| | NOD2 Variants and Antibody Response to Microbial Antigens in Crohn’s Disease Patients and Their Unaffected RelativesReceived 4 May 2006; accepted 26 October 2006. published online 14 November 2006. Background & Aims: The Cdcs1 locus of the C3Bir mouse confers severe colitis associated with a decrease in innate immune function and an increase in adaptive T-cell responses to commensal bacterial products. The aim of our study was to determine if defects in innate immunity are similarly associated with increased adaptive immune responses to microbial antigens in Crohn’s disease patients. Methods: Sera from 732 patients, 220 unaffected relatives, and 200 healthy controls were tested for antibodies to oligomannan, the Pseudomonas fluorescens–related protein, Escherichia coli outer membrane porin C, CBir1 flagellin, and DNA from the same subjects was tested for 3 Crohn’s disease–associated variants of the NOD2 gene, and 5 toll-like receptor (TLR) 2, 2 TLR4, and 2 TLR9 variants. The magnitude of responses to microbial antigens was examined according to variant status. Results: NOD2 variant carriage increased in frequency with increasing number of positive antibodies and increasing cumulative quantitative response as measured by quartile sum (P for trend, .0008 and .0003, respectively). Mean antibody and quartile sums were higher for patients carrying any NOD2 variant versus those carrying none (2.24 vs 1.92 and 10.60 vs 9.72; P = .0008 and P = 0.0003, respectively). The mean quartile sum was higher for unaffected relatives carrying any NOD2 variant versus those carrying none (10.67 vs 9.75, respectively; P = .02). No association was found between any TLR variant and the magnitude of response. Conclusions: Patients with Crohn’s disease and unaffected relatives carrying variants of the NOD2 gene have increased adaptive immune responses to microbial antigens. Crohn’s disease (CD) is a complex clinical and genetic disorder believed to relate to aberrant immunologic responses to commensal bacteria.1, 2 Animal models of CD have shown this response to a wide range of bacterial species.3, 4, 5, 6, 7, 8 In the interleukin-10 knockout mouse model, an inflammatory bowel disease–like phenotype can be demonstrated upon exposure to specific bacterial species in an otherwise germ-free environment.9 Several lines of evidence have implicated enteric bacteria in the pathogenesis of CD in humans. The use of antibiotics has been associated with an inconsistent treatment response in CD.10, 11, 12, 13, 14 Fecal diversion has been shown to decrease the recurrence of CD in the neoterminal ileum after resection, with subsequent instillation of the fecal stream in the excluded ileum leading to inflammatory lesions.15, 16 Moreover, one study has suggested that patients with CD with increased seroreactivity to microbial antigens are more likely to respond to antibiotics.17 A hyperresponsive adaptive immunologic response to microbial antigens has been characterized in patients with CD and is believed to be reflective of the underlying immunopathogenesis of this disorder. Measures of this adaptive immunologic response include antibodies to oligomannan (anti–Saccharomyces cerevisiae antibody [ASCA]), the Pseudomonas fluorescens–related protein (I2), Escherichia coli outer membrane porin C (anti-OmpC), and most recently CBir1 flagellin (anti-CBir1).18, 19, 20, 21, 22, 23 Whether this adaptive immunologic response is reflective of acquired characteristics or has an underlying genetically determined influence is an important question. Supporting the latter possibility is the observation that the prevalence of ASCA expression is elevated not only in patients with CD but in 9%–25% of unaffected relatives versus <5% of healthy controls.24, 25, 26, 27, 28, 29 More recently, a study by Mei et al showed that reactivity to anti-OmpC was present in 15.5% of unaffected relatives versus 6% of healthy controls, yielding a highly significant heritability estimate of 39%. Moreover, there were increased levels of serum expression in unaffected relatives compared with healthy controls even in those individuals with expression falling within the normal range, thus underscoring the fact that seroreactivity to microbial antigens is a quantitative trait.30 These studies thus suggest that the adaptive immunologic response reflects underlying genetic determinants. The concept of an underlying genetic determinant mediating the adaptive immunologic response to bacterial antigens has recently been described in an animal model. In the C3H/HeJBir (C3Bir) interleukin-10–deficient mouse model, the presence of a colitogenic cytokine deficiency induced colitis susceptibility (Cdcs1) allele was associated with an impairment of innate responsiveness to bacterial ligands such as CBir1 flagellin, flagellin X, lipoteichoic acid, muramyl dipeptide (MDP), and CpG oligodeoxynucleotides and a compensatory increase in adaptive CD4 T-cell response to CBir1 flagellin and flagellin X.31 In human studies, several genes or loci have thus far been described that may be associated with CD.32, 33, 34, 35, 36, 37, 38 The best characterized of these susceptibility genes is the innate immune gene NOD2.36, 37, 38 NOD2 is a member of a family of intracellular cytosolic proteins important in mediating the host response to bacterial antigens and is found in epithelial cells of the small and large intestine, as well as monocytes, macrophages, T and B cells, Paneth cells, and dendritic cells.39, 40, 41, 42 NOD2 senses MDP, a highly conserved component of bacterial peptidoglycan, which leads to the secretion of antibacterial substances such as α-defensins and the activation of nuclear factor κB.43, 44 At least 27 mutations of the NOD2 gene have been described, but the majority of susceptibility has been attributed to 3 common mutations, including the 2 missense mutations, R702W and G908R, and one frameshift mutation, 1007fs.37, 45, 46, 47 Variants in NOD2 are believed to result in a diminished innate immune response to MDP.48, 49 Moreover, variants in other pattern recognition receptors such as toll-like receptor (TLR) 2, TLR4, and TLR9 have been inconsistently associated with inflammatory bowel disease.50, 51, 52, 53, 54 In the C3Bir mouse model, a genetic defect in innate immunity accompanying the Cdcs1 allele results in a hyperresponsive adaptive immune response to bacterial ligands. In humans, loss-of-function mutations of the innate immune gene NOD2 could conceivably result in the same phenomenon, with a compensatory adaptive immunologic response to bacterial antigens. Similarly, defects in other innate immune genes, such as TLR genes, could have the same effect. We thus hypothesized that the presence of mutations in the NOD2 as well as a variety of TLR genes (ie, defects in innate immunity) would be associated with a greater serologic response to microbial antigens (ie, an increase in adaptive immunity). To approach this question, we used a large sample of patients with CD, their unaffected relatives, and healthy controls, whom we evaluated for serologic and genetic markers. The results of our studies reported herein show that the presence of a NOD2 variant, although not any TLR variant, is associated with both a greater qualitative and semiquantitative antibody response to microbial antigens in patients with CD. The same relationship was found in unaffected relatives of patients with CD. These results have important implications because they suggest an inherited and, therefore, genetic basis underlying innate and adaptive immune responses in CD and provide a potential pathophysiologic link to similar findings in rodent mucosal inflammation. These findings should facilitate disease-relevant rodent and human crossover genetic and functional studies. Materials and Methods  Patients The cohort of 732 unrelated patients was ascertained from patients assessed at Cedars-Sinai Medical Center from 1988 to 2005. This cohort included 303 patients previously reported55 but also included an additional 429 patients enrolled from the clinic or at the time of surgery. The diagnosis of CD was based on standard endoscopic, histologic, and radiographic features as previously described.55 In addition, a cohort of 220 unaffected relatives of patients with CD as well as 200 healthy controls were included. Many of these unaffected relatives and healthy controls have been previously described and studied for anti-OmpC status.30 All research-related activities were approved by the Cedars-Sinai Medical Center Institutional Review Board. Serologic Analysis and Classification All blood samples were taken at the time of consent and enrollment. Sera were analyzed for expression of ASCA, anti-I2, and anti-OmpC in a blinded fashion by enzyme-linked immunosorbent assay (ELISA) as previously described.23, 55 Antibody levels were determined and results expressed as ELISA units (EU/mL) that are relative to a Cedars-Sinai laboratory (immunoglobulin [Ig] A-I2, IgA-OmpC) or a Prometheus Laboratory standard (San Diego, CA; IgA and IgG ASCA) derived from a pool of patient sera with well-characterized disease found to have reactivity to these antigens. Quantitation of IgG anti-CBir1 reactivity was expressed in ELISA units derived based on a proportion of reactivity relative to a standardized positive control. Because ASCA can be expressed in both an IgA and an IgG class, positivity to ASCA was determined if either class of antibody was above the reference range. In determining a quantitative measure of ASCA, the reactivity was first log-transformed and standardized. The higher of 2 standardized units was then used to determine the quartile of reactivity. With the exception of determining variance (see Statistical Analysis), the magnitude of reactivity to the other 3 antigens was not standardized because each is represented by a single class of antibody. The magnitude of the serologic response to each antigen was divided into 4 equal quartiles in patients with CD, unaffected relatives, and healthy controls, evaluated as 3 separate cohorts, to determine quartile sum scores as previously described.22, 55 Figure 1 shows the patients with the serologic response to each antigen broken down by quartiles and assigned scores of 1–4 on the basis of their designated quartile. By adding individual quartile scores for each microbial antigen, a quartile sum (range, 4–16) was derived that represents the cumulative semiquantitative immune response toward all 4 antigens. The quartile ranking reflects the pool of individuals under study (ie, patient with CD, unaffected relative, or healthy control) and is not directly comparable between groups. Genotyping Three NOD2 variants that have been previously associated with CD37 (R702W, G908R, and 1007fs) were adapted to the TaqMan MGB (Applied Biosystems, Foster City, CA) genotyping platform as previously described.55, 56 Five TLR2 variants (intron, N199N, S450S, P631H, 3′-genomic), 2 TLR4 variants (D299G, S360N), and 2 TLR9 variants (5′-genomic, P545P) were similarly adapted to the TaqMan MGB genotyping platform (Table 1). Variants in the TLR genes were selected based on prior evidence of association with inflammatory bowel disease50, 51, 52, 53, 54 or by the use of Haploview and data from the International HapMap Project.57, 58 | | |  | Gene | Designation | Database SNP | Gene position | TaqMan MGB assay reagents | |
|---|
| TLR2 | | | | | | | | Intron | rs4696480 | 540 | C_27994607_10 | | | | N199N | rs3804099 | 29866 | C_22274563_10 | | | | S450S | rs3804100 | 30639 | C_25607727_10 | | | | P631H | rs5743704 | 31181 | C_25607736_10 | | | | 3′-genomic | rs2405432 | | C_16230373_10 | | | TLR4 | | | | | | | | D299G | rs4986790 | 13015 | C_11722238_20 | | | | S360N | rs4987233 | 13315 | C_43308516_10 | | | TLR9 | | | | | | | | 5′-genomic | rs187084 | 1656 | C_2301954_20 | | | | P545P | rs352140 | 5991 | C_2301952_10 | | | | |
Statistical Analysis We first assessed the relationship between carriage of an NOD2, TLR2, TLR4, and TLR9 variant and collective seroreactivity to microbial antigens both qualitatively and semiquantitatively (because no association was found between any TLR variant and seroreactivity, all subsequent analyses were conducted with only NOD2 variants). We then determined if any particular NOD2 variant was predominant and examined whether any particular antibody or combinations of antibodies was predominant in determining the relationship between NOD2 variants and seroreactivity. The contribution of NOD2 to collective seroreactivity was evaluated by calculating the percent of variance that could be attributed to the presence of NOD2 variants. Finally, we examined whether the presence of an NOD2 variant was related to seroreactivity to microbial antigens in unaffected relatives of patients with CD and healthy controls. Determination of the relationship of NOD2 variants to seroreactivity To determine the significance of increasing frequency of carriage of any NOD2 variants with increasing numbers of qualitatively positive antibodies and with increasing quartile sum (range, 4–16), the Cochran–Armitage trend test was performed.59 To test for differences in the mean quartile sum between those individuals with no NOD2 variant and those with any variant, Student t test was used because the distribution was approximately a normal distribution.59 One-way analysis of variance was performed to test the linear trend of mean quartile sum among those with 0, 1, and 2 NOD2 variants.59 One-way analysis of variance was used to test for a difference in seroreactivity associated with specific NOD2 variants and similarly when comparing mean quartile sum between differing TLR genotypes. Determination of the relative contribution of specific antibody or combinations of antibody positivity The nonparametric Mann–Whitney test was used to compare the level of seroreactivity between those individuals who carried and those who did not carry an NOD2 variant for each antibody.59 To identify whether there is a significant difference in the frequency of carriage of an NOD2 variant among groups within each set with single, double, and triple antibody positivity, χ2 analysis was performed.59 Determination of percent variance contribution by NOD2 In order to determine what proportion of the variation in the seroreactivity to microbial antigens was attributable to the presence of an NOD2 variant, a coefficient of determination (R2), defined as 1 − SS (regression)/SS (total) in analysis of variance, was used.59 Seroreactivity was defined, for this analysis, as the sum of the 4 standardized antibodies, where anti-OmpC = (log[anti-OmpC] − mean[log{anti-OmpC}])/SD(log[anti-OmpC]), and similarly for the other antibodies. All analyses were performed using SAS computer software (version 8.2; SAS Institute, Inc, Cary, NC). Results Serologic and Genetic Characteristics of the Study Population Table 2 shows the serologic and genetic (NOD2) characteristics of the 732-patient cohort. ASCA was detected in 50.4%, anti-I2 in 58.1%, anti-OmpC in 37.2%, and anti-CBir1 in 56.4%. Simple heterozygosity for a disease-predisposing NOD2 variant was detected in 194 patients (26.5%), compound heterozygosity for 2 NOD2 variants was detected in 23 patients (3.1%), and homozygosity for 2 NOD2 variants was detected in 16 patients (2.2%). | | | | Serologic and genetic characteristics | Cohort (n = 732) | |
|---|
| Serologic profile (%) | | | |  ASCA positive (n = 369) | 50.4 | | | Anti-I2 positive (n = 425) | 58.1 | | | Anti-OmpC positive (n = 272) | 37.2 | | | Anti-CBir1 positive (n = 413) | 56.4 | | | NOD2 genotype for R702W, G908R, 1007fs (%) | | | | No mutations (n = 499) | 68.2 | | | Heterozygous (n = 194) | 26.5 | | | Compound heterozygous (n = 23) | 3.1 | | | Homozygous (n = 16) | 2.2 | | | | |
NOD2 Variants, But Not Variants of TLR2, TLR4, or TLR9, Are Associated With Seroreactivity to Microbial Antigens in Patients With CD Our first approach was to determine if we could demonstrate an association between the presence of an NOD2 variant and seroreactivity to microbial antigens. First, the CD patient cohort was divided into 5 groups based on the number of antibodies (from 0 to 4) for which they were qualitatively positive and the proportion of patients with an NOD2 variant in each group was determined. Figure 2 shows that NOD2 variants were present with increasing frequency in patients with reactivity to an increasing number of microbial antigens, especially when there is reactivity to 2 or more antibodies. NOD2 variants were present in those with 0, 1, 2, 3, or 4 positive antibodies at a frequency of 23%, 24%, 36%, 34%, and 42%, respectively (P for trend = .0008). We next sought to investigate the association between the presence of NOD2 variants and semiquantitative seroreactivity by assessing the magnitude of the cumulative serologic response to all 4 antigens using quartile sums as described previously. Figure 3 shows that NOD2 variants were present at increasing frequency in patients with increasing cumulative semiquantitative immune response as reflected by individual quartile sums (P for trend = .0003). These analyses showed that as the serologic response increased, either qualitatively (by number of positive antibodies) or semiquantitatively (by magnitude of the cumulative serologic response), the likelihood of a patient carrying an NOD2 variant increased. Another approach to test this relationship was to compare the serologic response of those patients carrying an NOD2 variant with those carrying no variant. Table 3 shows that, in those patients carrying any NOD2 variant, the mean number of positive antibodies was higher than in those carrying no variant (2.24 ± 1.21 vs 1.92 ± 1.24, respectively; P = .0008). Moreover, those patients carrying any NOD2 variant had a higher mean quartile sum than those carrying no variant (10.60 ± 3.03 vs 9.72 ± 3.01, respectively; P = .0003). The mean quartile sum in individuals with and without any of the TLR2, TLR4, and TLR9 variants under study was compared in a similar fashion. Table 4 shows that there was no association between seroreactivity to microbial antigens and the TLR variants listed. | | | | | No NOD2 variant (n = 499) | Any NOD2 variant (n = 233) | P value | |
|---|
| Mean no. of antibody positivity | 1.92±1.24 | 2.24±1.21 | .0008 | | | Mean quartile sum (mean ± SD) | 9.72±3.01 | 10.60±3.03 | .0003 | | | | |
Because our data showed that the presence of a defective innate immune gene (NOD2) was associated with a hyperresponsive adaptive immunologic response, we next sought to determine if having 2 defective alleles would be associated with a greater response than having only one. Figure 4 shows that the mean quartile sum increased in parallel with increasing number of NOD2 variants (P for trend = .002). The Relationship of Specific NOD2 Variants to Seroreactivity to Microbial Antigens Different NOD2 variants are associated with differential degrees of altered sensing of MDP. The frameshift mutation 1007fs is associated with a more significant decrease in nuclear factor κB activity than the 2 missense mutations, R702W and G908R.38, 60 Therefore, we sought to determine if seroreactivity to microbial antigens varied according to which NOD2 variant was present in an individual. There was no significant difference in the cohort-specific mean quartile sum in individuals with CD with one or 2 1007fs, G908R, and R702W variants, respectively (10.11 ± 3.12, 10.63 ± 3.18, and 11.06 ± 2.78, respectively; P = .16). Increasing Cumulative Seroreactivity Rather Than Specific Antibody Combinations Are Associated With the Presence of an NOD2 Variant Our data thus indicated that the presence of an NOD2 variant was associated with an increased serologic response to microbial antigens both in terms of the number of positive antibodies and the cumulative response as measured by quartile sum. Our next question was whether any particular antibody or combinations of antibodies was the predominant factor in determining this relationship. We first examined the absolute level of response to each antibody individually rather than collectively to determine if the presence of any NOD2 variant was associated with higher individual reactivity. Table 5 shows that for each of the 4 antibodies, the magnitude of seroreactivity was higher when an NOD2 variant was present. Because there is a significant correlation among the expression of these antibodies in patients with CD,22, 55 we then divided the patients with CD into 16 mutually exclusive groups (Figure 5) based on all possible permutations of antibody positivity: no positive antibodies, single antibody positivity (4 groups in set 1), double antibody positivity (6 groups in set 2), triple antibody positivity (4 groups in set 3), and all antibodies positive. We then tested whether there was a significant difference among groups within each set where the groups had the same number of antibody positivity. There was no statistically significant difference in the frequency of NOD2 variants among groups within each set, implying that no single antibody or combination of antibody positivity was wholly responsible for the association between seroreactivity and variant status (Figure 5). If, for example, a given antibody or combination of antibody positivity was responsible for the association, it would be anticipated that the frequency of NOD2 carriage would be significantly greater in individuals with positivity to that antibody or combination. Therefore, these data indicate that the relationship between NOD2 variants and serologic response to microbial antigens reflects a cumulative effect rather than being driven by any particular antibody or antibody combination. After determining that the presence of an NOD2 variant was associated with both a qualitatively and semiquantitatively increased seroreactivity to microbial antigens, a calculation of variance was performed to determine what proportion of the variability in seroreactivity was attributable to the presence of an NOD2 variant. This calculation showed that 2.7% of the variability in the sum of the semiquantitative antibody levels was attributable to the presence of an NOD2 variant. The Presence of NOD2 Variants Is Significantly Related to Seroreactivity to Microbial Antigens in Unaffected Relatives of Patients With CD Both ASCA and anti-OmpC expression have been noted to be elevated in unaffected relatives of patients with CD, suggesting an underlying genetic determination of seroreactivity.24, 25, 26, 27, 28, 29, 30 To explore this concept further, our final approach was to determine if the presence of an NOD2 variant in unaffected relatives of patients with CD, and in a separate cohort of healthy controls, would be associated with a similarly greater adaptive immunologic response to microbial antigens. A quartile sum was again derived as previously described for patients with CD. The quartile sums in patients with CD, unaffected relatives, and healthy controls were based on the distribution of the magnitude of seroreactivity within each cohort; thus, the same quartile sum in a patient with CD or in a relative or healthy control is not representative of the same absolute magnitude of response and thus is not directly comparable. The magnitude of serologic response was significantly lower, as expected, in unaffected relatives and healthy controls compared with cases and generally fell within the normal range (data not shown). We utilized sera from 220 unaffected relatives of patients with CD (92% first-degree). Figure 6 shows that in the unaffected relatives, the mean quartile sum in those individuals carrying any NOD2 variant was higher than in those carrying no variant (10.67 ± 2.73 vs 9.75 ± 2.52; P = .02). The same analysis was undertaken using sera from 200 healthy controls. Again, the magnitude of seroreactivity was divided into quartiles based on the distribution specifically within this cohort. Cohort-specific quartile sums were again derived as previously described. Figure 7 shows that the mean quartile sum in healthy controls carrying any NOD2 variant (n = 24) showed a trend toward being higher than in healthy controls carrying no variant (n = 176) (10.79 ± 2.95 vs 9.69 ± 2.71; P = .07). Discussion The major etiologic hypothesis regarding CD is that it is likely related to a dysregulated immunologic response to enteric microorganisms. One manifestation of this dysregulated immunologic response is the expression of antibodies to microbial antigens. High levels of ASCA, anti-I2, and anti-OmpC have been associated with fibrostenosing and internal penetrating disease as well as the need for small bowel surgery.55 More recently, anti-CBir1 has been found to be independently associated with severe small bowel disease such as internal perforating and fibrostenosing disease.23 Indeed, defects in the innate immune gene NOD2 have also been found to be associated with a fibrostenosing clinical phenotype, suggesting a complex interaction between genetic susceptibility, the adaptive immunologic response, and clinical disease behavior.46, 55, 56 A suggestion of a link between the adaptive immunologic response and genetic susceptibility is supported by the finding of increased ASCA and anti-OmpC expression, both qualitatively and quantitatively, in unaffected relatives of patients with CD.24, 25, 26, 27, 28, 29, 30 Moreover, the recent finding that the presence of a defective innate immune gene locus (Cdcs1) that renders the host less responsive to microbial products and confers severe colitis in the interleukin-10–deficient C3Bir mouse in association with a hyperresponsive adaptive immunologic response to these same products supports the concept of a genetic link between innate and adaptive immunity and susceptibility to mucosal inflammation.31 We hypothesized that the presence of defective innate immune genes that render the host less responsive to bacterial products would be associated with a compensatory hyperresponsive adaptive immunologic response to microbial antigens. A previous study by Abreu et al found a borderline association between ASCA expression and the 1007fs NOD2 variant, while a study by Walker et al found no association between NOD2 variants and ASCA expression.56, 61 Similarly, a study by Arnott et al failed to show an association between ASCA, anti-I2, and anti-OmpC expression and NOD2 variants.62 However, this latter study included only 142 patients with CD and, thus, may have been underpowered to detect a relevant association. Moreover, the rate of NOD2 mutations in the Scottish population in this study was only 23.9%, lower than the 37.3% rate found in our previously studied North American cohort.55 Finally, recent studies by Annese et al and Cruyssen et al did demonstrate an association between NOD2 variant status and ASCA expression; however, they did not study the reactivity to other microbial antigens as in the present study.63, 64 In our study herein, we showed that patients with CD with a predominant qualitative (number of positive antibodies) and semiquantitative (absolute magnitude of response) serologic response to microbial antigens were more likely to carry an NOD2 variant (Figure 2, Figure 3). Moreover, we showed that patients with CD carrying an NOD2 variant had a higher qualitative and semiquantitative serologic response than patients carrying no variants (Table 3). This relationship was seen not only with the cumulative response to all 4 antibodies, but also with each antibody individually (Table 5). Because this had not been shown previously, we sought to explore whether the finding was reflective of a relationship between a specific antibody, particularly anti-Cbir1, because it had not been studied in this context previously, and NOD2 variant status. We were able to show that the association of seroreactivity to microbial antigens to NOD2 variant status was more a reflection of the cumulative semiquantitative response than any particular antibody or combination of antibodies. E coli, P fluorescens, and most flagellated bacterial species will express MDP as components of their bacterial cell walls. However, the increased expression of antibodies directed against bacterial and yeast antigens is likely a function of increased exposure of the mucosal immune system to a range of microbial antigens owing to diminished initial clearance, perhaps due to impaired secretion of defensins. Hence, a defect in MDP signaling via NOD2 variants could result in impaired defense against microbial species, with the subsequent development of antibodies to microbial antigens being a secondary phenomenon due to bacterial invasion and increased exposure of the mucosal immune system to a range of microbial antigens. The finding that only 2.7% of the variance in seroreactivity is attributable to the presence of an NOD2 variant is not surprising and does not detract from the relevance of our finding. This is in keeping with other complex genetic disorders such as insulin resistance and hypercholesterolemia. Approximately 6% of the variability in insulin clearance is due to variation in the gene for muscle-specific AMP deaminase, and as little as 6% of the variability in serum cholesterol is ascribable to different apolipoprotein E polymorphisms.65, 66 Furthermore, the variance we found in our study is likely the lower end of the true association between NOD2 variants and adaptive immunologic response, because we were only testing for the 3 most common variants (R702W, G908R, and 1007fs). More than 27 variants have been described, and in one large cohort more than 19% of disease-associated variants were not of the 3 most common mutations for which we tested.46 The innate immune system is complex and involves the sensing of bacterial products via many mechanisms, including not only NOD2 but TLRs, which act as pattern-recognition receptors serving to regulate the immunologic response of the host to enteric bacteria.2 There are 11 known mammalian TLRs, including TLR2, TLR4, TLR5, and TLR9, that sense bacterial lipoproteins, lipopolysaccharide, flagellin, and bacterial and viral CpG DNA, respectively.67 Defects in TLR signaling could also lead to diminished sensing and subsequent clearance of bacteria, leading to invasion and a compensatory adaptive immunologic response. Indeed, the TLR4 D299G polymorphism has been associated with both ulcerative colitis and CD in a Belgian study, whereas no association was found in a Scottish and Irish cohort.53, 68 In one Greek study, the presence of both TLR4 and NOD2 mutations was associated with increased susceptibility to inflammatory bowel disease, suggesting a synergistic effect.69 As previously discussed, variants of the genes for both TLR2 and TLR9 have also been associated with CD.50, 54 In this study, we were not able to show an association between variants in these TLR genes and seroreactivity to microbial antigens. This is not necessarily surprising because the association between variants in TLR genes and inflammatory bowel disease has been less consistent than the association between CD and functional variants of the NOD2 gene. Indeed, it has been suggested in a study by Oostenbrug et al that the D299G polymorphism is not causal but is in linkage with the true susceptibility variants of the TLR4 gene.51 This could serve as a potential explanation for the lack of association of seroreactivity to the D299G variant in our study. We would hypothesize that as we advance our understanding of defects in innate and adaptive immune response in CD, new gene defects will be characterized and new associations will be found with seroreactivity to microbial antigens paralleling our finding with NOD2 variants. Our finding that unaffected relatives carrying an NOD2 variant had a greater serologic response to microbial antigens than those carrying no variants further strengthens our conclusion. Because both seroreactivity and NOD2 variant status have been linked to disease severity, one argument could be that the association between NOD2 variant status and seroreactivity is a function of a common end point, and the relationship is only a surrogate. Arguing against this is the fact that in a study by Cruyssen et al,64 the association between ASCA and NOD2 variant status was independent of disease phenotype and that the unaffected relatives in our study have no apparent disease activity. However, it has been shown in a study by Thjodleifsson et al that 49% of unaffected relatives of patients with CD have elevated levels of fecal calprotectin, thus implying an element of subclinical intestinal inflammation.70 Subclinical intestinal inflammation could lead to increased mucosal permeability and subsequent exposure of the mucosal immune system to microbial antigens. However, if altered gut permeability is etiologic in determining the seroreactivity of unaffected relatives, then there would be no difference between those with and without NOD2 variants unless the presence of a variant itself was a determining factor. Therefore, this argues that the association between innate immune defects (NOD2) and adaptive immunologic response as measured by seroreactivity to microbial antigens is direct. The same relationship between NOD2 variant status and seroreactivity to microbial antigens was not statistically significant in healthy controls. However, only 12% of these individuals carried a variant; therefore, the sample size may have been too small to detect a significant difference (type II error). There was a trend toward healthy controls carrying an NOD2 variant having higher seroreactivity than those carrying no variant (Figure 7). This further supports the supposition that CD has a complex genetic basis and that a single innate immune defect is insufficient to cause disease but is nevertheless sufficient to be associated with an aberrant adaptive immunologic response. In summary, this study has shown that a significant degree of the variability in the adaptive immunologic response to CD-associated microbial antigens is due to the presence of a defective innate immune gene (NOD2). This relationship can be found in unaffected relatives of patients with CD and even perhaps in healthy controls as well. This supports the concept of a genetic basis for a link between innate immune defects and dysregulated, hyperresponsive adaptive immunity to microbial antigens in human CD, a link that parallels findings in rodent mucosal inflammation. Further studies can now explore this relationship between multiple innate or perhaps adaptive immune defects and adaptive immunologic response as new variants in innate and adaptive immune genes are described. Finally, this cumulative quantitative response could be used as a basis for targeting individuals in whom to search for novel genes associated with CD. References 1. 1Podolsky DK. Inflammatory bowel disease. N Engl J Med. 2002;347:417–429.
CrossRef
2. 2Cobrin GM, Abreu MT. Defects in mucosal immunity leading to Crohn’s disease. Immunol Rev. 2005;206:277–295. MEDLINE |
CrossRef
3. 3Strober W, Fuss IJ, Blumberg RS. The immunology of mucosal models of inflammation. Annu Rev Immunol. 2002;20:495–549. MEDLINE |
CrossRef
4. 4De Winter H, Cheroutre H, Kronenberg M. Mucosal immunity and inflammation (II. The yin and yang of T cells in intestinal inflammation: pathogenic and protective roles in a mouse colitis model). Am J Physiol. 1999;276:G1317–G1321. MEDLINE 5. 5Cong Y, Weaver CT, Lazenby A, Elson CO. Colitis induced by enteric bacterial antigen-specific CD4+ T cells requires CD40-CD40 ligand interactions for a sustained increase in mucosal IL-12. J Immunol. 2000;165:2173–2182. MEDLINE 6. 6Cong Y, Brandwein SL, McCabe RP, Lazenby A, Birkenmeier EH, Sundberg JP, et al. CD4+ T cells reactive to enteric bacterial antigens in spontaneously colitic C3H/HeJBir mice: increased T helper cell type 1 response and ability to transfer disease. J Exp Med. 1998;187:855–864. MEDLINE |
CrossRef
7. 7Cong Y, Weaver CT, Lazenby A, Elson CO. Bacterial-reactive T regulatory cells inhibit pathogenic immune responses to the enteric flora. J Immunol. 2002;169:6112–6119. MEDLINE 8. 8Brandwein SL, McCabe RP, Cong Y, Waites KB, Ridwan BU, Dean PA, et al. Spontaneously colitic C3H/HeJBir mice demonstrate selective antibody reactivity to antigens of the enteric bacterial flora. J Immunol. 1997;159:44–52. MEDLINE 9. 9Kim SC, Tonkonogy SL, Albright CA, Tsang J, Balish EJ, Braun J, et al. Variable phenotypes of enterocolitis in interleukin 10-deficient mice monoassociated with two different commensal bacteria. Gastroenterology. 2005;128:891–906. Abstract | Full Text |
Full-Text PDF (862 KB)
|
CrossRef
10. 10Prantera C, Zannoni F, Scribano ML, Berto E, Andreoli A, Kohn A, et al. An antibiotic regimen for the treatment of active Crohn’s disease: a randomized, controlled clinical trial of metronidazole plus ciprofloxacin. Am J Gastroenterol. 1996;91:328–332. MEDLINE 11. 11Prantera C, Kohn A, Zannoni F, Spimpolo N, Bonfa M. Metronidazole plus ciprofloxacin in the treatment of active, refractory Crohn’s disease: results of an open study. J Clin Gastroenterol. 1994;19:79–80. MEDLINE |
CrossRef
12. 12Prantera C, Kohn A, Mangiarotti R, Andreoli A, Luzi C. Antimycobacterial therapy in Crohn’s disease: results of a controlled, double-blind trial with a multiple antibiotic regimen. Am J Gastroenterol. 1994;89:513–518. MEDLINE 13. 13Rutgeerts P, Hiele M, Geboes K, Peeters M, Penninckx F, Aerts R, et al. Controlled trial of metronidazole treatment for prevention of Crohn’s recurrence after ileal resection. Gastroenterology. 1995;108:1617–1621. Abstract |
Full-Text PDF (526 KB)
|
CrossRef
14. 14Steinhart AH, Feagan BG, Wong CJ, Vandervoort M, Mikolainis S, Croitoru K, et al. Combined budesonide and antibiotic therapy for active Crohn’s disease: a randomized controlled trial. Gastroenterology. 2002;123:33–40. Abstract | Full Text |
Full-Text PDF (126 KB)
|
CrossRef
15. 15Rutgeerts P, Goboes K, Peeters M, Hiele M, Penninckx F, Aerts R, et al. Effect of faecal stream diversion on recurrence of Crohn’s disease in the neoterminal ileum. Lancet. 1991;338:771–774. Abstract |
CrossRef
16. 16D’Haens GR, Geboes K, Peeters M, Baert F, Penninckx F, Rutgeerts P. Early lesions of recurrent Crohn’s disease caused by infusion of intestinal contents in excluded ileum. Gastroenterology. 1998;114:262–267. Abstract | Full Text |
Full-Text PDF (421 KB)
|
CrossRef
17. 17Mow WS, Landers CJ, Steinhart AH, Feagan BG, Croitoru K, Seidman E, et al. High-level serum antibodies to bacterial antigens are associated with antibiotic-induced clinical remission in Crohn’s disease: a pilot study. Dig Dis Sci. 2004;49:1280–1286. MEDLINE |
CrossRef
18. 18Main J, McKenzie H, Yeaman GR, Kerr MA, Robson D, Pennington CR, et al. Antibody to Saccharomyces cerevisiae (bakers’ yeast) in Crohn’s disease. BMJ. 1988;297:1105–1106. 19. 19Quinton JF, Sendid B, Reumaux D, Duthilleul P, Cortot A, Grandbastien B, et al. Anti-Saccharomyces cerevisiae mannan antibodies combined with antineutrophil cytoplasmic autoantibodies in inflammatory bowel disease: prevalence and diagnostic role. Gut. 1998;42:788–791. MEDLINE 20. 20Sutton CL, Kim J, Yamane A, Dalwadi H, Wei B, Landers C, et al. Identification of a novel bacterial sequence associated with Crohn’s disease. Gastroenterology. 2000;119:23–31. Abstract | Full Text |
Full-Text PDF (281 KB)
|
CrossRef
21. 21Cohavy O, Bruckner D, Gordon LK, Misra R, Wei B, Eggena ME, et al. Colonic bacteria express an ulcerative colitis pANCA-related protein epitope. Infect Immun. 2000;68:1542–1548. MEDLINE |
CrossRef
22. 22Landers CJ, Cohavy O, Misra R, Yang H, Lin YC, Braun J, et al. Selected loss of tolerance evidenced by Crohn’s disease-associated immune responses to auto- and microbial antigens. Gastroenterology. 2002;123:689–699. Abstract | Full Text |
Full-Text PDF (309 KB)
|
CrossRef
23. 23Targan SR, Landers CJ, Yang H, Lodes MJ, Cong Y, Papadakis KA, et al. Antibodies to CBir1 flagellin define a unique response that is associated independently with complicated Crohn’s disease. Gastroenterology. 2005;128:2020–2028. Abstract | Full Text |
Full-Text PDF (288 KB)
|
CrossRef
24. 24Seibold F, Stich O, Hufnagl R, Kamil S, Scheurlen M. Anti-Saccharomyces cerevisiae antibodies in inflammatory bowel disease: a family study. Scand J Gastroenterol. 2001;36:196–201. MEDLINE |
CrossRef
25. 25Glas J, Torok HP, Vilsmaier F, Herbinger KH, Hoelscher M, Folwaczny C. Anti-saccharomyces cerevisiae antibodies in patients with inflammatory bowel disease and their first-degree relatives: potential clinical value. Digestion. 2002;66:173–177. MEDLINE |
CrossRef
26. 26Annese V, Andreoli A, Andriulli A, Dinca R, Gionchetti P, Latiano A, et al. Familial expression of anti-Saccharomyces cerevisiae Mannan antibodies in Crohn’s disease and ulcerative colitis: a GISC study. Am J Gastroenterol. 2001;96:2407–2412. MEDLINE |
CrossRef
27. 27Vermeire S, Peeters M, Vlietinck R, Joossens S, Den Hond E, Bulteel V, et al. Anti-Saccharomyces cerevisiae antibodies (ASCA), phenotypes of IBD, and intestinal permeability: a study in IBD families. Inflamm Bowel Dis. 2001;7:8–15. MEDLINE 28. 28Sutton CL, Yang H, Li Z, Rotter JI, Targan SR, Braun J. Familial expression of anti-Saccharomyces cerevisiae mannan antibodies in affected and unaffected relatives of patients with Crohn’s disease. Gut. 2000;46:58–63. MEDLINE |
CrossRef
29. 29Sendid B, Quinton JF, Charrier G, Goulet O, Cortot A, Grandbastien B, et al. Anti-Saccharomyces cerevisiae mannan antibodies in familial Crohn’s disease. Am J Gastroenterol. 1998;93:1306–1310. MEDLINE |
CrossRef
30. 30Mei L, Targan SR, Landers CJ, Dutridge D, Ippoliti A, Vasiliauskas EA, et al. Familial expression of anti-Escherichia coli outer membrane porin C in relatives of patients with Crohn’s disease. Gastroenterology. 2006;130:1078–1085. Abstract | Full Text |
Full-Text PDF (173 KB)
|
CrossRef
31. 31Beckwith J, Cong Y, Sundberg JP, Elson CO, Leiter EH. Cdcs1, a major colitogenic locus in mice, regulates innate and adaptive immune response to enteric bacterial antigens. Gastroenterology. 2005;129:1473–1484. Abstract | Full Text |
Full-Text PDF (695 KB)
|
CrossRef
32. 32Russell RK, Nimmo ER, Satsangi J. Molecular genetics of Crohn’s disease. Curr Opin Genet Dev. 2004;14:264–270.
CrossRef
33. 33Mathew CG, Lewis CM. Genetics of inflammatory bowel disease: progress and prospects. Hum Mol Genet. 2004;13:R161–R168. 34. 34Watts DA, Satsangi J. The genetic jigsaw of inflammatory bowel disease. Gut. 2002;50(Suppl 3):III31–III36.
CrossRef
35. 35Noble CL, Nimmo ER, Drummond H, Ho GT, Tenesa A, Smith L, et al. The contribution of OCTN1/2 variants within the IBD5 locus to disease susceptibility and severity in Crohn’s disease. Gastroenterology. 2005;129:1854–1864. Abstract | Full Text |
Full-Text PDF (198 KB)
|
CrossRef
36. 36Hugot JP, Laurent-Puig P, Gower-Rousseau C, Olson JM, Lee JC, Beaugerie L, et al. Mapping of a susceptibility locus for Crohn’s disease on chromosome 16. Nature. 1996;379:821–823. MEDLINE |
CrossRef
37. 37Hugot JP, Chamaillard M, Zouali H, Lesage S, Cezard JP, Belaiche J, et al. Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn’s disease. Nature. 2001;411:599–603. MEDLINE |
CrossRef
38. 38Ogura Y, Bonen DK, Inohara N, Nicolae DL, Chen FF, Ramos R, et al. A frameshift mutation in NOD2 associated with susceptibility to Crohn’s disease. Nature. 2001;411:603–606. MEDLINE |
CrossRef
39. 39Inohara N, Ogura Y, Nunez G. Nods: a family of cytosolic proteins that regulate the host response to pathogens. Curr Opin Microbiol. 2002;5:76–80. MEDLINE |
CrossRef
40. 40Gutierrez O, Pipaon C, Inohara N, Fontalba A, Ogura Y, Prosper F, et al. Induction of Nod2 in myelomonocytic and intestinal epithelial cells via nuclear factor-kappa B activation. J Biol Chem. 2002;277:41701–41705. MEDLINE |
CrossRef
41. 41Rosenstiel P, Fantini M, Brautigam K, Kuhbacher T, Waetzig GH, Seegert D, et al. TNF-alpha and IFN-gamma regulate the expression of the NOD2 (CARD15) gene in human intestinal epithelial cells. Gastroenterology. 2003;124:1001–1009. Abstract | Full Text |
Full-Text PDF (266 KB)
|
CrossRef
42. 42Lala S, Ogura Y, Osborne C, Hor SY, Bromfield A, Davies S, et al. Crohn’s disease and the NOD2 gene: a role for paneth cells. Gastroenterology. 2003;125:47–57. Abstract | Full Text |
Full-Text PDF (811 KB)
|
CrossRef
43. 43Ogura Y, Inohara N, Benito A, Chen FF, Yamaoka S, Nunez G. Nod2, a Nod1/Apaf-1 family member that is restricted to monocytes and activates NF-kappaB. J Biol Chem. 2001;276:4812–4818. MEDLINE |
CrossRef
44. 44Kobayashi KS, Chamaillard M, Ogura Y, Henegariu O, Inohara N, Nunez G, et al. Nod2-dependent regulation of innate and adaptive immunity in the intestinal tract. Science. 2005;307:731–734.
CrossRef
45. 45Shaw SH, Hampe J, White R, Mathew CG, Curran ME, Schreiber S. Stratification by CARD15 variant genotype in a genome-wide search for inflammatory bowel disease susceptibility loci. Hum Genet. 2003;113:514–521. MEDLINE |
CrossRef
46. 46Lesage S, Zouali H, Cezard JP, Colombel JF, Belaiche J, Almer S, et al. CARD15/NOD2 mutational analysis and genotype-phenotype correlation in 612 patients with inflammatory bowel disease. Am J Hum Genet. 2002;70:845–857. MEDLINE |
CrossRef
47. 47Sugimura K, Taylor KD, Lin YC, Hang T, Wang D, Tang YM, et al. A novel NOD2/CARD15 haplotype conferring risk for Crohn disease in Ashkenazi Jews. Am J Hum Genet. 2003;72:509–518. MEDLINE |
CrossRef
48. 48Inohara N, Ogura Y, Fontalba A, Gutierrez O, Pons F, Crespo J, et al. Host recognition of bacterial muramyl dipeptide mediated through NOD2 (Implications for Crohn’s disease). J Biol Chem. 2003;278:5509–5512. MEDLINE |
CrossRef
49. 49Li J, Moran T, Swanson E, Julian C, Harris J, Bonen DK, et al. Regulation of IL-8 and IL-1beta expression in Crohn’s disease associated NOD2/CARD15 mutations. Hum Mol Genet. 2004;13:1715–1725. MEDLINE |
CrossRef
50. 50Pierik M, Joossens S, Van Steen K, Van Schuerbeek N, Vlietinck R, Rutgeerts P, et al. Toll-like receptor-1, -2, and -6 polymorphisms influence disease extension in inflammatory bowel diseases. Inflamm Bowel Dis. 2006;12:1–8. MEDLINE |
CrossRef
51. 51Oostenbrug LE, Drenth JP, de Jong DJ, Nolte IM, Oosterom E, van Dullemen HM, et al. Association between Toll-like receptor 4 and inflammatory bowel disease. Inflamm Bowel Dis. 2005;11:567–575. MEDLINE |
CrossRef
52. 52Brand S, Staudinger T, Schnitzler F, Pfennig S, Hofbauer K, Dambacher J, et al. The role of Toll-like receptor 4 Asp299Gly and Thr399Ile polymorphisms and CARD15/NOD2 mutations in the susceptibility and phenotype of Crohn’s disease. Inflamm Bowel Dis. 2005;11:645–652. MEDLINE |
CrossRef
53. 53Franchimont D, Vermeire S, El Housni H, Pierik M, Van Steen K, Gustot T, et al. Deficient host-bacteria interactions in inflammatory bowel disease? (The toll-like receptor (TLR)-4 Asp299gly polymorphism is associated with Crohn’s disease and ulcerative colitis). Gut. 2004;53:987–992. MEDLINE |
CrossRef
54. 54Torok HP, Glas J, Tonenchi L, Bruennler G, Folwaczny M, Folwaczny C. Crohn’s disease is associated with a toll-like receptor-9 polymorphism. Gastroenterology. 2004;127:365–366. Full Text |
Full-Text PDF (73 KB)
|
CrossRef
55. 55Mow WS, Vasiliauskas EA, Lin YC, Fleshner PR, Papadakis KA, Taylor KD, et al. Association of antibody responses to microbial antigens and complications of small bowel Crohn’s disease. Gastroenterology. 2004;126:414–424. Abstract | Full Text |
Full-Text PDF (285 KB)
|
CrossRef
56. 56Abreu MT, Taylor KD, Lin YC, Hang T, Gaiennie J, Landers CJ, et al. Mutations in NOD2 are associated with fibrostenosing disease in patients with Crohn’s disease. Gastroenterology. 2002;123:679–688. Abstract | Full Text |
Full-Text PDF (141 KB)
|
CrossRef
57. 57The International HapMap Project. Nature. 2003;426:789–796.
CrossRef
58. 58Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics. 2005;21:263–265. 59. 59Armitage P, Berry G, Matthews JNS. Statistical methods in medical research. Malden, MA: Blackwell; 2005;. 60. 60Chamaillard M, Philpott D, Girardin SE, Zouali H, Lesage S, Chareyre F, et al. Gene-environment interaction modulated by allelic heterogeneity in inflammatory diseases. Proc Natl Acad Sci U S A. 2003;100:3455–3460. MEDLINE |
CrossRef
61. 61Walker LJ, Aldhous MC, Drummond HE, Smith BR, Nimmo ER, Arnott ID, et al. Anti-Saccharomyces cerevisiae antibodies (ASCA) in Crohn’s disease are associated with disease severity but not NOD2/CARD15 mutations. Clin Exp Immunol. 2004;135:490–496. MEDLINE |
CrossRef
62. 62Arnott ID, Landers CJ, Nimmo EJ, Drummond HE, Smith BK, Targan SR, et al. Sero-reactivity to microbial components in Crohn’s disease is associated with disease severity and progression, but not NOD2/CARD15 genotype. Am J Gastroenterol. 2004;99:2376–2384. MEDLINE |
CrossRef
63. 63Annese V, Lombardi G, Perri F, D’Inca R, Ardizzone S, Riegler G, et al. Variants of CARD15 are associated with an aggressive clinical course of Crohn’s disease—an IG-IBD study. Am J Gastroenterol. 2005;100:84–92. MEDLINE |
CrossRef
64. 64Cruyssen BV, Peeters H, Hoffman IE, Laukens D, Coucke P, Marichal D, et al. CARD15 polymorphisms are associated with anti-Saccharomyces cerevisiae antibodies in caucasian Crohn’s disease patients. Clin Exp Immunol. 2005;140:354–359. MEDLINE |
CrossRef
65. 65Goodarzi MO, Taylor KD, Guo X, Quinones MJ, Cui J, Li X, et al. Variation in the gene for muscle-specific AMP deaminase is associated with insulin clearance, a highly heritable trait. Diabetes. 2005;54:1222–1227. MEDLINE |
CrossRef
66. 66Motulsky AG, Brunzell JD. Genetics of coronary atherosclerosis. In: King RA, Rotter JI, Motulsky AG editor. The genetic basis of common disease. 2nd ed.. New York, NY: Oxford University Press; 2002;p. 105–126. 67. 67Cario E. Bacterial interactions with cells of the intestinal mucosa: Toll-like receptors and NOD2. Gut. 2005;54:1182–1193. MEDLINE |
CrossRef
68. 68Arnott ID, Nimmo ER, Drummond HE, Fennell J, Smith BR, MacKinlay E, et al. NOD2/CARD15, TLR4 and CD14 mutations in Scottish and Irish Crohn’s disease patients: evidence for genetic heterogeneity within Europe?. Genes Immun. 2004;5:417–425. MEDLINE |
CrossRef
69. 69Gazouli M, Mantzaris G, Kotsinas A, Zacharatos P, Papalambros E, Archimandritis A, et al. Association between polymorphisms in the Toll-like receptor 4, CD14, and CARD15/NOD2 and inflammatory bowel disease in the Greek population. World J Gastroenterol. 2005;11:681–685. MEDLINE 70. 70Thjodleifsson B, Sigthorsson G, Cariglia N, Reynisdottir I, Gudbjartsson DF, Kristjansson K, et al. Subclinical intestinal inflammation: an inherited abnormality in Crohn’s disease relatives?. Gastroenterology. 2003;124:1728–1737. Abstract | Full Text |
Full-Text PDF (157 KB)
|
CrossRef
⁎ Inflammatory Bowel Disease Center, Cedars-Sinai Medical Center, Los Angeles, California ‡ Medical Genetics Institute, Cedars-Sinai Medical Center, Los Angeles, California ∥ Department of Surgery, Cedars-Sinai Medical Center, Los Angeles, California § Division of Gastroenterology, Department of Medicine, Mount Sinai School of Medicine, New York, New York  Address requests for reprints to: Stephan R. Targan, MD, Division of Gastroenterology, Inflammatory Bowel Disease Center, and Immunobiology Institute, Cedars-Sinai Medical Center, 110 George Burns Road, Davis Building, Room 4063, Los Angeles, California 90048. fax: (310) 423-0224.
Supported by National Institutes of Health grant PO1 DK46763. PII: S0016-5085(06)02468-1 doi:10.1053/j.gastro.2006.11.013 © 2007 AGA Institute. Published by Elsevier Inc. All rights reserved. | |
|
FULL TEXT02468-1/journalimage?img=PIIS0016508506024681.gr1.lrg.gif&fig=fig1&kwhquery=&issn=0016-5085&ishighres=false&allhighres=false&free=yes','journalimage');)
02468-1/journalimage?img=PIIS0016508506024681.gr2.lrg.gif&fig=fig2&kwhquery=&issn=0016-5085&ishighres=false&allhighres=false&free=yes','journalimage');)
02468-1/journalimage?img=PIIS0016508506024681.gr3.lrg.gif&fig=fig3&kwhquery=&issn=0016-5085&ishighres=false&allhighres=false&free=yes','journalimage');)
02468-1/journalimage?img=PIIS0016508506024681.gr4.lrg.gif&fig=fig4&kwhquery=&issn=0016-5085&ishighres=false&allhighres=false&free=yes','journalimage');)
02468-1/journalimage?img=PIIS0016508506024681.gr5.lrg.gif&fig=fig5&kwhquery=&issn=0016-5085&ishighres=false&allhighres=false&free=yes','journalimage');)
02468-1/journalimage?img=PIIS0016508506024681.gr6.lrg.gif&fig=fig6&kwhquery=&issn=0016-5085&ishighres=false&allhighres=false&free=yes','journalimage');)
02468-1/journalimage?img=PIIS0016508506024681.gr7.lrg.gif&fig=fig7&kwhquery=&issn=0016-5085&ishighres=false&allhighres=false&free=yes','journalimage');)