Association of interferon-γ +874A polymorphism with the risk of developing cervical cancer in north-Indian population

Authors


Dr RD Mittal, Additional Professor (Biochemistry), Department of Urology, SGPGIMS, Raebareli Road, Lucknow-226014, Uttar Pradesh, India. Email ramamittal@yahoo.com, rmittal@sgpgi.ac.in

Abstract

Objective  Interferon gamma (IFN-γ) is a pro-inflammatory cytokine playing a pivotal role in both innate and adaptive immune responses. A single nucleotide polymorphism located in the first intron of the human IFN-γ gene can influence the secretion of cytokine. Therefore, we aimed to investigate the association of IFN-γ T/A gene polymorphism with the risk of cervical cancer.

Design  Case–control study.

Setting  Uttar Pradesh State in India.

Sample  Two hundred cases with histologically proven cancer of the cervix and healthy controls (= 230), age and ethnicity matched were recruited in this study.

Methods  Genotyping was performed for bi-allelic +874 (T/A) polymorphism of IFN-γ by amplification refractory mutation system method.

Main outcome measures  Low producer IFN-γ +874 AA genotype was associated with high risk for cervical cancer, which further modulated the increased risk in tobacco users.

Results IFN-γ AA genotype which is low producer of IFN-γ was associated with increased risk of cervical cancer (OR = 2.43, = 0.003). Allele A was at 1.54-fold increased risk of cervical cancer (OR=1.54, = 0.002). The AA genotype showed statistically significant risk with high stage (III + IV) of cervical cancer (OR = 4.99, = 0.001). In tobacco users, AA genotype showed significantly increased susceptibility to cervical cancer (OR = 5.08, = 0.010).

Conclusion  Variation in IFN-γ +874 AA genotype because of ethnicity in north-Indian population may represent an important susceptibility biomarker for cervical cancer risk as well as other diseases and should be explored further.

Introduction

Cervical cancer (OMIM number #603956) is the second most common cancer among women worldwide, with a disproportionate share of the mortality associated with this disease occurring in developing countries.1 Approximately 500 000 women worldwide develop cervical cancer each year.2 In India, the age standardised incidence from different regions has been seen to be 17.2 to 55 per 100 000.3 Cervical cancer is strongly associated with infection by oncogenic types of human papilloma virus (HPV).4 Although many women are infected with high-risk types of HPV, only a subset of infected women develops cervical cancer, suggesting that other cofactors including host genetic factors must be present for the development of malignancy. Although cervical cancer is multifactorial in origin, several epidemiological studies suggest that genetic predisposition may also play a significant role.5 Tobacco carcinogens have been associated with risk of cervical malignancy.6 Deoxyribonucleic acid (DNA) adducts of bulky aromatic compounds have been found with increased frequency in cervical epithelium of smokers, compared with nonsmokers, thus providing biochemical evidence that tobacco may act as a confounder in the aetiology of cervical cancer.7 Other risk factors associated with cervical cancer include high parity, an early age at first intercourse, multiple pregnancy or sexual partners, diet, use of oral contraceptives and low socio-economic status.

Interferon-gamma (IFN-γ) (OMIM: 147570) also known as macrophage-activating factor is a dimerized soluble cytokine. The IFN-γ gene on chromosome 12q24 spans approximately 5.4 kb and contains four exons. Like other cytokines, the IFN-γ coding region is invariant, with no reported polymorphisms.8 In contrast to interferon-α and -β which can be expressed by all cells, IFN-γ is secreted by CD4+ Th1 cells, cytotoxic CD8 cells, Th0-cells and activated natural killer cells. IFN-γ is an acid-labile protein which has antiviral, immunoregulatory and anti-tumour properties.9 It alters transcription in up to 30 genes producing a variety of physiological and cellular responses. IFN-γ binding to IFNGR1 & IFNGR2 (interferon gamma receptors) activates the JAK-STAT pathway by phosphorylation of Jak and thereby activating STAT transcription factor.10 JAK-STAT signalling pathway is the best-characterised and commonly accepted IFN signalling pathway.11 In addition, IFN-γ activates antigen presenting cells APCs and promotes Th1 differentiation by upregulating the transcription factor and inhibits the development of Th2 cells.12 The polarised type 1 cellular immunity plays a critical role in the host defence against tumour development and viral infection.13 Macrophage stimulation with IFN-γ induces direct antimicrobial and anti-tumour mechanisms as well as up-regulating antigen processing and presentation pathways. Thus, high level of IFN-γ production is typically associated with effective host defence against viral infection such as HPV (Figure 1A).

Figure 1.

 (A) Showing IFN-y pathway and hypothesis of study. (B) Representative gel picture of IFN-γ +874 T/A polymoiphism. Lane 1: 50-bp DNA Ladder; Lane 2, 3: AA (low producer) genotype; Lane 4, 5: AT (medium producer) genotype; Lane 6, 7: TT (high producer) genotype.

In cervical cancer patients, decreased intra-tumour expression of IFN-γ has been reported to be associated with the poor prognosis.14 Further, individuals with low IFN-γ producing genotypes were also reported to be at higher risk of developing tuberculosis.15

A single nucleotide polymorphism in the first intron of the human IFN gene, at 5′ end adjacent to a CA repeat region is at the position +874 T/A. The DNA sequence containing the +874T allele is the specific binding site (5′-ttcttacaacacaaaatcaaatct-3′) for the NF-kB transcription factor.16 The +874T allele correlates with high-IFN production and +874A allele correlates with low-IFN production as a response to stimuli.17

Therefore, we aimed to evaluate the association of IFN-γ +874(T/A) gene polymorphism for conferring susceptibility to cervical cancer, its clinical stages and modulation of risk after interaction with tobacco habits in a north-Indian population.

Methods

Study design

Subjects

In a retrospective study from January 2005 to April 2008, 200 cases of frank carcinoma of uterine cervix (all purely squamous cell carcinoma) confirmed by cervical biopsy were consecutively taken from different Gynaecology units of Lucknow hospitals. Controls (230) were cancer-free women unrelated, age, ethnicity matched, cervical cytology negative were recruited from routine colposcopy screening clinic belonging to the same institution. The cervical screening test performed was the ‘cervical smear test’. In this method, the cells scraped onto the brush are ‘smeared’ onto a glass slide. They are then sent to the laboratory where the cells on the glass slide are looked under a microscope to see if there are any abnormal cells present. If positive in smear test, colposcopy was performed which includes taking of a small piece of tissue from the cervix (biopsy) to make a more detailed assessment of the cells.

Clinical data were obtained by questionnaire, personal interviews and review of case records. Staging of carcinoma of cervix was performed by gynaecologist using guidelines of International Federation of Gynecology and Obstetrics (FIGO).18 Environmental exposures (tobacco usage) were recorded in a pre-designed questionnaire. The study was approved by local ethics committees of institute and all participants gave an informed signed consent. After obtaining informed consent, all individuals were personally interviewed using a predetermined questionnaire. The response rate for the interview was 80%. Information on age, occupation, region of origin, tobacco habit, age at menopause, parity, socio-economic status, use of contraceptives and the number of sexual partners were recorded. Tobacco habits included smoking of cigarette, bidi (Indian cigarette made of tobacco wrapped in tendu leaf diospyros melanoxylon) or hukka (Indian pipe). Smokeless tobacco use consisted of gutka (sweet flavourful chewing tobacco) or zarda (aromatic chewing tobacco). Alcohol consumption was also recorded. A 5 ml peripheral blood sample was collected in EDTA and stored at −70°C prior to DNA extraction.

Genotyping

Genomic DNA extraction was performed from peripheral blood leucocytes pellet using the standard salting-out method.19IFN-γ polymorphism +874T/A (rs no 2430561) was genotyped in subjects using amplification refractory mutation system PCR method. Primers for amplification were taken as described earlier.14 PCR was performed in a total volume of 12 μl with genomic DNA (100 ng), 0.25 mM each deoxynucleotide triphosphates, 1X PCR buffer containing 10 mM Tris-HCl, pH 8.6, 50 mM KCl, 1.5 mM MgCl2, 0.3 μM of each allele and 0.1 μM of control primer and 0.2 units of Taq polymerase (Bangalore Genei, Bangalore, India).

Genotyping was performed on 2% agarose gel using molecular weight markers and visualised after staining with ethidium bromide (Figure 1B). All reactions were carried out in an MJ Research 200 thermal cycler (MJ Research, Watertown, MA, USA). The laboratory personnel were blinded regarding the status of samples. Five percent of samples from both patients and controls were re-genotyped by other laboratory personnel and no discrepancy in genotyping was noticed.

Statistical analysis

The age variable was expressed as mean ± standard deviation (SD). The chi-square goodness-of-fit test was used for any deviation from Hardy–Weinberg equilibrium in controls. Differences in genotype prevalence and association between case and control groups were assessed by binary logistic regression model. Association was expressed as odds ratios (ORs) as risk estimates with 95% confidence intervals (95% CIs). The ORs were adjusted for confounding factors, such as age, cigarette smoking and alcohol consumption. Logistic regression analysis was used to fit statistical models to predict the association of IFN-γ with susceptibility to cervical cancer. Gene–environment interactions were examined between IFN-γ (+874 T/A) genotypes and use of tobacco between cases and controls. All statistical analysis was performed using SPSS software version 11.5 (SPSS, Chicago, IL, USA) and tests of statistical significance were two-sided and taken as significant when P-value was <0.05. Power of study was calculated using Quanto program version 1.0 (http://hydra.usc.edu/gxe/) with input of following variables: case–control study design, significance level <0.05 (two-sided), model of inheritance = log additive (for polygenic disease), allele frequency = 0.55 for IFN-γ A allele, genetic effect (OR) = 1.5–2.0. The study achieved 80% of power.

Results

The mean age (years ± SD) of patients and healthy controls were 44.8 ± 9.3 and 46.7 ± 9.9 respectively. There were no significant differences in ages and the ages of the individuals of the control group appeared to adequately match those of the patient groups (= 0.116). The demographical details with characteristics of cervical cancer patients are shown in Table 1. In cases, 34.2% were of medium/high socio-economic status compared with 76.8% of controls. This suggests that it is the lower socio-economic group which is more affected with the disease. The high percentage is higher among middle/high socio-economic group as they came for regular screening and routine checkups. Among all the demographic factors, only consumption of tobacco (31% versus 14%) and alcohol (7.3% versus 4.8%) was observed to be significantly higher in cases as compared with controls (< 0.001).

Table 1.   Demographic and tumour characteristics of the study population
 Patients n (%)Controls n (%)χ2P-value
  1. *The sum does not add up to the total because of some missing values.
    **Student’s t test for mean age comparison between cases and controls.

All women200 (100)230 (100)0.37
Age (years)**, mean (SD)45.8 (9.3)46.7 (9.9)
Sexual partner*
Single140 (86.9)194 (94.6)0.116
Multiple21 (13.1)17 (5.4)
Pregnancy status*
1–2 child38 (22.2)53 (28.8)0.789
2–4 child51 (29.6)62 (33.6)
>4 Child83 (48.2)69 (37.6)
Status of menopause*
Menstruating104 (63.4)142 (71.0)0.124
Non-menstruating60 (36.6)58 (29.0)
Contraceptive*
Present26 (16.4)36 (18.2)0.676
Absent132 (83.6)161 (81.8)
Socio-economic status*
High19 (11.8)22 (24.2)  0.553
Medium36 (22.4)91 (52.6)
Low106 (65.8)60 (23.2)
Tobacco habits
Nonuser138 (69.0)198 (86.0)0.001
User62( 31.0)32 (14.0)
Alcohol
No141 (92.7)177 (95.2)0.001
Yes11 (7.3)9 (4.8)
Tumour stage
Stage I32 (16.0)  
Stage II46 (23.0)  
Stage III114 (57.0)  
Stage IV8 (4.0)  

IFN-γ polymorphisms in cervical cancer

The genotype distribution of IFN-γ +874 T/A polymorphism in controls was in Hardy–Weinberg equilibrium. The genotype and allele frequencies of IFN-γ +874 T/A polymorphism in healthy controls and patients with cervical cancer are shown in Table 2. The frequency of low producer +874A was higher in cases (43.0%) as compared with controls (30.4%). Individuals with IFN-γ +874 AA genotypes were at higher risk of cervical cancer (= 0.003; OR = 2.43, 95%CI = 1.34–4.42). Allele A of IFN-γ was observed to be statistically and significantly associated with the risk of cervical cancer (= 0.002, OR = 1.54, 95% CI = 1.17–2.03).

Table 2.   Association of IFN-γ with the risk of cervical cancer in relation to genotypes considered as high or low producers of +874 IFN-γ
GenotypesCases
(200)
n (%)
Controls
(230)
n (%)
P-valueOR**95.0%CI
IFN-γ
TT23 (11.5)45 (19.6)Ref.
AT91 (45.5)115 (50.0)0.1241.560.882.78
AA86 (43.0)70 (30.4)0.0032.431.344.42
AllelesCases
(= 400)*
n (%)
Controls
(= 460)
*n (%)
    
  1. n* = total number of chromosome.
    **Age, tobacco and alcohol consumption adjusted odds ratios (OR) and 95% confidence intervals (CI).

IFN-γ
T137 (34.3)205 (44.6)Ref.
A263 (65.7)255 (55.4)0.0021.541.172.03

IFN-γ A874T polymorphism and clinical stages of cancer

The stages were merged because of low number of variables in each group. Stage I and II were grouped as low stage (carcinoma that extends beyond cervix but does not extend into pelvic wall) and stage III and IV as high stage (carcinoma has extended beyond true pelvis). Frequency of low producer AA genotype was highest in stage III + IV (52.1%). Low producer AA genotype of IFN-γ was observed to be at statistically significant elevated risk of cervical cancer (OR = 4.99; = 0.001, 95% CI = 1.86–13.3) in high stage (III + IV) (Table 3).

Table 3.   Association of IFN-γ 874 (T/A) genotype with stages of cervical cancer
Genotypes IFN-γLow stage (I + II)
n (%)
High stage (III + IV)
n (%)
P-value, OR (95% CI)*
  1. *Age, tobacco, alcohol consumption adjusted odds ratios (OR) and 95% confidence intervals (CI).

TT15 (18.6)8 (6.7)1 (ref)
AT42 (51.8)49 (41.2)0.091,2.28 (0.87–5.95)
AA24 (29.6)62 (52.1)0.001,4.99 (1.86–13.3)

Interaction of IFN-γ +874 A/T polymorphism with the use of tobacco

Interaction of IFN-γ +874 A/T genotypes with tobacco use (smoking or smokeless tobacco) was analysed to observe influence on cervical cancer risk. Frequency of IFN-γ +874 AA genotype was higher in cervical cancer patients among tobacco users (67.7% versus 40.3%). In case of tobacco users, AA genotype showed highly significant increased risk of cervical cancer (OR = 5.08; = 0.010, 95% CI = 1.47–17.5) (Table 4). However, no significance was observed among tobacco nonusers.

Table 4.   Association of IFN-γ 874 (T/A) genotype with use of tobacco in cervical cancer patients
GenotypesControlPatientsP value, OR (95% CI)*P value, OR (95% CI)**
IFN-γTTATAATTATAA
  1. *Comparing between TT and AT; **comparing between TT and AA.
    Age, alcohol adjusted P-value, odds ratios and 95% CIs were derived.

Nonuser33 (19.6)90 (53.6)45 (26.8)19 (13.8)75 (54.3)44 (31.9)0.241, 1.47 (0.77–2.72)0.128, 1.76 (0.85–3.48)
User12 (19.4)25 (40.3)25 (40.3)4 (6.5)16 (25.8)42 (67.7)0.319, 1.93 (0.52–7.06)0.010, 5.08 (1.47–17.5)

Discussion

Cytokines are known to be important in antiviral action. Interferon-gamma (IFN-γ) is a cytokine that plays an important role physiologically in promoting innate and adaptive immune response. The absence of IFN-γ production or cellular responsiveness in humans and experimental animals significantly predisposes the host to microbial infection, a result that validates the physiological importance of this cytokine in preventing infectious disease.20

The transcription factor NF-kB binds preferentially to the +874T allele.21 This preferential binding suggests that genetically determined variability in IFN-γ expression might be important for the inflammatory response activated through the NF-kB pathway. This transcription factor induces IFN-γ expression, and the T allele correlates with a high expression of IFN-γ also shown in vitro.16

The most salient feature of our work was the prevalence of high frequency of A allele, associated with low production of IFN-γ which was higher in cases (18.0%) as compared with controls (9.4%). This high prevalence of variant allele in our population could be because of ethnicity variation.22 The present study thus elucidated that individuals with IFN-γ +874 AA genotypes were at significantly increased risk of cervical cancer (OR = 2.43; = 0.003). A recent study by Linsengen et al. in a population of Brazilian women also indicated the occurrence of AA genotype to be higher in cervical intraepithelial neoplasia cases than in controls.23IFN-γ +874A allele has been previously reported to be associated with infectious disease such as tuberculosis, hepatitis B virus infection, and parvovirus infection revealing its potential role of function in host defence against microbial infections.24–26 It is possible that low IFN-γ production may impair their anti-viral response against tumour progression, rendering these individuals more susceptible to this virus infection. Similar results were observed of IFN-γ gene association with susceptibility to various other diseases.27,28

Association of the genotypes with disease stages was also carried out to explore the influence of variant genotypes on disease phenotype. Patients in stage III + IV of cervical cancer demonstrated IFN-γ +874 AA genotype association to be statistically significant at very elevated risk (= 0.001, OR = 4.49). However, in a study conducted in South African women, showed that, the AA genotype did not influence the development of invasive cervical cancer.29

We also analysed interaction of IFN-γ genotypes with tobacco exposure to investigate the modulation of risk. Epidemiological studies have also shown tobacco exposure as increased risk for the susceptibility to cervical cancer.29 Women who smoke are about twice as likely as nonsmokers to get cervical cancer. Tobacco by-products have been found in the cervical mucus of women who smoke. These carcinogenic substances damage the DNA of cervix cells and may contribute to the development of cervical cancer. Strong association of cigarette smoking has been observed with HPV expression and cervical neoplasia.30 We observed the low producers of IFN-γ, that is, AA genotype among tobacco users were significantly associated with elevated risk of cervical cancer (OR = 5.08, = 0.010) when compared with the controls. However, no significant results were observed among tobacco nonusers. This further complement the hypothesis that smoking may alter the cervical microenvironment and this altered microenvironment may enhance susceptibility to infectious agents or carcinogens. Compounds in smoke, such as 2-naphthylamine and 4-aminobiphenyl, can cause genotoxic events in the urothelium. In addition, the chemicals in cigarette smoke can result in carcinogenic events in bladder epithelium.31 Tobacco smoking enhances cellular proliferation and may have a synergistic effect on cervical carcinogenesis.32 These data may explain the possible role of the IFN-γ polymorphism in tobacco-induced cervical cancer. All these observations together suggest that the A allele is significantly risk factor.

Briefly, our study suggested that low producer IFN-γ +874 AA polymorphism is strongly associated with risk for cervical cancer which is further augmented with tobacco usage. However, functional studies of the polymorphism in cervical cancer, especially in biopsy samples of specific locations of tumour are required to draw definitive conclusions. The results of the study need to be reconfirmed in larger cohort from other populations to establish real association of IFN-γ genotypes with risk of cervical cancer.

Disclosure of interest

None.

Contribution to authorship

The concept of the paper was conceived by Dr Rama Devi Mittal and the work was carried by Ms Ruchika Gangwar. The manuscript draft was prepared by Ruchika Gangwar and finalised by Dr Rama Devi mittal. Saumya Pandey was involved with the collection of cervical cancer patients’ sample.

Details of ethics approval

The study was approved by local ethics committees of institute and all participants gave an informed signed consent. The reference number of the project is CST/D-3160 dated: 27.2.2008.

Funding

The study was supported in part by research grant from UP Council of Science and Technology (UPCST) India.

Acknowledgements

We are thankful to Dr B Mittal (Department of Genetics) for generously providing the samples. The study was supported by research grant from UP Council of Science and Technology (UPCST) India. RG is thankful to Council of Scientific and Industrial Research, New Delhi for JRF.

Ancillary