Metformin reverts age-associated transcriptional dysregulation in human and mouse adipose tissue
Ameya Kulkarni, Jessica Mar and Nir Barzilai
Albert Einstein College of Medicine, Bronx, New York, USA
Aging is the biggest risk-factor for metabolic, cardiovascular and neurodegenerative disorders, and is mediated by tissue-contextual changes manifesting through adaptive transcriptional programming. To understand the biology of aging and to target causal pathways, it is essential to investigate these transcriptional dynamics. With the advent of novel hypothesis that aging can be targeted with drugs like metformin, it is imperative to characterize response to metformin and prioritize candidate biomarkers for larger trials. We studied glucose-intolerant elderly participants in a placebo‐controlled crossover trial where they were treated with metformin for 6-weeks. In parallel, genetically heterogeneous mice were treated with 1000ppm metformin. Adipose tissues biopsied from human participants and mice, were subjected to bulk-RNA-Seq. Gonadal adipose from another cohort of metformin-treated C57BL/6J mice, was dissociated followed by 10x-chromium-based capture and sc-RNA-Seq for ~15000 cells. Using a multivariate model, we analyzed gene-expression, pathway-overrepresentation and predicted transcription factors dysregulated with aging and modulated by metformin. From scRNA-Seq data, we profiled single-cell dynamics of aging adipose tissue and analyzed metformin-induced response at the cellular resolution.
Metformin reverted age-associated dysregulation in extracellular-matrix remodeling, thermogenesis, mitochondrial respiration and inflammation. While each tissue had a transcriptional signature reflecting the aging process, we predicted metformin’s role in targeting transcriptional regulators like mTORC1, MYC and TGF. At the single-cell level, we profiled six cell-clusters, with the highest proportions of adipose-derived stem-cells, and macrophages. We also identified endothelial cells reflecting an adipocyte-progenitor signature. Finally, we found that aging induces transcriptional dysregulation primarily in macrophages, while metformin’s response is on B-cell-receptor signaling.
This work provides a proof-of-concept to prioritizing metformin for a large-scale study (TAME-trial) to delay age-related disorders. Furthermore, an integrated transcriptomic-metabolomic response to metformin, and the single-cell atlas of aging adipose tissue, provides better insights into the biology of aging and allows for research into precision therapeutics that can target aging.