Hypothesis

Aged hematopoietic stem cells accumulate mitochondrial iron, which sensitizes them to ferroptosis and shifts their transcriptional program toward myeloid lineage at the expense of erythroid differentiation. This iron‑driven ferroptotic bias remodels the bone marrow niche by releasing damage‑associated molecular patterns that activate mesenchymal stromal cells to overproduce M‑CSF, thereby amplifying osteoclastogenesis. Elevated M‑CSF not only erodes the osteoblastic niche but also feeds back to suppress erythropoietin (EPO) signaling in residual erythroid progenitors, creating a self‑reinforcing loop that explains age‑related EPO hyporesponsiveness despite normal circulating hormone levels.

Mechanistic Basis

1. Mitochondrial iron overload – Aging HSCs show increased expression of iron‑import genes (e.g., Tfrc) and reduced ferroportin, leading to labile Fe²⁺ accumulation that catalyzes lipid peroxidation via the Fenton reaction [https://pmc.ncbi.nlm.nih.gov/articles/PMC3292028/].
2. Ferroptosis priming – The lipid‑peroxidation burden activates ACSL4 and suppresses GPX4, pushing HSCs toward a ferroptosis‑prone state. Ferroptotic DAMPs (e.g., HMGB1, oxidized phospholipids) are released, which have been shown to stimulate MSCs to secrete M‑CSF [https://pmc.ncbi.nlm.nih.gov/articles/PMC12720310/].
3. Myeloid skew – Ferroptotic stress triggers NF‑κB and HIF‑1α pathways that upregulate myeloid transcription factors (PU.1, C/EBPα) while repressing GATA‑1 and KLF1, erythroid master regulators [https://journals.plos.org/plosbiology/article?id=10.1371%2Fjournal.pbio.0050215].
4. Niche disruption – MSC‑derived M‑CSF expands osteoclast precursors, accelerating bone resorption and degrading the osteoblastic niche that normally supports erythropoiesis via SCF and CXCL12 secretion [https://www.frontiersin.org/journals/hematology/articles/10.3389/frhem.2025.1525132/full].
5. Feedback on EPO signaling – Osteoclast‑derived cathepsin K and reactive oxygen species can oxidize the EPO receptor (EPOR) on erythroid progenitors, decreasing its surface expression and JAK2‑STAT5 activation, thereby blunting the erythropoietic response to physiologic or therapeutic EPO.

Testable Predictions

• Aged HSCs isolated from old mice will show higher mitochondrial labile iron, increased lipid ROS, and reduced GPX4 protein compared with young HSCs.
• Pharmacologic inhibition of ferroptosis (e.g., with liproxstatin‑1 or ferrostatin‑1) in aged HSCs will restore erythroid colony‑forming unit (CFU‑E) output in vitro without altering myeloid CFU‑GM numbers.
• Blocking M‑CSF signaling (using anti‑M‑CSF antibody or CSF1R inhibitor) in aged mice undergoing EPO therapy will preserve osteoblast numbers, increase bone marrow SCF/CXCL12 levels, and improve hemoglobin rise relative to EPO alone.
• Inducing osteoclast activation in young mice (via RANKL injection) will reproduce the EPO‑hyporesponsive phenotype, which can be rescued by ferroptosis inhibition in HSCs.

Experimental Approach

• Use longitudinal bone marrow iron imaging (Perl's Prussian blue + mitochondrial MitoTracker) in young (3 mo) vs aged (24 mo) mice.
• Measure lipid peroxidation (C11‑BODIPY) and GPX4 expression by flow cytometry in Lin⁻Sca1⁺c‑Kit⁺ (LSK) cells.
• Perform competitive transplantation of LSK cells treated with ferroptosis inhibitors into irradiated recipients, assess lineage contribution after 4 weeks.
• Administer EPO (5 U/g) ± CSF1R inhibitor (e.g., PLX3397) to aged mice for 2 weeks, monitor hemoglobin, reticulocyte count, bone histomorphometry, and serum M‑CSF.
• Validate EPOR oxidation by biotin‑switch assay and phospho‑STAT5 levels in CFU‑E cultures.

If ferroptosis‑driven myeloid bias and niche‑derived M‑CSF together underlie age‑related EPO hyporesponsiveness, targeting either arm should uncouple the vicious cycle and restore effective erythropoiesis in the elderly.