OUP user menu

Journal of Crohn's and Colitis: 10 (9)

Editor-in-Chief

Laurence J. Egan, Ireland

Associate Editors

Shomron Ben-Horin, IsraelSilvio Danese, ItalyPeter Lakatos, HungaryMiles Parkes, UKJesús Rivera-Nieves, USABritta Siegmund, GermanyGijs van den Brink, NLSéverine Vermeire, Belgium

6.585
5.586

Published on behalf of

A functional polymorphism in UGT1A1 related to hyperbilirubinemia is associated with a decreased risk for Crohn's disease

Hilbert S. de Vries, Rene H.M. te Morsche, Kevin Jenniskens, Wilbert H.M. Peters, Dirk J. de Jong
DOI: http://dx.doi.org/10.1016/j.crohns.2011.11.010 597-602 First published online: 1 June 2012

Abstract

Background: An imbalance between the production of reactive oxygen species (ROS) and their capturing by antioxidants results in oxidative stress, this may play an important role in the pathogenesis of inflammatory bowel disease (IBD). Since bilirubin is an important endogenous antioxidant, increased levels of bilirubin may protect against IBD. UDP-glucuronosyltransferase 1A1 (UGT1A1) is the only enzyme involved in the conjugation of bilirubin and the common UGT1A1*28 allele in the UGT1A1 gene, which is strongly associated with Gilbert's syndrome in Caucasians, results in elevated plasma bilirubin levels.

Aims: To test the hypothesis that the UGT1A1*28 allele is associated with lower disease susceptibility to, and disease behavior within, IBD. In addition, a possible altered risk for developing IBD-drug related side-effects was explored.

Methodology: Genomic DNA of 751 patients with IBD (209 patients with ulcerative colitis and 542 patients with Crohn's disease) and 930 healthy controls was genotyped for the UGT1A1*28 promoter polymorphism, and genotype distribution was compared between patients and controls. Genotype phenotype interactions were also investigated.

Results: Patients with Crohn's disease significantly less often bear the UGT1A1*28 homozygous genotype compared to the control group, with an odds ratio of 0.64, 95% CI: 0.42–0.98. The ulcerative colitis group showed no significant differences compared to controls.

Conclusion: The homozygous state of the UGT1A1*28 polymorphism, associated with higher serum bilirubin levels, may be protective for the development of Crohn's disease, suggesting that the anti-oxidant capacity of bilirubin may play a part.

Keywords
  • UDP-glucuronosyltransferase
  • Crohn's disease
  • Ulcerative colitis
  • Inflammatory bowel disease
  • Genetic polymorphism

1 Introduction

Over the last decades, extensive research has been performed to unravel the pathogenesis of inflammatory bowel disease (IBD), which encompasses both Crohn's disease (CD) and ulcerative colitis (UC). It appears that both genetic and environmental factors, including the enteric microbiota and food antigens as well as an altered innate and adaptive immune system, all may be involved.1 Oxidative stress is one of the pathways leading to cellular damage, which is also presumed to play an important role in the pathogenesis of IBD. Oxidative stress is the result of an imbalance between the production of reactive oxygen species (ROS), which are highly reactive due to the presence of unpaired electrons, and their removal by antioxidants.2 Notably of importance in IBD is the production of large amounts of ROS, known as respiratory burst, by neutrophils and macrophages after activation by pro-inflammatory agents like cytokines, immune complexes or bacterial products.3 In neutrophils derived from the intestinal mucosa of patients with IBD, an increased production of ROS has been observed.4 Furthermore in patients with IBD, intestinal subepithelial myofibroblasts exhibited an increased oxidative state related to elevated levels of interleukin-6.5 Oxidative stress might therefore be responsible for the increased synthesis of cytokines which accentuate and amplify the inflammatory state in patients with Crohn's disease.5 A recent study characterized the ROS generated in immune peripheral cells in patients with Crohn's disease, and showed a significant increase in H2O2, a potent ROS, in both active and remission phases of the disease, while anti-oxidative mechanisms catalyzing the decomposition of H2O2 were impaired.6

Intestinal damage in IBD has been suggested to be caused by an increased production of ROS and an impaired antioxidant capacity.2,7,8 Bilirubin, which is formed during the catabolism of heme, has for a long time been regarded only as a toxic waste product, causing jaundice or kernicterus when not properly cleared. However, bilirubin has since proved to be a powerful endogenous chain-breaking antioxidant, and a protective role of bilirubin has been suggested in coronary artery disease, respiratory diseases and cancer.914 A chronic mild form of unconjugated hyperbilirubinemia in the absence of liver disease or overt hemolysis is known as Gilbert's syndrome. In Caucasians, it is mainly caused by homozygosity for a polymorphism in the promoter region of the UTG1A1 gene, known as the UGT1A1*28 allele.15,16 Functional analyses of the transcriptional promoter activity demonstrated that the UGT1A1*28 allele reduces the transcription of the UGT1A1 gene up to 20% of normal.17 UGT1A1 is the only enzyme that catalyzes glucuronidation of bilirubin, which is the main determinant of elimination of bilirubin in humans.18 Genotypes including the UGT1A1*28 allele are associated with a reduced hepatic UGT1A1 enzyme activity up to 50%, with subsequent increased blood levels of unconjugated bilirubin, as compared with the most common UGT1A1*1 genotypes.15,19,20 Recent genome-wide association studies (GWAS) have also confirmed the substantial genetic influence of UGT1A1 variants on bilirubin levels.21,22

In 2006, Seiderer and co-workers reported on a patient suffering from both Crohn's disease and Gilbert's syndrome, who developed nodular regenerative hyperplasia following treatment with azathioprine, suggesting a predisposition of the UGT1A1*28 polymorphism for abnormal thiopurine interactions.23 In fact, the UGT1A1*28 allele is strongly associated with irinotecan toxicity, a drug often used in the treatment of metastatic colorectal cancer.24 We hypothesized that individuals bearing this allele, which is associated with reduced bilirubin conjugating enzyme activity and therefore increased bilirubin levels, are at a reduced risk for developing IBD. Since thiopurines (i.e. azathiopurine, 6-mercaptopurine and 6-thioguanine) are effective and widely used drugs in patients with IBD,25 we also aimed to explore the risk of thiopurine-induced side effects in patients bearing this polymorphism in a nested case approach.

2 Materials and methods

2.1 Patients

Patients with a diagnosis of IBD based on accepted clinical, endoscopic, radiological and histological findings,26 were recruited at the outpatient clinic of the Department of Gastroenterology, Radboud University Nijmegen Medical Center, being a tertiary referral center for IBD. A total number of 751 Caucasian patients (39% men, median age at diagnosis 26 years, range 4–76 years) were included, 542 patients with Crohn's disease (36% men, median age at diagnosis 24 years, range 4–76 years) and 209 patients with ulcerative colitis (48% men, median age at diagnosis 30 years, range 11–72 years). Furthermore, a total number of 930 healthy controls (43% men, median age 43 years, range 18–90 years), recruited from the referral region of our hospital, were included in the study. Blood samples were collected over a period of more than 10 years (1998–2010), while genotyping of all samples was performed in June 2010. Clinical characteristics of the patients, including disease extent according to the Montreal classification,27 family history of IBD, necessity of surgery and occurrence of extra-intestinal manifestations, were obtained at last follow up in 2010, as described before,28 and basic characteristics of both cases and controls are given in Table 1. Apart from clinical characteristics, we also collected data on thiopurine use and necessity to stop medication due to side effects. Side effects were categorized into gastro-intestinal intolerance (diarrhea, nausea and vomiting, abdominal pain), allergy (fever, skin rash, arthralgia/myalgia, Stevens Johnsons Syndrome), infection (Herpes zoster, fever of unknown origin, pneumonia), myelotoxicity (leukopenia and/or thrombocytopenia), hepatotoxicity (abnormal liver function tests) or pancreatic toxicity (abnormal pancreas enzymes, pancreatitis) and miscellaneous (fatigue, dizziness, generally unwell), based on laboratory results if appropriate. The mean follow up of the patient cohort with Crohn's disease was 17.4 years (standard deviation (SD) 10.6), while the mean follow up in the ulcerative colitis cohort was 15.1 years (SD 8.8). Prior to study recruitment, 2 patients had been diagnosed with Gilbert's syndrome, 1 patient with ulcerative colitis and 1 patient with Crohn's disease.

View this table:
Table 1

Basic characteristics of patients and controls.

Controls (n = 930)Crohn's disease (n = 542)Ulcerative colitis (n = 209)
Age, years (range)43 (18-90)44 (17-84)45 (19-82)
Sex
Male39643%20037%9746%
Female53457%34263%11254%

2.2 Genotyping

DNA of both patients and controls was isolated from whole blood by use of the Pure Gene DNA isolation kit (Gentra Systems, Minneapolis, MN).28 We determined the number of TA repeats in the promoter region of the UGT1A1 gene by using the polymerase chain reaction (PCR) conditions and primers as described before by Monaghan et al.16 Amplification was confirmed by agarose electrophoresis before fragments were resolved on 12% polyacrylamide gels (19:1 acrylamide/bisacrylamide; Biorad Laboratories, Veenendaal, The Netherlands) in Tris–borate–EDTA buffer. Electrophoresis was run at 400 V for 4 h and gels were stained with ethidium bromide for 15 min.19 Fragments of 98 bp indicate the UGT1A1*1 allele containing 6 TA repeats and fragments of 100 bp indicate the UGT1A1*28 allele, containing 7 TA repeats.29

2.3 Statistical analysis

The observed genotype frequencies were tested for deviation from the Hardy–Weinberg equilibrium using Fisher's exact test. Differences in UGT1A1 genotype distributions between patients and controls were determined by using the χ2 test, and odds ratios (ORs) with 95% confidence interval (95% CI) were calculated for genotypes associated with predicted normal, versus predicted altered UGT1A1 enzyme activities (variant genotypes). In addition, we calculated the generalized odds ratio (ORG), which utilizes the complete genotype distribution. It can be defined as the probability of a subject being more diseased relative to the probability of being less diseased, given that the more diseased subject has a higher mutational load.30 The ORG and 95% CI were calculated using the software “ORGGASMA” (downloaded from http://biomath.med.uth.gr). Lastly, we investigated whether an altered predicted enzyme activity influenced disease location or behavior, as well as for thiopurine side effects. Differences were calculated by using the χ2 test or Fisher's exact test when appropriate. Statistical analysis was performed using GraphPad Prism version 4.00 (GraphPad Software, San Diego, CA). The Bonferroni method was used to correct for multiple testing. All tables show the uncorrected p-values.

2.4 Ethics

The ethical committee of region Nijmegen and Arnhem reviewed and approved the protocol under number CWOM-nr 9804-0100. Verbal informed consent was obtained from each patient before study participation in agreement with the approval and all samples were anonymized. Since research data were collected anonymously, at least verbal informed consent was needed according to national regulations. Therefore, no written informed consent procedure was introduced at time of data collection.

3 Results

3.1 Distribution of the UGT1A1 genotypes

Distribution of the UGT1A1 genotypes in the control and patient groups, did fit the Hardy Weinberg equilibrium (p = 0.40 and p = 0.13, respectively). In patients with Crohn's disease, comparing the homozygous UGT1A1*28 and the homozygous UGT1A1*1 genotype, correlating with low and high bilirubin UGT enzyme activity respectively, revealed a significant difference (OR = 0.64, 95% CI 0.42–0.98; p = 0.04, Table 2). For patients with ulcerative colitis however, such difference was not observed. No differences were seen when Crohn's disease patients with the heterogenous genotype, corresponding with predicted intermediate enzyme activity, were compared to the homozygous UGT1A1*1 genotypes (Table 2). Furthermore, there was a decreased incidence of extra-intestinal manifestations in UGT1A1*28 allele barriers compared to wildtype carriers in patients with Crohn's disease (OR of 0.60, 95% CI 0.40–0.89; p = 0.009, Table 3), which however lost significance after Bonferroni correction (p = 0.09). The UGT1A1*28 allele had no effect on disease location or the need of disease related surgery in patients with ulcerative colitis (Table 4). The generalized odds ratio (ORG), which provides an estimate of the magnitude of the association between disease status and genotype, was calculated by comparing patients with Crohn's disease and controls and showed a non significant association (ORG = 0.83, 95% CI 0.69–1.01).

View this table:
Table 2

Distribution of the UGT1A1 genotypes and corresponding ORs in patients with IBD versus controls.

Genotype UGT1A1All patients with IBD (n = 751)p-valuePatients with CD (n = 542)p-valuePatients with UC (n = 209)p-valueControls (n = 930)
Number (%)OR (95% CI)Number (%)OR (95% CI)Number (%)OR (95% CI)Number (%)
*1/*1380 (50.6)Reference276 (50.9)Reference104 (49.8)Reference436 (46.9)
*1/*28319 (42.5)0.89 (0.73–1.09)0.28232 (42.8)0.89 (0.72–1.12)0.3487 (41.6)0.89 (0.65–1.22)0.47410 (44.1)
*28/*2852 (6.9)0.71 (0.49–1.03)0.0834 (6.3)0.64 (0.42–0.98)0.0418 (8.6)0.90 (0.52–1.56)0.7884 (9.0)

OR, Odds ratio; CI, confidence interval; IBD, inflammatory bowel disease; UC, ulcerative colitis; CD, Crohn's disease.

    View this table:
    Table 3

    UGT1A1 genotype–phenotype correlations in patients with Crohn's disease#.

    *1/*1 (n = 268)*1/*28 (n = 223)*28/*28 (n = 27)Odds ratio (95%CI)p-value
    Disease localization
    Ileal (L1)89839Reference
    Colonic (L2)714630.67 (0.42–1.06)0.09
    Ileocolonic (L3)10288150.98 (0.65–1.46)0.91
    Isolated upper disease+ (L4)151541.23 (0.59–2.56)0.59
    Disease behavior
    Non stricturing/Penetrating (B1)85699Reference
    Stricturing (B2)454201.02 (0.60–1.71)0.95
    Penetrating (B3)132106181.02 (0.69–1.52)0.91
    Age at diagnosis
    < 16 years (A1)24281Reference
    Between 17 and 40 years (A2)208171230.77 (0.43–1.37)0.38
    Above 40 years (A3)362170.64 (0.31–1.34)0.24
    Surgery164132190.97 (0.68–1.38)0.85
    Extra-intestinal92/24654/2016/260.60 (0.40–0.89)0.01
    Family history of IBD38/13231/1084/200.93 (0.54–1.60)0.80

    #The UGT1A1*1/*1 genotypes were compared with the patients bearing the UGT1A1*28 variant (both homo- and heterozygous).

    +Patients were classified as having disease localization in the upper gastrointestinal tract next to ileal, ileo-colonic or colonic localization.

    According to the Montreal Classification.

    Significant p-values are given in bold letters including corresponding odds ratios (OR) and 95% confidence intervals (CI). However, none of these p-values remained significant after Bonferroni correction (p > 0.05).

      View this table:
      Table 4

      UGT1A1 genotype–phenotype correlations in patients with ulcerative colitis.

      *1/*1*1/*28*28/*28Odds ratio (95%CI)p-value
      Disease localization (n = 176)
      Proctitis542Reference
      Left sided252961.17 (0.32–4.25)0.82
      Pancolitis573990. 70 (0.20–2.44)0.58
      Surgery (n = 194)30/9623/804/180.84 (0.45–1.55)0.58

      #The UGT1A1*1/*1 genotypes were compared with the patients bearing the UGT1A1*28 variant (both homozygous and heterozygous).

      According to the Montreal Classification.

        3.2 Possible relationship between UGT1A1 genotype and thiopurine-induced side effects

        We subsequently explored the possible relationship between the UGT1A1*28 allele and thiopurine induced side effects. Data on thiopurine use were available for 585 patients, 407 patients with Crohn's disease and 178 patients with ulcerative colitis. Of these patients, 343 patients used azathioprine, 53 patients used azathioprine in combination with 6-mercaptopurine (6-MP), 17 patients used azathioprine in combination with 6-thioguanine (6-TG), and 9 patients used azathioprine in combination with both 6-MP and 6-TG. A total of 126 thiopurine side effects were observed, 95 in patients with Crohn's disease and 31 in patients with ulcerative colitis. However, no significant association was observed between thiopurine related side effects and the presence of the UGT1A1*28 allele (OR = 1.17, 95% CI 0.77–1.79; p = 0.46). However, when subgroups were analyzed, gastrointestinal intolerance to thiopurine therapy was associated with the UGT1A1*28 allele (OR 2.26, 95% CI 1.11–4.62; p = 0.02, see Table 5). After Bonferroni correction however, this p-value did not remain significant (p = 0.18).

        View this table:
        Table 5

        Thiopurine side effects stratified by UGT1A1 genotype.

        *1/*1*1/*28*28/*28Odds ratio (95% CI)p-value
        All patients20818432
        No adverse event14312618Reference
        Total adverse events6558141.10 (0.73–1.65)0.65
        Gastrointestinal intolerance122422.15 (1.05–4.43)0.03
        Allergy241750.91 (0.49–1.70)0.77
        Infection3010.33 (0.03–3.22)0.62
        Myeloxicity9610.77 (0.28–2.13)0.62
        Hepatotoxicity8430.87 (0.31–2.46)0.79
        Pancreas toxicity5601.19 (0.36–3.99)1.00
        Miscellaneous4120.75 (0.16–3.39)1.00

        #The UGT1A1*1/*1 genotypes were compared with the genotypes bearing the UGT1A1*28 variant (both homozygous and heterozygous).

          4 Discussion

          In our cohort, homozygosity for the UGT1A1*28 allele, which correlates with low enzyme activity, exerted a protective effect against Crohn's disease. The 9.0% prevalence of homozygosity for the UGT1A1*28 allele in our control population is lower compared to homozygosity rates of 12% reported in other European studies,16,31 which might even more emphasize the significance of our finding. A detailed genotype–phenotype analysis revealed a weak association of the UGT1A1*28 allele with a decreased incidence of extra-intestinal manifestations, which however did not remain significant after Bonferroni correction. The results of our study could imply that an increased blood level of unconjugated bilirubin confers a protective effect on the development of Crohn's disease. In accordance with this hypothesis, Koutroubakis and co-workers showed that patients with inflammatory bowel disease had a decreased serum antioxidant capacity as compared to healthy controls, with bilirubin acting as one of the major endogenous anti-oxidants accounting for this antioxidant capacity.32 Furthermore, several animal studies have indicated a protective role of the products formed during heme catabolism, including heme oxygenase (HO), carbon monoxide (CO) and biliverdin–bilirubin, in the pathogenesis of inflammatory bowel disease.3335 Decreased levels of bilirubin have also been associated with other auto-inflammatory diseases in which oxidative stress also seems to pathogenically involved, including rheumatoid arthritis and systemic lupus erythematosus (SLE).3639

          In the catabolism of heme, the stress responsive enzyme HO-1 is involved, which together with HO-2 are the rate limiting enzymes in the formation of biliverdin, which successively is converted into bilirubin, free iron and CO.40 In an experimental colitis model administering dextran sodium sulfate (DSS) to rats, Berberat et al. showed that induction of intestinal HO-1 ameliorates manifestations of DSS induced colitis. Daily intraperitoneal injections of biliverdin resulted in a comparable protective effect, indicating that the generation of bilirubin by the up-regulation of HO-1 may be the key mediator in suppressing inflammatory damage in acute colitis.41 In rats treated with trinitrobenzene sulphonic acid (TNBS) to induce colitis, increased expression and activity of HO-1, as measured by formation of bilirubin, was observed after TNBS challenge.42

          In this study we explored the possible link between Gilbert's syndrome and the development of thiopurine-induced side effects, by investigation the association of the UGT1A1*28 allele and thiopurine related side effects.23 No association was found; while a sub analysis revealed that bearing the UGT1A1*28 allele was associated with gastrointestinal intolerance for thiopurines, however this association did not remain significant after correction for multiple comparisons. Furthermore, this nominally significant finding might also be explained as a by chance finding due to the small groups analyzed. Thiopurine prodrugs are converted intracellularly into cytotoxic 6-thioguanine nucleotides (6-TGNs) to exert their therapeutic effect, but UGTs seem to be not involved in the metabolism of the prodrugs. Therefore it is unclear how UGT1A1 might be involved in the occurrence of gastrointestinal intolerance. An enzyme which certainly is involved in the catabolism of thiopurines, the thiopurine S methyltransferase (TPMT), is a major determinant for 6-TGNs levels.43 High levels of 6-TGNs are associated with thiopurine side effects and different variants of TPMT exist which may result in altered TPMT activity. A recent meta-analysis showed an association between TPMT polymorphisms and thiopurine side effects in patients with IBD.44 However, not all side effects of thiopurine therapy could be explained by TPMT polymorphisms, which implies that other mechanisms, unrelated to TPMT, might play a role.45 A limitation of our study might be that we did not measure blood, serum or plasma levels of bilirubin in our patients, but only used the UGT1A1*28 allele as an indirect parameter for hyperbilirubinemia. However, multiple studies have confirmed the relationship between the UGT1A1*28 polymorphism and elevated plasma/serum/blood levels of bilirubin.16,18,46

          Homozygosity for the UGT1A1*28 allele exerted a protective effect against the development of Crohn's disease, as well as on the development of extra-intestinal manifestations of Crohn's disease in a Dutch cohort of IBD patients. Most likely this protective effect is due to the antioxidant capacity of the slightly elevated bilirubin levels. If firmly established, this might possibly lead to new strategies in the prevention or support treatment of Crohn's disease by supplementation with antioxidants.

          References

          View Abstract