P26-17. Increased binding to ALIX but no effect on viral growth for HIV-1 subtype C gag with PYKE insertion

Robert van Domselaar [1,2], Shambhu G. Aralaguppe [2], Ujjwal Neogi [2]
Affiliates: [1] Unit of Infectious Diseases and Dermatology, Department of Medicine Huddinge, Karolinska Institutet, Stockholm, Sweden. [2] Division of Clinical Microbiology, Department of Laboratory Medicine, Karolinska Institutet, Stockholm, Sweden.

HIV-1 is a major health burden with over 36 million infected people and 1 million deaths worldwide in 2015. Recently, we identified a PYKE tetrapeptide insertion in the HIV-1 type C gag-P6 domain within a subgroup of infected patients. This domain is associated with ALIX-binding, a functional property lost in HIV-1 type C. Host cell protein ALIX is known to facilitate retroviral budding. In this study, we examined whether the PYKE insertion in the gag-P6 domain restored ALIX-binding for HIV-1 type C and whether it affected virion release.

We used in silico analysis to predict gag:ALIX interactions. Cellular localization of gag and ALIX was assessed by immunofluorescence. To assess the gag:ALIX interaction in vitro, GFP-tagged ALIX was co-expressed with one of the codon-optimized gag variants (gag_wt, gag_PYKE, gag_PYQE) in cells. Lysates were subjected to GFP-pull down and immunoblotting. Finally, viral growth kinetic assays were performed to determine the fitness of viruses with or without the PYKE insertion.

In silico docking analysis showed that gag_PYKE could interact with ALIX using its positively charged amino acids and revealed a potential ubiquitination site at K485. Immunofluorescence showed predominant membrane localization for all gag variants and cytoplasmic localization for ALIX. Co-immunoprecipitation analysis showed increased binding to ALIX for gag_PYKE compared to gagwt. Abolishing the ubiquitination site by K485Q mutation (gag_PYQE) did not reverse the binding capacity compared to gag_PYKE. Using HIV-1 type C viruses expressing various patient-derived gag proteins showed no significant difference in viral fitness between viruses expressing gagPYKE compared to gagwt.

Although the PYKE insertion in the gag-domain of HIV-1 type C sequence did result in increased interaction of gag with ALIX, there was no clear significant difference in viral fitness between HIV-1 type C viruses with or without the PYKE insertion. Since PYKE insertion has also been associated with drug failure in HIV-1 patients, it will be clinically important to also evaluate drug responses against HIV-1 type C viruses with or without PYKE insertion.