phenotypic and genotypic co-receptor tropism testing in HIV-1 epidemic region of Tanzania where multiple non-B subtypes co-circulate

Abstract

Introduction: HIV-1 infects host cells in a systematic multi-step entry process, initiated by HIV-1 envelope surface glycoprotein (gp120) attachment to the host CD4 receptor then to a co-receptor, either CCR5 (R5) or CXCR4 (X4), followed by membrane fusion and the release of virus particles into the host cell cytoplasm. HIV-1 strains show differential preference of attachment to co-receptor that led to their classification into R5-tropic, X4-tropic and dual R5X4-tropic ones (Moser, 1997; Wilen et al., 2012; Chen, 2020). The co-receptor usage dynamics in HIV-1-infected patients was initially thought, based on studies using subtype B infections, that the R5-tropic virus dominates during the early phase of the disease and that almost half of these patients eventually harbor X4-tropic ones in later disease stages (Ndung’u et al., 2006). However, it is becoming evident that the co-receptor usage dynamics is highly dependent on the infecting subtypes (Schuitemaker et al., 2010). Specifically, patients with subtype D infection tend to harbor X4/R5X4 tropic viruses even in the early phase of their disease (Kaleebu et al., 2007); whereas those with subtype C rarely harbor X4/R5X4 tropic viruses at any disease stage (Ndung’u et al., 2006; Huang et al., 2007; Lin et al., 2011, 2012). In another aspect, antiviral activity of the entry inhibitors are generally dependent on viral co-receptor tropism (Donzella et al., 1998; Baba et al., 1999; Lalezari et al., 2003; Dorr et al., 2005; Lobritz et al., 2010); e.g., Maraviroc and AMD3100 inhibit the entry of R5-tropic and X4-tropic viruses, respectively (Donzella et al., 1998; Dorr et al., 2005). Therefore, testing for viral tropisms is important for not only understanding viral characteristics but also useful in the clinical setting, where Maraviroc is used as a salvage therapy. Co-receptor usage of a given viral envelope lineage can be definitively determined by a phenotypic assay, which requires specific amplification of the DNA of a full-length envelope gene, preparation of pseudovirus harboring those envelope glycoproteins, and infectivity assessment using CD4+ target cells expressing either R5 or X4 (Whitcomb et al., 2007; Lin et al., 2010; Raymond et al., 2010). This process is cost ineffective, highly laborious, and time consuming; and thereby it is not suitable when large numbers of samples are being tested for co-receptor usage. Instead, several genotypic prediction algorithms have been developed to infer the co-receptor usage based on amino acid sequence information on the hypervariable region 3 (V3) sequence of the envelope gene. However, many of the widely used genotypic prediction algorithms including Geno2pheno, WebPSSM, and PhenoSeq were developed mainly from subtype B information and show relatively poor performance when applied to non-B subtypes. Although algorithms optimized for non-B subtypes including SCOTCH (Löchel et al., 2018), CoRSeqV3-C (Cashin et al., 2013), and THETA (DImeglio et al., 2020) have been attempted, the co-receptor tropism prediction of co-circulating multiple subtypes and emerging inter-subtype recombinant forms (isRFs) remains challenging (Taylor et al., 2008). Thus, it could be of value to document the phenotypic co-receptor tropisms of a diverse panel of HIV-1 variants across multiple non-B subtypes and emerging isRFs that co-circulate in a given geographical region. In this study, we analyzed phenotypic tropism and genotypic prediction of viral co-receptor tropism in a diverse panel of 93 envelope clones isolated from 52 treatment-naïve, HIV-1-infected patients from Tanzania, where multiple non-B subtypes and various isRFs are known to co-circulate. Keywords: HIV-1, envelope, entry, co-receptor tropism, non-B subtypes, inter-subtype recombinant form.