The EMBO journalJul 6, 2026

SenFlag gene pattern identifies aging cells in mouse and human tissues through shared gene activity

Longevity & Aging Weekly Brief ↗PubMed ↗DOI ↗OA ↗

Updated Jul 7, 2026

Abstract

SenFlag identifies a rare yet progressively accumulating population of senescent cells in vivo across tissues in both mice and humans.

SenFlag is based on a streamlined gene signature derived from bulk and single-cell RNA-sequencing datasets.
It features a conserved transcriptional program marked by decreased expression of proliferation-related genes and increased expression of cell-cycle inhibitors.
The gene signature also includes lysosomal features, with higher levels of V-ATPase subunits and cathepsins.
SenFlag-positive cells are enriched in epithelial and endothelial tissues and show an increase with age and tissue injury.
These cells are reduced in datasets from senescence-targeting interventions, indicating the specificity of SenFlag in vivo.

Simplified

Identifying senescent cells via single-cell transcriptome profiling data remains challenging due to cellular heterogeneity and overlap with other cellular states. Here, we present SenFlag, a streamlined gene signature for enhanced identification of senescent cells based on integration of core gene expression features. SenFlag was derived through systematic assessment of bulk and single-cell RNA-sequencing datasets across multiple senescence models. It captures a conserved transcriptional program characterized by reduced expression of proliferation-associated genes and chromatin-associated genes (HMGB1/2, HMGN2), combined with upregulation of cell-cycle inhibitors (CDKN1A/CDKN2A) and of CCND1. Additionally, SenFlag incorporates lysosomal features, including increased expression of V-ATPase subunits and cathepsins. SenFlag identifies a rare but progressively accumulating population of senescent cells across tissues in both mice and humans in vivo, with enrichment in epithelial and endothelial compartments. SenFlag-positive cells increase with age and following tissue injury, and are reduced in datasets involving senescence-targeting interventions, supporting its specificity in vivo. Together, SenFlag provides a robust and interpretable signature for identifying senescent cells in single-cell datasets and facilitates the study of senescence across physiological and pathological contexts.