CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

Results: Phylogenetic trees showed that the transmitter’s sequence was usually more proximal, i.e. closer to the root than the recipient’s sequence (in 42/49, 76/83, and 70/82 TP in gag , pol , nef , respectively). Sequences from recipients were significantly more divergent from the consensus and MRCA than sequences from transmitters (Wilcoxon matched-pairs signed rank test, all p-values < 0.0001). When based on AA sequences, the divergence from either Con.C, Con.ZM, or the MRCA were larger in the sequences from the recipients than in those from the transmitters. The difference was significant in Pol (p < 0.0001) and Nef (p ≤ 0.036), but not in Gag (p ≤ 0.874). Similar to the distances constrained on the trees, measures based on pairwise distances showed that sequences from the recipients were more distant from the consensus sequences or the MRCA both at the nucleotide (all p-values < 0.0001) and AA level (0.044 ≤ p ≤ 0.136 for Gag; p < 0.0001 for Pol and Nef). Next, we identified sites that varied between transmitter and recipient and evaluated whether the residue in the recipient was the consensus. About half of the sites in the recipient had the consensus (Table 1), hence not suggesting a bias toward the consensus at transmission. Conclusions: Our phylogenetic analyses showed that sequences from the recipients were further away from the consensus than sequences from the transmitters, conforming to HIV’s diversification since the beginning of the epidemic. Our results suggest that the evidence for a selection bias for increased fitness upon transmission may not be generalizable when HIV sequences as a whole, rather than independent AA, are considered. Given that setpoints viral loads have remained stable over the epidemic, we conclude that a bias toward the consensus at transmission has a limited impact.

231 Mode of Sexual Transmission Can Exert an Impact on Shaping the HIV-1 Founder Virus Damien C. Tully 1 ; Colin B. Ogilvie 1 ; Rebecca E. Batorsky 1 ; Eric S. Rosenberg 2 ; Kenneth H. Mayer 3 ; Heiko Jessen 4 ; Sergei Kosakovsky Pond 5 ; Marcus Altfeld 6 ; Jonathan Carlson 7 ;Todd M. Allen 1 1 Ragon Inst of MGH, MIT, and Harvard, Cambridge, MA, USA; 2 Massachusetts General Hosp, Boston, MA, USA; 3 The Fenway Inst, Fenway Hlth, Boston, MA, USA; 4 HIV Clinic Praxis Jessen, Berlin, Germany; 5 Univ of California San Diego, San Diego, CA, USA; 6 Heinrich Pette Inst, Hamburg, Germany; 7 Microsoft Rsr, Redmond, WA, USA Background: While the global spread of HIV-1 has been fueled by sexual transmission the genetic determinants underlying the transmission bottleneck remains poorly understood. We sought to investigate whether the mode of sexual transmission has an influence on the multiplicity of infection and further examined whether other intrinsic differences exist between heterosexual (HSX) and men who have sex with men (MSM) founder viruses. Methods: Samples from 74 acute HIV-1 infected subjects were subjected to whole genome deep sequencing. Logistic regression coupled with cross validation was used to assess the performance of a newly adapted genetic-distance based measurement to relate quantitative ( env diversity) viral predicators to qualitative (single/multiple infection) outcomes. To distinguish the pattern of selection between MSM and HSX founder viruses a comparative codon-based phylogenetic framework was implemented and the transmission indices calculated. Lastly, a phylogenetic corrected approach was used to search for signature sites associated with MSM or HSX transmission. Results: Implementation of our genetic distance based approach to deep sequence data revealed that 83% of MSM infections in our cohort are established by a single founder virus. Combination of these data with previously published studies extended this analysis to a total of 354 subjects and further supported that there are no significant differences between the mode of transmission and the rate of multivariant infections. However comparison of envelope sequences revealed that HSX founder viruses exhibited a greater number of codon sites under positive selection, as well as higher transmission indices. Moreover, specific genetic signatures within MSM and HSX founder viruses were identified, with single polymorphisms within gp41 enriched among HSX viruses while MSM viruses appear to select for variants surrounding the CD4 binding site. Conclusions: These results illustrate how consequential the mode of sexual transmission is in shaping the genetic composition of the HIV-1 founder virus. While our findings suggest that the number of viral variants establishing infection is not influenced by the mode of transmission there are marked differences during the transmission bottleneck that influences the selection and genetic make-up of the founder genome. These data provide new insights into the HIV-1 bottleneck process that may in turn facilitate the design of a more effective HIV-1 vaccine or other therapeutic and prevention strategies. 232 Characterizing the Multiplicity of HIV Founder Variants During Sexual Transmission Antoine Chaillon 1 ; Sara Gianella 1 ; Susan J. Little 1 ; Christy Anderson 1 ; Gemma Caballero 1 ; Francis Barin 2 ; Sergei Kosakovsky Pond 1 ; Douglas D. Richman 1 ; Davey M. Smith 1 ; Sanjay R. Mehta 1 1 Univ of California San Diego, San Diego, CA, USA; 2 François-Rabelais Univ, Tours, France Background: Infections with multiple founder viral strains have been associated with faster HIV disease progression. Most previous studies have estimated the number of founder variants based on the analysis of viral populations sequenced from acutely HIV-infected individuals. We hypothesized that including sequences collected from source partners would improve resolution and characterization of transmission events. Methods: Blood plasma samples were collected from both the source and recipient in 31 epidemiologically and phylogenetically linked transmission pairs from the San Diego Primary Infection Resource Consortium (SD PIRC). All were men who have sex with men (MSM), sampled on average 64 days (range: 11-170) after the recipient’s estimated date of infection. Deep sequencing (454 FLX, Roche) of HIV-1 env (C2-V3) was performed for all samples. A combination of highlighter plots, tree topologies and sequence diversity were used to determine the number of founder viruses with and without the inclusion of source partner data. Each phylogenetically distinct clade with limited diversity (<2.5%) in the recipient viral population was deemed to arise from a single founder variant. Results: Using sequence data from the recipient partner, we were able to estimate the number of founders in 26/31 (84%) cases. Seven of these had high diversity (>2.5%) at baseline, consistent with the existence of multiple founder variants. Undetermined results were produced for 5 cases (16%) due to inconsistency among the applied methods to recipient data. By incorporating sequence data generated from the source partner, we were able to derive the number of founder variants for all 31 transmission pairs. Overall, 16/31 cases (51.6%) involved transmission of multiple founders and we found a moderate agreement between the two approaches with and without source data (73% congruent, k = 0.46, 95%IC=0.12-0.8). To further evaluate the transmission bottleneck, we focused on the 15 single founder transmissions, and identified four recipients (27%) who had

Poster Abstracts

90

CROI 2016