Introduction

Inflammatory bowel diseases (IBD) are common chronic relapsing inflammation disorders of the gastrointestinal tract with two main pathological entities: Crohn’s disease (CD) and ulcerative colitis (UC). Though the pathogenesis of IBD remains elusive, numerous epidemiological investigations strongly suggested that genetic factors may play a prominent role in the development of IBD [1].

With the help of genome-wide association techniques, our knowledge of the genetic basis of IBD advanced greatly [2]. The discovery of the first CD susceptibility gene nucleotide-binding oligomerization domain containing 2 (NOD2), also known as caspase recruitment domain family member 15 (CARD15), created a new era in IBD research. Since then, more and more single nucleotide polymorphisms (SNPs) have been identified to be associated with IBD. These included rs11209026 and rs11805303 in IL23R (which encodes a crucial subunit for the interleukin 23 receptor), rs2241880 in ATG16L1 (which encodes autophagy-related 16-like 1 gene), rs2542151 in PTPN2 (which encodes protein tyrosine phosphatase N2), rs10761659 in 10q21, rs13361189 and rs4958847 in IRGM (which encodes immunity-related GTPase family M), rs1050152 in OCTN1 (which encodes organic cation transporter gene cluster 1), and rs2631367 in OCTN2 (which encodes organic cation transporter gene cluster 2) [3–12].

Approximately 3.6 million people in the west are affected by IBD every year. In the past, few cases were reported in Asia. But in these years, the incidence of IBD increased continuously. Genetic studies of IBD carried out in China were limited and the major part of the genetic background of IBD in the Chinese population remains unclear [13, 14]. Therefore, we aimed to evaluate the contribution of those confirmed IBD susceptibility loci mentioned above to the Chinese population and also examine their possible phenotype–genotype relationships.

Materials and methods

Patient specimens

A total of 97 IBD patients (48 CD and 49 UC) of Chinese Han ethnicity from the departments of gastroenterology and general surgery and 50 control samples from healthy persons under routine health screening in the outpatient clinics of Nanfang Hospital, Guangzhou, China were consecutively recruited from January 2007 to January 2009.

All patients were unrelated and their medical histories were obtained from clinical records. The diagnosis of IBD was conducted by senior physicians based on standard clinical, endoscopic, radiologic, and histological criteria. Severity of IBD was classified according to the criteria of the modified Williams DAI standard and the Chinese Association of Digestive Diseases. The institutional ethics committee approved all protocols and all enrolled subjects gave their informed consents.

Genotyping and sequencing analysis

Genotypic DNA was obtained from peripheral blood leukocyte of CD, UC, and normal controls. DNA fragments were amplified by polymerase chain reaction (PCR) using specific primers (Table 1). The PCR conditions were as follows: initial denaturation at 94 °C for 4 min, followed by 30 cycles of denaturing at 94 °C for 45 s, annealing at the temperature (Table 1) for 45 s, extension at 72 °C for 1 min, and final incubation at 72 °C for 10 min. The PCR productions were confirmed by purification. DNA sequencing for P268S, R702W, G908R, 3020insC, rs11209026, rs11805303, rs2241880, rs1050152, and rs2631367 was carried out using an ABI377 automated sequencer by Invitrogen. Gene sequences from UC, CD, and normal controls were aligned with sequence in the GenBank (http://www.ncbi.nlm.nih.gov/BLAST). If a mutation was found, the corresponding exon will be reamplified and resequenced using both upstream and downstream sequencing primers for confirmation. Restriction fragment length polymorphisms (RFLP) assay was used for rs2542151, rs10761659, rs13361189, and rs4958847. The P268S variant was also confirmed by RFLP.

Table 1 PCR primers and conditions
Full size table

Statistical analysis

All analyses were performed using the software of SPSS13.0. Case–control analysis was performed by Fisher’s exact. Odds ratios (OR) were calculated with the corresponding chi square distribution test and 95 % confidence intervals (95 % CI). P < 0.05 was considered significant. The genotype frequencies for each variant were tested for deviation from the Hardy–Weinberg equilibrium by means of the chi square test.

Results

Identification of mutations associated with Chinese IBD patients

All the genotypic and allelic distributions analyzed in this study were in accordance with the Hardy–Weinberg equilibrium. The involvements of NOD2/CARD15 mutations in the etiology of CD were commonly seen in the west. To examine their possible associations with IBD in Chinese patients, all 12 exons of the NOD2/CARD15 gene were amplified by PCR. Then, the PCR products were subjected for both direct DNA sequence and RFLP analysis. However, no significant associations between three common SNPs (Arg702Trp, Gly908Arg, and Leu1007fsinsC) and CD were detected in our cohort (data not show). A novel P268S mutation was observed in 5 out of 48 CD patients, with codon changed from CCC to UCC (Table 2; Figs. 1 and 3a, b). This novel P268S mutation was only found in CD patients, but not in UC patients and the healthy control group (Table 2; χ 2 = 5.49, P < 0.05).

Table 2 Genotype distributions and allele frequencies of gene variants in IBD patients compared with healthy controls
Full size table
Fig. 1
figure 1

Forward sequence atlas of the PCR product which contains the P268S variant (indicated by arrows). a Wild-type CC, b heterozygote TC, c homozygote TT

Full size image

Another mutation, IL23R-rs11805303 (C → T), was detected in 32 out of 49 UC compared with 42 out of 50 in controls. Carriage of the T allele of IL23R-rs11805303 was found to have a decreased risk of UC (P = 0.010, under the genotypic model; P = 0.004, under the allelic model). The OR for the protective T allele was 0.27 (95 % CI, 0.10–0.74). Alternatively, the wild-type CC genotype may be considered as the risk genotype with an OR of 3.70 (Fig. 2). Moreover, the presence of a G allele at PTPN2-rs2542151 was significantly associated with an increased risk of UC (P = 0.001, under the genotypic model; P = 0.003, OR GG/TG vs. TT, 3.92; 95 % CI, 1.76–8.75, under the allelic model) (Fig. 3c, d). Other previously identified IBD-associated SNPs were not detected in the Chinese population (Table 2).

Fig. 2
figure 2

Forward sequence atlas of the PCR product which contains the rs11805303 variant (indicated by arrows). a Wild-type CC, b heterozygote CT, c homozygote TT

Full size image
Fig. 3
figure 3

Electrophoresis of PCR products (a, c) and products incised by enzymes (b, d). Lane M1 DNA marker I, lane M 1.5-kb DNA marker (50 bp ladder). a, b Analysis of P268S. b Lanes1–3 heterozygote CT, lane 4 homozygote TT, lane 5 wild-type CC. c, d Analysis of rs2542151. d Lane 2 wild-type TT, lane 6 homozygote GG. The other lanes: heterozygote TG

Full size image

Identification of association between genotype and clinical features

To examine whether the above-identified SNPs have any relations with clinical features, the occurrence of each SNP was compared with age at onset, gender, lesion location, disease severity, and so on. Intriguingly, the lesions of the five CD patients with P268S mutation were all located to the terminal ileum (P = 0.003) and accompanied with enteric cavity narrowing (P = 0.007). Four of these five patients were under 20 years of age at onset (P = 0.028), suggesting that P268S in Chinese CD patients might correlate with younger disease occurrence. However, P268S carriers were not related to gender (P = 0.521) or disease severity (P = 0.289). No significant associations between phenotype and genotype of both rs11805303 and rs2542151 were observed in Chinese patients with UC (Table 3).

Table 3 Relations between three variants and IBD clinical features
Full size table

Discussion

Identification of IBD susceptibility genes will provide key insights into pathogenic mechanisms and be beneficial to patients by proper interventions. During the past decades, a number of genes have been found to contribute to IBD susceptibility in White populations [2–12]. Here, we selected a total of eight genes and assessed the contribution of these established risk variants to the genetic susceptibility to IBD in a Chinese cohort.

Among all the IBD susceptibility genes, NOD2/CARD15 is the most well studied and has been shown to give rise to CD with a strong dissimilarity among different ethnicities [5]. Here, we genotyped the coding region of NOD2/CARD15 but failed to replicate associations of CD with the three SNPs (G908R, 1007fsCins, and R702W) which were commonly seen in White populations [10]. Our finding provides additional support for the limited role of these three SNPs in Asian CD patients [8, 14, 15].

Instead, our study confirmed that P268S was involved in the predisposition to CD in the Chinese population. Similar presence of P268S was also observed in Chinese Tu, Indian, and Pakistani populations, indicating it to be a common risk factor of CD among Asians [16]. Genetic heterogeneity of P268S exists among different ethnic groups, too. For instance, a North Indian UC study even observed significant association for P268S [17]. Moreover, P268S usually occurred in phase with the three SNPs and was regarded as a background polymorphism of NOD2/CARD15 in European ancestry [9]. P268S was also found to be in linkage disequilibrium (LD) with other mutations, such as IVS8-158(JW1) which acted as a potential “risk-conferring” variant in Jewish and Asian patients with CD [10, 11]. This raises the question whether P268S in LD with additional NOD2/CARD15 polymorphisms contributes to CD susceptibility in the Chinese Han population. Further studies are needed to explore the possible yet unidentified Chinese-specific variants in NOD2/CARD15 in a larger cohort.

CD is known to have a bimodal age distribution of disease onset, and recent studies revealed that P268S was associated with age onset of CD [18]. Our confirmation of CD patients carrying P268S polymorphism with a tendency of young disease occurrence adds to this predisposition. Furthermore, it has been recognized that patients with a younger age at onset or ileal involvement usually have a stricturing phenotype more frequently than other patients [7, 19]. And in our present study, P268S also conferred susceptibility to ileal lesions and enteric cavity narrowing. We, therefore, speculate that P268S may be used as a marker to represent specific clinical subgroups in Chinese CD patients. NOD2/CARD15 mutations may result in an impaired activation of NF-κB by altering the recognition of the bacterial lipopolysaccharide [5]. But a recent study revealed that P286S was not involved in the regulation of NF-κB signaling pathway [20]. As the exact mechanism by which P268S contributes to the development of CD remains unclear, further functional studies of P268S should be conducted to achieve a better understanding of NOD2-mediated CD pathogenesis.

Similar to previous reports, no statistically significant associations between UC and NOD2/CARD15 mutations were detected, extending the hypothesis that NOD2/CARD15 mutations which lead to defects in bacterial antigen processing are likely less important in UC than in CD [21].

CD and UC are thought to be distinct but related disorders. Unlike NOD2/CARD15, which is unique to CD, some susceptibility genes are common to both types of IBD [22–24]. IL-23R and PTPN2 are two of them.

IL-23 is believed to promote the differentiation of native CD4+ T cells into Th17 cells and production of several inflammatory mediators from both T cell and non-T cell sources. This cytokine can also restrain the activity of regulatory T cells. Its corresponding ligand IL-23R is thus of great importance for intestinal inflammation. IL23R-rs11209026 was found to be a risk factor for both CD and UC in the non-Jewish White population, and a magnitude association with UC was observed in a Jewish White cohort [4]. However, this SNP appears to be absent or very rare in Asians, including Japanese, Korean, and Indian [25, 26]. Similarly, we also failed to replicate the association of rs11209026 with neither UC nor CD in the Chinese population. Compared with other IL23R mutations, current studies focused on rs11805303 are limited. We found rs11805303 had a protective effect on UC susceptibility in our study. Moreover, this SNP was identified as a risk factor for CD, suggesting a common genetic background for these two diseases [27]. Although IL-23R signaling has a key role in the adaptive immune response, it remains unclear whether additional IL-23R variants are associated with IBD in the Chinese patients, and if so, how exactly these IL-23R polymorphisms modulate IBD susceptibility.

A defective PTPN2 function could lead to an enhanced production of proinflammatory mediators and elevated activity of pathways in the intestinal epithelium. The association between PTPN2-rs2542151 and CD was detected by multiple studies analyzing cohorts of European ancestry [12, 21]. Additionally, rs2542151 also showed an association with UC in German patients, as well as in ours [23, 25]. Pooling our data with previous studies, we conclude that IL-23R and PTPN2 may have a limited role in CD patients of East Asia. We further analyzed the interaction of PTPN2-rs2542151 with UC subphenotypes. However, we did not find any specific group of clinical features associated with rs2542151. One possible explanation may be that we overlooked gene–environmental interactions, since IBD are multifactorial diseases and can also be affected by environmental factors. For example, genetic studies carried out among Netherlands and New Zealand cohorts noticed that patients should be stratified according to their smoking behavior; otherwise, the moderate association between CD and rs2542151 can be missed [28]. Thus, environmental factors may need to be taken into account in future studies.

As for CD, a recent study demonstrated a novel role for rs2542151 in the pathogenesis of the disease by increasing the permeability of the intestinal epithelium [29]. Although epithelial barrier dysfunction was observed in both UC and CD, it is still unknown whether rs2542151 could also regulate intestinal epithelial permeability of UC. Biological studies of rs2542151 are required to assess its role in the development of UC.

Except for NOD2-P268S, IL23R-rs11805303, and PTPN2-rs2542151, we did not find other susceptibility loci related to IBD in Chinese people. Similarly, most of those possible associations of the variants with IBD reported in White populations were not detected in Asians, such as Japanese and Korean [13, 27].

A limitation of this study is that our sample size was relatively small. It is likely that the observed SNPs showing evidence of association may be explained by chance finding because studies with small sample size may decrease statistical power or increase the possibility of a false-positive finding. The limited sample size may also contribute to the lack of ability to discern significant differences in allele frequencies between patients and controls. Moreover, some susceptibility variants with low allele frequencies in the Chinese population may be missed. Thus, the results of our exploratory study should be viewed as preliminary findings and further replicated studies in larger cohorts will be required to determine the true interaction between genetic variants and IBD in the Chinese population.

In summary, we identified a novel P268S polymorphism in NOD2/CARD15 associated with Chinese CD patients. We also confirmed associations of PTPN2-rs2542151 and IL23R-rs11805303 with UC for the first time in the Chinese Han population. We conclude that P268S, rs11805303, and rs2542151 are possible candidates for conferring susceptibility to IBD in Chinese people. Our findings complement prior studies and suggest that the genetic background of IBD may vary among different ethnical groups. Further functional studies investigating the role of these genetic polymorphisms in the development of IBD are needed.