- Email: [email protected]
Interferon-γ gene+874T–A polymorphism is associated with tuberculosis and gamma interferon response
ARTICLE IN PRESS Tuberculosis (2007) 87, 225–230
Available at www.sciencedirect.com
journal homepage: http://intl.elsevierhealth.com/journals/tube
Interferon-c gene+874T–A polymorphism is associated with tuberculosis and gamma interferon response Nilgu ¨n Sallakcıa,1, Mesut Coskuna,1, Zafer Berberb, Fuat Gu ¨rkanc, ¨lnar Uysald, Sabin Bhujub, Ugur Yavuzera,e, Halil Kocamazc, Gu ˘ina, Mahavir Singhb, Olcay Yeg a
Akdeniz University Health Sciences Research Centre, Akdeniz University, Faculty of Medicine, Antalya, Turkey Department of Genome Analysis, GBF, German National Center for Biotechnology, Germany c Department of Pediatric Infectious Diseases, Dicle University, Diyarbakır, Turkey d Dıs- kapı Childrens Hospital, Ankara, Turkey e Department of Physiology, Akdeniz University Health Sciences Research Centre, Akdeniz University, Faculty of Medicine, Antalya, Turkey b
Received 26 April 2006; received in revised form 9 October 2006; accepted 17 October 2006
KEYWORDS Tuberculosis; Interferon g gene+874T–A polymorphism (IFN-g+874 T–A); ELISPOT; Gamma-Interferon
Summary Interferon-g is the most important cytokine in resistance to mycobacterial diseases and common variants of interferon-g gene could be related to tuberculosis susceptibility. We tested the hypothesis that the interferon-g+874T–A polymorphism is associated with tuberculosis disease, and affects the interferon-g reponse. We determined by pyrosequencing the distribution of the interferon-g+874T–A polymorphism in a Turkish population of 319 patients with pulmonary tuberculosis, 42 children with severe forms of tuberculosis and 115 healthy donors. We also analysed whether any correlation exists between this polymorphism and interferon-g response to Mycobacterium tuberculosis antigens by ELISPOT in 58 pulmonary tuberculosis cases, and the results were analysed according to the genotypes. We found that the minor allele (T) frequency was significantly lower in patients with pulmonary tuberculosis when compared to controls (P ¼ 0.024, OR ¼ 0.7), a similarly significant decrease in the frequency of TT genotype was observed in patients with pulmonary tuberculosis, compared to the control group (P ¼ 0.02, OR ¼ 0.49). IFN-g responses to PPD antigen in TT genotype was found to be significantly higher than the AA group (P40.001). Non-parametric correlation analysis of ELISPOT data showed significant reverse correlation in PPD, CFP10 and ESAT6 values and IFN-g +874 genotypes. These results show that the IFN-g +874T-A polymorphism is related to the IFN-g response and the magnitude of the response decreases during transition from TT- to TA and to AA genotypes. Our data suggest that similar to various Caucasian populations, in a Turkish population the
Corresponding author. Tel.: +90 2422496000. 1
˘in). E-mail address: [email protected] (O. Yeg These authors contributed equally.
1472-9792/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.tube.2006.10.002
ARTICLE IN PRESS 226
N. Sallakcı et al. IFN-g+874 T–A polymorphism is also associated with tuberculosis disease and affects the magnitude of the IFN-g response. & 2006 Elsevier Ltd. All rights reserved.
Introduction Only 10% of individuals infected with Mycobacterium tuberculosis develop clinical disease. The development of disease or resistance to it, is determined by various environmental and host genetic factors.1–4 Mutations in five of the genes related to the IL-12-gamma-interferon pathway have been documented to increase susceptibility to mycobacterial infections.5–9 These mutations are found within relatively rare families and exhibit clear Mendelian susceptibility to atypical mycobacterial infections. The majority of patients with tuberculosis (TB) are not affected by one of these rare deleterious mutations, but these studies showed the importance of these cytokines in resistance to mycobacterial disease and suggested that common gene variants of interferon gamma gene (IFN-g) could account for the differences in tuberculosis susceptibility. The +874T4A polymorphism of IFN-g encompasses the binding site for the transcription factor NFkB, which induces the gene expression. Therefore, it has been suggested that the +874 A and T alleles correlate with low and high IFN-g production, respectively.10–13 The aim of this study was firstly, to analyse the association between the +874T4A polymorphism of IFN-g gene and TB disease in a Turkish population and secondly, using the ELISPOT assay, to determine if any correlation exists between the +874 genotypes and IFN-g production capacity of mononuclear cells against M. tuberculosis antigens.
Subjects and methods Patients In this study, a total of 361 patients with various forms of TB (319 pulmonary TB, 35 TB meningitis and 7 miliary TB) and ethnically matched 115 healthy subjects, all of whom gave informed consent, were analysed. Patients with pulmonary TB (305 adults and 14 children) were being followed up by State Tuberculosis Control Centre in Antalya. Peripheral blood samples from patients with meningitis TB or miliary tuberculosis (MTB) were provided by Dicle University Department of Pediatric Infectious Diseases (29 samples) and Diskapi Childrens’ Hospital (6 samples) and Akdeniz University Department of Pediatrics (7 samples) whose ages varied between 4 months to 14 years. The diagnostic criteria for TB were defined as the presence of at least one of the following: (1) Clinical, radiological and laboratory findings consistent with TB disease and positive sputum or cerebrospinal fluid (CSF) smears for acid-fast bacilli (on at least two separate occasions for pulmonary TB, CSF biochemical and cellular findings consistent with TB, and one positive smear sought for TB meningitis cases), (2) culture positivity of sputum, bronchial lavage, pleural fluid and/or cerebrospinal fluid.
Patients who failed to provide conclusive evidence of TB disease or who had additional diseases such as HIV positivity, malnutrition, or diabetes mellitus were excluded from the study. The 115 healthy individuals comprising the control group were gathered from the living-related transplant donors at Akdeniz University Transplantation Centre. These individuals were particularly chosen because they go through excessive examination for the presence of any kind of diseases, including TB and none of them were reported to develop TB during the 2 years’ period during which this study was performed.
Genotyping For genotyping of IFN-g +874 polymorphism, pyrosequencing was employed. For this purpose, genomic DNA was isolated from peripheral blood using the genomic DNA purification Kit (MBI Fermentas, St Leon-Rot Germany). The genomic DNA encompassing the polymorphic site was amplified via PCR (Initial denaturation step for ten minutes at 95 1C, followed by 50 cycles of 30 s at 95 1C, 30 s at 60 1C and 30 s at 72 1C) using the biotin labelled forward (50 Biotin CGTTGCTCACTGGGATTTTG 30 ) and reverse primers (50 CAAACATGTGCGAGTGTGTG 30 ). Each PCR reaction contained; 50 ng of genomic DNA, 5 pmol of each primers, 1.5 mM MgCl2, 0.2 mM of each dNTPs and 1.25 U Ampli Taq Gold DNA polymerase (Qiagen) in a 25 ml of total reaction volume. After the amplification reaction, 15 ml of PCR product was mixed with 3 ml of streptavidin-coated sepharose beads (Amersham Pharmacia Biotech) in 37 ml of binding buffer (10 mM Tris pH 7.6, 2 mM NaCl, 1 mM EDTA and 0.1% Tween 20) and pyrosequencing was performed using the oligonucleotide (50 CACTCGCACATGTTTGG 30 ) at a 20 mM final concentration. The sequencing reactions and genotyping were performed at PSQ 96 A pyrosequencing machine (Biotage AB) and the data was analysed using PSQ tm HS96A software, according to the manufacturer’s instructions (www.pyrosequencing.com).
ELISPOT assays Peripheral mononuclear cells were isolated from freshly collected (maximum 2 h before the test) peripheral blood by density gradient and adjusted to 1.5 106 cells/ml in complete medium (RPMI-1640, containing 10% FCS, 1% penicillin/streptomycin, 1% Fungizone and 2% L-glutamine). The 96-well plates (Euroclone Ltd UK product code 050) were coated with anti-IFN-g antibody in PBS overnight at 4 1C. Next morning, the plates were washed with PBS and blocked with complete medium. Antigens were prepared in 100 ml of complete medium with a final concentration of 20 mg/ml of PPD (Stateus Serum Institut Denmark) and 10 mg/ ml of either ESAT6 or CFP10 (kindly provided by Lionex Diagnostic and Therapeutics GmbH-Germany). Each antigen solution was then mixed with 1.5 105 mononuclear cells,
ARTICLE IN PRESS Interferon-g polymorphism
227
plated into each well of the pre-prepared 96-well plates in duplicates and incubated for 18 h in an incubator at 37 1C, supplied with 5% CO2. At the end of the incubation period, the wells were washed with PBS containing 0.05% Tween-20, alkaline phosphatase- labelled anti-IFN-g antibody was added into each well and incubated for further 2 h. The spots were visualized by using streptavidin–Alkaline phosphatase conjugate in PBS containing 1% BSA. The reaction was stopped by washing with ddH2O extensively. Spots were counted after overnight incubation, using automatic reader (CTL GmbH- Germany) by predefined criteria. Results were expressed as means of the difference between the number of antigen-stimulated spots and unstimulated (medium only) spots. The ratio of antigen-stimulated cells to the unstimulated cells were presented as stimulation indexes. The cutoff for positivity was taken as at least twofold increase of spots in stimulated wells.
Statistical analyses Genotype data Hardy–Weinberg equilibrium was assessed using the genotype data by the w2 goodness-of-fit test. Genetic associations were evaluated both at the allelic and genotype level.14 Allele and genotype frequencies in cases and controls were compared using the w2 test and odds ratios (OR) and their 95% confidence intervals (CI) were calculated. Additive genetic model using all three genotype frequencies was tested using Armitage’s trend test and common odds ratio which provides a single parameter for changes from one genotype category to another was calculated as decsribed elsewhere.15 The genotype data analysis was performed online using the DeFinetti program for tests for deviation from Hardy–Weinberg equilibrium and tests for association (Strom TM & Wienker TF, http://ihg.gsf.de/cgi-bin/hw/ hwa1.pl).
ELISPOT data Statistical evaluation of differences among groups was performed by non-parametric Kruskall–Wallis variance analysis. Statistically significant results were further analysed by Mann–Whitney U test to identify the differences between two groups. Relationships between the genotypes and the ELISPOT data was tested by a non-parametric correlation test. All calculations were performed using the SPSS program, v.11 (SPSS Chicago, IL, USA). Table 1
A Allele T Allele AA AT TT
Results Genotyping for IFN-c +874T4A polymorhism Both cases and controls did not show any violation of Hardy–Weinberg equilibrium. Genotype frequencies of patients with pulmonary, meningitis or miliary TB and healthy controls for the IFN-g +874T4A polymorhism are given in Table 1. The minor allele (T) frequency was significantly lower in patients with pulmonary TB compared to controls (47.8% vs. 39%, P ¼ 0.024, OR ¼ 0.7, 95% CI ¼ 0.52–0.96). The Armitage trend test revealed a gradually increasing protection from the AA genotype through AT to TT (common odds ratio ¼ 0.7; P ¼ 0.02). Similar to the allele frequencies, a statistically significant decrease in the frequency of TT genotype was observed in patients with pulmonary TB, 15%, compared to the control group, 23%, (P ¼ 0.02, OR ¼ 0.49, 95% CI ¼ 0.26–0.91). The distribution of IFN-g +874T4A polymorphism in meningitis and miliary TB cases did not show a significant difference from the controls in allele frequencies (P ¼ 0.09). Comparison of the genotype frequencies, however, yielded a difference between cases and controls and reflected the protective association with the TT genotype seen in the pulmonary TB group (P ¼ 0.07; OR ¼ 0.57). The lack of statistical significance was due to smaller size of this group but the odds ratio was highly suggestive of a similar influence of the TT genotype.
Enumaration of IFN-c producing mononuclear cells In an attempt to establish a possible correlation between the capacity of mononuclear cells to produce IFN-g and different genotypes, ELISPOT assay was employed. For this analysis, we chose the patients who were diagnosed as having pulmonary TB for the first time, thus before any treatment was started. Mononuclear cells obtained from these patients were either unstimulated or stimulated with various M. tuberculosis antigens, namely PPD, CFP10 or ESAT6. The IFN-g producing cells were then counted as described in materials and method. The response of mononuclear cells towards these antigens and correlations with the IFN-g +874 polymorphism is shown in Table 2. Since the stimulation indices are calculated as the ratio of the number of stimulated cells to the unstimulated cells, the values of this ratio varies and there is no clear cut, biologically relevant, data to decide which response or ratio should be regarded as positive. To overcome this obstacle, we
IFN-g +874T4A allele and genotype frequencies (n, %) in cases and controls Pulmonary TB (n ¼ 319)
Severe forms of Tuberculosis (TB Meningitis (n ¼ 35) and Miliary TB (n ¼ 7))
Controls (n ¼ 115)
387 251 115 157 47
53 31 14 25 3
120 110 31 58 26
(60.7) (39.3) (36) (49) (15)
(63) (37) (33) (59.5) (7)
T allele and TT genotype frequencies were significantly lower in cases (see text).
(52.2) (47.8) (27) (50) (23)
ARTICLE IN PRESS 228
N. Sallakcı et al.
Table 2
ELISPOT results for each genotype
Genotype
N
PPD SP
PPDndx
CFP10 SP
CFP10ndx
ESAT-6 SP
ESAT6ndx
AA AT TT
24 21 13
38 (33) 68 (51) 92 (52)
3.1 (2.2) 5.1 (5.3) 6.0 (5.5)
29 (35) 40 (45) 48 (26)
2.2 (1.6) 2.1 (2.8) 3.2 (2.3)
25 (39) 32 (20) 35 (27)
2.1 (1.8) 2.6 (2.0) 3 (2.6)
SP: Means of spot numbers and standard deviations ( ) of the difference between the number of antigen-stimulated and unstimulated cells (150,00 mononucler cells added to each well). Stimulation index (ndx): ratio of the spot numbers of antigen-stimulated to unstimulated cells.
100
A/A
T/T
90
T/A
80 70 60 50 40 30 20 10 0
Figure 1
PPD
CFP10
ESAT-6
Percentage of positive response upon stimulation by different antigens in relation to genotypes.
accepted the cut-off value as 2.0 and any stimulation index value above the level of 2.0 was considered as positive as used by Lalvani et al.16,17 When we analysed the percentage of patients whose stimulation indices were above 2.0, responses to PPD and ESAT6 exhibited a gradual decrease from the patients with TT genotype (93% and 85%): TA (85% and 67%) and AA (75% and 42%) (Fig. 1). Thus, patients carrying a T allele seemed to respond better towards these antigens and produced greater amounts of IFN-g, whereas, the response is relatively weak in patients with AA genotype. Also for CFP10 antigen, patients with a TT genotype showed better response (70%) as compared to those with AA genotype (59%) but the differences were not statistically significant. In our experience and as has been reported previously,18 for the efficiency of ELISPOT assays, freshly collected blood should be used since the capacity of cytokine production of mononuclear cells diminishes to a great extent if the sample is not processed immediately. Due to this technical requirement, we were not able to perform ELISPOT assays on samples obtained from patients with miliary TB or TB meningitis, since most of them lived in other cities, from where transportation of samples would have taken more than 8 h. The statistical significance of the changes observed was assessed by non-parametric tests because spot numbers and indices did not show a normal distribution. IFN-g response to PPD antigen in TT genotype was found to be
significantly higher than that in the AA group for both spot numbers (P40.001) and PPD indices (P ¼ 0.011). However, comparisons of patients with TA genotype to those with AA genotype exhibited significant difference in PPD spot (P ¼ 0.02), but not in index values (P ¼ 0.1). We also looked for the relation of genotypes and ELISPOT data by nonparametric correlation analysis and found significant reverse correlation in PPD spot, PPD index, CFP10 spot and ESAT6 spot values and IFN-g +874 genotypes (PPD spot r ¼ 0.49, Po0.001), PPD index (r ¼ 0.33, P ¼ 0.012), CFP10 (r ¼ 0.29, P ¼ 0.027), ESAT6 (r ¼ 0.27, P ¼ 0.042). These results suggest that the IFN-g +874T4A polymorphism is related with IFN-g response and the magnitude of response decreases during transition from TT- to TA and to AA genotypes.
Discussion This is the first study exhibiting the genotype profile of the Turkish population for the IFN-g +874T4A polymorphism and its association with susceptibility to TB. In addition, using ELISPOT assays, we showed that this particular polymorphism affects the ability of mononuclear cells to produce IFNg, a well-known cytokine that plays important roles in immune response of the host towards mycobacterial antigens. We demonstrated that having a TT genotype for
ARTICLE IN PRESS Interferon-g polymorphism
229
the+874 position of the IFN-g gene decreases the risk of developing pulmonary TB by 30% whereas, the AA genotype causes 1.41-fold increase. Similar association was reported in Sicilian, Spanish, South African and Chinese populations.19–23 The genotypes for the IFN-g +874 region in the healthy control subjects revealed the genotype profile of Turkish population, which was very similar to Spanish, Sicilian and Croatian populations and as expected, showed variations from Chinese or South African colored populations (Table 3). Excepting the Croatian population,23 all of the above mentioned studies found a relation between this polymorphism and tuberculosis, either a protective effect of TT genotype (this study and Ref.19) or increased susceptibility with the AA genotype. The study performed on a Croatian population reported an association with severity of tuberculosis rather than susceptibility.23 By studying a different Caucasian population and finding similar results to those published previously, we demonstrated that the genetic profile of this particular polymorphism and its association with tuberculosis does not vary across ethnic boundaries. Furthermore, this relationship seems to persist in severe forms of tuberculosis such as TB meningitis or miliary tuberculosis (Table 1). There is convincing clinical and experimental evidence showing that IFN-g plays a key role in the control of mycobacterial infections. IFN-g gene-disrupted mice show increased susceptibility to TB and replacement of the gene directly in the lung confers the resistance.5,24,25 Moreover, it has been shown that IFN-g receptor deficiency results in increased susceptibility to mycobacterial infections in humans 6–9 and the clinical manifestations of tuberculosis are related with M. tuberculosis stimulated IFN-g production.26,27 Using ELISA methods Lopez-Maderuelo et al. 21 showed that IFN-g production in response to PPD antigen was associated with the IFN-g +874T4A polymorphism. We have chosen another approach to determine the functionality of this particular polymorphism by using ELISPOT assays. The ELISPOT assay measures the number of cells producing a particular protein and therefore is accepted to be more sensitive with respect to ELISA.16–18 In addition, we attempted to reveal the response of mononuclear cells to specific antigens of M. tuberculosis, such as CFP10 and ESAT6, besides PPD, which is known to contain many other components as well. CFP10 and ESAT6 are absent in the vaccine strain BCG and in several other mycobacteria. We have shown that depending on the genotype, the capacity of mononuclear cells to produce IFN-g changed significantly in
Table 3 Genotype frequencies (%) for IFN-g +874T4A polymorphism in different populations Population
AA
AT
TT
References
Chinese South African Sicillian Spanish Croatian Turkish
45.7 47 25.8 31 26 27
42 42 48.4 50 54.3 50
12.2 11 25.8 19 20 23
22
response to both ESAT6 and CFP10 mycobacterial antigens, and carrying a T allele was found to result in higher numbers of IFN-g producing mononuclear cells. This finding also supports our genotyping data that the TT genotype confers resistance to tuberculosis (Table 2). The results that we obtained using PPD antigen were similar to those previously published, indicating that our assay system is working reliably. The findings we presented in this study, together with the previously published data confirm the functionality of this polymorphism, with a consequence of altered IFN-g levels. However, the mechanism of altered gene expression associated with this specific polymorphism is still poorly understood. Specific binding of NFkB to DNA sequences containing the +874 T allele have been reported 11,12 and there is evidence that this could have functional effects on transcription of IFN-g genes and could influence the rate of IFN-g production.12,13 Although this could explain the higher number of IFN-g producing cells in individuals with a TT genotype, nevertheless it is worthwhile to mention that we observed a few cases of AA genotype exhibiting high numbers of IFN-g producing cells or vice versa, which implies the role of factors, other than NFkB, in regulation of IFN-g gene expression. There were significant changes in TT and AA genotypes distribution in severe forms of TB (meningitis or miliary TB, n ¼ 41), but we were not able to find a similar significant difference in allele frequency as found in pulmonary TB cases (Table 1). This could mainly be due to the small number of patients. However, the lack of age-matched controls should also be considered, since our control group consisted of adults whereas, the patients with meningitis or miliary TB were within the pediatric age group. Considering that, IFN-g +874 polymorhism has also been associated with longevity,28 we believe it is essential to analyse these two groups with an age-matched control group before reaching a conclusion. Although we and others showed that the IFN-g +874T4A polymorphism is associated with tuberculosis disease and effects the magnitude of IFN-g responses to mycobacterial antigens this relation explains only a minority of patients. The genetic component that increases susceptibility to TB disease probably involves multiple alleles on different genes and chromosomes. Further studies on immunoregulatory genes especially, genes on the IL-12—IFN-g axis, are needed.
Acknowledgements This study was supported by Akdeniz University Scientific Research Projects Units (No: 2004.04.0103.007). We thank doctors of Antalya State Tuberculosis Center for their collaboration and M. Tevfik Dorak MD PhD, University of Newcastle upon Tyne for his assistance with statistical analyses and reviewing this manuscript.
20 19 21 23
This study
References 1. Casanova JL, Abel L. Genetic dissection of immunity to mycobacteria the human model. Annu Rev Immunol 2002;20:581–620.
ARTICLE IN PRESS 230 2. Flynn JL, Chan J. Immunology of tuberculosis. Annu Rev Immunol 2001;19:93–129. 3. Guide SV, Holland SM. Host susceptibility factors in mycobacterial Infections. Infect Dis Clin N Am 2002;16:163–85. 4. Hill AVS. The immunogenetics of human infectious diseases. Annual Rev Immunol 1998;16:593–617. 5. Collins HL, Kaufmann SHE. The many faces of host response to tuberculosis. Immunology 2001;103:1–9. 6. Dorman SE, Picard C, Lammas D, Heyne K, van Diesel JT, Baretto R, et al. Clinical feautures of dominant and recessive interferon gamma receptor 1 deficiencies. Lancet 2004;364:2113–21. 7. Jouanguy E, Altare F, Lamhamedi S, Revy P, Emile J-F, Newport M, et al. Interferon-g-receptor deficiency in an infant with fatal bacille Calmette–Guerin infection. NEJM 1996;335:1956–61. 8. Newport MJ, Huxley CM, Huston S, Hawrylowicz CM, Oostra BA, Williamson R, et al. A mutation in the interferon-g-receptor gene and susceptibility to mycobacterial infection. NEJM 1996;335:1941–9. 9. Dorman SE, Holland SM. Mutation in the signal transducing chain of the interferon-g receptor and susceptibility to mycobacterial infection. J Clin Invest 1998;101:2364–9. 10. Bream JH, Carringhton M, O’Toole S. Polymorphism of the human IFN-g gene noncoding region. Immunogenetics 2000;51: 50–8. 11. Pravica V, Perrey C, Stevens A, Lee JH, Hutchinson IV. A single nucleotide polymorphism in the first intron of the human IFN-g gene. Human Immunol 2000;61:863–6. 12. Pravica V, Asderakis A, Perrey C, Hajeer A, Sinnott PJ, Huychinson IV. In vitro production of IFN-g correlates with CA repeat polymorphism in the human IFN-g gene. Eur J Immunogenet 1999;26:1–3. 13. Heinemeyer T, Wingender E, Reuter I, Hermjakop H, Kel AE, Kel OV, et al. Database on transcriptional regulation: TRANSFAC, TRRD and COMPEL. Nucleic Acids Res 1998;26:362–7. 14. Lewis CM. Genetic association studies: design, analysis and interpretation. Brief Bioinform 2002;3:146–53. 15. Sasieni PD. From genotypes to genes: doubling the sample size. Biometrics 1997;53:1253–61. 16. Lalvani A, Pathan AA, Durkan H, Wilkinson KA, Whelan A, Deeks JJ, et al. Enhanced contact tracing and spatial tracking of Mycobacterium tuberculosis infection by enumeration of antigen-specific T cells. Lancet 2001;9273:2017–21. 17. Ewer K, Deeks J, Alvarez L, Bryant G, Waller S, Andersen P, et al. Comparison of T-cell- based assay with tuberculin skin test
N. Sallakcı et al.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
for diagnosis of Mycobacterium tuberculosis infection in a school tuberculosis outbreak. Lancet 2003;9364:1168–73. Doherty TM, Demisse A, Menzies D, Andersen P, Rook G, Zumla A. Effect of sample handling on analysis of cytokine response to Mycobacterium tuberculosis in clinical samples using ELISA, ELISPOT and quantitative PCR. J of Immunol Methods 2005;298:129–41. Lio D, Marino V, Serauto A, Gioia V, Scola L, Crivello A, et al. Genotype frequencies of the +874 T–A single nucleotide polymorphism in the first intron of the interferon-g gene in a sample of Scilian patients affected by tuberculosis. Eur J Immunogenet 2002;29:371–4. Rossouw M, Nel HJ, Cooke GS, van Helden PD, Hoal EG. Association between tuberculosis and polymorphic NF (kappa) B binding site in the interferon gamma gene. Lancet 2003;361: 1871–2. Lopez-Maderuelo D, Arnalich F, Serantes R, Gonzalez A, Codoceo R, Madero R, et al. Interferon-g and interleukin-10 gene polymorphisms in pulmonary tuberculosis. Am J Respir Crit Care Med 2003;167:970–5. Tso HW, Ip WK, Chong WP, Tam CM, Chiang AK, Lau YL. Association of interferon gamma and interleukin 10 genes with tuberculosis in Hong Kong Chinese. Genes Immun 2005;6:358–63. Etokebe GE, Bulat-Kardum L, Johansen MS, Knezevic J, Balen S, Matakovic-Mileusnic N, et al. Interferon-g gene (T874A and G2109A) polymorphisms are associated with microscopy-positive tuberculosis. Scand J Immunol 2006;63:136–41. Cooper AM, Dalton DK, Stewart TA, Griffin JP, Russel DG, Orme IM. Disseminated tuberculosis in interferon g gene-disrupted mice. J Exp Med 1993;178:2243. Moreira AL, Tsenova L, Murray PJ, Freeman S, Bergtold A, Chiriboga L, et al. Aerosol infection of mice with recombinant BCG secreting murine IFN-g partially reconstitutes local protective immunity. Microb Pathogens 2000;29:175. Zhang M, Lin Y, Iyer DV, Gong J, Abrams JS, Barnes PF. T cell cytokine responses in human infection with Mycobacterium tuberculosis. Infect Immun 1995;63:3231–4. Sodhi A, Gong J-H, Silva C, Barnes PF. Clinical correlates of interferon-g production in patients with tuberculosis. Clin Infect Dis 1997;25:617–20. Lio D, Scola L, Crivello A, Bonafe M, Fransechi C, Olivieri F, et al. Allele frequencies of +874 T–A single nucleotide polymorphism at the first intron of interferon-g gene in a group of Italian centenarians. Exper Gerontology 2002;37:315–9.
Available at www.sciencedirect.com
journal homepage: http://intl.elsevierhealth.com/journals/tube
Interferon-c gene+874T–A polymorphism is associated with tuberculosis and gamma interferon response Nilgu ¨n Sallakcıa,1, Mesut Coskuna,1, Zafer Berberb, Fuat Gu ¨rkanc, ¨lnar Uysald, Sabin Bhujub, Ugur Yavuzera,e, Halil Kocamazc, Gu ˘ina, Mahavir Singhb, Olcay Yeg a
Akdeniz University Health Sciences Research Centre, Akdeniz University, Faculty of Medicine, Antalya, Turkey Department of Genome Analysis, GBF, German National Center for Biotechnology, Germany c Department of Pediatric Infectious Diseases, Dicle University, Diyarbakır, Turkey d Dıs- kapı Childrens Hospital, Ankara, Turkey e Department of Physiology, Akdeniz University Health Sciences Research Centre, Akdeniz University, Faculty of Medicine, Antalya, Turkey b
Received 26 April 2006; received in revised form 9 October 2006; accepted 17 October 2006
KEYWORDS Tuberculosis; Interferon g gene+874T–A polymorphism (IFN-g+874 T–A); ELISPOT; Gamma-Interferon
Summary Interferon-g is the most important cytokine in resistance to mycobacterial diseases and common variants of interferon-g gene could be related to tuberculosis susceptibility. We tested the hypothesis that the interferon-g+874T–A polymorphism is associated with tuberculosis disease, and affects the interferon-g reponse. We determined by pyrosequencing the distribution of the interferon-g+874T–A polymorphism in a Turkish population of 319 patients with pulmonary tuberculosis, 42 children with severe forms of tuberculosis and 115 healthy donors. We also analysed whether any correlation exists between this polymorphism and interferon-g response to Mycobacterium tuberculosis antigens by ELISPOT in 58 pulmonary tuberculosis cases, and the results were analysed according to the genotypes. We found that the minor allele (T) frequency was significantly lower in patients with pulmonary tuberculosis when compared to controls (P ¼ 0.024, OR ¼ 0.7), a similarly significant decrease in the frequency of TT genotype was observed in patients with pulmonary tuberculosis, compared to the control group (P ¼ 0.02, OR ¼ 0.49). IFN-g responses to PPD antigen in TT genotype was found to be significantly higher than the AA group (P40.001). Non-parametric correlation analysis of ELISPOT data showed significant reverse correlation in PPD, CFP10 and ESAT6 values and IFN-g +874 genotypes. These results show that the IFN-g +874T-A polymorphism is related to the IFN-g response and the magnitude of the response decreases during transition from TT- to TA and to AA genotypes. Our data suggest that similar to various Caucasian populations, in a Turkish population the
Corresponding author. Tel.: +90 2422496000. 1
˘in). E-mail address: [email protected] (O. Yeg These authors contributed equally.
1472-9792/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.tube.2006.10.002
ARTICLE IN PRESS 226
N. Sallakcı et al. IFN-g+874 T–A polymorphism is also associated with tuberculosis disease and affects the magnitude of the IFN-g response. & 2006 Elsevier Ltd. All rights reserved.
Introduction Only 10% of individuals infected with Mycobacterium tuberculosis develop clinical disease. The development of disease or resistance to it, is determined by various environmental and host genetic factors.1–4 Mutations in five of the genes related to the IL-12-gamma-interferon pathway have been documented to increase susceptibility to mycobacterial infections.5–9 These mutations are found within relatively rare families and exhibit clear Mendelian susceptibility to atypical mycobacterial infections. The majority of patients with tuberculosis (TB) are not affected by one of these rare deleterious mutations, but these studies showed the importance of these cytokines in resistance to mycobacterial disease and suggested that common gene variants of interferon gamma gene (IFN-g) could account for the differences in tuberculosis susceptibility. The +874T4A polymorphism of IFN-g encompasses the binding site for the transcription factor NFkB, which induces the gene expression. Therefore, it has been suggested that the +874 A and T alleles correlate with low and high IFN-g production, respectively.10–13 The aim of this study was firstly, to analyse the association between the +874T4A polymorphism of IFN-g gene and TB disease in a Turkish population and secondly, using the ELISPOT assay, to determine if any correlation exists between the +874 genotypes and IFN-g production capacity of mononuclear cells against M. tuberculosis antigens.
Subjects and methods Patients In this study, a total of 361 patients with various forms of TB (319 pulmonary TB, 35 TB meningitis and 7 miliary TB) and ethnically matched 115 healthy subjects, all of whom gave informed consent, were analysed. Patients with pulmonary TB (305 adults and 14 children) were being followed up by State Tuberculosis Control Centre in Antalya. Peripheral blood samples from patients with meningitis TB or miliary tuberculosis (MTB) were provided by Dicle University Department of Pediatric Infectious Diseases (29 samples) and Diskapi Childrens’ Hospital (6 samples) and Akdeniz University Department of Pediatrics (7 samples) whose ages varied between 4 months to 14 years. The diagnostic criteria for TB were defined as the presence of at least one of the following: (1) Clinical, radiological and laboratory findings consistent with TB disease and positive sputum or cerebrospinal fluid (CSF) smears for acid-fast bacilli (on at least two separate occasions for pulmonary TB, CSF biochemical and cellular findings consistent with TB, and one positive smear sought for TB meningitis cases), (2) culture positivity of sputum, bronchial lavage, pleural fluid and/or cerebrospinal fluid.
Patients who failed to provide conclusive evidence of TB disease or who had additional diseases such as HIV positivity, malnutrition, or diabetes mellitus were excluded from the study. The 115 healthy individuals comprising the control group were gathered from the living-related transplant donors at Akdeniz University Transplantation Centre. These individuals were particularly chosen because they go through excessive examination for the presence of any kind of diseases, including TB and none of them were reported to develop TB during the 2 years’ period during which this study was performed.
Genotyping For genotyping of IFN-g +874 polymorphism, pyrosequencing was employed. For this purpose, genomic DNA was isolated from peripheral blood using the genomic DNA purification Kit (MBI Fermentas, St Leon-Rot Germany). The genomic DNA encompassing the polymorphic site was amplified via PCR (Initial denaturation step for ten minutes at 95 1C, followed by 50 cycles of 30 s at 95 1C, 30 s at 60 1C and 30 s at 72 1C) using the biotin labelled forward (50 Biotin CGTTGCTCACTGGGATTTTG 30 ) and reverse primers (50 CAAACATGTGCGAGTGTGTG 30 ). Each PCR reaction contained; 50 ng of genomic DNA, 5 pmol of each primers, 1.5 mM MgCl2, 0.2 mM of each dNTPs and 1.25 U Ampli Taq Gold DNA polymerase (Qiagen) in a 25 ml of total reaction volume. After the amplification reaction, 15 ml of PCR product was mixed with 3 ml of streptavidin-coated sepharose beads (Amersham Pharmacia Biotech) in 37 ml of binding buffer (10 mM Tris pH 7.6, 2 mM NaCl, 1 mM EDTA and 0.1% Tween 20) and pyrosequencing was performed using the oligonucleotide (50 CACTCGCACATGTTTGG 30 ) at a 20 mM final concentration. The sequencing reactions and genotyping were performed at PSQ 96 A pyrosequencing machine (Biotage AB) and the data was analysed using PSQ tm HS96A software, according to the manufacturer’s instructions (www.pyrosequencing.com).
ELISPOT assays Peripheral mononuclear cells were isolated from freshly collected (maximum 2 h before the test) peripheral blood by density gradient and adjusted to 1.5 106 cells/ml in complete medium (RPMI-1640, containing 10% FCS, 1% penicillin/streptomycin, 1% Fungizone and 2% L-glutamine). The 96-well plates (Euroclone Ltd UK product code 050) were coated with anti-IFN-g antibody in PBS overnight at 4 1C. Next morning, the plates were washed with PBS and blocked with complete medium. Antigens were prepared in 100 ml of complete medium with a final concentration of 20 mg/ml of PPD (Stateus Serum Institut Denmark) and 10 mg/ ml of either ESAT6 or CFP10 (kindly provided by Lionex Diagnostic and Therapeutics GmbH-Germany). Each antigen solution was then mixed with 1.5 105 mononuclear cells,
ARTICLE IN PRESS Interferon-g polymorphism
227
plated into each well of the pre-prepared 96-well plates in duplicates and incubated for 18 h in an incubator at 37 1C, supplied with 5% CO2. At the end of the incubation period, the wells were washed with PBS containing 0.05% Tween-20, alkaline phosphatase- labelled anti-IFN-g antibody was added into each well and incubated for further 2 h. The spots were visualized by using streptavidin–Alkaline phosphatase conjugate in PBS containing 1% BSA. The reaction was stopped by washing with ddH2O extensively. Spots were counted after overnight incubation, using automatic reader (CTL GmbH- Germany) by predefined criteria. Results were expressed as means of the difference between the number of antigen-stimulated spots and unstimulated (medium only) spots. The ratio of antigen-stimulated cells to the unstimulated cells were presented as stimulation indexes. The cutoff for positivity was taken as at least twofold increase of spots in stimulated wells.
Statistical analyses Genotype data Hardy–Weinberg equilibrium was assessed using the genotype data by the w2 goodness-of-fit test. Genetic associations were evaluated both at the allelic and genotype level.14 Allele and genotype frequencies in cases and controls were compared using the w2 test and odds ratios (OR) and their 95% confidence intervals (CI) were calculated. Additive genetic model using all three genotype frequencies was tested using Armitage’s trend test and common odds ratio which provides a single parameter for changes from one genotype category to another was calculated as decsribed elsewhere.15 The genotype data analysis was performed online using the DeFinetti program for tests for deviation from Hardy–Weinberg equilibrium and tests for association (Strom TM & Wienker TF, http://ihg.gsf.de/cgi-bin/hw/ hwa1.pl).
ELISPOT data Statistical evaluation of differences among groups was performed by non-parametric Kruskall–Wallis variance analysis. Statistically significant results were further analysed by Mann–Whitney U test to identify the differences between two groups. Relationships between the genotypes and the ELISPOT data was tested by a non-parametric correlation test. All calculations were performed using the SPSS program, v.11 (SPSS Chicago, IL, USA). Table 1
A Allele T Allele AA AT TT
Results Genotyping for IFN-c +874T4A polymorhism Both cases and controls did not show any violation of Hardy–Weinberg equilibrium. Genotype frequencies of patients with pulmonary, meningitis or miliary TB and healthy controls for the IFN-g +874T4A polymorhism are given in Table 1. The minor allele (T) frequency was significantly lower in patients with pulmonary TB compared to controls (47.8% vs. 39%, P ¼ 0.024, OR ¼ 0.7, 95% CI ¼ 0.52–0.96). The Armitage trend test revealed a gradually increasing protection from the AA genotype through AT to TT (common odds ratio ¼ 0.7; P ¼ 0.02). Similar to the allele frequencies, a statistically significant decrease in the frequency of TT genotype was observed in patients with pulmonary TB, 15%, compared to the control group, 23%, (P ¼ 0.02, OR ¼ 0.49, 95% CI ¼ 0.26–0.91). The distribution of IFN-g +874T4A polymorphism in meningitis and miliary TB cases did not show a significant difference from the controls in allele frequencies (P ¼ 0.09). Comparison of the genotype frequencies, however, yielded a difference between cases and controls and reflected the protective association with the TT genotype seen in the pulmonary TB group (P ¼ 0.07; OR ¼ 0.57). The lack of statistical significance was due to smaller size of this group but the odds ratio was highly suggestive of a similar influence of the TT genotype.
Enumaration of IFN-c producing mononuclear cells In an attempt to establish a possible correlation between the capacity of mononuclear cells to produce IFN-g and different genotypes, ELISPOT assay was employed. For this analysis, we chose the patients who were diagnosed as having pulmonary TB for the first time, thus before any treatment was started. Mononuclear cells obtained from these patients were either unstimulated or stimulated with various M. tuberculosis antigens, namely PPD, CFP10 or ESAT6. The IFN-g producing cells were then counted as described in materials and method. The response of mononuclear cells towards these antigens and correlations with the IFN-g +874 polymorphism is shown in Table 2. Since the stimulation indices are calculated as the ratio of the number of stimulated cells to the unstimulated cells, the values of this ratio varies and there is no clear cut, biologically relevant, data to decide which response or ratio should be regarded as positive. To overcome this obstacle, we
IFN-g +874T4A allele and genotype frequencies (n, %) in cases and controls Pulmonary TB (n ¼ 319)
Severe forms of Tuberculosis (TB Meningitis (n ¼ 35) and Miliary TB (n ¼ 7))
Controls (n ¼ 115)
387 251 115 157 47
53 31 14 25 3
120 110 31 58 26
(60.7) (39.3) (36) (49) (15)
(63) (37) (33) (59.5) (7)
T allele and TT genotype frequencies were significantly lower in cases (see text).
(52.2) (47.8) (27) (50) (23)
ARTICLE IN PRESS 228
N. Sallakcı et al.
Table 2
ELISPOT results for each genotype
Genotype
N
PPD SP
PPDndx
CFP10 SP
CFP10ndx
ESAT-6 SP
ESAT6ndx
AA AT TT
24 21 13
38 (33) 68 (51) 92 (52)
3.1 (2.2) 5.1 (5.3) 6.0 (5.5)
29 (35) 40 (45) 48 (26)
2.2 (1.6) 2.1 (2.8) 3.2 (2.3)
25 (39) 32 (20) 35 (27)
2.1 (1.8) 2.6 (2.0) 3 (2.6)
SP: Means of spot numbers and standard deviations ( ) of the difference between the number of antigen-stimulated and unstimulated cells (150,00 mononucler cells added to each well). Stimulation index (ndx): ratio of the spot numbers of antigen-stimulated to unstimulated cells.
100
A/A
T/T
90
T/A
80 70 60 50 40 30 20 10 0
Figure 1
PPD
CFP10
ESAT-6
Percentage of positive response upon stimulation by different antigens in relation to genotypes.
accepted the cut-off value as 2.0 and any stimulation index value above the level of 2.0 was considered as positive as used by Lalvani et al.16,17 When we analysed the percentage of patients whose stimulation indices were above 2.0, responses to PPD and ESAT6 exhibited a gradual decrease from the patients with TT genotype (93% and 85%): TA (85% and 67%) and AA (75% and 42%) (Fig. 1). Thus, patients carrying a T allele seemed to respond better towards these antigens and produced greater amounts of IFN-g, whereas, the response is relatively weak in patients with AA genotype. Also for CFP10 antigen, patients with a TT genotype showed better response (70%) as compared to those with AA genotype (59%) but the differences were not statistically significant. In our experience and as has been reported previously,18 for the efficiency of ELISPOT assays, freshly collected blood should be used since the capacity of cytokine production of mononuclear cells diminishes to a great extent if the sample is not processed immediately. Due to this technical requirement, we were not able to perform ELISPOT assays on samples obtained from patients with miliary TB or TB meningitis, since most of them lived in other cities, from where transportation of samples would have taken more than 8 h. The statistical significance of the changes observed was assessed by non-parametric tests because spot numbers and indices did not show a normal distribution. IFN-g response to PPD antigen in TT genotype was found to be
significantly higher than that in the AA group for both spot numbers (P40.001) and PPD indices (P ¼ 0.011). However, comparisons of patients with TA genotype to those with AA genotype exhibited significant difference in PPD spot (P ¼ 0.02), but not in index values (P ¼ 0.1). We also looked for the relation of genotypes and ELISPOT data by nonparametric correlation analysis and found significant reverse correlation in PPD spot, PPD index, CFP10 spot and ESAT6 spot values and IFN-g +874 genotypes (PPD spot r ¼ 0.49, Po0.001), PPD index (r ¼ 0.33, P ¼ 0.012), CFP10 (r ¼ 0.29, P ¼ 0.027), ESAT6 (r ¼ 0.27, P ¼ 0.042). These results suggest that the IFN-g +874T4A polymorphism is related with IFN-g response and the magnitude of response decreases during transition from TT- to TA and to AA genotypes.
Discussion This is the first study exhibiting the genotype profile of the Turkish population for the IFN-g +874T4A polymorphism and its association with susceptibility to TB. In addition, using ELISPOT assays, we showed that this particular polymorphism affects the ability of mononuclear cells to produce IFNg, a well-known cytokine that plays important roles in immune response of the host towards mycobacterial antigens. We demonstrated that having a TT genotype for
ARTICLE IN PRESS Interferon-g polymorphism
229
the+874 position of the IFN-g gene decreases the risk of developing pulmonary TB by 30% whereas, the AA genotype causes 1.41-fold increase. Similar association was reported in Sicilian, Spanish, South African and Chinese populations.19–23 The genotypes for the IFN-g +874 region in the healthy control subjects revealed the genotype profile of Turkish population, which was very similar to Spanish, Sicilian and Croatian populations and as expected, showed variations from Chinese or South African colored populations (Table 3). Excepting the Croatian population,23 all of the above mentioned studies found a relation between this polymorphism and tuberculosis, either a protective effect of TT genotype (this study and Ref.19) or increased susceptibility with the AA genotype. The study performed on a Croatian population reported an association with severity of tuberculosis rather than susceptibility.23 By studying a different Caucasian population and finding similar results to those published previously, we demonstrated that the genetic profile of this particular polymorphism and its association with tuberculosis does not vary across ethnic boundaries. Furthermore, this relationship seems to persist in severe forms of tuberculosis such as TB meningitis or miliary tuberculosis (Table 1). There is convincing clinical and experimental evidence showing that IFN-g plays a key role in the control of mycobacterial infections. IFN-g gene-disrupted mice show increased susceptibility to TB and replacement of the gene directly in the lung confers the resistance.5,24,25 Moreover, it has been shown that IFN-g receptor deficiency results in increased susceptibility to mycobacterial infections in humans 6–9 and the clinical manifestations of tuberculosis are related with M. tuberculosis stimulated IFN-g production.26,27 Using ELISA methods Lopez-Maderuelo et al. 21 showed that IFN-g production in response to PPD antigen was associated with the IFN-g +874T4A polymorphism. We have chosen another approach to determine the functionality of this particular polymorphism by using ELISPOT assays. The ELISPOT assay measures the number of cells producing a particular protein and therefore is accepted to be more sensitive with respect to ELISA.16–18 In addition, we attempted to reveal the response of mononuclear cells to specific antigens of M. tuberculosis, such as CFP10 and ESAT6, besides PPD, which is known to contain many other components as well. CFP10 and ESAT6 are absent in the vaccine strain BCG and in several other mycobacteria. We have shown that depending on the genotype, the capacity of mononuclear cells to produce IFN-g changed significantly in
Table 3 Genotype frequencies (%) for IFN-g +874T4A polymorphism in different populations Population
AA
AT
TT
References
Chinese South African Sicillian Spanish Croatian Turkish
45.7 47 25.8 31 26 27
42 42 48.4 50 54.3 50
12.2 11 25.8 19 20 23
22
response to both ESAT6 and CFP10 mycobacterial antigens, and carrying a T allele was found to result in higher numbers of IFN-g producing mononuclear cells. This finding also supports our genotyping data that the TT genotype confers resistance to tuberculosis (Table 2). The results that we obtained using PPD antigen were similar to those previously published, indicating that our assay system is working reliably. The findings we presented in this study, together with the previously published data confirm the functionality of this polymorphism, with a consequence of altered IFN-g levels. However, the mechanism of altered gene expression associated with this specific polymorphism is still poorly understood. Specific binding of NFkB to DNA sequences containing the +874 T allele have been reported 11,12 and there is evidence that this could have functional effects on transcription of IFN-g genes and could influence the rate of IFN-g production.12,13 Although this could explain the higher number of IFN-g producing cells in individuals with a TT genotype, nevertheless it is worthwhile to mention that we observed a few cases of AA genotype exhibiting high numbers of IFN-g producing cells or vice versa, which implies the role of factors, other than NFkB, in regulation of IFN-g gene expression. There were significant changes in TT and AA genotypes distribution in severe forms of TB (meningitis or miliary TB, n ¼ 41), but we were not able to find a similar significant difference in allele frequency as found in pulmonary TB cases (Table 1). This could mainly be due to the small number of patients. However, the lack of age-matched controls should also be considered, since our control group consisted of adults whereas, the patients with meningitis or miliary TB were within the pediatric age group. Considering that, IFN-g +874 polymorhism has also been associated with longevity,28 we believe it is essential to analyse these two groups with an age-matched control group before reaching a conclusion. Although we and others showed that the IFN-g +874T4A polymorphism is associated with tuberculosis disease and effects the magnitude of IFN-g responses to mycobacterial antigens this relation explains only a minority of patients. The genetic component that increases susceptibility to TB disease probably involves multiple alleles on different genes and chromosomes. Further studies on immunoregulatory genes especially, genes on the IL-12—IFN-g axis, are needed.
Acknowledgements This study was supported by Akdeniz University Scientific Research Projects Units (No: 2004.04.0103.007). We thank doctors of Antalya State Tuberculosis Center for their collaboration and M. Tevfik Dorak MD PhD, University of Newcastle upon Tyne for his assistance with statistical analyses and reviewing this manuscript.
20 19 21 23
This study
References 1. Casanova JL, Abel L. Genetic dissection of immunity to mycobacteria the human model. Annu Rev Immunol 2002;20:581–620.
ARTICLE IN PRESS 230 2. Flynn JL, Chan J. Immunology of tuberculosis. Annu Rev Immunol 2001;19:93–129. 3. Guide SV, Holland SM. Host susceptibility factors in mycobacterial Infections. Infect Dis Clin N Am 2002;16:163–85. 4. Hill AVS. The immunogenetics of human infectious diseases. Annual Rev Immunol 1998;16:593–617. 5. Collins HL, Kaufmann SHE. The many faces of host response to tuberculosis. Immunology 2001;103:1–9. 6. Dorman SE, Picard C, Lammas D, Heyne K, van Diesel JT, Baretto R, et al. Clinical feautures of dominant and recessive interferon gamma receptor 1 deficiencies. Lancet 2004;364:2113–21. 7. Jouanguy E, Altare F, Lamhamedi S, Revy P, Emile J-F, Newport M, et al. Interferon-g-receptor deficiency in an infant with fatal bacille Calmette–Guerin infection. NEJM 1996;335:1956–61. 8. Newport MJ, Huxley CM, Huston S, Hawrylowicz CM, Oostra BA, Williamson R, et al. A mutation in the interferon-g-receptor gene and susceptibility to mycobacterial infection. NEJM 1996;335:1941–9. 9. Dorman SE, Holland SM. Mutation in the signal transducing chain of the interferon-g receptor and susceptibility to mycobacterial infection. J Clin Invest 1998;101:2364–9. 10. Bream JH, Carringhton M, O’Toole S. Polymorphism of the human IFN-g gene noncoding region. Immunogenetics 2000;51: 50–8. 11. Pravica V, Perrey C, Stevens A, Lee JH, Hutchinson IV. A single nucleotide polymorphism in the first intron of the human IFN-g gene. Human Immunol 2000;61:863–6. 12. Pravica V, Asderakis A, Perrey C, Hajeer A, Sinnott PJ, Huychinson IV. In vitro production of IFN-g correlates with CA repeat polymorphism in the human IFN-g gene. Eur J Immunogenet 1999;26:1–3. 13. Heinemeyer T, Wingender E, Reuter I, Hermjakop H, Kel AE, Kel OV, et al. Database on transcriptional regulation: TRANSFAC, TRRD and COMPEL. Nucleic Acids Res 1998;26:362–7. 14. Lewis CM. Genetic association studies: design, analysis and interpretation. Brief Bioinform 2002;3:146–53. 15. Sasieni PD. From genotypes to genes: doubling the sample size. Biometrics 1997;53:1253–61. 16. Lalvani A, Pathan AA, Durkan H, Wilkinson KA, Whelan A, Deeks JJ, et al. Enhanced contact tracing and spatial tracking of Mycobacterium tuberculosis infection by enumeration of antigen-specific T cells. Lancet 2001;9273:2017–21. 17. Ewer K, Deeks J, Alvarez L, Bryant G, Waller S, Andersen P, et al. Comparison of T-cell- based assay with tuberculin skin test
N. Sallakcı et al.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
for diagnosis of Mycobacterium tuberculosis infection in a school tuberculosis outbreak. Lancet 2003;9364:1168–73. Doherty TM, Demisse A, Menzies D, Andersen P, Rook G, Zumla A. Effect of sample handling on analysis of cytokine response to Mycobacterium tuberculosis in clinical samples using ELISA, ELISPOT and quantitative PCR. J of Immunol Methods 2005;298:129–41. Lio D, Marino V, Serauto A, Gioia V, Scola L, Crivello A, et al. Genotype frequencies of the +874 T–A single nucleotide polymorphism in the first intron of the interferon-g gene in a sample of Scilian patients affected by tuberculosis. Eur J Immunogenet 2002;29:371–4. Rossouw M, Nel HJ, Cooke GS, van Helden PD, Hoal EG. Association between tuberculosis and polymorphic NF (kappa) B binding site in the interferon gamma gene. Lancet 2003;361: 1871–2. Lopez-Maderuelo D, Arnalich F, Serantes R, Gonzalez A, Codoceo R, Madero R, et al. Interferon-g and interleukin-10 gene polymorphisms in pulmonary tuberculosis. Am J Respir Crit Care Med 2003;167:970–5. Tso HW, Ip WK, Chong WP, Tam CM, Chiang AK, Lau YL. Association of interferon gamma and interleukin 10 genes with tuberculosis in Hong Kong Chinese. Genes Immun 2005;6:358–63. Etokebe GE, Bulat-Kardum L, Johansen MS, Knezevic J, Balen S, Matakovic-Mileusnic N, et al. Interferon-g gene (T874A and G2109A) polymorphisms are associated with microscopy-positive tuberculosis. Scand J Immunol 2006;63:136–41. Cooper AM, Dalton DK, Stewart TA, Griffin JP, Russel DG, Orme IM. Disseminated tuberculosis in interferon g gene-disrupted mice. J Exp Med 1993;178:2243. Moreira AL, Tsenova L, Murray PJ, Freeman S, Bergtold A, Chiriboga L, et al. Aerosol infection of mice with recombinant BCG secreting murine IFN-g partially reconstitutes local protective immunity. Microb Pathogens 2000;29:175. Zhang M, Lin Y, Iyer DV, Gong J, Abrams JS, Barnes PF. T cell cytokine responses in human infection with Mycobacterium tuberculosis. Infect Immun 1995;63:3231–4. Sodhi A, Gong J-H, Silva C, Barnes PF. Clinical correlates of interferon-g production in patients with tuberculosis. Clin Infect Dis 1997;25:617–20. Lio D, Scola L, Crivello A, Bonafe M, Fransechi C, Olivieri F, et al. Allele frequencies of +874 T–A single nucleotide polymorphism at the first intron of interferon-g gene in a group of Italian centenarians. Exper Gerontology 2002;37:315–9.