Phenotypic Characterization of Replication-Impaired Lenacapavir-Resistant HIV Clinical Isolates

Abstract

Lenacapavir (LEN) is a potent, first-in-class long-acting HIV-1 capsid (CA) inhibitor, approved for treatment of people with multi-drug resistant HIV in combination with other antiretrovirals. In clinical studies with LEN, the CA resistance-associated mutations (RAMs) M66I, Q67H/K/N, K70H/N/R/S, N74D/H/K, A105S/T, and T107A/C/N/S were observed. Phenotypic analyses of these CA mutants were possible in single cycle (SC) PhenoSense Gag-Pro assays; however, most CA mutants exhibited severely impaired replication capacity (RC), rendering phenotyping with multicycle (MC) MT-2 cytopathic assays unsuccessful for many. Here, we developed and optimized a novel MC phenotyping assay using a Rev-dependent HIV reporter cell line, Rev-CEM-Luc/GFP (RevLun assay), to further characterize these replication-impaired LEN-resistant HIV CA mutants. HIV Gag-Protease fragments from patients' plasma samples with CA RAMs and associated site-directed mutants were cloned into the pXXLAI HIV molecular clone, and replicative viral supernatants were evaluated in MT-2 and RevLun MC assays, with readouts of cytopathic effect and reporter gene expression, respectively. Viruses were also evaluated in the SC assay. We successfully phenotyped CA mutants in RevLun that were noninfectious in MT-2 assay, including clinical isolates containing M66I in various genetic contexts and combinations of LEN RAMs. LEN susceptibility in the RevLun MC assay aligned with data in the SC PhenoSense Gag-Pro assay and MC MT-2 assay when available. All viruses with LEN RAMs remained sensitive to other HIV drug classes. Using a sensitive HIV-dependent reporter system enhanced our ability to assess the phenotypes of viruses with low RC, allowing for further investigation into LEN resistance and CA RAMs.

Keywords: HIV‐1; capsid; capsid resistance associated mutations; lenacapavir; phenotyping.

Figure 1

Phenotyping assays. (A) Phenotypic susceptibility of viruses to LEN and other compounds was assessed using the single cycle (SC) PhenoSense Gag‐Pro assay (Monogram Biosciences). GAG‐PR sequences were cloned in an env‐deleted pNL4‐3‐luciferase based resistance test vector (RTV), cotransfected with a plasmid producing the envelope proteins from the amphotropic murine leukemia virus (A‐MLV) in HEK293T cells to generate pseudotyped virus stocks and assayed for drug susceptibility in 96‐well plates. Susceptibility of the viruses generated to lenacapavir and other compounds was determined in two different 5‐day multicycle (MC) antiviral assay formats: (B) MT‐2 and (C) RevLun cell lines, with luciferase‐based readouts of cell viability or viral replication, respectively. Promega CellTiter‐Glo measures adenosine triphosphate to estimate the number of live cells, while Promega ONE‐Glo detects firefly luciferase reporter gene expression in mammalian cells. Therefore, luciferase activity in MT‐2 cells serves as an indicator of cell viability, while luciferase activity in RevLun cells indicates HIV replication.

Figure 2

Characterization of RevLun cells. (A) Evaluation of drug inhibition in MT‐2 assay and RevLun assay was carried out by luciferase assays. Signal/noise (S/N) range (maximum/minimum cell survival) was 4–7 for the MT‐2 assay for WT virus, which relies on viral cytopathicity for signal detection. Many CA mutants could not generate the sustained level of infection required to produce virally induced cytopathic effect in the MT‐2 MC assay (S/N ~ 1). Relative luminescence units between uninfected cells, cells infected with WT virus or CA SDM virus, and infected cells that were fully protected with the integrase strand‐transfer inhibitor (INSTI), dolutegravir (DTG) were compared across the two assays after 5 days. (B) The signal/noise range (maximum/minimum infectivity) was higher with the RevLun assay, at 50–120, and the assay can detect low levels of viral replication independent of cytopathic effects for CA mutants. Low background luminescence was observed from uninfected RevLun cells, comparable to infected but fully protected cells.

Figure 3

Correlation of EC₅₀ between RevLun and MT‐2 cells. (A) Phenotypic susceptibility of WT XXLAI virus to representative drugs belonging to the major classes of HIV‐1 inhibitors, including capsid inhibitor (CAI), integrase strand transfer inhibitor (INSTI), nucleoside or non‐nucleoside reverse transcriptase inhibitor (NRTI and NNRTI), and protease inhibitor (PI), were tested in MC MT‐2 and RevLun assays and analysed by linear regression. Comparisons of EC₅₀s obtained in MC MT‐2 assay versus MC RevLun assay are shown along with the corresponding correlation coefficient (r ²). RevLun, N = 24; MT‐2 cells, N = 30. (B) Time‐of‐addition experiments in RevLun cells to evaluate the inhibition of WT XXLAI virus by CAI (lenacapavir [LEN]), INSTI (dolutegravir [DTG]), NNRTI (efavirenz [EFV]), and PI (darunavir [DRV]) administered at different time points, with luciferase readouts on day 5. The drugs are denoted by different colors as described in the legend on the figure. AZT, azidothymidine; DRV, darunavir; DTG, dolutegravir; EC₅₀, half‐maximal effective concentration; EFV, efavirenz; LEN, lenacapavir; TAF, tenofovir alafenamide.

Figure 4

Correlation between assays for baseline and CA mutants. Comparison of lenacapavir (LEN) susceptibility of capsid mutants across multiple assays. Changes in susceptibility to LEN obtained in different assays for the mutants tested were plotted and analysed by linear regression. Comparison of (A) single‐cycle (SC) Gag‐Pro assay versus multi‐cycle (MC) MT‐2 assay, (B) MC RevLun assay versus MC MT‐2 assay, and (C) SC Gag‐Pro assay versus MC RevLun assay are shown along with the corresponding correlation coefficient (r²). Graphs include baseline, post‐baseline and SDM data from Table 1. Data points greater than maximal LEN concentration tested (Gag‐Pro assay: > 2797; MT‐2 and RevLun assays: > 1000 have been plotted at 2797 and 1000, respectively for purposes of visualization. N represents the number of viruses with paired correlation data for each assay. Total viruses with data available for Gag‐Pro assay = 35, MT‐2 assay = 20, and RevLun assay = 35. FC, fold change; LEN, lenacapavir; MC, multicycle; SC, single cycle.