The pregnane X receptor gene (PXR/NR1I2) regulates an array of genes involved in the response to xenobiotics.1,2 Dysregulation of this gene may critically influence intestinal barrier defence and susceptibility to inflammatory bowel disease (IBD).3 Recent data from Ireland have suggested strong associations between polymorphisms within the PXR/NR1I2 gene and IBD. Dring et al performed a case control study involving 422 patients with IBD (185 ulcerative colitis (UC) and 237 (Crohn's disease (CD)) and 350 healthy controls, using eight candidate polymorphisms in this gene.4 Highly significant associations were demonstrated with UC, CD, and IBD as a whole. This effect was most significant for the two individual single nucleotide polymorphisms (SNPs) in the promoter region of this gene; compared between the IBD cohort and controls, rs3814055/−23585 (p = 0.000008; odds ratio (OR) 1.62 (95% confidence interval (CI) 1.31–2.00)) and rs1523127/−24381 (p = 0.0002; OR 1.50 (95% CI 1.21–1.84)).
We have critically re‐evaluated the contribution of these allelic variants of rs1523127/−24381 of the PXR/NR1I2 gene as determinants of disease susceptibility and phenotype in the Scottish population. In addition, we also performed a gene wide association study using a haplotype tagging strategy to assess in detail the overall contribution of this gene to disease susceptibility. A total of 387 UC and 328 CD patients, together with 338 healthy controls (HC), were studied. This study was designed to have 98% power to replicate the previous association with the rs1523127/−24381 variant (p<0.05). In the haplotype analysis, five tagging SNPs (tSNPs) were selected using a multimarker criterion, haplotype r2>0.80 to predict all SNPs/haplotypes. This approach was described by Weale and Goldstein and was successfully applied in our previous study of the ABCB1/MDR1 gene.5,6,7,8 The exons, promoter region, and intronic boundaries were resequenced in 16 unrelated Centre d'Etude du Polymorphisme Humain (CEPH) individuals using 10 randomly spaced amplicons across the PXR/NR1I2 gene. Haplotypes and their respective frequencies were constructed by further genotyping in 32 CEPH trios.
No association was seen between rs1523127/−24381 (r2 = 0.96 with rs3814055/−23585) SNP and UC, CD, or IBD (A‐allelic frequency: 60.5% UC, 59.5% CD, 60.0% IBD, and 60.6% HC; p = 0.96, 0.69, and 0.82, respectively) (table 1). The significant rs6785049 (−7635) variant used in Dring's study was in strong linkage with one of the tSNPs used in this study (rs2472682, r2 = 0.86). In the Scottish dataset, this SNP (rs2472682) was not associated with IBD (T‐allele: p = 0.97, OR 1.00 (95% CI 0.82–1.22); TT genotype: p = 0.82, OR 0.96 (95% CI 0.62–1.47)). Log likelihood analyses comparing overall haplotypic distribution of the five tSNPs demonstrated no associations with UC, CD, or IBD (p = 0.90, 0.90, and 0.79, respectively). There were no associations observed between each of the five tSNPs and the common constructed haplotypes with UC, CD, and IBD, respectively. Genotype‐phenotype analyses using the Montreal classification did not show any associations with the studied variants or haplotypes (full data available on request).
Table 1 Allelic and genotype frequencies of the selected five tagging single nucleotide polymorphisms (tSNPs).
| dbSNP ID position | Allele (1/2) | UC | CD | IBD | HC | UC v HC 1 v 2 p value odds ratio 95% CI | UC v HC 1/1 v 2/2 p value odds ratio 95% CI | CD v HC 1 v 2 p value odds ratio 95% CI | CD v HC 1/1 v 2/2 p value odds ratio 95% CI | IBD v HC 1 v 2 p value odds ratio 95% CI | IBD v HC 1/1 v 2/2 p value odds ratio 95% CI |
|---|---|---|---|---|---|---|---|---|---|---|---|
| rs1523127 120983729 | AA | 139 (35.9%) | 102 (31.1%) | 241 (33.7%) | 119 (35.6%) | 0.96 | 0.91 | 0.69 | 1.00 | 0.82 | 1.00 |
| AG | 190 (49.1%) | 186 (56.7%) | 376 (52.6%) | 167 (50.0%) | 0.99 | 0.97 | 0.95 | 1.03 | 0.97 | 0.99 | |
| GG | 58 (15.0%) | 40 (12.2%) | 98 (13.7%) | 48 (14.4%) | 0.80–1.22 | 0.61–1.52 | 0.76–1.19 | 0.62–1.69 | 0.81–1.17 | 0.66–1.49 | |
| A | 468 (60.5%) | 390 (59.5%) | 858 (60.0%) | 405 (60.6%) | |||||||
| G | 306 (39.5%) | 266 (40.5%) | 572 (40.0%) | 263 (39.4%) | |||||||
| rs2461823 121002815 | CC | 152 (40.9%) | 132 (40.2%) | 284 (40.6%) | 119 (35.8%) | 0.38 | 0.71 | 0.46 | 0.80 | 0.38 | 0.66 |
| CT | 175 (47.0%) | 157 (47.9%) | 332 (47.4%) | 174 (52.4%) | 1.11 | 1.11 | 1.09 | 1.10 | 1.09 | 1.10 | |
| TT | 45 (12.1%) | 39 (11.9%) | 84 (12.0%) | 39 (11.7%) | 0.88–1.37 | 0.68–1.81 | 0.87–1.37 | 0.66–1.82 | 0.90–1.32 | 0.71–1.70 | |
| C | 479 (64.4%) | 421 (64.2%) | 900 (64.3%) | 412 (62.0%) | |||||||
| T | 265 (35.6%) | 235 (35.8%) | 500 (35.7%) | 252 (38.0%) | |||||||
| rs7643645 121008187 | TT | 143 (39.7%) | 128 (43.2%) | 271 (41.3%) | 128 (38.5%) | 0.79 | 0.90 | 0.82 | 0.62 | 0.73 | 0.91 |
| TC | 172 (47.8%) | 120 (40.5%) | 292 (44.5%) | 168 (50.6%) | 1.04 | 1.04 | 1.04 | 0.87 | 1.04 | 0.96 | |
| CC | 45 (12.5%) | 48 (16.2%) | 93 (14.2%) | 42 (12.6%) | 0.84–1.29 | 0.64–1.69 | 0.82–1.30 | 0.54–1.42 | 0.86–1.26 | 0.63–1.46 | |
| T | 458 (63.6%) | 376 (63.5%) | 834 (63.6%) | 424 (62.7%) | |||||||
| C | 262 (36.4%) | 216 (36.5%) | 478 (36.4%) | 252 (37.3%) | |||||||
| rs1464603 121009039 | AA | 172 (45.9%) | 153 (46.6%) | 325 (46.3%) | 167 (49.4%) | 0.61 | 1.00 | 0.72 | 1.00 | 0.62 | 1.00 |
| AG | 159 (42.5%) | 138 (42.1%) | 297(42.3%) | 130 (38.5%) | 0.94 | 0.98 | 0.96 | 1.01 | 0.95 | 0.99 | |
| GG | 43 (11.5%) | 37 (11.3%) | 80 (11.4%) | 41 (12.1%) | 0.75–1.17 | 0.61–1.58 | 0.76–1.21 | 0.62–1.67 | 0.78–1.15 | 0.65–1.52 | |
| A | 503 (67.2%) | 444 (67.7%) | 947 (67.5%) | 464 (68.6%) | |||||||
| G | 245 (32.8%) | 212 (32.3%) | 457 (32.5%) | 212 (31.4%) | |||||||
| rs2472682 121015342 | TT | 42 (11.0%) | 35 (11.1%) | 77 (11.0%) | 39 (11.7%) | 0.65 | 1.00 | 0.63 | 0.70 | 0.97 | 0.82 |
| TG | 177 (46.2%) | 129 (40.8%) | 306 (43.8%) | 140 (42.2%) | 1.06 | 1.00 | 0.94 | 0.90 | 1.00 | 0.96 | |
| GG | 164 (42.8%) | 152(48.1%) | 316 (45.2%) | 153(46.1%) | 0.84–1.32 | 0.62–1.64 | 0.74–1.19 | 0.54–1.50 | 0.82–1.22 | 0.62–1.47 | |
| T | 261 (34.1%) | 199 (31.5%) | 460 (32.9%) | 218 (32.8%) | |||||||
| G | 505 (65.9%) | 433 (68.5%) | 938 (67.1%) | 446 (67.2%) |
UC, ulcerative colitis; CD, Crohn's disease; HC, healthy controls; 95% CI, 95% confidence interval.
These negative data, based on a true candidate gene approach, combine the use of the single marker and multimarker haplotype tagging approach and set a statistical limit to the importance of the contribution of this gene to disease susceptibility in our population. Discordance between the findings of this study and those data presented in Dring's study require clarification and explanation. When the allelic frequencies and homozygosity rates of the rs1523127/−24381 SNP in these two studies are compared, it is of interest to note that differences in the conclusions between these studies may be driven primarily by differences in allelic and genotypic frequencies in healthy controls studied in these datasets (Scottish v Irish controls: A‐allele frequencies: 60.6% v 54.8%, respectively, p = 0.03, OR 1.27 (95% CI 1.02–1.58); AA‐genotype frequencies: 35.6% v 32.2%, respectively, p = 0.01, odds ratio 1.75 (95% CI 1.12–2.72)). Differences in frequencies in the Scottish and Irish patient populations were less marked (Scottish v Irish cases: A‐allele frequencies: 60.0% v 64.3%, p = 0.05, OR 1.20 (95% CI 1.00–1.44); AA‐genotype frequencies: 33.7% v 41.1%, p = 0.17, odds ratio 1.33 (95% CI 0.90–1.98)). Thus potentially a number of explanations exist, including cryptic population stratification, genetic heterogeneity, or a type 1 error in the initial dataset, each of which may explain the divergent findings in these two populations, in whom previous data had suggested consistent founder effects.9 Further studies in these and other Northern European populations may help to resolve this issue formally.
Footnotes
Dr Gwo‐Tzer Ho was supported by the Chief Scientist's Office, Scottish Executive, UK (2001‐04). Dr Albert Tenesa is funded by Cancer Research UK grant C348/A3758. Dr Elaine Nimmo and Mrs Hazel Drummond are supported by a program grant from the Wellcome Trust, C72789/Z/03/Z.
Conflict of interest: None declared.
References
- 1.Willson T M, Kliewer S A. PXR, CAR and drug metabolism. Nat Rev Drug Discov 20021259–266. [DOI] [PubMed] [Google Scholar]
- 2.Kliewer S A, Goodwin B, Willson T M. The nuclear pregnane X receptor: a key regulator of xenobiotic metabolism. Endocr Rev 200223687–702. [DOI] [PubMed] [Google Scholar]
- 3.Langmann T, Moehle C, Mauerer R.et al Loss of detoxification in inflammatory bowel disease: Dysregulation of pregnane X receptor target genes. Gastroenterology 200412726–40. [DOI] [PubMed] [Google Scholar]
- 4.Dring M M, Goulding C A, Trimble V I.et al The Pregnane X receptor locus is associated with susceptibility to inflammatory bowel disease. Gastroenterology 2006130341–348. [DOI] [PubMed] [Google Scholar]
- 5.Goldstein D B, Ahmadi K R, Weale M E.et al Genome scans and candidate gene approaches in the study of common diseases and variable drug responses. Trends Genet 200319615–622. [DOI] [PubMed] [Google Scholar]
- 6.Weale M E, Depondt C, Macdonald S J.et al Selection and evaluation of tagging SNPs in the neuronal‐sodium‐channel gene SCN1A: implications for linkage‐disequilibrium gene mapping. Am J Hum Genet 200373551–565. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Ho G T, Soranzo N, Nimmo E R.et al ABCB1/MDR1 gene determines susceptibility and phenotype in ulcerative colitis: discrimination of critical variants using a gene‐wide haplotype tagging approach. Hum Mol Genet 200615797–805. [DOI] [PubMed] [Google Scholar]
- 8.Ahmadi K R, Weale M E, Xue Z Y.et al A single‐nucleotide polymorphism tagging set for human drug metabolism and transport. Nat Genet 20053784–89. [DOI] [PubMed] [Google Scholar]
- 9.Arnott I D, Nimmo E R, Drummond H 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 20045417–425. [DOI] [PubMed] [Google Scholar]