Crohn's disease (CD) is a chronic inflammatory disorder, for which strong evidence of genetic susceptibility associates with specific chromosomal locations, including IBD5 on chromosome 5q31 (MIM# 606348)(Reference Rioux, Silverberg and Daly1). This region contains the Na-dependent organic cation transporter (OCTN1; alternatively called SLC22A4). A functional single nucleotide polymorphism L503F in exon 9 of OCTN1 (c. 1672 C>T) causes an amino acid substitution in the coding region of this gene, and the variant allele was claimed to be implicated in CD causality in a European population(Reference Peltekova, Wintle and Rubin2). However, this observation has not consistently replicated across different population groups, including a preliminary study in Canterbury, New Zealand(Reference Silverberg3, Reference Ferguson, Han and Huebner4). We suggest that such apparent contradictions of disease risk factors across different population groups might result from a gene–diet interaction.
The same dietary item that improves the condition in one individual with CD may exacerbate it in another(Reference Riordan, Hunter and Cowan5). For example, whereas most individuals show health benefits from eating a wide range of vegetables, this is not necessarily true for all individuals with CD or other forms of inflammatory bowel disease(Reference Jowett, Seal and Pearce6).
The main physiological substrate of the OCTN1 is l-ergothionine (ET), which is exclusively synthesised by certain soil fungi and bacteria, and sequestered by plants. Taubert et al. (Reference Taubert, Grimberg and Jung7) demonstrated that the L503F single nucleotide polymorphism in the OCTN1 gene leads to a gain of function polymorphism, resulting in a 3-fold higher substrate affinity and 50 % higher initial transport capacity at nanomolar levels of substrate. Because ET concentrations found in human tissues are in the nanomolar to micromolar range, these authors suggested that carriers of the variant L503F allele could accumulate higher ET concentrations in OCTN1-expressing cells. Thus, any potential gene–diet interactions with OCTN1 might be anticipated to involve foods with high levels of ET. The present preliminary (negative) New Zealand case–control study on OCTN1 and CD risk(Reference Ferguson, Han and Huebner4) has been extended to include a wider population group. We stratify their responses to vegetables, and report a novel gene–diet interaction in this New Zealand CD cohort.
Materials and methods
Cohort
Human subjects were a subset of a population-based inflammatory bowel disease cohort from Canterbury, New Zealand(Reference Gearry, Richardson and Frampton8), volunteers from Auckland District Health Board gastroenterology clinics or healthy volunteers from either area(Reference Burt, George and Upton9). Diagnosis of CD was clinically confirmed using Montreal classification criteria(Reference Sasieni10). Their phenotypes are described previously(Reference Gearry, Richardson and Frampton8, Reference Sasieni10). The study was conducted under ethical protocol MEC/04/12/011, authorised through the New Zealand Multi-Region Human Ethics Committee.
Applied Biosystems TaqMan®single nucleotide polymorphism genotyping
DNA was extracted from peripheral blood lymphocytes using standard methods. The single nucleotide polymorphism at position c. 1672 C>T (L503F) rs1050152 was genotyped using the TaqMan MGB diallelic discrimination system, and probes and oligonucleotides were obtained from Applied Biosystems (Foster City, CA, USA) using the Assay-by-Design product. Details can be provided on request.
Dietary questionnaire
A subject's reaction to each of a list of forty-four individual vegetables was scored on a five-point ordinal scale: ++ ‘definitely better’ to − − ‘definitely worse’. Data for which responses were ‘don't know’ or ‘makes no difference’ were not utilised in the analysis. The response of an individual to each vegetable was summarised by two numbers:
1 the percentage of individuals reporting ++ or + ‘beneficial’ responses;
2 the percentage of individuals reporting − − or − ‘adverse’ responses.
This allowed the data to be represented in two dimensions.
Statistical analysis
The allelic trend test(Reference Silverberg, Satsangi and Ahmad11) and Fisher's exact genotypic test were used to compare case and control allele frequencies of the OCTN1 variant using the Statistical Analysis System software package version 9.1 (SAS Institute Inc., Cary, NC, USA). An exact test was used to test for departures from Hardy–Weinberg equilibrium(Reference Wigginton, Cutler and Abecasis12). The dietary questionnaire data were analysed as percentages of individuals reporting a beneficial or an adverse response, as shown in Fig. 2. The percentage data are represented on a logit scale to stabilise variance of proportions. For data from the two extreme vegetables – mushroom and maize – the regression of the proportion reporting extreme responses in each genotype against the number of variant (T) alleles was performed using logistic regression. As reported, the data for mushrooms but not maize showed statistically significant evidence for an interaction (P < 0·025).
Results and discussion
Four hundred and ninety-nine New Zealand Caucasian subjects with CD and 370 controls provided DNA samples for OCTN1 genotyping. The case–control analysis of the present study database showed no significant differences in disease risk with increased frequency of the OCTN1 L503F variant genotype (data not shown). A significant proportion of the New Zealand population is Caucasian, typically being descendants of migrants from the UK. The present observation confirms data from our preliminary study(Reference Ferguson, Han and Huebner4), suggesting that variants in this gene are not associated with CD risk in this New Zealand group. This is despite their genetic background being comparable to that originally described by Rioux et al. (Reference Rioux, Silverberg and Daly1).
Crohn's disease subjects also completed a dietary questionnaire that characterised foods as alleviating symptoms, exacerbating symptoms or having no effect. The data from forty-four vegetable items were thus categorised, and the percentage of CD cases reporting benefits was plotted against the percentage showing adverse effects for each individual vegetable (Fig. 1). Two items stand out. Maize and mushrooms have high proportions reporting adverse effects, 49 and 39 %, respectively, and low proportions reporting beneficial effects, 1·8 and 2·0 % respectively.
Fig. 1 Percentage of Crohn's disease cases reporting ++ or + ‘beneficial’ responses compared with percentage reporting − − or − ‘adverse’ responses to forty-four vegetables. Key to vegetables in numeric order: maize; mushrooms; potato chips; leeks; fried potatoes; baked beans; chick peas; dried beans; green banana; onions; peas; celery; garlic; cucumber; raw tomatoes; tomato sauce; cabbage; gherkins; roasted potatoes; avocado; tomato puree; eggplant; silverbeet or spinach; broccoli or cauliflower; asparagus; salad greens; green beans; capsicum; cooked tomato; puha; taro; Chinese greens; Jerusalem artichokes; cassava; potatoes baked in skin; beetroot; lentils; watercress; parsnip; carrot; kumara; boiled potatoes; yams; pumpkin.
It is noteworthy that ET occurs at higher levels in mushrooms than in other major items in a common diet(Reference Ey, Schomig and Taubert13). These data led us to speculate that individuals carrying the variant OCTN1 allele could show enhanced sensitivity to ET-containing foods, especially mushrooms. Fig. 2 shows the frequency of individuals carrying different OCTN1 genotypes in relation to intolerance towards mushrooms. Although carrying a variant OCTN1 gene does not adversely affect most individuals on a normal diet, this variant confers sensitivity to mushrooms in CD cases (P < 0·025). Those individuals reporting mushroom intolerance show a strong bias towards carrying the variant allele. This statistically significant interaction was not seen for maize (P = 0·07), which does not contain ET.
Fig. 2 Frequency of Crohn's disease cases with specific genotypes reporting adverse effects from mushrooms.
Since ET is an antioxidant, high levels of ET might lead to an antioxidant overload in red blood cells or epithelium, leading to pro-oxidant effects and/or imbalance in immune reaction. An alternative possibility might involve effects on pro-inflammatory cytokines or heat shock protein 70(Reference Sakrak, Kerem and Bedirli14).
We believe that this is the first diet–gene interaction described in CD. Although this dietary sensitivity may be known to symptomatic CD subjects, the data begin to position us towards utilising genotyping as a diagnostic tool upon which to base informed dietary advice. It may be possible to use nutrigenomics principles, in order to develop foods especially suited to this group of individuals.
Acknowledgements
We thank Karen Munday and Philippa Dryland for their help with subject recruitment, Virginia Parslow for finalising the diagrams and Martin Philpott and Sarah Johnson for their helpful discussions. We also thank all the participants in the present study. Nutrigenomics New Zealand is a collaboration between AgResearch Ltd, Crop and Food Research, HortResearch and The University of Auckland. The programme is primarily funded through the Foundation for Research Science and Technology (Contract C02X0403). Other funding sources for the case and control DNA collection are the Canterbury Medical Research Foundation, Canterbury Liver and Digestive Diseases Trust and Christchurch Gastroenterology Research Trust. The present study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving human subjects/patients were approved by the Multi-Region Ethics Committee, Ministry of Health, New Zealand (MEC/04/12/011). Written informed consent was obtained from all subjects/patients. I. P., C. H. and L. R. F. designed research; C. H. and I. P. performed research; R. B. G. and M. L. B. contributed human subjects and DNA samples; P. S. D. and A. M. contributed analytic tools; C. M. T. and D. Y. H. analysed data; L. R. F. wrote the paper. None of the authors had any conflict of interest.
