Aging bone shows larger, more elongated hydroxyapatite (HA) crystals and a shift toward undercarboxylated osteocalcin (ucOC). We hypothesize that rising B‑type carbonate substitution in HA directly diminishes the binding affinity of carboxylated osteocalcin (Gla‑OC) for the crystal surface, turning Gla‑OC from a mineralization inhibitor into a ineffective bystander. This loss of Gla‑OC restraint allows unchecked HA growth along the c‑axis, while ucOC, which lacks the glutamate residues needed for HA interaction, fails to sequester phosphate and further fuels crystal elongation. The combined effect accelerates brittleness and fracture risk.

Mechanistic chain

1. Carbonate‑mediated surface alteration – B‑type carbonate replaces phosphate groups in the HA lattice, increasing the a‑lattice constant shrinkage and surface charge heterogeneity 1. This modifies the electrostatic landscape that Gla‑OC recognizes via its γ‑carboxyglutamate residues.
2. Gla‑OC binding decline – Surface plasmon resonance assays using synthetic HA peptides with varying B‑type carbonate content predict a >50 % drop in Gla‑OC association rate when carbonate exceeds 2 mol % (extrapolated from tooth data). Reduced Gla‑OC occupancy removes its steric hindrance on crystal faces, permitting lateral and longitudinal growth.
3. ucOC amplification – ucOC cannot bind HA regardless of carbonate state. Its accumulation, driven by vitamin K deficiency and pentose phosphate pathway suppression in senescent osteoblasts 4, leaves more crystal surface unprotected, exacerbating the growth promoted by the weakened Gla‑OC effect.
4. Feed‑forward to vascular calcification – The same HA surface changes that reduce Gla‑OC binding also increase affinity for circulating calcification inhibitors (e.g., fetuin‑A) being displaced, promoting ectopic mineral deposition in arteries—a link supported by observed correlations between bone ucOC and arterial calcium scores.

Testable predictions

• In human trabecular and cortical bone samples from donors aged 20‑90 yr, B‑type carbonate content (measured by FTIR ν₂ phosphate split) will inversely correlate with Gla‑OC binding affinity (determined by solid‑state NMR or SPR on powdered HA) and directly correlate with ucOC levels (ELISA).
• Ex vivo organ cultures of mouse calvariae treated with sodium bicarbonate to raise B‑type carbonate will show decreased Gla‑OC incorporation into the matrix (immunohistochemistry) and increased HA crystal length (TEM) unless supplemented with menaquinone‑4 (vitamin K2), which should rescue Gla‑OC carboxylation and normalize crystal size.
• In vivo, aged mice fed a vitamin K2‑enriched diet will exhibit lower B‑type carbonate, higher Gla‑OC/ucOC ratio, and improved bone mechanical properties (three‑point bending) compared with controls, despite similar osteoblast numbers.
• Serum ucOC adjusted for BMD (U‑osc/BMD ratio) will predict arterial calcification severity independent of traditional risk factors, reflecting the shared HA surface pathology.

Falsification If B‑type carbonate variation does not alter Gla‑OC binding kinetics, or if vitamin K2 supplementation fails to modify crystal morphology without changing ucOC levels, the hypothesis would be refuted. Likewise, a lack of correlation between the U‑osc/BMD ratio and vascular calcification would undermine the proposed mechanistic link to ectopic mineralization.

This framework integrates crystal chemistry, osteocalcin biology, and osteoblast metabolism to explain why bone becomes both more mineralized and more fragile with age, and it points to vitamin K2 as a potential intervention that targets the root cause rather than downstream symptoms.