Single-Cell Atlas Exposes How Aging Breaks Down the Entire Endocrine System

The endocrine system — the network of glands that produce hormones governing metabolism, stress, reproduction, growth, and sleep — undergoes profound changes with age. Yet until now, no study had systematically characterized those changes at single-cell resolution across multiple endocrine organs simultaneously. This study fills that gap, providing the first large-scale single-cell transcriptomic atlas of endocrine system aging in a mammalian model.

The team dissected eight endocrine organs from young (6-month-old) and aged (24-month-old) male and female mice: the pituitary gland, pineal gland, thyroid gland, adrenal gland, pancreatic islets, hypothalamus, ovary, and testis. Using single-cell RNA sequencing (scRNA-seq) for most organs and single-nucleus RNA-seq for the hypothalamus, they processed a total of 208,304 high-quality cells. After stringent quality control, 169,859 cells from six sex-shared organs and 38,445 cells from the two sex-specific gonads were analyzed. They identified 48 distinct cell types across the six shared organs, including 16 hormone-secreting cell subtypes.

Aging signatures were pervasive. The senescence marker Cdkn2a (p16) was upregulated across all six sex-shared endocrine organs. Senescence-associated secretory phenotype (SASP) gene scores and the SenMayo gene set were also significantly elevated (p ≤ 0.0001) in most organs. A total of 5,261 differentially expressed genes (DEGs) were identified between aged and young mice. The thyroid gland showed the greatest number of aging-associated DEGs at the organ level. At the cell-type level, immune cells — especially CD8+ T cells — and functional hormone-secreting cells accumulated the largest numbers of DEGs, reflecting that both immunity and secretory function are heavily disrupted by aging.

A central mechanistic discovery was the GZMK–MHC-I–UPR regulatory axis. Aging-associated exhausted CD8+ T cells, marked by high expression of Gzmk and Pdcd1, expanded significantly across multiple endocrine organs. GZMK, a serine protease secreted by these exhausted T cells, upregulated MHC-I expression on the surface of functional endocrine cells. This MHC-I elevation in turn triggered the unfolded protein response (UPR), a form of cellular stress linked to dysfunction and senescence. Critically, treating cells with small-molecule inhibitors of GZMK receptors — including nafamostat and alpha-1 antitrypsin — counteracted both UPR activation and cellular senescence in vitro, supporting a causal rather than merely correlational role for immune infiltration in endocrine aging.

Machine learning applied to the aging dataset identified CD59, a complement regulatory protein, as a novel surface marker that distinguishes aged functional endocrine cells from young ones across sex-shared organs. This offers a potential biomarker for endocrine aging assessments. For the sex-specific gonads, aging drove distinct pathology: in ovaries, theca cells showed dramatically elevated reactive oxygen species (ROS) signaling, while in testes, spermatogonia exhibited impaired DNA damage repair pathways. Both gonads also showed heightened immune infiltration with aging, consistent with the broader systemic pattern.

The study's implications are substantial. The finding that exhausted T cells mechanistically drive stress and senescence in hormone-secreting cells suggests that immunomodulatory strategies — including GZMK pathway inhibition — could be therapeutic targets for preserving endocrine function with age. The CD59 biomarker finding opens a new avenue for measuring endocrine aging non-invasively. For clinicians, the data provide a molecular framework explaining why hormonal output declines so broadly across glands simultaneously with aging, suggesting a shared immune-driven mechanism rather than purely gland-intrinsic deterioration.