Polymorphisms in the Toll-Like Receptor and the IL-23/IL-17 Pathways Were Associated with Susceptibility to Inflammatory Bowel Disease in a Danish Cohort

Background The inflammatory bowel diseases (IBD), Crohn’s disease (CD) and ulcerative colitis (UC), result from the combined effects of susceptibility genes and environmental factors. Previous studies have shown that polymorphisms in the Toll-like receptor (TLR), the apoptosis, the IL-23/IL-17 and the interferon gamma (IFNG) pathways are associated with risk of both CD and UC. Methods Using a candidate gene approach, 21 functional single nucleotide polymorphisms (SNPs) in 15 genes were assessed in a clinical homogeneous group of severely diseased ethnic Danish patients consisting of 624 patients with CD, 411 patients with UC and 795 controls. The results were analysed using logistic regression. Results The polymorphisms TLR5 (rs5744174) and IL12B (rs6887695) were associated with risk of CD, and TLR1 (rs4833095) and IL18 (rs187238) were associated with risk of both CD and UC (p<0.05). After Bonferroni correction for multiple testing, the homozygous variant genotype of TLR1 743 T>C (rs4833095) was associated with increased risk CD (OR: 3.15, 95% CI: 1.59–6.26, p = 0.02) and CD and UC combined (OR: 2.96, 95% CI: 1.64–5.32, p = 0.005). Conclusion Our results suggest that genetically determined high activity of TLR1 and TLR5 was associated with increased risk of both CD and UC and CD, respectively. This supports that the host microbial composition or environmental factors in the gut are involved in risk of IBD. Furthermore, genetically determined high activity of the IL-23/IL-17 pathway was associated with increased risk of CD and UC. Overall, our results support that genetically determined high inflammatory response was associated with increased risk of both CD and UC.


Introduction
Chronic inflammatory bowel diseases (IBD), Crohn's disease (CD) and ulcerative colitis (UC), are complex diseases that result from the interaction of numerous genetic and environmental factors [1]. Genetic association studies have identified polymorphisms in the Toll-like receptor (TLR) [2,3], the apoptosis [4][5][6], the IL-23/IL-17 [4][5][6] and the interferon gamma (IFNG) [4][5][6] pathways associated with susceptibility of CD and UC. As described below the inflammasome is connected to all of these pathways and it could be speculated that polymorphisms in the inflammasome pathway could be associated with CD or UC.
These pathways are involved in the autoimmune response to Pathogen-Associated Molecular Patterns (PAMPs) such as bacterial or viral DNA, flagellin or lipopolysaccharide (LPS). PAMPs can be bound by the membrane bound TLRs and induce inflammation by activating a number of pro-and anti-inflammatory cytokines [7]. PAMPs can also be recognized by Nodlike receptors (NLRs), which are intracellular receptors. NLRP1 and NLRP3 are members of the inflammasome protein complexes, which can activate the pro-protein caspase-1. The activation of caspase-1 can be suppressed by CARD8 [8]. Caspase-1 activates the pro-inflammatory cytokines IL-1β and IL-18, which are synthesized as pro-proteins [9]. In turn, IL-1β and IL-18 can induce the production of IL-17 [10]. NLRP1, NLRP3 and CARD8 have also been described to be involved in the apoptotic pathway, where overexpression of NLRP1 and NLRP3 stimulates apoptosis through activation of caspases [11,12] and CARD8 inhibit apoptosis by inhibiting the activation of caspase-1 [13].
In the IL-23/IL-17 pathway, IL-23 enhances the secretion of the pro-inflammatory cytokine IL-17, which in turn enhances the production of pro-inflammatory mediators such as IL-1β, IL-6, IL-8 and TNF-α [14]. IL-23 is a heterodimer composed of IL-12p40 and IL-23p19. IL-12p40 also acts as a subunit of the IL-12p35/IL-12p40 heterodimer cytokine IL-12. IL-12 can act pro-inflammatory by activating IFN-γ and anti-inflammatory by activating the anti-inflammatory cytokine IL-10 [15] and by inhibiting the pro-inflammatory cytokine IL-17 [14]. The IL-12 and IL-23 pathways also share the receptor subunit IL12R-β1, which is part of the IL12 receptor (IL12R-β1/IL12R-β2) and IL23 receptor (IL12R-β1/IL23R) heterodimer complex. The signalling through the IL12 and IL23 receptor also share intracellular mechanisms, where both receptors bind the Jak2 kinase, which in turn activates the STAT4 transcription factor. However, the resulting DNA binding STAT transcription complexes differ and involve the STAT3/ STAT4 heterodimer in IL23 signalling rather than the STAT4 homodimer in IL12 signalling [14].
IL-18, which can be activated by the inflammasome, can feedback activate the synthesis of IFN-γ. IFN-γ binds to the membrane bound IFNG receptor which is comprised of two ligandbinding IFNGR1 chains and two signal-transducing IFNGR2 chains. Binding of IFN-γ to the IFNG receptor recruits the Jak2 kinase. Jak2 initiates a kinase cascade, which ultimately activates the transcription factor STAT1 [16]. Another important activator of IFN-γ is the transcription factor T-bet (TBX21) [17].
In this study, we wanted to investigate whether functional single nucleotide polymorphisms (SNPs) in genes involved in the Toll-like receptor (TLR1, TLR5 and TIRAP), the inflammasome or apoptotic (NLRP1, NLRP3 and CARD8), the IL-23/IL-17 (IL12B, IL12RB1, IL12RB2, IL17A, IL18 and JAK2) and the IFNG (IFNGR1, IFNGR2 and TBX21) pathways were associated with risk of CD or UC in a Danish cohort of severely diseased patients. Knowing the biological effects of the studied polymorphisms allows a biological interpretation of the associations based on increased or decreased gene activity [3,18,19].

Materials and Methods Cohort
As described by Bank et al. an ethnic Danish cohort consisting of 624 patients with CD and 411 patients with UC was established [3]. The patients either received or were considered candidates to anti-tumor necrosis factor-α (anti-TNF) therapy (infliximab or adalimumab) and were thus considered to be a homogenous group of moderately to severely ill IBD patients. The control group consisted of 795 healthy blood donors recruited from Viborg, Denmark [20].
Linkage disequilibrium for the assessed SNPs in the same gene was calculated using the SNP Annotation and Proxy Search (SNAP) software [22]. In addition, linkage disequilibrium between the assessed SNPs in our study and known susceptibility loci in the same gene was calculated.

Power calculations
The Genetic Power Calculator was utilized for power analysis of discrete traits [23]. The lowest minor allele frequency (MAF) of the studied SNPs was 0.14. The 'high-risk allele frequency' was set to 0.14, the 'prevalence' was set to 0.00241 (CD), 0.00263 (UC) and 0.00504 (IBD) [24], D-prime was set to 1, type I error rate was set to 0.05 and number of cases and control:case ratio was based on Table 1. This cohort study had more than 80% chance of detecting a dominant effect with an odds ratio (OR) of 1.4 for CD, 1.5 for UC and 1.4 for IBD.

Statistical analysis
Logistic regression was used to compare genotype distributions among patients with CD, UC and IBD versus healthy controls. Odds ratio adjusted for age and gender were assessed ( Table 2). Odds ratio unadjusted (crude) and adjusted for age, gender and smoking status were included as supplementary Tables (S2 and S3 Tables). A chi-square test was used to test for deviation from Hardy-Weinberg equilibrium among the healthy controls.

Ethics Statement
The study was conducted in accordance with the Declaration of Helsinki and was approved by

Study population
Characteristics of the Danish patients with CD, UC and healthy controls are shown in Table 1.
The genotype distributions among the healthy controls deviated from Hardy-Weinberg equilibrium for TLR1 743 T>C (rs4833095) (p = 0.03). After correction for multiple testing, all SNPs studied were in Hardy-Weinberg equilibrium.

Polymorphisms associated with risk of IBD
In order to increase statistical power the analyses for CD and UC were combined (IBD). The studied polymorphisms generally showed the same direction of effect for CD and UC ( Table 2).
The biologic effects of the studied SNPs and OR for polymorphisms which have been shown to be associated with risk of CD, UC or IBD in other studies and in this study were summarized in S1 Table. Linkage disequilibrium Rs6887695 (IL12B) included in this study was in linkage disequilibrium with rs6556412 (IL12B) associated with CD in another study with r 2 = 0.96 and D' = 1.00. No linkage disequilibrium was found for the SNPs in TLR5 (rs5744174 and rs2072493), NLRP1 (rs878329 and    rs2670660), IL12B (rs3212217 and rs6887695), IL18 (rs1946518 and rs187238) or IFNGR2 (rs8134145, rs8126756 and rs17882748) (r 2 < 0.8).

Discussion
In this Danish cohort of severely ill patients, 4 functional polymorphisms in 4 genes involved in the Toll-like receptor (TLR1 and TLR5) and the IL-23/IL-17 (IL12B and IL18) pathways were found to be associated with risk of CD or UC (Fig 1).
In agreement with our study, the variant allele of the IL12B G>C (rs6887695) polymorphism has been associated with increased risk of CD, in the Welcome Trust Case Control Consortium and in a Japanese cohort [25,26]. Our study thereby confirms IL12B G>C (rs6887695) as a susceptibility locus for CD. IL12B encode the IL-12p40 subunit which is part of both the IL12 (IL-12p35/IL-12p40) and IL23 (IL-23p19/IL-12p40) heterodimers. All three subunits are up-regulated in patients with CD [27][28][29][30]. Furthermore, IL12B G>C (rs6887695) is in linkage disequilibrium with IL12B (rs6556412) which in another study has been associated with increased risk of CD [5] in agreement with the results in our study for rs6887695. In addition, the variant allele of IL12B G>C (rs6887695) has also been associated with increased risk of UC in a meta-analysis (OR: 1.11), and although the direction of association was the same in our study we were unable to confirm the result probably because of lack of power [31].
In contrast to our study, other case-control studies found opposite direction of association in other ethnic groups of patients with UC for IL18-137 G>C (rs187238) [32] or opposite direction of association in children with CD for TLR5 936 T>C (rs5744174) [33]. There is no consensus of how the stop codon polymorphism in CARD8 A>T (rs2043211) influence risk of CD [34,35].
To the best of our knowledge the polymorphism in TLR1 743 T>C (rs4833095) found to be associated with both CD and UC in our study, has not previously been described as a susceptibility polymorphism.
The biological interpretation of our results indicates that genetically determined high activity of TLR1 (rs4833095) was associated with increased risk of both CD and UC and genetically determined high activity of TLR5 (rs5744174) was associated with increased risk of CD. Both TLR1 and TLR5 can activate NFκB, a central regulator of inflammation regulating the expression of more than 150 genes including IL1B, IL12A (p35), IL12B (p40), IL17A, IL23A (p19) and IFNG [36]. The associations found in TLR1 and TLR5 further supports that the host microbial composition or environmental factors in the gut are involved in risk of IBD in agreement with other studies [2,3].
The pro-protein IL-18 can be activated by the inflammasome and IL-18 can downstream induce the production of the pro-inflammatory cytokine IL-17 [10]. In our study, the associations found for the polymorphism -137 G>C (rs187238) indicates that genetically determined low activity of IL-18 was associated with reduced risk of CD and UC. This suggests that lower activity of the IL-23/IL-17 pathway was associated with reduced risk of CD and UC.
The activity of IL-17 can also be up-regulated through the IL-23 pathway and down-regulated through the IL-12 pathway. Many proteins involved in the IL-12 pathway are also involved in the IL-23 pathway, including IL-12p40 and Jak2. Our study indicates that genetically determined low IL-12p40 level (a subunit of both IL-12 and IL-23) was associated with increased risk of CD (rs6887695), in accordance with other studies [25,26]. It could be speculated, that a lower level of the IL-12p40 subunit results in a lower activity of the IL-12 pathway, lower inactivation of the pro-inflammatory cytokine IL-17 and thus increased inflammatory response, which might explain the increased risk of CD.
Overall, the polymorphism in TLR5 (rs5744174) and IL12B (rs6887695) associated with CD, the polymorphism TLR1 (rs4833095) and IL18 (rs187238) associated with both CD and UC support that genetically determined high inflammatory response is associated with increased risk of both CD and UC. In addition, the polymorphism in IL12B (rs6887695) associated with CD and the polymorphism in IL18 (rs187238) associated with both CD and UC suggest that genetically determined high activity of the IL-23/IL-17 pathway was associated with increased risk of CD and UC. It could be speculated that IL-17 plays a key role as IL-17 can be inhibited by the IL-12 pathway and induced by the IL-23 pathway.
Studies have shown that mice deficient in the IL-23 pathway specific subunit (IL-23p19) or subunits shared with IL12 (IL-12p40 and IL12R-β1) were resistant to inflammation, whereas mice deficient in IL-12 pathway specific subunits (IL-12p35 and IL12R-β2) were more susceptible to inflammation [14]. Antibodies targeting the shared subunit IL-12p40 (Ustekinumab) have been shown to be more effective than placebo in treating patients with severe CD, especially those who failed to respond to anti-TNF therapy [37]. Based on our results it could be speculated, that a better result might be obtained by using antibodies targeting the IL-23 specific subunit IL-23p19, as our study indicates a protective effect of IL-12p40 through the IL-12 pathway.
The results in this study should be interpreted with care. TLR1 743 T>C (rs4833095) was not in Hardy-Weinberg equilibrium among the healthy controls, which is probable due to chance. When corrected adequately for multiple testing no deviation from Hardy-Weinberg equilibrium was found. In the light of the obtained P-values and the number of statistical tests performed we cannot exclude, that some of our positive findings may be due to chance. If the results were corrected for multiple testing the homozygous variant genotype of TLR1 743 T>C (rs4833095) was associated with increased risk of CD (p = 0.02) and marginally with UC (p = 0.07) and the combined homozygous and the heterozygous variant genotypes of IL12B G>C (rs6887695) were marginally associated with increased risk of CD (p = 0.08). We successfully assessed 18 polymorphisms and assuming a 5% acceptance level, one polymorphism would be expected to be associated with susceptibility by pure chance. In this study 5 polymorphisms were found to be associated with susceptibility and most of the found associations were biologically plausible. We cannot exclude that associations were not identified due to insufficient statistical power. The results should therefore be replicated in independent cohorts. A major strength was that this clinically homogeneous and well-characterised cohort was rather large including 1035 patients with IBD and 795 healthy controls. All the patients were considered for anti-TNF treatment and were therefore considered to have a severe disease course. Genetic determinants may be expected to be strong among severely ill cases [38].
In conclusion, 4 functional SNPs in 4 genes involved in regulation of inflammation were found to be associated with susceptibility of severe CD or UC. The SNP in TLR1 743 (rs4833095) has not previously been reported as susceptibility polymorphisms of both CD and UC (Fig 1). The IL12B G>C (rs6887695) polymorphism associated with risk of CD in our study has also been associated with risk of CD in other studies and should therefore be considered a confirmed risk locus. Our results suggest that genetically determined high activity of TLR1 and TLR5 was associated with increased risk of CD and UC. This supports that the host microbial composition or environmental factors in the gut are involved in risk of IBD. Furthermore, genetically determined high activity of the IL-23/IL-17 pathway was associated with increased risk of CD and UC. Overall, our results support that genetically determined high inflammatory response was associated with increased risk of both CD and UC.
Supporting Information S1 Table. The biologic effect of the studied single nucleotide polymorphism (SNP) and odds ratios (OR) for polymorphisms which have been shown to be associated with risk of Crohn's disease (CD), ulcerative colitis (UC) or inflammatory bowel disease (IBD) in other studies and in this study. (DOC) S2 Table. Odds ratios (OR) (unadjusted) for genotypes studied among healthy controls and patients with Crohns disease (CD), ulcerative colitis (UC) and combined inflammatory bowel disease (IBD). (DOC) S3 Table. Odds ratios (OR) (adjusted for age, sex and smoking status) for genotypes studied among healthy controls and patients with Crohns disease (CD), ulcerative colitis (UC) and combined inflammatory bowel disease (IBD). (DOC)