Hypothesis

EPO‑induced osteoclast activation in aged marrow releases bone‑matrix‑derived Wnt antagonists (e.g., sclerostin) that blunt erythroid progenitor EPOR signaling, creating a self‑limiting loop that exacerbates anemia.

Mechanistic rationale

• Aging bone marrow shows increased adiposity and vascular rarefaction, shifting HSCs toward osteoclast‑prone niches (3).
• EPO binds EPOR on osteoclast precursors, triggering a rapid M‑CSF surge that expands CD115+ cells and accelerates bone resorption (5).
• Osteoclast resorption liberates sequestered Wnt inhibitors such as sclerostin (SOST) and Dkk1 from the mineralized matrix; these proteins diffuse into the marrow microenvironment.
• Sclerostin/Dkk1 antagonize Wnt/β‑catenin signaling in erythroid progenitors, which is required for EPOR‑mediated survival and proliferation (2).
• Consequently, despite high circulating EPO, erythroid blunted response ensues, particularly in aged marrow where baseline Wnt activity is already low due to osteoblast loss.

Testable predictions

1. Biomarker prediction – Serum sclerostin levels will rise within 12–24 h after a therapeutic EPO dose in aged mice (or humans) and correlate inversely with reticulocyte response.
2. Pharmacologic prediction – Co‑administration of an osteoclast inhibitor (e.g., cathepsin K inhibitor Odanacatib or a neutralising anti‑sclerostin antibody) will blunt the EPO‑induced sclerostin rise and restore erythropoiesis without altering EPO pharmacokinetics.
3. Genetic prediction – HSC‑specific EPOR knockout will abolish the EPO‑driven osteoclast surge and prevent sclerostin release, confirming that the effect is mediated through EPOR on osteoclast precursors rather than a secondary inflammatory effect.

Experimental approach (murine)

• Use 20‑month‑old C57BL/6 mice; baseline bone marrow cellularity, adiposity, and serum sclerostin measured.
• Randomise to four groups (n=8 per group): (1) saline control, (2) EPO (250 U/kg, q48h ×3), (3) EPO + cathepsin K inhibitor (50 mg/kg daily), (4) EPO + anti‑sclerostin antibody (10 mg/kg twice weekly).
• Outcomes at 24 h post‑last dose: serum sclerostin (ELISA), osteoclast number (TRAP staining), EPOR‑pSTAT5 in Ter119+ cells (flow cytometry), reticulocyte percentage (CBC), and bone formation markers (P1NP).
• Expected: EPO alone ↑ sclerostin, ↓ pSTAT5, modest reticulocyte rise; addition of osteoclast blockade normalises sclerostin, rescues pSTAT5 and yields a ≥2‑fold greater reticulocyte increase.

Falsifiability

If osteoclast inhibition or sclerostin neutralisation fails to reduce serum sclerostin or improve erythropoietic response to EPO, the proposed Wnt‑antagonist feedback loop is unlikely to be a major mechanism of EPO resistance in aged marrow.

Translational implication

Repurposing bone‑targeted agents (anti‑sclerostin, cathepsin K inhibitors) alongside EPO could break the niche‑destructive cycle, improving anemia treatment in the elderly while limiting bone loss.