key: cord-0002186-u6dfp6gf authors: Toubiana, Julie; Courtine, Emilie; Tores, Frederic; Asfar, Pierre; Daubin, Cédric; Rousseau, Christophe; Ouaaz, Fatah; Marin, Nathalie; Cariou, Alain; Chiche, Jean-Daniel; Mira, Jean-Paul title: Association of REL polymorphisms and outcome of patients with septic shock date: 2016-04-08 journal: Ann Intensive Care DOI: 10.1186/s13613-016-0130-z sha: 47a7aa4af350e2de25a52dd78fbadb63a0da44f1 doc_id: 2186 cord_uid: u6dfp6gf BACKGROUND: cRel, a subunit of NF-κB, is implicated in the inflammatory response observed in autoimmune disease. Hence, knocked-out mice for cRel had a significantly higher mortality, providing new and important functions of cRel in the physiopathology of septic shock. Whether genetic variants in the human REL gene are associated with severity of septic shock is unknown. METHODS: We genotyped a population of 1040 ICU patients with septic shock and 855 ICU controls for two known polymorphisms of REL; REL rs842647 and REL rs13031237. Outcome of patients according to the presence of REL variant alleles was compared. RESULTS: The distribution of REL variant alleles was not significantly different between patients and controls. Among the septic shock group, REL rs13031237*T minor allele was not associated with worse outcome. In contrast, REL rs842647*G minor allele was significantly associated with more multi-organ failure and early death [OR 1.4; 95 % CI (1.02–1.8)]. CONCLUSION: In a large ICU population, we report a significant clinical association between a variation in the human REL gene and severity and mortality of septic shock, suggesting for the first time a new insight into the role of cRel in response to infection in humans. Despite continued efforts and significant advances in critical care medicine, septic shock remains a significant health problem with a mortality rate around 35 % [1, 2] . Septic shock is defined as sepsis accompanied by cardiovascular failure that is often a part of multiple organ dysfunction syndrome (MODS) [3] . Thus, septic shock represents an extreme manifestation of the host inflammatory response to severe infection. Transcription of inflammatory mediators such as cytokines, chemokines, adhesion molecules and reactive oxygen species is strongly activated by the transcriptional factor NF-κB and contributes to the development of MODS [4, 5] . NF-κB is an ubiquitous family of inducible dimeric homodimer or heterodimer transcriptional factors composed by five members: Rel (c-Rel), RelA (p65), RelB, NF-κB1 (p50/p105) and NF-κB2 (p52/p100) [6] . The role of RelA in severe infection is well established, as it is highly recruited to the promoter of pro-inflammatory genes in non-survivors of septic shock [7, 8] . The cRel subunit was the least studied member of the Rel family, but seems to have also a critical role in the antimicrobial host defense. Indeed, in vivo studies revealed that cRel is required for macrophage activation [9, 10] , adaptive immunity [11] and the control of lymphocyte proliferation [12, 13] . cRel is also a key regulator of numerous cytokines: IL-2, IL-3, IL-4, IL-6, IL-10, IL-13, IL-15, IL-21, IFN-γ, IFN-β, IFN-λ, MIP1-α and GM-CSF [9, [14] [15] [16] [17] [18] [19] . More recently, in a murine model of polymicrobial sepsis, Rel deficiency led to an increased mortality, an enhanced systemic inflammatory response and a sustained depletion of spleen lymphoid dendritic cells [9] [10] [11] [12] 20] . Furthermore, whole blood transcriptomics showed that cRel targets inflammatory and survival genes during sepsis [13, 20] . Moreover, genetic variants within the REL locus have been associated with inflammatory diseases or autoimmunity in Europeans [21] . Even if no study reports the importance of cRel in human sepsis, these elements highlight the potential importance of cRel for NF-κB targeted-immunomodulation in severe infections. Recently, the role of genetic factors influencing the susceptibility to or the severity of severe sepsis has been extensively studied. Several single nucleotide polymorphisms (SNPs) have been characterized in genes of NF-κB pathway proteins. For instance, SNPs in TLRs [22] [23] [24] , TIRAP [25] , IRAK1 [26, 27] , IκB [28] and NF-κB inducing kinase (NIK) [29] genes have been associated with severity of sepsis. However, the association between genetic variants in NF-κB subunits and severe infections has been poorly reported. Hence, the present study aims to test the hypothesis of an association between clinically significant REL genetic variants and severity of septic shock in a large cohort of well-defined intensive care unit (ICU) patients. This study was conducted prospectively in three medical ICUs in France. All three ICUs share similar severe sepsis management protocols based on international guidelines from the Surviving Sepsis Campaign for management of severe sepsis/septic shock [30] . The septic shock group was defined by usual criteria [31] . Briefly, patients were eligible for inclusion into the septic shock group (SS) if they had, within their stay in ICU, a clinical evidence of infection with two of four SIRS criteria (fever (>38.2 °C) or hypothermia (<36 °C); tachycardia (>90 beats/min); tachypnea (>20 breaths/min) or need of mechanical ventilation; white cell count >12 × 10 9 /L) and if, after an adequate fluid resuscitation, they required vasopressor infusion (norepinephrine, epinephrine or dopamine >8 μg/kg/min) to maintain a mean arterial pressure higher than 60 mmHg. Exclusion criteria included comorbidities highly associated with death in SS [32] : age above 85 years, cardiac failure (NYHA class III or IV), liver insufficiency (child C), bone marrow aplasia or leucopenia not related to septic shock (white blood cell count <0.50 × 10 9 /L), immunosuppression (HIV, current immunosuppressive therapy including steroids with equivalent prednisone >0.5 mg/kg per day) or ongoing cancer with undergoing treatment. The control group (C) was composed of patients hospitalized simultaneously in the three ICUs for other reasons than infection and who did not develop sepsis nor required any inotropic or vasopressor agents during their ICU stay. Similar exclusion criteria were used for the control and the SS groups. Patients were followed up throughout their ICU stay, and clinical and biological characteristics were prospectively collected: age, gender, SAPSII score and previous medical history of severe infection requiring hospitalization. For the SS group, characteristics of current infectious episode were also collected: primary sites of infection, infection-related microorganisms, development of multi-organ dysfunction syndrome (MODS) (defined as the presence of more than two organ system failures occurring simultaneously ICU stay) [3] , mechanical ventilation requirement estimated by ventilator-free day (VFD: time without mechanical ventilation within the ICU period censured to 28 days) [33] and ICU mortality. To minimize confounding factors due to ethnical differences, all patients selected in the study were Caucasians and had European origins. The Institutional Review Board of Cochin Hospital, Paris, France, approved the study, and informed consents have been obtained from the patients or their relatives. Two previously described SNP have been analyzed. The SNP rs842647 is a A → G transition located in the second intron of REL gene on chromosome 2 (chromosomic location 61119471). The SNP rs13031237 is a G → T transition located in the fourth intron of REL gene (chromosomic location 61136129). All genetic analyses were performed blinded from the clinical data. Genomic DNA was extracted from mononuclear cells using MagNA Pure Compact automate (Roche Diagnostics ® ). DNA extracts were then quantified and stored in code-barr tubes (2DCYPHER, ABgene ® ) to maintain anonymous status of the patients all along the study. Real-time PCR allelic discrimination assays were realized by TaqMan ® method on Abi 7900 (Applied Biosystems ® ). Probe and primer combinations were designed to discriminate the two REL SNPs (rs842647 and rs13031237). Quality control for genotyping was performed by automatic sequencing 12 patients carrying the different REL genotypes in order to confirm allelic discrimination results and also by re-genotyping 20 % of the entire cohort. All DNA samples showing discrepancy between the two analyses were definitively sequenced (n = 4). All data were analyzed by SPSS v11.5 and "R" v3 softwares. Both SNPs were tested for Hardy-Weinberg disequilibrium to check for stratification. In order to calculate the Šidák multiple testing correction, we first evaluate the effective number of independent tests (called Meff ) in the analysis by using the methodology proposed by Li and Ji [34] . This method aims to prevent from overcorrection due to possible linkage disequilibrium (LD) between the SNPs. Power calculation has been based on the frequency of the variant allele in the control population as proposed by Hattersley et al. [35] . Hence, for an incidence of the variant allele of 11 % in the control population and a power at 90 %, a 50 % increase in the case population with a type I error of 5 %, 886 individuals in each group appear to be sufficient to detect genetic susceptibility to SS (http://www.stat.ubc.ca/~rollin/stats/ssize/b2.html). For the second study assessing the prognostic value of the variant alleles in SS group, given the frequency of the variant genotype, and an expecting mortality rate at 35-40 % in the SS subgroup group, we considered that 870 SS patients were sufficient for a power at 0.90 (type I error at 0.05) to identify a 30 % difference in genotype frequency. Descriptive results of continuous variables were expressed as median and interquartile range reflecting population distribution. Variables were tested with Chisquare test for categorical data (sex, multi-organ failure, primary sites of infection, microorganisms, genotypes) and with Mann-Whitney U test for numerical data (age, SAPSII, VFD). A multivariate logistic regression model was used to determine the respective role of REL genotypes for susceptibility to SS and to ICU mortality. Confounding factors with a p value <0.05 were included in this model. Continuous variables were included without any transformation, and genotypes were considered as a factor (dichotomous unordered variate) to avoid the implicit dose effect when coding the genotypes 0, 1 and 2 according to the number of mutated alleles carried. Results were expressed as odds ratio (OR) and 95 % confidence interval (CI), and variable with p value <0.05 was defined as statistically significant. The total enrolled Caucasian population was composed by 1040 septic shock patients (SS) and 855 controls (C). In the C group, enrolled ICU patients were admitted for various non-infectious reasons (metabolic: 39 %, neurological: 29 %, respiratory: 22 %, cardiovascular: 6 % and surgical: 4 %) and did not develop severe sepsis and did not require vasopressor infusion during their ICU stay. C patients were younger than SS patients (46 vs . 66 years, respectively, p = 0.001), and females were more represented in C group (45 and 39 %, for C and SS, respectively, p = 0.006). Mortality rate and occurrence of MODS in the C group were 3 and 1 %, respectively. All SS patients received norepinephrine or epinephrine as first vasopressor. The main site of infection was the lung (58 %); microorganisms were identified in 74 % of the cases, mainly Gram-positive bacteria. Median SAPSII value of 54 and high percentage of patients with multiple organ dysfunctions (61 %) underlined the severity of the septic shock population. The ICU mortality rate of the SS group was 30 %. Hardy-Weinberg proportions were comparable to expected percentages regarding REL variants: REL rs842647 (p = 0.9) and REL rs13031237 (p = 0.7) in favor of homogeneity of population ethnicity. To determine whether REL rs842647 and rs13031237 SNPs were associated with septic shock susceptibility, genotype frequencies were determined for SS and C patients. As reported in Table 1 , no significant difference was found between the two groups. Moreover, these incidences were similar to those reported in the HapMap database-reported genotype distribution for European population (http:// www.ncbi.nhm.nih.gov/projects/SNP/snp_viewTable. cgi?pop=1409). Among the SS patients, general clinical characteristics were not significantly different between patients carrying REL minor allele and patients homozygous for major allele on both REL-analyzed loci ( Table 2 ). In order to study the link between REL SNPs and septic shock severity, we compared acute respiratory distress syndrome (ARDS) and MODS frequencies, and VFD value between patients carrying REL rs842647*G and rs13031237*T minor alleles and in those homozygous for the major alleles. As given in Table 2 , VFD values were Table 2 ]. For multiple testing correction, we calculated a Meff of 1.75 [34] leading to a corrected p value of 0.06. A similar trend was observed for ARDS (p = 0.08, Table 2 ). In SS group, mortality was not significantly different between patients carrying REL rs13031237*T minor allele and homozygous for the major allele of this SNP (30.2 vs. 29.8 %, for minor and major alleles, respectively, p = 0.9). In contrast, the presence of the rs842647*G minor allele was significantly associated with a higher mortality rate (33 vs. 27 %, for minor and major alleles, respectively, p = 0.03) (Fig. 1) The present study showed that septic shock patients carrying the rs842647*G minor allele had an over risk of MODS and mortality. In contrast, no association was found between the REL rs13031237*T allele and the severity of septic shock. This study was the first to investigate the importance of two polymorphisms within REL gene in a large European population of septic shock patients. Several human studies have suggested that these variants may have an effect on the inflammatory balance, as they have been associated with inflammatory and autoimmune diseases. Indeed, the intronic rs13031237 SNP in the REL gene was associated with susceptibility to rheumatoid arthritis [36] [37] [38] and psoriasis [39] . The intronic rs842647 SNP was linked to a higher risk of Crohn's disease, ulcerative colitis [40] and celiac disease [41, 42] and primary sclerosing cholangitis [43] . Genome-wide studies have also found that REL locus was associated with psoriasis [37, 44] , rheumatoid arthritis [37] , ulcerative colitis [45, 46] and Hodgkin's lymphoma [47] . However, functional and structural effects of these polymorphisms are still unknown and need to be investigated. Given that variant alleles are located on an intronic site, it is possible that these polymorphisms affect transcriptional efficiency of REL gene or these variants may be in strong linkage disequilibrium with a variant inside a neighbor gene. The higher rate of mortality observed in SS patients carrying rs842647*G might be linked to a higher inflammatory state, as they also developed more frequently MODS. They also tend to have more ARDS and lower VFDs however not significant, but this is most likely underpowered, as VFDs are not normally distributed. Mortality in septic shock was partially related to hyperactivation of NF-κB [7, 48] . In this setting, previous genetic studies on several gain of function SNPs in genes of receptors and signaling molecules upstream of NF-κB, such as TLR1 and IRAK1 [22, 26] , showed a significant association with severity of sepsis. These genetic factors might unbalance the fine-tune regulation toward a hyperinflammatory deleterious state. However, the exact role of cRel on inflammatory processes is less understood in humans. Recent studies have shown that cRel could be involved in autoimmunity, such as inflammatory arthritis [49] and autoimmune encephalomyelitis [50] . More recently, cRel was shown to have a key role in antimicrobial defense processes. Rel−/− mice are more susceptible to Leishmania major [9] or Toxoplasma gondii infections [10] , to viral infection by Influenza virus [11] , to bacterial infection by Listeria monocytogenes [12] and to polymicrobial sepsis [20] . cRel is probably important in pro-/ anti-inflammatory balance as Rel−/− mice seemed to have an enhanced inflammatory response [20] . The study design quality is important for a right interpretation of genetic association studies [35] . We tried to follow closely these quality criteria. First, it is important to select a SNP of a protein involved in the physiopathology of the disease. As already mentioned, REL seems to be an interesting gene to study because NF-κB plays a central role in physiopathology of sepsis, and recent studies show the importance of cRel in this context. However, one important limit of our study is the absence of data regarding the functional effect of these two SNPs. Functional data are needed to improve our understanding of how rs842647*G variant of REL is related to sepsis severity. Second, the population homogeneity has been controlled by limiting the study on European patients without severe comorbidities. The third item is probably one of the more controversial in the sepsis field: choice of a clearly defined phenotype to avoid confusion factors. Thus, we have selected only patients with septic shock whose diagnosis and treatment are standardized [3] , and these patients had no major comorbidity or immunosuppressive treatment, severe autoimmune diseases in particular, that could have been confounding the results. However, it is impossible to rule out effects of confounding factors or gene-environment interactions in our results, and septic shock is heterogeneous with regard to the source of infection. Sample size is essential for statistics quality in association study but is difficult to achieve in pure septic shock population. At our knowledge, our cohort is one of the largest ever published populations in this topic and is large enough to diminish type I error. It is important to consider that after correction for multiple testing, our result only reached near significance. Šidák/ Bonferroni correction assumes, however, that markers are independent, whereas the SNPs studied here are in LD and are therefore not truly independent from each other. As a result, though we tried to take into account LD by the calculation of Meff, the adjustment is likely to overcorrect in this case. We therefore consider that our preliminary results would need a validation in independent cohorts. Finally, this genetic association study is limited to one gene. Genome-wide association studies (GWAS) are now discovering new unsuspected genes that might have an impact on sepsis outcome [51] . The association between rs842647*G allele and severity of septic shock brings a new perspective on the role of cRel subunit of NF-κB in severe infections in humans. Better understanding of the genetic effects of NF-κBdependent inflammatory pathways is essential for further research on modulation of NF-κB activity by specific inhibitors, such as small molecule inhibitors of cRel, as an adjuvant treatment for sepsis [52, 53] . Further studies are needed to investigate the functional role of this REL polymorphism on the inflammatory processes observed in sepsis and to validate these encouraging results in independent cohort. Abbreviations MODS: multiple organ dysfunction syndrome; IL: interleukin; IFN: interferon; SNP: single nucleotide polymorphisms; ICU: intensive care unit; SS: septic shock group; C: control group; VFD: ventilator-free day; ARDS: acute respiratory distress syndrome; LD: linkage disequilibrium; TLR: Toll-like receptor. Severe sepsis and septic shock Epidemiology of severe sepsis Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock NF-kappa B activation as a pathological mechanism of septic shock and inflammation The two NF-kappaB activation pathways and their role in innate and adaptive immunity Characterization of elements determining the dimerization properties of RelB and p50 Predictive value of nuclear factor kappaB activity and plasma cytokine levels in patients with sepsis Role of NFkappaB in the mortality of sepsis The Rel subunit of NF-kappaB-like transcription factors is a positive and negative regulator of macrophage gene expression: distinct roles for Rel in different macrophage populations Cutting edge: identification of c-Rel-dependent and -independent pathways of IL-12 production during infectious and inflammatory stimuli Mice lacking the transcription factor subunit Rel can clear an influenza infection and have functional anti-viral cytotoxic T cells but do not develop an optimal antibody response Multiple hemopoietic defects and lymphoid hyperplasia in mice lacking the transcriptional activation domain of the c-Rel protein Genomewide analysis of gene expression in T cells to identify targets of the NF-kappa B transcription factor c-Rel The roles of c-rel and interleukin-2 in tolerance: a molecular explanation of self-nonself discrimination IL-6 rescues the hyporesponsiveness of c-Rel deficient B cells independent of Bcl-xL, Mcl-1, and Bcl-2 Interferon regulatory factor-4 activates IL-2 and IL-4 promoters in cooperation with c-Rel Regulation of the IL-21 gene by the NF-kappaB transcription factor c-Rel Regulation of IFN-lambda1 promoter activity (IFN-lambda1/IL-29) in human airway epithelial cells Nuclear factor kappaB subunits RelB and cRel negatively regulate Toll-like receptor 3-mediated beta-interferon production via induction of transcriptional repressor protein YY1 Critical role of cRel subunit of NF-kappaB in sepsis survival The c-Rel transcription factor in development and disease Toll-like receptor 1 polymorphisms affect innate immune responses and outcomes in sepsis A common dominant TLR5 stop codon polymorphism abolishes flagellin signaling and is associated with susceptibility to legionnaires' disease Relevance of mutations in the TLR4 receptor in patients with gram-negative septic shock A Mal functional variant is associated with protection against invasive pneumococcal disease, bacteremia, malaria and tuberculosis IRAK1 functional genetic variant affects severity of septic shock Variant IRAK-1 haplotype is associated with increased nuclear factor-kappaB activation and worse outcomes in sepsis IkappaB genetic polymorphisms and invasive pneumococcal disease A single nucleotide polymorphism in NF-kappaB inducing kinase is associated with mortality in septic shock Surviving Sepsis Campaign guidelines for management of severe sepsis and septic shock Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis Incidence, risk factors, and outcome of severe sepsis and septic shock in adults: a multicenter prospective study in intensive care units. French ICU Group for Severe Sepsis Statistical evaluation of ventilator-free days as an efficacy measure in clinical trials of treatments for acute respiratory distress syndrome Adjusting multiple testing in multilocus analyses using the eigenvalues of a correlation matrix What makes a good genetic association study? Confirmation of association of the REL locus with rheumatoid arthritis susceptibility in the UK population encoding a member of the NF-kappaB family of transcription factors, is a newly defined risk locus for rheumatoid arthritis Genome-wide association study meta-analysis identifies seven new rheumatoid arthritis risk loci An investigation of rheumatoid arthritis loci in patients with early-onset psoriasis validates association of the REL gene Genetic analysis of innate immunity in Crohn's disease and ulcerative colitis identifies two susceptibility loci harboring CARD9 and IL18RAP Coeliac disease-associated risk variants in TNFAIP3 and REL implicate altered NF-kappaB signalling Improving the estimation of celiac disease sibling risk by non-HLA genes Three ulcerative colitis susceptibility loci are associated with primary sclerosing cholangitis and indicate a role for IL2, REL, and CARD9 A genome-wide association study identifies new psoriasis susceptibility loci and an interaction between HLA-C and ERAP1 Genome-wide association identifies multiple ulcerative colitis susceptibility loci Ulcerative colitis-risk loci on chromosomes 1p36 and 12q15 found by genome-wide association study A genome-wide association study of Hodgkin's lymphoma identifies new susceptibility loci at 2p16.1 (REL), 8q24.21 and 10p14 (GATA3) Nuclear factor-kappaB activation in peripheral blood mononuclear cells in children with sepsis Distinct roles for the NF-kappaB1 (p50) and c-Rel transcription factors in inflammatory arthritis Critical roles of c-Rel in autoimmune inflammation and helper T cell differentiation Genome-wide association study of survival from sepsis due to pneumonia: an observational cohort study The IKK NF-kappa B system: a treasure trove for drug development A small-molecule c-Rel inhibitor reduces alloactivation of T cells without compromising antitumor activity This study was supported by national grant from the Ministry of Health (PHRC 2003), grant from CARISMA (Cochin Association for Research in inflammation, Sepsis and Molecular Advances), grant from SRLF (Société de Réanimation de Langue Française) and grant from SFAR (Société Française Anesthésie Réanimation. These funding organizations played no role in the design, execution and publication of the study. JT and JPM participated in the design of the study and the interpretation of the data. JDC, AC, PA, NM and CD participated in the recruitment of patients; EC and JT in the redaction and revision of the manuscript; and FT checked the statistics. CR and FO performed the genotyping experiments. All the authors reviewed the manuscript. All authors read and approved the final manuscript. The authors declare that they have no competing interests.