Functional host-genetic loci associated with pediatric HIV-disease progression in Uganda and Botswana.

Abstract

Background: According to UNAIDS, HIV/AIDS remains a leading cause of morbidity and mortality in children under the age of 15. Distinct conditions of HIV-disease progression have been observed in different populations globally; Rapid Progressors (RPs), who develop AIDS within 3 years of HIV infection, and Long-Term Non-Progressors (LTNPs) show no signs of HIV-disease even after 10 years post-infection and yet all are antiretroviral therapy (ART) naive. Thirdly is a group of individuals repeatedly get exposed to HIV, but they remain uninfected. The latter is due to a form of either host acquired or natural protection against the HIV. Therefore, HIV acquisition and rate of HIV-disease progression following exposure can be attributed to a complex interaction involving environmental, viral, and human host genetic factors. Africa has more than 2 million new cases of HIV infections every year of which more than half a million are children. HIV infected children differ from their adult counterparts in a number of ways such as; their route of HIV acquisition, clinical course, and disease pathophysiology. Importantly, their time-to-HIV-disease progression can easily be calculated unlike their adult counterparts. Studies on host genetic factors underlying diseases using advanced genetic and genomic technologies promise to transform our understanding and approach to management of human health and disease. Therefore, the overall objective of this thesis was to compare the distribution of host genetic mutations between HIV-infected RP and LTNP who were recruited under a pediatric cohort in Uganda and Botswana for whom the time of HIV acquisition and HIV-disease progression were well established. The specific objectives of this thesis were: a) Specific Objective One: To enable technology acquisition and knowledge transfer through training

b) Specific Objective Two: To identify functional host genetic variants associated with pediatric HIV-disease progression in Uganda and Botswana

c) Specific Objective Three: To identify new host genetic variants in chemokine receptors & their ligands in Uganda and Botswana

d) Specific Objective Four: To define the structure of linkage disequilibrium of statistically significant loci & visualize enriched gene networks/pathways

Methods: This work was nested within the Collaborative African Genomics Network (CAfGEN) project, whose aim is to identify host genetic variations associated with HIV/AIDS and TB disease progression. CAfGEN is an H3Africa collaborative consortium comprising expertise from the University of Botswana; Makerere University; Baylor College of Medicine Children’s Clinical Centres of Excellence (COEs) in Botswana, Uganda, and Swaziland; as well as Baylor College of Medicine, Houston, Texas. This collaboration offered the requirement training and samples utilised in this thesis. To identify functional host genetic mutations and biological networks/pathways associated with pediatric HIV-disease progression, we purposively selected 314 extreme phenotypes that included RPs (173) and LTNPs (141) from a well-clinically phenotyped pediatric cohort who were born maternally infected with HIV and performed Whole Exome Sequencing (WES), in which all the protein-coding genes in each individual’s genome (exome) were sequenced.

Results: The genomics training was successfully completed by CAfGEN. Furthermore, we identified ~23,000 common (Minor allele frequency > 5%) coding variants from the study participants and these were further tested for association with pediatric HIV-disease progression. Under an allelic model, we observed a missense damaging variant rs1493954 in MICALCL with a suggestive exome-wide significant enrichment in LTNPs. This association showed a trend toward a dominant model of association for the alternate allele (P=4.25x10-5). This variant in MICALCL has not been previously associated with Long-Term HIV-disease Non-progression. MICALCL is important in ERK/MAPK signalling. This signalling pathway plays an important role of integrating external signals from the presence of mitogens into signalling events that promote cell growth and proliferation in many mammalian cell types. I further focussed my analysis on alpha (CXC), beta (CC), gamma (C), delta (CX3C) chemokine and chemokine-receptor genes and performed single locus, rare-variant gene-based association tests, and logistic regression analyses to identify enriched genetic variants. None of the nonsynonymous variants observed among the above chemokine genes interrogated attained statistical significance (adjusted P-value < 0.05). The CCR5-Δ32, a gene variant of CC-type chemokine receptor 5 (CCR5) whose heterozygous genotype confers Long-Term HIV-disease Non-progression was not observed in our study cohort. Nonsynonymous variants in CCR6 showed the strongest nominal association (enriched among RPs; Odds Ratio = 2.9; 95% CI = 1.0 - 8.1; naïve P-value = 0.03) using both gene-based burden and non-burden tests, but this did not survive correction for multiple testing. Further evaluation of the predicted effects of the CCR6 mutations on protein function suggested that three of four mutations observed were neutral and one deleterious.

Conclusions: The CAfGEN training model successfully achieved technology acquisition, knowledge transfer and developed a critical mass of well-trained, highly-skilled, continent-based African genomic scientists. Our results show that; in the HIV-disease progression, certain common host genetic variants are enriched in extreme disease phenotypes and may influence pediatric HIV-disease progression rates in Ugandan and Batswana populations. Our findings suggest that nonsynonymous mutations in host chemokines and their ligands are not statistically associated with the rate of HIV-disease progression in this Ugandan and Batswana HIV-infected pediatric populations. There was no SNP that was in very strong linkage disequilibrium around of suggestive statistically significant pediatric HIV-disease progression loci and furthermore, no biologically enriched gene networks/pathways. Finally, using an extreme phenotype HIV infected pediatrics, we were unable to replicate the top HIV-disease progression statistically suggestive genetic markers in two in both Ugandan and Batswana populations.