Back

A transcriptional continuum links viral burden and host-cell biology in human lymphoid tissue

Mangold, J. F.; Schrode, N.; Fortune, T.; Keane, A. M.; Shroff, S.; Petri, S.; Tweel, B.; Beaumont, K. G.; Swartz, T. H.

2026-07-11 immunology
10.64898/2026.07.07.736525 bioRxiv
Show abstract

Human immunodeficiency virus (HIV-1) persistence in lymphoid tissue remains a major barrier to cure, yet infected cells are commonly represented using discrete categories that may obscure biologically meaningful heterogeneity. Using an ex vivo human tonsil explant model and single-cell RNA sequencing of more than 42,000 T cells, we show that HIV transcription spans a structured continuum and that viral transcriptional burden functions as an organizing axis of host-cell biology. Across the continuum, increasing HIV transcription is associated with coordinated remodeling of immune, inflammatory, metabolic, and redox-associated programs. Lower transcriptional tiers were enriched for innate sensing and inflammasome-associated responses, whereas higher tiers exhibited activation of oxidative phosphorylation and redox-buffering pathways. Antiretroviral therapy preferentially depleted highly transcriptionally active populations while preserving lower and intermediate tiers, resulting in compression rather than elimination of the continuum. Together, these findings provide a quantitative framework for interpreting HIV transcriptional heterogeneity within human lymphoid tissue and suggest that persistent viral activity reflects a spectrum of host-virus states rather than a single infected-cell population. By linking viral transcriptional burden to distinct host-cell programs, this framework may inform future strategies to reduce HIV persistence and its associated inflammatory consequences.

Matching journals

The top 1 journal accounts for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.