CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

193 ANALYSIS OF NEARLY FULL-GENOME HIV-1 SEQUENCES FROM UGANDA: RESULTS FROM PANGEA_HIV Gonzalo Yebra 1 , Deogratius Ssemwanga 2 , Dan Frampton 3 , Brian Magambo 2 , Nicholas Bbosa 2 , Frederick Lyagoba 2 , Rebecca N. Nsubuga 2 , Pontiano Kaleebu 2 , Andrew Leigh Brown 1 , for the PANGEA_HIV Consortium 1 Univ of Edinburgh, Edinburgh, UK, 2 MRC Uganda Virus Rsr Inst, Entebbe, Uganda, 3 Farr Inst of Hlth Informatics Rsr, London, UK Background: Nearly full genome sequencing (NFGS) of HIV is increasingly available in Western countries, but its application in sub-Saharan Africa is scarce. The PANGEA_HIV project is producing HIV NFGS in African sites, including Uganda. The Ugandan epidemic features a long-term co-circulation of subtypes A1 and D, with the consequent emergence of many recombinants. Methods: We collected 685 contemporary (2009-2014) samples from Ugandan populations (a rural cohort in Masaka, several Lake Victoria fisherfolk communities, and female sex workers from Kampala); as well as 53 historical samples taken in 1986 from AIDS patients. They were analysed using Illumina MiSeq next-generation sequencing; 609 (565 contemporary, 44 historical) yielded sequences ≥1Kb, with half of them (320, 52%) being NFGS. To identify recombinants, we adapted the SCUEAL subtyping tool for full-genome analysis. Maximum-likelihood phylogenies were built with RAxML to detect clusters using Cluster Picker. Drug-resistance mutations (DRM) and co-receptor usage were analysed using Stanford HIVdb and Geno2pheno tools, respectively. Results: Unique recombinant forms (URF) between subtypes A1 and D accounted for 169 (29.9%) of the contemporary sequences, followed by subtypes A1 (147, 26.0%) and D (113, 20.0%). The rest belonged to other subtypes (10 C, 1 G), other URF (70, 12.4%), and recombinants showing unclassified segments (55, 9.7%), most of them due to the presence of gaps. A1/D recombinants also predominated among the historical sequences (21, 47.7%), followed by subtype A1 (10, 22.7%) and other URF (10, 22.7%). We detected 54 clusters (44 pairs, 10 triplets) among contemporary sequences, showing a clustering rate of 21.1%. Most clusters (77.8%) involved subjects from the same population only; 63.0% involved subjects from the same sampling location only. Contemporary sequences showed low DRM rates at PR+RT (3.3%) and integrase (0.5%); historical sequences showed no DRM. X4 co-receptor usage, which confers resistance to entry inhibitors, was rare (0.5%) among contemporary samples but frequent (47.4%) among historical ones, probably due to an advanced infection stage. Conclusion: HIV recombinants dominate the Ugandan epidemics. Their prevalence analysing NFGS was higher than in earlier studies of partial sequences. We found low levels of resistance, as expected in low-income settings, and a low rate of phylogenetic clustering, with most clusters being intra-population. However, almost 40% of the clusters involved geographical mobility. 194 LTR GENETIC DIVERSITY AMONG HIV-2 ARV-NAÏVE PATIENTS IN THE ANRS CO5 HIV-2 COHORT Quentin Le Hingrat 1 , Benoit Visseaux 1 , Mélanie Bertine 1 , Fideline Collin 2 , Florence Damond 1 , Olivier Schwartz 3 , Sophie Matheron 1 , Diane Descamps 1 , Charlotte Charpentier 1 , for the ANRS HIV-2 CO5 Cohort 1 Bichat Hosp, Paris, France, 2 INSERM, Bordeaux, France, 3 Inst Pasteur, Paris, France Background: Long terminal repeat regions (LTR) include binding sites of transcription factors (TF) and are essential to HIV DNA transcription. The aim of this study was to assess associations between HIV-2 LTR genetic diversity and reservoir size. Methods: All plasma samples of HIV-2 ARV-naïve patients included in the French ANRS HIV-2 Cohort collected with PBMC during the year 2015 were assessed. LTR sequencing was performed using Sanger technology. HIV-2 total DNA quantification was performed using a real-time PCR assay with a limit of quantification (LOQ) of 6 c/PCR. The LTR transcriptional activity was assessed using a luciferase assay on HEK293T cells transfected by LTR-luciferase plasmids. Results: Among the 99 samples tested, LTR sequencing was successful in 65 (66%) including 27 HIV-2 group A and 38 group B, among them 8 had plasma viral load >40 c/mL. Demographic and immuno-virological characteristics were similar between group A and group B-infected patients. HIV-2 DNA was above the LOQ in 24 patients (37%) with a median of 2.04 log10c/106PBMC and was detectable below the LOQ in 39 patients (60%). The proportion of patients with a reservoir above the LOQ was significantly higher in group A than in group B (67% and 16%, respectively; p <0.001). Genetic distances showed the highest variability in the U3 region of the LTR. Variability was significantly higher in group B than in group A sequences (p<0.001). No specific LTR mutation could be associated with the size of reservoir. However, 4 group B sequences (11%) showed a deletion in the first binding site of Sp1 TF. Interestingly, transcriptional activity of this Sp1-deleted LTR is 2-fold less than that of the group A or B references. Binding sites of the following TF: PuB2, peri-κb and NFκB were conserved among group A and group B sequences. On the contrary, the region between the PuB1 and pets TF described in group A sequences was not observed in group B sequences. Furthermore, bioinformatics analysis identified two new potential binding sites of TF: IRF in group A sequences and C/EBP in group B sequences. Conclusion: In this first large analysis on HIV-2 LTR sequences we observed a high genetic variability in the U3 region of the LTR in which is located TF binding sites. The highest genetic variability and also the lack of some TF binding sites observed in group B sequences might be an explanation to the lowest proportion of patients with a reservoir above the LOQ observed in this group. 195 PHYLOGEOGRAPHIC STUDY OF HIV-2 GROUPS A AND B EARLY EPIDEMICS IN WESTERN AFRICA Benoit Visseaux 1 , Stéphane Hué 2 , Florence Damond 3 , Fideline Collin 4 , Alexandre Storto 1 , Genevieve Chene 4 , Sophie Matheron 3 , Charlotte Charpentier 3 , Diane Descamps 1 , for the French ANRS CO5 HIV-2 Cohort 1 Bichat Hosp, Paris, France, 2 London Sch of Hygiene & Trop Med, London, UK, 3 Bichat–Claude Bernard Hosp, Paris, France, 4 INSERM, Bordeaux, France Background: The early spread of HIV-2 in Western Africa has not been explored yet for group B and subtype A2 recently described using the French ANRS CO5 HIV-2 cohort datasets. In this study we performed the first phylogeographic study describing the early spreading patterns of HIV-2 subtypes A1, A2 and group B in the human population. Methods: All publicly available HIV-2 pol sequences including both time of sampling and patient’s country of birth (n=49 and 8 sequences for groups A and B, respectively) were added to 125 (group A) and 68 (group B) sequences from the ANRS cohort, sampled between 1994 and 2014. Phylogeographic reconstructions were performed under the best fitting combination of evolutionary, demographic and molecular clock models, using BEAST 1.8. The trees were assessed with maximum likelihood trees obtained using RAxML. Because of the large number of sequences sampled in France in the dataset, the patient’s country of birth was used to model the geographical dispersion instead of the sampling country. Results: The estimated time of the most common ancestors (tMRCA) of group A was 1945 [95% HPD 1935-53]. Subtype A1, mainly present in patient born Senegal, Gambia, Guinea-Bissau and Guinea, presents an early diversification in 1946 [1936-54] with two distinct epidemics in Guinea-Bissau and Senegal. Several later transmission events from Guinea-Bissau to Senegal are also observed. Subtype A2, mainly present in patients born in Ivory Coast and Mali, spreads latter than subtype A1 (1956 [1947-63]). Subtype A2 initially spreads initially in Ivory Coast and presents two introduction events in Mali in 1963 [1957-69] and 1967 [1960-74]. Group B was originally introduced in Ivory Coast in 1962 [1953-13]. Conclusion: This phylogeographic study is the first to reconstruct the early subtype A2 and group B dispersal and allows a better understanding of HIV-2 early epidemic in West Africa. These two HIV-2 clades rose at similar time but diversified latter than subtype A1. Both A2 and B clades originated in Ivory Coast, suggesting that a local historical or socio- demographic event may have triggered the dispersal of these viral strains. An early fonder effect for subtype A1 in Senegal occurred before the Guinea-Bissau independence war, suggesting that HIV-2 group A was already circulating in these countries before the war that contributed to further dispersal of HIV-2 within and outside West Africa.

Poster and Themed Discussion Abstracts

76

CROI 2017