Back

Low Transcriptional Complexity Cells Represent a Conserved, Aging-Relevant Maintenance State Overlooked by Single-Cell Transcriptomics

Bontempo, A.; Mathiyalagan, P.

2026-02-17 molecular biology
10.64898/2026.02.16.706190 bioRxiv
Show abstract

Whether mature tissues harbor transcriptionally quiet yet biologically functional cellular reservoirs remains largely unexplored. Here, using publicly available single-cell and single-nucleus transcriptomic datasets, we identify a previously unrecognized cellular state within fully differentiated human cell lineages characterized by low transcriptomic complexity (<1000 genes per cell) but preserved lineage identity and coherent functional gene expression. These "low-transcriptional" (low-T) states, often excluded by standard single-cell quality control thresholds or subsumed within major populations, are widespread across major organs including heart, brain, lung, and immune system, comprising substantial fractions of nearly all mature cell types. Despite reduced transcript abundance, low-T cells exhibit organized molecular programs distinct from high-T counterparts enriched in pathways related to cellular maintenance, metabolic resilience, survival, and aging, while lacking stress, apoptosis, senescence, or inflammation signatures. Low-T programs are conserved across mature cell lineages within organs but remain tissue-specific, revealing a hidden axis of cellular organization orthogonal to cell identity. Their abundance declines with age in brain and immune tissues, linking this state to organismal aging. Together, our findings uncover a transcriptionally quiescent yet functionally mature cellular state conserved across tissues. This previously overlooked population represents a biologically meaningful reservoir with implications for tissue maintenance, longevity, regenerative biology, and potential pharmacological aging interventions, and challenges conventional interpretations of transcriptional sparsity in single-cell genomics.

Matching journals

The top 5 journals account 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.