Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Int J Tuberc Lung Dis. Author manuscript; available in PMC 2010 July 12.
Published in final edited form as:
PMCID: PMC2902362

Association of SLC11A1 with tuberculosis interactions with NOS2A and TLR2 in African-Americans and Caucasians



Host defense factors may influence the development of active tuberculosis (TB).


To test variants in solute carrier family 11A, member 1 (SLC11A1), for an association with TB.


A mixed case-control study of TB cases, relatives or close contact controls, consisting of 474 African-Americans (243 families) and 381 Caucasians (192 families), examined 13 SLC11A1 polymorphisms for association with pulmonary TB using generalized estimating equations adjusting for age and gender.


Two associations were observed in Caucasians (rs3731863, P = 0.03, and rs17221959, P = 0.04) and one in African-Americans (rs3731865, P = 0.05). Multilocus analyses between polymorphisms in SLC11A1 and 11 TB candidate genes detected interactions between SLC11A1 and inducible nitric oxide synthase (NOS2A) in Caucasians (rs3731863 [SLC11A1] × rs8073782 [NOS2A], P = 0.009; rs3731863 [SLC11A1] × rs17722851 [NOS2A], P = 0.007) and toll-like receptor 2 (TLR2) in African-Americans (rs3731865 [SLC11A1] x rs1816702, P = 0.005).


No association was detected with 5′(GT)n promoter polymorphism previously associated with lower SLC11A1 expression, rs17235409 (D543N), or rs17235416 (3′ TGTG insertion/deletion polymorphism). SLC11A1 polymorphism rs3731865 was associated with TB in African-Americans, consistent with previous findings in West Africans. These results suggest that variants in SLC11A1 increase susceptibility to pulmonary TB and interact with other variants that differ by race.

Keywords: SLC11A1, tuberculosis, genetic epidemiology, epistasis, innate immunity

Tuberculosis (TB) is a global public health problem, resulting in 1.6 million deaths worldwide in 2005.1,2 About 10% of all people infected with Mycobacterium tuberculosis develop active pulmonary disease, suggesting differences in susceptibility to progression exist.3 Genetic susceptibility may play a role in determining which infected individuals develop active TB disease.

Several lines of evidence support a role for genetics in the development of pulmonary TB. Studies of twins show that infected identical twins were more likely to develop TB than infected fraternal twin pairs.4 Other studies have documented familial clustering of individuals with pulmonary TB, mostly from populations recently introduced to the M. tuberculosis pathogen.5,6 Further, mouse models of mycobacterial infection have identified several potential TB susceptibility or resistance loci; perhaps the most studied of these is Bcg/Lsh/Ity, or the natural resistance-associated macrophage protein 1 (Nramp1) locus. In mice, a substitution of aspartic acid for glycine at position 169 of this locus (Gly169Asp) increases resistance to infection by M. bovis (reviewed by North and Medina7). The human homolog of this gene, solute carrier family 11A, member 1 (SLC11A1), is a divalent cation transported located on chromosome 2q35.8 SLC11A1 may influence pathogen survival by regulating cation levels in vesicles or cells.9 The strong evidence that Nramp1 regulates mycobacterial disease development in mice made SLC11A1 an attractive candidate gene for studies of human genetic susceptibility to pulmonary TB.

Studies of SLC11A1 in humans with TB have primarily focused on four or five polymorphisms distributed across the gene. The original association study of SLC11A1 with TB in West Africans examined four polymorphisms: a GTn repeat in the 5′ promoter region, a four base-pair (TGTG) insertion/deletion (rs17235416) in the 3′ untranslated region (UTR), and two single nucleotide polymorphisms (SNPs) in intron 4 (rs3731865) and exon 15 (rs17235409, D543N).10 In the original study, the 5′GTn repeat and rs3731865 were in strong linkage disequilibrium (LD), as were rs17235409 and rs17235416. Both pairs of markers were significantly associated with TB, and carrying the risk alleles at both markers further increased association with disease. This association has been replicated in studies from Japan,11 Guinea-Conakry12 and South Africa.13 Studies focusing on just a subset of these SNPs have shown an association with the 5′GTn repeat in Gambians14 and non-Hispanic Caucasians,15 with rs3731865 in Koreans,16,17 and with rs17235409 or rs17235416 in Chinese18 and Koreans.16,17 Furthermore, a large Canadian aboriginal family with apparent autosomal dominant susceptibility to TB has been linked to SLC11A1.6 Some studies, however, have failed to find an association with one or more of these SNPs, suggesting that there is a variation in the strength of the association across populations.1821

In efforts to examine SLC11A1 more comprehensively for an association with pulmonary TB, we examined 13 SLC11A1 polymorphisms using tag and candidate markers rather than limiting our selection to previously studied variants. Our study population consisted of Caucasians of European descent and African-Americans sampled from the south-eastern United States and Argentina. We also tested SLC11A1 for interactions with polymorphisms in 11 other TB candidate genes. The primary goals of this study were to determine if SLC11A1 polymorphisms influence the risk of TB in our sample, and if these effects were modified by variants in other genes.



Participants were ascertained through the North Carolina or South Carolina TB Control Programs in the United States or as patients at the out-patient clinic at F J Muñiz Hospital in Buenos Aires, Argentina, between 2002 and 2006. Criteria for inclusion as TB cases were: 1) age ≥14 years and culture-confirmed pulmonary TB or 2) age <14 years and culture-confirmed or clinically diagnosed pulmonary TB that included a positive tuberculin skin test plus an infiltrate or hilar adenopathy on chest X-ray. Individuals were eligible to participate if their TB had been diagnosed in the past, or if they were currently receiving TB treatment. All TB cases remained eligible if they also had a diagnosis of extra-pulmonary TB. Family members of eligible TB cases who themselves had a history of TB were enrolled as part of a multi-case family if review of their records established a diagnosis of either pulmonary or extra-pulmonary TB. Thus, a small portion of our study subjects enrolled as part of a multi-case-family had extra-pulmonary TB only.

The diagnosis of a TB case was confirmed by a review of the medical records and laboratory reports, or, for US cases, by documentation through electronic surveillance databases used by each state’s TB program to document and report TB cases to the US Centers for Disease Control and Prevention (the TB Information Management System [TIMS]). TIMS comprises data collected by local jurisdictions using the Report of Verified Case of TB (RVCT) form. Severity of TB disease was assessed by the presence of acid-fast bacilli (AFB) in sputum smears or X-ray evidence of cavitary lesions. We attempted to document human immunodeficiency virus (HIV) status for all subjects. However, participation in this study did not require that the individual authorize review of HIV test results.

Unaffected individuals who were in close contact with cases during the infectious phase of the disease (household contacts such as spouses and partners, and relatives such as parents and siblings) were enrolled as controls. Informed consent was obtained from all subjects or their legal representatives before participation in the study. Human experimentation guidelines of the US Department of Health and Human Services and those of the participating research institutions were followed.

The protocol was approved by the Institutional Review Board at Duke University Medical Center, the North and South Carolina Departments of Public Health, Centro de Educación Médica e Investigaciones Clínicas Norberto Quirno (CEMIC), the F J Muñiz Hospital, Buenos Aires, Argentina, and the University of Miami Miller School of Medicine.

Marker selection and genotyping

Thirteen markers in the SLC11A1 gene were selected for analysis (Figure 1). Four of these (5′ (GT)n repeat polymorphism, rs3731865, rs17235409 and rs17235416) were studied in the original report by Bellamy et al.10 Two additional SNPs (rs2276631, a synonymous Phe66Phe polymorphism in exon 3, and rs17221959, a synonymous Gly249Gly polymorphism in exon 8) were examined by Abel et al.22 The remaining SNPs were tags selected from African (Yoruba) and Centre d’Etude du Polymorphisme Humain (CEPH) populations from the International HapMap project (,23 NCBI build 35 assembly HapMap phase II. The criteria used in tag SNP selection was a tag SNP 5′ upstream and 3′ downstream of SLC11A1, aminor allele frequency of 0.05, pairwise correlation coefficient (r2) of 0.80, and SNPs that met the criteria for aggressive tagging using 2- and 3-marker haplotypes in Haploview software. This approach was used to select the markers examined in the 11 additional TB candidate genes (Supplemental Table 1, available at inducible nitric oxide synthase (NOS2A), toll-like receptor 9 (TLR9), toll-like receptor 2 (TLR2), interferon-gamma receptor 1 (IFNGR1), tumor necrosis factor alpha (TNF-α), PARK2 co-regulated (PACRG), parkin isoform 1 (PARK2), toll-like receptor 4 (TLR4), interferon gamma (IFNG), vitamin D (1,25-dihydroxyvitamin D3) receptor (VDR), and ubiquitin protein ligase E3A (UBE3A).

Figure 1
SLC11A1 gene map with SNPs analyzed. 15 exons are labeled on the gene map of SLC11A1; untranslated regions are labeled in white and exons in black. The map is oriented 5′ to 3′. SNPs are indicated by vertical lines with SNP ’rs’ ...

The TaqMan allelic discrimination assay was used to genotype the 13 SNPs and one insertion/deletion polymorphism. Assays were obtained from the Applied Biosystems Assays on Demand (SNPs rs7576974, rs3731865, rs2290708, rs3816560, rs2279015, rs1059823) or Assays by Design (SNPs rs2276631, rs3731865, rs17221959, rs17235409, rs17235416, and rs13062) services. The 5′ (GT)n repeat polymorphism was genotyped using the fluorescent allele static scanning systems (FAAST)24 and previously published primer sequences and conditions.10,25

Statistical analysis

One TB case and control was selected from each family for tests of Hardy-Weinberg equilibrium (HWE) at each SNP using genetic data analysis (GDA) software.26 Pair-wise LD between SNPs was calculated using Haploview statistical software.27 Haplotype blocks were defined according to the algorithm described by Gabriel et al., which defines a block according to the 95% confidence interval (CI) of the D′ value for pairwise LD between SNPs.28 All analyses were conducted separately in Caucasians and African-Americans.

Genotypic tests of association and analyses of clinical data were performed using generalized estimating equations (GEE) implemented in SAS (Proc GENMOD) statistical software version 9.1 (SAS Institute, Cary, NC, USA) using the independence correlation matrix. GEE analyzes families regardless of structure, including concordant and discordant siblings, relative pairs, and unrelated close contacts. GEE has been shown to be a valid test of gene X gene and gene X environment interactions in mixed family and case-control data.29 Additive genotypic models were performed modeling the minor allele as the risk allele (0 vs. 1 vs. 2), and adjusting for potential confounders, age and sex for all analyses. For Caucasians, we tested for genotype-ascertainment site interactions by incorporating ascertainment site and a genotype-ascertainment site interaction term in all models to account for potential genetic heterogeneity between ascertainment sites.

Multilocus analysis between significant (P < 0.05 in single SNP analyses) SLC11A1 SNPs and SNPs in other TB candidate genes (minor allele frequency ≥0.05) were performed in GEE. Genotypes were modeled using additive risk models for both SNPs and an interaction term for the interactions between risk alleles of both SNPs. Interactions with the most significant associations (P ≤ 0.01 for the interaction term) were further analyzed stratifying by genotype. Odds ratios (OR) and 95% CIs were calculated for both single SNP and stratified GEE analyses.

To account for potential confounding by HIV status, analyses were repeated excluding individuals with TB who were also HIV-positive at diagnosis. All statistical tests were evaluated at a nominal significance level of α=0.05, unadjusted for multiple comparisons.


A description of the 855 individuals (435 families) analyzed in this study is presented in Table 1 and a summary of demographic characteristics is presented in Table 2. This data set consisted of 295 African-American cases and 179 African-American controls, 237 Caucasian cases (141 from the United States and 96 from Argentina) and 144 Caucasian controls (88 from the United States and 56 from Argentina). The only unrelated controls included were spouses and partners of enrolled cases. The remainder were exposed unaffected relatives of cases. In African-Americans and Caucasians, the majority of the families had only one case per family (86% in African-Americans and 81% in Caucasians). The average age at enrollment was respectively 44.4 ± 17.4, 52.9 ± 21.5, 45.84 ± 20.76, 54.73 ± 18.62, 34.38 ± 15.59 and 32.40 ± 15.38 years for African-American cases, African-American controls, US Caucasian cases, US Caucasian controls, Argentina Caucasian cases and Caucasian controls; 73.6% of African-American cases and 67.1% of Caucasian cases had documented HIV status (positive or negative). Among those cases with documented HIV status, 14.2% of African-Americans and 5.7% of Caucasians were HIV-positive. HIV status was not obtained on controls. The majority of cases and a minority of controls were male (African-American 66% of cases and 17% of controls; Caucasian 54% of cases and 33% of controls). As a result of our entry criteria (see Methods section), the majority of enrolled participants had pulmonary and extrapulmonary disease.

Table 1
Study Population
Table 2
Demographic characteristics of study participants

Single SNP association analyses

One SNP had a significant deviation from HWE in African-American controls (rs1059823, P = 0.002). The pair-wise LD (D′ and r2) for both African-American and Caucasian controls is presented on Figure 2. Similar patterns of LD were observed in cases (not shown). In general, markers were in low-to-moderate LD in African-Americans. In Caucasians, pair-wise LD was stronger for several SNP pairs, including the rs3731865, rs17235409 and rs17235416 SNPs that have previously been reported in strong LD in West Africans.10

Figure 2
LD plots for SLC11A1 unrelated founder controls. LD plots were generated in Haploview and are presented for: A) African-American controls D′; B) African-American controls r2; C) Caucasian controls D′; D) Caucasian controls r2. Within each ...

Single-locus GEE results stratified by race (Table 3) identified three significant genotypic associations. There were no statistically significant interactions between genotype and ascertainment site in analyses of Caucasian samples. One SNP in African-Americans, rs3731865 (P = 0.05), and two SNPs in Caucasians, rs3731863 (P = 0.03) and rs17221959 (P = 0.04), were significantly associated (P ≤ 0.05) with TB. Two SNPs are intronic, and rs17221959 is a synonymous coding SNP (G249G) in exon 8. There were no significant allele or genotype frequency differences between TB cases with and those without HIV, nor were there significant differences when analyses were stratified by HIV status (results not shown).

Table 3
SLC11A1 single locus GEE analyses

Multilocus analyses

The SNPs with significant (P < 0.05) single SNP associations were tested for interactions with SNPs in other TB candidate genes. We limited our interaction analyses to two SNP interactions due to sample size limitations. We observed strong interactions (P < 0.01) between SLC11A1 SNPs and SNPs in the NOS2A and TLR2 gene (Table 4); however, though the interactions varied by race. In African-Americans, SLC11A1 SNP rs3731865 (intronic) interacted with TLR2 SNP rs1816702 (intronic); (P = 0.005; Table 4). In Caucasians, there were two SLC11A1 interactions with NOS2A: SLC11A1 rs17221959 interacted with NOS2A promoter SNP rs8073782 (P = 0.009), and SLC11A1 intronic SNP rs3731863 interacted with NOS2A SNP rs17722851 (P = 0.007; Table 4). SLC11A1 rs17221959 and SLC11A1 rs3731863 are in moderate LD (D′=0.68 and r2=0.40) in Caucasians. The SNPs in NOS2A and TLR2 did not have statistically significant individual associations with TB.

Table 4
SLC11A1 statistically significant interaction analyses with GEE

We performed multi-locus diplotype analyses to determine which genotypic combination contributed to risk for developing active TB (Table 5). In African-Americans, the combination of TLR2 rs1816702 CC and SLC11A1 rs3731865 CG&GG genotypes resulted in the greatest increased risk, with an OR of 3.94 (95% CI 2.07–7.50, P < 0.0001). In Caucasians, the combination of NOSA rs8073782 CC and SLC11A1 rs3731863 CT&TT genotypes resulted in a reduced risk for disease, with an OR of 0.20 (95% CI 0.06–0.59, P = 0.004). Furthermore, the NOS2A rs17221951 AT&TT and SLC11A1 rs3731863 CC genotypes had an OR of 0.42 (95% CI 0.22–0.79, P = 0.007).

Table 5
Multilocus diplotype analyses for African-American and Caucasian interaction analyses


Our results suggest that NOS2A and TLR2 act through epistatic interactions as modifiers of SLC11A1-mediated TB risk. We observed one interaction in African-Americans (SLC11A1 rs3731865/TLR2 rs1816702) and two in Caucasians (SLC11A1 rs3731863/NOS2A rs8073782; SLC11A1 rs17221959/NOS2A rs8073782). In Caucasians, both SLC11A1 rs3731863 and rs17221959 are in moderate LD (r2=0.40, D′=0.68), suggesting that the interactions involving these SNPs may be capturing a single effect from a common functional SNP. The NOS2A SNP (rs8073782) is in the promoter region, a part of the gene previously observed to have several significant single SNP effects in West African populations.

Intracellular iron availability influences transcription of inducible nitric oxide synthase protein (iNOS) and iNOS availability due to SLC11A1-mediated mechanisms.30 SLC11A1 is involved in cation transport, particularly iron. It can transport iron into phagolysosomes leading to formation of reactive oxygen species by way of the Haber-Weiss and Fenton reactions.31 The release of iron upregulates iNOS production by activation of transcription factors such as STAT1 and interferon regulatory factor (IRF).32,33

The TLR2 intronic SNP that interacted with SLC11A1 in African-Americans has not been previously associated with TB. This SNP (rs1816702) had no main effect and would not have been associated with TB without the interaction with SLC11A1 rs3731865. To our knowledge, past research has not examined the joint role of TLR2 and SLC11A1. However, the absence of functional TLR2 in knockout mice has been associated with TB susceptibility,34 and prolonged TLR2 exposure is necessary to block TB attenuation of macrophage signaling.35 Our interactions could reflect a previously unknown interaction or relationship between SLC11A1 and TLR2.

The results of this study, the first to examine SLC11A1 and TB in African-Americans, indicate that SNPs in SLC11A1 are associated with risk of pulmonary TB in both African-Americans and a sample of Caucasians. We observed a significant association with the minor allele of rs3731865 (intron 4) in African-Americans, consistent with previous positive associations in West African populations.10,12 In agreement with results from Guinea-Conakry,12 no association was detected with other SNPs previously associated with TB in Africans: 5′ (GT)n repeat, rs17235409 and rs17235416.10,13 The 5′ (GT)n repeat is located within a potential enhancer element, and the 5′ (GT)9 allele (allele 207 in our sample) has been associated with higher levels of SLC11A1 expression.25 This, coupled with the lack a common functional variant in exons of SLC11A1,10 has led investigators to focus on 5′ (GT)9 as the variant of interest in SLC11A1. Our findings indicated strong D′ in both ethnic groups (African-American D′=0.90; Caucasian D′=0.94,) but variable r2 (African-Americans r2 = 0.38; Caucasians r2 = 0.90) between rs3731865 and the 5′ (GT)n repeat polymorphism, in contrast to the strong LD previously reported between the G allele at rs3731865 and 5′ (GT)9 (allele 207) in West Africans.10 The lack of an association between the 5′ (GT)n polymorphism and TB in this sample might be due to this weaker LD between rs3731865 and the 5′ (GT)n repeat; we also observed slightly different allele frequencies for both SNPs in our African-Americans and Caucasians relative to their West African population.

Although these results are interesting, there were some limitations to our study. First, we observed some baseline differences for the ratio of males to females and mean age at examination in cases and controls. We therefore adjusted for these covariates in our models. We also acknowledge that using two different ascertainment centers for the collection of our Caucasian dataset, one from the South-Eastern United States and another from Argentina, and the inclusion of a small number of HIV-positive cases, could confound the results. However, to account for these effects we adjusted for ascertainment site in our models and conducted sensitivity analysis by analyzing the dataset with and without HIV-positive individuals. HIV status did not influence the significance of our results. While we tested many individual SNP and two-way interactions, potentially increasing the number of false-positive results, we performed this analysis with the intention of trying to replicate previous associations observed in SLC11A1 and to generate new hypotheses with regard to how SLC11A1 may interact with other TB candidate genes.

The results of this study implicate SLC11A1 as a susceptibility gene for TB in both African-Americans and Caucasians. Taken together, the results for African-Americans and Caucasians suggest that multiple independent polymorphisms in SLC11A1 might influence the risk for pulmonary TB either independently or through interactions with other genes. These results could also explain why associations at SLC11A1 have replicated inconsistently, as it is possible that associations are only observed at SLC11A1 in the presence of specific genotype variants within other genes. Elucidation of the precise modes of interaction depends on isolating functional variants in SLC11A1, NOS2A and TLR2 that might explain these results.


The authors thank the study participants, without whom this study would have been impossible, the North Carolina TB control Nurse Consultants (M Allen, D Foster, J Luffman and E Zeringue) and county TB nurses who referred subjects to the study. They also thank M Fletcher, E Levine, E Little and C Poszik for assistance in recruiting participants in South Carolina, and C Linton, R Carney and A Mosher for recruiting participants in North Carolina. The work in this manuscript was supported by grant number R01 HL068534 from the National Heart, Lung and Blood Institute, National Institutes of Health (NIH). CDH acknowledges support from NIH K24-AI001833.


1. The Stop TB Strategy, World Health Organization. Tuberculosis. Geneva, Switzerland: WHO; 2008. [Accessed October 2008]. Programs and projects.
2. World Health Organization. Fact Sheet No. 104. Geneva, Switzerland: WHO; 2007. Tuberculosis.
3. Skamene E, Schurr E, Gros P. Infection Genomics: Nramp1 as major determinant of natural resistance to intracellular infections. Annu Rev Med. 1998;49:275–287. [PubMed]
4. Comstock G. Tuberculosis in twins: a re-analysis of the Prophit Survey. Am Rev Respir Dis. 1978;117:621–624. [PubMed]
5. Sousa AO, Salem J, Lee FK, et al. An epidemic of tuberculosis with a high rate of tuberculin anergy among a population previously unexposed to tuberculosis, the Yanomami Indians of the Brazilian Amazon. Proc Natl Acad Sci (USA) 1997;94:13227–13232. [PubMed]
6. Greenwood CMT, Fujiwara M, Boothroyd LJ, et al. Linkage of tuberculosis to chromosome 2q35 loci, including NRAMP1, in a large aboriginal Canadian family. Am J Hum Genet. 2000;67:405–416. [PubMed]
7. North RJ, Medina E. How important is Nramp1 in tuberculosis? Trends in Microbiology. 1998;6:441–443. [PubMed]
8. Goswami T, Bhattacharjee A, Babal P, et al. Natural-resistance-associated macrophage protein 1 is an H+/bivalent cation antiporter. Biochem J. 2001;354:511–519. [PubMed]
9. Mohamed HS, Ibrahim ME, Miller EN, et al. SLC11A1 (formerly NRAMP1) and susceptibility to visceral leishmaniasis in The Sudan. Eur J Hum Genet. 2004;12:66–74. [PubMed]
10. Bellamy R, Ruwende C, Corrah T, McAdam KPWJ, Whittle HC, Hill AVS. Variations in the Nramp1 gene and susceptibility to tuberculosis in West Africans. The New England Journal of Medicine. 1998;338:640–644. [PubMed]
11. Gao PS, Fujishima S, Mao XQ, et al. Genetic variants of NRAMP1 and active tuberculosis in Japanese populations. International Tuberculosis Genetics Team. Clin Genet. 2000;58:74–76. [PubMed]
12. Cervino AC, Lakiss S, Sow O, Hill AV. Allelic association between the NRAMP1 gene and susceptibility to tuberculosis in Guinea-Conakry. Ann Hum Genet. 2000;64:507–12. [PubMed]
13. Hoal EG, Lewis LA, Jamieson SE, et al. SLC11A1 (NRAMP1) but not SLC11A2 (NRAMP2) polymorphisms are associated with susceptibility to tuberculosis in a high-incidence community in South Africa. Int J Tuberc Lung Dis. 2004;8:1464–1471. [PubMed]
14. Awomoyi AA, Marchant A, Howson JM, McAdam KP, Blackwell JM, Newport MJ. Interleukin-10, polymorphism in SLC11A1 (formerly NRAMP1), and susceptibility to tuberculosis. J Infect Dis. 2002;186:1808–1814. [PubMed]
15. Ma X, Dou S, Wright JA, et al. 5′ dinucleotide repeat polymorphism of NRAMP1 and susceptibility to tuberculosis among Caucasian patients in Houston, Texas. Int J Tuberc Lung Dis. 2002;6:818–823. [PubMed]
16. Ryu S, Park YK, Bai GH, Kim SJ, Park SN, Kang S. 3′UTR polymorphisms in the NRAMP1 gene are associated with susceptibility to tuberculosis in Koreans. Int J Tuberc Lung Dis. 2000;4:577–580. [PubMed]
17. Kim JH, Lee SY, Lee SH, et al. NRAMP1 genetic polymorphisms as a risk factor of tuberculous pleurisy. Int J Tuberc Lung Dis. 2003;7:370–375. [PubMed]
18. Liu W, Cao WC, Zhang CY, et al. VDR and NRAMP1 gene polymorphisms in susceptibility to pulmonary tuberculosis among the Chinese Han population: a case-control study. Int J Tuberc Lung Dis. 2004;8:428–434. [PubMed]
19. Delgado JC, Baena A, Thim S, Goldfeld AE. Ethnic-specific genetic associations with pulmonary tuberculosis. J Infect Dis. 2002;186:1463–1468. [PubMed]
20. Soborg C, Andersen AB, Madsen HO, Kok-Jensen A, Skinhoj P, Garred P. Natural resistance-associated macrophage protein 1 polymorphisms are associated with microscopy-positive tuberculosis. J Infect Dis. 2002;186:517–521. [PubMed]
21. El Baghdadi J, Remus N, Benslimane A, et al. Variants of the human NRAMP1 gene and susceptibility to tuberculosis in Morocco. Int J Tuberc Lung Dis. 2003;7:599–602. [PubMed]
22. Abel L, Sanchez FO, Oberti J, et al. Susceptibility to leprosy is linked to the human NRAMP1 gene. J Infect Dis. 1998;177:133–145. [PubMed]
23. The International HapMap Consortium. The International HapMap Project. Nature. 2003;426:789–796. [PubMed]
24. Vance JM, Ben Othmane K. Methods of Genotyping. In: Haines JL, Pericak-Vance MA, editors. Approaches to gene mapping in complex human diseases. New York, NY, USA: Wiley-Liss; 1998. pp. 213–228.
25. Searle S, Blackwell JM. Evidence for a functional repeat polymorphism in the promoter of the human NRAMP1 gene that correlates with autoimmune versus infectious disease susceptibility. J Med Genet. 1999;36:295–299. [PMC free article] [PubMed]
26. Abecasis GR, Cookson WO. GOLD–graphical overview of linkage disequilibrium. Bioinformatics. 2000;16:182–183. [PubMed]
27. Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics. 2005;21:263–265. [PubMed]
28. Gabriel SB, Schaffner SF, Nguyen H, et al. The structure of haplotype blocks in the human genome. Science. 2002;296:2225–2299. [PubMed]
29. Hancock DB, Martin ER, Li YJ, Scott WK. Methods for interaction analyses using family-based case-control data: conditional logistic regression versus generalized estimating equations. Genet Epidemiol. 2007;31:883–893. [PubMed]
30. Weiss G, Werner-Felmayer G, Werner ER, Grunewald K, Wachter H, Hentze MW. Iron regulates nitric oxide synthase activity by controlling nuclear transcription. J Exp Med. 1994;180:969–976. [PMC free article] [PubMed]
31. Gutteridge JM, Shute DJ. Iron-dioxygen-dependent changes to the biological activities of bleomycin. J Inorg Biochem. 1981;15:349–357. [PubMed]
32. Dlaska M, Weiss G. Central role of transcription factor NF-IL6 for cytokine and iron-mediated regulation of murine inducible nitric oxide synthase expression. J Immunol. 1999;162:6171–6177. [PubMed]
33. Fritsche G, Dlaska M, Barton H, Theurl I, Garimorth K, Weiss G. Nramp1 functionality increases inducible nitric oxide synthase transcription via stimulation of IFN regulatory factor 1 expression. J Immunol. 2003;171:1994–1998. [PubMed]
34. Stenger S, Engele M, Bolcskei PL, Rollinghoff M, Wagner M. Induction of direct antimicrobial activity through mammalian toll-like receptors. Pneumologie. 2002;56:103–107. [German] [PubMed]
35. Arko-Mensah J, Julian E, Singh M, Fernandez C. TLR2 but not TLR4 signalling is critically involved in the inhibition of IFN-gamma-induced killing of mycobacteria by murine macrophages. Scand J Immunol. 2007;65:148–157. [PubMed]