HOXC10 loss drives epigenetic dysregulation of MSC hub gene network, promoting age-related senescence

Hypothesis

We propose that age‑dependent loss of HOXC10 in mesenchymal stromal cells (MSCs) triggers a cascade of chromatin accessibility changes at the promoters/enhancers of a hub gene set (TBX15, IGF1, GATA2, PITX2, SNAI1, VCAN), leading to their aberrant expression and driving a senescence‑associated secretory phenotype (SASP). This model extends the observed HOXC10 down‑regulation in aged skin fibroblasts [1] and the MSC hub‑gene dysregulation reported in aged versus young donors [2] by positioning HOXC10 as an upstream epigenetic gatekeeper whose loss converts positional memory noise into a deterministic pro‑aging transcriptional program.

Mechanistic reasoning

1. HOXC10 normally binds to AT‑rich sequences near developmental enhancers, recruiting PRC2 and maintaining a repressive chromatin state at mesenchymal lineage loci.
2. With age, reduced HOXC10 levels diminish PRC2 recruitment, increasing H3K27ac and decreasing H3K27me3 at hub‑gene regulatory regions, as detectable by ATAC‑seq and ChIP‑seq.
3. Open chromatin permits transcription factors such as TEAD (for TBX15) and SMADs (for SNAI1/VCAN) to access their sites, elevating hub‑gene transcription.
4. Elevated SNAI1 and VCAN reinforce a partial EMT/SASP state, while altered IGF1 and GATA2 signaling disrupt MSC paracrine support, creating a feed‑forward loop that further suppresses HOXC10 via inflammatory cytokines (e.g., IL‑6).
5. The resulting transcriptional shift reduces MSC multipotency and increases senescence markers (p16^INK4a, SA‑β‑gal), linking positional memory loss to functional decline.

Testable predictions

• Prediction 1: In MSCs from old donors, ATAC‑seq will show significantly increased accessibility at hub‑gene promoters compared with young donors, and this accessibility will inversely correlate with HOXC10 ChIP signal.
• Prediction 2: CRISPR‑a mediated upregulation of HOXC10 in aged MSCs will reduce hub‑gene expression, lower SASP factor secretion (IL‑6, IL-8), and decrease senescence-associated β‑galactosidase activity.
• Prediction 3: Single‑cell RNA‑seq of aged MSCs will reveal a bimodal distribution: a subpopulation with low HOXC10/high hub‑gene expression exhibiting elevated senescence scores, whereas the complementary subpopulation retains high HOXC10 and low hub‑gene expression.
• Prediction 4: Pharmacologic inhibition of EZH2 (to mimic loss of PRC2) in young MSCs will phenocopy the aged hub‑gene expression pattern and induce senescence, which can be rescued by HOXC10 overexpression.

Falsifiability

If any of the following observations hold, the hypothesis is falsified:

• No difference in chromatin accessibility at hub‑gene loci between young and old MSCs despite confirmed HOXC10 loss.
• Forced HOXC10 expression fails to modify hub‑gene transcription or senescence phenotypes in aged MSCs.
• Single‑cell analyses show no correlation between HOXC10 levels and hub‑gene expression or senescence signatures across MSC populations.
• EZH2 inhibition does not recapitulate the aged hub‑gene expression pattern.

Experimental outline (brief)

1. Obtain MSCs from young (≤30 yr) and old (>65 yr) donors; quantify HOXC10 mRNA/protein and perform ATAC‑seq, H3K27ac/H3K27me3 ChIP‑seq.
2. Use lentiviral CRISPR‑a (dCas9‑VP64) to elevate HOXC10 in old MSCs; assess hub‑gene qPCR, SASP cytokine ELISA, senescence assays.
3. Perform single‑cell multi‑ome (scATAC + scRNA) to map HOXC10 accessibility, hub‑gene expression, and senescence scores.
4. Treat young MSCs with EZH2 inhibitor (GSK126) and evaluate hub‑gene induction and senescence; rescue with HOXC10 overexpression.

This framework transforms correlative observations into a causal, mechanistically testable model of how HOXC10‑driven epigenetic drift propels MSC aging.