Associations between genetic variation in the T-cell signalling pathways and CD4+ T-cell count recovery after ART initiation in southern African cohort

Abstract

Background: Infection by the Human Immunodeficiency Virus (HIV) can be treated by means of anti-retroviral treatment (ART), but a substantial percentage of treated individuals experienced failure to reconstitute their CD4+ T-cells to normal levels (immunological failure). The aims of this study were to determine associations between genetic variation in six candidate genes and CD4+ T-cell recovery following two years of ART in a southern African cohort. of the six candidate genes, two play roles in T cell co-activation (CD28 and ICOS), three play roles in T cell inhibition (PDCD1, PD-L1 and CTLA4) and IL7-Rα affects T cell homeostasis. We also analysed association between the same genetic variants and four plasma markers: sCD27, a marker of T cell activation; sPD-1, a marker of chronic T cell activation and/or exhaustion; sCD163, a marker of macrophage type 2 activation; lastly sCD127, a soluble form of IL7-Rα. Methods: DNA from the WRHI001 cohort was analysed with permission from the Wits HREC (Human Research Ethics Committee). We identified the demographic and clinical factors influencing CD4+ T-cell counts and these were used as covariates during multivariate analysis of associations between single nucleotide polymorphisms (SNPs) and CD4+ T-cell count. A total of 28 SNPs in CD28, CTLA4, ICOS, PDCD1, PDL1 and IL7-Rα were successfully genotyped in 262 samples using Sequenom MassARRAY. Levels of sCD27, sCD163, sPD-1 and sIL7-Rα were measured by ELISA or Luminex techniques in 76 individuals from plasma samples at two years post-ART initiation. Demographic and clinical factors influencing plasma marker levels were identified and used as covariates during multivariate analysis of associations between SNPs and plasma markers, Plink v1.07 software was used to analyse all genetic associations. Results: We found that 9% of the total cohort experienced immunological failure defined as failure to reconstitute CD4+ T-cell count >200 cells/mm3, two years after ART initiation We found that male gender, younger age, use of tenofovir, higher weight and increased CD4+ T-cell count at baseline were significantly associated with higher CD4+ T-cell count after two years of ART. Of the 28 SNPs examined three were monomorphic and one SNP, rs11568821, had a minor allele frequency (MAF) = 0.006. Of the remaining 24 SNPs, eight were significantly different in frequency from LWK (Luhya in Webuye, Kenya) frequency. Analysis of linkage disequilibrium (LD) patterns of genes on chromosome 2 (CD28, CTLA4, ICOS and PDCD1) showed strong linkages (most D’=1) within genes but not across genes on chromosome 2 despite the short distances between them. SNPs in IL7-Rα were all in strong linkage. Weak linkage was found between PDL1 SNPs. Significant associations were observed between genetic variants in the T cell co-stimulation genes ICOS and CD28, and CD4+ T-cell recovery after ART initiation. However different SNPs in ICOS were implicated in the univariate vs. multivariate analyses. During univariate analysis, ICOS rs6761201 showed significant associations with CD4+ T-cell counts in allelic, recessive, genotypic and haplotypic models, whereas in multivariate analyses, ICOS rs10183087, rs1559931 and rs4404254 showed significant associations with CD4+ T-cell counts in dominant and haplotypic models. The SNP rs3181098 in CD28 that was significantly associated with CD4+ T-cell recovery in univariate analysis was not significant during multivariate analysis. One SNP rs10815225, in PD-L1, was associated with CD4+ T-cell counts in univariate analysis only. One SNP in the homeostasis IL7-Rα gene, rs11567705, was associated with CD4+ T-cell counts in multivariate analysis only. Two of the four soluble plasma markers tested had significant associations with CD4+ T-cell counts: significantly decreased sPD-1, and higher sCD163 were detected in individuals with poor CD4+ T-cell recovery. Associations were examined between variation in the six candidate genes and the four soluble markers. The same genetic variants rs6761201 in ICOS (rs6761201) and rs3181098 in CD28 that were significantly associated with CD4+ T-cell recovery in the larger cohort (n=262), were associated with sCD27 levels, a marker of T cell activation, in the cohort subset (n=69). However we could not demonstrate significantly higher T cell activation levels in those with lower CD4+ T-cell counts in the cohort subset. In both univariate and multivariate analyses, genetic variation in the IL7-Rα gene (rs6897932) was significantly associated with sIL7Rα levels, confirming previous literature in this regard. Conclusion: Genetic variation in the ICOS, CD28, PD-L1 and IL7-Rα genes influence CD4+ Tcell counts recovery after ART in a southern African cohort. The role of these SNPs in T cell activation and CD4+ T-cell recovery during ART use should be examined using further immunological methods. SNP rs6897932 in the IL7-Rα gene was confirmed to influence sIL7-Rα plasma levels. Our observation of higher sCD163 in those with poor CD4+ T-cell recovery suggest possibly higher innate (especially macrophage type 2) activation in those with poor CD4+ T-cell recovery. The genetic variants studied here do not directly relate to innate activation and other SNPs should be studied to examine this hypothesis.