SLC11A1 polymorphisms in inflammatory bowel disease and Mycobacterium avium subspecies paratuberculosis status

Abstract

AIM: To test for association of SLC11A1 with inflammatory bowel disease (IBD) and Mycobacterium avium subspecies paratuberculosis (MAP) status in a Caucasian cohort.

METHODS: Five hundred and seven Crohn’s disease (CD) patients, 474 ulcerative colitis (UC) patients, and 569 healthy controls were genotyped for SLC11A1 1730G>A and SLC11A1 469+14G>C using pre-designed TaqMan^® SNP assays. χ² tests were applied to test for association of single nucleotide polymorphisms (SNPs) with disease, and the presence of MAP DNA.

RESULTS: SLC11A1 1730G>A and SLC11A 1469+14G>C were not associated with CD, UC, or IBD. The SLC11A1 1730A minor allele was over-represented in patients who did not require immunomodulator therapy (P = 0.002, OR: 0.29, 95% CI: 0.13-0.66). The frequency of the SLC11A1 469+14C allele was higher in the subset of study participants who tested positive for MAP DNA (P = 0.02, OR: 1.56, 95% CI: 1.06-2.29). No association of SLC11A1 1730G>A with MAP was observed.

CONCLUSION: Although SLC11A1 was not associated with IBD, association with MAP suggests that SLC11A1 is important in determining susceptibility to bacteria implicated in the etiology of CD.

Key Words: NRAMP1; Crohn’s disease; Ulcerative colitis; IS900 polymerase chain reaction

INTRODUCTION

The solute carrier family 11 (SLC11A1) gene (also known as natural resistance associated macrophage protein 1, NRAMP1)[1] has been associated with susceptibility to intracellular pathogens since its initial identification in mice[2]. SLC11A1 encodes a divalent cation transporter that is located in endosome and phagosome membranes[3] of macrophages and monocytes within the liver, spleen and lungs[2,4]. This transporter plays a key role in mounting an effective immune response against intracellular pathogens[1,5] through its involvement in the acidification of the phagosomes[6], as well as the regulation of nitric oxide, interleukin-10[7] and vacuolar iron concentrations[8].

Given the pivotal roles that SLC11A1 plays in innate immunity, it is not surprising that the relationship between polymorphisms in SLC11A1 and a number of autoimmune and mycobacterial diseases has been explored. Associations have been found with leprosy[9], tuberculosis[10], rheumatoid arthritis[11], visceral leishmaniasis[12], multiple sclerosis[13], type 1 diabetes mellitus[14], and inflammatory bowel disease (IBD)[15-18]. Most of these disease associations have been with a promoter dinucleotide microsatellite (GT_n) that is known to affect SLC11A1 expression levels[19]. However, SLC11A1 also contains a number of single nucleotide polymorphisms (SNPs), including SLC11A1 1730G>A (rs17235409; D543N) and SLC11A1 469+14G>C (rs3731865; INT4G>C). The non-synonymous SNP 1730G>A is thought to alter the protein function[18], whereas the intronic SNP 469+14G>C has no known functional effect, but has been suggested to be in linkage disequilibrium with functional promoter polymorphisms[12].

SLC11A1 1730G>A and SLC11A1 469+14G>C have been tested for association with Crohn’s disease (CD) in two European cohorts. Although the smaller of the two studies found no association with CD, Gazouli et al[18] have reported a significant association of both SNPs with disease (SLC11A1 1730G>A P_genotypic = 0.0001, OR: 3.43, 95% CI: 1.95-5.93, SLC11A1 469+14G>C P_genotypic = 0.006, OR: 15.91, 95% CI: 0.92-273.46). The involvement of SLC11A1 in the handling and elimination of intracellular pathogens, as well as its association with mycobacterial diseases makes it a biologically plausible candidate risk gene for CD. The results of recent genome-wide association studies strongly suggest defects in genes involved in bacterial detection, handling, and elimination are central to CD pathogenesis. Furthermore the assertion, albeit controversial, that Mycobacterium avium subspecies paratuberculosis (MAP) is an initial trigger for CD provides an additional rationale to investigate SLC11A1 as a candidate risk gene for IBD. As a result, this study had two aims. The first was to attempt the first independent replication of the association of SLC11A1 1730G>A and SLC11A1 469+14G>C with IBD. The second aim was to use previously collected MAP IS900 data[20] to test for association of SLC11A1 genotypes with occurrence of MAP DNA in peripheral blood.

MATERIALS AND METHODS

Study participants

Patients were selected from a New Zealand Caucasian IBD cohort that had been recruited to investigate genetic and environmental factors that contribute to CD and UC etiology[20-24]. Detailed phenotypic data were available for members of this cohort including ancestry, location of disease, family history of IBD, age of onset, presence of extra-intestinal manifestations, and requirement for surgery. The MAP status of the CD patients in this cohort had been determined previously using IS900 polymerase chain reaction[20]. Randomly selected blood donors (n = 501) from Christchurch (New Zealand), including 180 who had been previously tested for MAP status[20] served as controls.

Genotyping

Genotyping of SLC11A1 1730G>A (rs17235409) and SLC11A1 469+14G>C (rs3731865) was performed in 384-well plates using the pre-designed Taqman^® SNP genotyping assays C_256352269_10 and C_1659793_10 (Applied Biosystems, Foster City, CA, USA) in a LightCycler^® 480 II (Hoffmann La Roche, Basel, Switzerland). Cycling conditions for rs17235409 were 10 min at 95°C, 40 cycles of 15 s at 92°C and 1 min at 60°C, and 30 s of cooling at 40°C. Conditions were the same for rs3731865, but annealing was at 66°C rather than 60°C. Results were analyzed using Lightcycler^® 480 software version 1.5.0. The accuracy of the genotyping assays was confirmed by repeat analysis of 13% of samples. Concordance between original and repeat genotype calls was 99%.

Statistical analysis

A web-based calculator (http://ihg2.helmholtz-muenchen.de/cgi-bin/hw/hwa1.pl) was used to test for deviations from Hardy-Weinberg Equilibrium (HWE). The χ² and OR analyses were performed using SPSS for Windows, version 13.0 (SPSS Inc., Chicago, IL, USA). Associations were considered significant if P was < 0.05. Post hoc power analysis demonstrated that our cohort had 90% power to detect a relative risk of 2.15 for SLC11A1 1730G>A (MAF_controls = 0.02, α = 0.05) and 99.8% power to detect a relative risk of 1.5 for SLC11A1 469+14G>C (MAF_controls = 0.30, α = 0.05).

Ethical considerations

All study participants provided written informed consent to be involved in ongoing IBD research, and ethical approval for this study was given by the Upper South Regional Ethics Committee (Canterbury, New Zealand).

RESULTS

Genotyping for SLC11A1 1730A>G and 469+14G>C was successful in 1468 (94.7%) and 1432 (92.4%) of study participants, respectively. No deviations from HWE were detected in cases or controls for either SNP (P > 0.05). The percentage minor allele frequency (MAF) of SLC11A1 1730G>A and SLC11A1 469+14G>C in our controls was 2% and 30%, respectively. We found no evidence of association of either SLC11A1 SNP with overall CD, UC or IBD susceptibility (Table 1). Similarly, the minor allele and genotype frequencies of SLC11A1 1730G>A and 469+14G>C did not associate with age of disease onset, disease behavior, disease location, or requirement for resectional surgery (all P values > 0.1, data not shown). A significantly higher frequency of the SLC11A1 1730A allele was seen in IBD patients who did not require immunomodulator therapy, compared to those who did require this treatment approach (P_IBD = 0.002, OR: 0.29, 95% CI: 0.13-0.66, P_CD = 0.03, OR: 0.38, 95% CI: 0.15-0.95, P_UC = 0.01, OR: 0.75, 95% CI: 0.71-0.79) (Table 2). There was no significant association of SLC11A1 1730G>A with MAP status, whereas the SLC11A1 469+14C allele was associated with increased incidence of MAP DNA in peripheral blood (P = 0.02, OR: 1.56, 95% CI: 1.06-2.23) in our cohort (Table 3).

Table 1 Genotype and allele frequencies of SLC11A1 1730G>A and 469+14G>C in New Zealand Crohn’s disease and ulcerative colitis patients, and healthy controls n (%).

Phenotype	Genotype			MAF	Allelic P value	Allelic OR (95% CI)
1730G>A	GG	GA	AA	A
CD (n = 495)	474 (96)	21 (4)	0	21 (2)	0.832	1.07 (0.57-2.00)
UC (n = 470)	450 (96)	20 (4)	0	20 (2)	0.827	1.07 (0.57-2.02)
HC (n = 503)	483 (96)	20 (4)	0	20 (2)
469+14G>C	GG	GC	CC	C
CD (n = 495)	265 (54)	192 (39)	38 (8)	268 (27)	0.153	0.83 (0.65-1.07)
UC (n = 451)	245 (54)	171 (38)	35 (8)	241 (27)	0.101	0.81 (0.62-1.04)
HC (n = 486)	238 (49)	204 (42)	44 (9)	292 (30)

MAF: Minor allele frequency; OR: Odds ratio; CI: Confidence interval; CD: Crohn’s disease; UC: Ulcerative colitis; HC: Healthy controls.

Table 2 Genotype frequencies of SLC11A1 1730G>A (rs17235409) in inflammatory bowel disease patients who have used/not used immunomodulators n (%).

Phenotype/immunomodulator status	Genotype			P value	OR (95% CI)
	GG	GA	AA
CD/never used IM (n = 217)	203 (94)	14 (6)	0	0.031	0.38 (0.15-0.95)
CD/have used IM (n = 278)	271 (98)	7 (2)	0
UC/never used IM (n = 356)	336 (94)	20 (6)	0	0.010	0.75 (0.71-0.79)
UC/have used IM (n = 114)	114 (100)	0	0
IBD/never used IM (n = 573)	539 (94)	34 (6)	0	0.002	0.29 (0.13-0.66)
IBD/have used IM (n = 392)	385 (98)	7 (2)	0

OR: Odds ratio; CI: Confidence interval; CD: Crohn’s disease; UC: Ulcerative colitis; IBD: Inflammatory bowel disease; IM: Immunomodulator.

Table 3 Distribution of SLC11A1 469+14G>C genotype by Mycobacterium avium subspecies paratuberculosis status¹ in New Zealand Caucasians n (%).

MAP DNA in blood	Genotype frequency		P value	OR (95% CI)
	GG	GC + CC
Present (n = 150)	66 (44)	84 (56)	0.02	1.56 (1.06-2.29)
Absent (n = 351)	193 (55)	158 (45)

¹Tested by IS900 polymerase chain reaction to detect the presence of Mycobacterium avium subspecies paratuberculosis (MAP) DNA in peripheral blood[20]. OR: Odds ratio; CI: Confidence interval.

DISCUSSION

Previous association of SLC11A1 1730G>A and 469+14G>C with mycobacterial infections and preliminary evidence of association with CD[10-12,25] suggest that SLC11A1 alters susceptibility to IBD. The primary aim of our study was to conduct the first independent replication of the association of SLC11A1 with CD. In contrast to the original study of Gazouli et al[18], we found no evidence of SLC11A1 1730G>A or 469+14G>C as risk factors for IBD, CD or UC (all P values > 0.8) (Table 1). Comparison of the MAFs for the two SLC11A1 SNPs revealed the existence of significant heterogeneity between Gazouli et al[18] and other studies for SLC11A1 1730A, and between populations of Northern versus Southern European ancestry for SLC11A1 469+14C. Our cohort and the cohort of Liu et al[26], which were composed primarily of individuals of Northern European ancestry, had SLC11A1 469+14C frequencies of 30% and 27% respectively. In contrast, the cohorts drawn from Southern European populations (Italian, Greek, and Turkish) exhibited significantly lower MAFs for this SNP. These differences in MAF distribution hint at the existence of a North-South gradient for SLC11A1, which could in turn explain the discordance between our study and that of Gazouli et al[18]. The occurrence of such gradients is not without precedence. The frequency of the CD-associated SNPs, R702W, G908R and 1007fs, within the nucleotide oligomerization binding domain 2 gene (NOD2, also known as CARD15) exhibits a strong North-South gradient within Europe. A recent meta-analysis of NOD2 association studies performed on European IBD cohorts has found that the MAFs and thus the contribution of these SNPs to CD risk increased significantly with decreasing latitude[27].

The minor allele of SLC11A1 1730G>A was found to be significantly over-represented in the subset of our IBD patients who had never used immunomodulators, and by inference had less severe disease (Table 2). However, we saw no association with other markers of disease severity in our cohort. Due to the very low minor allele frequency (no minor allele homozygotes were observed), this result requires replication in other large cohorts to rule out a type 1 error.

The second aim of this study was to test for association of SLC11A1 with MAP. The MAP status of 321 CD patients and 180 controls has been determined previously[20]. Combining these patients and controls, we found no association between MAP status and SLC11A1 1730G>A, but did find an association with SLC11A1 469+14G>C (P = 0.02, OR: 1.56, 95% CI: 1.06-2.29) (Table 3). Earlier studies[14,16] on smaller CD cohorts (n = 37 or 59) did not find any evidence of association of MAP status with SLC1A11 469+14G>C. However, this polymorphism has been associated with susceptibility to Mycobacterium tuberculosis[10], and additional variation within SLC11A1 has been associated with susceptibility to other mycobacterial diseases such as leprosy[9]. Our results provide preliminary evidence of an association of the SLC11A1 469+14C allele with susceptibility to MAP.

We conclude that although SLC11A1 could be a risk factor for IBD in some Southern European populations, we did not find an association of SLC11A1 469+14G>C or SLC11A1 1730G>A with IBD in our cohort that comprised primarily patients of Northern European ancestry. However, the significantly higher incidence of MAP DNA in the peripheral blood of SLC11A1 469+14C heterozygotes and homozygotes compared to SLC11A1 469+14G within our cohort suggests that this SLC11A1 SNP, although not directly influencing disease risk, might modify susceptibility to potential CD-causing bacteria.

COMMENTS

Background

The involvement of SLC11A1 in the handling and elimination of intracellular pathogens, as well as its association with mycobacterial diseases makes it a biologically plausible candidate risk gene for Crohn’s disease (CD). The suggestion that Mycobacterium avium subspecies paratuberculosis (MAP) is an initial trigger for CD provides an additional rationale to investigate SLC11A1 as a candidate risk gene for inflammatory bowel disease (IBD).

Research frontiers

A previous genetic association study has indicated that SLC11A1 is a susceptibility gene for IBD. The authors performed an independent replication of this study in a large population-based cohort of Northern European origin. They also tested for the association of these polymorphisms with MAP status.

Innovations and breakthroughs

This is believed to be the first study to examine the association of SLC11A1 polymorphisms in a well-powered cohort of Northern European origin. These findings indicate that SLC11A1 polymorphisms do not modify disease risk for IBD, but might influence disease behavior (through indirect markers of severity) and susceptibility to MAP, a putative pathogen in CD. The authors also note the disparity of allele frequency between populations of Northern and Southern European origin.

Applications

By understanding how SLC11A1 genotype influences the risk of colonization/infection with MAP, the authors might gain some insight into the contribution of this bacterium to IBD, and how defective clearance of MAP and other intracellular bacteria might be associated with modified disease risk.

Terminology

SLC11A1, solute carrier family 11 gene (also known as Natural Resistance Associated Macrophage Protein 1, NRAMP1) plays a key role in an effective innate immune response against intracellular pathogens. MAP is an intracellular bacterium that has been cited in several studies as a putative causal agent of CD.

Peer review

This paper provides interesting new results regarding the possible relationship between SLC11A1 polymorphisms and IBD risk. The study has been done carefully and thoroughly, and the paper is very well written. The lack of association of SLC11A1 and IBD risk in the study population (New Zealand Caucasians primarily of Northern European descent) is an important finding. The positive result that shows an association of an SLC11A1 allele and MAP status is novel and interesting.