymphocyte-associated antigen defined by monoclonal antibody HECA-452 on human Langerhans cells.\nJ Dermatol Sci\n11\n19\n27\n8867763\n14. RandolphGJOchandoJPartidaSNS\n2007\nMigration of Dendritic Cell Subsets and their Precursors.\nAnnu Rev Immunol\n15. SteinmanRMHemmiH\n2006\nDendritic cells: translating innate to adaptive immunity.\nCurr Top Microbiol Immunol\n311\n17\n58\n17048704\n16. CauxCDezutter-DambuyantCSchmittDBanchereauJ\n1992\nGM-CSF and TNF-alpha cooperate in the generation of dendritic Langerhans cells.\nNature\n360\n258\n261\n1279441\n17. GattiEVellecaMABiedermannBCMaWUnternaehrerJ\n2000\nLarge-scale culture and selective maturation of human Langerhans cells from granulocyte colony-stimulating factor-mobilized CD34+ progenitors.\nJ Immunol\n164\n3600\n3607\n10725716\n18. GeissmannFProstCMonnetJPDyMBrousseN\n1998\nTransforming growth factor beta1, in the presence of granulocyte/macrophage colony-stimulating factor and interleukin 4, induces differentiation of human peripheral blood monocytes into dendritic Langerhans cells.\nJ Exp Med\n187\n961\n966\n9500798\n19. HerbstBKohlerGMackensenAVeelkenHKulmburgP\n1996\nIn vitro differentiation of CD34+ hematopoietic progenitor cells toward distinct dendritic cell subsets of the birbeck granule and MIIC-positive Langerhans cell and the interdigitating dendritic cell type.\nBlood\n88\n2541\n2548\n8839846\n20. StroblHRiedlEScheineckerCBello-FernandezCPicklWF\n1996\nTGF-beta 1 promotes in vitro development of dendritic cells from CD34+ hemopoietic progenitors.\nJ Immunol\n157\n1499\n1507\n8759731\n21. BorkowskiTALetterioJJFarrAGUdeyMC\n1996\nA role for endogenous transforming growth factor beta 1 in Langerhans cell biology: the skin of transforming growth factor beta 1 null mice is devoid of epidermal Langerhans cells.\nJ Exp Med\n184\n2417\n2422\n8976197\n22. MohamadzadehMBerardFEssertGChalouniCPulendranB\n2001\nInterleukin 15 skews monocyte differentiation into dendritic cells with features of Langerhans cells.\nJ Exp Med\n194\n1013\n1020\n11581322\n23. MeradMManzMGKarsunkyHWagersAPetersW\n2002\nLangerhans cells renew in the skin throughout life under steady-state conditions.\nNat Immunol\n3\n1135\n1141\n12415265\n24. NestleFOZhengXGThompsonCBTurkaLANickoloffBJ\n1993\nCharacterization of dermal dendritic cells obtained from normal human skin reveals phenotypic and functionally distinctive subsets.\nJ Immunol\n151\n6535\n6545\n7504023\n25. ShortmanKNaikSH\n2007\nSteady-state and inflammatory dendritic-cell development.\nNat Rev Immunol\n7\n19\n30\n17170756\n26. LarreginaATMorelliAESpencerLALogarAJWatkinsSC\n2001\nDermal-resident CD14+ cells differentiate into Langerhans cells.\nNat Immunol\n2\n1151\n1158\n11702065\n27. GinhouxFTackeFAngeliVBogunovicMLoubeauM\n2006\nLangerhans cells arise from monocytes in vivo.\nNat Immunol\n7\n265\n273\n16444257\n28. PaluckaAKBanchereauJ\n2006\nLangerhans cells: daughters of monocytes.\nNat Immunol\n7\n223\n224\n16482165\n29. GustafsonJEklundCWallstromMZellinGMagnussonB\n2007\nLangerin-expressing and CD83-expressing cells in oral lichen planus lesions.\nActa Odontol Scand\n65\n156\n161\n17514517\n30. HasseusBJontellMBruneMJohanssonPDahlgrenUI\n2001\nLangerhans cells and T cells in oral graft versus host disease and oral lichen planus.\nScand J Immunol\n54\n516\n524\n11696204\n31. ParoliniSSantoroAMarcenaroELuiniWMassardiL\n2007\nThe role of chemerin in the colocalization of NK and dendritic cell subsets into inflamed tissues.\nBlood\n109\n3625\n3632\n17202316\n32. SantoroAMajoranaARoversiLGentiliFMarrelliS\n2005\nRecruitment of dendritic cells in oral lichen planus.\nJ Pathol\n205\n426\n434\n15714455\n33. ChenYGLuiHMLinSLLeeJMYingSY\n2002\nRegulation of cell proliferation, apoptosis, and carcinogenesis by activin.\nExp Biol Med (Maywood)\n227\n75\n87\n11815670\n34. JonesKLde KretserDMPatellaSPhillipsDJ\n2004\nActivin A and follistatin in systemic inflammation.\nMol Cell Endocrinol\n225\n119\n125\n15451576\n35. PhillipsDJJonesKLClarkeIJScheerlinckJPde KretserDM\n2005\nActivin A: from sometime reproductive factor to genuine cytokine.\nVet Immunol Immunopathol\n108\n23\n27\n16140391\n36. SulyokSWankellMAlzheimerCWernerS\n2004\nActivin: an important regulator of wound repair, fibrosis, and neuroprotection.\nMol Cell Endocrinol\n225\n127\n132\n15451577\n37. AttisanoLWranaJLMontalvoEMassagueJ\n1996\nActivation of signalling by the activin receptor complex.\nMol Cell Biol\n16\n1066\n1073\n8622651\n38. PhillipsDJde KretserDM\n1998\nFollistatin: a multifunctional regulatory protein.\nFront Neuroendocrinol\n19\n287\n322\n9799587\n39. SuginoKKurosawaNNakamuraTTakioKShimasakiS\n1993\nMolecular heterogeneity of follistatin, an activin-binding protein. Higher affinity of the carboxyl-terminal truncated forms for heparan sulfate proteoglycans on the ovarian granulosa cell.\nJ Biol Chem\n268\n15579\n15587\n8340384\n40. HubnerGHuQSmolaHWernerS\n1996\nStrong induction of activin expression after injury suggests an important role of activin in wound repair.\nDev Biol\n173\n490\n498\n8606007\n41. HubnerGWernerS\n1996\nSerum growth factors and proinflammatory cytokines are potent inducers of activin expression in cultured fibroblasts and keratinocytes.\nExp Cell Res\n228\n106\n113\n8892977\n42. BambergerCScharerAAntsiferovaMTychsenBPankowS\n2005\nActivin controls skin morphogenesis and wound repair predominantly via stromal cells and in a concentration-dependent manner via keratinocytes.\nAm J Pathol\n167\n733\n747\n16127153\n43. MunzBSmolaHEngelhardtFBleuelKBrauchleM\n1999\nOverexpression of activin A in the skin of transgenic mice reveals new activities of activin in epidermal morphogenesis, dermal fibrosis and wound repair.\nEmbo J\n18\n5205\n5215\n10508154\n44. WankellMMunzBHubnerGHansWWolfE\n2001\nImpaired wound healing in transgenic mice overexpressing the activin antagonist follistatin in the epidermis.\nEmbo J\n20\n5361\n5372\n11574468\n45. StoitznerPStosselHWankellMHoferSHeuflerC\n2005\nLangerhans cells are strongly reduced in the skin of transgenic mice overexpressing follistatin in the epidermis.\nEur J Cell Biol\n84\n733\n741\n16180311\n46. de GruijlTDSombroekCCLougheedSMOosterhoffDButerJ\n2006\nA postmigrational switch among skin-derived dendritic cells to a macrophage-like phenotype is predetermined by the intracutaneous cytokine balance.\nJ Immunol\n176\n7232\n7242\n16751366\n47. FumagalliMMussoTVermiWScuteraSDanieleR\n2007\nImbalance between activin A and follistatin drives postburn hypertrophic scar formation in human skin.\nExp Dermatol\n16\n600\n610\n17576240\n48. PiekEHeldinCHTen DijkeP\n1999\nSpecificity, diversity, and regulation in TGF-beta superfamily signaling.\nFaseb J\n13\n2105\n2124\n10593858\n49. de CaesteckerM\n2004\nThe transforming growth factor-beta superfamily of receptors.\nCytokine Growth Factor Rev\n15\n1\n11\n14746809\n50. KretzschmarMMassagueJ\n1998\nSMADs: mediators and regulators of TGF-beta signaling.\nCurr Opin Genet Dev\n8\n103\n111\n9529613\n51. Reis e SousaCStahlPDAustynJM\n1993\nPhagocytosis of antigens by Langerhans cells in vitro.\nJ Exp Med\n178\n509\n519\n8393477\n52. LarsenCPSteinmanRMWitmer-PackMHankinsDFMorrisPJ\n1990\nMigration and maturation of Langerhans cells in skin transplants and explants.\nJ Exp Med\n172\n1483\n1493\n2230654\n53. CumberbatchMDearmanRJGriffithsCEKimberI\n2000\nLangerhans cell migration.\nClin Exp Dermatol\n25\n413\n418\n11012599\n54. CerioRGriffithsCECooperKDNickoloffBJHeadingtonJT\n1989\nCharacterization of factor XIIIa positive dermal dendritic cells in normal and inflamed skin.\nBr J Dermatol\n121\n421\n431\n2576222\n55. FivensonDPNickoloffBJ\n1995\nDistinctive dendritic cell subsets expressing factor XIIIa, CD1a, CD1b and CD1c in mycosis fungoides and psoriasis.\nJ Cutan Pathol\n22\n223\n228\n7593815\n56. LenzAHeineMSchulerGRomaniN\n1993\nHuman and murine dermis contain dendritic cells. Isolation by means of a novel method and phenotypical and functional characterization.\nJ Clin Invest\n92\n2587\n2596\n8254016\n57. MeunierLGonzalez-RamosACooperKD\n1993\nHeterogeneous populations of class II MHC+ cells in human dermal cell suspensions. Identification of a small subset responsible for potent dermal antigen-presenting cell activity with features analogous to Langerhans cells.\nJ Immunol\n151\n4067\n4080\n8409386\n58. TurvilleSGCameronPUHandleyALinGPohlmannS\n2002\nDiversity of receptors binding HIV on dendritic cell subsets.\nNat Immunol\n3\n975\n983\n12352970\n59. SugermanPBSavageNWWalshLJZhaoZZZhouXJ\n2002\nThe pathogenesis of oral lichen planus.\nCrit Rev Oral Biol Med\n13\n350\n365\n12191961\n60. VermiWFacchettiFRiboldiEHeineHScuteraS\n2006\nRole of dendritic cell-derived CXCL13 in the pathogenesis of Bartonella henselae B-rich granuloma.\nBlood\n107\n454\n462\n16189275\n61. VulcanoMStruyfSScapiniPCassatellaMBernasconiS\n2003\nUnique regulation of CCL18 production by maturing dendritic cells.\nJ Immunol\n170\n3843\n3849\n12646652\n62. de GruijlTDLuykx-de BakkerSATillmanBWvan den EertweghAJButerJ\n2002\nProlonged maturation and enhanced transduction of dendritic cells migrated from human skin explants after in situ delivery of CD40-targeted adenoviral vectors.\nJ Immunol\n169\n5322\n5331\n12391253\n63. AroraDSCookeIEGanesanTSRamsdaleJManekS\n1997\nImmunohistochemical expression of inhibin/activin subunits in epithelial and granulosa cell tumours of the ovary.\nJ Pathol\n181\n413\n418\n9196439"
+}

batch_3/PMC2533402.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533402",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533402\nAUTHORS: Laura Savolainen, Liana Pusa, Hwa-Jung Kim, Heidi Sillanpää, Ilkka Seppälä, Tamara Tuuminen\n\nABSTRACT:\nBackgroundIn recent years T cell based interferon gamma release assays (IGRA) have been developed for immunodiagnosis of M. tuberculosis infection. At present these assays do not discriminate between disease and latency. Therefore, more promising antigens and diagnostic tools are continuously being searched for tuberculosis immunodiagnostics. The heparin binding hemagglutinin (HBHA) is a surface protein of M. tuberculosis which promotes bacterial aggregation and adhesion to non-phagocytic cells. It has been previously assumed that native, methylated form of this protein would be a promising antigen to discriminate latent from active infection.Methodology and Principal FindingsWe performed a pilot investigation to study humoral and T-cell mediated immunological responses to recombinant HBHA produced in M. smegmatis or to synthetic peptides in patients with recent or past tuberculosis, with atypical mycobacteriosis, or in healthy vaccinated individuals. The T cell reactivities to HBHA were compared to the respective reactivities towards Purified Protein Derivative (PPD) and two surface secreted proteins, ie. Early Secretory Antigen Target-6 (ESAT-6) and Culture Filtrate Protein-10 (CFP-10). Our pilot results indicate that methylated recombinant HBHA induced a strong T cell mediated immune response and the production of IgG and IgM-class antibodies in all patient groups, most surprisingly in young Finnish vaccinees, as well. We observed a positive correlation between the reactivities to HBHA and non-specific PPD among all studied subjects. As expected, ESAT-6 and CFP-10 were the most powerful antigens to discriminate disease from immunity caused by vaccination.ConclusionsOn the basis of results of this exploratory investigation we raise concerns that in countries like Finland, where BCG vaccination was routinely used, HBHA utility might not be sufficient for diagnostics because of inability to explicitly discriminate tuberculosis infection from immunoreactivity caused by previous BCG vaccination.\n\nBODY:\nIntroductionNewly introduced T cell based interferon gamma release assays (IGRA) are of a great diagnostic importance in distinguishing with high specificity the persons who are infected with Mycobacterium tuberculosis\n[1]. These assays, however, do not discriminate between disease and latent tuberculosis infection (LTBI) [2]. This limitation lowers their clinical utility e.g. in aging persons originating from a country with a previously high tuberculosis burden who present with pulmonary infiltrates or other manifestations suggestive of tuberculosis reactivation. Therefore, more promising antigens or diagnostic algorithms are continuously being searched to improve the current armament for tuberculosis (TB) immunodiagnostics. Mycobacterium tuberculosis heparin binding hemagglutinin (HBHA) is a virulence factor that promotes bacterial aggregation, adhesion to the heparan sulphate proteoglycans of nonphagocytic cells, and dissemination of tubercle bacilli from the lungs to other tissues in patients suffering from tuberculosis [3], [4]. Locht et al. [3] found that latently infected humans mount a strong Th1-type immune response to HBHA, whereas patients with active disease do not. Moreover, patients with active tuberculosis may develop a strong humoral response to native methylated HBHA [5], [6]. The diagnostic utility of HBHA-based interferon gamma release assays (IGRAs) has been further evaluated in Belgium [7], a country with low tuberculosis (TB) incidence where bacille Calmette et Guérin (BCG) vaccinations are rarely used. In that work, HBHA-based IGRA was not influenced by prior BCG vaccination and was significantly more sensitive than Early Secretory Antigen Target-6 (ESAT-6)-based IGRA.Because some earlier studies [5], [7] concluded that previous BCG vaccination does not influence the assay performance we wondered whether HBHA-based methods would retain their diagnostic potential also in Finland. Our country currently has a low incidence of TB (6,1/100 000 [8]) but our population bore a high TB burden as recently as four decades ago. Moreover, since early fifties till September 2006, 98% of the Finnish population have been vaccinated. The rationale for this exploratory investigation was to compare immunological responses to the recombinant methylated HBHA produced in Mycobacterium smegmatis\n[9], methylated synthetic peptides from HBHA, Purified Protein Derivative (PPD), and the proteins with the documented high specificity for M. tuberculosis infection ie. ESAT-6 and Culture Filtrate Protein Derivative (CFP-10). For this pilot study we enrolled a few tuberculosis (TB) patients with a variable degree of disease activity and healthy young and middle-age Finnish vaccinees who were practically free from any previous contact with M. tuberculosis and thus served as a valid control group.MethodsHuman subjects and disease definitionsFor this study we enrolled the subjects as follows:a) Active TB groupThose subjects in whom the symptoms of the disease started within one month, who were hospitalized, anti-tuberculosis treatment was commenced within 2–3 weeks and who had a recent positive bacteriological verification. In this group four presented with pulmonary tuberculosis, one with tuberculosis of gastrointestinal tract and one patient had tuberculosis of cervical vertebra. In these subjects the blood samples were obtained within the routine laboratory investigation.b) Inactive TB groupThis group consisted of outpatients who presented with no or mild symptoms and who were either partially treated for TB with PAS, streptomycin, and INH or thoracotomy in forties and fifties, in whom chest X-ray findings indicated lack of an active process. Patients with vertebral TB had destructive processes of their vertebra at early childhood. At adulthood their TB was confirmed by radiology and the thick needle aspiration biopsy that ruled out other diseases. At the time of investigation these patients did not receive anti-tuberculosis treatment. Culture or nucleic acid amplification analysis (NAA) were either negative or bacteriological analysis was not performed.c) Disease control groupFour patients with pulmonary manifestations in whom non-tuberculous mycobacterial (NTM) infection was diagnosed by a positive culture isolation. At the time of investigation these subjects were outpatients and their symptoms were mild (mainly cough).d) Healthy control groupFinnish-born University students (mean age 25 years) who on the interview did not admit any previous contact with a TB patient and thus served as an ideal negative control, and Finnish-born laboratory personnel (mean age 50 years), were enrolled. All subjects of this group were born after the routine BCG vaccination campaign was implemented in Finland. Repeated vaccinations were discontunued in Finland in 1990.Ethical considerationsFrom the subjects of the group a) a verbal informed consent was obtained. The verbal informed consent was sufficient because the patient–doctor relationship was earlier established (Dr. Pusa), the patients attended medical care for the routine doctor's check-up and venipuncture was a part of their current medical treatment. From the groups b), c) and d) a written informed consent was obtained before venipuncture. This study was approved by the Ethical Committee of the University Hospital of Helsinki, Internal Disease Department (Drno. 232/E5/07).Sample processingWhole blood was withdrawn and divided into portions. From one portion sera were prepared by conventional methods and frozen at −20°C until use. From another portion peripheral blood mononuclear cells (PBMC) were isolated by Ficoll Paque gradient (Amersham Biosciences Inc., Piscataway, USA) and frozen in the CTL (Cellular Technology Ltd., Cleveland, USA) media in a liquid nitrogen until use. The characteristics of the individuals enrolled in this study are presented in Table 1.10.1371/journal.pone.0003272.t001Table 1Characteristics of the individuals enrolled in the study.Patient groupsELISPOT (n)EIA (n)EthnicityAge range (yrs)GenderBacteriologyDiagnosisActive TB64Finns, n = 3 Persons from endemic area, n = 3 (2 with extrap. TB)31–82Female 2/6All bacterilogically confirmed: culture+ n = 5, of those, acid faststaining+ n = 3 NAA+ n = 1Tuberculosis: Pulmonary n = 4 GI channel n = 1 Cervical vertebra n = 1Inactive TB59Finns, n = 8 One person from an endemic area (extrap. TB)41–83Female 4/9At the time of examination culture and NAA negative or bacteriological analysis was not performedPartially treated for pulmonary TB n = 7 Partially treated/untreated TB of lumbar vertebra diagnosed radiologically n = 2Atypical mycobacteria4All Finns66–75All Female\nM. avium n = 2 M. intracellulare n = 1 M. abscessus n = 1co-morbidity with asthma, COPDVaccinated subjects students laboratory personnel7 84All Finns22–59 mean 25 mean 50Female 6/7 8Synthetic peptidesTwenty three 15-mer sequential peptides overlapping by nine amino acids and spanning the genomic sequence of the M. tuberculosis (H37Rv) HBHA-protein starting from 37th amino acid were synthesized (Proimmune, Oxford, UK; Alta Bioscience, Birmingham, UK). The average purity of the peptides was ∼73%. Three peptides from the protein C-terminus were chemically methylated at their lysine residues (see Legend of Fig. 3). The methylation was done on the PEPscreen synthesis platform. Fmoc-Lys(Me,Boc)-OH were pre-dissolved at a 0.5 M concentration, placed on the deck, coupled and deprotected in the same way as standard amino acids. Additionally, the peptides for serology were biotinylated at their N-terminus. The quality control for all peptides was accomplished by mass spectrometry analysis. The peptides were dissolved in sterile asetonitrile and stored at −20°C in aliquots of 1–4 mg/ml with ∼30% glycerol for serology and in PBS for the T-cell assays.Recombinant HBHA antigenRecombinant methylated HBHA was expressed in M. smegmatis (rMtb-HBHA). The pMV3-38 plasmid that contained the full-length HBHA open reading frame was kindly provided by Dr.G. Delogu (University of Sassari, Sassari, Italy) [9]. The protein from the bacterial extract was primarily purified by phosphocellulose chromatography, i.e., cation exchange, because most mycobacterial proteins do not bind to this resin at pH 7.0, and then further purified by Ni-NTA chromatography [10]. This two-step purification of rMtb-HBHA was highly effective. Contaminating bands on PAGE gels stained with Coomassie blue were not observed when 5–10 µg of the purified protein was loaded. The purified protein was stored at −20°C in aliquots until use in serology and T cell analysis.T ELISPOT analysisT-cell reactivity to ESAT-6, CFP-10 (the synthetic peptides were from T-SPOT ®-TB kit, Oxford Immunotec, Oxford, UK), PPD (Statens Seruminstitut, Copenhagen, Denmark) and rMtb-HBHA was assessed by Enzyme-Linked ImmunoSpot assay (Mabtech Inc. Cincinnati, USA). Peripheral blood mononuclear cells (PBMC) were purified from fresh blood samples by density gradient centrifugation and stored with CryoABC Kit Freezing media (Cellular Technology Ltd.) in liquid nitrogen until use. After cells were washed with RPMI (HaartBio, Helsinki, Finland) and resuspensed in the CTL Test Media (Cellular Technology Ltd.), the cell count was performed by blood count-analyzer (ADVIA-60 Closed Tube Automated Haematology System, Bayer, Germany). The cells were diluted at 2.5×106 PBMC/ml in CTL Test Media (Cellular Technology Ltd.). 250 000 PBMCs/well were stimulated in ELISPOT-plates in the presence of synthetic peptide-pool (10 µg/ml each), PPD (10 µg/ml), rMtb-HBHA (25 µg/ml). For viability test 50 000 PBMC/well were stimulated with phytohaemagglutinin (PHA) (Oxford Immunotech). CTL Test Media was used as a negative control. The plates were incubated at +37°C with 5% CO2 for 48 h. Thereafter the plates were washed and the analysis was performed according to the manufacturer's instruction. The spots were counted the ELISPOT-reader (Biosys, Lyngby, Denmark) and the net values were calculated by subtracting the readings of the media control. Spot-sizes and the cytokine values were examined with the ELISPOT-reader, AID EliSpot Software Version 4.0 (AID GmbH, Strasburg, Germany). The viability control test for the HBHA synthetic peptides was performed by exposing PBMC to the CFP-10 peptide mixture with asetonitrile added at concentrations of 15% and 30% (vs. only 2.5% in the final peptide solution). The addition of asetonitrile did not violate the reactivity to the CFP-10 antigen (data not shown).IgG and IgM determinationsThe presence of IgG- and IgM-class antibodies to synthetic peptides, rMtb-HBHA and PPD were determined by enzyme-linked immunosorbent assay (ELISA). Synthetic peptides from Borrelia burgdorferi VlsE-protein IR6 region and positive serum samples from Lyme disease patients were used as the method control [11]. 96-well microplates (Microlon high binding, Greiner, Frickenhausen, Germany) were coated either with streptavidin (Roche, Mannheim, Germany) in PBS, pH 7.5 (100 ng/well), or PPD in PBS, pH 7.5 (1 mg/well), or with rMtb-HBHA in 0.1 M bicarbonate buffer, pH 9.5 (250 ng/well) overnight at +4°C. The plates were blocked with 0.25% Human serum albumin (HSA) (Finnish Red Cross, Helsinki, Finland) in PBS for 1 h at +37°C and washed four times with PBS containing 0.05% Tween20. Biotinylated synthetic peptides (500 ng/well) and IR6 (200 ng/well) in PBS, pH 7.5 were incubated for 2 h at room temperature. Plates were washed as above and the human sera diluted 1∶100 in PBS containing 0.5% HSA, 10% FCS and 0.1% Tween20 and were incubated for 2 h at room temperature. After wash Alkaline phosphatase-conjugated anti-Human IgG or IgM antibody (Jackson ImmunoResearch, W. Baltimore, USA) was added at 1∶5 000 dilution in PBS containing 0.5% HSA and 0.1% Tween20. After 2 h of incubation at room temperature the plates were washed four times. 4-nitrophenylphosphate (Boehringer Mannheim, Germany), 1 mg/ml in 0.1 M diethanolamine-0.5 mM MgCl2 was added to each well and the reaction was stopped after 15 minutes with 0.1 M NaOH. The plates were read at 405 nm with iEMS Reader MF (Labsystems, Helsinki, Finland). IgM rheumatoid factor was controlled for all serum samples by immunonephelometric method at the HUSLAB, Unit of Immunology.Statistical analysisThe correlation was analyzed by the non-parametric Spearman test. The Receiver operating characteristic (ROC) curve analysis and the Area under the curve (AUC) with respective 95% confidence intervals (CI) for each antigen were calculated with the GraphPad Prism version 4.0 (GraphPad Software, Inc. San Diego, CA).The amino acid sequences of the HBHA from M. tuberculosis and M. bovis were searched with the Entrez Protein search engine and then the sequences were aligned with the Needle program, EBLOSUM62-Matrix (EMBOSS, [12]).ResultsCell-mediated immunityEmploying both techniques for immune reactivity study of rMtb-HBHA, we observed considerable interindividual variation in all studied groups (Fig. 1A–C and Fig. 3 A–B; D–F). These observations hold true irrespective of the measured parameters of immunoreactivity; neither the numbers of reactive cells, nor the amount of the IFNγ released, nor the optical densities in the IgG and IgM EIAs were discriminatory when comparing the groups. For example, the highest values of 430, 387, and 360 reactive cells/106 lymphocytes were observed in the groups of active TB, inactive TB and healthy vaccinees, respectively. Strikingly, half of the vaccinated persons reacted strongly against rMtb-HBHA and the reactivities were equally high as those of the TB-patients (Fig. 1A). When the comparisons between the groups were done by semiquantitative calculation of the amount of the produced cytokine, ie. the size and the intensity of the spot, the highest median value of the cytokine production was found in the vaccinated group (Fig. 1B). Persons with positive environmental mycobacteria culture result reacted similarly as the persons from other groups with a high interindividual variation (Fig. 1A and B). Positive correlation was observed then the reactivities to rMtb-HBHA and PPD, an indicator of vaccination or infection, were compared for all the tested samples (Fig. 1D). One healthy 60-year-old individual who was not vaccinated in his childhood showed humoral and cell-mediated immune responses to rMtb-HBHA as measured by both EIAs and ELISPOT assay, indicating immunization with so-called atypical mycobacteria. Noteworthy, this person had a negative tuberculin skin test and his lymphocytes did not recognise ESAT-6 and CFP-10 peptide mixtures in ELISPOT (data not shown). Furthermore, the obtained reactivities were reproducible and the measurements did not exceed the expected between-run imprecisions of CV% ≤15 and ≤40% for EIAs and ELISPOT, respectively. When the frequencies of reactive cells were compared between persons of different ages, no differences were noticed either (Fig. 1C), indicating that there was no waning of immunological memory towards HBHA with age. In other words, ELISPOT analysis produced similar patterns of reactivities between the studied groups with 2 out of 5; 3 out of 6; and 7 out of 15 being strong reactors in the groups of inactive TB, active TB, and vaccinees, respectively (Fig. 1A).10.1371/journal.pone.0003272.g001Figure 1The ability of rMtb-HBHA and PPD to induce the production of IFNγ was tested in the ELISPOT technique.The groups of patients with inactive TB (n = 5); active TB (n = 6), healthy young (n = 7) and middle-age (n = 8) vaccinated subjects and patients with isolation of so-called atypical mycobacteria (n = 4) and were enrolled. Comparison of the cell-mediated immunological responses to rMtb-HBHA was performed as the determination of the number of reactive cells per106 lymphocytes (A) and as the measurement of IFNγ production activities expressed as arbitrary units (B). Cell-mediated immunological responses in healthy vaccinated individuals was studied by division by age into two groups (C). The correlation of rMtb-HBHA with the PPD ELISPOT reactivities expressed as the frequencies of reactive cells per106 is presented in (D). Data are shown as individual reactivities; the horizontal bars represent arithmetic median values.As expected, when tested concurrently with a mixture of specific peptides derived from ESAT-6 and CFP-10 of M. tuberculosis, samples from patients with active and inactive TB did react in the ELISPOT assay, whereas samples from all healthy individuals, did not (Fig. 2 A and B).10.1371/journal.pone.0003272.g002Figure 2Production of IFNγ by lymphocytes stimulated with ESAT-6 (Panel A) and CFP-10 (Panel B) was tested in the ELISPOT technique.The groups of patients with inactive TB (n = 5), active TB (n = 6) and healthy vaccinated subjects (n = 13) were tested. Cell-mediated immunological responses were determined as the number of reactive cells per 106 lymphocytes. The lower dotted line is the level of positivity suggested by the manufacturer, the upper dotted line is the level of positivity adopted at the HUSLAB diagnostic laboratory. The area between the two dotted lines represents the so-called grey-zone, an area of uncertainty for interpretation that was calculated based on assay imprecision (data not published).Humoral immunity to HBHAHigh interindividual variations were observed in serology, ie. 3 out of 9 (Inactive TB) and 2 out of 4 (Active TB) were strong reactors in the IgG and IgM EIA whereas almost all vaccinees showed moderate to high reactivities in IgG and IgM EIAs (Fig. 3A–B, D–F). Noteworthy, the median value of IgM antibodies to rMtb-HBHA was the highest in the group of healthy BCG-vaccinated individuals. Synthetic peptides were not recognised by any of the tested sera in the IgG EIA (data not shown). When tested for IgM rheumatoid factor it was detectable only in one person in the TB-patients group. Interestingly, IgM-class antibodies were detectable in almost all of the individuals in the study, with an exception of two non-vaccinated infants (data not shown). The IgM-class antibodies recognised not only rMtb-HBHA but also the three 15-mer linear methylated peptides from the HBHA C-terminus (Fig. 3D–F). We believe that IgM-class antibodies alone reacting to the C-terminal peptides of HBHA probably indicate a non-specific reaction arising from heterophilic antibodies, for example, or immunological cross-reactivity with conserved sequences of environmental bacteria (e.g., rhodococcus, http://www.ncbi.nlm.nih.gov/blast/Blast.cg).10.1371/journal.pone.0003272.g003Figure 3The serological reactivities to synthetic peptides, rMtb-HBHA and PPD were tested in IgG and IgM ELISAs.The groups of patients with inactive TB (n = 9), active TB (n = 4) and healthy vaccinated individuals (n = 4) were enrolled. Serological responses to rMtb-HBHA measured by the IgG EIA (A), IgM EIA (B), and IgG correlation in the PPD and rMtb-HBHA ELISAs (C). IgM responses to methylated 15-mer linear peptides from the C terminus of HBHA: IELPKKAAPA[KMe]KAAP (D), AAPAKKAAPA[KMe]KAAA (E), and AAPAKKAAPA[KMe][KMe]AAA (F). Individual responses are presented as optical densities (OD405). The horizontal bars represent arithmetic median values.Comparison of the discriminatory power of rMtb-HBHA, PPD, ESAT-6 and CFP-10ROC curves were constructed for all tested antigens and the AUCs with respective 95% CI were compared for each tested antigen (Fig. 4A–D). As expected, ESAT-6 and CFP-10 produced ROC curves acceptable for diagnostics with AUCs ranging from 0.947 to 0.972 and a narrow 95% CI (0.84–1.052) (Fig. 4C–D). On the contrary, the AUC for rMtb-HBHA (0.636; 95% CI 0.391–0.886) was comparable to the one for PPD (0.736; 95%CI 0.531–0.941) (Fig. 4A–B) indicating in practice no discriminatory power between healthy vaccinees and persons with a TB infection. As a consequence of high interindividual variation in rMtb-HBHA immunoassays, the confidence interval for the respective AUC was wide.10.1371/journal.pone.0003272.g004Figure 4ROC curves were constructed to compare ELISPOT results when the cells were stimulated with rMtb-HBHA (A), PPD (B), and peptide mixtures of ESAT-6 and CFP-10 (C–D).The curves were established for infected (Active and Inactive TB) and the healthy control group. The calculated AUC and the respective confidence intervals (in brackets) are shown for each tested antigen.DiscussionThe potential of HBHA for diagnostics have been recently reported [3], [5], [6], [7], [10], [13]. In this pilot investigation we attempted to have an insight into the usefulness of this antigen for diagnostics in our fully vaccinated population. We aimed to compare immunoreactivity to rMtb-HBHA and HBHA methylated synthetic peptides with the immunoreactivity to the conventional PPD that have been used for decades with poor success because of inability to discriminate TB infection from immunity caused by vaccination. For comparison, we used two other secreted antigens, namely ESAT-6 and CFP-10. These antigens have been proven highly immunogenic and specific for M. tuberculosis infection and are absent from the majority of non-tuberculous mycobacteria and BCG substrain [for review see 1]. The major objective was to investigate whether immunological responses in persons with a proven contact with M. tuberculosis (Active and Inactive TB groups) would be quantitatively and qualitatively different from vaccinated individuals. In other words, we were interested to study the specificity of immune response to HBHA as a nominator of infection. The minor objective was to see whether immunological responses would differ in TB patients with a different degree of disease severity. We did not enrol a group of patients with a proven LTBI which may be considered as a limitation of our study. On the contrary, we enrolled young Finnish vaccinees that were unambiguously interpreted as free from LTBI and some patients with proven NTM infection. In our opinion, these two groups is a strength of our investigation. For analysis we used ELISPOT and EIA techniques.The rMtb-HBHA antigen used in this study was prepared as described by Delogu et al [9]. Using a limited clinical material we were not able to affirm the discriminatory power of rMtb-HBHA in serology. Zanetti et al [6] who referred to the same method of the recombinant HBHA expressed in M. smegmatis\n[9], showed that only forty-four out of 111 sera with active TB produced optical densities over the presumed cut-off level of 0.5 OD(405nm). In their study, surprisingly, the combined group of vaccinated people and patients with presumed latent tuberculosis infection did not produce IgG reactivities above the cut-off level. Using a larger cohort size than we did, Zanetti et al. [6] demonstrated a trend towards higher frequencies of responders in IgG serology in patients with active TB, however only 1/3 of the tested subjects were classified as serology positive. It seems however, that this frequency of positive results would not satisfy diagnostic needs. In our investigation we were able to detect IgM antibodies that recognized rMtb-HBHA and synthetic peptides of the C-terminus practically from all the tested individuals and also from one 60-year-old person who has never been vaccinated nor has a LTBI (data not shown). In the study of Shin et al. [10] who used immunoblot and EIA techniques to investigate IgM reactivity to rMtb-HBHA, the antibodies were detectable in early and chronic TB patients whereas healthy students, the controls, were non-responders. In their study, however, the vaccination status of the controls was not reported, therefore we are left with uncertainty how would this antigen be recognised in BCG vaccinees.Masungi et al. [5] used HBHA that was purified from a BCG substrain. As others, they also reported that negative control subjects and the BCG vaccinees did not produce anti-HBHA antibodies at detectable level and the reactivity towards HBHA in lymphocyte stimulation assay was non-significant. On the contrary, when lymphocytes were stimulated with PPD, the controls and the BCG vaccinees reacted strongly [5]. In the most recent study of Hougardy et al. [7] HBHA was also purified from a BCG substrain and the cell-mediated immunity was studied as Masungi et al. by lymphocyte stimulation and a subsequent measurement of secreted interferon gamma (IFNγ). In their study the authors used control subjects who had no history of a TB contact but in whom half were BCG vaccinated. Quite opposite to our findings they observed the steepest ROC when lymphocytes were stimulated with HBHA while the stimulation with ESAT-6 produced the least discriminatory ROC.Methylation of HBHA is crucial for effective T-cell immunity. Thus, native HBHA purified from BCG or M. tuberculosis H37Ra evoked a much stronger CD8+ and CD4+ mediated immune responses than the recombinant non-methylated HBHA produced in E. coli\n[13]. However, as indicated, recombinant M. smegmatis expression host was able to produce HBHA with a methylation pattern very similar to that of the native HBHA, as was evidenced by mass spectrometry analysis, amino acid sequencing, electrophoresis and recognition by the same monoclonal antibody [14]. There is however a possibility that folding of native HBHA produced by BCG or M. tuberculosis and of recombinant HBHA produced in M. smegmatis may be different which would cause minor differences in their immunoreactivities. The protein used in our study, rMtb-HBHA, was expressed in M. smegmatis and all attempts were made to avoid potential contamination that might have blurred the results. While it could be argued that minor M. smegmatis genome coded contaminant antigens might have caused the observed T cell responses in our study, the antibody responses (IgM) to the synthetic peptides support the view of weak specificity of the rMtb-HBHA responses in BCG vaccinated persons. In view of a longevity of immunological memory even without boosting [15], [16] it is not unexpected that our healthy vaccinees recognised rMtb-HBHA that has a 95.5% amino acid sequence homology to BCG HBHA [12].Results of immunological studies with a battery of synthetic peptides comprising mono- and dimethylated peptides of the C- terminus of HBHA were very disappointing. In this study IgM-class reactivities were detectable but in a very unpredictable fashion while IgG-class antibodies did not recognize peptides even methylated ones in any of the patient groups. The peptides did not evoke any T cell mediated response. Noteworthy, our study protocol confirmed that the non-responsiveness to the peptides was not attributed to the toxicity of asetonitrile used in the peptides solution. Temmerman et al [13] studied T cell immunity with a peptide scan analysis first with non-methylated peptides and then probing a methylated peptide. Only one non-methylated peptide produced some IFNγ in a portion of LTBI patients. The reproducibility of this finding was not, however, reported. In their hands the methylated peptide induced also some IFNγ production but only in combination with the recombinant protein. The authors speculated that the methylated peptide might need a protein carrier. We too have no solid explanation why synthetic peptides behaved so differently in cell mediated and humoral immunity studies compared to the recombinant protein. In fact, we know yet little about how the HBHA is chopped in the lysosome and how the methylated antigens are presented by MHC II molecules.Because our cohorts were too small to be able to pick-up minor differences in reactivity between the studied subjects we chose to use ELISPOT, the most sensitive and functional technique to study cell-mediated immunity. We assessed not only the frequencies of reactive lymphocytes (Fig. 1A) but the production of IFNγ as well (Fig. 1B). In this way we tried to do the analysis as close as feasible to that of Masungi et al. and Hougardy et al. However, in neither of the analyses could we prove HBHA superiority over the well-established antigens for diagnostics, namely ESAT-6 and CFP-10. As expected, these latter antigens possessed sufficient discriminatory power to separate infected individuals from vaccinated persons (Fig. 2A–B). The ROC analysis with a narrow 95% CI confirmed this conclusion (Fig. 4A–D). In fact, recent study by Chee et al. [17] showed that the two commercial methods for TB immunodiagnostics produced a poor agreement (κ = 0.257) when testing 270 patients with pulmonary TB. One was an ELISPOT- and the other was an EIA-based method. The authors speculated that the differences in results may be attributed to heterophilic antibody effects, non-specific IFNγ in the blood samples and lack of standardised lymphocyte counts in EIA-based technology compared to ELISPOT. The factors mentioned by Chee et al. [17] may also contribute to controversy between our pilot study and earlier studies [5], [7] where EIA-based methods have been applied.This pilot study casts some concerns about the possibility to generalize about diagnostic potential of HBHA. We have observed that samples from healthy individuals in a country with almost complete vaccination coverage and even from individuals who have never been vaccinated may exhibit immune reactivities to HBHA, indicating that natural immunization to this protein or to cross-reactive peptides may occur. Therefore, on the basis of i) the presence of closely related antigen in a BCG substrain; ii) observed good correlation with PPD in serology and cell-mediated assays, iii) comparable ROC analysis for HBHA and PPD; iv) evidence of reactivities of vaccinees without any previous risk of contraction of M. tuberculosis infection, HBHA by no means is superior to ESAT-6 and CFP-10. In our environment HBHA practically does not add to tuberculosis immunodiagnostics. In conclusion, our results emphasize that the search for new more promising antigens for TB diagnostics should continue.\n\nREFERENCES:\n1. MenziesDPaiMComstockG\n2007\nMeta-analysis: New tests for the diagnosis of latent tuberculosis infection: Areas of uncertainly and recommendations for research.\nAnn Intern Med\n146\n340\n350\n17339619\n2. DhedaKUdwadiaZFHuggetJFJohnsonMARookGAW\n2005\nUtility of the antigen-specific interferon-γ assay for the management of tuberculosis.\nCurr Opin Pulm Med\n11\n195\n202\n15818179\n3. LochtCHougardyJ-MRouanetCPlaceSMascartF\n2006\nHeparin-binding hemagglutinin, from an extrapulmonary dissemination factor to a powerful diagnostic and protective antigen against tuberculosis.\nTuberculosis\n86\n303\n309\n16510310\n4. PetheKAlonsoSBietFDeloguGBrennanMJ\n2001\nThe heparin-binding haemagglutinin of M. tuberculosis is required for extrapulmonary dissemination.\nNature\n412\n190\n193\n11449276\n5. MasungiCTemmermanSVan VoorenJ-PDrowartAPetheK\n2002\nDifferential T and B cell responses against Mycobacterium tuberculosis heparin-binding hemagglutinin adhesion in infected healthy individuals and patients with tuberculosis.\nJ Infect Dis\n185\n513\n520\n11865404\n6. ZanettiSBuaADeloguGPuscedduCMuraM\n2005\nPatients with pulmonary tuberculosis develop a strong humoral response against methylated heparin-binding hemagglutinin.\nClin Diagn Lab Immunol\n12\n1135\n1138\n16148186\n7. HougardyJ-MSchepersKPlaceSDrowartALechevinV\n2007\nHeparin-binding-hemagglutinin-induced IFN-γ release as a diagnostic tool for latent tuberculosis.\nPLoSone\n10\ne926\n8. National Public Health Institute\n2008\nInfectious diseases in Finland 2007. Publications of the National Public Health Institute B9 ISBN 978-951-740-812-7.\nhttp://www.ktl.fi/attachments/suomi/julkaisut/julkaisusarja_b/2008/2008b09.pdf. Accessed 2008 June 23\n9. DeloguGBuaAPuscedduCParraMFaddaG\n2004\nExpression and purification of recombinant methylated HBHA in Mycobacterium smegmatis.\nFEMS Microbiol Lett\n239\n33\n39\n15451098\n10. ShinA-RLeeK-SLeeJ-SKimS-YSongC-H\n2006\n\nMycobacterium tuberculosis HBHA protein reacts strongly with the serum immunoglobulin M of tuberculosis patients.\nClin Vaccine Immunol\n13\n869\n875\n16893986\n11. SillanpääHLahdenneJSarvasHArnezMSteereA\n2007\nImmune responses to borrelial VlsE IR6 peptide variants.\nInt J Med Microbiol\n297\n45\n52\n17234451\n12. LabargaAValentinFAnderssonMLopezR\n2007\nWeb Service at the European Bioinformatics Institute.\nNucleic Acids Research Web Service Issue 2007\n13. TemmermanSPetheKParraMAlonsoSRounetC\n2004\nMethylation-dependent T cell immunity to Mycobacterium tuberculosis heparin-binding hemagglutinin.\nNature medicine\n10\n935\n941\n14. PetheKBibaniPDrobecqHSergheraertCDebieA-S\n2002\nMycobacterial heparin-binding hemagglutinin and laminin-binding protein share antigenic methyllysines that confer resistance to proteolysis.\nPNAS\n99\n10759\n64\n12149464\n15. TuuminenTKekäläinenEMäkeläSAla-HouhalaIEnnisFA\n2007\nHuman CD8+ T cell memory generation in Puumala hantavirus infection occurs after the acute phase and is associated with boosting of EBV-specific CD8+ memory T cells.\nJ Immunol\n3\n1988\n95\n16. van EppsHTerajimaMMustonenJArstilaTPCoreyEA\n2002\nLong-lived Memory T lymphocytes responces after hantavirus infection.\nJ Exp Med\n196\n579\n588\n12208874\n17. CheeCBEGanSHKhinMarKVBarkhamTMKohCK\n2008\nComparison of sensitivities of two commercial gamma interferon release assays for pulmonary tuberculosis.\nJ Clin Microb\n46\n1935\n40"
+}

batch_3/PMC2533403.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533403",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533403\nAUTHORS: Pierre Corbeau\n\nABSTRACT:\nIn this review, a quick presentation of what interfering RNA (iRNA) are—small RNA able to exert an inhibition on gene expression at a posttranscriptional level, based on sequence homology between the iRNA and the mRNA—will be given. The many faces of the interrelations between iRNA and viruses, particularly HIV, will be reviewed. Four kinds of interactions have been described: i) iRNA of viral origin blocking viral RNA, ii) iRNA of viral origin downregulating cellular mRNA, iii) iRNA of cellular origin (microRNA) targeting viral RNA, and iv) microRNA downregulating cellular mRNA encoding cell proteins used by the virus for its replication. Next, HIV strategies to manipulate these interrelations will be considered: suppression of iRNA biosynthesis by Tat, trapping by the HIV TAR sequence of a cell component, TRBP, necessary for iRNA production and action, and induction by the virus of some microRNA together with suppression of others. Then, we will discuss the putative effects of these mutual influences on viral replication as well as on viral latency, immune response, and viral cytopathogenicity. Finally, the potential consequences on the human infection of genetic polymorphisms in microRNA genes and the therapeutic potential of iRNA will be presented.\n\nBODY:\nIntroductionThe discovery that cells produce small RNAs called interfering RNAs (iRNAs) that are able to inhibit in a sequence-specific way gene expression at a post-transcriptional level, a phenomenon called RNA interference, is a recent breakthrough in biology. This discovery has already had and will have many consequences on our knowledge of cell physiology and pathology, on our ability to regulate protein production at the laboratory, using tailored small iRNAs (siRNAs), and soon in medicine. Cohabitation of iRNAs physiologically produced by human cells, called microRNAs (miRNAs), and viral RNAs in the same cell results in interactions that may have an impact on virus replication, host cell physiology, and anti-microbial immune response. In this article, we will describe briefly the miRNA machinery, review the various connections existing between iRNAs and HIV RNAs, evaluate their consequences for both actors, and consider how we could interfere with these connections in order to regulate virus infection.Human Cellular miRNAsmiRNAs are transcribed by intergene regions, introns, and even exons, either as polycistronic transcripts if they are clustered or as monocistronic transcripts if they are not [1]–[3]. These transcripts, called pri-miRNAs, are imperfect RNA hairpins of hundreds to thousands of base pairs [4]. They are processed in the nucleus by the Rnase III endonuclease Drosha to a stem loop structure of about 60 base pairs, the pre-miRNA (Figure 1) [5]. The transfer of pre-miRNA from the nucleus to the cytoplasma is mediated by the nuclear export factor exportin 5 [6]. There, pre-miRNAs are cleaved by another Rnase III endonuclease, Dicer [7]. Dicer delivers an approximately 22–base pair duplex. One strand of this duplex, the mature miRNA, still bound to Dicer, is driven towards ribosome-free compartments of the cytoplasma called P-bodies (processing bodies) by an association of molecules called RISC (RNA-induced silencing complex), which includes the endonuclease Argonaute-2 (Ago-2) [8]. The targets of the miRNA-loaded RISC are the RNAs presenting with sequence homology with nucleotides 2–7 in the 5′ portion of the miRNA. Most of the time, the consequence is an inhibition of translation of the mRNA, and sometimes, particularly if the match between the mRNA and the miRNA is perfect, the mRNA is cleaved. Of note, TAR RNA–binding protein (TRBP), a cell protein initially discovered for its capacity to bind to the TAR sequence of HIV RNA, also binds to Dicer and Ago-2, and is necessary for the maturation of pre-miRNA into miRNA as well as for interfering RNA function [9],[10].10.1371/journal.ppat.1000162.g001Figure 1Biosynthesis and Activity of iRNAsA single miRNA targets at least 100 transcripts from various genes, and one mRNA may be targeted, at its 3′ end, by different miRNAs [11]. Thus, miRNAs, whose number has been estimated to be 340 [12], and more recently over 600 (http://microrna.sanger.ac.uk/cgi-bin/sequences/mirna_summary.pl?org=hsa), could regulate at least one-third of all human genes [13].Researchers have taken advantage of this pathway to induce the specific destruction of mRNAs in order to silence genes of choice. They do so either by directly transfecting siRNAs of approximately 21 base pairs, or by delivering transgenes encoding hairpin RNAs, small hairpin RNAs (shRNAs) processed by Dicer to siRNAs, with a stretch perfectly complementary with the target mRNA. Now, it happens that some viral RNAs adopt a stem loop conformation recognized by Drosha or Dicer as a suitable substrate and are processed to iRNAs, and we will call these viral iRNAs (viRNAs) to distinguish them from the endogenous cellular miRNAs. As the characteristics of Drosha and Dicer substrates are not fully defined, the fact that a viral RNA will or will not be processed to an iRNA cannot at present be accurately predicted. Likewise, the characteristics that make an mRNA a target for a given iRNA, including its own structure [14],[15], its accessibility, i.e., its localization in the cell and its association with other components of the cell [16], are not absolutely established. Consequently, any bioinformatic study predicting that a viral RNA will be processed into viRNA or that an mRNA will be silenced by an iRNA has to be confirmed experimentally. It must also be kept in mind that many studies analyze the effect on viral RNA of iRNAs at high concentrations obtained after transfection, and that such results must be confirmed in the course of the infection at physiological concentrations of iRNAs.RNA Interferences in the Cell/Virus ComplexViral RNAs and the miRNA machinery may interfere in various ways (Figure 2). First, the viRNAs generated by the cell from virus RNAs may target back viral RNAs (pathway 1), but also cell mRNAs that happen to share some sequence homology with them (pathway 2). Second, cellular miRNAs may recognize viral RNAs and silence them (pathway 3). Finally, the cell may produce miRNAs that control the expression of a cellular protein necessary for the virus life cycle (pathway 4). We will review these various possibilities, comparing examples of what is already known for other viruses with what is known for HIV.10.1371/journal.ppat.1000162.g002Figure 2Interactions between iRNAs, Cellular mRNAs, and HIV RNAsAutosilencing: viRNA against viral RNA (pathway 1)Obviously, some viRNAs will match exactly with some viral RNAs issued from the same genomic sequence, and induce their destruction. This is the case, for instance, in the course of simian virus 40 (SV40) infection. The virus encodes two RNAs with a stem loop structure that are transformed by the host cell into miR-S1s that are complementary to mRNA coding for the viral T antigen. Consequently, T antigen mRNA is degraded [17]. In this scenario, host miRNA machinery turns viral RNA against viral RNA. The same phenomenon has been looked for in HIV infection. A precursor [18] and a recent [19] study by two major groups in the field failed to identify iRNA of HIV origin in infected cells. Yet, a nef- and LTR-specific HIV miRNA able to inhibit LTR-driven transcription has been evidenced by another group [20],[21]. Moreover, a fourth group has reported that a stem loop HIV RNA can be processed by Dicer to a viRNA able to target env mRNA and that transfection of the corresponding shRNA inhibited over 80% of env mRNA production [22], a result that was challenged later [19]. Finally, Klase et al. [23] and Ouellet et al. [24] have shown that TAR is a source of viRNAs: Dicer interacts with TAR, and cleaves it to produce TAR-derived viRNAs able to exert gene downregulatory effects. The level of anti-HIV activity of this pathway, which is controversial, remains to be evaluated in the context of the infection. Anyway, the antiviral effect of HIV viRNAs is at best partial.viRNA against cell mRNA (pathway 2)It may happen that the host cell transforms a viral RNA into a viRNA with some degree of homology with a cell mRNA. In this case, the viRNA will inhibit the expression of the corresponding cellular gene. Such a possibility has been recently described during human cytomegalovirus (HCMV) infection. HCMV encodes an RNA transformed by the infected cell into a viRNA, hcmv-miR-UL112, that blocks the translation of major histocompatibility complex class I–related chain B (MICB). Of note, MICB is a stress-induced molecule expressed on HCMV-infected cells that is recognized by the natural killer (NK) cell activating receptor NKG2D [25]. Likewise, are there viRNAs of HIV origin able to target cell mRNAs? Computer-directed analyses have evidenced regions of base complementarity between HIV-1 sequences and human genes involved in HIV infection, e.g., CD4, CD28, CD40L, IL-2, IL-3, IL-12, and TNFβ\n[26]. Moreover, Bennasser et al. have predicted the existence of five HIV RNAs that could be transformed by Dicer into five viRNAs able to target various cellular mRNAs [27]. If these predictions are experimentally confirmed, HIV could also manipulate cellular biology and host immune response via viRNAs that were produced by the infected cell.Cellular miRNA against virus RNA (pathway 3)Sometimes the cell does not borrow viral RNAs to produce viRNAs with antiviral activity, but rather uses its own miRNAs. A striking example of this strategy is represented by the cell miR-32 that targets the open reading frame 2 of the primate foamy virus type 1, thereby inhibiting virus translation [28]. In this way, a cellular miRNA may directly target a virus RNA and block virus production. Interestingly, the same process results in the opposite effect for hepatits C virus (HCV). The endogenous miR-122, which is specifically produced by liver cells, binds to the 5′ noncoding region of HCV, but, unusually, this results in an increase in viral RNA replication through a mechanism that remains to be unraveled [29]. Interactions between cellular miRNAs and HIV RNAs may also exist, in so far as an in silico study predicted that five human miRNAs, expressed in T cells, might target nef, vpr, vif, and env RNAs [30]. Recently, Huang et al. have shown that miR-28, -125b, -150, -223, and -382, cellular miRNAs that are overexpressed in quiescent T4 lymphocytes, target sequences in the 3′ end of HIV-1 RNA, silencing thus almost all viral messengers [31]. Neutralizing these cellular miRNAs by transfecting specific antagonists into nonactivated T4 cells from patients with HIV under highly active antiretroviral therapy increased by 10-fold the in vitro efficiency of virus isolation. These observations strongly argue for a role of cellular miRNAs in HIV latency. This hypothesis, together with the suppression exerted by HIV on miRNA biosynthesis, might partly explain why HIV has not mutated to escape from the inhibitory effect of iRNAs of viral or cellular origin.Cellular miRNA targeting cell mRNA encoding proteins involved in virus replication (pathway 4)Finally, a fourth situation, where host miRNAs limit viral proliferation by acting on cellular mRNA, has been recently proposed for HIV [32]. Triboulet et al. have shown that the cellular miRNAs miR-17-5p and miR-20 silence the mRNA encoding the histone acetylase PCAF. PCAF has been previously presented as a host cofactor for Tat transactivation of HIV LTR, and as being recruited by Tat and remodeling the histone architecture in the vicinity of the LTR, promoting thereby HIV gene expression (Figure 3A). This is to say that human cells seem to permanently downregulate HIV-1 replication by depriving the infected cell of an endogenous enzyme that could be necessary for virus gene expression.10.1371/journal.ppat.1000162.g003Figure 3Inhibition of HIV Expression by the Silencing Exerted by miR-17-5p and miR-20 on PCAF (A), and Downregulation of This Inhibition by HIV (B)Regulation of the miRNA Pathway by the VirusThe few examples we have just reviewed suggest that, most of the time, the miRNA machinery works against the virus. It is not surprising then, that many viruses have elaborated strategies to subvert this machinery. And HIV is not the last one.Inhibition of miRNA production: viral suppressors of RNA silencingA strategy adopted by many viruses involves binding to a component of the miRNA machinery in order to block it. For instance, the adenoviral protein VA1 inhibits the nucleo-cytoplasmic transport of pre-miRNA by forming a complex with exportin 5 [33]. HIV seems to target two other factors of RNA silencing, Dicer and TRBP (Figure 4). Bennasser et al. have shown that purified Tat protein inhibits the capacity of Dicer to process double-stranded RNA to short iRNAs in vitro [22]. Moreover, the TAR and RRE sequences of HIV RNA, by recruiting TRBP, could competitively inhibit the effect of TRBP on pre-miRNA processing and on miRNA function [34],[35]. Yet, the capacity of Tat to hinder miRNA biogenesis has been recently challenged [19], and the impact on HIV infection of these inhibitions of miRNA expression remains to be quantified. Anyhow, if HIV encodes suppressors of RNA silencing, their effect is obviously incomplete.10.1371/journal.ppat.1000162.g004Figure 4Regulation of the miRNA Machinery by HIVThis double mechanism of suppression of RNA silencing strongly suggests that the interest of HIV is to counteract RNA interference; that is to say, the effect of the miRNA machinery is globally harmful for the virus. In support of this notion, a point mutation in Tat, which abrogates its inhibitory effect on Dicer, but not its transactivation effect on HIV LTR, results in a reduction in virus replication [22]. Likewise, Haasnoot et al. have shown that a Tat protein able to block Dicer activity is necessary for efficient HIV production [36]. Moreover, Triboulet et al. have reported that the inhibition of Drosha or Dicer using specific siRNA increases HIV replication in peripheral blood mononuclear cells from individuals with HIV [32].Viral up- and downregulation of miRNA concentrationsIn addition to globally diminishing miRNA production, viruses may also specifically regulate, positively or negatively, the level of expression of some miRNAs. This is the case for tobacco mosaic virus, where infection results in an increase in miR-156, -160, -164, -166, -169, and -171 in Nicotiana tabacum\n[37]. HIV infection also results in such a phenomenon. Triboulet et al. have reported that the infection of a lymphoid cell line by HIV-1 downregulates six miRNAs, including miR-17-5p and miR-20, and upregulates eleven others, including miR-122, -297, -370, and -373 [32]. The mechanism responsible for this selective regulation by HIV remains to be unveiled. During tobacco mosaic virus infection, it is the accumulation of two viral proteins, the movement protein and the coat protein, that is responsible for the change in miRNAs concentrations [37]. Of note, two of the miRNAs downregulated during HIV infection are precisely miR-17-5p and miR-20, which target the Tat cofactor PCAF. Thus, by reducing the amount of miR-17-5p and miR-20 available, the virus alleviates the negative control exerted by the cell on PCAF, thereby facilitating its own transcription (Figure 3B).Consequences of the Interactions between iRNAs and the Other RNAs during InfectionThus, viruses and the miRNA machinery of the host cell interact in various ways. These interactions may have consequences on the replication and the pathogenicity of the virus, but also on the immune response of the host.Consequences for virus replication/latencyInteractions between iRNAs and other RNAs in the infected cell may have consequences on the virus life cycle, mostly negative, with one exception, HCV. For HIV, viRNA complementary to env, nef, and/or LTR sequences [20]–[22] and cellular miRNAs targeting the 3′ end of HIV RNAs [31] could inhibit virus replication. Moreover, the cellular miR-17-5p and miR-20, through their repression of PCAF expression (Figure 3B), seem to exert the same effect [32]. In these various ways, RNA interference might downregulate HIV production. Logically, the virus tries to counter this effect by blocking Dicer activity and by hijacking TRBP.The global impact on in vivo infection of the interplay between virus and cell RNAs, and the relative importance of each interaction, remain to be determined. This interplay could influence the efficiency of HIV replication, and thereby the rate of disease progression, but it could also be involved in HIV latency and in the constitution of the viral reservoir. Finally, differences in the anti-HIV efficiency of RNA inteference could exist between individuals. These differences might be due to genetic polymorphisms in sequences regulating miRNA gene transcription or stability, as well as polymorphisms in cellular miRNA sequences resulting in variations in processing or targeting as already described for miR-138-2 and miR-30c-2 [38].Consequences on the immune responseInhibition of virus gene expression by RNA interference may downregulate virus replication in vitro, but in vivo, this downregulation might help the virus to escape from the immune response. This is the case for the above-mentioned inhibition exerted by the SV40 miRNAs miR-S1s on the T antigen (see “Autosilencing: viRNA against virus RNA (pathway 1)”). This inhibition occurs late enough in the virus life cycle not to hinder SV40 replication, but early enough to reduce T lymphocyte cytotoxicity and interferon-γ production triggered by the presence of the viral T antigen [17]. The targeting of cell mRNA by viRNA (pathway 2) may also result in the reduction of immune response. This possibility is illustrated by the inhibition of MICB, a ligand for NK cell activating receptor, by an HCMV miRNA (“viRNA against cell mRNA (pathway 2)”). This targeting is not innocent, since it results in a decrease in antiviral NK activity. Some cellular miRNAs are involved in innate immunity, e.g., miR-155 induced by Toll-like receptor ligands or interferon-β (IFN-β) [39] and miRNA let-7a mediating IL-6-induced cell survival [40]. Some cellular miRNAs are involved in specific immunity, e.g., miR-155 regulating genes driving functions of lymphocytes [41],[42], miR-150 controlling the differenciation of B cells via c-Myb [43], and miR-181a modulating the antigen sensitivity of T cells [44]. A striking example of the role played by miRNAs in immunity has been lately given by Pedersen et al. This group has reported that some of the anti-HCV effect of IFN-β is mediated by five cellular miRNAs, induced by IFN-β, that target HCV RNA [45]. Consequently, changes in miRNA expression induced by HIV might also alter the immune response.Consequences on the pathogenicityThese virus/cell, iRNA/RNA interactions could also have consequences on the pathogenicity of the infection. Because cellular miRNAs regulate major mechanisms of cell physiology, such as proliferation, differentiation, apoptosis, or tumorogenicity [46], targeting of cell mRNAs by viRNAs (pathway 2) and/or specific or global virus-induced disturbance of the cellular miRNA production might have pathogenic effects. The extent to which the modifications of host biology observed in individuals with HIV are mediated by RNA interference remains to be addressed, however.Therapeutic OpportunitiesBesides strategies using synthetic interfering RNAs (siRNAs or shRNAs) to target mRNA encoding cell proteins necessary for virus replication or directly target HIV RNA, the knowledge of HIV RNA/RNA silencing connections might pave the way for new therapeutic approaches. The fact that siRNAs or siRNAs antagonists modified chemically for stability and conjugated to cholesterol or liposomes have a specific, long-lasting, and ubiquitous effect when administrated intravenously [47],[48] is particularly encouraging for this kind of approach.As interfering RNAs disturb HIV replication, the possibility to act on the viral infection through miRNAs must be considered. At least two opposite strategies may be proposed. First, the miRNA machinery could be manipulated in order to hinder virus replication. Second, because of the involvement of miRNA in HIV latency, the miRNA machinery could be manipulated in order to provoke virus replication in the reservoir cells with the aim of lysing them. This second approach has the advantage of only requiring a brief treatment that is sufficient to induce HIV expression in infected, non-productive quiescent cells. To reach this goal, a possibility is to target host miRNA. For instance, the neutralization of miR-17 and miR-20 by specific inhibitors should release the inhibition exerted by these miRNAs on PCAF and hopefully trigger HIV replication in reservoir cells, resulting in the eradication of this population. A drawback of such a strategy is that the inhibition of a given miRNA will increase the level of expression of all of its mRNA targets with possible side effects. As our knowledge on the impact of the disturbance induced by HIV on host immunity and cell physiology increases, new therapeutic strategies may arise.Last, the manipulation of miRNAs in order to boost HIV production in vitro may ameliorate the efficiency of virus recovery from patients' cells [31],[32], and of virus mass production.ConclusionThe miRNA pathway appears to be a new branch of natural antiviral immunity. Understanding the complex relationship that exists between the miRNA pathway and HIV will enlighten the physiopathology of the infection and offer new therapeutic strategies. As miRNA might be involved in HIV replication and latency, pathogenicity, and immune response, manipulating miRNA might enable us to modulate all of these aspects of the infection. Much remains to be discovered with surprises on the way. For instance, the eventuality that miRNA might also regulate DNA transcription in mammal cells [49] by inducing DNA methylation [50] and/or other mechanisms [51],[52] could offer additional means to inhibit HIV replication, as suggested by recent data [23],[53].\n\nREFERENCES:\n1. Lagos-QuintanaMRauhutRMeyerJBorkhardtATuschlT\n2001\nNew microRNAs from mouse and human.\nRNA\n9\n175\n179\n2. Lagos-QuintanaMRauhutRLendeckelWTuschlT\n2001\nIdentification of novel genes coding for small expressed RNAs.\nScience\n294\n853\n858\n11679670\n3. LimLPGlasnerMEYektaSBurgeCBBartelDP\n2003\nVertebrate microRNA genes.\nScience\n299\n1540\n12624257\n4. LeeYJeonKLeeJTKimSKimVN\n2002\nMicroRNA maturation: stepwise processing and subcellular localization.\nEMBO J\n21\n4663\n4670\n12198168\n5. LeeJTAhnCHanJCholHKimJ\n2003\nThe nuclear RNase III Drosha initiates microRNA processing.\nNature\n425\n415\n419\n14508493\n6. YiRQinYMacaraIGCullenBR\n2003\nExportin-5 mediates the nuclear export of pre-microRNAs and short hairpin RNAs.\nGenes Dev\n17\n3011\n3016\n14681208\n7. HutvagnerGMcLachlanJPasquinelliAEBallintAETuschlT\n2001\nA cellular function for the RNA-interference enzyme Dicer in the maturation of the let-7 small temporal RNA.\nScience\n293\n834\n838\n11452083\n8. LiuJCarmellMARivasVMarsdenCGThompsonJM\n2004\nArgonaute2 is the catalytic engine of mammalian RNAi.\nScience\n305\n1437\n1441\n15284456\n9. ChendrimadaTPGregoryRIKumaraswamyENormanJCoochN\n2005\nTRBP recruits the Dicer complex to Ago2 for microRNA processing and gene silencing.\nNature\n436\n740\n744\n15973356\n10. HaaseADJaskiewiczLZhangHLainéSSackR\n2005\nTRBP, a regulator of cellular PKR and HIV-1 virus expression, interacts with Dicer and functions in RNA silencing.\nEMBO Rep\n6\n961\n967\n16142218\n11. KrekAGrunDPoyMNWolfRRosenbergL\n2005\nCombinatorial microRNA target predictions.\nNat Gen\n37\n495\n500\n12. LandgrafPRusuMSheridanRSewerAIovinoN\n2007\nA mammalian microRNA expression atlas based on small RNA library sequencing.\nCell\n129\n1401\n1414\n17604727\n13. LewisBPBurgeCBBartelDP\n2005\nConserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets.\nCell\n120\n15\n20\n15652477\n14. WesterhoutEMOomsMVinkMDasATBerkhoutB\n2005\nHIV-1 can escape from RNA interference by evolving an alternative structure in its RNA genome.\nNucleic Acids Res\n33\n796\n804\n15687388\n15. YoshinariKMiyagishiMTairaK\n2004\nEffects ln RNAi of the tight structure, sequence and position of the targeted region.\nNucleic Acids Res\n32\n691\n699\n14762201\n16. ChangJTaylorJM\n2003\nSusceptibility of human hepatitis delta virus RNAs to small interfering RNA action.\nJ Virol\n77\n9728\n9731\n12915586\n17. SullivanCSGrundhoffATTevethiaSPipasJMGanemD\n2005\nSV40-encoded microRNAs regulate viral gene expression and reduce susceptibility to cytotoxic T cells.\nNature\n435\n682\n686\n15931223\n18. PfefferSSewerALagos-QuintanaMSheridanRSanderC\n2005\nIdentification of microRNAs of the herpesvirus family.\nNat Methods\n2\n269\n276\n15782219\n19. LinJCullenBR\n2007\nAnalysis of the interaction of primate retroviruses with the human RNA interference machinery.\nJ Virol\n81\n12218\n12226\n17855543\n20. OmotoSItoMTsutsumiYIchikawaYIkuyamaH\n2004\nHIV-1 nef suppression by virally encoded microRNA.\nRetrovirology\n1\n44\n15601474\n21. OmotoSFujiiYR\n2005\nRegulation of human immunodeficiency virus 1 transcription by nef microRNA.\nJ Gen Virol\n86\n751\n755\n15722536\n22. BennasserYLeS-YBenkiraneMJeangK-T\n2005\nEvidence that HIV-1 encodes an siRNA and a suppressor of RNA silencing.\nImmunity\n22\n607\n619\n15894278\n23. KlaseZKalePWinogradRGuptaMVHeydarianM\n2007\nHIV-1 TAR element is processed by Dicer to yield a viral micro-RNA involved in chromatin remodeling of the LTR.\nBMC Mol Biol\n8\n63\n17663774\n24. OuelletDLPlanteILandryPBaratCJanelleME\n2008\nIdentification of functional microRNAs released through asymetrical processing of HIV-1 TAR element.\nNucleic Acids Res\n36\n2353\n2365\n18299284\n25. Stern-GinossarNElefantNZimmermannAWolfDGSalehN\n2007\nHost immune system gene targeting by a viral miRNA.\nScience\n317\n376\n381\n17641203\n26. CouturierJPRoot-BernsteinRS\n2005\nHIV may produce inhibitory microRNAs (miRNAs) that block production of CD28, CD4 and some interleukins.\nJ Theor Biol\n235\n169\n184\n15862587\n27. BennasserYLeS-YYeungMLJeangK-T\n2004\nHIV-1 encoded candidate micro-RNAs and their cellular targets.\nRetrovirology\n1\n43\n15601472\n28. LecellierC-HDunoyerPArarKLehmann-CheJEyquemS\n2005\nA cellular microRNA mediates antiviral defense in human cells.\nScience\n308\n557\n560\n15845854\n29. JoplingCLYiMLancasterAMLemonSMSarnowP\n2005\nModulation of Hepatitis C virus RNA abundance by a liver-specific micro-RNA.\nScience\n309\n1577\n1581\n16141076\n30. HariharanMScariaVPillaiBBrahmachariSK\n2005\nTargets for human encoded microRNAs in HIV genes.\nBiochem Biophys Res Commun\n337\n1214\n1218\n16236258\n31. HuangJWangFArgyrisEChenKLiangZTianH\n2007\nCellular microRNAs contribute to HIV-1 latency in resting primary CD4+ T lymphocytes.\nNat Med\n13\n1241\n1247\n17906637\n32. TribouletRMariBLinY-LChable-BessiaCBennasserY\n2007\nSuppression of microRNA-silencing pathway by HIV-1 during virus replication.\nScience\n315\n1579\n1582\n17322031\n33. LuSCullenBR\n2004\nAdenovirus VA1 noncoding RNA can inhibit small interfering RNA and microRNA biogenesis.\nJ Virol\n78\n12868\n12876\n15542639\n34. GatignolALainéSClerziusG\n2005\nDual role of TRBP in HIV replication and RNA interference: viral diversion of a cellular pathway or evasion from antiviral immunity?\nRetrovirology\n2\n65\n16253139\n35. ChristensenHSDaherASoyeKJFrankelLBAlexanderMR\n2007\nSmall interfering RNA against the TAR RNA binding protein, TRBP, a Dicer cofactor, inhibit Human Immunodeficiency Virus type 1 Long Terminal Repeat expression and viral production.\nJ Virol\n81\n5121\n5131\n17360756\n36. HaasnootJde VriesWGeutjesE-JPrinsMde HaanP\n2007\nThe Ebola virus VP35 protein is a suppressor of RNA silencing.\nPLoS Pathog\n3\ne86\ndoi:10.1371/journal.ppat.0030086\n17590081\n37. BazziniAAHoppHEBeachyRNAsurmendiS\n2007\nnfection and coaccumulation of tobacco mosaic virus proteins alter microRNA levels, correlating with symptom and plant development.\nProc Natl Acad Sci U S A\n104\n12157\n12162\n17615233\n38. IwayNNarabaH\n2005\nPoloymorphisms in human pre-miRNAs.\nBiochem Biophys Res Commun\n331\n1439\n1444\n15883035\n39. O'ConnellRMTaganovKDBoldinMPChengGBaltimoreD\n2007\nMicroRNA-155 is induced during the macrophage inflammatory response.\nProc Natl Acad Sci U S A\n104\n1604\n1609\n17242365\n40. MengFHensonRWehbe-JanekHSmithHUenoY\n2007\nThe microRNA let-7a modulates interleukin-6-dependent STAT-3 survival signaling in malignant human cholangiocytes.\nJ Biol Chem\n282\n8256\n8264\n17220301\n41. RodriguezAVigoritoEClareSWarrenMVCouttetP\n7\nRequirement of bic/microRNA-155 for normal immune function.\nScience\n316\n608\n611\n17463290\n42. ThaiTHCaladoDPCasolaSAnselKMXiaoC\n2007\nRegulation of the germinal center response by microRNA-155.\nScience\n316\n604\n608\n17463289\n43. XiaoCCaladoDPGallerGThaiTHPattersonHC\n2007\nmiR-150 controls B cell differentiation by targeting the transcription factor c-Myb.\nCell\n131\n146\n159\n17923094\n44. LiQJChauJEbertPJSylvesterGMinH\n2007\nmiR-181a is an intrinsic modulator of T cell sensitivity and selection.\nCell\n129\n147\n161\n17382377\n45. PedersenIMChengGWielandSVoliniaSCroceCM\n2007\nInterferon modulation of cellular microRNAs as an antiviral mechanism.\nNature\n449\n919\n922\n17943132\n46. AmbrosV\n2004\nThe functions of animal microRNAs.\nNature\n431\n350\n355\n15372042\n47. KrützfeldtJRajewskyNBraichRRajeevKGTuschlT\n2005\nSilencing of microRNAs in vivo with “antagomirs”.\nNature\n438\n685\n689\n16258535\n48. ZimmermannTSLeeACHAkincABramlageBBumcrotB\n2006\nRNAi-mediated gene silencing in non-human primates.\nNature\n441\n111\n114\n16565705\n49. MorrisKVChanSWJacobsenSELooneyDJ\n2004\nSmall interfering RNA-induced transcriptional gene silencing in human cells.\nScience\n305\n1289\n1292\n15297624\n50. CastanottoDTommasiSLiMLiHYanowS\n2005\nShort hairpin RNA-directed cytosine (CpG) methylation of the RASSF1A promoter in HeLa cells.\nMol Ther\n12\n179\n183\n15963934\n51. TingAHSchuebelKEHermanJGBaylinSB\n2005\nShort double-stranded RNA induces transcriptional gene silencing in human cancer cells in the absence of DNA methylation.\nNat Gen\n37\n906\n910\n52. VolpeTAKidnerCHallIMTengGGrewalSIS\n2002\nRegulation of heterochromatic silencing and histone H3 lysine-9 methylation by RNAi.\nScience\n297\n1833\n1837\n12193640\n53. SuzukiKShijuukuTFukamachiTZaundersJGuilleminG\n2005\nProlonged transcriptional silencing and CpG methylation induced by siRNAs targeted to the HIV-1 promoter region.\nJ RNAi Gene Silencing\n1\n66\n78\n19771207"
+}

batch_3/PMC2533405.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533405",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533405\nAUTHORS: Alexei Vazquez, Marcio A. de Menezes, Albert-László Barabási, Zoltan N. Oltvai\n\nABSTRACT:\nThe cell's cytoplasm is crowded by its various molecular components, resulting in a limited solvent capacity for the allocation of new proteins, thus constraining various cellular processes such as metabolism. Here we study the impact of the limited solvent capacity constraint on the metabolic rate, enzyme activities, and metabolite concentrations using a computational model of Saccharomyces cerevisiae glycolysis as a case study. We show that given the limited solvent capacity constraint, the optimal enzyme activities and the metabolite concentrations necessary to achieve a maximum rate of glycolysis are in agreement with their experimentally measured values. Furthermore, the predicted maximum glycolytic rate determined by the solvent capacity constraint is close to that measured in vivo. These results indicate that the limited solvent capacity is a relevant constraint acting on S. cerevisiae at physiological growth conditions, and that a full kinetic model together with the limited solvent capacity constraint can be used to predict both metabolite concentrations and enzyme activities in vivo.\n\nBODY:\nIntroductionUnderstanding an organism's metabolism at a system level and obtaining quantitative predictions for the different metabolic variables requires the identification and modeling of the physicochemical and regulatory constraints that are relevant at physiological growth conditions. Recently, there has been a surge of interest on how macromolecular crowding, i.e., the crowding of the cytoplasm by various molecular components, affects cellular function, including cell metabolism [1],[2].At the local scale it is well known that molecular crowding affects the rate of biochemical reactions, diffusion, protein folding and protein-protein association/dissociation [2],[3]. More recently, we have shown that macromolecular crowding acts also at a global scale by imposing a limited solvent capacity. Specifically, we have shown that a flux balance modeling framework that incorporates the limited solvent capacity is successful in predicting the maximum growth rate, the sequence of substrate uptake from a complex medium and, to an extent, the changes in intracellular flux rates upon varying growth rate of the bacterium, Escherichia coli\n[4],[5]. Yet, these studies were limited by the absence of a full kinetic model of E. coli cell metabolism, hindering our ability to investigate the impact of the solvent capacity constraint on in vivo metabolite concentrations and enzyme activities.During cellular metabolism the concentration of enzymes and metabolites are continuously adjusted in order to achieve specific metabolic demands. It is highly likely that during evolution global metabolic regulation has evolved such as to achieve a given metabolic demand with an optimal use of intracellular resources. However, the size of enzymes and intermediate metabolites are dramatically different. Enzymes are macromolecules that occupy a relatively large amount of space within a cell's crowded cytoplasm, while metabolites are much smaller. This implies that metabolite concentrations are likely to be adjusted to minimize the overall “enzymatic cost” (in terms of space cost).Here we study the validity of this hypothesis by focusing on the glycolysis pathway of the yeast, Saccharomyces cerevisiae, for which a kinetic model is available. We show that the maximum glycolysis rate determined by the limited solvent capacity is close to the values measured in vivo. Furthermore, the measured concentration of intermediate metabolites and enzyme activities of glycolysis are in agreement with the predicted values necessary to achieve this maximum glycolysis rate. Taken together these results indicate that the limited solvent capacity constraint is relevant for S. cerevisiae at physiological conditions. From the modeling point of view, this work demonstrates that a full kinetic model together with the limited solvent capacity constraint can be used to predict not only the metabolite concentrations, but in vivo enzyme activities as well.ResultsLimited Solvent Capacity ConstraintThe cell's cytoplasm is characterized by a high concentration of macromolecules [1],[2] resulting in a limited solvent capacity for the allocation of metabolic enzymes. More precisely, given that enzyme molecules have a finite molar volume vi only a finite number of them fit in a given cell volume V. Indeed, if ni is the number of moles of the i\nth enzyme, then(1)where V\n0 accounts for the volume of other cell components and the free volume necessary for cellular transport as well. Equation 1 can be also rewritten as(2)where ρi = nimi/V is the enzyme density (enzyme mass/volume), μi is the molar mass v\nspec is the specific volume, and φ = V\n0/V is the fraction of cell volume occupied by cell components other than the enzymes catalyzing the reactions of the pathway under consideration, including the free volume necessary for diffusion. The specific volume has been assumed to be constant for all enzymes, an approximation that has been shown to be realistic at least for globular proteins [6]. In this new form we can clearly identify the enzyme density (or mass, given that the volume is constant) as the enzyme associated variable contributing to the solvent capacity constraint. This choice is more appropriate than the enzyme concentration Ci = ni/V (moles/volume) because the specific volume is approximately constant across enzymes, while the molar volume can exhibit significant variations. For example, according to experimental data for several globular proteins [6], the molar volume exhibits a 70% variation while the specific volume is almost constant, with a small 2% variation.The solvent capacity constraint (Equations 1 and 2) thus imposes a limit to the amount of catalytic units (i.e., enzymes) that can be allocated in the cell cytoplasm. In the following we show that this in turn leads to a constraint for the maximum metabolic rate. The rate of the i\nth reaction per unit of cell dry weight (mol/time/mass) is given by(3)where Ai is the specific enzyme activity, Ci is the enzyme concentration in molar units, ki is the catalytic constant and M is the cell mass. The coefficient xi is determined by the specific kinetic model: it takes values in the range of 0≤xi≤1, and it is a function of metabolite concentrations. For example, if the i\nth reaction is described by Michaelis-Menten kinetics with one substrate then xi = Si/(Ki+Si), where Si is the substrate concentration and Ki is the equilibrium constant. More generally, xi is a function of the concentration of substrates, products and other metabolites regulating the enzyme activity. The fact that the reaction rates are proportional to the enzyme densities (Equation 3) suggests that the limited solvent capacity constraint (Equation 2) has an impact on the reaction rates as well. Indeed, from Equations 2 and 3 we obtain(4)where R is the cell metabolic rate (or pathway rate), ri = Ri/R is the rate of reaction i relative to the metabolic rate, and(5)where ρ = M/V is the cell density. We refer to ai as the crowding coefficients [4],[5], because they quantify the contribution of each reaction rate to molecular crowding. The crowding coefficient of a reaction i increases with increasing the enzyme's molar mass μi and decreases with increasing catalytic activity ki. It is also a function of the metabolite concentrations through xi.Hypothetical Three Metabolites PathwayTo illustrate the impact of the limited solvent capacity constraint, we first analyze a hypothetical example, in which we use the relative reaction rates as input parameters, and the metabolite concentrations are the variables to be optimized. Given the reaction rates and the “optimal” metabolite concentrations, the enzyme activities are determined by Equation 3. Finally, the maximum metabolic rate is computed using Equation 4.Consider a metabolic pathway consisting of two reversible reactions converting metabolite M1 into M2 (reaction 1) and M2 into M3 (reaction 2), catalyzed by enzymes e\n1 and e\n2, respectively (Figure 1, inset). The reaction rates per unit of cell mass, R\n1 and R\n2, are modeled by reversible Michaelis-Menten rate equations, using Equation 3 with(6)\n(7)where K\n1eq and K\n2eq are the equilibrium constants of reaction 1 and 2, respectively, Kim is the Michaelis-Menten constant of metabolite m in reaction i. From Equations 4 to 7 we finally obtain(8)For the purpose of illustration, we assume 1−ϕ = 0.01,  (mmol/h/min)−1 (as suggested by typical values reported in [5]), all Michaelis constants equal to 1 mM, and fixed pathway ends metabolite concentrations [M1] = 3 mM and [M2] = 1 mM. Furthermore, mass conservation for M2 implies that R\n1 = R\n2 = R (r\n1 = r\n2 = 1) in the steady state, where R is the pathway rate. When reaction 1 is close to equilibrium [M2]≈[M1]K\n1eq = 3 mM, the first term in the right hand side becomes very large, resulting in a small pathway rate (Figure 1). When reaction 2 is close to equilibrium [M2]≈[M3]/K\n2eq = 1 mM, the second term in the right hand side becomes very large, again resulting in a small pathway rate (Figure 1). At an intermediate [M2]* between these two extremes the pathway rate achieves its maximum. Therefore, given the solvent capacity constraint, there is an optimal metabolite concentration resulting in a maximum pathway rate.10.1371/journal.pcbi.1000195.g001Figure 1Hypothetical three metabolite pathway.The inset shows a hypothetical three metabolite-containing pathway with two reactions. The main panel displays the pathway rate as a function of the concentration of the intermediate metabolite. Of note, at an intermediate metabolite concentration [M2]*, the pathway rate achieves a maximum. The plot was obtained using the kinetic parameters indicated in the text.\nS. cerevisiae GlycolysisNext, we investigate whether the observation of an optimal metabolite concentration for maximum pathway rate extrapolates to a more realistic scenario. For this purpose we use the glycolysis pathway of the yeast S. cerevisiae (Figure 2A) as a case study. Glycolysis represents a universal pathway for energy production in all domains of life. In S. cerevisiae it has been studied extensively resulting in the description of a rate equation model for each of its reactions [7],[8]. In particular, we consider the kinetic model developed in [7] (see Methods). To compare our predictions with experimentally determined values we consider the glycolysis reaction rates and metabolite concentrations reported in [7] and the enzyme activities reported in [8].10.1371/journal.pcbi.1000195.g002Figure 2\nS. cerevisiae glycolysis.(A) Schematic representation of glycolysis in S. cerevisiae. Metabolites: GLCx, external glucose; GLC, glucose; G6P, glucose 6-phosphate; F6P, fructose 6-phosphate; FBP, fructose 1,6-bisphosphate; DHAP, glycerone phosphate; GAP, D-glyceraldehyde 3-phosphate; BPG, 1,3-bisphosphoglycerate; and PEP, phospho-enol-pyruvate. Reactions: hxt, glucose transport; hk, hexokinase; pgi, phosphogluco isomerase; pfk, phospho-fructokinase; ald, fructose 1,6-bisphosphate aldolase; tpi, triosephosphate isomerase; gapdh, D-glyceraldehyde 3-phosphate dehydrogenase; lpPEP, reactions from BGP to PEP; pk, pyruvate kinase; and g3pdh, glycerol 3-phosphate dehydrogenase. (B,C,D) Predicted glycolysis rate as a function of the concentrations of intermediary metabolites in the S. cerevisiae glycolysis pathway (in mM). The experimentally determined metabolite levels (from [7]) are indicated by the red triangles. The dashed lines indicate the concentration intervals resulting in 50% or more of the maximum rate.In analogy with the three metabolites case study (Figure 1), first we investigate the dependency of the glycolysis rate R, represented by the glucose uptake, on the concentration of a given metabolite. In this case we fix all other metabolite concentrations and all relative reaction rates (reaction rate/glycolysis rate) to their experimentally determined values. By doing so the predicted glycolysis rate is an implicit function of the free metabolite concentration alone, through Equation 4. For example, Figure 2B displays the maximum metabolic rate R as a function of the concentration of fructose-6-phosphate (F6P). R is predicted to achieve a maximum around a F6P concentration of 0.4 mM, close to its experimentally determined value of 0.5 mM [7] (red triangle in Figure 2B). Similar conclusions are obtained for D-glyceraldehyde-3-phosphate (GAP) (Figure 2C) and glycerone-phosphate (DHAP) (Figure 2D). This analysis corroborates that there is an optimal metabolite concentration maximizing R and, more importantly, that this concentration is very close to the experimentally determined metabolite concentrations. In all cases the measured metabolite concentrations are within the range of 50% or more of the maximum glycolysis rate.To further test the optimal metabolite concentration hypothesis, we perform a global optimization and simultaneously compute the optimal concentrations of the glycolysis intermediate metabolites. In this case we fix the concentrations of external glucose and co-factors and all relative reaction rates to their experimentally determined values. By doing so the predicted glycolysis rate is an implicit function of the glycolysis intermediate metabolite concentrations, through Equation 4. The optimal intermediate metabolite concentrations are those maximizing Equation 4. Figure 3A displays the predicted optimal metabolite concentrations as a function of their experimentally determined values (black symbols), the line representing a perfect match. The agreement is remarkably good given the wide range of metabolite concentrations. For phospho-enol-pyruvate (PEP), the predicted value is very sensitive to the model parameters, as indicated by the wide error bars. For fructose 1,6-biphosphate (FBP) the predicted value is smaller by a factor of five than the experimental value, but it is still within range. Taken together, these results indicate that the measured concentrations of intermediate metabolites in the S. cerevisiae glycolysis are close to the predicted optimal values maximizing the glycolysis rate given the limited solvent capacity constraint.10.1371/journal.pcbi.1000195.g003Figure 3Correlation between predictions vs. experimental data.(A) The predicted metabolite concentrations are plotted as a function of the experimentally determined values (black symbols). The error bars represent the standard deviations, upon generating 100 random sets of kinetic parameters. The solid line corresponds with the coincidence of measured and predicted values, indicating a strong correlation between them. (B) The predicted enzyme activities are plotted as a function of the experimentally determined values, measured in units of the glycolysis rate (black symbols). The error bars represent the standard deviations, upon generating 100 random sets of kinetic parameters. The solid line corresponds with the coincidence of measured and predicted values, indicating a strong correlation between them. In both cases, the red and blue symbols were obtained using the more general optimization objective R = (1−ϕ)/ΣN\ni = 1 (airi)H, with H = 0.1 and 10, respectively.Using the optimal intermediate metabolite concentrations we can make predictions for the enzyme activities as well. Indeed, from the first equality in Equation 3 it follows that(9)The reaction rates relative to the glycolysis rate ri are obtained from experimental data, while xi are obtained after substituting the predicted optimal metabolite concentrations on the reaction's kinetic models. Figure 3B displays the predicted enzyme activities (in units of the glycolysis rate) as a function of the experimentally determined values (black symbols), the line representing a perfect match. In most cases we obtain a relatively good agreement between experimentally measured and predicted values, with the exception of phosphofructokinase (pfk), for which the measured enzyme activities are significantly overestimated. Of note, for pyruvate kinase (pk) the predictions are significantly affected by the model parameters, as indicated by the wide error bars.The preceding analysis does not exclude the possibility that other constraints could result in a good agreement as well. To address this point we consider the more general optimization objective R = (1−ϕ)/ΣN\ni = 1 (airi)H, parametrized by the exponent H. Although this objective is not inspired by a biological intuition, it allows us to explore other possibilities beyond the original case H = 1. Figure 3 show our predictions for the case H = 0.1 (red symbols) and H = 10 (blue symbols), representing a milder and a stronger dependency with the crowding coefficients ai, respectively. For H = 0.1, 1.0 and 10 the predicted metabolite concentrations are in good agreement with the experimental values. Furthermore, when we allow sub-optimal metabolite concentrations resulting in a glycolysis rate below it s maximum our predictions are also in the range of the experimental values (see Protocol S1, Table IV). These results indicate that it is sufficient that the optimization objective is a monotonic decreasing function of the crowding coefficients. When the latter is satisfied the metabolite concentrations are up to a great extent constrained by the kinetic model.This is not, however, the case for the enzyme activities. For H = 0.1 and the enzymes pfk, tpi and pk, there is a large deviation from the perfect match line. For H = 10 and the enzymes tpi and pk, there is a large deviation from the perfect match line as well. Overall, H = 1 gives the better agreement between enzyme activity predictions and their measured values. In addition, it provides a clear biophysical interpretation of the solvent capacity constraint (H = 1).Finally, we use Equation 4 to compute the maximum glycolysis rate as determined by the limited solvent capacity constraint. The global optimization predicts the glycolysis rate R = (1−ϕ)×12.5 mmol/min/g dry weight. Taking into account that about 30% [9] of the cell is occupied by cell components excluding water, that proteins account for ∼45% of the dry weight [10], and that of these glycolytic enzymes account for ∼22% [11] of the protein mass we obtain 1−ϕ∼0.03. Therefore, given that glycolysis enzymes occupy only 3% of the cell volume, we obtain R∼0.38 mmol/min/g dry weight. This prediction is in very good agreement with the experimentally determined glycolysis rate of S. cerevisiae, ranging between 0.1 to 0.4 mmol/min/g dry weight [8],[12].DiscussionThe successful modeling of cell metabolism requires the understanding of the physicochemical constraints that are relevant at physiological growth conditions. In our previous work focusing on E. coli we have reported results indicating that the limited solvent capacity is an important constraint on cell metabolism, especially under nutrient-rich growth conditions [4],[5]. Using a flux balance approach that incorporates this constraint we predicted the maximum growth rate in different carbon sources [4], the sequence and mode of substrate uptake and utilization from a complex medium [4], and the changes in intracellular flux rates upon varying E. coli cells' growth rate [5]. More importantly, these predictions were in good agreement with experimentally determined values.Here we have extended the study of the impact of the limited solvent capacity by (i) considering a different organism (S. cerevisiae), and (ii) a full kinetic model of glycolysis. Using the full kinetic model of S. cerevisiae glycolysis, we have demonstrated that the predicted optimal intermediate metabolite concentrations and enzyme activites are in good agreement with the corresponding experimental values. Discrepancies were only observed in association with two different steps in the glycolysis pathway, namely the reaction catalyzed by pfk and the reactions between BPG and PEP. The experimental values measurements from cell extracts [8] and, except for potential experimental caveats, they represent phyiological conditions. We thus we believe that the larger deviations for these enzymes are determined by inconsistencies in the kinetic model equations and/or kinetic model parameters. Finally, the glycolysis rate achieved at the optimal metabolite concentrations is in the range of the experimentally measured values.From the quantitative modeling point of view our results indicate that a full kinetic model together with the solvent capacity constraint can be used to make predictions for the metabolite concentrations and enzyme activities. Thus, we propose the simultaneous optimization of intermediate metabolite concentrations, maximizing the metabolic rate given the solvent capacity, as a method to computationally predict the concentrations of a metabolic pathway's intermediate metabolites and enzyme activities. We have demonstrated the applicability of this method by computing the concentration of S. cerevisiae glycolysis intermediate metabolites, resulting in a good agreement with published data.The hypothesis that high concentration of macromolecules in the cell's cytoplasm imposes a global constraint on the metabolic capacity of an organism has been studied in the past [13],[14],[15]. In most cases [14],[15] it has been postulated that there is a bound to the total enzyme concentration (moles/volume). Yet, -to our knowledge-, no clear explanation has been provided to support that choice. In contrast, our starting postulate is an undeniable physical constraint, the total cell volume (Equation 1). Under this constraint, the enzyme molar volumes are the primary magnitude quantifying the enzymatic cost. In turn, since the enzyme-specific volumes are approximately constant, we can use the enzyme density (mass/volume) as an alternative measure of enzymatic cost.This modeling framework has advantages and disadvantages with respect to more traditional approaches based on dynamical systems modeling. As an advantage, our method does not require as input parameters the enzyme activities but rather make quantitative predictions for them. On the other hand, our method is based on a steady-state approximation. Therefore, in its present form, it cannot be used to understand dynamical processes, such as the observed metabolite concentration oscillations in S. cerevisiae cells when growing at high glucose concentrations [7].MethodsKinetic Model of GlycolysisWe use the S. cerevisiae glycolysis model reported in [7] (see Protocol S1 for details). The only modification is the extension of the phsophofructokinase (pfk) kinetic model from an irreversible to a reversible model.Catalytic Constants, Cell Density, Specific VolumeThe catalytic constants were obtained from experimental estimates for Saccharomyces carlsbergensis\n[16], except for glycerol 3-phosphate dehydrogenase that was obtained from an estimate for Eidolon helvum\n[17]. For the cell density we use an estimate reported for E. coli, ρ = 0.34 g/ml [18]. The specific volume was estimated for several proteins using the molar volumes and masses reported in [6], resulting in average of 0.73 ml/g and standard deviation of 0.02 ml/g. See Protocol S1 for details.Optimal Metabolite ConcentrationsThe optimal metabolite concentrations are obtained maximizing Equation 4 with respect to the free metabolite concentrations. In the case of Figure 2B–2D, all metabolite concentrations are fixed to their experimental values, except for the metabolite indicated by the X-axis. In the case of Figure 3A and 3B, all intermediate metabolite concentrations are optimized, keeping fixed the concentration of external glucose and cofactors (ATP, ADP, AMP, NADH, NAD). In both cases the reaction rates relative to the glycolysis rate (ri) were taken as input parameters, using the values reported in [7]. The maximization was performed using simulated annealing [19].Parameter SensitivityTo analyze the sensitivity of our predictions to the model parameters we have generated random sets of kinetic parameters, assuming a 10% variation of the fixed metabolite concentrations and all kinetic constants except for the catalytic activities. For the latter we assumed a larger variation of 50%, because they were estimated from a different organism. For each set of parameters we make predictions for the metabolite concentrations and enzyme activities. Figure 3 reports the mean values and standard deviations.Supporting InformationProtocol S1Details on the rate equation model used, the utilized model parameters, and the glycolysis rate and optimal metabolite concentrations.(0.10 MB PDF)Click here for additional data file.\n\nREFERENCES:\n1. EllisRJ\n2001\nMacromolecular crowding: obvious but underappreciated.\nTrends Biochem Sci\n26\n597\n604\n11590012\n2. MintonAP\n2006\nHow can biochemical reactions within cells differ from those in test tubes?\nJ Cell Sci\n119\n2863\n2869\n16825427\n3. EllisRJMintonAP\n2006\nProtein aggregation in crowded environments.\nBiol Chem\n387\n485\n497\n16740119\n4. BegQKVázquezAErnstJde MenezesMABar-JosephZ\n2007\nIntracellular crowding defines the mode and sequence of substrate uptake by Escherichia coli and constrains its metabolic activity.\nProc Natl Acad Sci U S A\n104\n12663\n12668\n17652176\n5. VazquezABegQKDe MenezesMAErnstJBar-JosephZ\n2008\nImpact of the solvent capacity constraint on E. coli metabolism.\nBMC Systems Biol\n2\n7\n6. LeeB\n1983\nCalculation of volume fluctuation for globular protein models.\nProc Natl Acad Sci U S A\n80\n622\n626\n6572909\n7. HynneFDanoSSorensenPG\n2001\nFull-scale model of glycolysis in Saccharomyces cerevisiae.\nBiophys Chem\n94\n121\n163\n11744196\n8. TeusinkBPassargeJReijengaCAEsgalhadoEvan der WeijdenCC\n2000\nCan yeast glycolysis be understood in terms of in vitro kinetics of the constituent enzymes? Testing biochemistry.\nEur J Biochem\n267\n5313\n5329\n10951190\n9. AlcázarEBRocha-LeaoMHDweckJ\n2000\nYeast intracellular water determination by thermogravimetry.\nJ Therm Anal Cal\n59\n643\n648\n10. SchulzeU\n1995\nAnaerobic physiology of Saccharomyces cerevisiae:\nTechnical University of Denmark\n11. KolkmanAOlsthoornMMHeeremansCEHeckAJSlijperM\n2005\nComparative proteome analysis of Saccharomyces cerevisiae grown in chemostat cultures limited for glucose or ethanol.\nMol Cell Prot\n4.1\n1\n11\n12. DuarteNCPalssonBOFuP\n2004\nIntegrated analysis of metabolic phenotypes in Saccharomyces cerevisiae.\nBMC Genomics\n54\n63\n13. BrownGC\n1991\nTotal cell protein-concentration as an evolutionary constraint on the metabolic control distribution in cells.\nJ Theor Biol\n153\n195\n203\n1787736\n14. HeinrichRSchusterS\n1996\nThe regulation of cellular systems\nNew York\nChapman & Hall\n15. KlippEHeinrichRHolzhütterHG\n2002\nPrediction of temporal gene expression. Metabolic opimization by re-distribution of enzyme activities.\nEur J Biochem\n269\n5406\n5413\n12423338\n16. BoiteuxAHessB\n1981\nDesign of glycolysis.\nPhilos Trans R Soc Lond B Biol Sci\n293\n5\n22\n6115423\n17. SchomburgIChangASchomburgD\n2002\nBRENDA, enzyme data and metabolic information.\nNucleic Acids Res\n30\n47\n49\n11752250\n18. ZimmermanSBTrachSO\n1991\nEstimation of macromolecule concentrations and excluded volume effects for the cytoplasm of Escherichia coli.\nJ Mol Biol\n222\n599\n620\n1748995\n19. PressWHFlanneryBPTeukolskySAVetterlingWT\n1993\nNumerical recipes in C: The art of scientific computing\nCambridge\nCambridge University Press"
+}

batch_3/PMC2533406.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533406",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533406\nAUTHORS: Manuela Mally, Hwain Shin, Cécile Paroz, Regine Landmann, Guy R. Cornelis\n\nABSTRACT:\nCapnocytophaga canimorsus, a commensal bacterium of the canine oral flora, has been repeatedly isolated since 1976 from severe human infections transmitted by dog bites. Here, we show that C. canimorsus exhibits robust growth when it is in direct contact with mammalian cells, including phagocytes. This property was found to be dependent on a surface-exposed sialidase allowing C. canimorsus to utilize internal aminosugars of glycan chains from host cell glycoproteins. Although sialidase probably evolved to sustain commensalism, by releasing carbohydrates from mucosal surfaces, it also contributed to bacterial persistence in a murine infection model: the wild type, but not the sialidase-deficient mutant, grew and persisted, both when infected singly or in competition. This study reveals an example of pathogenic bacteria feeding on mammalian cells, including phagocytes by deglycosylation of host glycans, and it illustrates how the adaptation of a commensal to its ecological niche in the host, here the dog's oral cavity, contributes to being a potential pathogen.\n\nBODY:\nIntroduction\nCapnocytophaga canimorsus (formerly Centers for Disease Control group DF-2) has been rarely but regularly isolated from dog or cat bite infections since its discovery in 1976 [1],[2]. C. canimorsus is a thin Gram-negative rod, found in the normal oral flora of dogs and cats. Clinical infections by C. canimorsus generally appear as fulminant septicemia, peripheral gangrene or meningitis [3],[4]. Splenectomy, alcohol abuse and immunosuppression have been associated with a number of cases, but more than 40% of the patients have no obvious risk factor [5]. Capnocytophaga belongs to the family of Flavobacteriaceae in the phylum of Bacteroidetes, which is taxonomically remote from Proteobacteria. The family of Bacteroidaceae contains many commensals of the mammalian intestinal system such as Bacteroides thetaiotaomicron and Bacteroides fragilis\n[6]. The family of Flavobacteriaceae includes a variety of environmental and marine bacteria, among which Flavobacterium johnsoniae is a common soil and freshwater bacterium studied for its gliding motility [7]. There are only a few examples of pathogenic bacteria belonging to this family. These are Flavobacterium psychrophilum, the causative agent of cold water disease in salmonid fish [8], Ornithobacterium rhinotracheale, a bacterial pathogen known for causing respiratory disease in poultry [9], and Riemerella anatipestifer which causes “duckling disease” in waterfowl and turkeys [10],[11]. The genus Capnocytophaga includes seven species found in normal human oral flora and C. canimorsus found in the normal flora of dogs and cats. More than 160 cases of C. canimorsus infections have been reported so far [12] but very few studies have addressed the molecular mechanisms of C. canimorsus pathogenesis. Recently, we showed that C. canimorsus 5 (Cc5) resists phagocytosis by macrophage cell line [13],[14]. We also showed that although C. canimorsus does not affect the viability of murine or human macrophages, it does not elicit proinflammatory cytokines and it even blocks the proinflammatory response to the LPS from enterobacteria [13]. In the course of such experiments, Cc5 exhibited robust growth, although it is usually considered fastidious for growth. In the present study, we show that a surface-localized sialidase plays a key role in initiating an extensive deglycosylation process of host cell glycan structures and that this feeding mechanism serves as a basis for growth and persistence of C. canimorsus in vivo.ResultsGrowth of C. canimorsus 5 requires direct contact with cellsWhen inoculated at a multiplicity of infection (moi) of 20 to J774.1 murine macrophages cultured in complete RPMI (cRPMI), which includes 10% fetal bovine serum (FBS), Cc5 multiplied about 100-fold during the 24 h of infection (Fig. 1A). This observation could be repeated with non-phagocytic human epithelial HeLa cells and with canine epithelial MDCK kidney cells (Fig. 1B). Surprisingly, growth was abolished when J774.1 macrophages were omitted and moreover, Cc5 was unable to grow in a transwell system, indicating that direct contact is required for bacterial growth (Fig. 1B). This implies that Cc5 may take advantage of some nutrient that is present on the host cell surface. Notably, Cc5 did not adhere tightly to cells and was not internalized (data not shown). We generated a transposon (Tn) mutant library using Tn4351 from B. fragilis\n[15],[16] and isolated a clone that was unable to grow in the presence of J774.1 cells, but grew normally on blood agar plates. Wild type (wt) and mutant bacteria grew equally well in serum enriched heart infusion medium (Fig. 1C). Impaired growth of this Tn mutant was not due to an increased phagocytic uptake by J774.1 since addition of cytochalasin D had little effect on bacterial growth (Fig. 1B).10.1371/journal.ppat.1000164.g001Figure 1Growth of C. canimorsus 5 is dependent on cell contact.(A) Viable counts of 2×106\nCc5 after 24 h in presence of J774.1 macrophages in RPMI supplemented with 10% FBS (moi = 20) (black) or in RPMI with FBS without cells (grey) and in a transwell system preventing physical contact between bacteria and macrophages in RPMI with FBS (white). (B) Viable counts of Cc5 and Tn mutant after 24 h culture with macrophages in RPMI and FBS (black), with macrophages in RPMI and FBS in addition of cytochalasin D (grey), with HeLa cells (light grey) and MDCK cells in DMEM and FBS (white). The grey dotted line represents the bacterial number inoculated. The difference is statistically significant between Cc5 and Tn mutant (2-tailed unpaired Student's t test p<0.05) in 3 or more experiments. (C) Growth curve of wt Cc5 (triangles) and Tn mutant (squares) in heart infusion broth (HIB) supplemented with 10% FBS, represented as the mean of 3 or more experiments with the error bars showing the s.d.Surface-localized sialidase is required for the growth of Cc5 in contact with cellsThe transposon inserted at codon 77 within a gene encoding a protein with similarity to bacterial sialidase, glycosylhydrolase that cleaves terminal sialic acid from glycoconjugates (Fig. 2A). The mutated gene, designated siaC, was found to be located downstream of genes encoding a predicted transcriptional regulator and a putative N-acyl-glucosamine epimerase (accession number: EU329392). The first gene downstream was found to start 148 bp further from the siaC stop codon (Fig. 2B). To exclude any polar effects of the Tn integration, we tested whether the downstream gene was transcriptionally linked to siaC. Total RNA was isolated from wt Cc5, reverse transcribed using two different primers annealing either at the end of siaC (5132) or at the end of the downstream gene (5129) and the cDNA was amplified by PCR. As shown in Fig 2C, even though transcripts were present for both genes separately, no transcript spanned siaC and the downstream gene. This result indicates that siaC is not transcriptionally linked to the downstream gene.10.1371/journal.ppat.1000164.g002Figure 2Identification of the Tn integration site and analysis of mRNA present in wt C. canimorsus 5.(A) Amino acid sequence of the C. canimorsus sialidase showing the signal peptide (italics) and the BNR/asp repeats (Ser/Thr-X-Asp-X-Gly-X-Thr-Trp/Phe) of bacterial sialidases (boxed). Domain predictions were analyzed by InterProScan [42]. The residues conserved in sialidases at the C-terminus are underlined and the tyrosine 488 is bold [43]. The Tn4351 integration site in SiaC at amino acid 77 is indicated, boxed in grey and bold. (B) Genetic locus of the sialidase gene (siaC) including its upstream genes, gntR-like gene (CAPCA_MM1) and putative N-acyl-glucosamine epimerase encoding gene (CAPCA_MM2); and downstream coding sequence (CAPCA_MM3). (C) Reverse transcription performed on total RNA with specific primers (5129 or 5132) followed by PCR to identify transcripts present in wt Cc5 (cDNA). PCR reactions were also performed using genomic DNA (gDNA) as template instead of cDNA as a positive control. As a control, reverse transcription was performed without reverse transcriptase in a parallel assay and used as template for the subsequent PCR reaction (-RT).While intact Cc5 bacteria cleaved 2′-(4-Methylumbelliferyl)-α-D-N-acetylneuraminic acid (MUAN), the Tn mutant could not, indicating that the mutated gene does indeed encode a sialidase (Fig. 3A). We engineered an expression shuttle vector by taking advantage of a cryptic plasmid isolated from another strain of C. canimorsus and the promoter of an insertion sequence from B. fragilis\n[16]. We constructed plasmids encoding full length (FL) SiaC, a variant deprived of the 21 N-terminal residues, predicted to be a signal peptide (Δ1–21), and a catalytic mutant (Y488C). Sialidase activity (Fig. 3A) and growth in the presence of J774.1 cells (Fig. 3B) was restored by introducing in trans siaC\nFL, but not siaC\nΔ1–21. Sialidase activity was not restored to wt levels by siaC\nY488C, but it was still significant (Fig. 3A), suggesting that this residual activity might account for elevated growth in comparison to the siaC mutant (Fig. 3B). Using a sarcosyl extraction method, SiaCFL and SiaCY488C were found to be associated with the outer membrane (Fig. 3C), whereas SiaCΔ1–21 was only detected in total cells (Fig. 3B). Indirect immunofluorescence using polyclonal anti-SiaC serum on paraformaldehyde fixed but unpermeabilized bacteria confirmed that SiaC is exposed on the bacterial surface unless the signal peptide is removed (Fig. 3D). Although it is surface exposed, no SiaC could be detected in the supernatant of infected J774.1 cultures, indicating that it is tightly associated with the outer membrane (Fig. 3C). Hence, surface-localized sialidase is required for growth of Cc5 at the expense of mammalian cells.10.1371/journal.ppat.1000164.g003Figure 3Surface localized sialidase is required for growth.(A) Sialidase activity of intact bacteria, measured with the substrate MUAN as the mean of triplicate measurements and s.d. of a representative experiment. (B) Viable counts after challenge with 2×106\nCc5 (black), siaC (light grey) or siaC complemented with plasmids containing siaC\nFL, siaC\nΔ1–21 and siaC\nY488C after 24 h in presence of J774.1 with the grey dotted line indicating the bacterial number inoculated. Sialidase was detected by immunoblotting with a polyclonal antibody against SiaC in total cells (TC). (C) Outer membrane protein fractions (OMP), cell free supernatants (SN) of the J774.1 cultures shown in (B) including as control TC of Cc5 were analyzed by immunoblotting for the presence of SiaC. (D) Surface localization of SiaC was tested by immunofluorescence on paraformaldehyde fixed but not permeabilized bacteria using anti-SiaC followed by anti- rabbit IgG conjugated to FITC.Growth is sustained by N-acetyl glucosamine (GlcNAc) and N-acetyl galactosamine (GalNAc) but not by sialic acidsSince sialidases cleave terminal sialic acid from glycoconjugates, we first tested whether the addition of sialic acids could restore growth of siaC. Addition of neither sialic acid (Neu5Ac, N-Acetyl-2,3-dehydro-2-deoxyneuraminic acid) nor its activated form (CMP-Neu5Ac, Cytidine-5′-monophospho-N-acetylneuraminic acid) restored growth of siaC in presence of J774.1. In contrast, growth could be restored by the addition of purified recombinant SiaC or neuraminidase/sialidase NanH from Clostridium perfringens to the culture medium, but not by the addition of the catalytically inactive SiaCY488C (Fig. 4A). This suggested that removal of terminal sialic acids from glycoconjugates is required to make other carbohydrates accessible. Indeed, N-acetyl glucosamine (GlcNAc) and N-acetyl galactosamine (GalNAc), common carbohydrate moieties of glycoconjugates, allowed growth of siaC in the presence of macrophages (Fig. 4B). Notably, addition of glucose (Glc), galactose (Gal), mannose (Man) or sialyl-lactose (N-acetylneuraminosyl-D-lactose) could not restore growth of siaC bacteria (Fig. 4C). As galactose (Gal) is a common sugar preceding GlcNAc in glycan molecules, we next tested addition of N-acetyl lactosamine (LacNAc), a disaccharide consisting of β-D-Gal β(1→4) GlcNAc. LacNAc also restored the growth defect of siaC indicating the presence of an active β-galactosidase releasing monosaccharides Gal and GlcNAc in wt and siaC Cc5 (Fig. 4B).10.1371/journal.ppat.1000164.g004Figure 4Aminosugars but not sialic acids sustain growth of C. canimorsus.\nViable counts after challenge with 2×106 wt Cc5 (black) or siaC (grey) grown for 24 h with J774.1 in cRPMI (control) or in the same condition with the addition of Neu5Ac, Neu5Ac- CMP, 12.5 ng/ml enzyme SiaCFL, SiaCY488C or NanH from C. perfringens (A) or with the addition of GalNAc, GlcNAc or LacNAc (B) or with the addition of mannose, galactose, glucose or sialyl-lactose (C). Mean values from 3 or more experiments and s.d. are shown including statistical difference between wt Cc5 and siaC with * p<0.05, ** p<0.01 and *** p<0.001 for each pair of columns (2-tailed unpaired Student's t test). The grey dotted line indicates the bacterial number inoculated.Sialidase desialylates macrophage and epithelial cell surfacesJ774.1 macrophages were incubated with either wt or siaC bacteria and thereafter analyzed for lectin binding to investigate desialyation process on the macrophage cell surface. We used Sambucus nigra agglutinin (SNA), which recognizes terminal sialic acids (2- 6 or 2- 3) linked to Gal or to GalNAc, and peanut agglutinin (PNA), a lectin specific for Gal (β 1–3) GalNAc, a disaccharide often forming the core unit of O-linked glycoconjugates (Fig. 5A). As shown in Fig. 5B, wt bacteria greatly reduced the amount of sialic acids (SNA panel) and Gal (β 1–3) GalNAc (PNA panel) at the cell surface, while siaC bacteria had no effect on glycans masked by sialic acids. When cells were treated simultaneously with purified SiaC and siaC bacteria, neither sialic acid nor Gal (β 1–3) GalNAc were detected, indicating that siaC bacteria are still proficient in extensive deglycosylation of exposed glycans chains. The same deglycosylation of cell surfaces was observed when HeLa epithelial cells were used (Fig. 5C).10.1371/journal.ppat.1000164.g005Figure 5\nC. canimorsus desialylates macrophage and epithelial cell surfaces.(A) The targets of the lectins used in this study are schematically represented (adapted from [44]). Surface carbohydrates of J774.1 macrophages (B) or HeLa epithelial cells (C) were analyzed by lectin binding after 2 h of infection with 4×107 wt (Cc5) or siaC bacteria. Cells were fixed with paraformaldehyde and incubated for 1 h with lectin SNA, which recognizes terminal sialic acids (2- 6 or 2- 3) linked to Gal or to GalNAc or PNA that binds to the disaccharide Gal 1–3 GalNAc only after removal of terminal sialic acids. SiaC was added to cells alone or with siaC bacteria at 100 ng/ml. Biotinylated lectins were visualized by FITC conjugated streptavidin.Sialidase inhibitor N-Acetyl-2,3-dehydro-2-deoxyneuraminic acid (Neu5Ac2en) can inhibit growth of Cc5\nAs bacterial and viral sialidases can share common ASP boxes that interact with sialic acid, we postulated that common sialidase inhibitor might have sufficient specificity for the active site in SiaC to inhibit growth of Cc5 wt bacteria in presence of cells. We tested the anhydro sialate derivative Neu5Ac2en, which is known to inhibit many viral and bacterial neuraminidases [17],[18]. Approximately 150 cfu/ml of wt Cc5 were inoculated to a culture of J774.1 macrophages in the presence of 1mM Neu5Ac2en and growth was monitored after 2, 6, 10 and 24 h (Fig. 6A). Between 2 and 24 h post infection, counts of wt Cc5 were significantly reduced to values close to the siaC mutant (Fig. 6B). These data indicate that Neu5Ac2en has affinity for the active site of SiaC and restricts the growth of Cc5 in the presence of J774.1 cells.10.1371/journal.ppat.1000164.g006Figure 6The sialidase inhibitor Neu5Ac2en decreases growth of wt C. canimorsus 5 in presence of macrophages.(A) Viable counts of approximately 150 bacteria grown in cRPMI in the presence of J774.1 cells for 2, 6, 10 and 24 h: siaC bacteria (light grey); wt Cc5 bacteria with 1mM Neu5Ac2en (grey, dotted line); wt Cc5 in the absence of inhibitor (black). Mean values from 4 experiments and s.d. are shown including statistical difference between Cc5 and siaC in grey or between Cc5 and Cc5 treated with Neu5Ac2en in black with * p<0.05, ** p<0.01 and *** p<0.001 (2-tailed unpaired Student's t test). (B) Data from viable counts (mean) shown in (A) is represented as the fold difference compared to wt Cc5. The statistical difference is depicted from (a) with * p<0.05, ** p<0.01 and *** p<0.001.\nCc5 but not siaC is able to persist in murine tissue cagesTo test whether sialidase could play a role during C. canimorsus systemic infection, we selected a murine tissue cage infection model [19]. Around 107 cfu of wt or siaC Cc5 bacteria were injected directly into Teflon cages, which had been subcutaneously implanted in C57BL/6 mice. Colony forming units (cfu) counts of wt decreased on day 2 and 5. However, on day 9 they increased by 1 to 3 logs in 4 out of 5 mice, and were able to persist in 3 of 5 mice after 27 days post infection with more than 107 bacteria per ml fluid. The siaC bacteria were undetectable after the second day in 5 out of 5 infected mice (Fig. 7A). After infection, the total number of leukocytes in tissue cage fluid (1.8×104 +/− 1.3×104 leukocytes/µl, mean +/− standard deviation, s.d.) did not significantly increase and was not related to the bacterial load, suggesting that Cc5 infection did not lead to strong leukocyte recruitment. This was in agreement with the suppression of inflammation, which we previously reported [13]. In mixed infections, the competitive index of siaC bacteria was 9.7×10−4, 5.8×10−7 and 4.7×10−7 on day 5, 9 and 14, respectively. As observed during infection with wt Cc5 alone, 3 mice out of 5 that were infected by both strains developed a persistent infection (Fig. 7B).10.1371/journal.ppat.1000164.g007Figure 7The sialidase mutant is hypo-virulent in a tissue cage mouse infection model.Tissue cages were implanted in C57BL/6 mice and infected with 107\nCc5 wt and siaC bacteria (n = 5) singly (A) or in competition (B). Bacteria were counted in tissue cage fluid (TCF) during 27 days (Cc5 = black circles; siaC = open circles). Individual values are shown; horizontal lines indicate the median value of each group. The black dotted line is the detection limit of 20 bacteria per ml TCF. (A) Cfu numbers between groups were significantly different on days 2, 5 and 9 with p<0.01 and on days 14 and 27 with p<0.05 (Mann Whitney test). (B) 107 cfu Cc5 and erythromycin resistant siaC were inoculated at a 1∶1 ratio. Bacterial numbers were analyzed for 27 days (n = 5). Viable counts between wt and siaC were significantly different on day 2, 5 and 9 with p<0.01 and on day 14 with p<0.05. (C) Ex vivo isolated leukocytes were resuspended in serum free RPMI and inoculated at a moi of 20 (2×106 bacteria) or 0.2 (2×104 bacteria) indicated with grey dotted lines and bacterial viable count was monitored after 24 h. Values represent the mean using TCF cells from 3 uninfected mice. TCF leukocytes consist of 68% +/− 4.8% polymorphonuclear neutrophils (PMNs), 18% +/− 3.2% monocytes and 9.1% +/− 3.7% macrophages. Wt and siaC numbers were significantly different with p<0.05 (*) and p<0.001 (**) using 2-tailed unpaired student's t test. (D) In vitro, Cc5 and siaC were tested in heart infusion broth with FBS inoculated at a 1∶1 ratio with approximately 100 bacteria total and bacterial growth was monitored for 2, 6, 10 and 24 h. (E) Viable counts after challenge with 2×106 (grey dotted line) Cc5 (black) or siaC (grey) grown for 24 h with J774.1 in cRPMI singly (control) or at a 1∶1 ratio (cross-feeding).The fluid from uninfected control cages was collected and the leukocytes and liquid were tested separately for their capacity to sustain growth of Cc5. Interestingly, wt Cc5 did not grow in the presence of the cell-free liquid (data not shown) but they grew in presence of leukocytes whereas siaC bacteria did not (Fig. 7C). Both strains grew equally well in heart infusion broth supplemented with 10% FBS, indicating a similar fitness in vitro (Fig. 1C). Mixed cultures in heart infusion broth supplemented with 10% FBS showed comparable growth of wt and siaC bacteria. Both strains reached 106 cfu/ml after 24 h (Fig. 7D).Our data from mixed infection in mice suggest that there is no cross-feeding of nutrients between wt and siaC Cc5 (Fig. 7B). We thus tested whether there would be cross-feeding between wt and siaC bacteria when inoculated to J774.1 cultures. When wt and siaC were inoculated together at 1∶1 ratio to J774.1 cells, wt Cc5 reached 108 cfu/ml while siaC bacteria only reached 3×106 cfu/ml 24 h post infection (Fig. 7E).Taken together, these results demonstrate that SiaC plays an essential role in allowing persistence of wt Cc5 in this tissue cage model and that clearance of siaC bacteria is not due to a growth defect per se but to an altered interaction of the mutant with the host. Since sialidase is surface-exposed, one could consider the possibility that it alters the susceptibility to complement. Hence, we checked the susceptibility of wt and siaC Cc5 to mouse complement and found no difference (data not shown). It is also very unlikely that siaC bacteria have an increased sensitivity to killing by mouse leukocytes. Indeed, we tested phagocytosis and killing by human polymorphonuclear leukocytes and found no difference between wt and siaC bacteria (Manuscript in preparation). Hence, we conclude that the role of sialidase in infected mice is essentially nutritional.DiscussionSialic acids are a family of nine carbon acid sugars among which Neu5Ac is one of the most widespread variants. Sialic acids are predominantly found at the terminal position of cell-surface and secreted eukaryotic glycan structures and are involved in many physiological processes including binding to microbes and down-regulation of innate immunity [20]–[22]. Therefore it is not surprising that sialic acids play a role in a variety of host microbe interactions. Several pathogens have evolved ways to expose sialic acid on their surface and hence to escape complement killing and opsonization by mimicry. Sialic acids are incorporated into capsules by E. coli K1 [23], Group B Streptococci\n[24], Serogroups B, C, W135 and Y Neisseria meningitidis\n[25]. The lipooligosaccharide of Neisseria gonorrhoeae, Neisseria meningitidis and Haemophilus influenzae are also sialylated [26]. In this case, a bacterial sialyltransferase uses CMP-Neu5Ac from the host as a substrate [26]. Sialic acids can also be synthesized from lactate by Neisseria itself, demonstrating a close link between metabolism and evasion of innate immune defenses [27].Besides molecular mimicry, many microbes can utilize sialic acids as a source of carbon and nitrogen like E. coli K1, H. influenzae or C. perfringens\n[28]. Their metabolism comprises a permease for uptake and a neuraminiate lyase for conversion to N-acetyl mannosamine, which is either degraded or used in sialic acid biosynthesis. A number of commensal and pathogenic bacteria are also endowed with a sialidase, a glycosylhydrolase that cleaves sialic acid from sialo-glycoconjugates. Bacterial sialidases have been thought since a long time to contribute to virulence in bacteria that colonize mucosal surfaces such as Vibrio cholerae, Streptococcus pneumoniae, group B streptococci, C. perfringens and B. fragilis but the exact role of sialidase on virulence remains controversial [29]. Recently, it was shown that a sialidase is involved in the formation of Pseudomonas aeruginosa biofilms and hence contributes to colonization of the lungs during the initial stages of infection in cystic fibrosis patients [30]. In S. pneumoniae, a sialidase initiates an extensive deglycosylation of different host proteins, including IgA1 and human secretory component [31]. Furthermore, the sequential action of exoglycosidases sustains growth of S. pneumoniae on human α-1 acid glycoprotein, though growth is not as robust as on sucrose and lactose. Although the genetic analysis suggests that sugars from the glycan chain would sustain growth, this has not been shown directly [32].In the present study, which is among the very first on the pathogenesis mechanisms of C. canimorsus, we demonstrate that a sialidase allows C. canimorsus to feed on glycan chains from glycoproteins. The role of sialidase is not to supply sialic acid since growth of the sialidase-deficient mutant could not be restored by adding sialic acid to the culture medium. Thus, we have a situation similar to that of S. pneumoniae: the role of sialidase is to provide access to masked sugars of surface-exposed glycoproteins. Growth of the sialidase-deficient mutant could be restored by amino sugars like GalNAc, GlcNAc and LacNAc but not by glucose, galactose, mannose or sialyl-lactose, indicating that the nutritional requirements of C. canimorsus are very different from those of S. pneumoniae.\nOur study thus confirms the importance of a sialidase to initiate a deglycosylation process for bacterial metabolism. Moreover, in comparison with S. pneumoniae, C. canimorsus uses sialidase to feed on glycoproteins exposed at the surface of epithelial cells or even of macrophages, in spite of the fact that they do not adhere to these cells. The observation of extracellular bacteria specifically feeding on the surface of epithelial cells is not unprecedented. It has been described for B. thetaiotaomicron, a major commensal from the intestine, which feeds on fucosylated intestinal cells. Colonization by B. thetaiotaomicron even triggers the appearance of fucosyltransferase and fucosylated glycan expression [33]. Recent studies showed that host acquired fucose is incorporated by B. fragilis into capsular polysaccharide or glycoproteins, which in turn provides a survival advantage in the mammalian intestinal ecosystem [34]. As for C. canimorsus, it is likely that the capacity to feed on HeLa cells reflects the adaptation to feed on buccal epithelial cells.Sialidase, which is pivotal in this feeding process, is surface localized and this surface localization is a prerequisite for unmasking glycan structures. It is not common to find enzymes anchored into the outer membrane, facing the outside of Gram-negative bacteria but there are examples like pullulanase a 116-kDa isoamylase of Klebsiella oxytoca\n[35]. Not surprisingly, SiaC is endowed with an N-terminal signal sequence, which turned out to be critical for its targeting. Sialidase thus crosses the cytoplasmic membrane via the Sec pathway but we have at present no explanation on how it crosses the outer membrane and remains anchored. It is probably not by a C. canimorsus specific mechanism since sialidase appeared to be also surface-exposed when expressed in E. coli (unpublished data). Sialidase could be a lipoprotein, like pullulanase. Alternatively, sialidase could be a surface-anchored auto-transporter protein like the Y. enterocolitica YadA [36]. However, the fact that the C-terminus of sialidase is not involved in the surface localization (unpublished data) argues against this hypothesis. Work in progress tries to address the question of how sialidase is anchored in the outer membrane.Unlike what is observed with pullulanase, our data indicate that extremely little sialidase is released from C. canimorsus. This observation is in perfect agreement with the fact that C. canimorsus needs to be in direct contact with cells to feed on them. It also makes sense in the context of the mouth commensal microflora. Indeed, the oral cavity is occupied by some 500 different bacterial strains [37],[38], creating a fierce competition for nutrition. The fact that C. canimorsus does not release this enzyme suggests that C. canimorsus maximizes the benefit of sialidase by not sharing this fitness factor with competing bacteria. In agreement with this hypothesis, there is no cross-feeding when wt and siaC Cc5 bacteria are inoculated together in the presence of macrophages. This implies that wt C. canimorsus must be extremely efficient in capturing the aminosugars that it extracts from the surface of cells and we hypothesize that C. canimorsus has dedicated high affinity transporters for these in its outer membrane.Extracellular C. canimorsus replicated very efficiently not only when they were in direct contact with HeLa cells but also with J774.1 macrophages. Thus, C. canimorsus not only resists phagocytosis by cultured macrophages [13],[14], but they even take advantage of macrophages whose normal function is to engulf and kill microbes. To our knowledge, this is the very first report of a pathogen that can feed on phagocytic cells. This observation suggests that sialidase could contribute to virulence. We used a mouse tissue cage model in which the readout is bacterial persistence and we observed a dramatic difference in persistence between wt and sialidase-deficient C. canimorsus. Even more, we gained evidence that in vivo, C. canimorsus also feeds on phagocytes. These observations confirm our hypothesis that sialidase contributes to virulence, at least in the mouse model. It seems reasonable to extrapolate that it also plays a critical role during human infections. We would however be reluctant to call sialidase a virulence factor since it most probably evolved as a fitness factor for commensalism in the dog's mouth. Nevertheless, the mouse experiment shows that it may become a persistence factor if C. canimorsus is introduced in the tissues from another host. Our study thus shows once again the link between metabolism and virulence, as already well documented in studies on Salmonella\n[39], Listeria\n[40] and Neisseria\n[27]. However, unlike what was seen with Salmonella, there seem to be no or very little redundancy in the in vivo metabolism of C. canimorsus since the loss of sialidase had dramatic consequences on growth. It is interesting to observe that nutrition in vivo may be quite specific in spite of a very rich nutritional environment. Indeed, only GlcNAc and GalNAc could rescue growth while glucose had no effect and galactose was even deleterious. This difference could result from the fact that unlike Salmonella, C. canimorsus is a commensal highly adapted to its niche and only exceptionally a pathogen. Specialization is probably the hallmark of a bacterium that is primarily a commensal and only rarely a pathogen. Finally, C. canimorsus represents one more example illustrating that the distinction between commensals and pathogens is illusive. Commensalism and pathogenesis are two faces of the same coin.Influenza neuraminidases have been successfully targeted with chemotherapeutic inhibitors for prophylaxis and treatment [41]. Given the wide prevalence and important role of sialidases in microbial infections, inhibition of bacterial sialidases could also provide a mechanism to prevent bacterial spreading during infections. Here, we observed a significant inhibition of the growth of C. canimorsus in the presence of macrophages by Neu5Ac2en. These preliminary data indicate that microbial sialidases could indeed serve as an attractive drug target to prevent bacterial dissemination.Materials and MethodsBacterial strains and growth conditions\nC. canimorsus 5 was routinely grown on Heart Infusion Agar (HIA; Difco) supplemented with 5% sheep blood (Oxoid) for 2 days at 37°C in presence of 5% CO2. Bacteria were harvested by gently scraping colonies off the agar surface, washed and resuspended in PBS. C. canimorsus was also grown in Heart Infusion Broth (Difco) supplemented with 10% (v/v) fetal bovine serum (FBS; Invitrogen) for approximately 24 h without shaking in an 37°C incubator with 5% CO2. Selective agents were added at the following concentrations: erythromycin, 10 µg/ml; cefoxitin, 10 µg/ml; gentamicin, 20 µg/ml; ampicillin, 100 µg/ml.Cell Culture and InfectionMurine monocyte-macrophage J774A.1 cells (ATCC TIB-67) were cultured in RPMI 1640 (Invitrogen) supplemented with 10% (v/v) FBS (Invitrogen), 2 mM L-glutamine and 1 mM sodium pyruvate. Human epithelial HeLa cells (ATCC CCL-2) and canine epithelial MDCK kidney cells (ATCC CCL-34) were grown in DMEM (Invitrogen) with 10% (v/v) FBS. Cells were seeded in medium without antibiotics at a density of 105/cm2 and cultured at 37°C in humidified atmosphere containing 5% CO2. Unless otherwise indicated, infection was performed after 15 h at a moi of 20 representing 2×106 bacteria per ml in each well at 37°C.Monosaccharides and disaccharides (Sigma Aldrich) were added to 0.1% (w/v) final concentration. Neu5Ac and CMP- Neu5Ac were added to 0.01% final concentration.\nCc5 was pretreated with 1mM Neu5Ac2en at 37°C for 30 min. Subsequently, infection of J774.1 was carried out in presence of 1 mM Neu5Ac2en during 24 h.Arbitrarily Primed PCRPrimers specific to the ends of the transposon and primers of random sequence that may anneal to chromosomal DNA sequences in close proximity to the transposon insertions were used in two rounds of PCR before sequencing. The first round of amplification was carried out in 50 µl containing 100 ng of genomic DNA, 1.5 mM MgCl2, 200 µM of primers 5′ CAGAATTCTGTTGCATTTGCAAGTTG 3′ complementary to Tn4351 and 5′ggccacgcgtcgactagtacNNNNNNNNNNacgcc3′, 2.5 U of DNA polymerase (DyNAzymeII, Finnzymes), 200 µM of each dNTP, in 10 mM Tris HCl (pH 8.3) for 6 cycles (94°C for 1 min, 30°C for 1 min, 72°C for 2 min) and 30 cycles (94°C for 1 min, 45°C for 1 min, 72°C for 2 min) and final 10 min at 72°C. 10 µl of PCR product containing random fragments was used as template in a second round of 30 cycles of amplification (94°C for 30 sec, 45°C for 30 sec, 72°C for 1 min) using primers 5′ CAGAATTCTGTTGCATTTGCAAGTTG 3′ and 5′ GGCCACGCGTCGACTAGTAC 3′, from the 5′ of the random primer. PCR products were purified using NucleoSpin® from Machery Nagel. 20- 50 ng of random sized products were sequenced using an ABI sequencer. The Tn integration site was further confirmed by using primers on chromosomal DNA by sequencing towards the Tn integration site. Primers used were 5′ AATTGTTGTAACGATTGTCG 3′ or 5′ GCGAAGCGTTATCCCAAAGC 3′ complementary to the siaC sequence in a sequencing reaction containing 2 µg genomic DNA of siaC, betaine 0.25 M and BigDye Terminator Ready Reaction (PE Biosystems) with an initial denaturation step for 5 min and subsequent 99 cycles (95°C for 30 sec, 50°C for 20 sec, 60°C for 4 min).RNA isolation and reverse transcription (RT) PCR\nCc5 were grown for 2 days on HIA blood plates. RNA was isolated from 5×108 bacteria by a hot phenol/chloroform extraction method followed by DNase I (Amersham Pharmacia) treatment (0.5 U/µg RNA) for 2 h at 37°C. RNA was further cleaned by using a RNeasy kit (Quiagen) and stored at −80°C until use. An additional DNase I digest was introduced with 0.25 U/µg RNA for 15 min at 37°C and stopped by addition of final 2.5 mM EDTA and heat inactivation at 75°C for 10 min. Subsequent reverse transcription was performed with 50 U Superscript II/µg RNA in RT buffer (Invitrogen), 10 mM DTT and 50 µM specific primer (5129: 5′ GGGTAATCCGCACTTGTCGGG3′ or 5132: 5′ GTTTAGTTCTTGATAAATTCC 3′) for 60 min at 42°C and stopped at 70°C for 10 min. 10% of cDNA preparation or of a preparation made without addition of reverse transcriptase was subjected to PCR using following primer combinations: 4130 (5′ GGGTAACAACAAAAACCACTG 3′)+5129; 4132 (5′ TATAAGAATAATTGGTGGGC 3′)+5129; 4130+5132. 100 ng of genomic DNA from Cc5 was used as a positive control of the PCR reactions.Construction of complementation and expression plasmidsFull length siaC was amplified with 5′ CATACCATGGGAAATCGAATTTTTTATCTT 3′ and 5′ GTTCTAGAGAGTTCTTGATAAATTCCTCAACTG 3′ primers and cloned into the E. coli- C. canimorsus shuttle vector pMM47.A [16] with NcoI and XbaI, leading to the insertion of a glycine at position 2 and a C- terminal histidine 6× tag in plasmid pMM52 (siaC\nFL). Forward primer 5′ AAAGCCATGGGAAACGTAATCGGCGGAGGCG 3′ was used with the same reverse primer to construct pMM50 (siaC\nΔ1–21), deleting the first 63 bp of siaC, but still including methionine and glycine at position 1 and 2, respectively, and using a C-terminal His 6× tag. The catalytic mutation in siaC of was introduced by site directed mutagenesis with an inverse PCR on pMM52, using primers 5′ GAAGGATTTGGGTGTTCGTGTATGTCG 3′ and 5′ CGACATACACGAACACCCAAATCCTTC 3′ leading to pMM59 (siaC\nY488C). Plasmids derived from pMM47.A contained the cfxA gene originating from Bacteroides sp. and could be selected in C. canimorsus with 10 µg/ml cefoxitin [16]. The beta-lactamase also present on pMM47.A was used as a selection marker in E. coli.The cDNAs encoding SiaCΔ1–21 (pHS2) were subsequently amplified using 5′ GGAATTCCATATGAACGTAATCGGCGGAGGC 3′ plus 5′ CGCGGATCCCTAGTTCTTGATAAATTCCTC 3′ and cloned into the expression vector pET15b(+) (Novagen). Plasmid pHS3 encoding SiaCΔ1–21,Y488C was constructed by site directed mutagenesis on template pHS2 using the same primers as described for pMM59. All constructs were sequenced with an ABI sequencer. The sequence of SiaC was deposited at GenBank (accession number: EU329392).Purification of recombinant SiaC and immunoblottingExpression of siaC constructs in E. coli BL21(DE3) was induced with 0.5 mM isopropyl-β-D-1-thiogalactopyranoside at A600 = 0.5 for 3 h. Proteins were purified by affinity chromatography using chelating Sepharose (Pharmacia) charged with NiSO4 according to the manufacturer's instructions. Samples were analyzed by SDS-PAGE by the system of Laemmli, and immunoblotted. Polyclonal serum from rabbit was generated against recombinant SiaCΔ1–21. The antigen was injected at days 0, 14, 28, and 56 with a final bleeding at day 80 (Laboratoire d'Hormonologie, Marloie, Belgium).MUAN hydrolysis107 bacteria were incubated with 0.006% 2′-(4-Methylumbelliferyl)-α-D-N-acetylneuraminic acid (MUAN) in 0.25 M sodium acetate pH 7.5 at 37°C for 3 min. Reactions were stopped with 50 mM Na2CO3 pH 9.6 and fluorescence was determined at 445 nm with a Wallac Victor2 1420 Multilabel counter (Perkin Elmer).Outer Membrane PreparationBacterial cells resuspended in PBS containing DNase and RNase (10 µg/ml), were sonicated on ice. Unbroken cells were removed at 3000× g for 15 min, and total membranes were collected at 20 000× g for 30 min at 4°C. The membranes were suspended in PBS and sarcosyl (N-Lauroylsarcosine sodium salt, Sigma) was added to a final concentration of 1% (v/v). After incubation on ice for 1 h, membranes were collected at 20 000× g for 30 min and resuspended in electrophoresis sample buffer and analyzed by SDS-PAGE by the system of Laemmli.Immunofluorescence of bacteria107 bacteria were incubated on poly-D-lysine (BD) coated glass slides for 1 h at 37°C and subsequently fixed with 3% paraformaldehyde for 15 min. Anti- SiaC polyclonal serum (1∶500) and a FITC conjugated secondary antibody (Goat Anti- Rabbit IgG, Southern Biotech) was used at 1 µg/ml and fluorescence was measured with a Leica DMIRE2 microscope. Pictures were taken with a digital camera (Hamamatsu Photonics) and analyzed with OpenLab software (version 3.1.2) and Adobe Photoshop CS3 (version 10.0.1).Lectin Staining105 J774.1 macrophages or HeLa epithelial cells were seeded on poly-D-lysine coated slides. Infection was carried out with 4×107 bacteria for 2 h. Uninfected cells were alternatively treated with purified recombinant SiaC at 100 ng/ml. Cells were fixed with 3% paraformaldehyde for 15 min. Biotinylated lectins SNA and PNA (Vector Laboratories) were incubated with cells at 2 µg/ml and 2.5 µg/ml, respectively, for 1 h. After washing with PBS, cells were treated with 1 µg/ml fluorescein conjugated streptavidin (Vector Laboratories) and fluorescence was determined on mounted slides (Vectashield, Vector Laboratories).Mice and tissue cage infection model12 week-old male C57BL/6 mice were maintained under pathogen-free conditions in the Animal Facility of the Department of Research, University Hospital Basel. Animal experiments were performed in accordance with the guidelines of the Swiss veterinary law. Teflon tissue cages were implanted subcutaneously in the back of anesthetized mice as previously described [19]. The cages consisted of closed Teflon cylinders (10 mm diameter, 30 mm length, internal volume 1.84 ml) with 130 regularly spaced 0.2 mm holes. 2 weeks after surgery, 200 µl of bacterial suspension was injected percutaneously into the cage. Prior to infection, sterility of the tissue cage was verified. Tissue cage fluid (TCF) was sampled at day 2, 5, 9, 14 and 27 and examined for leukocytes and bacterial viable counts. Leukocytes from TCF were quantified with a Coulter counter (Coulter Electronics) and differentiated by Diff-Quick (Medion Diagnostics) Wright staining of cytospins and examined under light microscopy. The percentage of viable leukocytes was assessed by trypan blue exclusion.The survival of siaC bacteria in the competition experiment was compared directly with wt Cc5 in individual animals giving a 1∶1 ratio of wt to mutant bacteria. The number of mutant (erythromycin resistant) and wt bacteria recovered from the TCF of animals was established by plating to media with and without erythromycin. The competitive index was calculated as the (number of mutant/wild-type bacteria recovered from animals)/(number of mutant/wild-type bacteria in the inoculum).Statistical analysisFor growth experiments, means and standard deviation (s.d.) were calculated and statistical significance was evaluated by using a two- tailed, unpaired Student's t test. Differences were determined to be significant when p<0.05. For in vivo experiments, individual mouse values are shown including the median value of each group. Mann Whitney test with the post hoc Bonferroni correction was used for comparison between Cc5 and siaC cfu numbers during infection.\n\nREFERENCES:\n1. BoboRANewtonEJ\n1976\nA previously undescribed gram-negative bacillus causing septicemia and meningitis.\nAm J Clin Pathol\n65\n564\n569\n1266816\n2. BrennerDJHollisDGFanningGRWeaverRE\n1989\n\nCapnocytophaga canimorsus sp. nov. (formerly CDC group DF-2), a cause of septicemia following dog bite, and C. cynodegmi sp. nov., a cause of localized wound infection following dog bite.\nJ Clin Microbiol\n27\n231\n235\n2915017\n3. PersCGahrn-HansenBFrederiksenW\n1996\n\nCapnocytophaga canimorsus septicemia in Denmark, 1982–1995: review of 39 cases.\nClin Infect Dis\n23\n71\n75\n8816132\n4. Le MoalGLandronCGrollierGRobertRBurucoaC\n2003\nMeningitis due to Capnocytophaga canimorsus after receipt of a dog bite: case report and review of the literature.\nClin Infect Dis\n36\ne42\n46\n12539089\n5. LionCEscandeFBurdinJC\n1996\n\nCapnocytophaga canimorsus infections in human: review of the literature and cases report.\nEur J Epidemiol\n12\n521\n533\n8905316\n6. CoyneMJComstockLE\n2008\nNiche-specific features of the intestinal bacteroidales.\nJ Bacteriol\n190\n736\n742\n17993536\n7. McBrideMJ\n2004\n\nCytophaga-flavobacterium gliding motility.\nJ Mol Microbiol Biotechnol\n7\n63\n71\n15170404\n8. DuchaudEBoussahaMLouxVBernardetJFMichelC\n2007\nComplete genome sequence of the fish pathogen Flavobacterium psychrophilum.\nNat Biotechnol\n25\n763\n769\n17592475\n9. SchuijffelDFvan EmpelPCPenningsAMvan PuttenJPNuijtenPJ\n2005\nSuccessful selection of cross-protective vaccine candidates for Ornithobacterium rhinotracheale infection.\nInfect Immun\n73\n6812\n6821\n16177359\n10. SegersPMannheimWVancanneytMDe BrandtKHinzKH\n1993\n\nRiemerella anatipestifer gen. nov., comb. nov., the causative agent of septicemia anserum exsudativa, and its phylogenetic affiliation within the Flavobacterium-Cytophaga rRNA homology group.\nInt J Syst Bacteriol\n43\n768\n776\n8240957\n11. SubramaniamSHuangBLohHKwangJTanHM\n2000\nCharacterization of a predominant immunogenic outer membrane protein of Riemerella anatipestifer.\n\nClin Diagn Lab Immunol\n7\n168\n174\n10702488\n12. TierneyDMStraussLPSanchezJL\n2006\n\nCapnocytophaga canimorsus mycotic abdominal aortic aneurysm: why the mailman is afraid of dogs.\nJ Clin Microbiol\n44\n649\n651\n16455937\n13. ShinHMallyMKuhnMParozCCornelisGR\n2007\nEscape from immune surveillance by Capnocytophaga canimorsus.\nJ Infect Dis\n195\n375\n386\n17205476\n14. MeyerSShinHCornelisGR\n2008\n\nCapnocytophaga canimorsus resists phagocytosis by macrophages and blocks the ability of macrophages to kill other bacteria.\nImmunobiology: in press\n15. CooperAJKalinowskiAPShoemakerNBSalyersAA\n1997\nConstruction and characterization of a Bacteroides thetaiotaomicron recA mutant: transfer of Bacteroides integrated conjugative elements is RecA independent.\nJ Bacteriol\n179\n6221\n6227\n9335266\n16. MallyMCornelisGR\n2008\nGenetic tools for Capnocytophaga canimorsus.\nAppl Environ Microbiol: in press\n17. SchreinerEZbiralEKleineidamRGSchauerR\n1991\n2,3-Didehydro-2-deoxysialic acids structurally varied at C-5 and their behaviour towards the sialidase from Vibrio cholerae.\nCarbohydr Res\n216\n61\n66\n1797392\n18. HoyerLLRoggentinPSchauerRVimrER\n1991\nPurification and properties of cloned Salmonella typhimurium LT2 sialidase with virus-typical kinetic preference for sialyl alpha 2----3 linkages.\nJ Biochem\n110\n462\n467\n1769974\n19. KristianSALauthXNizetVGoetzFNeumeisterB\n2003\nAlanylation of teichoic acids protects Staphylococcus aureus against Toll-like receptor 2-dependent host defense in a mouse tissue cage infection model.\nJ Infect Dis\n188\n414\n423\n12870123\n20. SchauerR\n2000\nAchievements and challenges of sialic acid research.\nGlycoconj J\n17\n485\n499\n11421344\n21. TravingCSchauerR\n1998\nStructure, function and metabolism of sialic acids.\nCell Mol Life Sci\n54\n1330\n1349\n9893709\n22. VarkiNMVarkiA\n2007\nDiversity in cell surface sialic acid presentations: implications for biology and disease.\nLab Invest\n87\n851\n857\n17632542\n23. BarryGT\n1959\nDetection of sialic acid in various Escherichia coli strains and in other species of bacteria.\nNature\n183\n117\n118\n24. WesselsMRRubensCEBenediVJKasperDL\n1989\nDefinition of a bacterial virulence factor: sialylation of the group B streptococcal capsule.\nProc Natl Acad Sci U S A\n86\n8983\n8987\n2554337\n25. BhattacharjeeAKJenningsHJKennyCPMartinASmithIC\n1975\nStructural determination of the sialic acid polysaccharide antigens of Neisseria meningitidis serogroups B and C with carbon 13 nuclear magnetic resonance.\nJ Biol Chem\n250\n1926\n1932\n163259\n26. MandrellREApicellaMA\n1993\nLipo-oligosaccharides (LOS) of mucosal pathogens: molecular mimicry and host-modification of LOS.\nImmunobiology\n187\n382\n402\n8330904\n27. ExleyRMShawJMoweESunYHWestNP\n2005\nAvailable carbon source influences the resistance of Neisseria meningitidis against complement.\nJ Exp Med\n201\n1637\n1645\n15897277\n28. VimrERKalivodaKADeszoELSteenbergenSM\n2004\nDiversity of microbial sialic acid metabolism.\nMicrobiol Mol Biol Rev\n68\n132\n153\n15007099\n29. CorfieldT\n1992\nBacterial sialidases–roles in pathogenicity and nutrition.\nGlycobiology\n2\n509\n521\n1472757\n30. SoongGMuirAGomezMIWaksJReddyB\n2006\nBacterial neuraminidase facilitates mucosal infection by participating in biofilm production.\nJ Clin Invest\n116\n2297\n2305\n16862214\n31. KingSJHippeKRWeiserJN\n2006\nDeglycosylation of human glycoconjugates by the sequential activities of exoglycosidases expressed by Streptococcus pneumoniae.\nMol Microbiol\n59\n961\n974\n16420364\n32. BurnaughAMFrantzLJKingSJ\n2007\nGrowth of Streptococcus pneumoniae on human glycoconjugates is dependent upon the sequential activity of bacterial exoglycosidases.\nJ Bacteriol\n33. BryLFalkPGMidtvedtTGordonJI\n1996\nA model of host-microbial interactions in an open mammalian ecosystem.\nScience\n273\n1380\n1383\n8703071\n34. CoyneMJReinapBLeeMMComstockLE\n2005\nHuman symbionts use a host-like pathway for surface fucosylation.\nScience\n307\n1778\n1781\n15774760\n35. PugsleyAPKornackerMGRyterA\n1990\nAnalysis of the subcellular location of pullulanase produced by Escherichia coli carrying the pulA gene from Klebsiella pneumoniae strain UNF5023.\nMol Microbiol\n4\n59\n72\n2181241\n36. KoretkeKKSzczesnyPGruberMLupasAN\n2006\nModel structure of the prototypical non-fimbrial adhesin YadA of Yersinia enterocolitica.\nJ Struct Biol\n155\n154\n161\n16675268\n37. KroesILeppPWRelmanDA\n1999\nBacterial diversity within the human subgingival crevice.\nProc Natl Acad Sci U S A\n96\n14547\n14552\n10588742\n38. PasterBJBochesSKGalvinJLEricsonRELauCN\n2001\nBacterial diversity in human subgingival plaque.\nJ Bacteriol\n183\n3770\n3783\n11371542\n39. BeckerDSelbachMRollenhagenCBallmaierMMeyerTF\n2006\nRobust Salmonella metabolism limits possibilities for new antimicrobials.\nNature\n440\n303\n307\n16541065\n40. GoetzMBubertAWangGChico-CaleroIVazquez-BolandJA\n2001\nMicroinjection and growth of bacteria in the cytosol of mammalian host cells.\nProc Natl Acad Sci U S A\n98\n12221\n12226\n11572936\n41. von ItzsteinM\n2007\nThe war against influenza: discovery and development of sialidase inhibitors.\nNat Rev Drug Discov\n6\n967\n974\n18049471\n42. QuevillonESilventoinenVPillaiSHarteNMulderN\n2005\nInterProScan: protein domains identifier.\nNucleic Acids Res\n33\nW116\n120\n15980438\n43. RoggentinPRotheBKaperJBGalenJLawrisukL\n1989\nConserved sequences in bacterial and viral sialidases.\nGlycoconj J\n6\n349\n353\n2562507\n44. VarkiA\n2007\nGlycan-based interactions involving vertebrate sialic-acid-recognizing proteins.\nNature\n446\n1023\n1029\n17460663"
+}

batch_3/PMC2533656.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533656",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533656\nAUTHORS: Beatrix Groneberg-Kloft, Carolin Kreiter, Tobias Welte, Axel Fischer, David Quarcoo, Cristian Scutaru\n\nABSTRACT:\nBackgroundHistorical, social and economic reasons can lead to major differences in the allocation of health system resources and research funding. These differences might endanger the progress in diagnostic and therapeutic approaches of socio-economic important diseases. The present study aimed to assess different benchmarking approaches that might be used to analyse these disproportions. Research in two categories was analysed for various output parameters and compared to input parameters. Germany was used as a high income model country. For the areas of cardiovascular and respiratory medicine density equalizing mapping procedures visualized major geographical differences in both input and output markers.ResultsAn imbalance in the state financial input was present with 36 cardiovascular versus 8 respiratory medicine state-financed full clinical university departments at the C4/W3 salary level. The imbalance in financial input is paralleled by an imbalance in overall quantitative output figures: The 36 cardiology chairs published 2708 articles in comparison to 453 articles published by the 8 respiratory medicine chairs in the period between 2002 and 2006. This is a ratio of 75.2 articles per cardiology chair and 56.63 articles per respiratory medicine chair. A similar trend is also present in the qualitative measures. Here, the 2708 cardiology publications were cited 48337 times (7290 times for respiratory medicine) which is an average citation of 17.85 per publication vs. 16.09 for respiratory medicine. The average number of citations per cardiology chair was 1342.69 in contrast to 911.25 citations per respiratory medicine chair. Further comparison of the contribution of the 16 different German states revealed major geographical differences concerning numbers of chairs, published items, total number of citations and average citations.ConclusionDespite similar significances of cardiovascular and respiratory diseases for the global burden of disease, large input and output imbalances have been revealed in the present study which point to a need for changes in funding policies. The present study supplies data that could be used for decision making in the field of health systems funding.\n\nBODY:\nBackgroundDiseases of the cardiovascular system play an important role in health care. They have a great impact on the burden of disease. This burden of disease is defined as the impact of a health care problem in an area measured by financial cost, mortality, morbidity, or other indicators. Quantification is often performed using Disability-adjusted life years (DALYs) [1] or Quality-adjusted life years (QALYs) [2,3]. These measures combine the burden due to both death and morbidity into one index. With regard to the different disorders listed in global and national burden of disease rankings, also diseases of the respiratory play a prominent role [4]. In this respect, four out of the ten most common causes of death are respiratory diseases [5]. In view of the enormous socio-economic burden, it should be anticipated that a large proportion of health research funding is allocated to this field of medicine.In contrast to these features, health system resources and research funding policy are often debated as being disproportional. Numerous publications discussed this issue i.e. in the field of neurosciences [6], cardiovascular medicine [7], gastroenterology [8], genetics [9] or stem cell research [10-12]. These areas are heavily funded by governmental and non-governmental sources and there are various statements concerning policy guidelines available [13-18].For the high income country Germany, especially the fields of respiratory medicine and cardiology are interesting areas for health and research funding allocation policy. In this respect, the present study aimed to 1) identify and compare different output figures 2) relate these figures to selected input figures 3) provide data in relation to geographical information.MethodsOutput benchmarking data sourceData for output benchmarking (published items and citations) was retrieved from the biomedical database Web of Science (Thomson Institute for Scientific Information, ISI) [19,20].Search strategiesFor the different searches, phrases joined together with Boolean operators, i.e. AND, OR and NOT were used.Time frameA time frame was set and all entries between the years 2002 and 2006 were analysed.Input benchmarking data sourceData for input benchmarking was retrieved from internet searches and the German Lung White Book [21]. All full professorships/chairs (W3/C4 salary level) of medical school departments of cardiology and respiratory medicine were identified (begin of analysis 2007-08-01, last update 2008-4-30) and related to the respective German states. In this respect, the numbers of full professorships/chairs were calculated for each of the 16 German states (i.e. 3 chairs for cardiology in the state of Berlin and 0 for the state of Brandenburg) and density equalizing mapping performed. The numbers were related to each German state since German medical schools are financed directly by the single states and not by the Federal Republic of Germany. Using the calculation for each state, an input analysis was possible for cardiology and respiratory medicine.Output quantity analysis: Comparison of number and origin of publications in relation to the fieldTo perform a comparison between respiratory medicine and cardiology, published items were screened. All full professors/chairmen of medical school departments of cardiology and respiratory medicine were identified by name and their publications between 2002 and 2006 were recorded (begin of analysis 2007-08-01, last update 2008-4-30). For density equalizing procedures, only the publication type \"article\" was used. The entries of all full professors/chairmen for each of the 16 German states were added in order to establish a formula for each of the 16 German states for geographical distribution.Output quality analysis: Comparison of citations in relation to the fieldTo perform a qualitative comparison between respiratory medicine and cardiology, published items were related to their citations. Parallel to the quantity analysis, all full professors/chairmen of medical school departments of cardiology and respiratory medicine were identified by name and the numbers of citations of their publications between 2002 and 2006 were recorded (begin of analysis 2007-08-01, last update 2008-4-30). For density equalizing procedures, only the publication type \"article\" was used. The citations of all full professors/chairmen of each state were added in order to establish a formula for each of the 16 German states for geographical distribution.Density-equalizing mappingThe method of density-equalizing mapping was used following a recently described method [22] basing on Gastner and Newman's algorithm [23]. In brief, territories were re-sized from the original size (additional file 1) according to a particular variable, i.e. the number of published items or the citations. For the re-sizing procedure the area of each state was scaled in proportion to its total number of published items or citations.ResultsInput benchmarking: Spatial distribution of full professorships/chairsLarge differences were found in the input benchmark \"number of full professorships/chairs\" (highest level – W3/C4) of cardiology versus respiratory medicine. For cardiology, it was found that the total number of 34 medical school/faculties in Germany have established 36 full professorships/chairs (W3/C4 salary level) of cardiology. In this respect, the medical faculty of the Charité in Berlin has three independent departments of cardiology that are directed by three separate full professorships/chairs of cardiology. Also, Munich has two full professorships/chairs of cardiology which belong to two separate medical faculties. Every other university has an own department of cardiology (Fig. 1a). Density equalizing mapping calculations visualizes the geographical distribution. Every state apart from Brandenburg and Bremen has financed a medical school department for cardiology. In these states, there are no medical schools (Fig. 1a).Figure 1Density equalizing mapping of full professorships/chairs in relation to single German states in 2008. Cardiology (A) vs. Respiratory Medicine (B). Greyscales encode number of professorships per state.In contrast to these input figures, the area of respiratory medicine is represented by only 8 independent clinical full professorships/chairs (W3/C4 level) of respiratory medicine (fig. 1b). Two out of the eight are situated in Hesse at the same faculty of medicine (Justus-Liebig-University Giessen). The following states do not finance an independent full clinical professorship/chair of respiratory medicine: Berlin, Brandenburg, Bremen, Hamburg, Bavaria, Saxony, Saxony-Anhalt, Thuringia which leads to a major distortion in the density equalizing map (fig. 1b).Quantity output benchmarking: Total numbers and spatial distribution of published items per stateThe comparison of respiratory medicine and cardiology concerning the benchmark of total numbers of published items of full professors of each German state demonstrated large quantitative differences between the two fields of medicine and the different states.For the German full professorship of clinical cardiology, an overall number of 2708 published items was found. The ranking was headed by North Rhine-Westfalia (#1 with 610 published items), followed by Bavaria (#2 with 413), Baden-Württemberg (#3 with 369), Berlin (#4 with 292), Saxony (#5 with 219), Lower Saxony (#6 with 163), Hesse (#7 with 134), Mecklenburg-Western Pomerania (#8 with 122), Hamburg (#9 with 88), Rhineland-Palatinate (#10 with 86), Schleswig-Holstein (#11 with 72), Thuringia (#12 with 69), Saxony-Anhalt (#13 with 60) and Saarland (#14 with 11). The states Brandenburg and Bremen do not have a medical faculty. Density equalizing mapping approaches were used to analyse the distribution and it was found that the states Lower Saxony, Saxony-Anhalt, Bremen and Brandenburg were distorted (Fig. 2a) in comparison to their natural shape (additional file 1).Figure 2Density equalizing mapping of total numbers of published items of the full professorships (C4/W3) per German state between 2002 and 2006. Cardiology (A) vs. Respiratory Medicine (B). Greyscales encode total numbers of published items per state in the publication category \"article\".For respiratory medicine, lower numbers were recorded in general with an overall number of 453 published items for all 8 chairs. In this field, Hesse led the field with a total number of 255, followed by Lower Saxony (66), North Rhine-Westfalia (52), Mecklenburg-Western Pomerania (36), Baden-Württemberg (25), and Schleswig-Holstein (19). The states Bavaria, Berlin, Hamburg, Rhineland-Palatinate, Saarland, Saxonia, Saxonia-Anhalt and Thuringia do not have an independent full professorship at the C4/W3 salary level despite having cardiology professorships. Bremen and Brandenburg do not have medical faculties and therefore neither cardiology nor respiratory full professorships. Density equalizing mapping calculations led to a strong distortion with Hesse and Lower Saxony dominating the map (Fig. 2b) in comparison to the natural shape (additional file 1).Quality output benchmarking: Citation numbers and spatial distribution of published items per stateLarge differences were also present between the two fields with regard to output quality benchmarking. In this respect, the total number of citations of cardiology articles was 48337 versus 7290 citations of respiratory medicine articles. The average citation per item also differed with 17.85 for cardiology articles and 4.34 for respiratory articles.To perform a detailed spatial comparison between respiratory medicine and cardiology, citations were related to the states of origin.In the field of cardiology, the citation ranking was partly different from the publication number ranking: Parallel to the publication number ranking, North Rhine-Westfalia headed this analysis with 10825 citations, followed by Saxony (#2 with 8078 citations), Bavaria (#3 with 6003), Baden-Württemberg (#4 with 5499), Lower Saxony (#5 with 4486), Berlin (#6 with 4351), Mecklenburg-Western Pomerania (#7 with 1826), Rhineland-Palatinate (#8 with 1695), Hesse (#9 with 1653), Schleswig-Holstein (#10 with 1585), Hamburg (#11 with 1066), Saxony-Anhalt (#12 with 469), Thuringia (#13 with 464) and Saarland (#14 with 337). The states Brandenburg and Bremen do not have a medical faculty.Density equalizing mapping calculations to led slight distortions of the map (Fig. 3a) in comparison to the natural shape (additional file 1).Figure 3Density equalizing mapping of citation numbers of the full professorships (C4/W3) per German state between 2002 and 2006. Cardiology (A) vs. Respiratory Medicine (B). Greyscales encode total numbers of citations per state in the publication category \"article\".For respiratory medicine, lower numbers of citations were recorded in general with an overall number of 7290 citations for all 8 chairs.In specific, Hesse ranked #1 with 5442 citations, followed by Lower Saxony (#2 with 682), North Rhine-Westfalia (#3 with 357 citations), Baden-Württemberg (#3 with 335), and Schleswig-Holstein (#4 with 172). The other states Bavaria, Berlin, Hamburg, Rhineland-Palatinate, Saarland, Saxonia, Saxonia-Anhalt and Thuringia do not have an independent full professorship at the C4/W3 salary level despite having cardiology professorships. Density equalizing mapping calculations again led to strong distortions in the map with Hesse dominating (Fig. 3b) in comparison to the natural shape (additional file 1) and to the shape in the cardiology citation density-equalizing map (Fig. 3a).Quality output benchmarking: Spatial distribution of average citations per itemDensity equalizing mapping approaches were also used to assess differences in the spatial distribution of average citations per item in both fields. These calculations based on the number of publications and citations in relation to the states of origin.In the field of cardiology, the calculations led to stronger distortions in the density-equalizing map than in the publication and citation number analysis (Fig. 4a). In this respect, the average citation per item analysis listed Saxony at the first position (36.89 citations per published item) followed by Saarland (30.64), Lower Saxony (27.52), Schleswig-Holstein (22.01), Rhineland-Palatinate (19.71), North Rhine-Westfalia (17.75), Mecklenburg-Western Pomerania (14.97), Berlin (14.9), Baden-Württemberg (14.9), Bavaria (14.54), Hesse (12.34), Hamburg (12.11), Saxony-Anhalt (7.82), Thuringia (6.72). Thus, the density equalizing map was dominated by Saxony and Saarland (Fig. 4a) in comparison to the natural shape (additional file 1).Figure 4Density equalizing mapping of average citations per published item of the full professorships (C4/W3) per German state between 2002 and 2006. Cardiology (A) vs. Respiratory Medicine (B). Greyscales encode average citations per article per state in the publication category \"article\".For respiratory medicine, Hesse was also ranked on first position in this analysis (Fig. 4b) with an average citation per published item of 21.34. Hesse was followed by Baden-Württemberg (#2 with 13.4 citations per item), Lower Saxony (#3 with 10.33), Schleswig-Holstein (#4 with 9.05), Mecklenburg-Western Pomerania (#5 with 8.39) and North Rhine-Westfalia (#6 with 6.87).DiscussionNumerous publications indicate that current settings for health system and research funding need review. Reasons are potential imbalances in the existing policy for funding allocation. The present study addressed this issue using Germany as a model high income country and the two socio-economic important fields of cardiovascular and respiratory medicine.Methodologically, we used both output and input benchmarking. Output benchmarking was divided into the quantitative measure of total number of published items (publications type \"article\") and the qualitative measure of citations. The later feature is partly debated as not being a very good tool to assess research quality [24,25] but other tools such as the H-index also bear limitations [26].In terms of input parameters, the present study is limited to the number of full professorships/chairs for cardiology vs. respiratory medicine per state. For Germany, this can be used as an indicator for governmental funding since the states are responsible for the financial support of the medical schools. In this respect, the state ministers for research and education are usually also responsible to establish full professorships/chairs. A further useful figure would have been to assess the funding for the two fields by federal funding institutions such as the German Research Council (DFG), the federal ministry for Education and research (BMBF) or the European Union and the industry [27,28]. However, the precise funding from these sources is not accessible since some institutions and departments do not uncover these figures. In specific, industry funding is often not published as demonstrated by the tobacco industry funding policies [29,30]. Therefore, the present study was limited to monitor only the state financial input in terms of established independent full professorships/chairs at medical faculties.A further potential bias within the methodology of the present study is related to the issue of linguistic differences as previously discussed [31]. In this respect, the present analyses encompassed all languages included in the data bases. The majority of publications is published in English and it is difficult for non-English journals to get included in the data bases. Therefore, numerous scientific publications in languages other than English are not accessible. However, the major German cardiovascular (Z Kardiol – Clinical Research in Cardiology) and respiratory journals (Pneumologie) are included in the data base. Also, it is generally accepted that German scientists publish their high quality research in scientific journals that use English as language.Large differences were present between the two fields: All medical faculties had chairs for cardiology. At the Berlin medical faculty, three cardiology chairs were present but not an independent single chair for respiratory medicine. This field was subordinated and headed by a full professorship for cardiology and a full professorship for infectious diseases. The presence of three independent cardiology chairs in Berlin is most probably due to historical reasons since this faculty was divided into two faculties during period of the Berlin wall and reunified in 2002/2003 [32].In striking contrast to the high number of cardiology chairs, only 8 chairs for respiratory medicine were present in Germany. The regional distribution as assessed by density equalizing mapping demonstrated a focus in North Rhine-Westfalia and Hesse. The largest state Bavaria did not have a chair of respiratory medicine.After the demonstration of an imbalance in the financial input (36 cardiology versus 8 respiratory medicine state-financed clinical university departments), the present study aimed to analyze potential imbalances in output figures. Therefore, the quantitative measure \"number of publications\" and the qualitative measures of \"overall numbers of citations\" and \"average citations per published item\" were used. In general, the imbalance in financial input is paralleled by an imbalance in overall quantitative output figures. I.e. the 36 cardiology full professorships published 2708 articles in comparison to 453 articles published by the 8 respiratory medicine full professorships. This is a ratio of 75.2 articles per cardiology chair and 56.62 articles per respiratory medicine chair. A similar trend is also present in the qualitative measures. Here, the 2708 cardiology professorship publications were cited 48337 times which is an average citation of 17.85 per publication. The average number of citations per cardiology chair was 1342.69. For respiratory medicine, the 453 publications were cited 7290 times. This is an average citation number of 16.09 per publication and a ratio of 911.25 citations per respiratory medicine chair.Interestingly, the citations per state and the number of publications per state varied to a large extend between the different states and the two fields of internal medicine. For respiratory medicine, the maximal number of publications per state was 255 for the 2 chairs in Hesse. These 255 publications were cited 5442 times which is an average citation per published item of 21.34. This is a ratio of 127.5 publications per professorship in Hesse and a ratio of 2721 citations per professorship in Hesse. By contrast, the best ratios in the field of cardiology were found for Saxony. Here, the ratio of publications per professorship was 109.5 and the ratio of citations per professorship was 4039.Closer analysis revealed that the most cited publications for the Saxony cardiology professorships were articles in which the full professor was co-author [33] whereas the most cited publications for the Hesse respiratory medicine professorships were senior authorships [34-36].The reasons for the presently analyzed imbalances are numerous: I.e. the high income country Germany is known to have an extremely low ratio of respiratory physicians in comparison to other European countries (as indicated in the European Lung White Book [37]. Therefore, a lower number of respiratory specialists may lead to a lower research activity. 2) The number of full professorships and department chairs for respiratory medicine at the highest level (C4/W3) is disproportional in Germany in comparison to other countries since there are 36 chairs for cardiology but only 8 for respiratory medicine. This imbalance leads to a lower research activity with a lower number of publication entries in the database.An important issue is related to the reason for this difference of 8 vs. 36 university chairs at German medical schools. Two major reasons may account for the imbalance: 1) History: in the times of tuberculosis at the beginning of the 20th century, respiratory disorders were treated in remote hospitals but not in university medical schools. When the faculties started to create new chairs for internal medicine after the second world war, respiratory medicine capacities was not present at the medical faculties but in remote hospitals and the denomination of the chairs was directed towards cardiology. As a result, respiratory medicine is now underrepresented at German faculties in comparison to i.e. the UK. 2) Economics: Interventional and diagnostic procedures in cardiology such as left heart catheter offer a larger financial benefit to the faculties than respiratory interventional and diagnostic procedures [38,39]. Therefore, economic features may direct the faculties to the direction of cardiology professorships. Future studies should analyse these imbalances in closer detail.It is difficult to interpret how input imbalance affects on the output ratios. I.e. the allocation of public and private funding to a specific field such as cardiology and the consecutive concentration of financial resources in specific fields may lead to an increase of research actors or promotion of networking between outside institutes in this area. This may then lead to increased research activities resulting in production of higher-impact publications, eventually, obtaining more funding. Eventually a circle structure may appear that leads to the phenomenon that the rich areas automatically get richer [40].In conclusion, the present study used input and output benchmarking in combination with density equalizing mapping to assess differences in the two important fields of cardiovascular and respiratory medicine. Germany was used a model high income country. A major imbalance in the state financial input was present with 8 respiratory medicine versus 36 cardiovascular state-financed full clinical university departments at the C4/W3 salary level. This difference in the state financial input was paralleled by large differences in overall quantitative output figures with 2708 published cardiology articles in comparison to 453 respiratory medicine in the period between 2002 and 2006. However, there was also a difference between the two fields in the qualitative citation analysis. Here, cardiology publications had an average citation of 17.85 per publication whereas the respiratory medicine publication had an average citation of 16.09 per publication. This small difference might be due to the fact that a larger number of professorships lead to a larger number of networking collaborations and citations. Despite a high significance of both cardiovascular and respiratory diseases for the burden of disease, large differences are present in Germany. This should be realized for health policy and research funding allocation.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsBGK, AF, TW, DQ, and CS contributed to the conception and design of the study, BGK and CK performed the analysis, BGK and CS prepared the first draft and all authors contributed to the writing of the manuscript.Supplementary MaterialAdditional file 1Map of the 16 German states. This geographic map of the 16 German states and their geographical position in Europe can be used as a matrix for the comparison with the density equalizing mappings in figures 1, 2, 3, 4.Click here for file\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2533664.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533664",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533664\nAUTHORS: Wen Ji Yuan, Takao Yasuhara, Tetsuro Shingo, Kenichiro Muraoka, Takashi Agari, Masahiro Kameda, Takashi Uozumi, Naoki Tajiri, Takamasa Morimoto, Meng Jing, Tanefumi Baba, Feifei Wang, Hanbai Leung, Toshihiro Matsui, Yasuyuki Miyoshi, Isao Date\n\nABSTRACT:\nBackgroundParkinson's disease (PD) is a neurological disorder characterized by the degeneration of nigrostriatal dopaminergic systems. Free radicals induced by oxidative stress are involved in the mechanisms of cell death in PD. This study clarifies the neuroprotective effects of edaravone (MCI-186, 3-methyl-1-phenyl-2-pyrazolin-5-one), which has already been used for the treatment of cerebral ischemia in Japan, on TH-positive dopaminergic neurons using PD model both in vitro and in vivo. 6-hydroxydopamine (6-OHDA), a neurotoxin for dopaminergic neurons, was added to cultured dopaminergic neurons derived from murine embryonal ventral mesencephalon with subsequet administration of edaravone or saline. The number of surviving TH-positive neurons and the degree of cell damage induced by free radicals were analyzed. In parallel, edaravone or saline was intravenously administered for PD model of rats receiving intrastriatal 6-OHDA lesion with subsequent behavioral and histological analyses.ResultsIn vitro study showed that edaravone significantly ameliorated the survival of TH-positive neurons in a dose-responsive manner. The number of apoptotic cells and HEt-positive cells significantly decreased, thus indicating that the neuroprotective effects of edaravone might be mediated by anti-apoptotic effects through the suppression of free radicals by edaravone. In vivo study demonstrated that edaravone-administration at 30 minutes after 6-OHDA lesion reduced the number of amphetamine-induced rotations significantly than edaravone-administration at 24 hours. Tyrosine hydroxylase (TH) staining of the striatum and substantia nigra pars compacta revealed that edaravone might exert neuroprotective effects on nigrostriatal dopaminergic systems. The neuroprotective effects were prominent when edaravone was administered early and in high concentration. TUNEL, HEt and Iba-1 staining in vivo might demonstrate the involvement of anti-apoptotic, anti-oxidative and anti-inflammatory effects of edaravone-administration.ConclusionEdaravone exerts neuroprotective effects on PD model both in vitro and in vivo. The underlying mechanisms might be involved in the anti-apoptotic effects, anti-oxidative effects, and/or anti-inflammatory effects of edaravone. Edaravone might be a hopeful therapeutic option for PD, although the high therapeutic dosage remains to be solved for the clinical application.\n\nBODY:\nBackgroundParkinson's disease (PD) is a neurodegenerative disorder characterized by slowly progressive degeneration of DA neurons in the substantia nigra pars compacta, with subsequent damage of nerve terminals accompanied by dopamine (DA) depletion in the striatum [1]. Although the neuropathological hallmarks of PD are well described, the etiology remains still undefined. However, accumulative evidences revealed many biochemical processes and molecular mechanisms as mediators of neuronal cell death in PD. Notably oxidative stress and mitochondrial dysfunction might be an important pillar of pathogenesis of PD [2].6-hydroxydopamine (6-OHDA) is widely used for experimental models of PD [3]. It damages cells with dopaminergic neuronal attribute, including human neuroblastoma SH-SY5Y [4], PC12 cells derived from rat pheochromocytoma [5] and rat ventral mesencephalic neurons [6]. Furthermore, it is also a specific neurotoxin for DA neurons in vivo [2,7]. Intracellular lipids, proteins or DNA are damaged with consequent impairment of cell function induced by 6-OHDA. Mitochondrial oxidative phosphorylation with subsequent energy deprivation and excrement of 6-OHDA-auto-oxidation, including quinones and hydrogen peroxide (H2O2) are deeply involved in the cytotoxic processes [8]. As above described, mitochondrial dysfunction and oxidative stress might play important roles in the pathogenesis of PD [2], thus indicating that the experimental model using 6-OHDA might have essential mechanisms in common with PD. Furthermore, anti-oxidant agents, such as catalase, vitamin E, N-acetyl cysteine, ascorbic acid and pyruvate might exert neuroprotection for 6-OHDA-treated DA neurons [9].Edaravone (3-methyl-1-phenyl-2-pyrazolin-5-one) is a potent scavenger of hydroxyl radicals, and is useful for patients suffering from ischemic stroke [10,11], with the involvement of peroxidation leading to neuronal cell death [12]. Neuroprotective effects of edaravone are explored using head trauma [13] and spinal cord ischemia [14]. Recent study demonstrated that edaravone suppress the production of nitric oxide and reactive oxygen species by activated microglia [15]. In both cerebral ischemia and PD, free radicals might be one of the critical pathogenesis which accelerates progression of disease. These results suggest that edaravone might have neuroprotective effects on 6-OHDA-treated DA neurons and might on slowly degenerated DA neurons in PD patients through anti-oxidative mechanisms.In this study, first we explored the neuroprotective effects of edaravone on 6-OHDA-induced toxicity against murine ventral mesencephalic (VM) cell cultures and the underlying mechanisms. After confirming the effects in vitro, edaravone was intravenously administered to 6-OHDA-lesioned PD model of rats and evaluated behaviorally and immunohistochemically.MethodsIn vitro model of Parkinson's diseaseCell preparationMurine DA neurons were cultured as described previously with minor modifications [16]. Tissue blocks of the ventral mesencephalon containing DA neurons were dissected from murine embryo (C57/B6) on day 14 of gestation after cervical dislocation with consequent trituration into single cell suspension. Cells were plated in mixed hormone MEM (MHM) supplemented with 1% fetal bovine serum at a density of 1 × 105 cells/well on poly-D-ornithine and fibronectin-coated glass slides in 24-well plates (Nunc, Frankfurt, FRG). Cultures were maintained at 37 degrees C in an atmosphere of 5% CO2 plus 95% air and with 100% relative humidity. Forty-eight hours after initial plating, the medium was exchanged and the cells were used for the further experiments. The average number of mesencephalic neurons was 0.42 ± 0.07 × 105 cells/well at the beginning of the experiment with TH-immunoreactivity in 37 ± 12% of total cells.Administration of 6-OHDA and edaravoneEdaravone (MCI-186, 3-methyl-1-phenyl-2-pyrazolin-5-one) was kindly provided from Mitsubishi Pharma (Japan). It was dissolved in 0.5 ml of 1 N NaOH and 8 ml of distilled water, and adjusted to pH 7 by addition of 1 N HCl. At 48 hours after the initial plating, the cultured cells were exposed to 40 μM 6-OHDA (Sigma) or PBS for 30 minutes, and then added 10-6, 10-5, 10-4 or 10-3 M edaravone, or saline as a control at 37 degrees C. The cells were incubated for 18 hours and developed to immunocytochemical investigations.ImmunocytochemistryCells were fixed with 4% paraformaldehyde (PFA) for 30 minutes and then washed three times for 5 minutes in PBS. They were incubated overnight at 4 degrees C with an antibody directed against tyrosine hydroxylase (TH, rabbit polyclonal IgG, 1: 500, Chemicon) with 10% normal goat serum (Vector). After several rinses in PBS, cells were incubated at room temperature for 30 minutes in sheep anti-rabbit IgG FITC conjugate (1: 500, Sigma) and 4',6-diamidino-2-phenylindole, dilactate (DAPI, 1: 500, Molecular Probes). The cells were then washed three times in PBS and mounted on albumin-coated slides and embedded with cover glass. After photographically captured, immunoreactive neurons were counted per high power field view selected at random (n = 3 in each well, 10,000 μm2). Six wells were assigned to each group for statistical analyses. Control studies involved exclusion of primary antibody substituted with 10% normal goat serum in PBS. No immunoreactivity was observed in these controls.TUNEL staining and HEt stainingIn order to explore the involvement of apoptosis in this study, a modified method for terminal deoxynucleotidyl transferase-mediated biotinylated UTP nick end labeling (TUNEL, Roche) and DAPI staining was also used. After edaravone (10-6-10-3 M) or saline were administered into separate series of 6-OHDA-treated DA neurons, cells were fixed at 18 hours as described in the previous section. TUNEL staining was performed according to the manufacturer's instruction.In order to detect the early production of superoxide anions after 6-OHDA addition, hydroethidine (HEt), selectively oxidized to ehidium by superoxide anions, was used. HEt (1 mg/ml in PBS) was administered to 6-OHDA-treated DA neuronal cell culture at 0, 10, 20 or 30 minutes after edaravone- or saline- administration. After 5-minute incubation with HEt, the cells were washed 3 times in PBS, fixed with PFA, washed with PBS containing DAPI and finally embedded with cover glass. The cells were observed using a fluorescent microscope at an excitation of 355 nm and an emission of 450 nm for HEt and stained cells were counted as described above [17].In vivo model of Parkinson's diseaseSubjectsWe used adult female Sprague-Dawley rats (Charles River, Japan) weighing 250–300 g at the beginning of the experiment, according to approved guidelines of the institutional animal care and use committee of Okayama University. They were housed two per cage in a temperature and humidity controlled room, maintained on a 12-hour light/dark cycle, and they had free access to food and water.Surgical proceduresSeventy eight rats were deeply anesthetized with sodium pentobarbital (30 mg/kg, i.p.) and placed in a stereotaxic instrument (Narishige, Japan). After pre-treatment of desipramine (25 mg/kg, i,p., Sigma), 20 μg of 6-OHDA (4 μl of 5 μg/μl dissolved in saline containing 0.2 mg/ml ascorbic acid; Sigma) was injected into the right striatum with a 28-gauge Hamilton syringe into the following coordinates: 1.0 mm anterior to the bregma, 3.0 mm lateral to the sagittal suture, and 5.0 mm ventral to the surface of the brain with tooth-bar set at 0 mm [18]. The injection rate was 1 μl/minute, and the syringe was left in place for an additional 5 minutes before being retracted slowly (1 mm/minute). At 30 minutes or 24 hours after 6-OHDA lesion, 30, 100, or 250 mg/kg of edaravone or saline (2 ml) were intravenously administered slowly from the right femoral vein.Behavioral testingAll rats were tested with amphetamine (2.5 mg/kg, Dainippon-Seiyaku, Japan) at 1 and 2 weeks after 6-OHDA lesion, and rotational behaviors were assessed for 60 minutes with a video camera. Full 360 degrees turns in the direction ipsilateral to the lesion were counted.Fixation and SectioningAt 2 weeks after 6-OHDA lesion, rats were deeply anesthetized with sodium pentobarbital (100 mg/kg), perfused from the ascending aorta with 200 ml of cold PBS, followed by 100 ml of 4% PFA in PBS. Brains were removed and post-fixed in the same fixative for 2 days followed by 30% sucrose in phosphate buffer (PB) until to be sunk completely. Six series of coronal sections were cut at a thickness of 40 μm with a freezing microtome and stored at -20 degrees C.ImmunohistochemistryFree floating sections for TH immunohistochemistry were blocked by 0.3% hydroxygen peroxide in methanol for 3 minutes with subsequent incubation in 1.5% normal goat serum (Vector). Sections were then incubated overnight at 4 degrees C with rabbit anti-TH (1: 1,000; Chemicon) antibody with 10% normal goat serum. After several rinses in PBS, sections were incubated for 30 minutes in biotinylated donkey anti-rabbit IgG (1: 1,000, Jackson) then for 30 minutes in avidin-biotin-peroxidase complex (1: 200, Vector). Subsequently the sections were treated with 3, 4-diaminobenzidine (DAB, Sigma) and hydroxygen peroxide, mounted on albumin-coated slides and embedded with cover glass.TUNEL and HEt staining were also performed to investigate the involvement of anti-apoptotic effects and radical scavenging activity of edaravone using 14 rats receiving saline or 250 mg/kg of edaravone-administration at 30 minutes after 6-OHDA lesioning and sacrificed at 5 days after 6-OHDA lesioning. Furthermore, in order to reveal the effects of edaravone on the inflammation induced by 6-OHDA-administration, immunofluorescent Iba-1 staining was also performed. Rabbit anti-Iba-1 antibody (1: 100, Wako Pure Chemical Industries, Osaka, Japan) was used as the primary antibody and Alexa Fluor 594 (Molecular Probes) as the secondary antibody.Morphological analysisThe density of TH-positive fibers and Iba-1-positive microglia in the striatum of rats receiving edaravone- or saline-infusion was determined and analyzed as described previously with a computerized analysis system (Olympus Sp-1000, Japan) [16,19] using 3 serial coronal section at the bregma level. Two areas adjacent to the needle tract of lesioned side and symmetrical contralateral side were analyzed, respectively. For counting the number of TH-positive neurons, every fifth 40 μm-thick coronal tissue section through the substantia nigra pars compacta (SNc) was explored using 3 coronal sections respectively at -4.8 and -5.3 mm to the bregma. The number of TH-positive cell bodies in the SNc was counted and used for the statistical analyses.Statistical AnalysisThe data obtained were evaluated statistically using analysis of variance (ANOVA) and subsequent post hoc Scheffe's F-test or Mann-Whitney's U test. Statistical significance was preset at p < 0.05.ResultsEdaravone promotes the survival of DA neurons in vitroWe began our investigations into the neuroprotective capacity of edaravone on 6-OHDA-treated DA neurons in vitro. Exposure of 40 μM 6-OHDA resulted in a significant loss of TH-positive neurons to 30.2 ± 2.5% relative to the unexposed control (Fig. 1). Edaravone-administration (10-4 and 10-3 M) significantly reduced the loss of DA neurons induced by 6-OHDA (81.1 ± 3.5 and 73.6 ± 2.4%), compared to the 6-OHDA-treated DA neurons with 0, 10-6 and 10-5 M edaravone, although 10-6 and 10-5 M edaravone did not exert significant reduction of the cell loss (35.4 ± 1.9 and 35.8 ± 1.7%, One way ANOVA, F5, 102 = 125, p < 0.0001, Fig. 1).Figure 1Neuroprotective effects of edaravone on 6-OHDA-treated DA neurons in vitro. Left column: TH staining of DA neurons (a: non-6-OHDA-treated DA neurons) demonstrates that the number of surviving 6-OHDA-treated DA neurons significantly increased by the treatment with 10-4 and 10-3 M edaravone (e and f), compared to that with 10-6 and 10-5 M (c and d) edaravone as well as control without edaravone-administration (b). Scale bar: 30 μm. Right column: The graph demonstrates the number of surviving DA neurons by edaravone-administration. Data are shown as mean percentages of the cell number relative to the number of DA neurons without 6-OHDA-treatment +S.E. *p < 0.01 vs. 6-OHDA-treated DA neurons without edaravone-administration and those with low dose edaravone (10-6 and 10-5 M) by ANOVA.In order to determine that edaravone suppressed cell death through apoptosis, DA neurons were exposed to 40 μM 6-OHDA and then added 10-6, 10-5, 10-4 or 10-3 M edaravone with subsequent counting the number of swelling apoptotic cells exhibiting agglutinated and fragmented TUNEL-positive nuclei. Edaravone treatment (10-3 M) significantly reduced the number of TUNEL-positive apoptotic cells to 60.9 ± 1.7 and 82.1 ± 0.8% relative to that of 6-OHDA-treated DA neurons without edaravone treatment, although 10-6 and 10-5 M did not suppress apoptosis (96.2 ± 0.7 and 97.3 ± 0.8%). 10-4 M edaravone significantly suppressed apoptosis of DA neurons, compared to 6-OHDA-treated DA neurons without edaravone treatment (One way ANOVA, F4, 45 = 48, p < 0.0001, Fig. 2).Figure 2Reduced TUNEL-positive apoptotic 6-OHDA-treated DA neurons with edaravone-administration in vitro. Upper column: TUNEL-positive 6-OHDA-treated DA neurons with 10-3 M edaravone (b) significantly decreased, compared to those without edaravone-administration (a). Scale bar: 60 μm. Lower column: The graph demonstrates that TUNEL-positive 6-OHDA-treated DA neurons decreased by 10-4 and 10-3 M edaravone-administration. Data are shown as mean percentages of the cell number relative to the 6-OHDA-treated DA neurons without edaravone-administration +S.E. *p < 0.01 vs. 6-OHDA-treated DA neurons without edaravone-administration and those with low dose edaravone (10-6 and 10-5 M) by ANOVA. **p < 0.01 vs. 6-OHDA-treated DA neurons without edaravone-administration.In order to demonstrate the production of superoxide anions, HEt staining was performed. Edaravone treatment (10-3 M) significantly reduced the number of HEt-positive cells (21.5 ± 0.9, 29 ± 1.1, 31 ± 2.0, and 34 ± 1.7 cells/10,000 μm2 at 0, 10, 20, and 30 minutes after edaravone administration), compared to the untreated 6-OHDA-exposed cells (24.3 ± 1.7, 35.7 ± 1.7, 45.7 ± 3.3, and 62.5 ± 0.9 cells/10,000 μm2 at 0, 10, 20, and 30 minutes, Repeated Measures of ANOVA, F3, 18 = 21, p < 0.0001 and posthoc t-tests of p's < 0.01 for 10, 20, and 30 minutes after edaravone-administration, Fig. 3)Figure 3Reduced number of HEt-positive cells by edaravone-treatment in vitro. Left column: Edaravone treatment (10-3 M) significantly reduced the number of HEt-positive cells at 10, 20, 30 minutes after edaravone-administration, compared to the untreated 6-OHDA-exposed cells. (untreated 6-OHDA-treated DA neurons at 0 and 30 minutes (a and b); 6-OHDA-treated DA neurons with 10-3 M edaravone at 0 and 30 minutes (c and d). Scale bar: 30 μm. Right column: The graph demonstrates that HEt-positive cells significantly decreased by edaravone-administration. Data are shown as the mean cell number ± S.E. Dotted line: HEt-positive cells without edaravone, Full line: HEt-positive cells with 10-3 M edaravone. *p < 0.05 vs. 6-OHDA-treated DA neurons without edaravone-administration by ANOVA.Behavioral analyses in vivoNext, we proceeded to the in vivo study using PD model of rats. There were no significant changes in the spontaneous behavior of rats receiving edaravone-administration at 30 minutes after 6-OHDA lesion (30 mg/kg: determined by the dose for ischemic stroke) or saline (data not shown). As shown in Fig. 4, in PD model of rats receiving intravenous saline-infusion, the number of amphetamine-induced rotations increased over time at 1 and 2 weeks (9.8 ± 1.1 and 11.1 ± 0.6 turns/hour). However, rats receiving 250 mg/kg of edaravone-administration at 30 minutes after 6-OHDA lesion showed a significant reduction of the rotational number (3.8 ± 0.9 and 2.3 ± 0.7 turns/hour at 1 and 2 weeks), although 30 and 100 mg/kg of edaravone did not exert significant effects (30 mg/kg: 9.9 ± 1.8 and 10.4 ± 1.9 turns/hour, 100 mg/kg: 6.5 ± 1.5 and 7.0 ± 1.4 turns/hour at 1 and 2 weeks, Repeated Measures of ANOVA, F3, 24 = 8.8, p < 0.0001 and posthoc t-tests of p's < 0.01 for both time periods, Fig. 4).Figure 4Edaravone ameliorated the amphetamine-induced rotational behavior of PD model of rats. Left graph: Rats receiving 250 mg/kg of edaravone infusion at 30 minutes after 6-OHDA lesion showed a significant reduction of the rotational number, although 30 and 100 mg/kg of edaravone did not exert significant effects. Data are shown as the mean rotational number per minute ± S.E. *p < 0.05 vs. rats receiving 30 mg/kg of edaravone and those without edaravone-administration. Right column: Edaravone administration (250 and 100 mg/kg) at 24 hours after 6-OHDA lesion significantly suppressed the rotational behavior, compared to rats receiving saline. Data are shown as the mean rotational number per minute ± S.E. *p < 0.05 vs. rats receiving 30 mg/kg of edaravone and those without edaravone-administration.After confirmation of neuroprotective effects of intravenous administration of edaravone (250 mg/kg) at 30 minutes after 6-OHDA lesion on 6-OHDA-treated rats behaviorally, edaravone-administration at 24 hours were explored. Edaravone-administration (250 and 100 mg/kg) at 24 hours after 6-OHDA lesion significantly suppressed the rotational behavior (250 mg/kg: 5.5 ± 0.4 and 4.3 ± 0.6 turns/hour, 100 mg/kg: 6.0 ± 1.9 and 6.3 ± 2.4 turns/hour at 1 and 2 weeks), compared to rats receiving saline (10.2 ± 0.8 and 12.2 ± 0.4 turns/hour at 1 and 2 weeks, Repeated Measures of ANOVA, F3, 22 = 9.0, p = 0.0005 and posthoc t-tests of p's < 0.01 for both time periods, Fig. 4). Edaravone-administration (250 mg/kg) at 30 minutes significantly ameliorated rotational behavior, compared to that at 24 hours after 6-OHDA lesion (Repeated Measures of ANOVA, F1, 12 = 4.5, p = 0.04 and posthoc t-tests of p's = 0.04 for both time periods). Thus, edaravone significantly ameliorated the rotational behavior when it was administered earlier and in higher concentration.TH immunohistochemistry in the striatum and the SNcAt 2 weeks after edaravone-()administration, TH staining was performed to evaluate the preserved DA fibers in the striatum and DA neurons in the SNc (Fig. 5). The density of TH-positive fibers in the 6-OHDA-lesioned striatum was compared with the contralateral side using a modified method of computerized image analysis system [19]. The preservation of TH-positive fibers in the striatum of rats receiving edaravone was significantly greater (30 minutes: 23.3 ± 1.5, 41 ± 1.2 and 65.2 ± 1.6%; 24 hours: 20.2 ± 0.9, 38.2 ± 1.7 and 51.4 ± 1.1% relative to the intact side at the dose of 30, 100 and 250 mg/kg, respectively) than those receiving the saline (30 minutes: 12.1 ± 0.5, 24 hours: 8.9 ± 0.3%, Repeated Measures of ANOVA, F3, 16 = 425, p < 0.0001 and posthoc t-tests of p's < 0.01 for all groups, Fig. 6).Figure 5Preserved TH-positive fibers in the striatum and neurons in the SNc of rats receiving edaravone-administration. Photomicrographs demonstrate that 100 and 250 mg/kg of edaravone preserved TH immunoreactivity in the striatum and SNc (edaravone-administration at 30 minutes after 6-OHDA lesion, 30 mg/kg: c and i, 100 mg/kg: d and J, 250 mg/kg: e and k, edaravone administration at 24 hours after 6-OHDA lesion, 250 mg/kg: f and l), compared to the untreated 6-OHDA-lesioned rats (b and h). TH staining of the intact side of the striatum and SNc: a and h, TH staining of the striatum: a-f, and the SNc: g-l. Scale bar: 120 μm in a-f, 480 μm in g-l.Figure 6Edaravone-administration exerted neuroprotective effects in a dose-responsive manner immunohistochemically in vivo. Left graph: TH staining of the striatum demonstrates neuroprotective effects of edaravone on 6-OHDA-lesioned striatal DA fibers in a dose-responsive manner. Data are shown as the percentages of TH-positive fibers relative to the intact side ± S.E. *p < 0.05 vs. rats in all other groups, **p < 0.05 vs. rats receiving 30 mg/kg of edaravone and those without edaravone-administration, ***p < 0.05 vs. rats without edaravone-administration, ****p < 0.05 (rats receiving edaravone at 30 minutes after 6-OHDA lesion (30 min group) vs. rats with edaravone at 24 hours (24 h group)). Right graph: TH staining of the SNc demonstrates neuroprotective effects of edaravone on 6-OHDA-lesioned striatal DA fibers in a dose-responsive manner. Data are shown as the percentages of TH-positive neurons relative to the intact side ± S.E. *p < 0.05 vs. rats in all other groups, **p < 0.05 vs. rats receiving 30 mg/kg of edaravone and those without edaravone-administration.The number of TH-positive neurons in the ipsilateral SNc of rats was analyzed as percentages relative to the number of counted DA neurons in the intact side. The preservation of TH-positive neurons in the SNc of rats receiving edaravone (100 and 250 mg/kg) was significantly greater (30 minutes: 18.5 ± 1.4, 33.5 ± 1.4 and 53.9 ± 1.4%; 24 hours: 16.4 ± 0.9, 31.8 ± 1.0 and 48.3 ± 2.3% relative to the intact side at the dose of 30, 100 and 250 mg/kg, respectively) than those receiving the saline (30 minutes: 12.8 ± 1.6, 24 hours: 15 ± 0.8%, Repeated Measures of ANOVA, F3, 16 = 324, p < 0.0001 and posthoc t-tests of p's < 0.01, Fig. 6). DA fibers in the striatum of rats receiving edaravone at 30 minutes after 6-OHDA lesion was significantly preserved, compared to those with edaravone-administration at 24 hours (p's < 0.001), although DA neurons in the SNc of both time periods were not significantly different (p's = 0.11).TUNEL and HEt staining for anti-apoptotic and anti-oxidative effectsThe percentages of TUNEL positive cells in the SNc of rats receiving 250 mg/kg of edaravone at 30 minutes after 6-OHDA lesion significantly decreased (9.6 ± 1.0%), compared to those of rats without edaravone treatment (26.7 ± 5.7%, Repeated Measures of ANOVA, F1, 12 = 3.4, p = 0.034 and posthoc t-tests of p's < 0.05, Fig. 7). The percentages of TH and HEt double positive cells per TH positive cells of rats receiving 250 mg/kg of edaravone at 30 minutes after 6-OHDA lesion significantly decreased (12.9 ± 1.5%), compared to those of rats without edaravone treatment (37.7 ± 3.4%, Repeated Measures of ANOVA, F1, 12 = 32, p < 0.001 and posthoc t-tests of p's < 0.05, Fig. 7).Figure 7Anti-apoptotic and anti-oxidative effects of edaravone. Upper column: TUNEL staining revealed that edaravone administration (250 mg/kg) at 30 minutes after 6-OHDA lesion significantly decreased the percentages of TUNEL positive cells in the SNc (b), compared to rats without edaravone-administration (a). Scale bar: 60 μm The graph demonstrates the significant differences (c). Data are shown as the percentages of TUNEL positive cells +S.E. *p < 0.05 vs. the lesion side of edaravone-administered rats and the intact side. Lower column: TH and HEt double staining revealed that edaravone administration (250 mg/kg) at 30 minutes after 6-OHDA lesion significantly decreased the percentages of TH and HEt double positive cells (e), compared to rats without edaravone-administration (d; green: TH, red: HEt; Scale bar: 60 μm). The graph demonstrates the significant differences (f). Data are shown as the percentages of HEt positive cells +S.E. *p < 0.05 vs. the lesion side of edaravone-administered rats and the intact side.Iba-1 immunohistochemistry for the affected inflammationIba-1 staining was performed to evaluate anti-inflammatory effects of edaravone through the microglia. Edaravone administration (250 mg/kg) at 30 minutes after 6-OHDA lesion significantly suppressed the number of Iba-1-positive cells (178 ± 4.2 cells/10,000 μm2; non-edaravone-administered group: 252 ± 14 cells/10,000 μm2), thus indicating that edaravone suppressed inflammation induced by 6-OHDA with the decrease of activated microglia (p < 0.05, Mann-Whitney's U test, Fig. 8).Figure 8Anti-inflammatory effects of edaravone. Upper column: Iba-1 staining revealed that edaravone administration (250 mg/kg) at 30 minutes after 6-OHDA lesion significantly suppressed the number of Iba-1-positive cells (b), compared to rats without edaravone-administration (a), thus indicating anti-inflammatory effects of edaravone. Scale bar: 60 μm. Lower column: The graph demonstrates that 250 mg/kg of edaravone significantly suppressed the microglial proliferation, compared to rats without edaravone-administration. Data are shown as the cell number of Iba-1-positive cells +S.E. *p < 0.05 vs. rats without edaravone administration.DiscussionIn this study, neuroprotective effects of edaravone on 6-OHDA-treated murine ventral mesencephalic DA neurons were clarified in vitro. Anti-apoptotic effects through scavenging radicals might play an important role in the underlying mechanisms of neuroprotective effects of edaravone. In parallel, neuroprotective effects of edaravone on 6-OHDA-lesioned PD model of rats were demonstrated behaviorally and immunohistochemically. Edaravone might exert the neuroprotective effects on DA neurons. TUNEL, HEt and Iba-1 staining suggested the involvement of anti-apoptotic, anti-oxidative and anti-inflammatory effects of edaravone.Anti-apoptotic effects of edaravoneNeuroprotective effects of edaravone might be mediated by anti-apoptotic effects. Against ischemic reperfusion, edaravone prevents cell death and the release of cytochrome c with subsequent pathological apoptosis through Bcl-2 upregulation by inhibiting the opening of the mitochondrial permeability transposition pore [20,21]. In parallel, edaravone might reduce Fas-associated death domain protein and subsequently suppress apoptotic cell death in cerebral infarct [22]. Furthermore, edaravone might alleviate dysfunction of endoplasmic reticulum with subsequent cell death in cerebral ischemia [23]. Edaravone also reduces nitric oxide-induced apoptosis by inhibiting activation of MAP kinase in astroctes [24]. Related to 6-OHDA-toxicity, apoptosis is induced by down-regulation of Bcl-2 with activation of caspases in thymocytes [25], which might be suppressed by edaravone. Activated microglia damages surrounding cells by the paracrine of various cytokines. Using co-culture of neuronal cells and microglia, neuronal cell death by the peroxynitrite donor, SIN-1 (N-morpholinosydnonimine) is significantly suppressed by 10-4 M edaravone [15]. In our study, the number of TUNEL-positive apoptotic cells and HEt-positive cells decreased using both in vitro and in vivo model of PD. The number of activated microglia of rats receiving 250 mg/kg of edaravone decreased, suggesting that the reduced cytotoxic cytokines might suppress apoptosis synergistically. The underlying mechanisms of the neuroprotection of edaravone might be involved in the hypothesis above described.Characteristics of our studyRecently, the similar study was reported using 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-treated mice [26]. MPTP activated microglial activation both in the striatum and SNc with the increase of 3-nitrotyrosine, a biomarker of peroxynitrite production, in the SNc, but not in the striatum. Intraperitoneal 3 mg/kg of edaravone significantly ameliorate the behavioral scores, however the neuroprotective effects might be limited in the SNc. Using the animal model of cerebral infarct and head trauma, Dohi and colleagues also demonstrated the neuroprotective effects of low dose of edaravone [13]. In our study, the neuroprotective effects of edaravone (100 and 250 mg/kg) was demonstrated both in the striatum and SNc, although 30 mg/kg did not exert any neuroprotective effects, except for the histological amelioration in the striatum. Additionally, the behavioral amelioration by 100 mg/kg of edaravone-administered at 30 minutes and 24 hours after 6-OHDA lesioning did not show time-dependency. Furthermore, the lower dosage of edaravone (3 and 10 mg/kg) exerted no neuroprotective effects in our pilot study (data not shown). These discrepancies of the results might be due to the alteration of edaravone-activity and affinity over time after lesioning, the characteristics of the behavioral test, or the differences of the toxin (MPTP vs. 6-OHDA), of the administration route (i.p. vs. i.v.), of the animal species (mice vs. rat), and of the detailed regimen. For the safe clinical application, some amelioration of the drug, including the enhanced action for neurons specifically, because edaravone-administration even at clinical dosage might result in severe side effects [27].Until now several studies demonstrated neuroprotective effects of pre-treatment of edaravone against metamphetamine-toxicity on striatal dopaminergic degeneration [28] and post-ischemic dopaminergic dysfunctions [29]. One of the remarkable characteristics of our study also lie in the time-dependent effects of edaravone, that is, the earlier (at 30 minutes after 6-OHDA lesion) administration might exert significantly stronger neuroprotective effects than the later one (at 24 hours), mimicking the clinical settings, although the later administration still displayed the behavioral and histological amelioration. In the future, the effects of repeated administration of edaravone on PD model should be clarified.Therapy for PD in the future including edaravone-administrationThe established therapy for PD is medication using L-DOPA (dihydroxyphenylalanine), DA agonist and various drugs of different mechanisms, surgeries including electrical stimulation and ablation [30]. Fetal cell transplantation and GDNF infusion [31] are also hopeful, although the recent double-blinded randomized controlled trials questioned us the efficacy of these therapy [32,33]. In the nearest preceding years, neural transplantation might be a hopeful therapeutic option for PD [34] with recent development in the stem cell biology [35-40]. When edaravone is used for PD patients, several advantages might be recognized in combination with other therapeutic options. As edaravone extends the therapeutic time window for ischemic patients in combination with tissue plasminogen activator [41], edaravone might ameliorate the survival of transplanted cells [42] as well as scavenge free radicals in PD. Edaravone might also suppress inflammatory reaction induced by surgical procedures including electrical stimulation and cell transplantation.ConclusionNeuroprotective effects of edaravone on 6-OHDA-treated DA neurons were clarified in vitro. Anti-apoptotic effects and radical scavenging activity might be involved in the underlying mechanisms of neuroprotective effects of edaravone. Neuroprotective effects of edaravone were then demonstrated using animal model of PD. Edaravone might be a hopeful therapeutic option for PD, although several critical issues remain to be solved, including high therapeutic dosage of edaravone for the safe clinical application in the future.Authors' contributionsWJY is involved in acquisition of data and drafting the manuscript. TS, TY, TA and ID designed the study, analyzed the data and revised the manuscript. KM, MK and YM performed in vivo experiments including surgeries and animal care. TU, TM and MJ performed in vitro experiments including immunocytochemical investigations. NT and TB performed immunohistochemical investigations. FW and LH performed additional experiments in the revised manuscript. All authors read and approved the final manuscript.\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2533696.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2533696",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2533696\nAUTHORS: Desigan Kumaran, Richa Rawat, S. Ashraf Ahmed, Subramanyam Swaminathan\n\nABSTRACT:\nThe seven antigenically distinct serotypes of Clostridium botulinum neurotoxins, the causative agents of botulism, block the neurotransmitter release by specifically cleaving one of the three SNARE proteins and induce flaccid paralysis. The Centers for Disease Control and Prevention (CDC) has declared them as Category A biowarfare agents. The most potent among them, botulinum neurotoxin type A (BoNT/A), cleaves its substrate synaptosome-associated protein of 25 kDa (SNAP-25). An efficient drug for botulism can be developed only with the knowledge of interactions between the substrate and enzyme at the active site. Here, we report the crystal structures of the catalytic domain of BoNT/A with its uncleavable SNAP-25 peptide 197QRATKM202 and its variant 197RRATKM202 to 1.5 Å and 1.6 Å, respectively. This is the first time the structure of an uncleavable substrate bound to an active botulinum neurotoxin is reported and it has helped in unequivocally defining S1 to S5′ sites. These substrate peptides make interactions with the enzyme predominantly by the residues from 160, 200, 250 and 370 loops. Most notably, the amino nitrogen and carbonyl oxygen of P1 residue (Gln197) chelate the zinc ion and replace the nucleophilic water. The P1′-Arg198, occupies the S1′ site formed by Arg363, Thr220, Asp370, Thr215, Ile161, Phe163 and Phe194. The S2′ subsite is formed by Arg363, Asn368 and Asp370, while S3′ subsite is formed by Tyr251, Leu256, Val258, Tyr366, Phe369 and Asn388. P4′-Lys201 makes hydrogen bond with Gln162. P5′-Met202 binds in the hydrophobic pocket formed by the residues from the 250 and 200 loop. Knowledge of interactions between the enzyme and substrate peptide from these complex structures should form the basis for design of potent inhibitors for this neurotoxin.\n\nBODY:\nIntroductionClostridium botulinum neurotoxins (CNTs) are the most potent toxins known to humans since even one billionth of an ounce is fatal. Seven antigenically distinct botulinum neurotoxins are produced by the bacterium Clostridium botulinum and they share considerable sequence homology, and structural and functional similarity [1]–[3]. They are produced as inactive single chains of molecular mass 150 kDa and released as active dichains, a heavy chain (HC, 100 kDa) and a light chain (LC, 50 kDa) held together by an interchain disulfide bond [4]–[7]. HC comprising two distinct domains is responsible for binding to neuronal cells and translocation into cytosol. LC is the catalytic domain cleaving one of the three proteins forming the SNARE complex (Soluble N-ethylmaleimide-sensitive fusion protein attachment protein receptors) required for docking and fusion of vesicles containing neurotransmitters to target cells [8]–[12]. The SNARE complex formation is prevented when any of the SNARE proteins is cleaved and accordingly blocks neurotransmitter release leading to flaccid paralysis and eventual death.Catalytic domains of BoNTs are zinc proteases and cleave SNARE proteins with stringent substrate specificity though they share significant sequence similarity. BoNT/A and BoNT/E cleave the synaptosomal-associated 25 kDa protein (SNAP-25) while BoNT/B, /D, /F, and /G cleave the vesicle-associated membrane protein (VAMP). BoNT/C is the only one that has dual substrate specificity, viz SNAP-25 and syntaxin [13]. The enhanced substrate specificity of CNTs is due to the recognition of substrates at remote sites called exosites in addition to the active site [14].The potency and the ease with which these toxins can be produced make them potential bioweapons and bioterrorism agents. The Centers for Disease Control and Prevention (CDC) has declared them as Category A biowarfare agents. Currently, while experimental vaccines are available, only an equine trivalent antitoxin is available for post-exposure therapeutics with a limited therapeutic window [15]. One of the most effective ways a drug can act is by blocking the site where the substrate binds to toxin and accordingly the crystal structure of substrate-enzyme complex is essential to map out a strategy. Even though crystal structure of SNAP peptide (146–206)-inactive enzyme complex is available, it lacks interactions at the active site since the enzyme used was an inactive double mutant [14]. Here we present for the first time the structure of the substrate peptide, QRATKM containing the scissile peptide bond, bound to the active enzyme. This crystal structure reveals interesting features of complex formation which can help in designing efficient drug molecules to prevent or treat botulism. It is remarkable that this natural substrate peptide is not cleaved by the enzyme. In addition, we are also reporting the crystal structure of RRATKM, a variant of the substrate peptide, in complex with the enzyme. Though both are weak inhibitors, RRATKM is a better inhibitor than QRATKM.Materials and MethodsProtein expression and purification\nClostridium botulinum neurotoxin serotype A truncated light chain (residues 1 to 424), Balc424, was expressed in E. coli and purified to homogeneity using size exclusion chromatography, as described previously [16]. The purified enzyme in 20 mM HEPES, 2 mM DTT, 200 mM NaCl, pH 7.4 was stored at −20°C until used. Amides of the peptides, QRATKM and RRATKM, were custom synthesized by Peptide 2.0 Inc., Chantilly, VA20153, USA. The stock solutions of the peptides were prepared with the above mentioned buffer.Crystallization and data collectionBalc424-QRATKM and Balc424-RRATKM complex crystals were grown using a range of protein/peptide molar ratio (1∶5 to 1∶30). Both QRATKM and RRATKM complex crystals were grown by sitting drop vapor diffusion at room temperature. Briefly, 3 µl of the protein solution (15 mg/ml) was mixed with an equal volume of a reservoir solution containing 20% PEG 8000, 100 mM sodium cacodylate, pH 6.5, 5% ethylene glycol and 200 mM ammonium sulfate. Thick plate-like crystals were obtained in five days and were flash frozen with liquid nitrogen using 20% ethylene glycol as cryoprotectant. The X-ray intensity data for both complex crystals were collected at X29 beamline of National Synchrotron Light Source (NSLS) using ADSC QUANTUM 315 detector. Balc424-QRATKM and Balc424-RRATKM complex crystals diffracted to 1.5 Å and 1.6 Å, respectively and belonged to the P21 space group with one molecule in the asymmetric unit (Table 1). All data were processed using the HKL2000 suite [17].10.1371/journal.ppat.1000165.t001Table 1Crystal data and refinement statistics of Balc424 with substrate peptide complexesName/codeQRATKMRRATKMCell dimensionsa (Å)49.1450.87b66.2066.58c64.8265.06β (°)99.1098.3Space groupP21\nP21\nResolution range (Å)Overall50–1.550–1.6Last shell1.54–1.501.65–1.6# unique reflections64,56253,272Completeness (%)(Overall/Last shell)97.4/80.095.4/76.3Rmerge\n1 overall/last shell0.066/0.250.059/0.15<I/σ(I)> overall/last shell17/2.023.0/3.0\nRefinement Statistics\nResolution (Å)50–1.550–1.6R factor2/Rfree (%)18.0/20.020.0/22.0R.M.S deviation from idealityBond lengths (Å)0.0050.005Bond angles (°)1.21.2Average B-factors (Å2)Main chain14.021.1Side chain16.223.0Waters22.629.7Ions27.035.3Substrate peptide31.028.4Number of atomsProteins3,4233,423Waters375375Ions (Zn2+/SO4\n2−)1/51/10Ligands5052Surface area Å2(total/buried)993/7261023/739(by substrate peptide)Residues (%) in the coreregion of φ-ψ plot91.289.01Rmerge = ∑j(|Ih−<I>h|)/∑Ih, where <Ih> is the average intensity over symmetry equivalents2R-factor = ∑|Fobs−Fcalc|/∑|Fobs|Structure determinationThe structures of the complexes were determined by Fourier Synthesis using the acetate bound Balc424 (Protein Data Bank id 3BWI) as model followed by rigid-body refinement and simulated annealing. The composite omit map and the difference Fourier showed interpretable electron density for these hexapeptides. The best results were obtained with data collected from crystals grown with 1∶25 (protein/peptide) molar ratio. The peptide models were built with O [18] and further refined with CNS [19] until convergence. The final refinement statistics are shown in Table 1. Models were validated with the Ramachandran plot using PROCHECK [20].Activity assayThe proteolytic activity of balc424 was determined by HPLC using P[187–203] synthetic peptide as reported previously [21]; [22]. Briefly, balc424 enzyme (550 nM) was incubated with the 17-mer peptide (1mM) at 37°C for 30 min in the assay buffer (50 mM HEPES, 0.25 mM ZnCl2, 5.0 mM DTT, pH 7.2). IC50 values were determined by varying the concentration of inhibitors. The experimental data were analyzed using equation 1, where I is the inhibitor concentration, y is the percent inhibition, with a slope factor (s) of 1.0.(1)\nCoordinates and structure factors have been deposited to the Protein Data Bank. BALC424-QRATKM (3DDA) and BALC424-RRATKM (3DDB). The SwissProt accession number for BoNT/A is P10845.ResultsCrystal structure of Balc424 with QRATKMThe crystal structure has been determined to 1.5Å resolution. The model refined with R and R free of 18.4 and 20.1%, respectively. The final refined model contains 423 protease residues, 6 substrate residues, one sulfate and one zinc ions and 375 waters. More than 91% of residues are within the most allowed region of the Ramachandran plot. The electron density in the residual map (Fo-Fc) was well defined for the hexapeptide and QRATKM could be modeled unambiguously except for the side chains of K and M (Figure 1A). It appears that K could take two rotamer positions. This is the first time an uncleavable substrate bound structure of an active botulinum neurotoxin has been reported and it has helped in unequivocally defining S1 to S5′ sites. Most notably, the amino nitrogen and carbonyl oxygen of P1 residue (Gln197) chelate the zinc ion (Figures 2 and 3). The amino nitrogen has replaced the nucleophilic water as was shown earlier [16].10.1371/journal.ppat.1000165.g001Figure 1Hexapeptides and inhibition plots.Electron density maps (blue mesh) for bound substrate (QRATKM) and its variant (RRATKM) are shown in A and B, respectively. The electron density is from composite omit maps (2|Fo|-|Fc|) and contoured at 1σ level. QRATKM (green) and RRATKM (gray) peptides are shown in ball and stick model. Zinc, oxygen and nitrogen atoms are shown as magenta, red and blue spheres, respectively. Carbon atoms are shown in green (QRATKM) and black (RRATKM). Figures were prepared using Molscript, Raster3D and Bobscript [34]–[36]. Distribution of B factors for the QRATKM and RRATKM are shown in C and D, respectively. The peptide atoms are colored according to B factor with RGB (Red-Green-Blue) color ramp with blue and red corresponding to the lowest (17 Å2) and highest (50 Å2). Pymol (DeLano, W.L. The PyMOL Molecular Graphics System (2002) on World Wide Web http://www.pymol.org) was used to prepare C and D figures. Inhibition of Balc424 catalytic activity at increasing concentrations (µM) of QRATKML (E) and of RRATKML (F).10.1371/journal.ppat.1000165.g002Figure 2Binding of the substrate peptide in the active site of Balc424.(A) Figure shows the substrate peptide (QRATKM) binding in the active site of Balc424. Balc424 is shown in solid blue-colored surface. Substrate peptide is shown in sphere (CPK) model. Carbon, oxygen, nitrogen, sulfur and zinc atoms are shown in green, red, blue, yellow and magenta, respectively. (B). Superposed stick models of QRATKM and RRATKM are shown in green, and gray, respectively. Carbon atoms are shown in green (QRATKM) and gray (RRATKM). Balc424 is shown in semi-transparent surface (blue) representation with secondary elements at the active site pocket. Only a few substrate-binding residues are shown as markers. Pymol (DeLano, W.L. The PyMOL Molecular Graphics System (2002) on World Wide Web http://www.pymol.org) was used to prepare these figures.10.1371/journal.ppat.1000165.g003Figure 3Structure of Balc424 in complex with the QRATKM, a segment of substrate SNAP-25.(A). Stereo view of the active site shows the molecular interactions between the substrate peptide (QRATKM) and protease (Balc424). Protease and substrate residues are shown in cyan and green, respectively. Blue color ribbon represents the protease secondary elements at the vicinity of the active site. (B). Stereo view of the active site center of Balc424 (cyan stick) with RRATKM (gray ball and stick). Oxygen, nitrogen, sulfur and zinc atoms are shown in red, blue, yellow and magenta, respectively. Hydrogen bonds are depicted as black dash lines while zinc co-ordination is shown in solid line.Crystal structure of Balc424 with RRATKMThe crystal structure of Balc424 with a substrate analog RRATKM has been determined to 1.6Å resolution. The R and R free for the final refined model are 20.1 and 21.2%, respectively. The final refined model contains 423 residues of protease, 6 residues of substrate analog peptide, two sulfate ions, one zinc ion and 375 waters. More than 90% of residues are within the most allowed region of the Ramachandran plot. The substrate analog could be modeled unambiguously in the residual map (Fo-Fc) (Figure 1B). Except for some minor variations of side chain orientations, the hexapetide RRATKM binds similar to the substrate peptide QRATKM (Figures 2, 3 and 4). As in the case of QRATKM, the P1 (Arg197) amino group and the carbonyl oxygen chelate the catalytic zinc and the nucleophilic water has been replaced. P1-P5′ residues occupy identical subsites as in QRATKM. This kind of interaction seems to be common with all peptide analog inhibitors [16] and probably plays a dominant role in inhibiting the catalytic activity.10.1371/journal.ppat.1000165.g004Figure 4Schematic diagram of the molecular interactions between the Balc424 and substrate peptides.The interactions between Balc424 active site residues and the substrate peptides are shown. (A) QRATKM and (B) RRATKM. Black, red, blue, and yellow colored circles represent carbon, oxygen, nitrogen, and sulfur atoms, respectively. For clarity, zinc co-ordination and water molecules involved in the interactions at the active site are not shown. (C). A Schematic diagram representing S1 to S5′sites. Residues of the enzyme forming the subsites and substrate peptide are shown in red and blue, respectively. Proteolytic site is shown in cyan colored double-headed arrow. Figures A and B were prepared with Ligplot [37]. ChemDraw ultra (CambridgeSoft, Inc) was used to prepare figure C.Though we have shown earlier that short tetrapeptides (analogs of substrate) are good inhibitors (nM range), the hexapeptides are weak inhibitors [16]. The IC50 of QRATKM and RRATKM are 133 and 95 µM, respectively (Figures 1E and 1F).DiscussionMapping of S1–S5′ subsitesThe side chain of P1-Q197 is exposed to the solvent region but makes a hydrogen bond with Glu164 OE1 (Figures 3 and 4). However, it is stabilized by various other interactions as well. N and O chelate zinc while O is also hydrogen bonded to Tyr366 OH which stabilizes the substrate and positions it for catalytic activity. Mutation of Tyr366 to Phe or Ala resulted in dramatic decrease in activity [23]; [24]. The amino nitrogen which has replaced the nucleophilic water is hydrogen bonded to Glu224 OE1 and OE2 (the latter through a water molecule). It is known that variation in P1 does not affect the catalytic activity, probably due to most of the interactions being with the main chain atoms [25]–[28]. Mutation of Glu164 to Gln only had a marginal effect on the catalytic activity [23]. The only difference between QRATKM and RRATKM is at P1 residue. This was based on our previous experience with tetrapeptides [16] since the positive charge on Arg197 better complements the charge in the active site cavity. While P1-Gln197 makes a hydrogen bond with Glu164, P1-Arg197 makes a salt bridge interaction with Glu164 thus making it more strongly bound (Figures 3 and 4). There are additional interactions with a sulfate ion nearby but this may be an artifact of crystallization. Other than this, residues from 198 to 202 in both structures superpose well except for minor variations in side chain orientations (Figure 2B). The following discussion on subsites S1′ to S5′ applies equally for the both structures.P1′-Arg198 occupies the S1′ site formed by Arg363, Thr220, Asp370, Thr215, Ile161 and Phe194. Phe163, though slightly farther, also forms part of this subsite. The amino nitrogen and carbonyl oxygen of P1′ are hydrogen bonded to Phe163 O and Arg363 NH2 (Figures 3 and 4). These two interactions stabilize the substrate binding. When Arg363 is mutated to Leu or Ala, the activity decreases by 620 and ∼80 fold, respectively [23]; [24]. In addition, the guanidinium group of P1′ Arg198 forms salt bridges with Asp370 and P1′-Arg198 NE forms a hydrogen bond with Ile161 O. The salt bridge interaction between P1′-Arg198 and Asp370 is crucial since mutation of Asp370 reduced the catalytic activity by 250–600 fold [23]; [29]. The other major interaction is the stacking of guanidinium group of P1′-Arg198 with Phe194 (Figure 3). This stacking interaction also plays a major role in the activity since Balc424 Phe194Ala has ∼100 fold less activity [29]. Accordingly, both the electrostatic and hydrophobic interactions are crucial for catalytic activity. The S1′ site is fairly big and gives enough flexibility for Arg198. In substrate analog tetrapeptide inhibitor complexes, it takes various rotamer positions [16]. In BoNT/A arginine hydroxamate complex structure, Arg hydroxamate occupies the S1′ site. But Zn is chelated by the carbonyl oxygen and the hydroxamate group. Also the direction of the peptide N to C is reversed [30].S2′ site is formed by Arg363, Asn368 and Asp370, while S3′ subsite is formed by Tyr251, Leu256, Val258, Tyr366, Phe369 and Asn388. P3′-Thr200 OG makes a hydrogen bond with Tyr251 OH. P4′-Lys201 is exposed to the solvent region. In the present crystal structure the side chain density for this residue is weak probably due to high thermal factors (Figures 1C and 1D). However, one of the rotamer positions could form a hydrogen bond with Gln162 OE1. This does not form a hydrogen bond in the complex structure of BoNT/A-SNAP-25 (146–206) (PDB id = 1XTG). Instead Glu257 is close by, about 4.5Å. S5′ site is made of Tyr251, Phe369, Leu256, Ser254 and Phe 423. P5′-Met202 occupies this hydrophobic pocket (Figure 4C).Comparison of SNAP-25(146–206) and QRATKM at the active siteThe crystal structure of SNAP-25 (146–206) peptide with an inactive double mutant (Pdb id = 1XTG) had identified the exosites as recognition sites distant from the active site [14]. However, the region of SNAP-25 peptide near the active site was disordered and could not be modeled very well. Comparison of the C-alpha position of the corresponding residues in the present structure shows that the C-alpha positions of these six residues are shifted. C-alphas of 197, 198, 199, 200, 201 and 202 are 4.34, 3.84, 3.55, 3.13, 5.69 and 6.12Å for the corresponding C-alphas in the present structures (Figure 5A). In the absence of Tyr366, SNAP25 residues near the active site move towards 250 loop increasing the distance from catalytic zinc. When the wild type light chain is used, the SNAP peptide is closer to the catalytic zinc and the 170 loop. This shift is probably due to either the disorder or the inactive mutant in 1XTG. One possibility is that since residues corresponding to α-exosites are missing in the short peptide, the whole peptide could have slid down. But this possibility is less likely since the β-exosite interaction is maintained in both the structures. Though the C-alpha atom of P5′ in the current structure and 1XTG are farther apart, the side chains occupy the same place. We conclude that this shift is due to the loss of interaction of SNAP-25 with Tyr366 which has been mutated to Phe in 1XTG. Because of this difference, P4′-Lys201 has potential interaction with Gln162 of the enzyme rather than Glu257. The length of the anti-parallel β sheet formed near the 250 loop (β-exosite) in 1XTG (13 Å) is almost double the length as in QRATKM (6.5 Å) (Figure 5A). Based on the above observations, the subsites as identified in this structure truly represent the substrate-enzyme complex interactions.10.1371/journal.ppat.1000165.g005Figure 5Comparison of Balc424 complexed with segments of SNAP-25 (197–202) and (146–204) and an inhibitor (N-Ac-CRATKML).(A) and (B) show the superposition of Balc424 with SNAP-25 (197–202) and (146–204). Green and red represent SNAP-25 (197–202) and SNAP-25 (146–204), respectively. Superposition of Balc424 with substrate peptide (SNAP-25 [197–202]) and N-Ac-CRATKML peptide at the vicinity of the active site are shown in C and D. N-Ac-CRATKML peptide complex is shown in brown color. Zinc binding and selective substrate binding residues are in stick model.Though the overall conformation of the enzyme in 1XTG and the current structure is very similar (RMSD is ∼1 Å for 400 Cα atoms), loops 200 and 250 vary significantly (Figure 5B). This conformational change may be either due to the recognition of α-exosites in 1XTG or just an artifact of crystal packing. In the current structure, loops 200, 250 and 370 pack together tightly whereas in the 1XTG, 200 loop moved away. The C-alphas of Pro206 (within 200 loop) in 1XTG and QRATKM complex are ∼12 Å apart.Comparison of QRATKM with N-Ac-CRATKMLRecently, the structure of a complex between the BoNT/A-LC and an inhibitory peptide N-Ac-CRATKML has been reported [31]. Though the direction of the polypeptide is the same, the inhibitory peptide (N-Ac-CRATKML) is shifted down by one residue compared to the substrate peptide QRATKM (Figure 5C). This appears to be due to the effect of oxidation of Cys and the N-terminal blocking acetyl group. The cysteine is oxidized to sulfenic form. Both the sulfur and the OH group chelate the zinc ion unlike in QRATKM complex where the carbonyl oxygen and amino nitrogen of P1 residue chelate zinc (Figure 5D). As a consequence, the acetyl group takes the C-alpha position of P1′ (Arg198) and P1′ arginine moves to P2′ alanine's place. Moreover, P1 carbonyl oxygen interacts with Arg363 instead of Tyr366. In QRATKM, P1' arginine forms salt bridge with Asp370 through guanidinium:carboxylate pair whereas in the N-Ac-CRATKML it is through a single NE and OD1 interaction. Interestingly, even though the C-alpha position has moved, Arg198 side chain takes a different rotamer position made possible by the size of the cavity and stays in the same pocket. In addition, P4' lysine interacts with Tyr366 while in the substrate peptide (QRATKM) it interacts with Glu162. Hence the positioning of the inhibitory peptide (N-Ac-CRATKML) may not represent the substrate binding position as in QRATKM structure. In both cases the enzyme does not undergo significant conformational changes as it did in the structure of SNAP-25 (146–206) peptide complex [14].Roles of substrate amino acid residues spanning the cleavage siteN-Ac-CRATKML is a fairly good inhibitor (Kι 1.9 µM) [28]. But when the N terminal Cys is replaced with 2-mercapto-3-phenylpropionyl (mpp) the Ki improved to 300nM. Keeping this as a control various truncations were done [27]. Truncating the last three residues of the mpp derivative (KML) increased the Ki 100-fold while deletion of only the last two increased it only by ∼13-fold. The importance of Lys201 of the substrate may be attributed to the potential hydrogen bond the terminal side chain atom (NZ) makes with Gln162. Mutation of Lys201 to Ala increased the Ki 10 fold suggesting that the Lys side chain interaction is crucial. When Thr200 of the substrate was mutated to Ala, Ki increased only marginally since the hydrogen bond with OG was lost. However, it is not clear from the present structure why Ala199Val will increase the Ki <10 fold. A simple modeling shows that the S2' subsite is big enough to accommodate a Val. Mutation of Arg198 to Lys increases Ki by more than 1000 fold. This is because both the salt bridge and stacking interactions are lost. It appears stacking may be important since ionic interaction between Lys201 and Asp370 is still possible. Though the present hexapeptide lacks Leu203, truncation of this peptide had no effect on Ki.Recognition and binding of substrate by Balc424Saturation mutation studies based on the crystal structure of BoNT/A with SNAP-25 (146–206) has been used to define two regions, active site (AS) domain and binding site (B) domain in SNAP-25 [14]; [29]. SNAP-25 residues 193–202 form AS while residues 156–181 form B. Our hexapeptides form part of AS only. In the same work, two minimal length peptides have been tested for catalytic activity, D193EANQRATK201 (SN/A1) and A195NQRATK201 (SN/A2) (the numbers correspond to our numbering scheme). While SN/A1 was cleavable by BoNT/A, SN/A2 was not, suggesting that the N terminal DEAN is required for cleavage. This probably explains why QRATKM which lacks DEAN was not cleaved in our case even though we used up to 1∶30 ratio of Balc424 to peptide. However, the major reason for the peptide not being cleaved is the amino group chelating zinc. Any extension beyond in the N terminal direction would change the character of this amino group and may not be able to chelate zinc. However, the earlier study used GST fusion protein to express the short peptide and might have some effect in binding to the enzyme. This is supported by the facts that I192DEANQRATKKMLGSG207 had 1/5th the activity compared to wild type [22] and the mutants A195C and N196C in the 17-mer SNAP-25 substrate peptide [28] insignificantly affected Km and kcat.The current structure confirms our earlier model for catalytic mechanism [16]. Glu224 acts as the general base in abstracting a proton from the nucleophilic water and also helps in shuttling protons to the leaving group. In addition, the roles of Arg363 and Tyr366 are to stabilize the substrate for proper positioning and orientation as the carbonyl oxygens of P1 and P1' are hydrogen bonded to Tyr366 and Arg363. Tyr366 further stabilizes the oxyanion role of P1 carbonyl oxygen. Another molecular mechanism for BoNT/A recognition and cleavage of SNAP-25 has been proposed [29]. In that mechanism P5 (Asp193) residue of SNAP25 is supposed to make the initial contact with the enzyme at the α-exosites by forming a salt bridge with Arg177. This in turn aligns P4'-Lys201 to form a salt bridge with Glu257. These interactions are supposed to broaden the active site and allow P1'-Arg198 to dock into the S1' site by both electrostatic and hydrophobic interactions. The current structure does not support such a mechanism. First, the substrate peptide is able to dock into S1' site even though the peptide lacks substrate residues upstream of P1. Second, the S1' site of Balc424 with and without bound peptide is similar and there is no indication of any change in shape or size. Third, there is no possibility for Lys201 to make hydrogen bond contact with Glu257. Accordingly, our crystallographic data show that Balc424 is well positioned for peptide binding and catalytic action without having to undergo a conformational change. However, the interaction of P4' with S4' substrate may be disrupted after cleavage and help the substrate to leave allowing uncleaved peptide to bind in its place. But there is no experimental or mutational evidence for that.Implication for drug designEven though botulinum neurotoxins are declared category A biowarfare agents, effective drugs are yet to be developed. Antibody therapeutics is emerging but more than one antibody may be needed to contain the effect of a single serotype [32]. An equine antitoxin is also available for post exposure therapeutics. Small molecule inhibitors are being developed but the active site of botulinum neurotoxin is large and it would be better to have larger molecules or strongly binding peptidomimetic inhibitors to block the active site. The current structure where S1 to S5' sites have been mapped unequivocally will be a good starting point. This would at least give a serotype specific inhibitor that could be transformed into an effective drug for botulinum neurotoxin A. We have shown that the P1 residue could be changed to Arg without affecting the binding efficiency and in fact it has proved to be a better inhibitor since it complements the charge in that region. It is known that changing it to cysteine improves binding [27]. However, oxidation of Cys may cause a problem. The structural environment of P1 also suggests that an amino acid containing an aromatic ring may be better suited as it would improve stacking interactions. The hexapeptide could be extended by one residue at the N terminus. However, it might affect the chelation of zinc by P1 amino group. The requirement of P1' Arg is crucial for BoNT/A activity. However, changing it to Tyr will still keep the stacking interaction though the salt bridge would be lost. Arg198Ala abolishes the activity without affecting the Km value [33]. S2' site also suggests that it can tolerate bigger hydrophobic, aromatic residue. It is possible to introduce modifications in the peptides to bring rigidity, specificity and resistance from proteases. There are endless possibilities that can be tried with the information provided by this structure. Our biochemical assays with full length and truncated balc (balc424) do not show much variation and hence the results are equally applicable to both. It is desirable to have a broad spectrum inhibitor to be effective across the serotypes and this structure will be a starting point.\n\nREFERENCES:\n1. LacyDBTeppWCohenACDasGuptaBRStevensRC\n1998\nCrystal structure of botulinum neurotoxin type A and implications for toxicity.\nNat Struct Biol\n5(10)\n898\n902\n9783750\n2. SimpsonLL\n1986\nMolecular pharmacology of botulinum toxin and tetanus toxin.\nAnnu Rev Pharmacol Toxicol\n26\n427\n453\n3521461\n3. SwaminathanSEswaramoorthyS\n2000\nStructural analysis of the catalytic and binding sites of Clostridium botulinum neurotoxin B.\nNat Struct Biol\n7(8)\n693\n699\n10932256\n4. DasGuptaBRRasmussenS\n1983\nAmino acid composition of Clostridium botulinum type F neurotoxin.\nToxicon\n21(4)\n566\n569\n6353671\n5. DasGuptaBRRasmussenS\n1983\nPurification and amino acid composition of type E botulinum neurotoxin.\nToxicon\n21(4)\n535\n545\n6353669\n6. SathyamoorthyVDasGuptaBR\n1985\nSeparation, purification, partial characterization and comparison of the heavy and light chains of botulinum neurotoxin types A, B, and E.\nJ Biol Chem\n260(19)\n10461\n10466\n4030755\n7. SathyamoorthyVDasGuptaBR\n1985\nPartial amino acid sequences of the heavy and light chains of botulinum neurotoxin type E.\nBiochem Biophys Res Commun\n127(3)\n768\n772\n3985955\n8. MenestrinaGSchiavoGMontecuccoC\n1994\nMolecular mechanisms of action of bacterial protein toxins.\nMol Aspects Med\n15(2)\n79\n193\n7984032\n9. MontecuccoCSchiavoG\n1994\nMechanism of action of tetanus and botulinum neurotoxins.\nMol Microbiol\n13(1)\n1\n8\n7527117\n10. MontecuccoCSchiavoGTugnoliVde GrandisD\n1996\nBotulinum neurotoxins: mechanism of action and therapeutic applications.\nMol Med Today\n2(10)\n418\n424\n8897436\n11. OgumaKFujinagaYInoueK\n1995\nStructure and function of Clostridium botulinum toxins.\nMicrobiol Immunol\n39(3)\n161\n168\n7603360\n12. SimpsonLL\n2000\nIdentification of the characteristics that underlie botulinum toxin potency: Implications for designing novel drugs.\nBiochemie\n82\n943\n953\n13. SchiavoGMatteoliMMontecuccoC\n2000\nNeurotoxins affecting neuroexocytosis.\nPhysiol Rev\n80(2)\n717\n766\n10747206\n14. BreidenbachMABrungerAT\n2004\nSubstrate recognition strategy for botulinum neurotoxin serotype A.\nNature\n432(7019)\n925\n929\n15592454\n15. SobelJ\n2005\nBotulism.\nClin Infect Dis\n41(8)\n1167\n1173\n16163636\n16. KumaranDRawatRLudivicoMLAhmedSASwaminathanS\n2008\nStructure and substrate based inhibitor design for clostridium botulinum neurotoxin serotype A.\nJ Biol Chem\n283\n18883\n18891\n18434312\n17. OtwinowskiZMinorW\n1997\nProcessing of X-ray diffraction data collected in oscillation mode.\nMethods Enzymol\n276\n307\n326\n18. JonesTAZouJYCowanSWKjeldgaardM\n1991\nImproved methods for building protein models in electron density maps and the location of errors in these models.\nActa Crystallogr A\n47 (Pt 2)\n110\n119\n2025413\n19. BrungerATAdamsPDCloreGMDeLanoWLGrosP\n1998\nCrystallography & NMR system: A new software suite for macromolecular structure determination.\nActa Crystallogr D Biol Crystallogr\n54(Pt 5)\n905\n921\n9757107\n20. LaskowskiRA\n2001\nPDBsum: summaries and analyses of PDB structures.\nNucleic Acids Res\n29(1)\n221\n222\n11125097\n21. RawatRAshraf AhmedSSwaminathanS\n2008\nHigh level expression of the light chain of botulinum neurotoxin serotype C1 and an efficient HPLC assay to monitor its proteolytic activity.\nProtein Expr Purif\n60\n165\n169\n18482846\n22. SchmidtJJBostianKA\n1995\nProteolysis of synthetic peptides by type A botulinum neurotoxin.\nJ Protein Chem\n14(8)\n703\n708\n8747431\n23. AhmedSAOlsonMALudivicoMLGilsdorfJSmithLA\n2008\nIdentification of Residues Surrounding the Active Site of Type A Botulinum Neurotoxin Important for Substrate Recognition and Catalytic Activity.\nProtein J\n24. BinzTBadeSRummelAKolleweAAlvesJ\n2002\nArg(362) and Tyr(365) of the botulinum neurotoxin type a light chain are involved in transition state stabilization.\nBiochemistry\n41(6)\n1717\n1723\n11827515\n25. BurnettJCRuthelGStegmannCMPanchalRGNguyenTL\n2007\nInhibition of metalloprotease botulinum serotype A from a pseudo-peptide binding mode to a small molecule that is active in primary neurons.\nJ Biol Chem\n282(7)\n5004\n5014\n17092934\n26. SchmidtJJBostianKA\n1997\nEndoproteinase activity of type A botulinum neurotoxin: substrate requirements and activation by serum albumin.\nJ Protein Chem\n16(1)\n19\n26\n9055204\n27. SchmidtJJStaffordRG\n2002\nA high-affinity competitive inhibitor of type A botulinum neurotoxin protease activity.\nFEBS Lett\n532(3)\n423\n426\n12482605\n28. SchmidtJJStaffordRGBostianKA\n1998\nType A botulinum neurotoxin proteolytic activity: development of competitive inhibitors and implications for substrate specificity at the S1' binding subsite.\nFEBS Lett\n435(1)\n61\n64\n9755859\n29. ChenSKimJJBarbieriJT\n2007\nMechanism of substrate recognition by botulinum neurotoxin serotype A.\nJ Biol Chem\n282(13)\n9621\n9627\n17244603\n30. SilvaggiNRBoldtGEHixonMSKennedyJPTziporiS\n2007\nStructures of Clostridium botulinum Neurotoxin Serotype A Light Chain complexed with small-molecule inhibitors highlight active-site flexibility.\nChem Biol\n14(5)\n533\n542\n17524984\n31. SilvaggiNRWilsonDTziporiSAllenKN\n2008\nCatalytic Features of the Botulinum Neurotoxin A Light Chain Revealed by High Resolution Structure of an Inhibitory Peptide Complex.\nBiochemistry\n47\n5736\n5745\n18457419\n32. MarksJD\n2004\nDeciphering antibody properties that lead to potent botulinum neurotoxin neutralization.\nMov Disord\n19\nSuppl 8\nS101\n108\n15027061\n33. ChenSBarbieriJT\n2006\nUnique substrate recognition by botulinum neurotoxins serotypes A and E.\nJ Biol Chem\n281(16)\n10906\n10911\n16478727\n34. EsnoufRM\n1997\nAn extensively modified version of MolScript that includes greatly enhanced coloring capabilities.\nJ Mol Graph Model\n15(2)\n132\n134\n112��133\n9385560\n35. EsnoufRM\n1999\nFurther additions to MolScript version 1.4, including reading and contouring of electron-density maps.\nActa Crystallogr D Biol Crystallogr\n55(Pt 4)\n938\n940\n10089341\n36. MerrittEAMurphyME\n1994\nRaster3D Version 2.0. A program for photorealistic molecular graphics.\nActa Crystallogr D Biol Crystallogr\n50(Pt 6)\n869\n873\n15299354\n37. WallaceACLaskowskiRAThorntonJM\n1995\nLIGPLOT: A program to generate schematic diagrams of protein-ligand interactions.\nProt Eng\n8\n127\n134"
+}

batch_3/PMC2535585.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2535585",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2535585\nAUTHORS: Olawale O Ogunsemi, Olatunde Odusan, Michael O Olatawura\n\nABSTRACT:\nBackgroundThe aim of this study is to evaluate the effect of a psychiatric label attached to an apparently normal person on the attitude of final year medical students at a Nigerian university.MethodsA questionnaire with sections on demographic information, a single-paragraph case description illustrating a normal person, a social distance scale and questions on expected burden was used to elicit responses from 144 final year medical students who have had previous exposure to psychiatric posting. The students consisted of two randomly assigned groups; group A received a case description with a psychiatric label attached while group B received the same case description but without a psychiatric label.ResultsA total of 68 (47.2%) of the students responded to the questionnaire with the attached psychiatric label, while 76 (52.8%) responded to the questionnaire without the attached label. There was no statistical difference in age (p = 0.187) and sex (p = 0.933) between the two groups of students. The students who responded to the questionnaire with the attached psychiatric label would not rent out their houses (p = 0.003), were unwilling to have as their next-door neighbour (p = 0.004), or allow their sister to get married (p = 0.000) to the man depicted in the case description compared with those that responded to the questionnaire without label. This group also felt that the man would exhaust them both physically (p = 0.005) and emotionally (p = 0.021) in any relationship with him.ConclusionThese results strengthen the view that stigma attached to mental illness is not limited to the general public; medical students are also part of the stigmatising world. There is, therefore, a need to incorporate issues concerning stigma and its reduction as a core component of the mental health curriculum of medical schools.\n\nBODY:\nIntroductionIn most societies mental illness carries a substantial stigma [1,2]. The mentally ill are often blamed for bringing on their own illnesses, while others may see them as victims of unfortunate fate, religious and moral transgression, or even witchcraft. Such stigma may lead to denial on the part of the family that one of their members is psychiatrically ill. Some families may hide or overprotect a member with mental illness, thus keeping the person from receiving potentially effective care.Stigma remains a powerful negative attribute in all social relations. It is considered as an amalgamation of three related problems: a lack of knowledge (ignorance), negative attitudes (prejudice), and exclusion or avoidance behaviours (discrimination) [3,4].The mentally ill are labelled as different from other people and are viewed negatively by others. Stigmatisation can lower a person's self esteem, contribute to disrupted family relationships, and affect employability [5]. It is a barrier to the provision of mental health services by health planners [6].Many studies have demonstrated that persons labelled as mentally ill are perceived with more negative attributes and rejection regardless of their behaviour [7-9]. Research has shown that people who are labelled as mentally ill associate themselves with society's negative conceptions of mental illness, and that society's negative reactions contribute to the incidence of mental disorder. [10]. However, other studies have demonstrated that negative societal reactions are the result, rather than the cause, of mental illness [11].Individuals who perpetuate stigma are likely to socially distance themselves from persons with mental illness. Social distance may manifest itself in such discriminatory practices as, for example, not renting property to or hiring people who have psychiatric disabilities [5,12].Stigmatising views about mental illness are not limited to uninformed members of the general public; even well-trained professionals from most mental health disciplines subscribe to stereotypes about mental illness [13,14]. Medical students have been shown to have stigmatising attitudes toward mental illness which they hold onto in their professional lives [15]. Therefore, research on attitudes toward mental illness, specifically of those in mental health related fields, is necessary to ensure quality care to persons with mental illness. This is important because interventions directed at these target groups may be more cost effective than interventions directed at the general public [16].This study aimed to evaluate the effect of a psychiatric label attached to an apparently normal person on the attitude of final year medical students in a Nigerian university.MethodsThis was a cross sectional questionnaire based study conducted among the final year medical students of Olabisi Onabanjo University, Ogun State. Participation was on a voluntary basis. A questionnaire containing demographic information, a single-paragraph case description illustrating a normal person, a social distance scale and questions on expected burden was used to elicit response from the students.The students were randomly assigned into two groups using their matriculation numbers. Group A received a case description with a psychiatric label attached while group B received the same case description but without a psychiatric label.The case description is as follows 'Mr AB is a young man who can express his feelings and thoughts among those close to him, although he sometimes gets anxious while talking in a group consisting of strangers. He gets along all right with his family most of the time. Generally he also gets along with other people. Compared to those of his age, his life can be considered as organised. He is generally an optimistic and happy person. In summary, he establishes a good balance between his social life and study'. The students were assigned to one of the two conditions of the case description. One condition involved adding the sentence 'This young man has been diagnosed as having mental illness by the doctor who examined him' to the end of the case description. In the second condition no psychiatric label was attached to the case description.Each case description was followed by 16 questions to be rated on a 4-point scale ranging from definitely agree to definitely disagree. The questions from 1 to 13 were designed to measure social distance between oneself and the person depicted in the case description while questions 14, 15 and 16 assess the possible burden expected from a mentally ill person one may associate with. The case description and the questionnaire were modified versions of those used in previous studies that concerned psychiatric label and attitude to mental illness [17].The data derived from the responses of the students were analysed using SPSS v.10 (SPSS Inc., Chicago, IL, USA). Results are presented in frequencies and percentages. The Chi square test was used to determine statistical difference between proportions while the Student t test was used to determine the statistical difference between means. A p value less than 0.05 was considered as statistically significant.ResultsA total of 144 students responded to the questionnaire out of a class of 167. Thus, the response rate was 86.2%. In all, 81 (56.2%) were males while 63 (43.8%) were females. A total of 68 (47.2%) of the students responded to the questionnaire with the attached psychiatric label (male 55.9%, female 44.1%) while 76 (52.8%) responded to the questionnaire without the attached psychiatric label (male 56.6%, female 43.4%) (p = 0.933). The mean (SD) age of the students that responded to the questionnaire with the attached psychiatric label was 27.07 (3.33) years compared with 26.96 (2.18) years for those that responded to the questionnaire without the psychiatric label (p = 0.187).Table 1 shows the responses of the students to the questions about the man depicted in the case description. The students that responded to the questionnaire with the attached psychiatric label were significantly more unwilling to rent out their houses to the man depicted in the case description compared to those that responded to the questionnaire without the attached label (p = 0.003). Similarly, they were unwilling to have him as their next-door neighbour (p = 0.004) or have him as their barber or hairdresser (p = 0.000) compared with the group that responded to the questionnaire without the attached label. They were also not willing to share an office with him (p = 0.000) or allow their sister to get married to him (p = 0.000).Table 1responses of the students to the person depicted in the case descriptionLabel attached frequency, % (n = 68)No label attached: frequency, % (n = 76)p ValueUncomfortable sitting close to him on public transport36 52.931 40.80.144Disturbed by shopping from a market which he runs13 19.116 21.10.774Willing to let your house to him39 57.461 80.30.003Ill at ease by his working as a gateman at your house31 45.625 32.90.112Disturbed participating in a social gathering to which he has been invited13 19.120 26.30.308Willing to play cards with him at a social gathering54 79.450 65.80.087Willing to chat with him on political matters at a social gathering41 60.349 64.50.614Willing to tell him about your own private problems19 27.927 35.50.305Disturbed by his becoming your next-door neighbour20 29.408 10.50.004Will have my hair cut/styled by him if he was a barber/hairdresser26 38.257 75.00.000Disturbed by working in the same place as him03 04.105 06.60.616Will be worried sharing the same room with him if you work at the same place34 50.012 15.80.000Disturbed by your sister wanting to marry him49 72.128 36.80.000Will be an emotional burden on you in your friendship with him17 25.009 11.80.021Will exhaust your physical energy in your friendship with him20 29.409 11.80.005Your friendship with him will have a negative influence on your mental health08 11.710 13.20.884Significantly, the students that responded to the questionnaire with the attached label felt that the man in the case description will exhaust them both physically (p = 0.005) and emotionally (p = 0.021) in their relationship with him, compared with those that responded to the questionnaire without the psychiatric label.DiscussionThis study set out to investigate the effect of a psychiatric label attached to an apparently normal person in a case description on the attitude of final year medical students toward psychiatrically ill patients. The students have had previous clinical exposure to psychiatry in the course of their medical training. The finding in this study indicated that a label of mental illness on the person depicted in the case description elicited negative attitude that resulted in the students wanting to maintain a significant distance from the person that was labelled mentally ill. The results provided strong support for the influence of labelling on certain attitudes. These attitudes were more obvious in circumstances that could bring a closer relationship between the respondents and the person depicted in the case description. They were not willing to have him as their barber/hairdresser, they would discourage their sister from planning to get married to him, and they were uncomfortable with the thought of sharing an office with him. These findings are consistent with previous studies on the influence of psychiatric label on attitude towards mental illness [8,9,17]. Furthermore, the students felt that friendship with the labelled person would be a burden on them physically and emotionally. This could further worsen the social distance between them and the labelled person. These stigmatising attitudes have been shown to increase psychological distress in people labelled to be mentally ill [18]. Moreover such attitudes may inhibit help seeking among individuals with a mental disorder [19,20] and provide barriers to their successful reintegration into the society [21].The findings in this study provide support for an earlier report by Adewuya and Makanjuola [22] on the attitudes of students generally toward the mentally ill in a Nigerian university. This however, challenges studies where less stigmatisation of mental illness was reported for non-Western cultures especially of Asian and African countries [23,24]. Although a dearth of research on this issue was given for the observation in these cultures, Fabrega, however, noted that lack of differentiation between psychiatric and non-psychiatric disorders in non-Western cultures could be an important factor for less stigmatisation [23,24]. It is however important to note that this study was conducted in a group of students who are medically inclined, and hence they should be able to differentiate between psychiatric and non-psychiatric disorders.Although a considerable number of studies have consistently reported improvement in attitude of medical students toward psychiatry after clinical exposure [25,26], follow-up studies on such students have queried the sustenance of the observed improvement in attitude toward psychiatry over time even before they eventually graduate from medical school [27]. Thus, the finding of a stigmatising attitude of the final year medical students in this study may be a reflection of this decay. The challenge then will be to find a way of sustaining the initial improvement reported in the literature. The focus will be to provide a more cost effective approach of educating the medical students on stigma reduction in mental health. This is particularly important because stigma involves different but related facets [3,4]. This is however, hindered by deficiencies in the mental health curriculum in medical schools where little or no attention is given to stigma as an issue. Moreover, the common medical textbooks in psychiatry fail to devote elaborate attention to issues on stigmatisation of mental illness.In conclusion, medical students are not exonerated from the list of people that express stigmatising attitude toward the mentally ill. There is therefore the need to equip the students with more knowledge on stigma reduction in mental health.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsOOO conceived the study, and participated in its design, acquisition, analysis and interpretation of data, and in the drafting of the manuscript. OO participated in its coordination, statistical analysis and helped to draft the manuscript. MOO participated in the design of the study, its coordination and the draft of the manuscript. All authors read and approved the final manuscript.\n\nREFERENCES:\nNo References"
+}

batch_3/PMC2535586.json

@@ -0,0 +1,4 @@
+{
+    "id": "PMC2535586",
+    "text": "This is an academic paper. This paper has corpus identifier PMC2535586\nAUTHORS: Sanjeev Kumar, Nikhil Agrawal, Rahul Khanna, AK Khanna\n\nABSTRACT:\nAgressive angiomyxoma is a rare mesenchymal neoplasm. It mainly presents in females. We here present a case of angiomyxoma presenting as huge abdominal lump along with gluteal swelling. Case note is described along with brief review of literature.\n\nBODY:\nBackgroundAggressive angiomyxoma is a rare mesenchymal tumor occurring predominantly in the pelvi-perineal region of females. We report such a case presented as abdominal as well as gluteal lump, a very unusual presentation. Patient underwent laparotomy and tumor was successfully excised.Case presentationAn eighteen year old unmarried female presented with progressive distension of abdomen and swelling in left gluteal region for last ten months. It was associated with mild dull aching pain in lower abdomen and bilateral flanks. She was having normal menstrual history. Examination revealed distended abdomen and a non-tender, diffuse, firm and dull mass was felt all over abdomen. Free fluid in peritoneum could not be demonstrated. There was another 10 × 8 cm boggy swelling on upper postero-medial aspect of left thigh having cross fluctuation with abdominal swelling. On ultrasonography 30 × 18 × 16 cm complex cystic mass was found occupying whole abdomen and pelvis with internal septations and echoes displacing bowel loops posteriorly with mild right sided hydro-ureteronephrosis. CECT scan (Figure 1) showed heterogeneous soft tissue mass, adjacent to coccyx, involving left gluteal region and extending superiorly into pelvis and abdomen with anterolateral displacement of urinary bladder and lateral displacement of bowel with right hydronephrosis. Rectum was displaced anteriorly and to the right. The lesion was complex with variable soft tissue attenuation (+23.0 HU). Fat plane between the mass lesion and pelvic musculature and abdominal wall muscles was intact. Fine needle aspiration cytology was inconclusive and trucut needle biopsy showed round to spindle shaped cells in a loose to fibrous stroma with fair number of intervening vessels, suggestive of spindle cell tumor.Figure 1CECT of the abdomen.Patient underwent laparotomy and the tumor was excised. The tumor was found to be bilobed, weighing about 15 Kg. (Figure 2). Rectum was displaced anteriorly and to right. Right ureter was displaced laterally with hydroureter. Tumor was extending into thigh through left obturator foramen. Urinary bladder was adherent to mass and displaced anteriorly. Uterus and fallopian tubes were normal, tumor was found to be originating from rectovaginal septum. Liver was normal and there was no ascites or lymphadenopathy.Figure 2Intraoperative photograph of the tumor.The histopathology of the specimen showed capillary and cavernous vascular spaces stuffed with blood and separated by edematous fibrous and myxomatous tissue. The fibrous tissue is in form of interlacing or parallel bands of collagen with edema. The myxomatous tissue comprise of stellate cells with fibrillary in mucinous setting (Figure 3). Immuno-histochemistry was positive for vimentin and desmin while negative for actin and myosin. These findings were consistent with the diagnosis of aggressive angiomyxoma.Figure 3Microscopic photograph of the tumor.DiscussionAggressive angiomyxoma is an uncommon mesenchymal neoplasm occurring predominantly in the pelvi-perineal region of adults, first described in 1983 by Steeper and Rosai [1]. About 90% of patients are women, usually of reproductive age [2]. A few cases have been described in males, usually in scrotum. It presents as a painless, poorly circumscribed gelatinous vulvar mass and clinically simulates a bartholin gland cyst or an inguinal hernia. On gross examination the tumors are lobulated, soft to rubbery, solid masses. The cut surface reveals a glistening, soft homogeneous appearance. Recurrent tumors show more prominent areas of hemorrhage and fibrosis. Histologically angiomyxoma is a mesenchymal tumor, composed of fibroblasts within a strong myxoid background. Vascular proliferation is also prominent, and virtually no mitoses are present [3]. The vast majority of cases demonstrate positivity for desmin in the myxoid bundles and/or stromal cells, while actins and CD34 may be variably positive [3]. The estrogen and progesterone receptor positivity suggests that aggressive angiomyxoma might be hormone dependent as rapid growth has been observed during pregnancy. The tumor grows slowly, and its benign nature is suggested by the histology and by the fact that it shows no tendency to metastasize. However, it is locally aggressive and tends to recur (36–72%) after resection [4]. Imaging of these tumors is important to determine extent and, thus, the optimal surgical approach. Sonography shows a mass that is hypoechoic or appears frankly cystic. Angiography usually shows a generally hypervascular mass. It has a characteristic appearance on CT and MR imaging and these techniques reveal the extent of the tumor as well. On CT, the tumor has a well-defined margin and attenuation less than that of muscle. On T2-weighted MR imaging, the tumor has high signal intensity [4]. Treatment is usually surgery in form of wide local excision. Preoperative angiographic embolization, preoperative external beam irradiation and intraoperative electron beam radiotherapy are useful to decrease the chances of local recurrence [5]. Hormonal treatment with a gonadotropin-releasing hormone agonist can be applied for small primary aggressive angiomyxomas in addition to adjuvant therapy for residual tumors [6].ConclusionAlthough a rare diagnosis, aggressive angiomyxoma can present with unusual features. Detailed radiological examination is helpful in suspecting the problem, but histology is gold standard for diagnosis. Wide excision is curative and prognosis of such tumor is good.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsAKK and RK: operating surgeons, SK: collected clinical details including photographs, summarized the case history and prepared final draft, NA: conducted a literature search and prepared first draft. All authors read and approved the final manuscript.ConsentA fully informed written consent was obtained from the patient for the publication of this case report and accompanying images. A copy of the written consent can be sent to Editor-in-Chief of this journal if article is accepted for publication.\n\nREFERENCES:\nNo References"
+}