Hypothesis

Aging bone exhibits a paradox: increased mineralization accompanied by reduced mechanical strength. We propose that the loss of carboxylated osteocalcin (cOC) function—driven by rising undercarboxylated osteocalcin (ucOC) and declining vitamin K status—removes a key molecular brake on hydroxyapatite (HA) crystal growth. When ucOC predominates, its low affinity for HA leaves nascent mineral platelets unchecked, allowing them to coalesce into larger, less perfect crystals. Concurrently, accumulation of carboxymethyl‑lysine (CML) advanced glycation end‑products increases collagen electronegativity, promoting mineral nucleation but also imposing strain on the collagen‑mineral interface. The combined effect is hypermineralized yet brittle bone, where excess mineral size and matrix strain outweigh any gains in mineral density.

Mechanistic Rationale

1. Osteocalcin–HA interaction – γ‑Carboxylation of osteocalcin creates negatively charged glutamate residues that enhance binding to the positively charged HA surface (PubMed 9504950). Carboxylated osteocalcin (cOC) therefore adsorbs onto HA crystal faces, sterically hindering further ion addition and limiting crystal thickness (Front Endocrinol 2021). In contrast, ucOC lacks these residues, exhibits lower HA affinity, and cannot effectively cap growing crystals.

2. CML‑driven collagen changes – CML modification lysines adds carboxymethyl groups, increasing collagen’s net negative charge (PubMed 40497659). This heightened electronegativity attracts calcium ions, accelerating mineral deposition onto collagen fibrils. However, the altered charge distribution also induces lateral strain in collagen‑HA composites, reducing nanoindentation hardness despite higher mineral content.

3. Coupled outcome – When ucOC fails to bind HA, the CML‑enhanced nucleation proceeds without the physicochemical brake normally supplied by cOC. Crystal platelets therefore grow beyond the physiological 40‑60 nm length, producing larger, more imperfect HA domains that impede crack‑deflection and increase brittleness. The increased crystal size also reduces the surface‑to‑volume ratio available for remodeling, further entrenching aged mineral.

4. Feedback via osteoclast signaling – ucOC can activate the GPRC6A receptor on osteoclasts, promoting resorption (Cell Metab 2007). Declining ucOC with age may blunt this signal, decreasing resorption‑coupled formation and preserving older, larger crystals. Vitamin K supplementation, by raising cOC, could restore both HA inhibition and osteoclast‑mediated turnover.

Testable Predictions

• Prediction 1: In aged human bone biopsies, higher ucOC/cOC ratios will correlate with larger HA crystal thickness (measured by TEM) and lower nanoindentation hardness, independent of BMD.
• Prediction 2: Vitamin K₂ supplementation in aged mice will decrease ucOC, increase cOC, reduce HA crystal size (TEM), and improve hardness and fracture resistance, without markedly altering BMD.
• Prediction 3: Osteocalcin‑knockout mice exhibiting chronic ucOC deficiency will show exaggerated CML‑induced crystal growth and brittleness, which can be rescued by transgenic expression of carboxylated osteocalcin.

Experimental Approach

1. Human cohort – Collect iliac crest biopsies from participants >60 y (n≈60). Quantify total osteocalcin, ucOC, and cOC by ELISA; measure CML via immunoblot; assess HA crystal dimensions using transmission electron microscopy; perform nanoindentation for hardness and modulus.
2. Intervention study – Randomized, double‑blind trial of aged mice (18 mo) receiving vitamin K₂ (MK‑7) or control for 12 weeks. Endpoints: serum ucOC/cOC, femoral HA crystal size (TEM), mechanical testing (three‑point bend), and CML content.
3. Genetic model – Oc‑null mice crossed with a knock‑in expressing only carboxylated osteocalcin (cOC‑only). Compare to wild‑type and Oc‑null under identical CML‑challenge (e.g., hyperglycemic diet).

Falsifiability

If ucOC/cOC ratio shows no correlation with HA crystal size or mechanical properties, or if vitamin K₂ fails to alter crystal dimensions despite shifting ucOC/cOC, the hypothesis would be refuted. Similarly, if Oc‑null mice do not exhibit exacerbated CML‑mediated brittleness, the proposed protective role of cOC would be unsupported.

Broader Implications

Confirming this mechanism would reposition vitamin K not merely as a coagulation factor but as a regulator of bone material quality through osteocalcin‑mediated crystal homeostasis. It would open therapeutic avenues targeting the ucOC/cOC axis—either via vitamin K supplementation, gamma‑carboxylase activators, or osteocalcin mimetics—to mitigate age‑related bone fragility independent of BMD gains.