Hypothesis

Age‑dependent enlargement of hydroxyapatite (HA) crystallites sterically hinders the binding of fully carboxylated osteocalcin (cOC) to the bone matrix, increasing the pool of undercarboxylated osteocalcin (ucOC) that leaks into circulation. This circulating ucOC then acts as an endocrine signal that exacerbates bone resorption via RANKL upregulation in osteoblasts, creating a feed‑forward loop that accelerates brittleness independent of vitamin K status.

Mechanistic Basis

1. Crystallite geometry and OC binding sites – cOC binds HA through its Gla domains, which require specific nanometer‑scale surface features for optimal coordination 1. As HA crystallites grow along the c‑axis (from ~8.9 nm in newborn to ~18.1 nm in adult vertebral bone) the surface‑to‑volume ratio drops, reducing the density of high‑affinity nucleation sites for cOC. Cryo‑EM shows VKGC carboxylates OC residues processively, making OC highly sensitive to vitamin K fluctuations 3. When binding sites are scarce, even fully carboxylated OC remains unbound, shifting the equilibrium toward ucOC.
2. ucOC as a maladaptive signal – Unbound ucOC retains its ability to interact with GPRC6A on osteoblasts and osteoclast precursors. In aging bone, elevated serum ucOC has been linked to frailty and increased resorption markers 4. We propose that ucOC‑GPRC6A signaling stimulates osteoblast expression of RANKL, thereby increasing osteoclastogenesis and net bone loss.
3. Decoupling from vitamin K supplementation – Clinical trials show vitamin K reduces circulating ucOC by ~60 % but fails to improve bone density or fracture risk 2. This disconnect is explained if the primary driver of bone loss is not ucOC levels per se, but the matrix‑bound OC deficit caused by enlarged HA crystals. Supplementing vitamin K cannot restore binding sites lost to crystallite growth.

Testable Predictions

• Prediction 1: In vitro, synthetic HA particles with crystallite sizes mimicking neonatal bone (<10 nm) will bind significantly more cOC (measured by ELISA) than particles sized to mimic aged bone (>15 nm) at equal vitamin K concentrations.
• Prediction 2: Mice fed a high‑phosphate diet to accelerate HA crystal growth will exhibit lower bone‑cOC content, higher serum ucOC, and increased trabecular resorption (TRAP staining) despite normal vitamin K levels. Rescue with a crystal‑growth inhibitor (e.g., pyrophosphate) should restore cOC binding and reduce ucOC without altering vitamin K status.
• Prediction 3: Human vertebral biopsies from older donors will show an inverse correlation between HA crystallite width (via synchrotron XRD) and osteocalcin immunoreactivity within the matrix, while serum ucOC positively correlates with crystallite width.
• Prediction 4: Administration of a GPRC6A antagonist in aged mice with large HA crystals will blunt the ucOC‑driven increase in RANKL expression and mitigate bone loss, even when crystallite size remains large.

Falsification would occur if increasing crystallite size does not reduce matrix‑bound cOC, or if modulating ucOC signaling fails to affect bone resorption despite changes in crystal dimensions. This hypothesis shifts focus from vitamin K alone to the physical properties of the mineral matrix as a gatekeeper of osteocalcin’s dual structural and endocrine roles.