Hypothesis

The aged bone marrow niche does not merely permit passive decline of hematopoietic stem cells (HSCs); it actively measures HSC metabolic fitness and eliminates those falling below a threshold, akin to synaptic pruning of inefficient neurons. This niche‑driven metabolic pruning is mediated by stromal CCL5 and endothelial NF‑κB signaling that senses low mitochondrial oxidative phosphorylation (OXPHOS) and elevated ROS in HSCs, triggering their removal or myeloid skewing. Consequently, erythropoiesis suffers as a side‑effect of losing OXPHOS‑competent, lymphoid‑biased HSCs, while myeloid‑biased, glycolytic HSCs persist because they are tolerated by the inflammatory niche. Enhancing HSC mitochondrial fitness should therefore resist niche‑eviction, restore balanced output, and improve erythropoiesis without the bone‑loss side effects of high‑dose EPO.

Novel Mechanistic Insight

1. Metabolic checkpoint in the niche – Aged mesenchymal stromal cells upregulate CCL5 in response to HIF‑1α stabilization caused by niche hypoxia and adipocyte accumulation. CCL5 acts as a “danger” signal that binds CCR5 on HSCs with low OXPHOS, activating NF‑κB in both stroma and hematopoietic cells.
2. ROS‑dependent feedback – HSCs with compromised mitochondria produce excess ROS, which oxidizes cysteine residues on the CCL5‑CCR5 complex, prolonging signaling and reinforcing NF‑κB activation. This creates a feed‑forward loop that selectively amplifies pruning of metabolically weak HSCs.
3. Parallel to neuronal pruning – Just as microglia complement tagging marks weakly active synapses for removal, the aged niche uses CCL5/CCR5 as a tag for HSCs failing a metabolic fitness test, linking inflammatory remodeling to a quality‑control program.

Testable Predictions

• Prediction 1: In aged mice, HSCs with low MitoTracker Red signal (indicating low membrane potential) will show higher CCR5 phosphorylation and greater proximity to CCL5‑expressing stromal cells compared to high‑MitoTracker HSCs (p<0.01, immunofluorescence co‑localization).
• Prediction 2: Chronic treatment of aged mice with the NAD⁺ booster nicotinamide riboside (NR) or the mitophagy inducer urolithin A will increase HSC OXPHOS, reduce CCR5 signaling, and shift the HSC transcriptome toward lymphoid/erythroid programs, decreasing myeloid bias by ≥30% (flow cytometry for CD150⁺CD48⁻Lin⁻Sca1⁺cKit⁺ subsets).
• Prediction 3: Genetic deletion of Ccr5 in HSCs will blunt niche‑driven myeloid skewing without affecting stromal CCL5 production, preserving bone volume (µCT trabecular thickness unchanged) while improving EPO‑stimulated hemoglobin rise (+1.5 g/dL vs. wild‑type aged controls).
• Prediction 4: Combining low‑dose EPO (5 U/kg) with NR supplementation will achieve hemoglobin levels comparable to high‑dose EPO alone (≈14 g/dL) but prevent the trabecular bone loss seen with high‑dose EPO (BV/TV loss <5% vs. 20% in high‑dose EPO group).

Falsifiability

If any of the above predictions fail—e.g., NR does not alter HSC mitochondrial fitness or myeloid bias, CCR5 loss does not rescue erythropoiesis, or low‑dose EPO + NR still causes bone loss—the core claim that the aged niche actively prunes HSCs based on metabolic fitness would be refuted. Conversely, confirmation would support the notion that therapeutic strategies targeting HSC metabolism, rather than blunt EPO elevation, can ameliorate geriatric anemia by working with, not against, the niche’s intrinsic quality‑control mechanism.