CROI 2025 Abstract eBook

Abstract eBook

Oral Abstracts

ORAL ABSTRACTS

100

5’Leader Defects in HIV Plasma Clones Drive 80% of Persistent Viremia on Long-Term ART Julia R. Box 1 , Filippo Dragoni 1 , Angelica Camilo-Contreras 1 , Feng Yun Yue 2 , Vitaliy Matveev 2 , Eileen Scully 1 , Joyce L. Jones 1 , Mario Ostrowski 2 , Colin M. Kovacs 2 , Francesco R. Simonetti 1 1 The Johns Hopkins University School of Medicine, Baltimore, MD, USA, 2 University of Toronto, Toronto, Canada Background: Residual viremia (RV) persists despite years of effective ART due to the daily activation of latently infected cells. Some people on long-term ART develop nonsuppressible viremia (NSV), characterized by persistently detectable HIV RNA that complicates clinical care, raising concerns for virological failure and HIV transmission. Recent work from our group and others showed that proviruses with defects in the 5’Leader (5’L) can also cause NSV. However, their overall contribution to persistent viremia remains unknown. Methods: We enrolled 30 participants with intermittent or persistent plasma HIV RNA above the limit of quantification by clinical assays (20-1500 copies/ mL) despite no history of adherence issues. We isolated virus-associated RNA by plasma centrifugation and performed single genome sequencing covering the U5- gag and protease-reverse transcriptase (P6-RT) regions. Sequences were analyzed for drug-resistance mutations, clonality, and 5’L defects. To benchmark the frequency of major splicing donor (MSD) mutations pre-ART, we used 2238 5’L RNA sequences from the LANL HIV database. The impact of 5’L defects on viral infectivity was tested by site-directed mutagenesis of pNL4-3. Results: We obtained 376 U5-gag and 358 P6-RT HIV RNA sequences from 29 participants with a median of 28 sequences per participant. No drug-resistance mutations relevant for current ART were detected. On average, 90% of sequences were identical, with a median of 2 unique variants per participant. Estimates of clonality by U5-gag and P6-RT strongly correlated (Spearman r=0.77, P<0.0001). We detected HIV RNA with 5’L defects in 27/29 participants, contributing to an average of 78% (sd±34%) of defective sequences per participant. Among 65 unique 5’L variants recovered, 22 were intact, 18 had deletions of 2-98 nucleotides involving the MSD, and 25 had single mutations in three conserved positions (HXB2 G744, T745, G748) necessary for MSD-spliceosome binding. These mutations abolished viral replication in vitro and were significantly enriched in participants with NSV compared to the pre-ART sequences (Chi square, p<0.00001). Conversely, the polymorphic T742C substitution showed no significant difference in frequency (p=0.35). Conclusions: HIV RNA contributing to NSV is highly clonal and enriched in narrowly focused 5’L defects that prevent de novo invention events. These findings can inform the clinical management of NSV and support the development of assays to distinguish intact from defective HIV RNA in plasma. BACH2-Driven Tissue Resident Memory Programs Promote HIV-1 Persistence Yulong Wei 1 , Haocong Katherine Ma 1 , Michelle E. Wong 1 , Liza Konnikova 1 , Pablo Tebas 2 , Ricardo Morgenstern 2 , Emmanouil Papasavvas 3 , Luis J. Montaner 3 , Ya-Chi Ho 1 1 Yale University, New Haven, CT, USA, 2 University of Pennsylvania, Philadelphia, PA, USA, 3 The Wistar Institute, Philadelphia, PA, USA Background: Tissue resident memory T cells (TRMs) protect the gut mucosal barrier against microbial invasion by exerting rapid and localized effector functions while restraining excessive inflammation. Yet, pathogens can evade immune clearance and persist. The gut is a major HIV-1 reservoir in people living with HIV-1 (PLWH). However, given the rarity of HIV-1-infected cells (<0.1%) and the lack of selection markers, how HIV-1-infected cells survive and evade immune clearance in the gut remain unknown. Methods: To determine mechanisms of HIV-1 persistence in the gut, we coupled single-cell DOGMA-seq and TREK-seq to simultaneously capture chromatin accessibility, transcriptome, 156 surface proteins, T cell receptor, HIV-1 DNA and HIV-1 RNA in the same cells from colon biopsy samples of ten PLWH under antiretroviral therapy (ART) and five uninfected donors.

Results: We identified 49,205 CD4+ T cells, including 99 HIV-1-infected cells, and 51,539 CD8+ T cells in the gut. Gene regulatory network analysis and multiomic profiling revealed BACH2 as the leading transcription factor that shaped gut T cell programs. While IRF drove effector T cell responses and clonal proliferation, BACH2 restrained effector programs and increased cytokine receptor expression to maintain long-term survival of tissue resident memory T cells (TRMs). TRMs accounted for half (55.1%) of gut CD4+ T cells, but the majority (80.81%) of HIV-1-infected CD4+ T cells were TRMs. HIV-1-infected cells had increased BACH2 and RORC transcription factor activity, TRM (CD49a, CD69, CD103) surface protein expression, Th17 ( CCR6 ), and long-lived memory ( IL7R ) gene expression (q < 0.05, Wilcoxon rank-sum test). In vitro validation revealed HIV-1 preferential infection and persistence in CCR6+CD4+ TRMs (q < 0.05, Wilcoxon rank-sum test). Conclusions: Overall, BACH2-driven TRM program restrained effector T cell functions and maintained long-lived TRMs. The long-lived nature of CD4+ TRMs allowed HIV-1-infected cells to upregulate survival cytokine receptors, escape activation-induced cell death, and persist in the gut. Our work identified BACH2 driven TRM program as a new therapeutic target for HIV-1 eradication strategies. Longitudinal Co-Evolution of HIV-1 Reservoir Cells and Immune Effector Cells During Long-Term ART Weiwei Sun 1 , Ce Gao 1 , Toong Seng Tan 1 , Ciputra A. Hartana 1 , Rebecca Hoh 2 , Melissa Buitrago 2 , Douglas Robbins 2 , Viva Tai 2 , Michael Peluso 2 , Steven G. Deeks 2 , Xu G. Yu 1 , Mathias Lichterfeld 1 1 Ragon Institute of MGH, MIT and Harvard, Cambridge, MA, USA, 2 University of California San Francisco, San Francisco, CA, USA Background: Recent studies suggest that HIV-1 reservoir cells are subject to immune selection pressure, but the host immune responses that are able to engage and target viral reservoir cells are unknown. To infer mechanisms underlying immune selection activity against viral reservoir cells, we conducted a parallel, longitudinal phenotypic evaluation of HIV-1 reservoir cells and antiviral immune cells. Methods: Longitudinal peripheral blood samples collected at ~1 year on ART (T1), at ~10 years on ART (T2), and at ~20 years on ART (T3) of 5 long-term ART treated individuals were studied. Phenotypic and proviral sequencing (PheP-Seq) was used for the analysis of 73 markers on HIV-1 reservoir cells and CITE-seq was conducted for profiling of 177 markers on host immune cells. Results: Using PheP-seq, we analyzed a total of 1,075,240 individual memory CD4 (mCD4) T cells from all three timepoints combined (2433 represented HIV infected cells; 265 cells harbored genome-intact HIV-1 and 356 included clonally expanded reservoir cells). Cells harboring clonal intact proviruses integrated into heterochromatin regions increased from 2.5% at T1 to 25% at T3. At the earliest timepoint, reservoir cells harboring intact proviruses showed few phenotypic alterations relative to HIV-uninfected cells. After 20 years of ART, the following markers were significantly (FDR-adjusted p<0.01) upregulated on mCD4 T cells containing intact HIV-1: CD127, HLA-G, 2B4, CD161, HLA-E, KLRG1, CD28, and CD48; all of these markers can be associated with reservoir persistence and resistance to immune-mediated killing. Correspondingly, CITE-Seq demonstrated statistically significant (FDR-adjusted p<0.01) downregulation of the activating markers Siglec-7, CD161, NKG2D, NKp46, DNAM1, NKp30 on NK cells on T3 relative to T1, while the inhibitory receptors KIR2DL2/3 were upregulated. Phenotypic changes in CD8 T cells over two decades of ART were more limited and included significant upregulation of activation (CD38, CD151) and immune inhibition (CD39, CD93) markers. Conclusions: These data suggest a longitudinal evolution of viral reservoir cells with phenotypic features of resistance to immune-mediated killing, paired with a progressive increase of NK cells and, to a lesser extent, CD8 T cells with signs of immune inhibition and dysfunction. Phenotypic changes in reservoir cells and effector cells may jointly contribute to an immune microenvironment conducive to viral persistence.

Oral Abstracts

102

101

15

CROI 2025