Bone strength is determined by bone mass, geometry, and material quality. Mechanisms determining bone mass have been well-studied but what maintains material quality remains poorly understood. We have previously provided evidence that EphrinB2 in osteocytes may mediate one pathway supporting bone material quality. Knockdown of EphrinB2 in osteocytes led to lower bone strength without any difference in bone mass or geometry. The bone strength defect originated from greater levels of collagen and mineral in cortical bone suggesting that EphrinB2 in osteocytes limits their accrual. In this thesis, I investigate the cellular and molecular downstream effects of EphrinB2-deletion in osteocytes in vitro and assess differences in the bone material quality due to absence of EphrinB2 in osteocytes in vivo and in vitro.

Since our initial data suggested that greater autophagy in osteocytes may cause the greater mineral content I sought to determine the role of EphrinB2 in autophagy. I showed that EphrinB2-Fc treatment limits autophagosome numbers and that stimulation of autophagy can increase mineralisation in osteocytes in vitro. However, further investigation showed that EphrinB2-deficiency in osteocytes leads to lower lysosome content and cis-Golgi apparatus fragmentation compared to control cells. This suggests that autophagosomes accumulate in EphrinB2-deficient osteocytes due to impaired degradation by lysosomes. Thus, I suggest that EprhinB2 deletion in osteocytes limits lysosome formation.

To address the low lysosome content in EphrinB2-deficient osteocytes, I next investigated the potential dysregulated pathways supporting lysosome formation in these cells, including the transcription factor EB (TFEB) and mTOR activity specifically mTORC1. I found that EphrinB2-deficiency in osteocytes resulted in lower Tfeb mRNA and TFEB protein levels, possibly caused by greater mTORC1 activity, an inhibitor of TFEB activity. Proteomics analysis additionally showed lower levels of bone material-degrading enzymes such as cathepsins and matrix metalloproteinases, which are direct downstream targets of TFEB. However, Tfeb knockout in osteocytes did not mimic the greater mineral levels observed when EphrinB2 was absent in osteocytes suggesting that the downregulation of TFEB is not the primary mechanism by which EphrinB2 deletion leads to mineral accrual.

Since the bones of mice with EphrinB2 deficient osteocytes exhibit lower bone strength due to a defective material, and my data showed lower levels of bone matrix-degrading enzymes, I next sought to determine the effects of EphrinB2-deficiency in osteocytes on the bone matrix. I found that, when cultured in a collagenous matrix, absence of EphrinB2 in osteocytes led to less contraction of their collagenous environment compared to control cells. This could explain additional observations I made in vivo, where bones with EphrinB2-deficiency in osteocytes exhibited thicker and less parallel-oriented collagen fibre bundles. When I assessed crystal length in cortical bone in vivo I observed no differences between mice with EphrinB2-deficient osteocytes and controls. This suggests that the origin of the lower bone strength caused by the deficiency of EphrinB2 in osteocytes may be due to a gradual formation of a defective collagenous matrix which provides more binding space for mineral crystals to nucleate and this leads to brittle bones.

Having initially shown that stimulating autophagy using rapamycin can increase mineral deposition in vitro I wanted to investigate whether the stimulation of autophagy can be used as a therapeutical tool to improve bone material quality. For this I used a mouse of osteogenesis imperfecta (OI) and tested whether the removal of toxic collagen I aggregates in osteoblasts could improve the severity of the OI phenotype using the FDA-approved autophagy-inducing drug carbamazepine (CBZ). Neither short- (3 weeks) nor long-term treatment (6 weeks) improved bone size, strength, or material quality in growing OI mice. Additionally, long-term treatment with CBZ had no effect on bone material quality, but led to lower bone strength in control mice caused by lower cortical and trabecular bone mass. These data suggest that CBZ-induced autophagy does not improve the OI phenotype but inhibits osteoblast function and bone growth in young individuals.

In summary, I showed that EphrinB2 deficient osteocytes have a lower lysosome content and fragmented cis-Golgi apparatus and produce lower levels of bone matrix-degrading proteins compared to control cells in vitro. Additionally, EphrinB2- deficiency in osteocytes led to less collagen arrangement in vitro and thicker and less-parallel oriented collagen fibre bundles in bones in vivo without any difference in mineral crystal length compared to control bone. This suggests that EprhinB2 in osteocytes is essential for lysosome formation and supports lysosomal enzyme production and thereby maintains collagen fibre integrity. This reveals a novel physiological role for osteocytes to gradually limit mineral and collagen accrual in bone tissue to maintain bone strength.