Donor-specific anti-HIV/SIV immunity mediated by APOBEC3 enzymes

Texas Biomedical Research Institute, San Antonio TX

Investigators

Linked publications & trials

Abstract

PROJECT SUMMARY Human APOBEC3 enzymes, particularly APOBEC3D, APOBEC3F, APOBEC3G, and stable haplotypes of APOBEC3H, can induce G>A mutations in the HIV-1 genome. Often, these mutations, particularly those inflicted by APOBEC3G, which is a potent anti-HIV enzyme, generate stop codons and lead to HIV-1 inactivation. Studies have shown that APOBEC3 enzymes can also induce sub-lethal levels of mutations that lead to HIV-1 diversification and drug resistance. Mutations inflicted by different APOBEC3 enzymes and/or different variants of the same APOBEC3 enzyme vary substantially in terms of both their extent and sequence context. For example, APOBEC3H haplotype II often induces many G>A mutations. By contrast, APOBEC3H haplotype I, which is mostly expressed in European and Asian populations, induces little/no changes in the HIV-1 genome. Our analysis of all reported HIV-1 sequences from ~37,000 patients show that the extent and pattern of viral hypermutation is highly patient-specific. Additionally, our data indicate that hypermutation patterns are different between human and nonhuman primate models such as rhesus macaque. We hypothesize that variations in both APOBEC3 and HIV-1 genes are responsible for the observed differential hypermutation patterns. Our preliminary data suggest these variations create a complex cascade of patient-specific interactions between HIV- 1 and APOBEC3 enzymes. In support of this hypothesis, we have identified a patient-specific interaction between APOBEC3H and one of the other APOBEC3 enzymes. This interaction is induced by the HIV-1 protease processing of APOBEC3H Haplotype II splice variant SV200. Additionally, we have generated data that point to a significant defect in APOBEC3G mRNA splicing in non-human primates used frequently as HIV-1 models. Furthermore, our data indicate that variations in multiple HIV-1 proteins, not only Vif, contribute to patient-specific hypermutation patterns. We propose to combine computational and experimental techniques to determine the link between these viral and host variations and differential hypermutation profiles.