Bone tissue, the main material the skeleton consists of, has remarkable properties. Two macroscopically different types are distinguished. The first type is cortical or compact bone, which is a rather dense tissue. The second type is trabecular or cancellous bone. This is a porous tissue with a complex three-dimensional structure consisting of struts and plates, called trabeculae. Already in 1892 Wolff found that the orientation of these trabeculae coincides with the direction of the stress trajectories. He proposed that bone loading is somehow sensed and that bone adapts its structure accordingly. This principle of functional adaptation is generally known as ‘Wolff’s Law’. It occurs in conditions of disuse when bone is lost, and in intense physical usage when bone mass increases, but also during growth, after fracture healing and in relation with implant incorporation, when the orientation of trabeculae changes. The ability of the bone to adapt to mechanical loads is brought about by continuous bone resorption and bone formation. If these processes occur at different locations, the bone structure is altered. This is called modeling. In a homeostatic equilibrium resorption and formation are balanced. In that case old bone is continuously replaced by new tissue, ensuring the maintenance of mechanical integrity of bone tissue without any global changes in the architecture. This is called remodeling. The modeling and remodeling processes are conducted by specialized bone-resorbing and bone-forming cells, called osteoclasts and osteoblasts respectively. The pathways by which mechanical forces are expressed in osteoclast and osteoblast activity is currently one of the main unresolved issues in bone mechanobiology.

For this dissertation we studied the modeling and remodeling processes as modulated by mechanical forces. We demonstrated that many of the phenomena as they occur in the bone tissue can be explained by a remarkably simple theory, based on the assumptions that osteocytes, which reside in the bone tissue matrix, are mechanosensitive cells capable of sending biochemical messengers to the bone surface. These signals stimulate osteoblasts to form bone whereas they inhibit resorption by osteoclasts.

Applying computer simulations we demonstrated that this control mechanism explains how the bone is modeled in growth towards a mature 3D trabecular structure with realistic characteristics in terms of morphology and physiology, as compared to actual trabecular bone. Eventually a homeostatic steady state is reached in which the mechanical integrity of the bone structure is maintained by ongoing resorption and formation, precisely as it occurs in mature trabecular bone. We demonstrated that the theory also mimics how the structure adapts to alternative loading conditions. It describes how reduced loads lead to bone loss due to trabecular thinning and loss of trabecular connectivity, and how increased loads lead to trabecular thickening. In addition, it also predicts that new loading directions cause trabeculae to gradually realign accordingly.

An important asset of our theory is that it relates cell activity to local trabecular structure. We showed how this feature can be applied to investigate the effects of altered osteoclast, osteoblast and osteocyte activities. We investigated the pathways by which estrogen deficiency after menopause causes the typical patterns of bone loss as they occur in postmenopausal osteoporotic women. These pathways are largely unknown, but we found that the assumption that estrogen deficiency enhances only osteoclast resorption is sufficient to explain the increased remodeling rates and the rapid, yet transient, loss of bone mass due to loss of complete trabeculae, whereas the thickness of remaining trabeculae is preserved. The enhanced osteoblast bone formation could be explained as a secondary effect due to ‘mechanical’ coupling of formation to resorption, indicating that estrogen does not necessarily have a direct, separate effect on osteoblasts. In a second study we investigated the medical intervention effects on osteoporotic bone. Osteoporotic patients are often treated with anti-resorptive drug administration to arrest the degeneration of the bone structure. The morphological consequences of such a treatment are poorly understood. We applied our theory to investigate these effects. A clinically important result is that our theory predicts that preventive anti-resorptive drug administration reduces the rate of bone loss. Compared to late treatment it does not preserve more bone mass on the long term, but it does preserve trabecular connectivity, and thus fracture resistance.

If we consider the modeling and remodeling processes at the cellular level, then it appears that osteoclasts and osteoblasts closely collaborate in what is called a “Basic Multicellular Unit”, or BMU. In cortical BMU’s, osteoclasts excavate cylindrical tunnels in the predominant loading direction of the bone. They are followed by osteoblasts, filling the tunnel, creating secondary osteons of renewed tissue. Trabecular bone remodeling is mainly a surface event, in which osteoclasts dig a trench rather than a tunnel and also here they are followed by bone forming osteoblasts creating hemiosteons. That osteoblasts follow osteoclasts in such a coordinated manner indicates that a coupling mechanism must exist. Its precise nature, however, is currently uncertain. Another unresolved issue in bone mechanobiology is the mechanism that guides osteoclast resorption in its directionality. We refined the scale of our computational theory in terms of spatial and temporal dimensions to address these unresolved issues and demonstrated that the cellular communication pathways that are proposed also provide an explanation for the osteoclast resorption direction and coupling of formation to resorption as observed in remodeling BMU’s of both cortical and trabecular bone.

Summarizing these results, we conclude that bone adaptation according to Wolff’s Law, but also the coordinated fashion by which osteoclasts and osteoblasts collaborate in the remodeling BMU’s of cortical bone and trabecular bone, can all be explained by one coherent cellular level theory for bone modeling and remodeling.