- Email: [email protected]
Interferon-γ receptor 1 promoter polymorphisms: population distribution and functional implications
Clinical Immunology 112 (2004) 113 – 119 www.elsevier.com/locate/yclim
Interferon-g receptor 1 promoter polymorphisms: population distribution and functional implications Sergio D. Rosenzweig, a Alejandro A. Scha¨ffer, b Li Ding, a Rachel Sullivan, a Balasz Enyedi, a Jae-Joon Yim, a James L. Cook, c James M. Musser, d and Steven M. Holland a,* a
Laboratory of Host Defenses, National Institute of Allergy and Infectious Diseases, NIH, DHHS, Bethesda, MD, 20892-1886, USA b Computational Biology Branch, National Center for Biotechnology Information, NIH, DHHS, Bethesda, MD, 20892-6075, USA c Section of Infectious Diseases, Department of Medicine, University of Illinois at Chicago, Chicago, IL, 60607, USA d Laboratory of Human Bacterial Pathogenesis, National Institute of Allergy and Infectious Diseases, NIH, DHHS, Hamilton, MN, 59840, USA Received 4 August 2003; accepted with revision 22 March 2004
Abstract Different polymorphisms have been described in the minimal promoter region (MPR) of the interferon-g receptor 1 (IFNGR1), a molecule that plays a critical role in mycobacterial control. We sequenced the IFNGR1 MPR from African American, Caucasian and Korean controls, and from mycobacteria-infected patients. Six different single nucleotide polymorphisms (SNPs) were detected in the IFNGR1 MPR. The three ethnic groups showed different SNP distribution patterns, but no significant differences were detected between mycobacterial cases and controls. Two polymorphisms were found in all populations (G-611A, T-56C). We cloned the four allelic variants (var) of haplotype G-611A/ T-56C into a luciferase reporter vector and determined their promoter activity. Polymorphisms at position -611 had a stronger effect on the promoter activity than those at position -56, and constructs carrying G-611 produced a stronger promoter activity than -611A constructs. The IFNGR1 MPR is a polymorphic region with at least two SNPs influencing its activity, but these are not associated with increased mycobacterial susceptibility. Published by Elsevier Inc. Keywords: Single nucleotide polymorphisms; Mycobacteria; Pulmonary; Disseminated; Haplotype; Avium
Introduction Genetic susceptibility factors have been shown to play critical roles in susceptibility to mycobacterial infections (reviewed in [1,2]). Mutations in interferon-g receptor 1 (IFNGR1, OMIM *107470), interferon-g receptor 2 (IFNGR2, OMIM *147569), interleukin-12 receptor h1 (IL12RB1, OMIM *601604), interleukin-12 p40 (IL12B, OMIM *161561), signal transducer and activator of transcription 1 (STAT1, OMIM *600555), and NF-kB essential modulator (NEMO, OMIM *300248) confer markedly increased susceptibility to mycobacterial infections. Of these, mutations in IFNGR1 are the most prevalent. Never* Corresponding author. Immunopathogenesis Section, Laboratory of Host Defenses, National Institute of Allergy and Infectious Diseases, NIH Building 10, Room 11N103 10, Center Drive MSC 1886, Bethesda, MD 20892-1886. Fax: +1-301-402-4369. E-mail address: [email protected] (S.M. Holland). 1521-6616/$ - see front matter. Published by Elsevier Inc. doi:10.1016/j.clim.2004.03.018
theless, fewer than 50% of individuals with disseminated mycobacterial infections have an identified underlying gene defect. Both exonic and intronic mutations have been described in IFNGR1; most of which lead to abolition of protein expression (recessively inherited nonsense mutations) or normal levels of expression of dysfunctional proteins (recessively inherited missense mutations). Jouanguy et al. [3] reported a distinct group of dominantly inherited nonsense mutations that result in overaccumulation of a truncated IFNGR1. Therefore, the level of IFNGR1 expression probably plays a critical role in IFNg responsiveness. Because IFNGR1 mutations have been highly associated with severe disseminated mycobacterial disease, we sought IFNGR1 promoter mutations or polymorphisms that might be associated with pulmonary mycobacterial disease. In the process of performing this search, we identified six polymorphisms in the IFNGR1 minimal promoter region (MPR); two of which were novel. To understand and
114
S.D. Rosenzweig et al. / Clinical Immunology 112 (2004) 113–119
explore functional effects of some of polymorphisms, we created IFNGR1 promoter-driven luciferase constructs. Here, we show the distribution of IFNGR1 MPR polymorphisms in several ethnic populations and in mycobacteria-infected cases and controls. We also show the effects of different haplotypes on baseline receptor expression.
Cases, controls, materials, and methods Cases Individuals with clinical and microbiological criteria for pulmonary tuberculosis (PTB; microbiologically confirmed tuberculosis limited to the lungs), disseminated nontuberculous mycobacterial infection [(NTM); must be present in at least two sites, at least one outside the lungs], and pulmonary nontuberculous mycobacterial infection (NTM infection limited to the lungs), but without any known susceptibility factor for mycobacterial disease, were included in this study. Individuals with known genetic underlying conditions associated with increased susceptibility to mycobacterial infection (e.g., mutations in IFNGR1, IFNGR2, IL-12Rh1, IL-12p40, NEMO, or STAT1) were excluded. Patients with HIV, cancer, lymphoproliferative disorders, or those undergoing chemotherapy or immunosuppressive treatment were also excluded. In families with more than one affected member, only one was included. All patients gave written informed consent. Pulmonary tuberculosis. Seventy-six African Americans and 70 Caucasians with pulmonary tuberculosis (PTB) from Houston, TX, were included in the study. Samples from Korean PTB patients were not available at the time of the study. Disseminated and pulmonary nontuberculous mycobacterial infections Twenty-five Caucasian individuals with disseminated nontuberculous mycobacterial (DNTM) infection and 53 Caucasians with pulmonary nontuberculous mycobacterial (PNTM) infection, all of them followed at the Laboratory of Host Defenses, NIH, were included in this study. Controls One hundred and fourteen healthy African American adults (37 from the NIH Blood Bank and 77 from Houston, TX), 128 healthy Caucasian adults (81 from the NIH Blood Bank and 47 from Houston, TX), and 80 healthy Korean adults (Seoul National University College of Medicine, Seoul, Korea) constituted our control populations. All control individuals were negative for HIV serology and self-evaluated as healthy adults by questionnaires. Neither sputum nor blood cultures for mycobacteria were determined in healthy volunteers. No information regarding PPD status was recorded.
Materials and methods: PCR and sequencing The full IFNGR1 minimal promoter region (-692 to -1) was bidirectionally amplified and sequenced from Cases and Controls using two sets of primers producing two overlapping amplimers. Distal segment (412 bp): forward primer, 5V GCACAAGCGCTGA AGGACTTAG 3V; reverse primer, 5VGGTTTCCTGGGGTTGTGCA AG 3V. Proximal amplimer: forward primer, 5VG ATTGTGGCTCGGATTGTGGCTCGGCTGTGGC 3V; reverse primer, 5V GGTACCTGAGGACGGCCCCA 3V. The PCR products were treated with shrimp alkaline phosphatase (SAP; USB, Cleveland, OH)) and exonuclease I (Exo; USB) before sequencing. Briefly, 0.037 U of SAP and 0.19 U of Exo per Al PCR were incubated at 37jC during 5V followed by 15V incubation at 72jC. Sequencing was performed with Dye Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems, Foster City, CA) and analyzed with Sequencher 3.1.1 software. Single nucleotide polymorphisms (SNPs) analysis For comparison of the nucleotide diversity between the ethnic groups, we used the normalized number of variant (var) sites, which corrects for sample size [4]:
h¼K
X n1
I1 L
ð1Þ
i¼1
where K is the observed number of var alleles, L is the length of the IFNGR1 MPR (692), and n is the number of control chromosomes analyzed [Caucasian = 256 (128 individuals); African American = 228 (114 individuals); Korean = 160 (80 individuals); altogether = 644 (322 individuals)]. Functional studies Constructs. The IFNGR1 minimal promoter region from a Caucasian control, whose sequence matched the NCBI reference sequence for the interferon-g receptor a-chain gene (NCBI, GI: 632535), was cloned into the pGL3-Basic Luciferase Reporter Vector (Promega, Madison, WI). Briefly, the IFNGR1 minimal promoter region (MPR) was amplified from genomic DNA: forward primer (-707): 5V GCAATGTGGCTCGAGTGGAATAGGAG 3V; reverse primer (+18): 5V GAGAAAGAGGAGAAGCTTGCTGCCTACC 3V. Forward primer (-707) creates an XhoI restriction site at position -692 (in italics), and reverse primer creates a HindIII restriction site at position 1 of the ORF (in italics). The amplicon and the vector were digested with XhoI and HindIII (Invitrogen, Carlsbad, CA) and ligated with TaKaRa DNA Ligation Kit (TaKaRa Shuzo, Japan). Transformed DH5a competent cells were grown on LB/ ampicillin media. Plasmid DNA was extracted (Minipreps
S.D. Rosenzweig et al. / Clinical Immunology 112 (2004) 113–119
115
Table 1a Genotype distribution of the IFNGR1 MPR polymorphisms
Data are presented as ‘‘homozygous Wt/heterozygous/homozygous var’’ for each of the groups analyzed (for polymorphism – 470delTT, the deletion was designated as the var allele). Shaded, SNPs where var alleles were found.
DNA Purification System, Promega) and the insert sequenced. The construct with G-611 and, a T-56 is designated wild-type (Wt)/Wt. Construct Wt/Wt was used as the template into which 611A and -56C were separately introduced by site-directed mutagenesis (QuickChange Site-Directed Mutagenesis Kit, Stratagene, La Jolla, CA) following the manufacturer’s protocol. This process gave two new constructs, -611A/T56 (var/Wt) and G-611G/-56C (Wt/var). Using var/WT construct as a template, we introduced mutation -56C by site-directed mutagenesis leading to -611A/-56C (var/var). All mutations and inserts were verified by sequencing. Cell lines, transfections, and luciferase assays. Cell lines K562 (human erythroleukemia line, ATCC CCL-243) and PLB-985 (human acute nonlymphocytic leukemia line) [5] were grown in RPMI 1640 media (Invitrogen) supplemented with 10% heat-inactivated fetal calf serum (Gemini BioProducts, Woodland, CA), 2 mM L-glutamine (HyClone, Logan, UT), and antibiotics (Pen/Strept, Invitrogen). Cells were split every 3 days. Both cell lines naturally express significant levels of IFNgR1 as detected by flow cytometry (not shown).
K562 cells were lipofectin-transfected (Invitrogen) with 1 Ag of IFNGR1 MPR-luciferase constructs (Wt/Wt, Wt/ var, var/Wt, and var/var) and 0.5 Ag of h-Gal, following the producer’s protocol for transfecting suspension cell lines. Cells were grown for 24 h, harvested, lysed, and luciferase and h-Gal activities measured with the DualLight kit (Tropix, Bedford, MA). The assay was performed three independent times in triplicate and once in duplicate. PLB-985 cells (8 106) were transfected with 2 Ag of the IFNGR1 MPR-luciferase constructs (Wt/Wt, Wt/var, var/Wt, and var/var) and 1 Ag of pRL-CMV Renilla (Promega) by electroporation (3 pulses, 2 ms, 400v) in 4-mm gap cuvettes (Bio-Rad, Hercules, CA). Cells were grown for 72 h, harvested, lysed, and luciferase and Renilla activities measured with the Dual-Luciferase kit (Promega). The assay was performed three independent times in triplicate. Luciferase indices [LI = (luciferase activity/normalization agent activity) K] were calculated for each construct in each experiment and compared to the pGL3-Basic LI (designated 1).
Table 1b Statistical analysis of the genotype distribution of IFNGR1 MPR polymorphisms
Control populations are compared to each other in the upper part of the table; Caucasian cases and controls are compared in the middle part of the table; and African American cases and controls are compared in the lower part of the table. Depending on the size of the samples analyzed, v2 test or Fisher exact test was applied to compare genotype distribution between groups in 2 3 tables. Nominal P values (uncorrected for the number of tests) were computed. Significant P value for both tests < 0.05 (shaded). AA indicates African American; Cau, Caucasian; Kor, Korean; Con, control; PTB, pulmonary tuberculosis; PNTM, pulmonary nontuberculous mycobacteria; DNTM, disseminated nontuberculous mycobacteria.
116
S.D. Rosenzweig et al. / Clinical Immunology 112 (2004) 113–119
Statistical analysis Distributions of the polymorphisms found in the IFNGR1 MPR were compared between the three control populations (African American, Caucasian, and Korean), between the case groups, and between the case groups and their respective ethnic-matched controls. v2 test or Fisher exact test was applied, depending on the size of the samples analyzed and nominal P values (uncorrected for the number of tests) computed (significant P value for both tests, < 0.05). Allele frequency and Hardy-Weinberg equilibrium were determined for each polymorphism analyzed. All African American and Caucasian samples were combined for logistic regression analyses using the glm and related functions in S-Plus.
The analyses were done separately on the -611, -72, and 56 SNPs. The outcome was presence or absence of mycobacterial disease. The frequency of one allele or the other in the individual’s ethnic control was a control variable. For each SNP, there were three separate analyses in which the single predictor variables were (1) presence of the first allele, (2) presence of the second allele, and (3) number of copies of the first allele. To assess the representation of haplotypes, we used the method of Fallin et al. [6], which relies on estimating frequencies of haplotypes by the expectation maximization (EM) method. Since this method may not be the most accurate for inferring haplotypes [7], we used several other approaches for inferring haplotypes from the IFNGR1 SNPs, because the problem of haplotype
Fig. 1. IFNGR1 promoter activity in K562 cells (a) and PLB cells (b). Luciferase index (LI) of IFNGR1 promoter G-611A/T-56C allelic variants (gray bars) and pGL3-Basic (black bars) are compared. pGL3-Basic LI was defined as 1. LI = (luciferase activity / normalizing agent activity) K. (K562 normalizing agent, h-Gal; PLB normalizing agent, pRL-Renilla; K, constant value). Significant P values are shown.
S.D. Rosenzweig et al. / Clinical Immunology 112 (2004) 113–119
inference is not fully solved, and competing methods have been proposed [8]. For the data shown here, we inferred haplotypes in the following ways: (1) the method of Stephens et al. [7] as implemented in a program called PHASE, applied to all SNPs, followed by extracting the haplotypes on the two SNPs of interest; (2) PHASE applied to only the two SNPs of interest; and (3) the EM method, as implemented in a program by Fallin, applied to the two SNPs of interest. Since the most likely haplotypes are not computed explicitly in this program, we modified the source code (generously provided by Dr. D. Fallin) to store and print the most likely haplotypes. Luciferase assay results between the four different constructs were compared by two-tailed t test.
Results All the observed genotype frequencies were in HardyWeinberg equilibrium. The male-to-female ratio in the control populations ranged from 1.5 to 2.1. Among cases, the ratios ranged from 0.17 to 0.92. Age and gender were comparable across groups, with the exception of pulmonary NTM, which has a female preponderance. However, no significant differences for IFNGR1 MPR SNP distribution were detected across genders in any of the case or control groups analyzed. We identified six SNPs in the IFNGR1 MPR of the three populations analyzed (African American, Caucasian, and Korean) (Table 1a). Two of these SNPs are novel (T-255C and C-169A). For each polymorphic position, we designated the wild-type (Wt) nucleotide as that reported by Merlin et al. [9] (NCBI, Nucleotide Accession Sequence, GI: 632535); other nucleotides at the same position were considered ‘‘variants’’ (var). Only one var allele was found per SNP identified. We determined the frequency of each genotype (homozygous Wt, heterozygous, and homozygous var) for every SNP position and compared them between the ethnic control groups and between the mycobacterial cases and their respective controls (Table 1a). Polymorphisms G-611A and T-56C were present in all groups; SNP C-72T was present in African Americans and Caucasians, and the other SNPs were found in African Americans (-470delTT), Koreans (T-255C), or Caucasians (C-169A). All SNPs had variant allele frequencies over 2%, except for C-169A (frequency, 1.3%) (Table 1b). The nucleotide diversity values found were as follows: Caucasian = 9.44 104; African American = 9.63 104; Korean = 7.67 104; altogether 12.31 104. Polymorphisms G-611A and T-56C were found in all ethnic groups and showed a significantly different ethnic distribution when they were analyzed independently and in predicted haplotypes. Polymorphism G-611A was overrepresented in Caucasian controls and statistically differently distributed between Caucasian and African American controls ( P value < 0.001), and between Caucasian and Korean
117
controls ( P value < 0.001). This difference was mainly dependent on the increased presence of allele G-611 in Caucasians. No statistical difference was found between African American and Korean controls for G-611A distribution ( P value = 0.16). Similarly, T-56C distribution in Caucasian controls was statistically different from African American ( P value = 0.003) and Korean controls ( P value = 0.016) because of the increased prevalence of T-56 in Caucasians. No difference was detected when T-56C was compared between African American and Korean controls ( P value = 0.616). Because of allele-72T overrepresentation in Caucasian controls, polymorphism C-72T followed the G-611A and T-56C distribution pattern. Caucasians were statistically different from African Americans and Koreans, and Koreans and African Americans were not statistically different (Table 1b). The predicted haplotype distributions for polymorphisms G-611A and T-56C were similar to those in the individual SNP analysis. The ‘‘omnibus likelihood ratio test’’ for predicted haplotype distribution was significantly different between Caucasian and African American controls ( p value < 0.0001) and between Caucasian and Korean controls ( p value < 0.0001), but not different between African American and Korean controls ( p value 0.149) (Figs. 1a and b). The most prevalent predicted haplotype for SNPs-611/-56 in African Americans and Koreans was-611A/-56C (above 50% in both populations), and G-611/T-56 in Caucasians (43%; 3.5 and 4.6 times more frequent in Caucasian than in African American and Korean controls, respectively) (Table 2a). When the polymorphisms were compared between mycobacterial cases and controls or between Caucasian cases, no consistent differences in genotype and haplotype distribution [Table 2b] were found. In addition, no other differences than those detected between African American controls and Caucasian controls were detected when African American PTB and Caucasian PTB were compared. Differences between each group, such as those within or between different ethnic groups, were determined identically as stated in Cases, controls, materials, and methods. Importantly, none of the most prevalent polymorphisms in African Americans and Caucasians (G-611A, C-72T, and T-56C)
Table 2a Predicted haplotype distribution for SNPs G-611A/T-56C
Predicted haplotype distribution analyzed by Fallin and PHASE methods gave very similar estimates. Results by Fallin’s method are shown. Shaded, most prevalent predicted haplotype for each group analyzed.
118
S.D. Rosenzweig et al. / Clinical Immunology 112 (2004) 113–119
Table 2b Omnibus likelihood ratio test for predicted haplotype distribution
Significant P value < 0.05 (shaded). AA indicates African American; Cau, Caucasian; Kor, Korean; Con, control; PTB, pulmonary tuberculosis; PNTM, pulmonary nontuberculous mycobacteria; DNTM, disseminated nontuberculous mycobacteria.
were statistically significant predictors of mycobacterial disease by logistic regression analysis. IFNGR1 promoter activity was tested in the myeloid cell lines K562 and PLB; both of which express IFNgR1 (not shown). All four allelic variants of the IFNGR1 promoter showed a stronger promoter activity than the pGL3-Basic vector (range: K562, 3.22- to 8.31-fold; PLB, 289- to 705fold) (Figs. 1a and b). In K562 cells, promoters carrying G611 (constructs G-611/T-56 and G-611/-56C) showed a stronger promoter activity than those carrying a-611A (611A/T-56 and-611A/-56C). Construct G-611/T-56 showed a significantly stronger promoter activity than -611A/T-56 and-611A/-56C. Alleles at position -56 (T or C) did not significantly affect IFNGR1 promoter activity (Fig. 1a). In PLB cells, constructs carrying G-611 (G-611/T-56 and G-611/-56C) showed stronger promoter activity than those carrying-611A (-611A/T-56 and-611A/-56C). The G-611/56C construct showed significantly stronger promoter activity than -611A/T-56,-611A/-56C, and G-611/T-56, as well (Fig. 1b). Polymorphisms at -56 showed no functional effects when an A was found at position -611, but relevant differences were detected when a G was found at the same position (Fig. 1b).
Discussion The IFNGR1 MPR is highly polymorphic. Taking all of our control subjects together, we found six polymorphisms in the 692 bp analyzed (1 in 115 bp), yielding a nucleotide diversity rate (h) of 12.35 104. This represents a higher incidence than that described for noncoding regions by Cargill et al. [4] (1 SNP per 354 bp; h = 5.43 104) and Halushka et al. [10] (1 SNP per 211 bp; h = 8.5 104), as determined by analysis of 106 and 75 genes, respectively, in other multiethnic cohorts. It is noteworthy that, although the h in the IFNGR1 MPR in our group is surprisingly high, more than 15% of the genes screened by a Cargill et al. [4] and Halushka et al. [10] had nucleotide diversity rates z 12 104. Richardson et al. [11] reported a similar increased nucleotide diversity rate for the TNFa promoter, another gene involved in mycobacterial control
[12]. In 1947, Haldane [13] recognized the importance of microorganisms in the evolution of the human genome as drivers of natural selection. The high rate of genetic polymorphism found in the IFNGR1 MPR and the TNFa promoter suggests that these regions of the genome are highly plastic and have been under intense evolutionary pressure [14]. When each ethnic group was analyzed independently, the SNP distribution diverged. This phenomenon has been previously described [4,10,15]. Only two SNPs were shared by all three racial groups, and each group showed certain SNP absent in the others. Koreans showed fewer SNPs than African Americans and Caucasians. Due to the limited sample size in our study, finding a polymorphism in only one group does not necessarily imply that it is private to that particular ethnicity. SNPs common to all ethnic groups may represent ‘‘old SNPs’’ that arose before population divergence, while those particular to some ethnicities—‘‘young SNPs’’—may have appeared later. The age of an SNP must be taken into account in the evaluation of disease association, since the results could be influenced by it; while ‘‘young SNPs’’ are less prevalent and have larger genomic segments in linkage disequilibrium, ‘‘old SNPs’’ are more prevalent and have undergone more recombinations with other haplotypes. On average, frequency is used as a surrogate for age [10,16]. In our cohorts, neither ‘‘old’’ nor ‘‘young SNPs’’ were found to be associated with mycobacterial disease. Our luciferase reporter system showed that SNP G-611A had a more prominent functional effect than SNP T-56C and that constructs carrying G-611 had stronger promoter activity than those carrying -611A. Recently, Ju¨liger et al. [17] explored the functional effects of SNP T-56C in stimulated B cell lines with a luciferase reporter system and electrophoretic mobility shift assays. Since they do not report their sequence at -611 and we found that G-611A is more influential than T-56C, their luciferase results are not easily compared to ours. However, we agree that under baseline conditions, T-56C has little effect on promoter activity. We do not yet have a mechanistic explanation for how these promoter polymorphisms exert their effect on the expression of IFNGR1 or, conceivably, other genes. Computer programs predict a GATA-1/TFIID consensus binding site around position-611 and an AP-2/AP-4 consensus binding site around position -56 [18 –20]. Recently, reports of the association of IFNGR1 promoter polymorphism and infectious disease susceptibility have been published. Koch et al. [21] reported that in the Mandika, the major Gambian ethnic group, heterozygosity for IFNGR1 polymorphism T-56C was associated with protection from development of or death from cerebral malaria. Interestingly, they also reported another polymorphism (T-270C) that we did not find in our study. However, Ju¨liger et al. [17] studied a pediatric Gabonese population with mild or severe malaria but did not find any difference in the IFNGR1 T-56C SNP distribution between these
S.D. Rosenzweig et al. / Clinical Immunology 112 (2004) 113–119
groups. Thye et al. [22] reported that allele T-56, along with two other IFNGR1 coding sequence polymorphisms, was associated with a higher rate of anti-Helicobacter pylori IgG antibodies. Controversial results regarding pulmonary tuberculosis susceptibility and IFNGR1 coding sequence polymorphisms have also been described [23,24]. Three groups of mycobacteria-infected cases were examined in Caucasians (PTB, PNTM, and DNTM) and one in African Americans (PTB). Since the three infections examined here are clinically distinct, we have shown comparisons of Caucasian cases to Caucasian controls for each type separately. However, even combining all three types of cases (PTB, PNTM, and DNTM) in an association analysis of IFNGR1 polymorphisms did not yield significant associations either (data not shown). Independent of the way the information was compiled, no consistent differences were found between cases and controls among Caucasians or African Americans. These data suggest that IFNGR1 MPR polymorphisms, even those affecting function, are not significantly associated with mycobacterial infections in Caucasians or pulmonary tuberculosis in African Americans. IFNGR1 MPR SNPs may not play a critical role in mycobacterial susceptibility. Based on our luciferase reporter assays and previously reported data, we found that, although haplotypes differ in terms of activity, they are not associated with mycobacterial infections. This is not terribly surprising, since individuals heterozygous for IFNgR1 deficiency do not appear to have increased susceptibility to mycobacterial diseases [1,2]. However, other phenomena, such as autoimmune processes, may be linked to IFNGR1 MPR polymorphisms [25,26]. Further studies are warranted to determine the specific factors involved in G-611A functional effects and whether these SNPs are relevant to other IFNg pathway-associated diseases. References [1] S.E. Dorman, S.M. Holland, Interferon-g and interleukin-12 pathway defects and human disease, Cytokine Growth Factor Rev. 11 (2000) 321 – 333. [2] J.-L. Casanova, L. Abel, Genetic dissection of immunity to mycobacteria: the human model, Annu. Rev. Immunol. 20 (2002) 581 – 620. [3] E. Jouanguy, S. Lamhamedi-Cherradi, D. Lammas, et al., A human IFNGR1 small deletion hotspot associated with dominant susceptibility to mycobacterial infection, Nat. Genet. 21 (1999) 370 – 378. [4] M. Cargill, D. Altshuler, J. Ireland, P. Sklar, K. Ardlie, N. Patil, C.R. Lane, E.P. Lim, N. Kalyanaram, J. Namesh, L. Ziaugra, L. Friedland, A. Rolfe, J. Warrington, R. Lipshutz, G.Q. Daley, E.S. Lander, Characterization of single-nucleotide polymorphisms in coding regions of human genes, Nat. Genet. 22 (1999) 231 – 238. [5] K.A. Tucker, M.B. Lilly, L. Heck Jr., T.A. Rado, Characterization of a new human diploid myeloid cell line (PLB-985) with granulocytic and monocytic differentiation capacity, Blood 70 (1987) 372 – 378. [6] D. Fallin, A. Cohen, L. Essioux, et al., Genetic analysis of case/ control data using estimated haplotype frequencies: application to APOE locus variation and Alzheimer’s disease, Genome Res. 11 (2001) 143 – 151.
119
[7] M. Stephens, N. Smith, P. Donnelly, A new statistical method for haplotype reconstruction from population data, Am. J. Hum. Genet. 68 (2001) 978 – 989. [8] D. Gusfield, Inference of haplotypes from samples of diploid populations: complexity and algorithms, J. Comp. Biol. 8 (2001) 305 – 323. [9] G. Merlin, B.-J.M. van der Leede, K. McKune, et al., The gene for the ligand binding chain of the human interferon gamma receptor, Immunogenetics 45 (1997) 413 – 421. [10] M.K. Halushka, J.-B. Fan, K. Bentley, L. Hsie, N. Shen, A. Weder, R. Cooper, R. Lipshutz, A. Chakravarti, Patterns of single-nucleotide polymorphisms in candidate genes for blood-pressure homeostasis, Nat. Genet. 22 (1999) 239 – 247. [11] A. Richardson, F. Sisay-Joof, H. Ackerman, S. Usen, P. Katundu, M. Molyneux, M. Pinder, D. Kwiatkowski, Nucleotide diversity of the TNF gene region in an African village, Genes Immun. 2 (2001) 343 – 348. [12] J.L. Flynn, M.M. Goldstein, J. Chan, K.J. Triebold, K. Pfeffer, C.J. Lowenstein, R. Schreiber, T.W. Mak, B.R. Bloom, Tumor necrosis factor-a is required in the protective immune response against Mycobacterium tuberculosis in mice, Immunity 2 (1995) 561 – 572. [13] J.B.S. Haldane, Disease and evolution, La Ricerca Sci. 19 (1947) 68 – 76 (Suppl.). [14] R. Lazarus, D. Vercelli, L.J. Palmer, W.J. Klimecki, E.K. Silverman, B. Richter, A. Riva, M. Ramoni, F.D. Martinez, S.T. Weiss, D.J. Kwiatkowski, Single nucleotide polymorphisms in innate immunity genes: abundant variation and potential role in complex human disease, Immunol. Rev. 190 (2002) 9 – 25. [15] E. Zietkiewicz, V. Yotova, M. Jarnik, M. Korab-Laskowska, K.K. Kidd, D. Modiano, R. Scozzari, M. Stoneking, S. Tishkoff, M. Batzer, D. Labuda, Nuclear DNA diversity in worldwide distributed human populations, Gene 205 (1997) 161 – 167. [16] A. Chakravarti, Population genetics—Making sense out of sequence, Nat. Genet. 21 (1999) 56 – 60 (Suppl.). [17] S. Ju¨liger, M. Bongartz, A.J.F. Luty, P.G. Kremsner, J.F.J. Kun, Functional analysis of a promoter variant of the gene encoding the interferon-g chain I, Immunogenetics 54 (2003) 675 – 680. [18] http://www.cbil.upenn.edu/tess/. [19] http://www.genomatix.de. [20] A. Aranburu, R. Carlsson, C. Persson, T. Leanderson, Transcription factor AP-4 is a ligand for immunoglobulin-n promoter E-box elements, Biochem. J. 354 (2001) 431 – 438. [21] O. Koch, A. Awomoyi, S. Usen, M. Jallow, A. Richardson, H. Jeremy, M. Pinder, M. Newport, D. Kwiatkowski, IFNGR1 gene polymorphisms and susceptibility to cerebral malaria, J. Infect. Dis. 185 (2002) 1684 – 1687. [22] T. Thye, G.D. Burchard, M. Nilius, B. Mu¨ller-Myhsok, R.D. Horstmann, Genomewide linkage analysis identifies polymorphism in the human interferon-g receptor affecting Helicobacter pylori infection, Am. J. Hum. Genet. 72 (2003) 448 – 453. [23] D.A. Fraser, L. Bulat-Kardum, J. Knezevic, P. Babarovic, N. Matakovic-Mileusnic, J. Dellacasagrande, D. Matanic, J. Pavelic, Z. Beg-Sec, Z. Dembic, Interferon-g receptor-1 gene polymorphism in tuberculosis patients from Croatia, Scand. J. Immunol. 57 (2003) 480 – 484. [24] M.J. Newport, A.A. Awomoyi, J.M. Blackwell, Polymorphisms in the interferon-g receptor-1 gene and susceptibility to pulmonary tuberculosis in the Gambia, Scand. J. Immunol. 58 (2003) 383 – 385. [25] H. Nakashima, H. Inoue, M. Akahoshi, Y. Tanaka, E. Ogami, S. Nagano, et al., The combination of polymorphisms within interferon-g receptor 1 and 2 associated with the risk of systemic lupus erythematosus, FEBS Lett. 453 (1999) 187 – 190. [26] Y. Tanaka, H. Nakashima, C. Hisano, T. Kohsaka, Y. Nemoto, H. Niro, et al., Association of the interferon-gamma receptor variant (Val14Met) with systemic lupus erythematosus, Immunogenetics 49 (1999) 266 – 271.
Interferon-g receptor 1 promoter polymorphisms: population distribution and functional implications Sergio D. Rosenzweig, a Alejandro A. Scha¨ffer, b Li Ding, a Rachel Sullivan, a Balasz Enyedi, a Jae-Joon Yim, a James L. Cook, c James M. Musser, d and Steven M. Holland a,* a
Laboratory of Host Defenses, National Institute of Allergy and Infectious Diseases, NIH, DHHS, Bethesda, MD, 20892-1886, USA b Computational Biology Branch, National Center for Biotechnology Information, NIH, DHHS, Bethesda, MD, 20892-6075, USA c Section of Infectious Diseases, Department of Medicine, University of Illinois at Chicago, Chicago, IL, 60607, USA d Laboratory of Human Bacterial Pathogenesis, National Institute of Allergy and Infectious Diseases, NIH, DHHS, Hamilton, MN, 59840, USA Received 4 August 2003; accepted with revision 22 March 2004
Abstract Different polymorphisms have been described in the minimal promoter region (MPR) of the interferon-g receptor 1 (IFNGR1), a molecule that plays a critical role in mycobacterial control. We sequenced the IFNGR1 MPR from African American, Caucasian and Korean controls, and from mycobacteria-infected patients. Six different single nucleotide polymorphisms (SNPs) were detected in the IFNGR1 MPR. The three ethnic groups showed different SNP distribution patterns, but no significant differences were detected between mycobacterial cases and controls. Two polymorphisms were found in all populations (G-611A, T-56C). We cloned the four allelic variants (var) of haplotype G-611A/ T-56C into a luciferase reporter vector and determined their promoter activity. Polymorphisms at position -611 had a stronger effect on the promoter activity than those at position -56, and constructs carrying G-611 produced a stronger promoter activity than -611A constructs. The IFNGR1 MPR is a polymorphic region with at least two SNPs influencing its activity, but these are not associated with increased mycobacterial susceptibility. Published by Elsevier Inc. Keywords: Single nucleotide polymorphisms; Mycobacteria; Pulmonary; Disseminated; Haplotype; Avium
Introduction Genetic susceptibility factors have been shown to play critical roles in susceptibility to mycobacterial infections (reviewed in [1,2]). Mutations in interferon-g receptor 1 (IFNGR1, OMIM *107470), interferon-g receptor 2 (IFNGR2, OMIM *147569), interleukin-12 receptor h1 (IL12RB1, OMIM *601604), interleukin-12 p40 (IL12B, OMIM *161561), signal transducer and activator of transcription 1 (STAT1, OMIM *600555), and NF-kB essential modulator (NEMO, OMIM *300248) confer markedly increased susceptibility to mycobacterial infections. Of these, mutations in IFNGR1 are the most prevalent. Never* Corresponding author. Immunopathogenesis Section, Laboratory of Host Defenses, National Institute of Allergy and Infectious Diseases, NIH Building 10, Room 11N103 10, Center Drive MSC 1886, Bethesda, MD 20892-1886. Fax: +1-301-402-4369. E-mail address: [email protected] (S.M. Holland). 1521-6616/$ - see front matter. Published by Elsevier Inc. doi:10.1016/j.clim.2004.03.018
theless, fewer than 50% of individuals with disseminated mycobacterial infections have an identified underlying gene defect. Both exonic and intronic mutations have been described in IFNGR1; most of which lead to abolition of protein expression (recessively inherited nonsense mutations) or normal levels of expression of dysfunctional proteins (recessively inherited missense mutations). Jouanguy et al. [3] reported a distinct group of dominantly inherited nonsense mutations that result in overaccumulation of a truncated IFNGR1. Therefore, the level of IFNGR1 expression probably plays a critical role in IFNg responsiveness. Because IFNGR1 mutations have been highly associated with severe disseminated mycobacterial disease, we sought IFNGR1 promoter mutations or polymorphisms that might be associated with pulmonary mycobacterial disease. In the process of performing this search, we identified six polymorphisms in the IFNGR1 minimal promoter region (MPR); two of which were novel. To understand and
114
S.D. Rosenzweig et al. / Clinical Immunology 112 (2004) 113–119
explore functional effects of some of polymorphisms, we created IFNGR1 promoter-driven luciferase constructs. Here, we show the distribution of IFNGR1 MPR polymorphisms in several ethnic populations and in mycobacteria-infected cases and controls. We also show the effects of different haplotypes on baseline receptor expression.
Cases, controls, materials, and methods Cases Individuals with clinical and microbiological criteria for pulmonary tuberculosis (PTB; microbiologically confirmed tuberculosis limited to the lungs), disseminated nontuberculous mycobacterial infection [(NTM); must be present in at least two sites, at least one outside the lungs], and pulmonary nontuberculous mycobacterial infection (NTM infection limited to the lungs), but without any known susceptibility factor for mycobacterial disease, were included in this study. Individuals with known genetic underlying conditions associated with increased susceptibility to mycobacterial infection (e.g., mutations in IFNGR1, IFNGR2, IL-12Rh1, IL-12p40, NEMO, or STAT1) were excluded. Patients with HIV, cancer, lymphoproliferative disorders, or those undergoing chemotherapy or immunosuppressive treatment were also excluded. In families with more than one affected member, only one was included. All patients gave written informed consent. Pulmonary tuberculosis. Seventy-six African Americans and 70 Caucasians with pulmonary tuberculosis (PTB) from Houston, TX, were included in the study. Samples from Korean PTB patients were not available at the time of the study. Disseminated and pulmonary nontuberculous mycobacterial infections Twenty-five Caucasian individuals with disseminated nontuberculous mycobacterial (DNTM) infection and 53 Caucasians with pulmonary nontuberculous mycobacterial (PNTM) infection, all of them followed at the Laboratory of Host Defenses, NIH, were included in this study. Controls One hundred and fourteen healthy African American adults (37 from the NIH Blood Bank and 77 from Houston, TX), 128 healthy Caucasian adults (81 from the NIH Blood Bank and 47 from Houston, TX), and 80 healthy Korean adults (Seoul National University College of Medicine, Seoul, Korea) constituted our control populations. All control individuals were negative for HIV serology and self-evaluated as healthy adults by questionnaires. Neither sputum nor blood cultures for mycobacteria were determined in healthy volunteers. No information regarding PPD status was recorded.
Materials and methods: PCR and sequencing The full IFNGR1 minimal promoter region (-692 to -1) was bidirectionally amplified and sequenced from Cases and Controls using two sets of primers producing two overlapping amplimers. Distal segment (412 bp): forward primer, 5V GCACAAGCGCTGA AGGACTTAG 3V; reverse primer, 5VGGTTTCCTGGGGTTGTGCA AG 3V. Proximal amplimer: forward primer, 5VG ATTGTGGCTCGGATTGTGGCTCGGCTGTGGC 3V; reverse primer, 5V GGTACCTGAGGACGGCCCCA 3V. The PCR products were treated with shrimp alkaline phosphatase (SAP; USB, Cleveland, OH)) and exonuclease I (Exo; USB) before sequencing. Briefly, 0.037 U of SAP and 0.19 U of Exo per Al PCR were incubated at 37jC during 5V followed by 15V incubation at 72jC. Sequencing was performed with Dye Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems, Foster City, CA) and analyzed with Sequencher 3.1.1 software. Single nucleotide polymorphisms (SNPs) analysis For comparison of the nucleotide diversity between the ethnic groups, we used the normalized number of variant (var) sites, which corrects for sample size [4]:
h¼K
X n1
I1 L
ð1Þ
i¼1
where K is the observed number of var alleles, L is the length of the IFNGR1 MPR (692), and n is the number of control chromosomes analyzed [Caucasian = 256 (128 individuals); African American = 228 (114 individuals); Korean = 160 (80 individuals); altogether = 644 (322 individuals)]. Functional studies Constructs. The IFNGR1 minimal promoter region from a Caucasian control, whose sequence matched the NCBI reference sequence for the interferon-g receptor a-chain gene (NCBI, GI: 632535), was cloned into the pGL3-Basic Luciferase Reporter Vector (Promega, Madison, WI). Briefly, the IFNGR1 minimal promoter region (MPR) was amplified from genomic DNA: forward primer (-707): 5V GCAATGTGGCTCGAGTGGAATAGGAG 3V; reverse primer (+18): 5V GAGAAAGAGGAGAAGCTTGCTGCCTACC 3V. Forward primer (-707) creates an XhoI restriction site at position -692 (in italics), and reverse primer creates a HindIII restriction site at position 1 of the ORF (in italics). The amplicon and the vector were digested with XhoI and HindIII (Invitrogen, Carlsbad, CA) and ligated with TaKaRa DNA Ligation Kit (TaKaRa Shuzo, Japan). Transformed DH5a competent cells were grown on LB/ ampicillin media. Plasmid DNA was extracted (Minipreps
S.D. Rosenzweig et al. / Clinical Immunology 112 (2004) 113–119
115
Table 1a Genotype distribution of the IFNGR1 MPR polymorphisms
Data are presented as ‘‘homozygous Wt/heterozygous/homozygous var’’ for each of the groups analyzed (for polymorphism – 470delTT, the deletion was designated as the var allele). Shaded, SNPs where var alleles were found.
DNA Purification System, Promega) and the insert sequenced. The construct with G-611 and, a T-56 is designated wild-type (Wt)/Wt. Construct Wt/Wt was used as the template into which 611A and -56C were separately introduced by site-directed mutagenesis (QuickChange Site-Directed Mutagenesis Kit, Stratagene, La Jolla, CA) following the manufacturer’s protocol. This process gave two new constructs, -611A/T56 (var/Wt) and G-611G/-56C (Wt/var). Using var/WT construct as a template, we introduced mutation -56C by site-directed mutagenesis leading to -611A/-56C (var/var). All mutations and inserts were verified by sequencing. Cell lines, transfections, and luciferase assays. Cell lines K562 (human erythroleukemia line, ATCC CCL-243) and PLB-985 (human acute nonlymphocytic leukemia line) [5] were grown in RPMI 1640 media (Invitrogen) supplemented with 10% heat-inactivated fetal calf serum (Gemini BioProducts, Woodland, CA), 2 mM L-glutamine (HyClone, Logan, UT), and antibiotics (Pen/Strept, Invitrogen). Cells were split every 3 days. Both cell lines naturally express significant levels of IFNgR1 as detected by flow cytometry (not shown).
K562 cells were lipofectin-transfected (Invitrogen) with 1 Ag of IFNGR1 MPR-luciferase constructs (Wt/Wt, Wt/ var, var/Wt, and var/var) and 0.5 Ag of h-Gal, following the producer’s protocol for transfecting suspension cell lines. Cells were grown for 24 h, harvested, lysed, and luciferase and h-Gal activities measured with the DualLight kit (Tropix, Bedford, MA). The assay was performed three independent times in triplicate and once in duplicate. PLB-985 cells (8 106) were transfected with 2 Ag of the IFNGR1 MPR-luciferase constructs (Wt/Wt, Wt/var, var/Wt, and var/var) and 1 Ag of pRL-CMV Renilla (Promega) by electroporation (3 pulses, 2 ms, 400v) in 4-mm gap cuvettes (Bio-Rad, Hercules, CA). Cells were grown for 72 h, harvested, lysed, and luciferase and Renilla activities measured with the Dual-Luciferase kit (Promega). The assay was performed three independent times in triplicate. Luciferase indices [LI = (luciferase activity/normalization agent activity) K] were calculated for each construct in each experiment and compared to the pGL3-Basic LI (designated 1).
Table 1b Statistical analysis of the genotype distribution of IFNGR1 MPR polymorphisms
Control populations are compared to each other in the upper part of the table; Caucasian cases and controls are compared in the middle part of the table; and African American cases and controls are compared in the lower part of the table. Depending on the size of the samples analyzed, v2 test or Fisher exact test was applied to compare genotype distribution between groups in 2 3 tables. Nominal P values (uncorrected for the number of tests) were computed. Significant P value for both tests < 0.05 (shaded). AA indicates African American; Cau, Caucasian; Kor, Korean; Con, control; PTB, pulmonary tuberculosis; PNTM, pulmonary nontuberculous mycobacteria; DNTM, disseminated nontuberculous mycobacteria.
116
S.D. Rosenzweig et al. / Clinical Immunology 112 (2004) 113–119
Statistical analysis Distributions of the polymorphisms found in the IFNGR1 MPR were compared between the three control populations (African American, Caucasian, and Korean), between the case groups, and between the case groups and their respective ethnic-matched controls. v2 test or Fisher exact test was applied, depending on the size of the samples analyzed and nominal P values (uncorrected for the number of tests) computed (significant P value for both tests, < 0.05). Allele frequency and Hardy-Weinberg equilibrium were determined for each polymorphism analyzed. All African American and Caucasian samples were combined for logistic regression analyses using the glm and related functions in S-Plus.
The analyses were done separately on the -611, -72, and 56 SNPs. The outcome was presence or absence of mycobacterial disease. The frequency of one allele or the other in the individual’s ethnic control was a control variable. For each SNP, there were three separate analyses in which the single predictor variables were (1) presence of the first allele, (2) presence of the second allele, and (3) number of copies of the first allele. To assess the representation of haplotypes, we used the method of Fallin et al. [6], which relies on estimating frequencies of haplotypes by the expectation maximization (EM) method. Since this method may not be the most accurate for inferring haplotypes [7], we used several other approaches for inferring haplotypes from the IFNGR1 SNPs, because the problem of haplotype
Fig. 1. IFNGR1 promoter activity in K562 cells (a) and PLB cells (b). Luciferase index (LI) of IFNGR1 promoter G-611A/T-56C allelic variants (gray bars) and pGL3-Basic (black bars) are compared. pGL3-Basic LI was defined as 1. LI = (luciferase activity / normalizing agent activity) K. (K562 normalizing agent, h-Gal; PLB normalizing agent, pRL-Renilla; K, constant value). Significant P values are shown.
S.D. Rosenzweig et al. / Clinical Immunology 112 (2004) 113–119
inference is not fully solved, and competing methods have been proposed [8]. For the data shown here, we inferred haplotypes in the following ways: (1) the method of Stephens et al. [7] as implemented in a program called PHASE, applied to all SNPs, followed by extracting the haplotypes on the two SNPs of interest; (2) PHASE applied to only the two SNPs of interest; and (3) the EM method, as implemented in a program by Fallin, applied to the two SNPs of interest. Since the most likely haplotypes are not computed explicitly in this program, we modified the source code (generously provided by Dr. D. Fallin) to store and print the most likely haplotypes. Luciferase assay results between the four different constructs were compared by two-tailed t test.
Results All the observed genotype frequencies were in HardyWeinberg equilibrium. The male-to-female ratio in the control populations ranged from 1.5 to 2.1. Among cases, the ratios ranged from 0.17 to 0.92. Age and gender were comparable across groups, with the exception of pulmonary NTM, which has a female preponderance. However, no significant differences for IFNGR1 MPR SNP distribution were detected across genders in any of the case or control groups analyzed. We identified six SNPs in the IFNGR1 MPR of the three populations analyzed (African American, Caucasian, and Korean) (Table 1a). Two of these SNPs are novel (T-255C and C-169A). For each polymorphic position, we designated the wild-type (Wt) nucleotide as that reported by Merlin et al. [9] (NCBI, Nucleotide Accession Sequence, GI: 632535); other nucleotides at the same position were considered ‘‘variants’’ (var). Only one var allele was found per SNP identified. We determined the frequency of each genotype (homozygous Wt, heterozygous, and homozygous var) for every SNP position and compared them between the ethnic control groups and between the mycobacterial cases and their respective controls (Table 1a). Polymorphisms G-611A and T-56C were present in all groups; SNP C-72T was present in African Americans and Caucasians, and the other SNPs were found in African Americans (-470delTT), Koreans (T-255C), or Caucasians (C-169A). All SNPs had variant allele frequencies over 2%, except for C-169A (frequency, 1.3%) (Table 1b). The nucleotide diversity values found were as follows: Caucasian = 9.44 104; African American = 9.63 104; Korean = 7.67 104; altogether 12.31 104. Polymorphisms G-611A and T-56C were found in all ethnic groups and showed a significantly different ethnic distribution when they were analyzed independently and in predicted haplotypes. Polymorphism G-611A was overrepresented in Caucasian controls and statistically differently distributed between Caucasian and African American controls ( P value < 0.001), and between Caucasian and Korean
117
controls ( P value < 0.001). This difference was mainly dependent on the increased presence of allele G-611 in Caucasians. No statistical difference was found between African American and Korean controls for G-611A distribution ( P value = 0.16). Similarly, T-56C distribution in Caucasian controls was statistically different from African American ( P value = 0.003) and Korean controls ( P value = 0.016) because of the increased prevalence of T-56 in Caucasians. No difference was detected when T-56C was compared between African American and Korean controls ( P value = 0.616). Because of allele-72T overrepresentation in Caucasian controls, polymorphism C-72T followed the G-611A and T-56C distribution pattern. Caucasians were statistically different from African Americans and Koreans, and Koreans and African Americans were not statistically different (Table 1b). The predicted haplotype distributions for polymorphisms G-611A and T-56C were similar to those in the individual SNP analysis. The ‘‘omnibus likelihood ratio test’’ for predicted haplotype distribution was significantly different between Caucasian and African American controls ( p value < 0.0001) and between Caucasian and Korean controls ( p value < 0.0001), but not different between African American and Korean controls ( p value 0.149) (Figs. 1a and b). The most prevalent predicted haplotype for SNPs-611/-56 in African Americans and Koreans was-611A/-56C (above 50% in both populations), and G-611/T-56 in Caucasians (43%; 3.5 and 4.6 times more frequent in Caucasian than in African American and Korean controls, respectively) (Table 2a). When the polymorphisms were compared between mycobacterial cases and controls or between Caucasian cases, no consistent differences in genotype and haplotype distribution [Table 2b] were found. In addition, no other differences than those detected between African American controls and Caucasian controls were detected when African American PTB and Caucasian PTB were compared. Differences between each group, such as those within or between different ethnic groups, were determined identically as stated in Cases, controls, materials, and methods. Importantly, none of the most prevalent polymorphisms in African Americans and Caucasians (G-611A, C-72T, and T-56C)
Table 2a Predicted haplotype distribution for SNPs G-611A/T-56C
Predicted haplotype distribution analyzed by Fallin and PHASE methods gave very similar estimates. Results by Fallin’s method are shown. Shaded, most prevalent predicted haplotype for each group analyzed.
118
S.D. Rosenzweig et al. / Clinical Immunology 112 (2004) 113–119
Table 2b Omnibus likelihood ratio test for predicted haplotype distribution
Significant P value < 0.05 (shaded). AA indicates African American; Cau, Caucasian; Kor, Korean; Con, control; PTB, pulmonary tuberculosis; PNTM, pulmonary nontuberculous mycobacteria; DNTM, disseminated nontuberculous mycobacteria.
were statistically significant predictors of mycobacterial disease by logistic regression analysis. IFNGR1 promoter activity was tested in the myeloid cell lines K562 and PLB; both of which express IFNgR1 (not shown). All four allelic variants of the IFNGR1 promoter showed a stronger promoter activity than the pGL3-Basic vector (range: K562, 3.22- to 8.31-fold; PLB, 289- to 705fold) (Figs. 1a and b). In K562 cells, promoters carrying G611 (constructs G-611/T-56 and G-611/-56C) showed a stronger promoter activity than those carrying a-611A (611A/T-56 and-611A/-56C). Construct G-611/T-56 showed a significantly stronger promoter activity than -611A/T-56 and-611A/-56C. Alleles at position -56 (T or C) did not significantly affect IFNGR1 promoter activity (Fig. 1a). In PLB cells, constructs carrying G-611 (G-611/T-56 and G-611/-56C) showed stronger promoter activity than those carrying-611A (-611A/T-56 and-611A/-56C). The G-611/56C construct showed significantly stronger promoter activity than -611A/T-56,-611A/-56C, and G-611/T-56, as well (Fig. 1b). Polymorphisms at -56 showed no functional effects when an A was found at position -611, but relevant differences were detected when a G was found at the same position (Fig. 1b).
Discussion The IFNGR1 MPR is highly polymorphic. Taking all of our control subjects together, we found six polymorphisms in the 692 bp analyzed (1 in 115 bp), yielding a nucleotide diversity rate (h) of 12.35 104. This represents a higher incidence than that described for noncoding regions by Cargill et al. [4] (1 SNP per 354 bp; h = 5.43 104) and Halushka et al. [10] (1 SNP per 211 bp; h = 8.5 104), as determined by analysis of 106 and 75 genes, respectively, in other multiethnic cohorts. It is noteworthy that, although the h in the IFNGR1 MPR in our group is surprisingly high, more than 15% of the genes screened by a Cargill et al. [4] and Halushka et al. [10] had nucleotide diversity rates z 12 104. Richardson et al. [11] reported a similar increased nucleotide diversity rate for the TNFa promoter, another gene involved in mycobacterial control
[12]. In 1947, Haldane [13] recognized the importance of microorganisms in the evolution of the human genome as drivers of natural selection. The high rate of genetic polymorphism found in the IFNGR1 MPR and the TNFa promoter suggests that these regions of the genome are highly plastic and have been under intense evolutionary pressure [14]. When each ethnic group was analyzed independently, the SNP distribution diverged. This phenomenon has been previously described [4,10,15]. Only two SNPs were shared by all three racial groups, and each group showed certain SNP absent in the others. Koreans showed fewer SNPs than African Americans and Caucasians. Due to the limited sample size in our study, finding a polymorphism in only one group does not necessarily imply that it is private to that particular ethnicity. SNPs common to all ethnic groups may represent ‘‘old SNPs’’ that arose before population divergence, while those particular to some ethnicities—‘‘young SNPs’’—may have appeared later. The age of an SNP must be taken into account in the evaluation of disease association, since the results could be influenced by it; while ‘‘young SNPs’’ are less prevalent and have larger genomic segments in linkage disequilibrium, ‘‘old SNPs’’ are more prevalent and have undergone more recombinations with other haplotypes. On average, frequency is used as a surrogate for age [10,16]. In our cohorts, neither ‘‘old’’ nor ‘‘young SNPs’’ were found to be associated with mycobacterial disease. Our luciferase reporter system showed that SNP G-611A had a more prominent functional effect than SNP T-56C and that constructs carrying G-611 had stronger promoter activity than those carrying -611A. Recently, Ju¨liger et al. [17] explored the functional effects of SNP T-56C in stimulated B cell lines with a luciferase reporter system and electrophoretic mobility shift assays. Since they do not report their sequence at -611 and we found that G-611A is more influential than T-56C, their luciferase results are not easily compared to ours. However, we agree that under baseline conditions, T-56C has little effect on promoter activity. We do not yet have a mechanistic explanation for how these promoter polymorphisms exert their effect on the expression of IFNGR1 or, conceivably, other genes. Computer programs predict a GATA-1/TFIID consensus binding site around position-611 and an AP-2/AP-4 consensus binding site around position -56 [18 –20]. Recently, reports of the association of IFNGR1 promoter polymorphism and infectious disease susceptibility have been published. Koch et al. [21] reported that in the Mandika, the major Gambian ethnic group, heterozygosity for IFNGR1 polymorphism T-56C was associated with protection from development of or death from cerebral malaria. Interestingly, they also reported another polymorphism (T-270C) that we did not find in our study. However, Ju¨liger et al. [17] studied a pediatric Gabonese population with mild or severe malaria but did not find any difference in the IFNGR1 T-56C SNP distribution between these
S.D. Rosenzweig et al. / Clinical Immunology 112 (2004) 113–119
groups. Thye et al. [22] reported that allele T-56, along with two other IFNGR1 coding sequence polymorphisms, was associated with a higher rate of anti-Helicobacter pylori IgG antibodies. Controversial results regarding pulmonary tuberculosis susceptibility and IFNGR1 coding sequence polymorphisms have also been described [23,24]. Three groups of mycobacteria-infected cases were examined in Caucasians (PTB, PNTM, and DNTM) and one in African Americans (PTB). Since the three infections examined here are clinically distinct, we have shown comparisons of Caucasian cases to Caucasian controls for each type separately. However, even combining all three types of cases (PTB, PNTM, and DNTM) in an association analysis of IFNGR1 polymorphisms did not yield significant associations either (data not shown). Independent of the way the information was compiled, no consistent differences were found between cases and controls among Caucasians or African Americans. These data suggest that IFNGR1 MPR polymorphisms, even those affecting function, are not significantly associated with mycobacterial infections in Caucasians or pulmonary tuberculosis in African Americans. IFNGR1 MPR SNPs may not play a critical role in mycobacterial susceptibility. Based on our luciferase reporter assays and previously reported data, we found that, although haplotypes differ in terms of activity, they are not associated with mycobacterial infections. This is not terribly surprising, since individuals heterozygous for IFNgR1 deficiency do not appear to have increased susceptibility to mycobacterial diseases [1,2]. However, other phenomena, such as autoimmune processes, may be linked to IFNGR1 MPR polymorphisms [25,26]. Further studies are warranted to determine the specific factors involved in G-611A functional effects and whether these SNPs are relevant to other IFNg pathway-associated diseases. References [1] S.E. Dorman, S.M. Holland, Interferon-g and interleukin-12 pathway defects and human disease, Cytokine Growth Factor Rev. 11 (2000) 321 – 333. [2] J.-L. Casanova, L. Abel, Genetic dissection of immunity to mycobacteria: the human model, Annu. Rev. Immunol. 20 (2002) 581 – 620. [3] E. Jouanguy, S. Lamhamedi-Cherradi, D. Lammas, et al., A human IFNGR1 small deletion hotspot associated with dominant susceptibility to mycobacterial infection, Nat. Genet. 21 (1999) 370 – 378. [4] M. Cargill, D. Altshuler, J. Ireland, P. Sklar, K. Ardlie, N. Patil, C.R. Lane, E.P. Lim, N. Kalyanaram, J. Namesh, L. Ziaugra, L. Friedland, A. Rolfe, J. Warrington, R. Lipshutz, G.Q. Daley, E.S. Lander, Characterization of single-nucleotide polymorphisms in coding regions of human genes, Nat. Genet. 22 (1999) 231 – 238. [5] K.A. Tucker, M.B. Lilly, L. Heck Jr., T.A. Rado, Characterization of a new human diploid myeloid cell line (PLB-985) with granulocytic and monocytic differentiation capacity, Blood 70 (1987) 372 – 378. [6] D. Fallin, A. Cohen, L. Essioux, et al., Genetic analysis of case/ control data using estimated haplotype frequencies: application to APOE locus variation and Alzheimer’s disease, Genome Res. 11 (2001) 143 – 151.
119
[7] M. Stephens, N. Smith, P. Donnelly, A new statistical method for haplotype reconstruction from population data, Am. J. Hum. Genet. 68 (2001) 978 – 989. [8] D. Gusfield, Inference of haplotypes from samples of diploid populations: complexity and algorithms, J. Comp. Biol. 8 (2001) 305 – 323. [9] G. Merlin, B.-J.M. van der Leede, K. McKune, et al., The gene for the ligand binding chain of the human interferon gamma receptor, Immunogenetics 45 (1997) 413 – 421. [10] M.K. Halushka, J.-B. Fan, K. Bentley, L. Hsie, N. Shen, A. Weder, R. Cooper, R. Lipshutz, A. Chakravarti, Patterns of single-nucleotide polymorphisms in candidate genes for blood-pressure homeostasis, Nat. Genet. 22 (1999) 239 – 247. [11] A. Richardson, F. Sisay-Joof, H. Ackerman, S. Usen, P. Katundu, M. Molyneux, M. Pinder, D. Kwiatkowski, Nucleotide diversity of the TNF gene region in an African village, Genes Immun. 2 (2001) 343 – 348. [12] J.L. Flynn, M.M. Goldstein, J. Chan, K.J. Triebold, K. Pfeffer, C.J. Lowenstein, R. Schreiber, T.W. Mak, B.R. Bloom, Tumor necrosis factor-a is required in the protective immune response against Mycobacterium tuberculosis in mice, Immunity 2 (1995) 561 – 572. [13] J.B.S. Haldane, Disease and evolution, La Ricerca Sci. 19 (1947) 68 – 76 (Suppl.). [14] R. Lazarus, D. Vercelli, L.J. Palmer, W.J. Klimecki, E.K. Silverman, B. Richter, A. Riva, M. Ramoni, F.D. Martinez, S.T. Weiss, D.J. Kwiatkowski, Single nucleotide polymorphisms in innate immunity genes: abundant variation and potential role in complex human disease, Immunol. Rev. 190 (2002) 9 – 25. [15] E. Zietkiewicz, V. Yotova, M. Jarnik, M. Korab-Laskowska, K.K. Kidd, D. Modiano, R. Scozzari, M. Stoneking, S. Tishkoff, M. Batzer, D. Labuda, Nuclear DNA diversity in worldwide distributed human populations, Gene 205 (1997) 161 – 167. [16] A. Chakravarti, Population genetics—Making sense out of sequence, Nat. Genet. 21 (1999) 56 – 60 (Suppl.). [17] S. Ju¨liger, M. Bongartz, A.J.F. Luty, P.G. Kremsner, J.F.J. Kun, Functional analysis of a promoter variant of the gene encoding the interferon-g chain I, Immunogenetics 54 (2003) 675 – 680. [18] http://www.cbil.upenn.edu/tess/. [19] http://www.genomatix.de. [20] A. Aranburu, R. Carlsson, C. Persson, T. Leanderson, Transcription factor AP-4 is a ligand for immunoglobulin-n promoter E-box elements, Biochem. J. 354 (2001) 431 – 438. [21] O. Koch, A. Awomoyi, S. Usen, M. Jallow, A. Richardson, H. Jeremy, M. Pinder, M. Newport, D. Kwiatkowski, IFNGR1 gene polymorphisms and susceptibility to cerebral malaria, J. Infect. Dis. 185 (2002) 1684 – 1687. [22] T. Thye, G.D. Burchard, M. Nilius, B. Mu¨ller-Myhsok, R.D. Horstmann, Genomewide linkage analysis identifies polymorphism in the human interferon-g receptor affecting Helicobacter pylori infection, Am. J. Hum. Genet. 72 (2003) 448 – 453. [23] D.A. Fraser, L. Bulat-Kardum, J. Knezevic, P. Babarovic, N. Matakovic-Mileusnic, J. Dellacasagrande, D. Matanic, J. Pavelic, Z. Beg-Sec, Z. Dembic, Interferon-g receptor-1 gene polymorphism in tuberculosis patients from Croatia, Scand. J. Immunol. 57 (2003) 480 – 484. [24] M.J. Newport, A.A. Awomoyi, J.M. Blackwell, Polymorphisms in the interferon-g receptor-1 gene and susceptibility to pulmonary tuberculosis in the Gambia, Scand. J. Immunol. 58 (2003) 383 – 385. [25] H. Nakashima, H. Inoue, M. Akahoshi, Y. Tanaka, E. Ogami, S. Nagano, et al., The combination of polymorphisms within interferon-g receptor 1 and 2 associated with the risk of systemic lupus erythematosus, FEBS Lett. 453 (1999) 187 – 190. [26] Y. Tanaka, H. Nakashima, C. Hisano, T. Kohsaka, Y. Nemoto, H. Niro, et al., Association of the interferon-gamma receptor variant (Val14Met) with systemic lupus erythematosus, Immunogenetics 49 (1999) 266 – 271.