CROI 2024 Abstract eBook

Abstract eBook

Poster Abstracts

462

Differential Gene-Specific Selection Pressures on HIV-1 Defective Proviruses During ART Thuy Nguyen 1 , Mary-Elizabeth Zipparo 1 , Lindsey Adams 1 , Annemarie Glassey 1 , Ulisses Santamaria 2 , Catherine A Rehm 3 , Jessica Earhart 4 , Wei Shao 1 , Chuen-Yen Lau 5 , Frank Maldarelli 1 1 National Cancer Institute, Frederick, MD, USA, 2 Leidos Biomedical Research, Inc, Frederick, MD, USA, 3 National Institute of Allergy and Infectious Diseases, Bethesda, MD, USA, 4 National Institutes of Health, Bethesda, MD, USA, 5 National Cancer Institute, Bethesda, MD, USA Background: The error-prone nature of HIV-1 replication and host factors generate defective genomes. Defective proviruses persist, can be transcribed, and translated contributing to HIV-1 pathogenesis. During antiretroviral therapy (ART), most defective proviruses contain large deletions and hypermutations precluding the investigation of gene-specific selection pressures. We studied selection pressures on whole HIV-1 genomes and subgenomes of >7kb near-full length (NFL) defective proviruses prior to following ART. Methods: Blood cells were collected from 11 HIV-1 suppressed participants prior to and at multiple timepoints on ART (4-20 years). We performed single genome sequencing of >7kb proviruses and identified defective proviruses with stop codons, insertions/deletions or hypermutations. We collapsed identical sequences and compared the whole genomic and subgenomic characteristics of defective proviruses from pretherapy to ART. We predicted the Major Histocompatibility (MHC) class I binding, immunogenicity, and sensitivity to Broadly Neutralizing Antibodies (bNabs) Results: We obtained 705 NFL defective sequences. Proportions of hypermutant proviruses were not significantly different prior (median 38.38%) to following ART (26.86%). We observed significant selection for proviruses with defects on major spliced donor/packaging signal (MSD/PSI) (33.33% prior to 84.44% on ART) and gag (40% to 75%) but against proviruses with defects on env (66.67% to 21.25%) (p-values: 0.001-0.006). However, Rev-response elements (RRE) remained very conserved during ART; a median (interquartile range) of 0 (0-5.8%) proviruses carried RRE defects. Patterns of gag defects mainly involved start codon or deletion while deletion was the only defect pattern on env. HLA-associated mutations emerged or significantly shifted in env of 5/11 participants, but these mutations did not cause changes in the predicted MHC binding affinity or immunogenicity. Significant changes in predicted bNAbs sensitivity were observed for env of 2/11 participants. Conclusion: NFL defective proviruses are subjected to significant differential selection pressures on MSD/PSI, gag and env under ART. Gag defects are likely linked to deletions on MSD/PSI suggesting the selection for proviruses with impaired replication and infectivity. Selection against deletions on env is possibly linked to HLA escape mechanism but did not lead to significant changes in MHC binding or immunogenicity suggesting the possible involvement of other immune and viral factors. In-Depth Proviral Sequence Analysis Reveals Conserved Reservoir Landscape Composition Across Group M Hannah J MacLeod 1 , Elizabeth A. Ferrer 1 , Hanna R. Marks 1 , Mian Cai 1 , Qiuyan Ma 1 , Evgenia Aga 2 , Grace M. Aldrovandi 3 , Ronald J. Bosch 2 , Albine Martin 1 , Gregory M. Laird 1 1 Accelevir Diagnostics, Baltimore, MD, USA, 2 Harvard University, Cambridge, MA, USA, 3 University of California Los Angeles, Los Angeles, CA, USA Background: The biology of HIV-1 replication and epidemiology of transmission led to a diverse landscape of viral subtypes heterogeneously distributed across the globe. The majority of people with HIV-1 (PWH) are infected with non-subtype B virus, yet the landscape of persistent HIV-1 is understudied outside of subtype B. In-depth analysis of persistent HIV-1 in PWH across subtypes is essential to globalize HIV-1 cure research and clinical trials. Critically, such analysis is also foundational to the design and validation of molecular assays for latent HIV-1, such as a cross-Group M IPDA. Methods: We performed an extensive proviral sequencing campaign on samples from hundreds of PWH on ART with broad geographic and Group M coverage. We optimized and employed a new semi-automated, high throughput, near-full length single-genome PCR (NFL-PCR) workflow to generate proviral amplicons paired with improved pipelines for proviral assembly, alignment, and defect mapping. Proviral quantity and landscape composition were compared between subtypes and CRFs across Group M. Results: Our analysis revealed highly consistent proviral landscapes across Group M. In all subtypes, the majority of proviruses analyzed harbored fatal deletions, which on average spanned half the genome and most frequently

Bayesian mixture model was developed to incorporate different IS associated characteristics of CD8+ T-cell pressure. Results: The top 10% of expanded clones harbored integrations in recurrent integration genes previously identified in clinical studies, such as STAT5B, BACH2, MKL1, as well as genes observed in rare HIV-related tumors, including STAT3 and LCK. In the absence of CD8+ T cell pressure, HIV integrations were enriched in genes involved in cell cycle and cellular signaling pathways (p.adj<1e-6). In the presence of CD8+ T-cells, integrations frequently targeted genes involved in chromatin organization, mRNA processing, modification, and trafficking, as well as SUMOylation (p.adj<1e-9). The Bayesian model determined a CD8+ T-cell selection score of each mouse, thus functioning as a tool for quantifying CD8+ T-cell selective pressure, and highlighted clonality as a prominent characteristic associated with CD8+ T-cell selection. Conclusion: Our findings support a role for CD8+ T-cell pressure in selecting for particular clones of infected cells. Enrichments in gene pathways may reflect accessibility for integration, but also raises the possibility of CD8+ T-cell selection for insertional mutagenesis events that might modulate the transcription of genes involved in HIV expression. HIV Proviruses Among CMV-Reactive Cells Are Abundant, Defective, and Polyclonal After 4 Years of ART Filippo Dragoni 1 , Maura Manion 2 , Hao Zhang 3 , Angelica Camilo-Contreras 1 , Frank Maldarelli 4 , Irini Sereti 2 , Francesco R. Simonetti 1 1 The Johns Hopkins University School of Medicine, Baltimore, MD, USA, 2 National Institute of Allergy and Infectious Diseases, Bethesda, MD, USA, 3 The Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA, 4 National Cancer Institute, Frederick, MD, USA Background: Since clonal expansion plays a major role in the maintenance of HIV-infected CD4+ T cells, antigen exposure and immune reconstitution can affect reservoir dynamics. Chronic CMV infection is characterized by percolating antigen expression and progressive inflationary T-cell memory. Although HIV infected, CMV-reactive cells are often dominated by large clones, which factors lead to their selection remain unclear. We assessed the contribution of immune reconstitution, CMV-related disease, and CMV-DNAemia to HIV persistence. Methods: We studied 16 participants living with HIV and CMV, enrolled in two NIAID clinical trials. All had a CD4 nadir <50cells/µL. CMV-driven immune reconstitution inflammatory syndrome (IRIS) and end-organ disease were present in 3 and 3 participants, respectively. PBMCs from week 192 on ART were CD8-depleted and stimulated with CMV antigens. Reactive cells (positive for CD69/CD154/CD137) were plate-sorted at limiting dilution and subjected to whole genome amplification. HIV+ wells were tested for LTR, IPDA, proviral sequencing, and integration site analysis. Bulk TCRseq was performed on cells reactive to CMV and CD3/CD28. Results: CMV DNAemia was detected in 11/16 participants but decreased below LOD upon CD4 recovery. Only 4/16 participants reached CD4 counts >500 cells/ uL by week 192. CMV-reactive cells were highly variable but detectable in 15/16 participants (median 2.7%, range 0.6-18.6%). We observed a high frequency of HIV DNA among CMV-reactive cells, with a median of 4328 [2464-5654] proviruses/million cells. IPDA showed 2-fold lower total DNA (p=0.0006) and rare intact genomes, suggesting that the majority of proviruses are highly defective or solo LTR. Based on gag and env sequencing, we detected only a few identical proviruses in most individuals. Among those expanded, we found proviruses in loci linked to heterochromatin (ZNF84) and insertional mutagenesis (STAT5B). TCR sequencing revealed marked clonality of CMV reactive cells compared to the CD3/CD28 control (p<0.0001). Across all analyses, we detected no correlation with CMV DNAemia, IRIS, or CMV disease. Conclusion: We observed high infection frequency among CMV-reactive cells regardless of immune reconstitution and CMV DNAemia. The discrepancy in clonality between total CMV-responding cells and those infected suggests a preferential expansion of uninfected cells, likely selected during ART, while older infected clones may require longer time to be substantially shaped by proliferation.

461

Poster Abstracts

463

CROI 2024 118