CXCL10 is overexpressed in active tuberculosis patients compared to M. tuberculosis-exposed household contacts

Abstract

Introduction: Variability in clinical outcome of tuberculosis infection is dependent, among other factors, on variation in host immunological response to the infection, which is modulated, in part by genetic variations present in the host. We undertook a study to identify host factors associated with such clinical variability.

Study design and methods: A comparative study between groups of active TB patients vs. clinically normal household contacts, family members living under the same roof with the patients for a long period of time, was carried out. We screened 22 candidate cytokines and chemokines in the plasma of 119 pairs ("discovery set") of TB patients and their asymptomatic household contacts. Identified associations were validated in an independent cohort of 78 patient-household contact pairs ("validation set"). Validated associations were further cross-validated by gene expression assays using RT-PCR and in-vitro whole blood stimulation by mycobacterial antigens ESAT6 and Rv2031c, two well-characterized antigens that are expressed in active and latent phases of disease, respectively. In a concomitant SNP association study, we have sequenced the validated gene in these patients and household contacts.

Result: CXCL10 was found to be the most significantly (p = 0.0002) elevated chemokine - discovered and validated -- in patients' plasma compared to their household contacts. We found that CXCL10 was overexpressed by 5-fold at the RNA level in patients compared to asymptomatic household contacts (p = 0.004). On stimulation of whole blood collected from normal healthy volunteers with mycobacterial antigens ESAT6 and Rv2031c, we found that production of CXCL10 by ESAT6 was significantly higher (p = 2.8 × 10^-12) than Rv2031c. The production of CXCL10 was 20-fold more than IFN-γ, the most widely validated cytokine, by ESAT6 stimulation (p = 4.6 × 10^-8). One of the polymorphisms in promoter of CXCL10, rs4508917 (-1447 A > G), was identified as a proteinQTL (pQTL). Reduced expression of CXCL10 was observed among individuals with GG genotype, but the reduction was statistically significant only among controls, but not among patients. Among patients, the expression level was very high compared to the controls irrespective of the genotypes at this locus.

Conclusion: Plasma level of CXCL10 is predictive of the active phase of TB infection.

Keywords: CXCL10; DNA polymorphisms; Immune responses; Mycobacterial antigens; Tuberculosis.