- Open Access
MicroRNAs implicated in dysregulation of gene expression following human lung transplantation
© Zhang et al.; licensee Springer. 2013
- Received: 18 April 2013
- Accepted: 6 August 2013
- Published: 8 August 2013
Lung transplantation remains the only viable treatment option for the majority of patients with advanced lung diseases. However, 5-year post-transplant survival rates remain low primarily secondary to chronic rejection. Novel insights from global gene expression profiles may provide molecular phenotypes and therapeutic targets to improve outcomes after lung transplantation.
Whole-genome gene expression profiling was performed in a cohort of patients that underwent lung transplantation as well as healthy controls using the Affymetrix Human Exon 1.0ST Array. To explore the potential roles of microRNAs (miRNAs) in regulating lung transplantation-associated gene dysregulation, miRNA expression levels were also profiled in the same samples using the Exiqon miRCURY LNA Array.
In a cohort of 18 lung transplant patients, 364 dysregulated genes were identified in Caucasian patients relative to normal individuals without pulmonary disorders. Pathway enrichment analysis of the dysregulated genes pointed to Gene Ontology biological processes such as “defense response”, “immune response” and “response to wounding”. We then compared the expression profiles of potential regulating miRNAs, suggesting that dysregulation of a number of lung transplantation-associated genes (e.g., ATR, FUT8, LRRC8B, NFKBIA) may be attributed to the dysregulation of their respective regulating miRNAs.
Following human lung transplantation, a substantial proportion of genes, particularly those genes involved in certain biological processes like immune response, were dysregulated in patients relative to their healthy counterparts. This exploratory analysis of the relationships between miRNAs and their gene targets in the context of lung transplantation warrants further investigation and may serve as novel therapeutic targets in lung transplant complications.
- Lung transplant
- Gene expression
- Gene ontology
For many patients with end-stage lung diseases, lung transplantation is often the only remaining viable therapeutic measure . The number of lung transplants is ~1,500 on average a year in the United States, which represents ~45% of lung transplants performed world-wide . Compelling data have documented the beneficial impact of lung transplantation on functional status, hemodynamics, and quality of life. Less compelling, however, is the demonstration of a survival benefit due to significant constraints on long-term survival .
Although short-term survival has improved via improved surgical techniques, donor preservation and immunosuppressive agents, long-term survival remains reduced after lung transplantation. The major cause of decreased long-term survival is bronchiolitis obliterans syndrome (BOS), a physiological measure of chronic rejection after lung transplantation. Approximately 50% of lung transplant recipients will develop BOS by five years post transplantation. However, the pathogenesis of BOS has not been clearly elucidated. Both alloimmune-dependent and -independent factors have been suggested to contribute to BOS pathogenesis. These factors include acute rejection, lymphocytic bronchiolitis, acute infectious etiologies and gastroesophageal reflux disease.
In addition, an individual patient’s genetic make-up may also contribute to the prognosis after lung transplantation as well as to the development of various complications. Particularly, global gene expression profiling has been used to identify unique expression signatures in organ transplant biopsies that may help distinguish various outcomes such as acute rejection, acute dysfunction without rejection and well-functioning transplants with no rejection history [3–5]. For example, gene expression in bronchoalveolar lavage cell samples from lung transplant recipients with and without acute rejection on simultaneous lung biopsies was examined, and specific expression patterns were demonstrated at defined time points after transplantation in allografts . Though not definitive and comprehensive, these studies showed the potential power of whole-genome microarrays to identify biomarkers of acute/chronic transplant rejection and development of other complications.
Notably, gene expression itself has been demonstrated to be a complex and quantitative trait that varies within and between natural human populations [6–11] and is controlled by various genetic, epigenetic and non-genetic factors [12–15]. MicroRNAs (miRNAs), small (21-25 nt) non-coding RNA molecules, have emerged as a novel class of gene regulators that may affect various complex phenotypes including disease susceptibility and drug response [16, 17]. Integrating whole-genome mRNA and miRNA profiles, therefore, could help elucidate the complex cellular response and its mechanisms in lung transplant patients, and provide novel biomarkers for the outcomes of lung transplantation. Specifically, we compared whole-genome transcriptional expression data profiled using the Affymetrix Human Exon 1.0ST Array (exon array) in peripheral blood mononuclear cells (PBMCs) from lung transplant patients and normal individuals. We searched for any enriched pathways or biological processes among the dysregulated genes in lung transplant patients. We further demonstrated that miRNAs could potentially play a critical role in determining the gene expression dysregulation observed in lung transplant patients.
Subjects and collection of PBMC samples
Summary of the study cohort
African American controls
Age (mean ± SD) years
56 ± 18
65 ± 12
48 ± 9
75 ± 7
Gender (F) (%)
Type of transplant
Time from transplant(mean ± SD) months
10 ± 9
11 ± 4
Obtaining exon array data
Total RNA was extracted from PBMCs and prepared using standard molecular biology protocols. RNA concentration and purity was determined before gene expression profiling using the Affymetrix Human Exon 1.0ST Array (exon array) (Affymetrix, Inc., Santa Clara, CA). The microarray labeling, hybridization and processing was performed at the University of Chicago Microarray Core Facility according to the manufacturer’s protocol.
Processing of exon array data
We used the experimental probe masking workflow provided by the Affymetrix Power Tools v.1.12.0 (http://www.affymetrix.com/) to filter the probeset (exon-level) intensity files by removing probes that contain known SNPs in the dbSNP database v129 [18, 19]. The resulting probe signal intensities were quantile normalized over all samples, summarized with the robust multi-array average (RMA) algorithm  and log2 transformed with a median polish  for ~22,000 transcript clusters (gene-level) with the core set (i.e., with RefSeq-supported annotations) . Adjustment for possible batch effect was conducted by COMBAT (http://jlab.bu.edu/ComBat/) . We consider a transcript cluster to be reliably expressed if the DABG (detection above ground)  p-value computed by the Affymetrix Power Tools was less than 0.01 in at least 80% of the samples in each test group (healthy controls or patients) in each population, respectively. We further limited our analysis set to the genes with unambiguous annotations by Affymetrix. Totally, 11,461 and 11,576 transcript clusters in the Caucasian American and African American samples, respectively, met these criteria and were further analyzed. We have deposited the raw and processed exon array data in the NCBI Gene Expression Omnibus (GEO) (Accession Number: GSE49081).
Obtaining miRNA expression data
The expression levels of miRNAs were profiled using the Exiqon miRCURY LNA Array v10.0 (∼700 human miRNAs, updated to miRBase 11.0 annotation)  (Exiqon, Inc., Denmark). Briefly, total RNA from PBMCs was extracted and prepared according to manufacturer’s protocol. Array hybridization was performed by Exiqon with the quantified signals background corrected using normexp with offset value 10 based on a convolution model  and normalized using the global Lowess regression algorithm. In total, 318 miRNAs and 309 miRNAs were found to be expressed in the Caucasian American samples and the African American samples, respectively (i.e., present in at least 80% of total samples in each population).
Identifying genes dysregulated in patients with lung transplants
We excluded genes on chromosomes X and Y to avoid the potential confounding effect of gender. SAM (Significance Analysis of Microarrays) , implemented in the samr library of the R Statistical Package , was used to identify differential genes between patients who underwent lung transplantation and healthy controls in the Caucasian American and African American samples, respectively, as well as between patients with and without development of BOS. Transcripts with a greater than 1.5 fold-change and q-value  less than 0.01 (i.e., 1% FDR, false discovery rate) were deemed significantly dysregulated. We searched for any enriched pathways and biological processes among the differential genes relative to the respective analysis set using the DAVID (Database for Annotation, Visualization and Integrated Discovery) tool [29, 30]. The following databases were included: KEGG (Kyoto Encyclopedia of Genes and Genomes) , BioCarta (http://www.biocarta.com/), Reactome , PANTHER , and Gene Ontology (GO) . Due to the exploratory nature of this study, we chose to use a relatively lenient cutoff, i.e., FDR < 25% after the Benjamini-Horchberg procedure  and a minimum of 5 differential genes in a pathway or biological process, for the DAVID analysis.
Identifying relationships between dysregulated genes and potential regulating miRNAs
The differential genes were then searched against the MicroCosm Targets  (i.e., miRanda algorithm) through our ExprTarget database (http://www.scandb.org/apps/microrna/)  for potential regulating miRNAs (miRanda p < 0.0001). Only human miRNAs that are expressed in these samples (318 miRNAs in the Caucasian samples; 309 miRNAs in the African American samples) were included in the analysis. The expression patterns of those miRNAs and their corresponding gene targets were compared between patients and normal controls using standard t-test. Significant miRNA-mRNA relationships (i.e., negative association between miRNA and mRNA at t-test p < 0.05) were further confirmed using linear regression. The Pearson correlation coefficients and the associated p-values (cutoff p < 0.05) were calculated using the lm library of the R Statistical Package .
Identifying genes dysregulated in patients with lung transplants
Enriched pathways among dysregulated genes
Some enriched pathways and biological processes among the dysregulated genes in Caucasian American patients
GO:0009617 ~ response to bacterium
GO:0006955 ~ immune response
GO:0009611 ~ response to wounding
GO:0006952 ~ defense response
GO:0050817 ~ coagulation
GO:0007596 ~ blood coagulation
GO:0007599 ~ hemostasis
GO:0009617 ~ response to bacterium
GO:0006955 ~ immune response
GO:0006952 ~ defense response
GO:0007610 ~ behavior
GO:0010033 ~ response to organic substance
GO:0009611 ~ response to wounding
Identifying potential regulating miRNAs for the dysregulated genes
We searched for potential regulating miRNAs for the dysregulated genes in lung transplant patients based on the predictions of the miRanda algorithm . Among the 292 down-regulated genes in the Caucasian American patients, 178 miRNA-mRNA relationships corresponding to 95 expressed miRNAs and 78 genes were identified, while 74 miRNA-mRNA relationships corresponding to 40 expressed miRNAs and 31 genes were identified in the 72-up-regulated genes (miRanda p < 0.0001). In comparison, nine miRNAs were identified for the single down-regulated gene, SMOX (encoding spermine oxidase) in the African American patients (miRanda p < 0.0001).
Some dysregulated genes in the Caucasian American patients are negatively associated with potential regulating miRNAs
p (miRanda) a
p (t-test) b
p (linear regression) c
r 2(linear regression) d
transcription factor 4
leucine rich repeat containing 8 family, member B
chromosome 14 open reading frame 2
chromosome 14 open reading frame 135
ataxia telangiectasia and Rad3 related
pyrin and HIN domain family, member 1
transcription factor 4
carbonic anhydrase I
nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, alpha
nuclear factor, interleukin 3 regulated
dedicator of cytokinesis 4
polo-like kinase 2
polo-like kinase 2
Lung transplantation is associated with major complications such as infection, acute rejection and chronic rejection characterized by BOS . Elucidating the complex cellular and physiological response after lung transplantation will be critical to understanding the pathogenesis of acute and chronic complications after lung transplantation. To our knowledge, this is the first study to assess the relationships between dysregulated genes and potential gene regulators of miRNAs in patients that underwent lung transplantation.
Approximately 3% of the analyzed genes were differentially expressed in Caucasian patients with lung transplants, indicating systematical dysregulation of certain genes in these patients, potentially implicating their outcomes after lung transplantation. Using DAVID [29, 30], these lung transplant-associated genes were found to be enriched in a number of known pathways and GO  biological processes including “immune response”, “defense response”, “response to wounding”, “hemostasis” and “blood coagulation” (Table 2). Interestingly, biological processes such as “blood coagulation” and “hemostasis” were enriched among down-regulated genes, while biological processes such as “immune response”, “defense response”, “response to bacterium”, and “response to wounding” were enriched among up-regulated genes. Lung transplant patients are routinely anticoagulated to prevent thrombosis, and given antibiotic prophylaxis to prevent infections and immunosuppressants to prevent organ rejection. It appears that after lung transplantation and relevant treatments, genes related to blood coagulation were significantly down-regulated in patients, while genes related to the aftermath of a major surgery including “response to wounding” were significantly up-regulated in patients.
Notably, many of these pathways shared a significant number of genes, displaying the complex interactions of several biological processes after lung transplantation. For example, all of the seven up-regulated genes including FOS (encoding FBJ murine osteosarcoma viral oncogene homolog), IL8 (encoding interleukin 8), IL1B (encoding interleukin 1, beta), involved in “inflammatory response” were also involved in “response to wounding”. In addition, many genes related to the response to bacterial infection and lipopolysaccharide (LPS) were up-regulated in patients such as NFKBIA, FOS, PTGS2 (encoding prostaglandin-endoperoxide synthase 2), ADM (encoding adrenomedullin), SOCS3 (encoding suppressor of cytokine signaling 3), TRIB1 (encoding tribbles homolog 1) and IL1B. Among them, FOS, PTGS2 and IL1B are genes involved in “response to glucocorticoid stimulus”, which was also enriched in the up-regulated genes and likely reflected treatment effect. Besides these up-regulated immune response genes, 20 other immune response genes such as CCR4 (encoding chemokine receptor 4), CD86 (encoding CD86 molecule), were down-regulated in patients. Since lung transplant patients were treated continuously with immunosuppressive drugs, the dysregulation of some of these immune response genes could be due to the on-going immunosuppressant treatments. Obviously, some immune response genes such as FOS, PTGS2, IL1B could be induced by immunosuppressive drugs (e.g., glucocorticoids), while some other immune response genes could be suppressed by drug treatments. Dysregulation of these genes after transplantation provides more insight regarding the interactions of various biological processes and may ultimately provide biomarkers of the various complications related to outcomes after transplantation.
Among the 364 differential genes in patients, a number of genes showed an expression pattern correlated with their potential regulating miRNAs (Table 3). Using a linear regression model, we demonstrated that expression of specific miRNAs was significantly correlated with the expression levels of their potential gene targets. For example, ATR and LRRC8B (encoding leucine rich repeat containing 8 family, member B) were down-regulated in transplant patients. Their expression levels were significantly correlated with their potential regulating miRNA, respectively (Figure 2). The majority of the identified miRNA-mRNA relations could be confirmed using linear regression. Notably, the gene NFKBIA was found to be negatively associated with its potential regulating miRNA has-miR-381 (Table 3). NFKBIA is also involved in “response to wounding” and “response to LPS”, suggesting that miRNAs may contribute to these biological processes in lung transplant patients. Our results suggest that the complex dynamics of dysregulated genes in these patients may be partially attributed to the differential expression of their potential regulating miRNAs following lung transplantation, as well as relevant treatments such as immunosuppressive drugs and anticoagulants.
We recognize that there are some limitations to this exploratory study. Firstly, the sample size was small and therefore, this analysis must be validated with a larger cohort of patients. Secondly, some potential confounding factors (e.g., types of immunosuppressive agent that the patients are taking) might influence gene dysregulation. In addition, we were unable to compare our findings between different ethnic populations, as well as derive more robust conclusions for BOS-associated dysregulation, given the small sample size. However, given the exploratory nature of this analysis, our primary goal was to determine the putative relationships between dysregulated genes and regulating miRNAs in patients after lung transplantation. Indeed, we were able to show a significant number of miRNA-mRNA relationships, suggesting that the regulation of gene targets by these miRNAs in the context of lung transplantation warrant further investigation, and could ultimately serve as novel therapeutic targets in lung transplant complications.
Despite improved short-term survival rates, long-term survival rates after lung transplantation remain dismal. Examination of global mRNA and miRNA, an important class of gene regulators, expression profiles in lung transplant patients may provide novel insights into the pathogenesis of transplantation-associated complications. We showed the presence of a significant number of dysregulated genes, particularly those genes involved in pathways and biological processes such as immune response and defense, in the PBMCs derived from a cohort of patients after lung transplantation. The contribution of miRNAs in regulating these differential genes was also demonstrated. The relationships between miRNAs and those dysregulated genes in the context of lung transplantation warrant further investigation, and may serve as novel therapeutic targets in lung transplantation-associated complications.
This work was supported by an NIH grant, HL05864. The funding body had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
- Kotloff RM: Does lung transplantation confer a survival benefit? Curr Opin Organ Transplant 2009,14(5):499–503. 10.1097/MOT.0b013e32832fb9f8PubMedView ArticleGoogle Scholar
- McCurry KR, Shearon TH, Edwards LB, et al.: Lung transplantation in the United States, 1998–2007. Am J Transplant 2009,9(4 Pt 2):942–958.PubMedView ArticleGoogle Scholar
- Flechner SM, Kurian SM, Head SR, et al.: Kidney transplant rejection and tissue injury by gene profiling of biopsies and peripheral blood lymphocytes. Am J Transplant 2004,4(9):1475–1489. 10.1111/j.1600-6143.2004.00526.xPubMed CentralPubMedView ArticleGoogle Scholar
- Lande JD, Patil J, Li N, Berryman TR, King RA, Hertz MI: Novel insights into lung transplant rejection by microarray analysis. Proc Am Thorac Soc 2007,4(1):44–51. 10.1513/pats.200605-110JGPubMed CentralPubMedView ArticleGoogle Scholar
- Lu BS, Yu AD, Zhu X, Garrity ER Jr, Vigneswaran WT, Bhorade SM: Sequential gene expression profiling in lung transplant recipients with chronic rejection. Chest 2006,130(3):847–854. 10.1378/chest.130.3.847PubMedView ArticleGoogle Scholar
- Zhang W, Duan S, Kistner EO, et al.: Evaluation of genetic variation contributing to differences in gene expression between populations. Am J Hum Genet 2008,82(3):631–640. 10.1016/j.ajhg.2007.12.015PubMed CentralPubMedView ArticleGoogle Scholar
- Stranger BE, Forrest MS, Dunning M, et al.: Relative impact of nucleotide and copy number variation on gene expression phenotypes. Science 2007,315(5813):848–853. 10.1126/science.1136678PubMed CentralPubMedView ArticleGoogle Scholar
- Stranger BE, Nica AC, Forrest MS, et al.: Population genomics of human gene expression. Nat Genet 2007,39(10):1217–1224. 10.1038/ng2142PubMed CentralPubMedView ArticleGoogle Scholar
- Spielman RS, Bastone LA, Burdick JT, Morley M, Ewens WJ, Cheung VG: Common genetic variants account for differences in gene expression among ethnic groups. Nat Genet 2007,39(2):226–231. 10.1038/ng1955PubMed CentralPubMedView ArticleGoogle Scholar
- Morley M, Molony CM, Weber TM, et al.: Genetic analysis of genome-wide variation in human gene expression. Nature 2004,430(7001):743–747. 10.1038/nature02797PubMed CentralPubMedView ArticleGoogle Scholar
- Cheung VG, Conlin LK, Weber TM, et al.: Natural variation in human gene expression assessed in lymphoblastoid cells. Nat Genet 2003,33(3):422–425. 10.1038/ng1094PubMedView ArticleGoogle Scholar
- Zhang W, Ratain MJ, Dolan ME: The HapMap resource is providing new insights into ourselves and its application to pharmacogenomics. Bioinform Biol Insights 2008,2(1):15–23.PubMed CentralPubMedGoogle Scholar
- Cheung VG, Spielman RS: Genetics of human gene expression: mapping DNA variants that influence gene expression. Nat Rev Genet 2009,10(9):595–604. 10.1038/nrg2630PubMed CentralPubMedView ArticleGoogle Scholar
- Gilad Y, Rifkin SA, Pritchard JK: Revealing the architecture of gene regulation: the promise of eQTL studies. Trends Genet 2008,24(8):408–415. 10.1016/j.tig.2008.06.001PubMed CentralPubMedView ArticleGoogle Scholar
- Moen E, Zhang X, Mu W, et al.: Genome-wide variation of cytosine modifications between European and African populations and the implications for complex traits. Genetics 2013,194(4):987–996. 10.1534/genetics.113.151381PubMed CentralPubMedView ArticleGoogle Scholar
- Zhang W, Dolan ME: The emerging role of microRNAs in drug responses. Curr Opin Mol Ther 2010,12(6):695–702.PubMed CentralPubMedGoogle Scholar
- Zhou T, Garcia JG, Zhang W: Integrating microRNAs into a system biology approach to acute lung injury. Transl Res 2011,157(4):180–190. 10.1016/j.trsl.2011.01.010PubMed CentralPubMedView ArticleGoogle Scholar
- Sherry ST, Ward MH, Kholodov M, et al.: dbSNP: the NCBI database of genetic variation. Nucleic Acids Res 2001,29(1):308–311. 10.1093/nar/29.1.308PubMed CentralPubMedView ArticleGoogle Scholar
- Duan S, Zhang W, Bleibel WK, Cox NJ, Dolan ME: SNPinProbe_1.0: a database for filtering out probes in the Affymetrix GeneChip human exon 1.0 ST array potentially affected by SNPs. Bioinformation 2008,2(10):469–470. 10.6026/97320630002469PubMed CentralPubMedView ArticleGoogle Scholar
- Irizarry RA, Hobbs B, Collin F, et al.: Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 2003,4(2):249–264. 10.1093/biostatistics/4.2.249PubMedView ArticleGoogle Scholar
- Pruitt KD, Tatusova T, Maglott DR: NCBI reference sequences (RefSeq): a curated non-redundant sequence database of genomes, transcripts and proteins. Nucleic Acids Res 2007,35(Database issue):D61-D65.PubMed CentralPubMedView ArticleGoogle Scholar
- Johnson WE, Li C, Rabinovic A: Adjusting batch effects in microarray expression data using empirical Bayes methods. Biostatistics 2007,8(1):118–127. 10.1093/biostatistics/kxj037PubMedView ArticleGoogle Scholar
- Affymetrix: Exon array background Correction. Affymetrix Whitepaper. 2005. http://media.affymetrix.com/support/technical/whitepapers/exon_background_correction_whitepaper.pdf Google Scholar
- Griffiths-Jones S, Saini HK, van Dongen S, Enright AJ: miRBase: tools for microRNA genomics. Nucleic Acids Res 2008,36(Database issue):D154-D158.PubMed CentralPubMedGoogle Scholar
- Ritchie ME, Silver J, Oshlack A, et al.: A comparison of background correction methods for two-colour microarrays. Bioinformatics 2007,23(20):2700–2707. 10.1093/bioinformatics/btm412PubMedView ArticleGoogle Scholar
- Tusher VG, Tibshirani R, Chu G: Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci USA 2001,98(9):5116–5121. 10.1073/pnas.091062498PubMed CentralPubMedView ArticleGoogle Scholar
- R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2005.Google Scholar
- Tibshirani R, Hastie T, Narasimhan B, Chu G: Diagnosis of multiple cancer types by shrunken centroids of gene expression. Proc Natl Acad Sci USA 2002,99(10):6567–6572. 10.1073/pnas.082099299PubMed CentralPubMedView ArticleGoogle Scholar
- da Huang W, Sherman BT, Lempicki RA: Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 2009,4(1):44–57.PubMedView ArticleGoogle Scholar
- Dennis G Jr, Sherman BT, Hosack DA, et al.: DAVID: database for annotation, visualization, and integrated discovery. Genome Biol 2003,4(5):P3. 10.1186/gb-2003-4-5-p3PubMedView ArticleGoogle Scholar
- Kanehisa M, Goto S, Kawashima S, Okuno Y, Hattori M: The KEGG resource for deciphering the genome. Nucleic Acids Res 2004,32(Database issue):D277-D280.PubMed CentralPubMedView ArticleGoogle Scholar
- Croft D, O’Kelly G, Wu G, et al.: Reactome: a database of reactions, pathways and biological processes. Nucleic Acids Res 2011,39(Database issue):D691-D697.PubMed CentralPubMedView ArticleGoogle Scholar
- Mi H, Lazareva-Ulitsky B, Loo R, et al.: The PANTHER database of protein families, subfamilies, functions and pathways. Nucleic Acids Res 2005,33(Database issue):D284-D288.PubMed CentralPubMedView ArticleGoogle Scholar
- Ashburner M, Ball CA, Blake JA, et al.: Gene ontology: tool for the unification of biology. The gene ontology consortium. Nat Genet 2000,25(1):25–29. 10.1038/75556PubMed CentralPubMedView ArticleGoogle Scholar
- Benjamini Y, Hochberg Y: Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Statist Soc B 1995, 57: 289–300.Google Scholar
- Gamazon ER, Im HK, Duan S, et al.: Exprtarget: an integrative approach to predicting human microRNA targets. PLoS One 2010,5(10):e13534. 10.1371/journal.pone.0013534PubMed CentralPubMedView ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.