Molecular Epidemiology Of Sepsis Susceptibility And Genetic Association Study In Patients With Major Trauma

BackgroundMajor trauma, a public health problem worldwide, is the leading cause of death forpersons younger than45years of age. One of the most serious complications of majortrauma is sepsis, which often induces sequential multiple organ dysfunction syndrome(MODS). Therefore, preventing sepsis and subsequent organ dysfunction is crucial in thetreatment of surviving patients with major trauma. Because of this, establishing an earlydiagnosis of sepsis is an essential prerequisite. Accumulating evidence indicates that geneticvariants, particularly single nucleotide polymorphisms (SNPs), are critical determinants forinter-individual differences in both inflammatory responses and clinical outcomes in traumapatients. Delineating the variation in genes and associated differences in response to traumamay contribute to the development of new genetically tailored diagnostic and therapeuticinterventions that will improve outcome in the patients with major trauma.Objective of our research is to identify the genetic variants associated with the risk ofposttraumatic sepsis. First, we reviewed literatures in the field of sepsis genetic associationstudies, and comprehensively evaluated the possible relevance of reported geneticpolymorphisms by means of meta-analysis, which will provide basis for future riskprediction study of posttraumatic sepsis. Second, we selected some importantimmune-inflammation related genes and explored the clinical relevance of these genepolymorphisms in trauma patients in clinical multiple centers, which will provide novelgenetic markers for future risk prediction study.Materials and Methods1. We searched the PubMed, Medline, Embase, Web of Science, and HuGE databaseswith the terms ‘(sepsis OR "septic shock" OR septicemia) AND (polymorphism ORvariation OR mutation)’ to find all studies that investigated associations between geneticvariants and sepsis risk published before24December2013. For genetic variants with atleast two data soures, we conducted meta-analysis under the dominant, recessive, and allelic genetic model, respectively. Moreover, meta-analysis stratified by ethnicity was done if datapermitted. Evaluation of meta-analysis results included the tests of heterogeneity, sensitivityanalyses, and examination for bias. Finally, we applied the Venice criteria to evaluate theepidemiological credibility of all significant associations identified by meta-analysis.2. Genetic variation data for entire genes of NLRP3, SOCSs and pre-miRNA wereobtained from HapMap database for healthy Chinese Han Beijing (CHB) population. Wepartitioned them in “bins” inferred according to the r2linkage disequilibrium (LD) statistic(threshold≥0.8) and a maximally informative tagSNP was selected from each haplotype binusing TAGster. Furthermore, the potential functional SNPs within pre-miRNA were selectedthrough the analysis of miRNA secondary structure and mRNA targets.3. The selected tagging or potential functional SNPs from candidate genes weredetected by pyrosequencing method among1408unrelated severe trauma patients fromChongqing (n=806), Yunnan (n=286), and Zhejiang (n=316) districts, respectively.Genotype distribution of each selected SNP was determined by counting. Hardy-Weinbergequilibrium of each SNP among trauma patients was assessed by χ2test. Three geneticmodels (allele dose, dominant, and recessive) were used. The association of SNPs withsepsis risk was determined by χ2analysis. The association between SNPs and MOD scoreswas determined using one-way analysis of variance. We performed linear regressionanalysis to quantify the allele dose effect of MOD score and logistic regression to estimatethe relative risk of sepsis with age, sex, and ISS to adjust for possible confounding effects.4. The peripheral blood collected from the patients with major trauma within2hoursafter admission and healthy blood donors were incubated with100ng/ml lipopolysaccharide(LPS) at37°C for4hours. The cytokines’ levels in the plasma from trauma patients weredetermined with enzyme-linked immunosorbent assay (ELISA) and the expression of CISin the leukocytes from healthy blood donors after LPS stimulation was performed byreal-time PCR.5. The possible effect of rs2027432(-1017G/A) on NLRP3promoter activity andrs414171(-237A/T) on CIS promoter activity were investigated using a reporter gene assaysystem. The wild-type and mutated sequences of NLRP3or CIS promoter were inserted intoa promoterless pGL3-Basic vector, respectively, which were transiently transfected intoHuman model cell lines and luciferase activities were measured with a Luminoskan Ascent luminometer at48hours posttransfection. In addition, the effect of rs4919510G/C on theexpression of mature miR-608was investigated using TaqMan probes in transfected cellswith the miR-608expression vectors.Results1. From8796reviewed articles,232were included based on predefined inclusioncriteria. Among the main meta-analyses for123variants within56genes,21variants within17genes showed nominally significant association (P<0.05) for sepsis risk. Strongassociations with sepsis were detected for five variants (NOD2rs2066844C/T, LBPrs2232618T/C, TNFA rs361525G/A, IL-10rs1800894, SETPD rs12219080). Moderateassociations were found for six variants (TLR1rs5743611G/C, RAGE rs1800625T/C, MD2rs11465996C/G, IL-12B rs2195940, PBEF rs61330082T/C, SFTPD rs1998374). No variantwas graded strong for cumulative epidemiological evidence of association with sepsis riskusing Venice criteria. Five variants were graded moderate and the remaining16variantswere graded weak. Through stratified meta-analyses by ethnicity, one additional variant(IL-10rs1800896) was identified to be nominally associated with sepsis risk and graded itweak for cumulative evidence of association with sepsis risk. Among the independent studyfor177variants within94genes,42variants within31genes showed nominally significantassociation (P<0.05) for sepsis risk. To gain definitive proof of the relationships, furthertrials with adequate patient numbers and proper power would be required.2. Six tagSNPs were selected from35SNPs within the entire NLRP3gene in CHBpopulation through construction of haplotype bins. Similarly, twelve tagSNPs were selectedfrom145SNPs within seven SOCS genes according to the r2LD statistic. Moreover, ninepotential functional SNPs from26pre-miRNA SNPs by the analysis of miRNA secondarystructure and mRNA targets. In total, twenty-seven SNPs from NLRP3, SOCSs, andpre-miRNAs were selected.3. Among the six tagSNPs of NLRP3, two of them (rs2027432-1017G/A andrs12048215+5134A/G) were well associated with the development of sepsis or MODS inChongqing cohort, showing that the patients carrying A allele had significantly higher sepsismorbidity or MOD scores and higher LPS-induced IL-1β production. However, nocombined effects were found between these two polymorphisms. In addition, the-1017Aallele could significantly enhance the promoter activities of the NLRP3gene. 4. Among the twelve tagSNPs of SOCSs, the rs414171-237A�'T of CIS andrs3748726+49083T�'C of SOCS7showed lower sepsis morbidity and MOD scores in bothChongqing and Zhejiang cohort, respectively and synergistically. For rs414171polymorphism, after exposure to LPS, CIS mRNA expression levels in peripheralleukocytes with AT and TT genotypes were significantly lower than in AA homozygotes.Moreover, there were significant lower plasma TNF-α and higher IL-10expression levelsfor the rs414171T allele. The promoter luciferase activity assay further confirmed that-237T could decrease the activity of the CIS promoter.5. Among the nine selected SNPs with potential functional significance, only one(miR-608rs4919510) was found to be strongly associated with a higher risk of developingsepsis and MODS in all three independent cohorts from Chongqing, Yunnan, and Zhejiangdistricts, respectively. An even stronger association was observed when combing these threestudy cohorts together. In the genotype-phenotype correlation analysis, the rs4919510polymorphism showed a significant correlation with a lower production ofanti-inflammatory cytokines and a higher production of pro-inflammatory cytokines. Invitro experiments further indicated that the G�'C variant could significantly enhance theexpression of mature miR-608.ConclusionsWe found some genetic variants which might influence the risk of sepsis throughmeta-analysis and systematic review for the first time. Furthermore, through multicenterprospective association study, we identified that five novel polymorphisms (rs2027432andrs12048215of NLRP3, rs414171of CIS, rs3748726of SOCS7, and rs4919510of miR-608)may affect the development of sepsis and MODS after trauma. These genetic variants mightbe used as biomarkers for the predictors of sepsis or MODS development, which are criticalfor the early discovery and personal therapy of sepsis.