Association of single nucleotide polymorphisms in IFNGR1 and IFNGR2 genes with bovine tuberculosis 

Bovine Tuberculosis (bTB) is a zoonotic disease of cattle caused by Mycobacterium bovis, an obligate aerobic, facultative intracellular bacillus that resides in host alveolar macrophages. It is transmitted mainly via aerosolized respiratory discharges containing infectious bacilli, with the natural site of infection being the respiratory tract. It causes extensive suffering on cows and poses health risks to the human population that consumes products of animal origin. bTB has been reported in almost every country of the world but is endemic in Africa and Indian sub-continent (Mukherjee, 2006, Perry et al., 2013, Prakash et al., 2015). A loss of 10-25% in milk production efficiency has been reported in bTB infected animals (Boland et al., 2010, Dhama et al., 2013). An estimated $3 billion annual loss to global livestock production due to bTB has been reported (Garnier et al., 2003). bTB is a factor that undermines the development of the dairy industry and international commerce (Cousins, 2001). bTB infection has an incubation period ranging from several weeks to months thus cattle may become infectious long before they exhibit clinical signs or lesions (Neill et al., 2001). Therefore the strategy of early detection and removal of bTB infected animals is mainstay of disease control. Currently, no effective vaccine exists for bTB and treatment regimen available is not cost effective and generally not recommended. The eradication of bTB by slaughtering of affected animals is also difficult because of the socio-economic condition of farmers and the social customs or religious taboos. Thus development of genetic resistance in cattle can be an ideal approach to control this zoonotic disease as genetic gain is cumulative and permanent. Host susceptibility to bTB has been reported to be heritable in cattle with significant heritability estimates of 0.27 (Bermingham et al., 2009) and 0.18 (Brotherstone et al., 2010). Higher resistance to bovine tuberculosis has been reported among Bos indicus than Bos taurus (Ameni et al., 2007). This indicates presence of genetic variation for susceptibility to bTB in cattle and that genetic selection could be used to exploit this variation in order to reduce the prevalence of bTB infection in dairy herds. Candidate gene approach can serve as a useful tool in identifying resistant superior genotypes for the production of new resistant animal population (Allen et al., 2010, Bermingham et al., 2014, Tsairidou et al., 2014).
       
The host interferon gamma receptor genes are important candidates for genetic association studies as they play major role in regulating the immune response against Mycobacterium spp. Bovine Interferon gamma receptor 1 (IFNGR1) gene localized on BTA9 encodes the ligand-binding chain while IFNGR2 gene localized on BTA1 encodes the non-ligand-binding partner of the heterodimeric interferon gamma receptor (IFNGR) found on macrophages. Cellular response to proinflammatory cytokine Interferon Gamma (IFNG) is mediated by IFNGR which activates downstream signal transduction cascades, ultimately leading to the regulation of gene expression to induce macrophages to kill intracellular pathogens. This steers the host immune response towards a Th1 bias favoring macrophage effector function (Lin and Flynn, 2010). A number of SNPs in candidate genes (TLR2, TLR6, TLR9, SP110, CARD15, CXCR3, IL8, IL23R, SLC11A1, CD14) involved in recognition of components of Mycobacterium bovis and subsequent activation of both innate and adaptive immune response have been investigated as potential strong candidates for genetic basis of resistance (Bhaladhare et al., 2018, Xue et al., 2018, Mishra et al., 2018, Wang et al., 2018, Shukla et al., 2017, Bhaladhare et al., 2016, Chauhan et al., 2016, Baqir et al., 2016, Baqir et al., 2015, Baqir et al., 2014). However, reports on significant association of IFNGR1 and IFNGR2 polymorphism with susceptibility/resistance to bTB in cattle are unavailable even though mutations in these genes have been reported to be associated with TB in humans (Stein et al., 2007, Hijjakata et al., 2012). Therefore the objectives of this study were to genotype a resource population tested for bTB infection and to investigate the potential association between bTB and four SNPs in IFNGR1 and IFNGR2 genes in cattle.

Resource population
 
For the current investigation, resource population comprised 245 cattle including Indigenous (Koshi, Sahiwal, Gir)/Non-descript and crossbred from Shri Mataji Gaushala, Barsana. All animals were maintained under similar feeding and managemental practices and had an equal opportunity of infection. Animals were screened for the presence of bTB by Single intradermal tuberculin test, wherein increase in thickness of skin after 72 h of intradermal injection of tuberculin antigen was noted to develop case (tuberculin positive) and Control (tuberculin negative) resource panel. An intradermal inoculation of 0.1 ml of tuberculin PPD antigen on neck region was carried out. The skin thickness was measured with vernier calipers before and 72 h after inoculation. Based on thickness, cattle were classified into three groups: Those showed marked swelling and skin thickness more than 4 mm (positive), skin thickness <4 mm and >2 mm (inconclusive) and no reaction >2 mm (negative). The inconclusive animals were not included in the present investigation. A case and control resource panel of 35 positive and 45 negative animals was developed.
 
Sample collection and Isolation of Genomic DNA
 
From each of case and control animals, 5 ml of blood was collected from a jugular vein in tubes containing 2.7% EDTA and stored at -20°C. DNA was isolated from whole blood using Promega Wizard® Genomic DNA Purification Kit as per recommended protocols. The DNA concentration was determined using Qubit Fluorometer. DNA quality was also assessed by 1% submarine agarose gel electrophoresis. 1 µl of genomic DNA was resolved on 1% agarose gel stained with ethidium bromide or SYBR® Safe DNA gel stain, and quantification was made by comparing the intensity of the band with the intensity of a known quantity of lambda DNA. Only thick DNA band and without smearing were chosen for further processing.
 
Genotyping of SNPs
 
Primers for the 1 SNP in IFNGR1 (rs109049057) and 2 SNPs in IFNGR2 (rs109579937 and rs110689128) genes were designed using OligoAnalyzer (Integrated DNA Technology software) software for amplification of the loci. The detail of primers and restriction enzymes are being presented in Table 1. Concerned amplicons were amplified under the optimized PCR condition. The PCR reaction was carried out in 25 µl volume which included 1 µL of each primers (forward and reverse), 1.5 µL MgCl₂, 5 µl buffer, 0.2 µL dNTPs, 0.125 µl to 0.25 µl Taq polymerase, 1 µl genomic DNA and Nuclease free water 15.05 -15.175 µl. The cycling program used for amplification having following steps; initial denaturation (94°C for 4 min), followed by 35 cycles of 30 s at 94°C, 30 s at annealing temperature (Table 1), 30 s at 72°C and final extension of 5 min at 72°C. The PCR products were resolved in 2.4% agarose gel and visualized under UV light after staining with ethidium bromide. Restriction digestion was carried out in 25 µL reaction volume which included 20 µL of PCR product, 1.5 U of restriction enzyme, 2.5 µL of 10x buffer and NFW to make volume up to 25 µL and incubated at recommended temperature as proscribed by manufacturer for 16 hours. The restriction enzyme digestion was made at the optimized conditions and the restriction digested products were resolved in 3.5% agarose gel and visualized under UV light after staining with ethidium bromide. Mass genotyping of all case-control resource population for all three SNPs was done by using PCR-Restriction Fragment Length Polymorphism (PCR-RFLP).
 

 
Statistical analysis
 
Initially in univariate logistic regression analysis, the non-genetic factors like age (two levels), sex (two levels) and breed (two levels) were fitted and found that none of these effects were significantly affecting the Single intradermal tuberculin test result. The association between various allelic variants with bTB tolerance/susceptibility was worked out by suitable statistical techniques using different procedures of SAS 9.3. The PROC LOGISTIC procedure of SAS 9.3 was used to find association of allelic and genotypic frequencies with bTB. The Odds Ratio (OR) of genotypes was calculated in affected population versus their contemporary genotypes. The PROC ALLELE procedure of the SAS 9.3 used for the estimation of polymorphism information content (PIC), Hardy Weinberg Equilibrium (HWE) and heterozygosity.

All non-genetic factors (breed, age and sex) had non-significant (p < .05) effect on the tuberculin test. The case-control population was genotyped by using PCR-RFLP for the three SNPs. All the three SNPs under investigation (rs109049057, rs109579937 and rs110689128) revealed polymorphism. The allelic frequencies and the genotypic frequencies in Case and Control populations at three SNP loci and their effect on susceptibility to infection along with ODDs Ratio (OR) have been shown in Table 2 and 3 respectively. The chi square test revealed that the population was not in HWE for all three SNP loci investigated. All three loci revealed moderate estimates of PIC and allelic diversity while medium (rs109049057) to high (rs110689128 and rs109579937) estimates were found for heterozygosity. PIC, Heterozygosity, Allelic diversity and probabilities of the population being in HWE for the three SNPs is presented in Table 4.
 

 

 

          
At SNP locus rs109049057, three genotypes were identified i.e. AA (311 bp and167 bp), AG (478 bp, 311 bp and 167 bp) and GG (478 bp) (Fig 1). The probability values showed that the genotype (P <0.01) as well as allele (P <0.01) had significant effect on occurrence of bTB.  The OR of A verse G was 0.19(0.09-0.4; 95% CI). The OR of AA verses GG and AG verse GG were >999.99 (<0.01 - >999.99); 95% CI) and >999.99 (<0.01 - >999.99); 95% CI) respectively. The odds of AA and AG verses GG were close to infinity, revealing that AA and AG genotypes seem to confer the susceptibility to bTB in comparison to GG genotype. Similarly the ODDs of A allele verses G allele
 

 
Statistical analysis
 
Initially in univariate logistic regression analysis, the non-genetic factors like age (two levels), sex (two levels) and breed (two levels) were fitted and found that none of these effects were significantly affecting the Single intradermal tuberculin test result. The association between various allelic variants with bTB tolerance/susceptibility was worked out by suitable statistical techniques using different procedures of SAS 9.3. The PROC LOGISTIC procedure of SAS 9.3 was used to find association of allelic and genotypic frequencies with bTB. The Odds Ratio (OR) of genotypes was calculated in affected population versus their contemporary genotypes. The PROC ALLELE procedure of the SAS 9.3 used for the estimation of polymorphism information content (PIC), Hardy Weinberg Equilibrium (HWE) and heterozygosity.
 
RESULTS AND DISCUSSION
All non-genetic factors (breed, age and sex) had non-significant (p < .05) effect on the tuberculin test. The case-control population was genotyped by using PCR-RFLP for the three SNPs. All the three SNPs under investigation (rs109049057, rs109579937 and rs110689128) revealed polymorphism. The allelic frequencies and the genotypic frequencies in Case and Control populations at three SNP loci and their effect on susceptibility to infection along with ODDs Ratio (OR) have been shown in Table 2 and 3 respectively. The chi square test revealed that the population was not in HWE for all three SNP loci investigated. All three loci revealed moderate estimates of PIC and allelic diversity while medium (rs109049057) to high (rs110689128 and rs109579937) estimates were found for heterozygosity. PIC, Heterozygosity, Allelic diversity and probabilities of the population being in HWE for the three SNPs is presented in Table 4.
       
At SNP locus rs109049057, three genotypes were identified i.e. AA (311 bp and167 bp), AG (478 bp, 311 bp and 167 bp) and GG (478 bp) (Fig 1). The probability values showed that the genotype (P <0.01) as well as allele (P <0.01) had significant effect on occurrence of bTB.  The OR of A verse G was 0.19(0.09-0.4; 95% CI). The OR of AA verses GG and AG verse GG were >999.99 (<0.01 - >999.99); 95% CI) and >999.99 (<0.01 - >999.99); 95% CI) respectively. The odds of AA and AG verses GG were close to infinity, revealing that AA and AG genotypes seem to confer the susceptibility to bTB in comparison to GG genotype. Similarly the ODDs of A allele verses G allele was 0.19 (0.09 – 0.4; 95% CI), showing that A allele is more related with bTB resistance in cattle as compared to G allele. This SNP lies in the exonic region of IFNGR1 and is non-synonymous, resulting to change of amino acid Isoleucine to Leucine. The amino acid substitution might play a role in susceptibility/ resistance of cattle for bovine tuberculosis. At SNP locus rs109579937, three genotypes i.e. CC (107 bp and 259 bp), TC (366 bp, 107 bp and 259 bp) and TT (366 bp) were observed (Fig 2). The probability values showed that the genotype (P = 0.05) as well as allele (P = 0.31) had no significant effect on occurrence of bTB. The ODDs ratio of C verses T was 0.72(0.38-1.34; 95% CI) .Where as ODDs ratio of CC verses TT and TC verses TT were <0.01 (<.01 - >999.99; 95% CI) and 0.76 (0.24 – 2.41; 95% CI) respectively. At SNP locus rs110689128, two genotypes i.e. TC (491 bp, 143 bp and 348 bp) and TT (491 bp) were observed (Fig 3). The probability values showed that the genotype (P =0.17) as well as allele (P =0.64) had non- significant effect on occurrence of bTB. The ODDs ratio of C verses T was 1.16(0.63-2.14; 95% CI), where as ODDs ratio of TC verses TT were 3.86 (0.43 – 34.63; 95% CI) respectively.
 

 

 
               
This is first report of association of SNP in IFNGR1 gene with the susceptibility to bTB. However, there are several reports in human that mutation in IFNGR1 and IFNGR2 causes impaired immune against mycobacterium tuberculosis (Doffinger et al., 2000, Newport et al., 1996). At IFNGR1 SNP loci -56C/T loci, CC genotype was associated with TB in African populations (Cooke et al., 2006, Stein et al., 2007). However, non significant association of SNPs in IFNGR1 gene with brucellosis and MAP infection has been previously reported (Prakash et al., 2014, Pant et al., 2011). Hijjakata et al., (2012) found an association of one intronic region SNP i.e. rs2834213 and one IFNGR2 3’UTR SNP i.e. rs1059293 with susceptibility to TB in human. Since, IFNGR1 and 2 genes have key roles in immunity against M. bovis thus more number of SNPs validations in a larger population may reveal biomarkers susceptibility of bovine tuberculosis.

Thanks are due to Director, Indian Veterinary Research Institute, Izatnagar for providing necessary funding and facilities to carry out this work.