Integrase polymorphism and HIV-1 group O diversity

HIV-1 is subdivided into three groups: M (major, HIV-M), O (outlier, HIV-O) and N (non-M/non-O). HIV-O is mainly found in Cameroon where it represents 1% of the HIV infections. The high genetic divergence between groups M and O viruses and the large intragroup O diversity [1,2] lead to major consequences in diagnosis and follow-up of the HIV-O infected patients [3,4]. Moreover, HIV-O show natural residues similar to mutations implicated in group-M resistance: most of HIV-O exhibit a Y181C mutation in the reverse transcriptase which makes them naturally resistant to nonnucleoside reverse transcriptase inhibitor [5]. Conversely, the systematic presence of N42D mutation in the gp41 does not seem to impact the in-vitro sensitivity to T20/enfuvirtide [6,7]. The protease shows a high level of natural polymorphism corresponding to a dozen of secondary resistance-associated mutations in group M, whose impact needs to be assessed [8,9]. The limitation of nonnucleoside reverse transcriptase inhibitor use in HIV-O infected patient makes the new integrase inhibitors (INI) a promising alternative. Raltegravir (MK-0518, Isentress, Merck, Whitehouse Station, New Jersey, USA) and elvitegravir (GS-9137, Gilead, Foster City, California, USA) represent this new class of antiretroviral agents, whose safety and efficacy when associated with an optimized background regimen has been demonstrated in patients infected by HIV-M or HIV-2 [10–12]. But mutations have already been described as associated with INI in-vitro or in-vivo resistance (Table 1) [13,14].Table 1: Integrase polymorphism of HIV-O observed at group-M resistance-associated positions.The aim of this study was to analyse HIV-O polymorphism and diversity of the Integrase region, and to determine whether group O viruses present naturally occurring resistance-associated mutations as observed for some other molecules. Seven HIV-O isolates and 89 clinical samples of treatment-naïve or experimented but INI-naive HIV-O patients living in Cameroon (33), Gabon (one) and France (55) were analysed. Nested reverse transcriptase (RT)-PCR was developed to amplify a 603 bp fragment encoding Integrase amino 19 to 219, a region which encompasses the catalytic core domain (amino acids 51 to 212). RNA extracted from plasma or diluted supernatant was used for a RT-PCR with the CTCCAYCCAGACAARTGGAC (U) and CCYTCACCTTTCCACAGGAG (L) primers followed by a nested PCR with ATAGAYCAGGCACARGAAGATC (U) and GGGTCTCTGCTRTCTCTGTAATA (L) primers. PCR products were sequenced using the nested PCR primers as previously described [15]. Sequences were aligned for phylogenetic analyses using MEGA, version 4 [16]. The phylogenetic tree obtained with the 96 sequences and 21 HIV-O sequences from the Los Alamos National Laboratory (LANL) database (http://www.hiv.lanl.gov) presents a ‘comet-like’ tree topology, different from the well symmetric double star topology classically observed in HIV-M (Fig. 1). Group O organization seems more complex than the previously described clades A to C [1] or clusters I to V [2]. Seventy-six percent of the newly characterized isolates are included in clade A/(Ia+Ib+Iu+IV+V+U), while B/II and C/III clusters represent 7.5% and 5% of the strains, respectively. Two sub-clusters, a1 and a3, are becoming significant, representing 22% and 14% of the new strains, respectively and being supported by high bootstrap values (Fig. 1a). The remaining 11.5% sequences are out of any clustering. This asymmetrical tree topology and the high phylogenetic complexity are in agreement with the hypothesis of different epidemiological mechanisms between HIV-O and HIV-M viruses [17], but possibly among HIV-O themselves, due to the presence of many divergent strains with high branch length as well as strong founder effects (Fig. 1a).Fig. 1: Phylogenetic tree of 117 HIV-1 group O integrases. (a) Phylogenetic analysis of the unambiguously aligned 603 bp sequences was performed using the Neighbour-joining method, and evolutionary distances were computed using the Kimura two-parameter method. One thousand bootstrap replicates were performed to assess the reliability of the branching order; bootstrap values are shown only when significant (>70). The 96 integrase sequences obtained here are represented in black, while the database sequences are in grey. The six patients showing the 157Q or 157EQ mutation are indicated in large font. Roques classification: clades A (including sub-clusters a1, a2 and a3), B and C. Yamaguchi's classification: clusters I (subdivided in Ia, Ib and Iu), II, III, IV and V. (b) Phylogenetic tree of 55 HIV-1 group M integrases. Phylogenetic analysis of the LANL group M reference set sequences covering the same 603 bp region was performed as described in (a).Fig. 1: (Continued).The 96 HIV-O amino acids sequences were compared to the corresponding HXB2 sequence, showing a mean p-distance (number of mutations/sequence length) of 0.17 (pairwise values ranged from 0.15 to 0.20). Thirty-three positions described in INI resistance were analysed: 14 were nonpolymorphic, but atypical residues (with unknown signification) were found at 12 positions. Natural mutations were observed at 11 positions mainly associated with in-vitro resistance, with high prevalence (>96%) for four of them (Table 1). Interpretation using ANRS and Stanford algorithms (http://www.hivfrenchresistance.org, http://hivdb.stanford.edu) indicated resistance to raltegravir for six strains, due to the natural presence of E157Q mutation or E157E/Q mixture. This glutamine substitution has been shown to impair the activity of integrase [18], but the HIV-O specific plasma viral loads [19] of these six patients ranged from 2.9 to 5.2 log copies/ml, indicating efficient replication and suggesting the existence of compensative mutations. Presence of E157Q mutation was associated neither with a particular cluster (Fig. 1a), nor with a former treatment, as it has already been shown that emergence of some INI resistance-associated mutations can be linked to reverse transcriptase mutations selected under nucleoside reverse transcriptase inhibitor treatment [18]. In conclusion, our results suggest that HIV-O present natural mutations associated with in-vitro or in-vivo resistance to INIs, which are not linked to a particular cluster or clade. Their impact needs to be evaluated by phenotypic and clinical assays. Acknowledgements We thank the Institut de Veille Sanitaire (Saint Maurice, France), Agence Nationale de recherche sur le Sida et les Hépatites Virales (Paris, France) and Rouen University Hospital for financial support. The RES-O group includes all physicians and biologists implicated in the surveillance of HIV-O infection in France.