The cell sources for bone tissue engineering are required to produce bone-like materials.
There have been several studies reporting the successful culture of bone using osteoblasts (OBs),
mesenchymal stem cells (MSCs) and embryonic stem cells (ESCs). Previous work using specific
calcium staining or expression of osteoblastic genes/markers indicated that mineralised bone-like
tissue had formed, detailed structural observations are lacking. The key to understanding the
mechanical properties of bone can be related to the specific structure, organisation and chemistry
of the apatite mineral crystals and the collagen fibrils. This PhD project aimed to study and
compare the ultrastructure of bone cultured from three different sources; mouse OBs, MSCs and
ESCs using a combination of high resolution imaging and analytical TEM based techniques,
including electron energy loss spectroscopy. Protocols were developed to preserve the structure
and chemistry of bone during sample preparation for electron microscopy imaging and analysis.
The mechanisms of bone formation by each cell source, mediated by the cells themselves, were
also investigated.
A difference in the structure and organisation of the bone produced by each cell source
was revealed. OBs and MSCs produced the mineralised bone nodules with a similar chemistry
and organisation of apatite crystals and collagen fibrils. In comparison, ESC nodules failed to
produce banded collagen fibrils in the same culture condition as the OBs and ESCs (culture
medium supplement with ascorbic acid, β -glycerophosphate and dexamethasone), and therefore
lacked the association between mineral aggregates and collagen fibrils which is characteristic of
native bone. During mineralisation, the OBs sequestered calcium and phosphate inside the
membrane bound vesicles. Inside the cells, vesicles containing calcium and phosphorus appeared
to have been transported to the extracellular matrix. Mineralised globules, not bound by a
membrane, were intimately associated with the collagen fibrils, suggesting that the mineral had
escaped from the vesicles and was actively mineralising the collagen. Mitochondria have been
found to play a role in the mineralisation process, as they stored amorphous calcium phosphate
granules during mineralisation. The calcium phosphate containing mitochondria occasionally
fused with calcium-containing intracellular vesicles. This fusion may suggest a direct transport
for calcium and possibly phosphate ions from mitochondria to the vesicles. Mineralisation by the
MSCs appeared to follow a similar pathway. In comparison, the ESCs produced less calcified
mitochondria and vesicles which suggested that ESCs did not differentiate into a great number of
OB-like cells and this also may explain the differences in the structure and chemistry of bone
surrounding these cells. In addition, EELS analysis of the OB nodules detected silicon
compounds present in the OB nodule. Silicon may be present as silicic acid and unidentified Si-species which might associate with extracellular proteins, however further work is needed to
discern whether or not the silicon is a contaminant.