BoneKEy-Osteovision | Commentary
Death and taxes: Glucocorticoids and bone cell apoptosis
DOI:10.1138/2001019
As osteoblasts form bone, they become surrounded by the extracellular matrix (ECM) they secrete that subsequently is mineralized. At this point they become osteocytes that form a syncytial network through gap junctions and canaliculi that permit the passage of extracellular fluid. Whereas newly formed osteocytes have many structural features of osteoblasts with an abundant and well-organized granular endoplasmic reticulum and a large Golgi apparatus, characteristics of active protein-synthesizing cells, osteocytes located at increasing distances from active bone-forming surfaces possess a scanty granular endoplasmic reticulum and a smaller Golgi apparatus. Nevertheless, osteocytes are not “inactive” cells in view of evidence indicating that mature osteocytes have a role in transduction of signals of mechanical loading, thereby acting as the mechano-sensors in bone (). Although osteocytes are long-lived cells, some of them die as shown by the presence of empty lacunae in “inactive” bone (). Frost () used the term “micropetrosis” to describe the pathology where the empty lacunae are surrounded by hypermineralized bone. Increased osteocyte death has been observed in specimens of iliac bone from patients with estrogen deficiency () and in the generalized osteoporosis induced by glucocorticoid (GC) excess (). Osteocyte death associated with the use of high doses of GCs has also been observed in the focal damage in the epiphyses of long bones termed “aseptic necrosis” or “osteonecrosis”.
In order to gain insights into the mechanisms of GC-induced osteocyte death, Eberhardt, Yeager-Jones and Blair studied the bone abnormalities induced by high doses of methylprednisolone acetate in rabbits. In a report this month (), they observed in femurs from treated animals an unusual pattern of tetracycline labeling in ECM deep in the trabeculae, a pattern not seen in untreated animals. In addition, they found that the presence of this damage to the ECM was correlated with death of cells in the subarticular bone in the treated animals, assessed using TUNEL-staining. In a related, recent study of bone from the femoral heads of patients who underwent total hip replacement for GC-induced “osteonecrosis”, Weinstein, Nicholas and Manolagos () found abundant apoptotic osteocytes and bone lining cells in areas adjacent to the subchondral fracture crescent. Weinstein et al., (), however, observed few apoptotic cells in bone from traumatic or sickle cell disease-induced osteonecrosis and proposed that the pathology that results from GC excess is the consequence of direct apoptosis-inducing effects of GCs rather than necrosis from loss of blood supply.
What are possible mechanisms to account for these apoptosis-inducing effects of GCs? It is currently assumed that the biological activities of GCs in suppressing inflammation or tumor growth are mediated via binding to the GC receptor (GR) and, after ligand-mediated dimerization, the GR binds to conserved sequence motifs to positively or negatively regulate specific gene transcription (). There may also be mutual interferences between the GR and and other transcription factors, such as AP-1 and NFκB, that are independent of GR binding to DNA but are dependent upon interaction with these factors. Collagenase genes, such as those that encode MMP-1 and MMP-13, are among the genes whose expression is regulated through AP-1; repression of collagenase expression by GCs appears to be mediated through negative interference between AP-1 and GR. In contrast to GCs, parathyroid hormone (PTH) increases collagenase production by skeletal cells such as osteoblasts and stromal cells by increasing transcription of the MMP-13 gene (). Of the several candidate transcription factor recognition sequences in the MMP-13 regulatory region, the runt domain (Osf2/Cbfa1) site and the AP-1 site are both necessary for PTH activation of MMP-13 gene transcription. Further confirmation of the importance of the runt domain/Osf2/Cbfa1 element in regulating collagenase gene transcription has been obtained by Porte et al. () who introduced mutations into the CCACA motifs required for the osteoblast-specific splice variant of Cbfa1 and these mutations interfered with PTH inducibility of MMP-13 as well as transactivation by Cbfa1. Furthermore, MMP-13 expression is reduced in c-fos -/- mouse embryos and absent in Cbfa1-/- embryos (). The implication of these findings is that PTH/ PTHrP is important in upregulation of bone collagenase and cleavage of bone type I collagen. Does PTH affect bone cell apoptosis? Jilka et al. () showed that the increase in bone mass that accompanies the administration of PTH to normal or osteopenic mice results not from increased generation of osteoblasts but from prevention of osteoblast apoptosis. It follows that such anti-apoptotic effects of PTH might be mediated through collagenase degradation of type I collagen and the pro-apoptotic effects of GCs might be mediated through inhibition of MMP gene transcription.
Many cells bind to type I collagen through the α2 β1 integrin and “liganding” α2 β1 induces procollagenase gene transcription followed by secretion and activation of latent (pro)enzyme. Proteolytic attack on type I collagen then can result in unwinding of the cleaved ends to reveal a cryptic binding site(s) for the αv β3 integrin (). In melanoma cells, for example, “liganding” this cryptic binding site by the αv β3 integrin promotes an adhesion-dependent survival signal necessary for cells to normally progress through the cell cycle. Failure to bind through αv β3 results in apoptosis, possibly mediated by p53 and p21WAF1/CPI1() and other intracellular mediators. We have made use of a mouse model in which a mutation (r) was targeted to both alleles (r/r) of Col1a1 that encodes amino acid substitutions in the α1(I) chains that result in resistance to collagenase cleavage in the helical domain of all 3 α chains in the native type I collagen heterotrimeric molecules. As early as 2 weeks of age, empty osteocyte lacunae were evident in the calvariae and long bones from r/r mice and the number of empty lacunae increased with increasing age (). Many persisting osteocytes as well as periosteal osteoblasts in r/r calvariae were TUNEL positive, whereas few TUNEL positive cells were seen in wild type calvariae. We also obtained evidence that collagenase cleavage takes place in peri-osteocytic ECM in wild type but not in r/r calvariae. Thus, normal osteocytes (and osteoblasts) could utilize signals generated by collagenase cleavage of type I collagen that uncovers cryptic epitopes to which αv β3 integrin can bind. Such signals would not be generated in the r/r mice. Perhaps the pro-apoptotic effects of GCs are exerted through similar mechanisms secondary to inhibition by GCs of collagenase gene transcription and decreased cleavage of collagen in the ECM of bone.
This work is licensed under a
Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.