The Actions and Interactions of Sex Steroids and Growth Factors/Cytokines on the Skeleton

The discoveries that sex steroid receptors for estrogen [estrogen receptor (ER)], androgen [androgen receptor (AR)], and progesterone [progesterone receptor (PR)] are present in the bone-forming osteoblasts (OB) and bone-resorbing osteoclasts (OCL), and that these cells are targets for sex steroid action, have generated much excitement in the field of bone biology. The subsequent discoveries that the sex steroids regulate the production of growth factors and cytokines in these OB and OCL cells and that these factors mediate much of the steroid action on the skeleton have added to the excitement and complexity of the system. Because of the required brevity of this minireview, we will discuss only selected aspects of this field using a limited number of references and review articles. For more information in this field, readers are referred to several recent books and chapters that review bone biology, bone diseases, and the role of sex steroids in bone biology and disease (1–8). As a brief background to this unique system, the skeletal maintenance involves a continuous remodeling at discrete sites termed bone-remodeling units (BRU). As outlined in Fig. 1, the multinucleated bone-resorbing OCLs first dissolve bone at these BRU sites, resulting in resorption cavities. While the mechanism(s)/signal(s) that initiates resorption at a given site is unknown, there appears to be a marked regulation of overall bone resorption via the regulation of the OCL differentiation and activity (5–8). These processes are discussed in more detail below. The bone-forming OBs are then recruited by unknown signals, to replace the previously resorbed bone. The identities of the primary factors that couple these processes of resorption to formation to maintain bone mass are also addressed later in this review and are still the subject of much debate. During periods of skeletal growth in children, the increase in bone formation largely involves bone modeling (accretion of bone at surfaces in the absence of bone resorption). The response to mechanical stresses also utilizes mostly modeling. During early adulthood (20–40 yr), bone resorption and formation are in balance and bone mass is maintained. The increased serum estrogen (E) level, which occurs at the onset of puberty in girls, is accompanied by an increase in growth velocity and, ultimately, in the closure of the epiphyseal growth plate and the cessation of linear growth (9). E is the primary hormone responsible for maintaining bone mass in adult women and may serve a similar role along with androgen, in adult men. Finally, during periods of bone loss, such as in postmenopause in women and aging in both genders, bone resorption outpaces bone formation and, in some cases, the bone loss is sufficient to cause osteoporosis. Schematic of the Steroid Actions and Interactions with Growth Factors/Cytokines in Bone Cells at Bone Resorption and Formation Site This model is self-explanatory. The solid arrows represent pathways with significant support in the literature. The broken lines represent potential pathways not currently well identified or characterized (4, 5, 6, 7, 8 ). Schematic of the Steroid Actions and Interactions with Growth Factors/Cytokines in Bone Cells at Bone Resorption and Formation Site This model is self-explanatory. The solid arrows represent pathways with significant support in the literature. The broken lines represent potential pathways not currently well identified or characterized (4, 5, 6, 7, 8 ). It is currently accepted that in postmenopausal females, a deficiency of E results in osteoporosis, whereas in aging males, the deficiency of E and androgens (A) may lead to bone loss (1–3). Ovariectomized (OVX) animals and postmenopausal women experience increased bone turnover, increased bone resorption, and an overall decrease in bone mass. On the other hand, E replacement therapy inhibits this bone loss by inhibiting bone turnover and bone resorption, which prevents the development of osteoporosis (1, 3). Although, the complete mechanisms of action of E in this protection against bone loss are still unclear, new studies are beginning to elucidate the molecular processes involved. The following review summarizes some of the important molecular events in the actions of sex steroids, primarily estrogens, which have been recently reported on the skeleton and bone cells, as well as the role that growth factors and cytokines play in this action. For more details, the readers are referred to several recent reviews that emphasize various aspects of steroid action on bone biology (4–8). That bone cells, especially OB cells, contain ER is fairly well established. Using sensitive techniques, such as immunofluorescence, semiquantitative RT-PCR, in situ hybridization, and in situ RT-PCR, both the ER mRNA and ER protein have been identified in periosteal cells, OBs, osteocytes, and OCLs, at levels less than the reproductive organs, but equivalent to those in liver (9–13). Moreover, Bodine et al. (14), using RT-PCR and primary cultures of rat calvarial OBs, found that ER concentrations increased markedly during OB differentiation. It is fairly well established that E has major effects on OB cells. As summarized in Table 1, E acts directly on OB cell proliferation and expression of genes coding for enzymes, bone matrix proteins, hormone receptors, transcription factors, as well as growth factors/cytokines (4–8, 15, 16). However, results on the specific genetic and metabolic alterations induced by E in these in vitro systems have been conflicting. As outlined in Table 2, these discrepancies could be caused by differences in the receptor concentrations, animal species, skeletal locale of the OB cells, the stages of OB differentiation, and the concentrations of the two ER species (ER-α and ER-β) in the OB cells. Although E inhibits the proliferation of mature human OB cells in culture (15, 16), the possibility that E may stimulate OB proliferation under some circumstances has not been excluded. In mature OB cells in culture, E has been shown to induce the synthesis of transforming growth factor-β (TGF-β), insulin-like growth factor-I (IGF-I), and IGF-binding proteins, and to inhibit the synthesis of interleukin (IL)-1, IL-6, and IL-11 (5–8). E deficiency increases the production of IL-1 and IL-6, as well as macrophage-colony stimulating factor (M-CSF) in bone marrow stromal cells. Further, E has also been reported to inhibit the IL-1 or tumor necrosis factor-α (TNF-α)-induced IL-6 production in human OB and rat bone marrow stromal cells (5–8, 17, 18). More recently, E has been shown to regulate the synthesis of bone matrix proteins, bone morphogenetic protein-6 (BMP-6), osteoprotegerin (OPG), as well as transcription factors, such as TIEG (TGF-β inducible early gene), nuclear factor-κB and c/EBP-β, and c-fos in human OBs (5, 8, 19–22). E has also been shown to regulate the activities of the estrogen response element (ERE)-reporter genes and the IGF-I gene promoter-reporter genes transfected into primary rat or human OB cells in culture, as well as the inositol phosphate receptor gene promoter-reporter in normal and transformed OB cells (5–8). Reported Actions of Estrogen on Gene Expression in Osteoblast Cells in Culture ↑↑↔ means there are conflicting data but the evidence is stronger for induction. ↓↓↔ represents the same but the evidence is stronger for a reduction. Reported Actions of Estrogen on Gene Expression in Osteoblast Cells in Culture ↑↑↔ means there are conflicting data but the evidence is stronger for induction. ↓↓↔ represents the same but the evidence is stronger for a reduction. Suspected Explanations of Differential Actions of E on Osteoblasts Suspected Explanations of Differential Actions of E on Osteoblasts The E regulation of bone mass and remodeling may be initially dependent on mature bone cell activity or on the regulation of bone cell populations (i.e. osteoblastogenesis and osteoclastogenesis). As suggested by Jilka et al. (6, 23), it is possible that changes in osteoblastogenesis induced by E deficiency, paracrine factor production, or cell populations may, in turn, regulate osteoclastogenesis. Indeed, SAMP-6 and OSF2 (−/−) mice, which have decreased or no osteoblastogenesis, respectively, also show decreased osteoclastogenesis (for review see Ref. 6). These results, combined with the finding that E deficiency results in increased osteoblastogenesis as well as osteoclastogenesis in mouse bone, suggests that osteoclastogenesis and increased bone remodeling caused by E deficiency may occur in response to the initial osteoblastogenesis (6, 23). Whether this phenomena is due to changes in bone marrow stromal cell activity or simply populations or to changes in the mature OB activities and cell populations is currently unknown. At least some of the variable actions of E on OB cells may be attributed to different concentrations of ER in the OB lines or to different ratios of ER-α and β. In humans, Smith et al. (24) described a man with an inactive mutated ER-α who had osteopenia and open epiphyses, thereby demonstrating the importance of E in maintaining bone mass in men. In contrast, ER-α knock-out mice displayed only a 10% reduction in mineralization, but otherwise had normal skeletal development (25), suggesting that another ER species or nonreceptor mechanisms may be playing major roles in the E action in skeletal tissues. Significantly, a separate species, ER-β, has been identified in humans and in rat tissues, including OB cells, and found to have partial homology to the extensively studied ER-α (26, 27). The ligand-binding affinities and DNA-binding domains of ER-α and ER-β are similar, but their transactivation domains, tissue distribution, and molecular sizes differ significantly. In some cell systems, using transfection studies with reporter genes, the ER-β was shown to be a weaker regulator of gene transcription than is the well known ER-α (26–28). Further, the ER-α was shown to activate, whereas ER-β inhibits, AP-1 regulatory elements (28). While both ER-α and ER-β are expressed in rat and human OB cells, the two species are differentially expressed during OB differentiation (13, 29). Although the exact functions of ER-β and ER-α in the human skeleton must await further studies, their differential expression in bone cells may help explain the different E responses among the various OB preparations, as well as among the different actions of the selective ER modulators (SERMs), described below. The major action of E on the skeleton in vivo is believed to be an inhibition of bone resorption, mostly by indirect actions of E. This indirect E action involves regulation of growth factor and cytokine production in OBs and their precursor cells, which, in turn, regulate OCL differentiation and activity (discussed further below). However, some direct actions of E on mature OCL activities have been reported. E must have some direct actions on OCLs since avian OCL cells have been shown to contain E receptors, and E has been shown to regulate bone-resorbing activity and specific gene expression by these cells (10, 30). Table 3 summarizes the results from several laboratories, which demonstrated that E directly inhibits bone-resorbing activity in avian, rabbit, and human OCL via the regulation of specific gene activities (5, 10, 30–32). E was shown to antagonize the induction of a novel 150-kDa superoxide dismutase-related membrane glycoprotein by OCL differentiating agents (31). E was also shown by that laboratory to induce an IL-1 decoy receptor and to decrease the expression of an IL-1-signaling gene (32). Other studies using murine and rabbit OCL have substantiated the above avian OCL studies (33, 34). Not only was ER found in rabbit OCL but E inhibited (and anti-E blocked) effects on OCL-resorptive activity, including the production of bone resorbing enzymes (33). E-induced decreases in bone resorption may also be caused by the reduced OCL numbers that are due to E-induced apoptosis (33, 34). This direct E induction of apoptosis may also explain the E inhibition of bone resorption by reducing the OCL life span and population. The E induction of apoptosis may be further enhanced in vivo since TGF-β, whose production by OB is enhanced by E, was also shown to induce OCL apoptosis in these cultures (34). It should be mentioned, however, that some laboratories have not been able to detect ERs in, nor E effects on, OCL cells (5, 35). As with the OB cells (Table 2), these contradictory results may potentially be explained by the age of the animal (embryonic, neonate, or adult) or differences in the OCL cell preparations used (5). Reported Actions of Estrogens on Gene Expression in Osteoclast Cells ↓↓↔ represents the same but the evidence is stronger for a reduction. Reported Actions of Estrogens on Gene Expression in Osteoclast Cells ↓↓↔ represents the same but the evidence is stronger for a reduction. Recently, it has become clear that many actions of E on bone cells are mediated by paracrine factors. Only a few examples of this mechanism are described here, with references to more extensive reviews (2, 4–8). As mentioned at the beginning of this minireview, the coupling between the OB and OCL at the BRU sites infers a cross-talk by chemical signals (factors) between the OB and OCL cells and their precursors (see Fig. 1 and reviews in Refs. 4–8). As outlined in Fig. 2, recent studies suggest that OB production of bone-resorbing cytokines, such as interleukin (IL)-1, TNF-α, M-CSF, IL-6, and PGE2 may mediate bone loss after E deficiency via acting on OCL cell differentiation and activity (6–8). Previous studies have shown that the production of IL-1 and a specific competitive inhibitor of IL-1 action, the IL-1 receptor antagonist, IL-1ra, by peripheral blood monocytes is increased in oophorectomized and postmenopausal women and is suppressed by estrogen replacement therapy (36). Moreover, the of the to decreased bone resorption and bone loss the of the receptor in or of the animals with the increase in bone resorption and bone loss after in both IL-1 and and their receptors play important roles in the bone loss after E Recently, was reported to induce the expression in cells of an inhibitor of OCL differentiation as well as and have been shown to increase the production of the an of OCL differentiation, in murine OB cells The important roles of and are discussed in the of the and between OB and OCL via and Other Growth Factors/Cytokines of the and between OB and OCL via and Other Growth Factors/Cytokines other studies that IL-6 a role in bone loss after E deficiency studies have shown that E IL-6 production by murine bone marrow stromal and cells. Moreover, an IL-6 the increase in osteoclastogenesis induced by in In support of a role for this mice reported to be against (i.e. bone loss (6–8). only the Fig. a that several bone-resorbing cytokines, such as TNF-α, IL-6 (and IL-11 and in osteoclastogenesis and OCL activity, and bone resorption after E E is believed to the production of these cytokines in OBs and other cells which, in turn, to inhibit OCL activity or osteoclastogenesis. It should be however, that studies from other laboratories have not shown a regulation of these factors by E (2, This has the that may be novel of which are described below. It should be mentioned that IL-6, and both well as of their gene in OBs have been shown to their IL-1 and stimulate as well as IL-6 production, and the production of IL-6 and are suppressed by E. changes in of these cytokines could the system. to is the that E the production of a decoy receptor for termed and decreases the expression of the IL-1-signaling receptor gene (32). In the expression of the gene be increased and that of the decoy receptor be which should the OCL to Recently, a novel a of the has been identified by a rat a of studies, is believed to play a major role in bone biology The a and is as a the rat protein and a human have been and both are of osteoclastogenesis (see Fig. mice that a of with decreased osteoclastogenesis Further, mice have markedly increased OCL activity and osteoporosis characterized by decreased bone mass and studies have also identified the for is a cytokine that is by a of cells, including bone marrow stromal cells and is to which was recently identified as a factor cell growth and cell As outlined in Fig. 2, many that is the an from both marrow stromal cells and OBs to the OCL cells to their differentiation into mature in the of M-CSF, replace the for stromal cells in the model for in vitro osteoclastogenesis The osteoclastogenesis effects of are in vitro and in vivo by E of human OB cells increased mRNA levels and protein and these effects inhibited by the E antagonist, or of bone-resorbing cytokines IL-6, in animal prevents sex hormone bone loss is unique in to other cytokines osteoclastogenesis production is by E, by gene transfection results in and results in osteoporosis the evidence is that the acts to IL-6, and PGE2 and may be the the effects of cytokines on the regulation of and an important of the of E deficiency on bone cell More studies to this possibility are are two other growth factor that are believed to also play major roles in bone cell coupling and E and are found in concentrations in the bone only brief review of their but more extensive reviews are (4, 5, 8, In vivo studies have demonstrated that is in and and that of into caused increased bone mass (4, 8, into the of induced the differentiation of periosteal cells into OB cells of mice increased calvarial and bone OBs and OCLs have been demonstrated to in and this production has been shown to be induced by E (5). The in turn, have been shown to regulate gene expression and other activities of the OBs and OCLs (5, The an extensive of the and are an in growth and of this are to in bone and has been shown to be by E whereas E has no on the other IGF-I and are also found in concentrations in the are in OB cells, and are by E (4, 5, E also of the for the which in the activity of the (4, The exact roles of and the in cell differentiation, and bone loss caused by E in the skeleton to be with estrogens, including such as and are not only new into the actions of E in bone, but also are new to osteoporosis the effects of the These have actions and serve as E or the action of E in bone cells and blood E action in reproductive cells. Other such as are E in bone but have a on tissue While the receptor at the same site and with the same as E, induce different changes in the receptor which, in turn, which domains on the ER will be In the with different to different of gene expression than the For E the expression of the IL-6 gene not by direct of the ER with the on but by and inhibiting the activity of transcription factors, such as nuclear and which are required for IL-6 gene transcription In has been shown to the gene for transforming growth in human cells transfected with a mutated ER-α that the DNA-binding in the ER-α Finally, both ER-α and ER-β with AP-1 sites in via with the transcription factors, and However, is ER-α transcription from the AP-1 ER-β inhibits transcription (26, In contrast, in the of such as ER-β transcription from the AP-1 whereas ER-α inhibits the of ER-α or ER-β, or different ratios of the may have a significant on the response of the OBs or OCLs to the The exact mechanisms of action of these to be but may well new into E action in the skeleton and other tissues. The possible role of E in bone biology should be mentioned, since these become the major serum in postmenopausal is to the primary in the of postmenopausal The is then primarily at the or The to the production of an E that the in in cell as E Further, but not decreased bone turnover and growth in similar to E, but had much less of an than E on and there is some evidence that women who less than are at for osteoporosis and show bone loss than women who have levels of Further, the of to with bone These results that has E activities on the skeleton and may play a role in the (i.e. protection of postmenopausal osteoporosis. Only a brief of the actions of the other two of sex steroids, androgens and be The increase in bone mass at puberty in is with increases in of bone formation and is to suggesting that androgen production to the increase in bone mass As in the of E, there is evidence that androgens inhibit bone resorption and have receptors in the bone-resorbing OCLs and inhibit their activity in is with increased bone resorption and bone loss human OB and OCL cells have been demonstrated to have and has been shown to have effects on normal and transformed cells in OB cell lines on the differentiation of the it is that androgens on OB cells to regulate their activity and differentiation studies have shown that an androgen response reporter and regulate OB cell proliferation and bone matrix protein production In and inhibited IL-6 production and the IL-6 receptor gene activity in murine bone stromal cells as well as IL-6 production by human OB cells In to and there is some evidence that androgens may also have significant effects on bone The effects of progesterone on bone are much less clear than those of E or However, have been identified in primary human cells, human cells, and OB cells (for a see Ref. studies using progesterone to bone loss, however, have found variable results, with some studies an of progesterone in of or bone loss, whereas other studies have found no significant of progesterone on postmenopausal bone loss studies are required to the exact role of progesterone and androgens in bone biology. 3 summarizes the potential pathways of the E regulation of bone resorption, as described in this Although the action of E in the skeleton appears to occur at the of bone resorption due to the regulation of OCL activity as well as differentiation of the OCL has to the role of the OB cells. of the growth factors/cytokines that mediate E action on bone resorption as well as OCL differentiation and activity are by OB cells and are under E in the production of these factors or in the regulation of this production by E, be in the overall bone activity found in and bone This review only a of the field of sex steroids and cytokine action on bone. are to further in the more extensive reviews and new in the osteoporosis of women and and the a in into new both and to and bone loss, will be a and for in the to in this the molecular actions and of the sex steroids, and growth factors/cytokines on the skeleton and bone to be This field will present for and in the to in the E of Bone Resorption The IL-6, and inhibit apoptosis and the life span of In contrast, increases apoptosis (5, ). in the E of Bone Resorption The IL-6, and inhibit apoptosis and the life span of In contrast, increases apoptosis (5, ). This was by and