The Role of Collagen Cross-linking in Craniofacial and Long Bone Mineralization and Healing

Abstract

Compromised collagen and mineral can lead to decreases in bone quantity and quality in a variety of diseases that differentially affect craniofacial and long bones. Since bone is a composite - with mineral lending stiffness and collagen lending toughness - maintaining balance of both constituents is critical to the physical integrity and sustained function of bone. Experimental models of perturbed collagen cross-links utilize beta-aminoproprionitrile (BAPN), a dose-dependent inhibitor of the lysyl oxidase enzyme that catalyzes formation of the enzymatic collagen cross-links – pyridinolines and pyrroles (mature) as well as aldimines (immature). The ratios of immature to mature cross-links correlate with bone strength and toughness in long bones. However, a full cross-link profile of craniofacial bone and distinction between craniofacial and long bones have not been investigated. Likewise, site-specific compositional and mechanical consequences of collagen perturbation are lacking. The central hypothesis of this thesis is that collagen cross-linking inhibition compromises bone quality in a bone-specific manner (long bone or craniofacial bone) in normal growth (aim 1) and in osseous wound healing (aim 2), and that a full panel of bone cross-links can be detected in serum (aim 3). During normal growth in the mouse femur and mandible, bone composition (mineral and collagen) showed an anatomical dependence including ~30% higher carbonation in the femur and ~67% higher mature/immature collagen ratio in the mandible. A highly sensitive liquid chromatography/mass spectrometry method was developed which allowed for discovery of the advanced glycation end-product, carboxymethyl-lysine (CML) in mouse bones for the first time. Accumulation of CML was bone dependent, with ~22X higher levels in the mandible than the femur. BAPN caused a significant increase in carbonation (4.4%) in the femur, but no change in the mandible and no change in tissue level mechanics. BAPN significantly decreased mandibular mature cross-links (~22-38% for subsets of mature cross-links) but not in the femur. This altered response to BAPN highlights the need to further understand how altered collagen differentially affects composition in craniofacial and long bones. To study the effects of BAPN on bone healing and mechanical properties, sub-critical osseous defects were created in the femur and maxilla of mice subjected to collagen perturbation via BAPN. BAPN decreased tissue level mechanics (stiffness, hardness, and Young’s modulus, viscoelastic properties) in newly formed bone differently between the maxilla and femur. Yet, there was no change in bone volume with BAPN, only differences in healing between the femur (~60% increase in bone volume between 7-14 days) and maxilla (no significant change). BAPN initially decreased the number of osteoclasts in the femoral defect osteoclasts (7 days), then increased (14 days), with no change in the maxilla. This site-specific response of mechanics and osteoclasts to local changes in collagen matrix may make lysyl oxidase mediated collagen cross-linking a potential therapeutic target for controlling site specific osteoclast response. Additionally, serum detectible cross-links were measured as a potential minimally invasive assay to assess direct bone cross-links. Described here is the first report of immature cross-links measured in both mouse and human serum, creating the most comprehensive bone collagen cross-link analysis in serum to date. Overall, this thesis work has advanced the understanding the role of collagen cross-linking in mineralization during normal growth and healing, and the differential effects in craniofacial and long bone. Also established, was a new potential assay for measuring bone cross-links in serum.