Genomic Variations Within DEFB1 Are Associated with the Susceptibility to and the Fatal Outcome of Severe Sepsis in Chinese Han Population

SHORT COMMUNICATION Genomic variations within DEFB1 are associated with the susceptibility to and the fatal outcome of severe sepsis in Chinese Han population

Q-X Chen, C Lv, L-X Huang, B-L Cheng, G-H Xie, S-J Wu and X-M Fang Department of Anesthesiology, The First Affiliated Hospital, College of Medicine, Zhejiang University, Hangzhou, China

Sepsis is a systemic inflammatory response syndrome to infection. Human b-defensin 1 (DEFB1) is a multifunctional mediator in infection and inflammation, which has been largely explored in ex vivo studies. The present case–control study was designed to investigate whether DEFB1 genomic variations are associated with the susceptibility to and the outcome of severe sepsis in 211 patients with severe sepsis and 157 ethnic-matched healthy controls. After correcting for multiple testing, the À44G/C was the only polymorphism found to show significant associations with both the susceptibility to and the fatal outcome of severe sepsis (P ¼ 0.0049, odd ratio (OR) 1.971 and P ¼ 0.002, OR 2.406, respectively). Haplotype À20A/À44C/À52G showed a protective role against severe sepsis (P ¼ 0.0066, OR 0.6751), whereas haplotype À20G/À44G/À52G served as a risk factor for the fatal outcome of severe sepsis (P ¼ 0.0052, OR 2.427). These findings provide further evidence that b-defensin 1 may play a role in the pathogenesis of severe sepsis. Genes and Immunity (2007) 8, 439–443; doi:10.1038/sj.gene.6364401; published online 17 May 2007

Keywords: human b-defensin 1; DEFB1; SNP; haplotype; severe sepsis; association study

Introduction adaptive immunity.16 Evidence suggests that DEFB1 may play an important role during the course of infection as Sepsis is an infection-initiated systemic inflammatory well as inflammation.15,17,18 reaction syndrome resulting from the interactions be- The cluster of human defensin genes including a- tween environmental and genetic factors.1,2 Studies defensin genes and at least four b-defensin genes is confirmed that single nucleotide polymorphisms (SNPs) located in chromosome 8p22–23.5 This region containing in some particular genes, such as tumor necrosis factor a number of genes related to the innate immunity and locus, interleukin-1 receptor antagonist, plasminogen the nervous system shows a strikingly high polymorphic activator inhibitor-1 and so on, play a critical role in rate in the human population.19 Recent studies found sepsis.3,4 that the genes coding a-defensins 1 (DEFA1) and 3 Defensins are cysteine-rich cationic antimicrobial (DEFA3) as well as b-defensins 2 (DEFB4), 3 (DEFB103) polypeptides with three disulfide bridges and are and 4 (DEFB104) exhibit copy number polymorphisms, expressed in plants, insects and mammals.5,6 Human b- while the DEFB1 gene was a single copy gene containing defensin 1 (DEFB1, hBD-1) was the first characterized several SNPs.20–23 The SNPs in DEFB1 have been human b-defensin identified in hemofiltrate.7 DEFB1 is associated with the pathogenesis of asthma and chronic constitutively expressed mainly in kidney, keratinocytes, obstructive pulmonary disease (COPD) as well as 8,9 oral cavity, as well as in urogenital tract. The in vitro infectious disease.24–28 However, the genetic predisposi- expression of DEFB1 can be upregulated by LPS, IL-1b, tion of individuals carrying these SNPs to sepsis remains interferon-g (IFN-g) and arginine.10–12 In vivo, elevated unknown. Hence, the present study was designed to levels of DEFB1 in body fluid were observed in investigate the association of genomic variants within individuals suffering from sepsis and infectious dis- DEFB1 and severe sepsis. eases.13,14 Moreover, the inducible expression of DEFB1 showed interindividual variability upon infection or LPS stimulation.10,15 In addition, DEFB1 has shown chemoat- Results tractant properties for dendritic cells and memory T cells via chemokine receptor CCR6, which links innate and All the enrolled patient and controls were Han Chinese. In patients with severe sepsis, 46.9% individuals have experienced surgical or traumatic history. Severe acute Correspondence: Dr X-M Fang, Department of Anesthesiology, The pancreatitis (18.5%) was the most frequent initial First Affiliated Hospital, College of Medicine, Zhejiang University, diagnosis, followed by gastrointestinal perforation or Qingchun Road 79, Hangzhou 310003, China. E-mail: [email protected] intestinal fistula after abdominal operation (17.5%), Received 31 January 2007; revised 9 April 2007; accepted 16 April pneumonia (17.1%), trauma (14.2%) and biliary duct or 2007; published online 17 May 2007 liver infection (10.4%). While lung (42.6%), abdomen SNPs in DEFB1 and severe sepsis Q-X Chen et al 440 (19.9%) and blood (17.5%) were the main source of age and gender as covariates confirmed the above infection. For treatment of severe sepsis, surgical inter- associations (Po0.05, OR 1.735, 95% CI 1.030–2.958 and ventions were applied for 44.5% patients. Meanwhile, Po0.05, OR 2.741, 95% CI 1.128–6.660, respectively). broad-spectrum antibiotics were widely used in all the Linkage disequilibrium (LD) analysis showed that patients. The most commonly administered antibiotics SNPs À52A/G, À44C/G and À20A/G were in strong were Tienam (Imipenem plus Cilastatin, 40.3% patients), LD. Three most common haplotypes were found in the Sulperazon (Sulbactam plus Cefoperazone, 39.8% pa- cohort (Table 2). The haplotype À20G/À44G/À52G was tients), Vancocin (Vancomycin Hydrochloride for intra more frequent in non-survivors than in survivors (17.1 vs venous, 33.2% patients) and Tazocin (Piperacillin, 28.9% 7.7%; P ¼ 0.0052, OR 2.427, 95% CI 1.307–4.507). How- patients). Antifungal agents were used in 37.4% patients. ever, the frequency of haplotype À20A/À44C/À52G in The allele frequency and genotype frequency of the control and severe sepsis group was 41.7 and 32.0%, two groups were shown in Table 1. Of note, the mutant respectively (P ¼ 0.0066, OR 0.6751, 95% CI 0.4850– allele (A allele) of SNP 1654G/A, which had been 0.8902). reported in another study,27 was not detected in the present cohorts. The genotype-allele distribution for the assayed loci followed Hardy–Weinberg equilibrium in Discussion both patients and controls (P40.05). After correcting for In this study, the À44G/C variation was associated with multiple testing, the frequency of À44G allele in patients both the susceptibility to and the fatal outcome of severe with severe sepsis was significantly higher than that in sepsis. Furthermore, the À20G/À44G/À52G haplotype controls (15.64 vs 8.60%; P ¼ 0.0049, OR 1.971, 95% CI was associated with the fatal outcome of severe sepsis, 1.227–3.166). Furthermore, in severe septic patients, the whereas the À20A/À44C/À52G haplotype showed a frequency of À44G-allele was 20.91% in non-survivors, protective role against severe sepsis. To our knowledge, while it was 9.90% in survivors (P ¼ 0.002, OR 2.406, 95% this is the first association study to investigate the genetic CI 1.368–4.232). The logistic regression model including impact of DEFB1 polymorphisms on sepsis.

Table 1 Allelic and genotype frequencies of variations À1816A/G, À390A/T, À52A/G, À44C/G, À20A/G and 1654G/A in patients with severe sepsis and controls

SNPs Severe sepsis (n ¼ 211) Control (n ¼ 157) P-valuea Survivor (n ¼ 101) Non-survivor (n ¼ 110) P-valuea N (%) N (%) N (%) N (%)

À1816A/G AA 157 (74.41) 114 (72.61) — 82 (81.19) 75 (68.18) — AG 44 (20.85) 36 (22.93) — 16 (15.84) 28 (25.46) — GG 10 (4.74) 7 (4.46) 0.89 3 (2.97) 7 (6.36) 0.087 A 358 (84.83) 264 (84.08) — 180 (89.11) 178 (80.91) — G 64 (15.17) 50 (15.92) 0.78 22 (10.89) 42 (19.09) 0.021

À390A/T AA 63 (29.86) 43 (27.39) — 36 (35.65) 27 (24.55) — AT 95 (45.02) 75 (47.77) — 43 (42.57) 52 (47.27) — TT 53 (25.12) 39 (24.84) 0.84 22 (21.78) 31 (28.18) 0.193 A 221 (52.37) 161 (51.27) — 115 (56.93) 106 (48.18) — T 201 (47.63) 153 (48.73) 0.77 87 (43.07) 114 (51.82) 0.079

À52A/G AA 62 (29.38) 39 (24.48) — 34 (33.66) 28 (25.45) — AG 98 (46.44) 75 (47.77) — 48 (47.53) 50 (45.46) — GG 51 (24.18) 43 (27.39) 0.58 19 (18.81) 32 (29.09) 0.169 A 222 (52.61) 153 (48.73) — 116 (57.42) 106 (48.18) — G 200 (47.39) 161 (51.27) 0.30 86 (42.58) 114 (51.82) 0.064

À44C/G CC 153 (72.51) 130 (82.80) — 83 (82.18) 70 (63.64) — CG 50 (23.70) 27 (17.20) — 16 (15.84) 34 (30.91) — GG 8 (3.79) 0 (0) 0.011 2 (1.98) 6 (5.45) 0.009 C 356 (84.36) 287 (91.40) — 182 (90.10) 174 (79.09) — G 66 (15.64) 27 (8.60) 0.0049 20 (9.90) 46 (20.91) 0.002

À20A/G AA 29 (13.74) 33 (21.02) — 14 (13.86) 15 (13.64) — AG 91 (43.13) 73 (46.50) — 45 (44.56) 46 (41.82) — GG 91 (43.13) 51 (32.48) 0.058 42 (41.58) 49 (44.54) 0.905 A 149 (35.31) 139 (44.27) — 73 (36.14) 76 (34.54) — G 273 (64.69) 175 (55.73) 0.014 129 (63.86) 144 (65.46) 0.760

Abbreviation: SNP, single nucleotide polymorphism. aSignificant P-value, after correction for multiple comparisons, is in bold italics. The adjusted P-value in allelic frequency and genotype frequency at the 0.05 significance level after Bonferroni correction for five SNPs is 0.005.

Genes and Immunity SNPs in DEFB1 and severe sepsis Q-X Chen et al 441 Table 2 Haplotype analysis of SNPs À20A/G, À44C/G and À52A/G in patients with severe sepsis and controls

Haplotype À20/À44/À52 Severe sepsis Control P-valuea Survivor Non-survivor P-valuea

G/C/A 0.473 0.465 0.8231 0.513 0.436 0.1067 A/C/G 0.320 0.417 0.0066 0.323 0.319 0.9369 G/G/G 0.126 0.083 0.0711 0.077 0.171 0.0052

Abbreviation: SNP, single nucleotide polymorphism. aSignificant P-value, after Bonferroni correction for multiple comparisons, is in bold italics. The adjusted P-values for the three major haplotypes formed by the three SNPs are 0.0166 for the 0.05 significance level.

Defensins are a family of secreted antimicrobial syndrome, it is unlikely that a single polymorphism peptides that directly interfere with bacterial mem- would result in a particular phenotype of sepsis. The branes. In addition, defensins have been implicated in DEFB1 gene is located on chromosome 8p22–23 which the regulation of various inflammatory and immunolo- contains a cluster of a-defensin and b-defensin genes. gical processes.5,6 In vitro, b-defensin peptides showed These defensins display important biological functions in protective action against LPS-mediated effects via abol- immune response to infection and inflammation.5,6 ishing induction of mitogen-activated protein kinase.29 DEFB4-encoding hBD-2 peptide shows high Gram- Elucidation of the pathophysiological role of the human negative antimicrobial activity, while DEFB103-encoding b-defensin 1 peptide in infectious and inflammatory hBD-3 peptide is more effective on Gram-positive diseases such as sepsis needs to be confirmed in an bacteria.5,6 Furthermore, defensins mediated adaptive additive genetic animal model. Since the members of the immune response via Toll-like receptor.5 Meanwhile, in defensin family have overlapping effect, a single gene vivo study showed a protective role of overexpressed b- knockout model will compromise the investigation of the defensin 2 against sepsis via anti-inflammation activity.31 biologic role of the target gene. Disruption of the entire Studies have shown that DEFB4 and DEFB103 as well defensin gene cluster is hampered by currently available as DEFB104 were polymorphic in copy number varia- techniques.5 Meanwhile, human defensins and their tion,20–22 and the genomic copy number of DEFB4 homologues in animals may not be completely equiva- correlated with the level of its transcript.20 Copy number lent either in structure or in function.5,6 The lack of fully polymorphism of defensin may have a crucial role in representative animal models increases the importance inflammatory disease, which has been demonstrated in of analysis of the impact of defensin gene polymor- Crohn’s disease.32 Therefore, the functional SNP À44C/ phisms on human health. G and the related haplotypes within the DEFB1 gene in Previous studies have shown that the inducible mRNA this study, as in any genetic association study, may not be level of DEFB1 appeared as interindividual variability,10 causative but may be in strong linkage disequilibrium and production of human b-defensin 1 peptide was with separate causative polymorphisms such as copy genetically regulated.30 The À44C/G variation, which number polymorphisms in other defensin genes. The was located in the 50-UTR of DEFB1, has been implicated À44C-G SNP in DEFB1 may be in synergy with these in the pathogenesis of infections with HIV and Candida.24,28 defensins or other immune components to modulate host In this study, it is noteworthy that, even with a standard immune response. Further associated studies should Bonferroni correction for multiple tests on the five SNPs address the role of defensin copy number polymor- and three haplotypes, the P-values for the association of phisms in the pathogenesis of sepsis. À44G/C and haplotype À20G/À44G/À52G as well as The current study used three different methods for haplotype À20A/À44C/À52G with related phenotypes genotyping. Since À52A/G, À44C/G and À20A/G are still significant. This strongly supports the significance variations are located very closely, sequencing is the of the present findings. Using reporter gene constructs best way for the primary measurement. Meanwhile, analysis, a very recent study demonstrated that the À44G/ PCR-allele-specific amplification (PCR-ASA) was C polymorphism did influence the promoter activity of adopted for detection of À390A/T for financial reason. DEFB1,30 which may lead to a variation in transcription or Further genetic epidemiological studies or clinical translation of DEFB1, or in stability of DEFB1 mRNA. application for assay development using the standard Therefore, individuals carrying different alleles at À44 loci method (sequencing or TaqMan assay) should be of DEFB1 may show different host immune response to implemented. infection and ultimately influence the pathophysiological On the basis of a well-matched case–control associa- course of sepsis. tion study, the present investigation identified that Previous studies have reported a protective role of DEFB1 genetic polymorphisms were not only associated À44C-G SNP in HIV infection and oral Candida with susceptibility to severe sepsis, but also related to the infection in Caucasian.24,28 However, the present study clinical course of severe sepsis. These findings suggest found that the À44C-G SNP was a risk factor for that DEFB1 is involved in immune response central to incidence and fatal outcome of severe sepsis in Han the pathophysiology of severe sepsis. Consistent with Chinese. Such conflicting results have been reported in this view were the findings indicating that defensins did another genetic association study of DEFB1 polymorph- suppress the release of proinflammatory cytokines and isms and asthma.26 The discrepancy may be contributed adhesion molecules in vivo.29,31 Thus, DEFB1 could both by ethnic distribution, target disease studied and other play a direct antimicrobial role and have a modulating unmeasured covariates. Since sepsis is a polygenic effect on the inflammatory response.

Genes and Immunity SNPs in DEFB1 and severe sepsis Q-X Chen et al 442 In summary, DEFB1 plays an important role in the host eosinophil-derived neurotoxin in host defense. Annu Rev immune response to severe sepsis. The genomic varia- Immunol 2004; 22: 181–215. tion À44G/C within DEFB1 showed a positive associa- 6 Ganz T. Defensins: antimicrodial peptides of innate immunity. tion with the development of severe sepsis. Further Nat Rev Immunol 2003; 3: 710–720. studies correlating À44G/C variation with level of 7 Bensch KW, Raida M, Magert HJ, Schulz-Knappe P, Fors- human b-defensin 1 peptide in serum and a replicate smann WG. hBD-1: a novel b-defensin from human plasma. association design with samples from other ethnicities to FEBS Lett 1995; 368: 331–335. confirm the reported findings are currently going on. 8 Goldman MJ, Anderson GM, Stolzenberg ED, Kari UP, Zasloff M, Wilson JM. Human beta-defensin-1 is a salt-sensitive antibiotic in lung that is inactivated in cystic fibrosis. Cell 1997; 88: 553–560. Subjects and methods 9 Valore EV, Park CH, Quayle AJ, Wiles KR, McCray Jr PB, After the study was approved by the local Ethics Ganz T. Human beta-defensin-1: an antimicrobial peptide of urogenital tissues. J Clin Invest 1998; 101: 1633–1642. Committee, informed written consent was obtained from 10 Fang XM, Shu Q, Chen QX, Book M, Sahl HG, Hoeft A et al. all subjects or a first degree relative of the patient. A total Differential expression of alpha- and beta-defensins in human of 211 patients with severe sepsis (123 men and 88 peripheral blood. Eur J Clin Invest 2003; 33: 82–87. women, age 60.6717.5 years) and 157 ethnic-matched 11 Joly S, Organ CC, Johnson GK, McCray Jr PB, Guthmiller JM. healthy controls (96 men and 61 women, age 58.4718.3 Correlation between beta-defensin expression and induction years) were enrolled in the study. All the subjects were profiles in gingival keratinocytes. Mol Immunol 2005; 42: Han Chinese. The diagnosis of sepsis and severe sepsis 1073–1084. met the criteria recommended by the American College 12 Sherman H, Chapnik N, Froy O. Albumin and amino acids of Chest Physicians and Society of Critical Care Medicine upregulate the expression of human beta-defensin 1. Mol Consensus Conference.33 Patients with immunosuppres- Immunol 2005; 43: 1617–1623. sion or AIDS were excluded. 13 Schaller-Bals S, Schulze A, Bals R. Increased levels of antimicrobial peptides in tracheal aspirates of newborn Genomic DNA was extracted from peripheral whole infants during infection. Am J Respir Crit Care Med 2002; 165: blood and used as a template for next analysis. The SNPs 992–995. À1816A/G (rs2741136), À52A/G (rs1799946), À44C/G 14 Hiratsuka T, Nakazato M, Ihi T, Minematsu T, Chino N, (rs1800972) and À20A/G (rs11362) were analyzed using Nakanishi T et al. Structural analysis of human beta-defensin-1 standard direct sequencing. The À390A/T (rs2738182) and its significance in urinary tract infection. Nephron 2000; 85: variation was assayed by mean of PCR-ASA. The 34–40. polymorphism 1654G/A (rs2738047) was detected using 15 Dommisch H, Acil Y, Dunsche A, Winter J, Jepsen S. TaqMan assay. The genotypes of 30 randomly selected Differential gene expression of human beta-defensins (hBD- samples were confirmed by a second method. 1, -2, -3) in inflammatory gingival diseases. Oral Microbiol All SNPs data were evaluated for Hardy–Weinberg Immunol 2005; 20: 186–190. 16 Yang D, Chertov O, Bykovskaia SN, Chen Q, Buffo MJ, equilibrium. Allele frequency and genotype frequency Shogan J et al. b-defensins: linking innate and adaptive between the defined groups were compared using immunity through dendritic and T cell CCR6. Science 1999; 2 Fisher’s exact test or w -test. Logistic regression was 286: 525–528. used to model the effect of the related genotype on 17 Sun L, Finnegan CM, Kish-Catalone T, Blumenthal R, Garzino- incidence and outcome of severe sepsis when gender and Demo P, La Terra Maggiore GM et al. Human b-defensins age were included as covariates. Linkage disequilibria suppress human immunodeficiency virus infection: potential and haplotype were analyzed using the Haploview 2.05 role in mucosal protection. J Virol 2005; 79: 14318–14329. program. The Bonferroni method was used to correct for 18 Wehkamp J, Harder J, Weichenthal M, Mueller O, Herrlinger multiple comparisons where applicable. Po0.05 was KR, Fellermann K et al. Inducible and constitutive beta- considered statistically different. defensins are differentially expressed in Crohn’s disease and ulcerative colitis. Inflamm Bowel Dis 2003; 9: 215–223. 19 Nusbaum C, Mikkelsen TS, Zody MC, Asakawa S, Taudien S, Garber M et al. DNA sequence and analysis of human Acknowledgements chromosome 8. Nature 2006; 439: 331–335. 20 Hollox EJ, Armour JA, Barber JC. Extensive normal copy This work was financially supported by National number variation of a beta-defensin antimicrobial-gene Natural Science Foundation of China (No. 30471662). cluster. Am J Hum Genet 2003; 73: 591–600. The authors disclose no financial interests that are 21 Aldred PM, Hollox EJ, Armour JA. Copy number polymorphism relevant to the research or constitute a conflict of interest. and expression level variation of the human alpha-defensin genes DEFA1 and DEFA3. Hum Mol Genet 2005; 14: 2045–2052. 22 Chen Q, Book M, Fang X, Hoeft A, Stuber F. Screening of copy number polymorphisms in human beta-defensin genes using References modified real-time quantitative PCR. J Immunol Methods 2006; 308: 231–240. 1 Cohen J. The immunopathogenesis of sepsis. Nature 2002; 420: 23 Linzmeier RM, Ganz T. Copy number polymorphisms are not 885–891. a common feature of innate immune genes. Genomics 2006; 88: 2 Hotchkiss RS, Karl IE. The pathophysiology and treatment of 122–126. sepsis. N Engl J Med 2003; 348: 138–150. 24 Braida L, Boniotto M, Pontillo A, Tovo PA, Amoroso A, 3 Lin MT, Albertson TE. Genomic polymorphisms in sepsis. Crovella S. A single-nucleotide polymorphism in the human Crit Care Med 2004; 32: 569–579. beta-defensin 1 gene is associated with HIV-1 infection in 4 Cobb JP, O’Keefe GE. Injury research in the genomic era. Italian children. AIDS 2004; 18: 1598–1600. Lancet 2004; 363: 2076–2083. 25 Leung TF, Li CY, Liu EK, Tang NL, Chan IH, Yung E et al. 5 Yang D, Biragyn A, Hoover DM, Lubkowski J, Oppenheim JJ. Asthma and atopy are associated with DEFB1 polymorphisms Multiple roles of antimicrobial defensins, cathelicidins, and in Chinese children. Genes Immun 2006; 7: 59–64.

Genes and Immunity SNPs in DEFB1 and severe sepsis Q-X Chen et al 443 26 Levy H, Raby BA, Lake S, Tantisira KG, Kwiatkowski D, 30 Sun CQ, Arnold R, Fernandez-Golarz C, Parrish AB, Lazarus R et al. Association of defensin beta-1 gene poly- Almekinder T, He J et al. Human b-defensin-1, a potential morphisms with asthma. J Allergy Clin Immunol 2005; 115: chromosome 8p tumor suppressor: control of transcription 252–258. and induction of apoptosis in renal cell carcinoma. Cancer Res 27 Matsushita I, Hasegawa K, Nakata K, Yasuda K, Tokunaga K, 2006; 66: 8542–8549. Keicho N. Genetic variants of human beta-defensin-1 and 31 Shu Q, Shi Z, Zhao ZY, Chen Z, Yao HP, Chen QX et al. chronic obstructive pulmonary disease. Biochem Biophys Res Protection against pseudomonas aeruginosa pneumonia and Commun 2002; 291: 17–22. sepsis-induced lung injury by overexpression of b-defensin 2 28 Jurevic RJ, Bai M, Chadwick RB, White TC, Dale BA. Single- in rats. Shock 2006; 26: 365–371. nucleotide polymorphisms (SNPs) in human beta-defensin 1: 32 Fellermann K, Stange DE, Schaeffeler E, Schmalzl H, high-throughput SNP assays and association with Candida Wehkamp J, Bevins CL et al. A chromosome 8 gene-cluster carriage in type I diabetics and nondiabetic controls. J Clin polymorphism with low human beta-defensin 2 gene copy Microbiol 2003; 41: 90–96. number predisposes to Crohn disease of the colon. Am J Hum 29 Motzkus D, Schulz-Maronde S, Heitland A, Schulz A, Genet 2006; 79: 439–448. Forssmann WG, Jubner M et al. The novel b-defensin DEFB123 33 Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, Cook D prevents lipopolysaccharide-mediated effects in vitro and et al. 2001SCCM/ESICM/ACCP/ATS/SIS International sepsis in vivo. FASEB J 2006; 20: E997–E1004. definitions conference. Crit Care Med 2003; 31: 1250–1256.

Genes and Immunity