Non-synonymous Substitutions in HIV-1 GAG Are Frequent in Epitopes Outside the Functionally Conserved Regions and Associated With Subtype Differences

Abstract

In 2019, 38 million people lived with HIV-1 infection resulting in 690,000 deaths. Over 50% of this infection and its associated deaths occurred in Sub-Saharan Africa. The West African region is a known hotspot of the HIV-1 epidemic. There is a need to develop an HIV-1 vaccine if the HIV epidemic would be effectively controlled. Few protective cytotoxic T Lymphocytes (CTL) epitopes within the HIV-1 GAG (HIV_gagconsv) have been previously identified to be functionally conserved among the HIV-1 M group. These epitopes are currently the focus of universal HIV-1 T cell-based vaccine studies. However, these epitopes’ phenotypic and genetic properties have not been observed in natural settings for HIV-1 strains circulating in the West African region. This information is critical as the usefulness of universal HIV-1 vaccines in the West African region depends on these epitopes’ occurrence in strains circulating in the area. This study describes non-synonymous substitutions within and without HIV_gagconsv genes isolated from 10 infected Nigerians at the early stages of HIV-1 infection. Furthermore, we analyzed these substitutions longitudinally in five infected individuals from the early stages of infection till after seroconversion. We identified three non-synonymous substitutions within HIV_gagconsv genes isolated from early HIV infected individuals. Fourteen and nineteen mutations outside the HIV_gagconsv were observed before and after seroconversion, respectively, while we found four mutations within the HIV_gagconsv. These substitutions include previously mapped CTL epitope immune escape mutants. CTL immune pressure likely leaves different footprints on HIV-1 GAG epitopes within and outside the HIV_gagconsv. This information is crucial for universal HIV-1 vaccine designs for use in the West African region.

Accelerating Bayesian inference of dependency between mixed-type biological traits

“…Another pair of mutations that potentially shows a similar interaction is H28X and M30X, which have a positive and negative partial correlation with VL, respectively. These mutations have indeed been observed to co-occur in gag epitopes from longitudinally followed-up patients [36]. Fig 1B keeps all the other compensatory mutation pairs in Fig 1A such as A146X-I147X and A163X-S165X that find confirmation in experimental studies [37,38].…”

Hamiltonian zigzag accelerates large-scale inference for conditional dependencies between complex biological traits

“…Another pair of mutations that potentially shows a similar interaction is H28X and M30X, which have a positive and negative partial correlation with VL, respectively. These mutations have indeed been observed to co-occur in gag epitopes from longitudinally followed-up patients (Olusola et al, 2020). Figure 2b keeps all the other compensatory mutation pairs in Figure 2a such as A146X-I147X and A163X-S165X that find confirmation in experimental studies (Troyer et al, 2009;Crawford et al, 2007).…”