All New Zealand study participants gave their informed written consent and approval for the study was obtained from the Upper and Lower South Regional Ethics Committees of New Zealand (Approval ID: CTY/03/01/011; Date: 05/06/2007). For the meta-analysis, genotype data from non-New Zealand subjects were obtained from review of published studies (Table 1).

Study participants were selected from a population-based study of genetic and environmental determinants of the aetiology of IBD in the Canterbury region of New Zealand which has been described in detail elsewhere[20]. The population of Canterbury is largely of Northern European (United Kingdom and Irish) origin and participants were included in the current study if they were of self-reported European ancestry. Diagnosis of IBD was made by standard criteria[21]. Detailed phenotypic data according to the Montreal Classification system were available in addition to information regarding the presence of extra-intestinal manifestations of disease, history of immuno-modulator use and the need for surgery[21]. The control group comprised 540 healthy Caucasian New Zealanders over the age of 17 years with no history of inflammatory disorders[22].

DNA was collected from peripheral blood samples of the IBD patients and controls using guanidinium isothiocyanate-chloroform extraction[23]. Genotyping of MIF -173G > C (rs755622) was performed using a pre-designed Taqman^® SNP genotyping assay (assay ID: C_2213785_10) as per the manufacturer’s instructions (Applied Biosystems, Foster City, CA, United States). The MIF CATT_5-8 repeat, was amplified in a total volume of 10 μL containing 2 mM dNTPs (Fisher Biotec), 2 mM MgCl₂, 10% betaine (Sigma, St Louis, Missouri, United States), 0.5 μmol/L of the primers MIFrepf (5’FAM-gcctgtgatccagttgctgccttgtc3’) and MIFrepr (5’ccactaatggtaaactcggggaccat3’), 1 U of Platinum Taq DNA Polymerase (Invitrogen, CA, United States), and 20 ng of genomic DNA. The forward primer, MIFrepf, was labelled to enable subsequent resolution on a DNA Analyzer (Applied Biosystems 3730xl DNA Analyzer). The polymerase chain reaction (PCR) conditions comprised an initial denaturation step of 2 min at 94 °C; followed by 30 cycles of 94 °C for 30 s, 65 °C for 30 s, 72 °C for 30 s; and a final extension step of 1 min at 72 °C. PCRs were diluted by the addition of 20 μL of water. Amplification of the MIF promoter was assessed by 3% agarose gel electrophoresis. To accurately size the PCR products and thus reliably assign CATT genotypes, 2 μL of each diluted PCR product was mixed with 1 μL of GeneScan™ 500 LIZ^® Size Standard (Applied Biosystems, Foster City, CA, United States) and resolved on an ABI 3730xl DNA Analyzer. Samples containing 5, 6, 7, or 8 CATT repeats yielded PCR products of 152, 156, 160 and 164 bp, respectively. Genotyping results were then analysed using Peak Scanner™ Software version 1 (Applied Biosystems, Foster City, CA, United States). Accuracy of the TaqMan SNP and CATT assays was confirmed by repeat analysis of 10% of samples. The concordance between original and repeat genotypes was 100%.

The software package PLINK[24] was used to test for deviations from Hardy-Weinberg Equilibrium (HWE), and to conduct single marker and haplotype association tests. The statistical significance of the observed allele and genotype associations were determined by Pearson’s χ². Results were considered statistically significant if P was < 0.05. Bonferroni’s correction was applied to adjust for multiple testing (0.05/18 tests, adjusted P < 0.0028). As CATT₅ is reported to have less promoter activity than longer repeat alleles[9,25], tests of association were performed on full genotype and following simplification to 55, 5X or XX, where X = CATT_6-8, as previously described[9,26,27].

All published studies of MIF -173G > C and MIF CATT_5-8 in IBD were identified using PubMed and subsequently searching the references of all PubMed-identified studies. Imputed genotypes for MIF -173G > C were generated from the Wellcome Trust Case Control Consortium (WTCCC)[28] and National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Crohn’s disease (CD) datasets[29,30] using IMPUTE and the HapMap release [NCBI dbSNP Build 131 (Apr 2010, hg37.1)]. A quality threshold of 0.95 was set for imputation of both datasets.

Meta-analyses were conducted for the MIF SNP -173G > C in STATA (Stata Statistics/Data Analysis Software, version 8, Stata Corporation, Texas, United States) using the Mantel-Haenszel (MH) method with either a random or a fixed effects model depending on the presence or absence respectively of significant heterogeneity between studies (P < 0.05). Meta-analyses were considered significant if the pool MH P value was < 0.05. As published CATT_5-8 frequency data was limited to a single Japanese ulcerative colitis (UC) dataset[18], meta-analysis was not conducted for this variant.

Nine hundred and fifty-seven New Zealand patients (CD 467, UC 464, IBD-U 26) and 483 New Zealand controls were successfully genotyped for both MIF promoter polymorphisms. Haplotype analysis of these polymorphisms found the MIF CATT₅/-173C haplotype occurred at a higher frequency in controls compared to IBD patients (0.6 vs 0.01; P = 0.03, OR = 0.22, 95%CI: 0.05-0.99). However, after bonferroni correction for multiple testing (0.05/18 tests) no association of haplotype with susceptibility to CD, UC or IBD, or with disease phenotype, behaviour or clinical course was found.

Meta-analysis was undertaken in order to determine the overall influence of MIF -173G > C on IBD susceptibility in Caucasians and East Asians. Published summary genotype data were available for five Caucasian[12,14,16,17], and three East Asian datasets[13,15,18]. In addition, MIF-173G > C genotypes were imputed from both the WTCCC and NIDDK Caucasian CD datasets. As the contribution that specific loci make to IBD susceptibility differs significantly with race, the East Asian and Caucasian datasets were meta-analysed separately. The Breslow-Day test revealed the existence of significant heterogeneity (P_het≤ 0.05) among the Caucasian datasets but not among the East Asian datasets (P_het = 0.5). As a result a random effects model was applied to the Caucasian meta-analysis whilst a fixed effects model was applied to the East Asian meta-analysis. No cumulative evidence of association of MIF -173G > C with UC, CD or all IBD risk was detected in either Caucasians or East Asians (Figure 1).

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Meta-analyses of migration inhibitory factor-173G > C in Caucasian and East Asian inflammatory bowel disease datasets. Meta-analyses of migration inhibitory factor (MIF)-173G > C were performed by the Mantel-Haenszel method using a random effects model. On each Forest plot the 95%CI of the individual datasets are represented by horizontal lines and the total number of study participants in each dataset is proportional to the size of the square. The diamond represents the pooled OR with 95%CI delineated by the diamond’s width. A: The combined Caucasian ulcerative colitis (UC) dataset; B: The combined East Asian UC datasets; C: The combined Caucasian Crohn’s disease dataset had 100% power to detect an effect size of OR = 1.40 at α = 0.05.

MIF is a pleiotropic pro-inflammatory cytokine implicated in the pathophysiology of numerous diseases including IBD[5,9,31-33]. Promoter polymorphisms in MIF influence both basal and disease-associated MIF expression, and MIF expression has been shown to have profound effects on disease phenotype in experimental models[5,7,34]. In vitro MIF -173C allele exhibits greater expression than the MIF -173G allele in T lymphoblast cells, and expression of MIF increases with increasing length of the CATT repeat in COS-7 fibroblast like cells[9]. Despite these effects on expression, previous association studies investigating these polymorphisms in IBD have given discordant results[12-19].

The MIF-173C allele has been significantly associated with increased UC susceptibility in Spanish and Polish datasets[16,17], but not in German[14] or Indian[19] datasets. The MIF-173 C/C genotype was significantly associated with pan-colitis compared with left-sided or distal disease (OR = 10.78, 95%CI: 1.34-86.62. P = 0.0074) in one Japanese UC dataset (n = 221)[15], but not a second Japanese dataset[18]. With respect to the length polymorphism, MIF CATT₅/CATT₅ genotype has been found to confer a protective effect against IBD in Japanese individuals aged over 20 years (OR = 0.33, 95%CI: 0.14-0.82. P = 0.013)[18], whilst the CATT₇/CATT₇ genotype was significantly associated with both a chronic continuous phenotype (OR = 5.49, 95%CI: 1.19-25.3, P = 0.015) and distal disease location (OR = 6.10, 95%CI: 1.32-28.3, P = 0.0091) in Japanese subjects. However, no association of CATT repeat length with UC risk was observed in two Spanish datasets[16].

The story is similar for CD. Of the previous studies to have investigated the association between MIF-173G > C and CD, four found no association[12,13,16,17], and one reported a protective effect of the minor allele, MIF-173C[14,16]. Furthermore, in one dataset, despite no association with overall CD susceptibility, further analysis revealed MIF-173C influenced CD phenotype; conferring a protective effect against the development of upper GI CD (OR = 0.31, 95%CI: 0.118-0.789, P = 0.01)[12].

Our study sought to resolve the discordance observed between previous investigations by conducting the single largest Caucasian dataset (1006 New Zealand cases, 540 New Zealand controls) study of MIF promoter polymorphisms, and employing meta-analysis. Single-site analysis found no evidence of association of MIF promoter polymorphisms with overall IBD susceptibility in New Zealand Caucasians. Haplotype analysis found preliminary evidence that a rare haplotype, CATT_5/-173C, which is the least active haplotype in vitro[9], confers protection against IBD (OR = 0.22, 95%CI: 0.05-0.99, P_unadjusted = 0.03). However, as the functional effects of MIF promoter polymorphisms are cell type specific and relevant experiments in monocytes, the primary cell type responsible for MIF dependent mucosal inflammation in IBD are lacking, no firm conclusions can be drawn from these observations[5].

Subsequent meta-analyses with additional Caucasian CD (3373 patients, 5543 controls), Caucasian UC (794 patients, 1499 controls), and Asian UC (416 patients, 789 controls) datasets also detected no association of MIF-173G > C with CD or UC in Caucasians, or with UC in East Asians (Figure 1). However, our Caucasian meta-analysis is limited by the presence of significant heterogeneity among the datasets, thus it is not possible to completely rule out the possibility that MIF promoter polymorphism may alter susceptibility to IBD in some Caucasian populations. In contrast, no significant heterogeneity was detected in the East Asian datasets (P_het = 0.50), therefore the discordance previously observed is likely to be due to sample size rather than to differences in baseline characteristics of the study participants.

In conclusion, based on the results of our study, neither MIF-173G > C nor MIF CATT_5-8 are risk factors for IBD in New Zealand Caucasians, nor is MIF-173G > C a risk factor for UC in East Asians. The results of meta-analysis do not support MIF-173G > C as a susceptibility factor for IBD in Caucasian datasets however, because significant heterogeneity exists between the populations investigated, the possibility that a significant effect exists in a subgroup of these populations cannot be excluded. Additional well-powered studies in European Caucasian datasets would be required to clarify this point.

Macrophage migration inhibitory factor (MIF) is a pro-inflammatory cytokine implicated in the pathophysiology of inflammatory arthritis, systemic lupus erythematosus, atherosclerosis, and cancer. Evidence from animal models and human inflammatory bowel disease (IBD) indicates that MIF is also an important mediator of IBD; however few studies have considered its mechanism of action in these diseases. In recent years, evidence regarding the role of MIF in inflammatory disease has accumulated rapidly, and efforts have subsequently turned to the development of specific anti-MIF therapies for use in human disease.

An international phase one trial of a neutralizing anti-MIF antibody is currently underway for lupus nephritis. On this background authors believe that critical appraisal of the role of MIF in IBD is needed in order for IBD patients to benefit from developments in this field. Two functional variants of the MIF promoter region have been identified. Despite clear association with other inflammatory diseases, the contribution of these polymorphisms to IBD risk and phenotype remains uncertain due to previous studies being under-powered.

This paper describes the largest single cohort study of MIF promoter polymorphisms to date, as well as meta-analysis of all available genotype data on Caucasians and East Asians. Their population based cohort study had 93% power, and the meta-analyses 100% power, to detect an effect size of OR = 1.40 (α = 0.05). They found no evidence of association of MIF promoter polymorphisms with IBD.

This study is important as it provides clear evidence that genetic variants of MIF do not influence IBD susceptibility. Their findings are in agreement with recent research that indicates the contribution of MIF to IBD pathophysiology is not determined at the level of nuclear transcription.

The role of MIF in the pathogenesis of IBD has been intensively discussed during last couple of years, including the promoter polymorphism of this gene. It is also seen as one of the possible novel therapeutic targets for the management of IBD. The group has contributed to the research in this field by contributing data about the prevalence of the MIF promoter polymorphism in large population of patients and healthy controls. Furthermore, they have meta-analyzed this new data with previously published data, and concluded that there is no evidence of association of MIF promoter polymorphisms with IBD. Paper is well structured, comprehensive and clearly written.