Single-cell capture of on-ART SIV transcription reveals TGF-β–mediated metabolic control of viral latency
Romaila Abd-El-Raouf, Jakob Harrison-Gleason, Jinhee Kim, Ching Man Wai, Kayla L. Yerlioglu, Catarina Ananias-Saez, Alec Ksiazek, Jeffrey T. Poomkudy, Mariluz Araínga, Deepanwita Bose, Claudia Cicala, James Arthos, Francois J. Villinger, Ramon Lorenzo-Redondo, Elena Martinelli
Romaila Abd-El-Raouf, Jakob Harrison-Gleason, Jinhee Kim, Ching Man Wai, Kayla L. Yerlioglu, Catarina Ananias-Saez, Alec Ksiazek, Jeffrey T. Poomkudy, Mariluz Araínga, Deepanwita Bose, Claudia Cicala, James Arthos, Francois J. Villinger, Ramon Lorenzo-Redondo, Elena Martinelli
View: Text | PDF
Research Article AIDS/HIV Immunology Metabolism

Single-cell capture of on-ART SIV transcription reveals TGF-β–mediated metabolic control of viral latency

Abstract

We previously demonstrated that blocking TGF-β with galunisertib, a safe, orally available small drug, reactivated latent SIV in vivo by shifting T cells toward a transitional effector phenotype. Here, we investigated the mechanisms underlying this effect using single-cell RNA sequencing, metabolic profiling, and high-dimensional spectral flow cytometry of samples from SIV-infected, antiretroviral therapy–treated (ART-treated) macaques before and after galunisertib. To characterize virus-transcribing, infected cells during ART, we developed a novel, sensitive SIV Transcripts Capture Assay (SCAP) that detected 127 SIV-expressing cells within lymph node single-cell transcriptome libraries. Galunisertib drove broad metabolic reprogramming in CD4+ T cells, with transcriptional upregulation of inflammatory and mitochondrial biosynthesis pathways, confirmed by Seahorse profiling. Metabolomics revealed increased energy metabolites and amino acids and enhanced metabolic flux without proliferation. SIV transcript–positive cells before galunisertib were metabolically quiescent compared with cells without detectable viral transcripts. After galunisertib, virus-expressing cells showed a dramatic metabolic activation, with upregulation of glycolysis, fatty acid metabolism, and TNF-α signaling. High-dimensional flow cytometry demonstrated effects beyond CD4+ T cells, including fewer tissue-resident memory T cells, but more inflammatory macrophages. In conclusion, SCAP represents a specific tool for characterizing rare SIV-infected cells transcribing virus during ART, and it reveals TGF-β as a key mediator of viral latency in vivo through metabolic suppression.

Authors

Romaila Abd-El-Raouf, Jakob Harrison-Gleason, Jinhee Kim, Ching Man Wai, Kayla L. Yerlioglu, Catarina Ananias-Saez, Alec Ksiazek, Jeffrey T. Poomkudy, Mariluz Araínga, Deepanwita Bose, Claudia Cicala, James Arthos, Francois J. Villinger, Ramon Lorenzo-Redondo, Elena Martinelli

×

Figure 3

Lymph node cells transcribing SIV on ART display a higher metabolic signature after galunisertib.

Options: View larger image (or click on image) Download as PowerPoint
Lymph node cells transcribing SIV on ART display a higher metabolic sign...
(A) Schematic process for the identification and filtering of vRNA-positive cells from merged lymph node whole-transcriptome library (98,000 cells) and the SCAP analysis (which yielded 706 cells). Five hundred fifty-two cells had a corresponding barcode in the whole-transcriptome library analyzed with standard Parse pipeline. Only 127 cells passed QC (features between 50 and 8,000 and total UMI between 450 and 200,000). (B and C) UMAP projection of the CD4+ T cell cluster from Figure 1 showing manual annotation of distinct cell subsets (B) and highlighted SIV transcript–positive cells (vRNA+) (C). (D and E) The frequencies of cells, CD4+ T cells (D) and vRNA+ (E), in each CD4+ T cell subset before (BC1) and after (AC1) galunisertib are shown compared by negative binomial mixed-effects models and BH FDR correction. (F) Comparison of quiescence module scores calculated on genes included in Supplemental Table 3 in SIV+ cells before (BC1) and after (AC1) galunisertib by Wilcoxon signed-rank test (*P ≤ 0.05). (G) Volcano plot of DEGs in vRNA+ T cells (AC1 vs. BC1) using MAST hurdle model (P ≤ 0.001; abs log2FC ≥ 1.5). Labeled genes have a P less than or equal to 0.001. (H) Bubble plot showing DEGs with a differential expression of abs log2FC ≥ 1.5 and P ≤ 0.001. (I and J) Enriched hallmark (I) and KEGG metabolic (J) pathways in vRNA+ CD4+ T cells based on GSEA (BC1 vs. AC1). (I) Asterisks indicate pathways reaching FDR-adjusted significance (q < 0.1).