| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
,25-Dihydroxyvitamin D3: Implications in Cell Growth and Differentiation
Departments of Medicine (A.G., R.K.), Dermatology (M.R.P.), and Biochemistry and Molecular Biology (M.R.P., R.K.), Mayo Clinic and Foundation, Rochester, Minnesota 55905
Correspondence: Address all correspondence and requests for reprints to: Rajiv Kumar, M.D., Departments of Medicine, Biochemistry and Molecular Biology, Mayo Clinic and Foundation, 200 First Street SW, 911A Guggenheim Building, Rochester, Minnesota 55905. E-mail: rkumar{at}mayo.edu
 |
Abstract |
|---|
 Top Abstract I. IntroductionII. 1{alpha},25(OH)2D3 as a...III. Effect of...IV. Impact of 1{alpha},25(OH)2D3...V. Summary and Future...References |
|---|
,25-dihydroxyvitamin D3 [1,25(OH)2D3] is involved in the local control and regulation of cellular growth and differentiation in various tissues, including epidermis (keratinocytes) and bone (osteoblasts and osteoclasts). In this review, the impact of 1,25(OH)2D3 on growth factor/cytokine synthesis and signaling is discussed, particularly as it pertains to bone cells and keratinocytes. 1,25(OH)2D3 not only regulates growth factor/cytokine synthesis but may also alter growth factor signaling. Recently discovered examples for such interactions are the interactions between the vitamin D receptor and the mothers against decapentaplegic-related proteins that function downstream of TGFß receptors. Inhibitory effects of 1,25(OH)2D3 on keratinocytes through TGFß activation and IL-1, IL-6, and IL-8 suppression may provide a rationale for its beneficial effects in the treatment of hyperproliferative skin disorders, whereas stimulatory effects through the epidermal growth factor-related family members and platelet-derived growth factor may be operative in its beneficial effects in skin atrophy and wound healing. Modulation of cytokines and growth factors by 1,25(OH)2D3 during bone remodeling plays an important role in the coupling of osteoblastic bone formation with osteoclastic resorption to maintain bone mass.
,25(OH)2D3 as a Regulator of Cell Growth and Differentiation: Overview ,25(OH)2D3 on Growth Factors and Cytokines: Role in Keratinocyte Growth and Differentiation and Epidermal Pathologies
,25(OH)2 D3 ,25(OH)2D3 function in the skin ,25(OH)2D3 and Growth Factors in Local Control of Bone Cell Development and Function
|
I. Introduction |
|---|
|
TopAbstractI. IntroductionII. 1{alpha},25(OH)2D3 as a...III. Effect of...IV. Impact of 1{alpha},25(OH)2D3...V. Summary and Future...References |
|---|
,25 dihydroxyvitamin D3 [1,25(OH)2D3] act on various cell types to modulate development, growth, and differentiation. All steroid hormones mediate these effects by binding to structurally related intracellular receptors, which in turn interact with defined sequences on regulatable genes to modulate their activity (1). Growth factors and cytokines act through membrane-bound receptors, the stimulation of which result in transcription of several genes involved in cell growth and differentiation (2, 3). The transition from proliferation to differentiation is a complex process requiring tightly coordinated control mechanisms.
In the past several years, a substantial amount of evidence has accumulated concerning the interaction of 1,25(OH)2D3 with growth factors/cytokines in the regulation of coordinated cell growth and differentiation. The effects of 1,25(OH)2D3 may be either antagonized or enhanced by growth factors/cytokines depending on the type of the cell, stage of differentiation, and duration of exposure to the steroid in vitro. The synthesis of growth factors/cytokines as well as their receptors can be altered by 1,25(OH)2D3 in various cell types (4, 5). In this review, the impact of 1,25(OH)2D3 on growth factor/cytokine synthesis and signaling is discussed particularly as it pertains to bone cells and keratinocytes.
The keratinocyte is an excellent model for the study of cellular differentiation in vitro (6, 7). It has recently become apparent that human keratinocytes are organized into stem cells, transit-amplifying cells, and terminally differentiated cells (8, 9, 10). The hierarchical organization of keratinocytes is maintained in cell culture conditions (10). It has been shown that 1,25(OH)2D3 is able to induce keratinocyte differentiation and suppress keratinocyte growth (11, 12). However, mitogenic effects of the sterol on keratinocytes have also been observed (13, 14, 15, 16, 17). This discrepancy seems to be related to concentrations of the hormone and its effects on the synthesis of various local factors implicated in the growth and differentiation of these cells.
It is well known that steroid hormones interact with autocrine/paracrine factors to exert their effects in bone (18, 19, 20, 21). 1,25(OH)2D3 also affects synthesis and signaling of growth factors/and cytokines in bone. This interaction has an important impact on osteoblast and osteoclast development and function and, therefore, coordinates control of bone remodeling.
|
II. 1,25(OH)2D3 as a Regulator of Cell Growth and Differentiation: Overview
|
|---|
|
TopAbstractI. IntroductionII. 1{alpha},25(OH)2D3 as a...III. Effect of...IV. Impact of 1{alpha},25(OH)2D3...V. Summary and Future...References |
|---|
,25(OH)2D3, is the regulation of calcium and phosphorus metabolism, it also plays important roles in the regulation of growth and differentiation in various cells and tissues, including bone cells and keratinocytes (4), cartilage (22), and hematopoietic cells (23). The observation that 1,25(OH)2D3 suppresses cellular growth and induces differentiation of myeloblasts and promyelocytes into macrophages (24), and that it inhibits the proliferation of malignant melanoma cells (25), initially showed that the sterol functioned in areas distinct from its classical transcellular calcium transport function. In a general sense, 1,25(OH)2D3 inhibits the proliferation of the cells and induces them toward a more differentiated state in most cell types, such as osteoblasts (26). In dendritic cells, however, 1,25(OH)2D3 promotes a persistent state of immaturity leading to inhibition of cellular immune responses (27, 28). Therefore, cell culture conditions, cell types, species differences, as well as interactions with growth factor/cytokine signaling are important determinants of the diversity of the biological actions mediated by the sterol.
The effects of 1,25(OH)2D3 are primarily mediated by the vitamin D receptor (VDR), which is a member of the nuclear receptor superfamily (29, 30, 31, 32). Nuclear receptors activate transcription by binding to response elements that consist of hexameric core binding motifs in promoters of their target genes (33). The VDR is a 48-kDa zinc finger protein (34, 35), which activates transcription by binding to vitamin D response elements (VDREs) within the promoter of vitamin D-responsive genes, either as a homodimer (33, 36) or a heterodimer with the retinoid acid X receptor- (37, 38, 39, 40), retinoic acid receptor (41, 42), or thyroid hormone receptors (43).
1,25(OH)2D3 regulates the expression of growth factor/cytokine synthesis as well as receptor expression and modulates growth and differentiation in many cell types including epithelial, mesenchymal, neural, vascular endothelial, immune cells, and chondrocytes (Refs. 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68 and Table 1
). As seen in Table 1, the increased growth factor and cytokine synthesis by 1,25(OH)2D3 generally leads to the inhibition of cellular growth and, in some instances, induction of differentiation. The sterol also exerts immunosuppressive effects through local factors (Table 1). In this review, we will confine our remarks to its effects on growth factors/cytokines using keratinocytes and bone cells as a paradigm. We will also provide information concerning the molecular mechanisms of the interaction of 1,25(OH)2D3 with growth factor/cytokine synthesis and signaling when data are available.
|
|
III. Effect of 1,25(OH)2D3 on Growth Factors and Cytokines: Role in Keratinocyte Growth and Differentiation and Epidermal Pathologies
|
|---|
|
TopAbstractI. IntroductionII. 1{alpha},25(OH)2D3 as a...III. Effect of...IV. Impact of 1{alpha},25(OH)2D3...V. Summary and Future...References |
|---|
. It has become apparent recently that human keratinocytes are organized into stem cells, transit-amplifying cells, and terminally differentiated cells (8, 9, 10). Stem cells have a high proliferative capacity and express ß1,4 integrins. In the presence of adequate levels of calcium, keratinocytes progress from a proliferative status to a differentiated state characterized by irreversible growth arrest, induction of differentiationspecific gene products, and the ability to synthesize cornified envelopes. The cornified envelope is formed by the cross-linking of precursor molecules such as involucrin and lorucrin into an insoluble, durable membrane-associated scaffold by the membrane-bound enzyme transglutaminase. During the progress of differentiation, a successive increase occurs in involucrin, transglutaminase, and cornified envelope formation (69, 70). With the appearance of the cornified envelopes, previously up-regulated transglutaminase and 25-hydroxyvitamin D3-1-hydroxylase activities fall concomitantly, and a reciprocal increase in 25-hydroxyvitamin D3-24-hydroxylase activity occurs (69). Stem cells play a central role in maintaining keratinocyte turnover and epidermal integrity and mass balance. In normal epidermis, the majority of stem cells are quiescent (arrested in Go), but in hyperproliferative skin disorders such as psoriasis, they enter the cell cycle at an enhanced rate (71, 72). An increased rate of stem cell proliferation is believed to be responsible, in part, for epidermal hyperplasia. The proliferative state of stem cells is maintained by various growth factors, primarily those of the epidermal growth factor (EGF) family [such as amphiregulin (AR), TGF, and heparin-binding EGF], as well as TGFß, IGF-I, and keratinocyte growth factor (KGF or FGF7) and cytokines [IL-1, IL-6, IL-8, ß-interferon (IFN)] (72, 73, 74).
|
,25(OH)2D3,25(OH)2D3 induces keratinocyte differentiation. This effect is mediated via the VDR, whose expression in the skin had been noted earlier (75). 1,25(OH)2D3 alters growth of human keratinocytes both in vivo and in vitro (14, 76, 77). We have previously shown that 1,25(OH)2D3 has a biphasic effect on keratinocyte growth that is concentration dependent (14). In defined medium, free of serum, sterol, and pituitary extract, 1,25(OH)2D3 suppresses keratinocyte growth at concentrations greater than 10-8 M, whereas a stimulation occurs at concentrations less than 10-9 M (14). In serum-containing medium, the sterol inhibits keratinocyte growth at all concentrations.
The administration of 1,25(OH)2D3 to keratinocytes is associated with an induction of 25-hydroxyvitamin D3-24-hydroxylase activity, inhibition of 25-hydroxyvitamin D3-1- hydroxylase activity (76), an induction in skin calcium-binding protein synthesis (77), and inhibition of PTHrP production (78, 79). Cultured human keratinocytes do not possess receptors for PTHrP (80) and, therefore, they are not considered as a target for this peptide. However, PTHrP is implicated in the regulation of keratinocyte growth and differentiation in vivo. In a recent study, the absence of PTHrP appeared to result in the reduction of the basal keratinocyte compartment and premature acquisition of suprabasal and granular differentiation markers, whereas overexpression of the peptide generated reciprocal findings (81). These findings suggest that the inhibition of PTHrP by 1,25(OH)2D3 contributes to some extent to the growth-inhibitory and prodifferentiative effects of the sterol on keratinocytes. The growth-inhibitory and prodifferentiative effect of 1,25(OH)2D3 is associated with a decrease in c-myc mRNA levels (12). 1,25(OH)2D3 increases involucrin and transglutaminase activity in keratinocytes, and it increases cornified envelope formation (11, 82, 83, 84, 85, 86). An increase in protein kinase C (PKC) activity and translocation of PKC to cellular membranes has been described (87, 88).
The level of expression of the VDR is a major determinant of cellular 1,25(OH)2D3 responsiveness in keratinocytes. It has been suggested that VDR expression in normal human keratinocytes is linked to cell cycle control (89). VDR levels are markedly decreased by arresting the cells in G0, at the G1-S border, or during the mitotic metaphase. In each of these cases, VDR is rapidly induced when the cells are stimulated to reenter cell cycle. This indicates that VDR expression is restricted to cycling cells but not to a particular cell cycle phase. Proliferating cells should therefore be considered as preferential vitamin D target cells. This is in accord with the identification of the basal keratinocyte as the main VDR-containing cell in the epidermis.
Although previous findings have indicated an antiproliferative effect of 1,25(OH)2D3 on keratinocytes ex vivo (90) and in vitro (12, 91), there are also studies reporting a mitogenic effect mediated by the sterol (13, 14, 15, 16, 17). Similar discrepancies also exist in the clinical effects of the sterol. 1,25(OH)2D3, calcipotriene (MC 903), and noncalcemic vitamin D analogs used in the treatment of hyperproliferative skin disorders such as psoriasis (92) reduce cell proliferation in such conditions. In contrast, topical application of these compounds has also been reported to result in epidermal hyperplasia of human and mouse skin (93, 94), enhanced cutaneous healing in rats (95), and reversal of glucocorticoid-induced epidermal atrophy in mice (96). As discussed below, however, the discrepancies with respect to its effects in epidermis and skin may be reconciled when its local interactions with various growth factors are taken into account. Moreover, the fluctuation of VDR levels according to the proliferation and differentiation state of the cells is a parameter that may help to explain the paradoxical effects of the sterol in vitro and in vivo. 1,25(OH)2D3 inhibits keratinocyte growth in psoriatic lesions due to the dominance of rapidly proliferating cells with up-regulated VDR expression. It has been shown that 1,25(OH)2D3-dependent growth inhibition is reversed into a mitogenic effect in cells that are growth arrested or committed to differentiate (16). In this respect, the sterol stimulates epidermal proliferation of normal skin in which the cells are mostly quiescent and have down-regulated VDR expression.
2. Hair follicle keratinocytes.
1,25(OH)2D3 has an important impact on hair follicle biology. The VDR is expressed in the epidermal keratinocytes and the mesenchymal dermal papilla cells that are components of the hair follicle (97). VDR expression in the hair follicle is increased during late anagen (hair growth phase) and catagen (regression phase), correlating with decreased proliferation and increased differentiation of the keratinocytes (97). It is well known that VDR knockout mice develop alopecia. It has recently been shown that alopecia in these mice is due to a defect in anagen initiation after the period of hair morphogenesis, which ends during the third week of life (98). Hair reconstitution assays performed in athymic nude mice show that VDR expression in dermal papilla cells is not essential for a normal response to anagen initiation but VDR expression by the hair follicle keratinocytes is required for this process (99). The hair cycle defect that leads to alopecia is corrected in VDR knockout mice by transgenic expression of human VDR in hair follicle keratinocytes (100).
C. Growth factors and cytokines in 1,25(OH)2D3 function in the skin
1,25(OH)2D3 modulates the growth and differentiation of epidermal keratinocytes and affects inflammatory processes in skin by altering the synthesis and signaling of several growth factors/cytokines, as summarized in Table 2.
|
, heparin binding-EGF, and AR (107, 108, 109). Keratinocytes also express several cognate receptors for these EGF-related ligands. The main receptor that is activated in response to TGF, AR, and heparin binding-EGF is the type I human EGF receptor (110). In addition to the human EGF-receptor HER1 (also called ErbB1), three other transmembrane proteins that belong to the same family, HER2, HER3, and HER4, may cooperate in signal transduction by EGF receptor (EGFR) ligands. With the exception of HER4, members of the ErbB family are expressed by the keratinocytes (111).
In a recent study by Garach-Jehoshua et al. (13), 1,25(OH)2D3 was found to enhance the proliferation of human keratinocyte cell line HaCaT in the absence of exogenous growth factors. It was demonstrated that HER1/HER3 heterodimers were the major mitogenic signals in 1,25(OH)2D3-stimulated cells. 1,25(OH)2D3 did not affect the levels of the proteoglycan-dependent EGFR ligands AR and heparin-binding EGF. In contrast, a significant increase in the cellular contents of HER1, 2, and 3 was observed, which seemed to be responsible for the stimulatory effect of 1,25(OH)2D3 on autonomous proliferation (Fig. 2 and Ref. 13). Furthermore, 1,25(OH)2D3-stimulated proliferation was suppressed by a specific inhibitor of EGFR tyrosine kinase and a EGFR-neutralizing antibody (Fig. 3 and Ref. 13).
|
|
,25(OH)2D3 has variable effects on HER/ErbB levels in various cellular systems (12, 112, 113). The receptor numbers have been determined by radioreceptor assays in these studies (12, 112, 113). The inability of the radiolabeled EGF binding assays to detect the receptors occupied by the autocrine and juxtacrine ligands defined by Garach-Jehoshua et al. (13) might be another explanation for the variability in HER/ErbB levels described previously (12, 85, 86). It is likely that the proliferative action of the sterol in skin seems to be related to its action through EGF-related/HER signaling pathways.
2. TGFß.
TGFß is a 25-kDa peptide that has pleiotropic effects within the skin and epidermis. Although it can stimulate the growth and metabolic activity of fibroblasts (114, 115), it is a major growth-inhibitory autocrine factor expressed by keratinocytes (116, 117, 118). TGFß, which is expressed by human keratinocytes, arrests keratinocyte growth in a reversible and G1 phase-specific manner (119). Overexpression of a constitutively active form of TGFß targeted to epidermis markedly inhibits keratinocyte proliferation without affecting the terminal differentiation markers discussed above (120). In vivo studies demonstrate hyperproliferative epidermis in animals with targeted disruption of the TGFß gene (121). In psoriatic plaques, TGFß is detectable but restricted to suprabasal cells (122, 123).
Kim et al. (124) showed that human keratinocytes cultured in the presence of 1,25(OH)2D3 produce increased amounts of TGFß1 and -2. More active than latent TGFß1 was produced after exposure to higher doses of 1,25(OH)2D3. Neutralizing antibodies to TGFß blocked the 1,25(OH)2D3mediated antiproliferative effect more than 50%. Enhanced secretion of TGFß1 and -2 after 1,25(OH)2D3 treatment has also been demonstrated in cultured mouse keratinocytes (125). More recently, we demonstrated that the inhibition of keratinocyte growth by 1,25(OH)2D3 was associated with a time- and dose-dependent increase in the TGFß2 concentrations in the supernatant medium of cultured human keratinocytes (91). However, in contrast to the previous observations (124, 125), we did not observe a change in TGFß1 concentrations. We also observed that antibodies directed against TGFß partially blocked the suppression of cellular growth mediated by the sterol (91). Taken together, these results clearly indicate that 1,25(OH)2D3 inhibits keratinocyte growth by a mechanism that involves, at least in part, an increase in the release of TGFß. This mechanism may provide a rationale for the beneficial effects of the sterol in hyperproliferative as well as preneoplastic epidermal skin disorders.
3. Platelet-derived growth factor (PDGF).
PDGF plays an important role in wound healing (126). It stimulates proliferation of fibroblasts and muscle cells, promotes collagen and extracellular matrix synthesis, and acts as a chemoattractant for fibroblasts, monocytes, and neutrophils (127, 128, 129). In the process of wound healing, PDGF functions synergistically with many growth factors (126). It is a disulfide-linked dimer with a molecular mass of approximately 30 kDa. The subunits of the dimer are made up of two related chains (A and B), which are encoded by different genes. There are three PDGF isoforms; two homodimeric AA and BB and one heterodimeric AB. PDGF exerts its bioactivity via cellular receptors of 180 kDa that belong to the Ig superfamily (130). Although platelets and macrophages are two major sources of PDGF, AA, BB, and AB isoforms are synthesized by keratinocytes (101, 131). However, PDGF-receptors are not expressed in normal epidermis or keratinocytes. Zhang et al. (101) have shown that PDGF-AB production is up-regulated by 1,25(OH)2D3 at a time when cellular growth is suppressed by the sterol in cultured human keratinocytes. These results suggest that the increase in PDGF production is regulated coordinately with the increase in growth-inhibitory factors such as TGFß that play a direct role in inhibiting keratinocyte growth. The findings that keratinocytes are a source of PDGF and can be modulated by 1,25(OH)2D3 should foster further investigation of the biological functions of these cells and vitamin D in normal epidermal growth control, dermal-epidermal interaction, and wound healing.
4. IL-1.
Keratinocytes produce the proinflammatory cytokine IL-1 (102, 132, 133). Treatment of unstimulated keratinocytes with 1,25(OH)2D3 showed small effects on IL-1 production and secretion (102). However, when the cells were stimulated with TNF, IL-1 secretion was enhanced and this enhancement was significantly inhibited by 1,25(OH)2D3 (102). The down-regulation of enhanced IL-1 secretion by proinflammatory cytokine TNF may therefore be one of the explanations for the beneficial effects of the sterol in inflammatory skin disorders like psoriasis.
5. IL-6.
IL-6 is an inflammatory cytokine produced and secreted by keratinocytes and infiltrating mononuclear cells implicated in psoriatic skin lesions (103, 134). It has been demonstrated that 1,25(OH)2D3 inhibits the expression of IL-6 from keratinocytes (103) and mononuclear cells (135). Inhibition of IL-6 release from both keratinocytes and infiltrating inflammatory cells also contributes to the beneficial effects of 1,25(OH)2D3 in psoriasis.
6. IL-8.
IL-8 is a member of a family of human leukocyte chemotactic and activating cytokines (136, 137). Apart from being chemotactic for neutrophils (138), T lymphocytes (136), and keratinocytes (139), it also induces the proliferation of keratinocytes (139). In psoriasis, increased IL-8 receptor expression has been detected in both neutrophils (140) and keratinocytes (141). These data indicate an altered regulation of the IL-8 receptor expression in psoriasis. Taking into account the beneficial therapeutic effects of 1,25(OH)2D3 and its analogs in psoriasis, it is speculated that 1,25(OH)2D3 interferes with the expression of IL-8 and its receptor to mediate its antiproliferative and immunosuppresive actions. Accordingly, Larsen et al. (62) have shown that 1,25(OH)2D3 inhibits IL-1-induced IL-8 production and mRNA expression in keratinocytes, fibroblasts, and leukocytes, but not in endothelial cells. Optimal concentrations of the sterol to exert these effects varied between 10-10 and 10-11 M. With respect to the effects on IL-8 receptor expression, Kemeny et al. (104) were able to demonstrate that 1,25(OH)2D3 caused a dose-dependent decrease in IL-8 binding to cultured keratinocytes. Taken together, these data suggest that the inhibition of keratinocyte IL-8 and IL-8 receptor expression by 1,25(OH)2D3 may contribute to its therapeutic effects in psoriasis.
7. TNF.
TNF is produced by keratinocytes (142) and promotes their differentiation with only a modest antiproliferative effect (143). Geilen et al. (105) have demonstrated that 1,25(OH)2D3 stimulates TNF mRNA expression from human keratinocyte cell line HaCaT. The stimulation of TNF from the cells is associated with increased sphingomyelin hydrolysis and ceramide formation. Because ceramide is an important mediator of cell differentiation (144), the TNF-mediated stimulation of sphingomyelin hydrolysis in keratinocytes might contribute to the prodifferentiative actions of the sterol on these cells.
8. RANTES.
RANTES (regulated on activation, normal T expressed and secreted, small inducible cytokine 5A, SCY5A) is a chemokine that specifically attracts eosinophils, T lymphocytes of memory phenotype, and monocytes in vitro (145, 146). Upon cytokine stimulation, epidermal and oral keratinocytes are induced to produce RANTES (147). Fukuoka et al. (106) have investigated the implication of RANTES in psoriatic skin lesions. They found that RANTES was present in the intercellular spaces between epidermal keratinocytes in these lesions. Stimulation with TNF and IFN
synergistically increased the RANTES production of the cells. Moreover, they were able to demonstrate that tacalcitol, an active vitamin D3 analog, inhibited the RANTES production in cultured normal epidermal keratinocytes (106). Taken together, these observations suggest that the regulation of RANTES expression by keratinocytes partly accounts for the action of the sterol as an antipsoriatic drug.
|
IV. Impact of 1,25(OH)2D3 and Growth Factors in Local Control of Bone Cell Development and Function
|
|---|
|
TopAbstractI. IntroductionII. 1{alpha},25(OH)2D3 as a...III. Effect of...IV. Impact of 1{alpha},25(OH)2D3...V. Summary and Future...References |
|---|
,25(OH)2D3 on osteoblast growth and differentiation: an overview.
1,25(OH)2D3 has a biphasic effect on osteoblasts: it abrogates or stimulates the normal developmental pathway or gene expression profiles, depending upon whether it is given, respectively, during the proliferation or differentiation stage. It has been shown that 1,25(OH)2D3 stimulates the proliferation of nonadherent rat marrow-derived osteoprogenitor cells (150). 1,25(OH)2D3 given during the proliferative period of rat calvaria osteoblast cultures inhibits proliferation and down-regulates collagen synthesis and alkaline phosphatase activity; osteocalcin expression is not stimulated (151) and nodule formation is blocked (152), suggesting a stop in osteoblast differentiation. 1,25(OH)2D3 administered in vivo also decreases collagen synthesis and procollagen mRNA in mature osteoblasts (153). 1,25(OH)2D3 treatment of mature osteoblasts results in up-regulation of osteoblast-associated genes, such as osteocalcin and osteopontin, and in stimulation of calcium accumulation (154). Thus, at this stage of differentiation, 1,25(OH)2D3 may promote further maturation of the osteoblast. In summary, despite the controversy in the effects of 1,25(OH)2D3 on osteoblast growth in vitro, general agreement is in favor of its antiproliferative and prodifferentiative properties.
2. Growth factors in 1,25(OH)2D3 function.
1,25(OH)2D3 alters growth factor synthesis, release, and signaling in osteoblasts (Table 3).
|
The biological effects of TGFß are mediated through cell-specific transmembrane serine/threonine kinase receptors in osteoblasts (162). The heterodimeric interaction of liganded type I and type II receptors is needed for signal transduction and, consequently, mediation of the different biological responses (162). TGFß signal transduction and transcriptional control are mediated in the cells by a family of receptor substrates, the Smad proteins (mothers against decapentaplegic-related proteins) that translocate to the nucleus and act as transcription factors (163). Smad3 plays an essential role in TGFß-mediated signal transduction. It has been shown that targeted disruption of Smad3 in mice markedly blunts the cellular effects of TGFß (164).
If primary human osteoblastic cell cultures are treated with 1,25(OH)2D3 and TGFß, they stimulate spindle-shaped cells to become stellate in appearance and increase the number of cytoplasmic processes (165). TGFß increases 3H-thymidine incorporation and 1,25(OH)2D3 reduces this effect. In contrast, procollagen type I synthesis and secretion are increased in a dose-dependent manner in the presence of TGFß, and these effects are not suppressible by 1,25(OH)2D3. TGFß and 1,25(OH)2D3 each marginally increase alkaline phosphatase activity whereas cotreatment synergistically increases alkaline phosphatase activity in a dose-and time-dependent manner (165). TGFß suppresses 1,25(OH)2D3-induced osteocalcin expression and secretion (165, 166). In summary, these data suggest that TGFß stimulates the human osteoblast to actively produce collagen matrix and proliferate. 1,25(OH)2D3 stimulation may induce cells to advance to an endstage where cell proliferation is reduced and osteocalcin expression is promoted.
Evidence is now available with regard to the mechanisms of 1,25(OH)2D3-TGFß interaction and the role of TGFß in 1,25(OH)2D3 function in osteoblasts. Finkelman et al. (167) have previously shown that there were significant reductions in skeletal TGFß but no differences in IGF-I, IGF-II, or osteocalcin in vitamin D-deficient rats compared with vitamin D-replete ones. To determine whether 1,25(OH)2D3 increased TGFß production by bone cells, they treated mouse calvaria for 6 d and mouse osteoblasts for 2 d with 10 nM 1,25(OH)2D3 (167). They observed a nearly 100% increase in TGFß production with 1,25(OH)2D3 treatment. Similarly, we have shown that treatment with 1,25(OH)2D3 increased the transcription rate of TGFß2 in cultured human fetal osteoblasts (Ref. 168 and Fig. 4). Antibodies directed against TGFß partially blocked the antiproliferative effect by 1,25(OH)2D3 (168). Furthermore, 1,25(OH)2D3 stimulated the mRNA expression of TGFß type I and type II receptors (Ref. 168 and Fig. 5). Taken together, these data suggest that antiproliferative effects of 1,25(OH)2D3 on osteoblasts is mediated, at least in part, by TGFß. The release of TGFß, which also influences the activity of the adjacent osteoclasts, plays an important role in bone remodeling.
|
|
,25(OH)2D3 enhances the TGFß expression have been recently highlighted by several studies. Because VDREs, to which the VDR binds, are comprised of direct repeats, palindromes, or inverted palindromes of the hexameric core binding motif RRKNSA (R = A or G, K = G or T, S = C or G) (33, 43, 169), we investigated whether 1,25(OH)2D3 induces TGFß2 expression via such repeat elements. We transiently transfected human fetal osteoblasts (hFOB cells) with a series of TGFß2 promoter/human GH reporter constructs (170). We identified two imperfect direct repeat sequences (TGTAGAACAAGTAGA and AATGAAGTTGGTGGA) in the TGFß2 promoter where the heterodimers of the VDR and RXR bind and, therefore, mediate the effects of 1,25(OH)2D3 (Ref. 170 and Fig. 6).
|
,25(OH)2D3 of TGFß is the stimulation of AP-1 (activator protein-1) via a genomic action. Takeshita et al. (171) have shown that 1,25(OH)2D3 markedly enhances the transient activity of TGFß1-induced AP-1 binding to the 12-O-tetradecanoylphorbol-13-acetate response element in a clonal mouse osteoblastic cell line, MC3T3-E1. The synergistic activity was markedly suppressed by VDR antisense oligonucleotides before treatment. Furthermore, 1,25(OH)2D3 synergism of the TGFß1induced expression of c-jun, but not c-fos, was enhanced in VDR expression vector-transfected cells (171).
In recent years, evidence has accumulated with respect to the interactions between 1,25(OH)2D3 and events downstream of TGFß receptor stimulation. Yanagisawa et al. (172) showed that Smad3, a regulatory Smad, acts as a coactivator of VDR and positively regulates the vitamin D signaling pathway. These data suggest that TGFß and 1,25(OH)2D3 signaling pathways by which osteoblast gene expression is induced converge on Smad3. The consequences of the targeted disruption of Smad3 in osteoblasts were recently evaluated by Borton et al. (173). They showed that TGFß can no longer inhibit the differentiation of osteoblasts in Smad3 null mice; nevertheless, stimulation of the proliferation remains intact (173). The clinical consequence of the lack of Smad3 in these mice was osteopenia, which is attributed to a decreased rate of bone formation. The reduction in bone formation is associated with increased osteocyte number and apoptosis (173). These results suggest that inhibition of osteoblast differentiation and stimulation of osteoblast proliferation is regulated through divergent mechanisms by TGFß. Whether 1,25(OH)2D3 signaling is implicated in the diversity of these TGFß actions on osteoblasts is currently not known. In view of the complex interplay between TGFß and 1,25(OH)2D3, it is highly likely that other factors downstream from the signaling pathway are also operative. As an example, Yanagi et al. (174) have recently demonstrated that Smad7, an inhibitory Smad, abrogates the Smad3-mediated potentiation of VDR function. However, more work is required to clarify molecular mechanisms of the interplay between TGFß and 1,25(OH)2D3 in the context of inhibitory Smads and other Smad coactivator and corepressor partners with which Smads cooperate.
b. IGFs and IGF binding proteins (IGFBPs).
IGF-I and II are traditionally known as the mediators of GH action (175). In addition to their major roles in the growth and differentiation of cartilage, they directly affect bone. IGF-I and IGF-II were found to stimulate cell multiplication as well as collagen synthesis in fetal rat calvaria (176, 177) and in primary osteoblast-enriched cell cultures (178, 179). Studies have shown that IGF-I is also a mediator for some of the anabolic effects of PTH in calvaria cultures (180). IGF-I and -II synthesis and secretion have been detected in the conditioned media of organ cultures (181), osteoblast-like cell cultures (178), and preosteoblastic human bone marrow stromal cells (182). IGF receptor expression has been demonstrated in clonal osteoblastic MC3T3-E1 cells (183). Therefore, it is clear from these observations that IGFs locally mediate for osteoblast development and function. The synthesis and release of IGFs during bone resorption suggest a role of these growth factors in the coupling of bone resorption with formation.
It has previously been reported that there is a synergistic interaction between IGF-I and 1,25(OH)2D3 with respect to the stimulation of alkaline phosphatase activity in osteoblastic cell lines (184). Observations vary with regard to the effects of 1,25(OH)2D3 on IGF synthesis by osteoblasts and preosteoblastic cells. Kurose et al. (183) have reported that IGF-I production of the murine clonal osteoblastic cells is not affected by 1,25(OH)2D3 treatment. In accordance with this observation, Kveiborg et al. (182) recently reported that 1,25(OH)2D3 (dose range 10-10 to 10-7 M) exerts no effect on either IGF-I or IGF-II mRNA expression in preosteoblastic human bone marrow cells. However, Chenu et al. (185) showed that 1,25(OH)2D3, in a time- and dose-dependent manner, increased IGF-I release from human osteoblast-like cells in short-term cultures. Linkhart and Keffer (186) have found that 1,25(OH)2D3 stimulates IGF-I but inhibits IGF-II release by cultured neonatal mouse calvarial cells, suggesting differential regulation of IGF subtypes by the sterol. The results of these studies suggest that the effect of 1,25(OH)2D3 on IGF synthesis and release are divergent, possibly depending on the differences in cell types, culture conditions, and differentiation status of the cells.
1,25(OH)2D3 may also exert modulatory effects on IGF receptor expression by osteoblasts. For example, Kurose et al. (183) found that the affinity and hormone binding capacity of VDRs were not affected by IGF-I in the MC3T3-E1 osteoblastic cell line. However, 1,25(OH)2D3 significantly induced IGF-I receptor expression. This was evident by Scatchard analysis, which revealed a significant increase in IGF-I binding sites by 50% after treatment for 3 d with 5 x 10-11 M 1,25(OH)2D3, without any change in affinity (183).
The bioactivity of IGFs is regulated through their interaction with a group of high-affinity IGFBPs. At least six IGFBPs (IGFBPs 1–6) have been identified and were found to have a variety of biological effects that are both IGFBP and cell type specific (187). The activities of IGFBPs are regulated through changes in their rates of degradation by multiple specific and nonspecific proteases and protease inhibitors (188, 189). The exact functions of the different IGFBPs in bone metabolism are complex and are not yet fully understood. However, increasing evidence suggests that they may modulate the activities of IGFs. For example, exogenous IGFBP-5 enhances the mitogenic potential of IGF-I (190) or IGF-II (191) added to mouse osteoblast-like cells in vitro and has an important role in the storage of IGF-II in bone matrix (191). IGFBP-1 (192), IGFBP-3 (193), and IGFBP-4 (194) have been shown to suppress the functions of IGF actions in human osteoblast-like cells. IGFBP-2 has been shown to potentiate (195) or antagonize (196, 197) the IGF-mediated effects, and its role in osteoblast biology is not precisely known.
In human preosteoblastic bone marrow stromal cells, 1,25(OH)2D3 has been found to increase steady-state mRNA levels of IGFBPs 2, 3, and 4 (182). Western ligand blot analyses revealed consistently the stimulation of secretion of the IGFBPs (182). 1,25(OH)2D3 stimulates IGFBP-5 mRNA expression in cultured neonatal calvarial cells (198) and the osteoblastic UMR-106 cell line (199). Increased mRNA expression and secretion of IGFBP-3 and 4 have also been observed in human osteoblastic cells in several other studies (200, 201, 202, 203). As a common point, the increase in the expression of these IGFBPs was not associated with increased proteolytic activity. The increase in mRNA and protein levels of IGFBPs requires continuous exposure to 1,25(OH)2D3 for more than 48 h (182), suggesting that these changes are secondary to effects on immediate responsive genes. As no consensus VDRE has been found in the promoter regions of the IGFBPs, it is plausible that 1,25(OH)2D3 effects are mediated through other signaling pathways such as cAMP or AP-1 transcription factors, which have been shown to mediate some of the genomic effects on components of the IGF system (204, 205).
Given the inhibitory effects of IGFBPs in IGF actions on bone cells (206, 207), it is plausible that the inhibitory effects of 1,25(OH)2D3 on bone formation may be mediated, at least in part, by the enhanced production of inhibitory IGFBPs. However, the regulatory roles of IGFBPs in the effects of 1,25(OH)2D3 on bone cell biology are complex and deserve further evaluation. Based on earlier observations that addition of excessive amounts of IGFBPs to cell cultures resulted in variable effects, more physiological approaches are required. This will rely on the assessment of endogenous levels of IGFBPs and development of IGFBP-specific neutralizing antibodies that will prevent interaction with IGFs.
c. Nerve growth factor (NGF).
NGF belongs to a family of neurotrophic factors named neurotrophins, which includes brain-derived neurotrophic factor and neurotrophins 3, 4, and 5 (208). NGF was initially described on the basis of its trophic effects on sympathetic and some sensory neurons in the peripheral nervous system (208). However, its role now extends to cells belonging to the endocrine and immune systems. In the mouse fibroblastic cell line L929, the expression of the NGF gene has been shown to be positively regulated by an activator of PKC, the phorbol 12-myristate 13-acetate (209), and 1,25(OH)2D3 (51). The presence of NGF in the embryonic bone of the chick has been shown by immunochemistry (210), as has the expression of NGF family neurotrophins in the MC3T3-E1 mouse osteoblastic cell line (211). The regulation of NGF expression by 1,25(OH)2D3 in osteoblasts was first studied by Jehan et al. (212). In the ROS 17/2.8 rat osteoblastic cell line, they showed that 1,25(OH)2D3 caused a dose-dependent increase in NGF mRNA expression. Binding of radiolabeled NGF displayed the presence of low-affinity receptors for the growth factor in these cells (212). In the rat ROS17/2.8 osteoblastic cell line, we have shown that treatment with 1,25(OH)2D3 for 6 h resulted in a dose-dependent increase in NGF expression (213).
The biological significance of the 1,25(OH)2D3-induced NGF expression by osteoblastic cells has not yet been clarified. Regeneration after limb amputation can be successfully induced by transplanting nervous tissue to the lesioned limb (214, 215), and symphatectomy may result in the loss of a trophic influence important in the regulation of osteogenesis (216). More recently, it was shown that topical application of NGF improves fracture healing in rats (217). However, whether NGF produced by bone cells is associated with the nerve supply of bone, is directly implicated in bone formation, or acts on bone marrow cells to promote proliferation remains to be further explored.
The molecular mechanisms by which 1,25(OH)2D3 induces NGF expression in osteoblasts have not yet been fully understood. The NGF gene contains an AP-1 site within the first intron, 35 bp downstream from the transcription initiation site (218). This AP-1 site has been shown to be functionally required for basal level NGF transcription in fibroblasts (219). Members of the fos and jun family of transcription factors are known to bind to these AP-1 sequences. Indeed, lesion-induced increases in NGF expression have been mediated by c-fos (220). In a previous study from our laboratory (213), we have shown that 1,25(OH)2D3 induces NGF expression in ROS 17/2.8 osteoblastic cells indirectly by increasing AP-1 binding activity (213).
d. Vascular endothelial growth factor (VEGF).
Osteoblasts are located in proximity to endothelial cells. It is highly likely that a mutual communication exists between endothelial cells and parenchymal cells, as demonstrated in liver and thyroid (221, 222). Osteoblasts and preosteoblastic progenitor cells have been found to locate adjacent to endothelial cells in blood vessels at sites of new bone formation (223). Vascular invasion is a prerequisite for endochondral bone formation and fracture healing (224). Factors produced by endothelial cells may therefore affect osteoblast function or differentiation and vice versa (225, 226). As a clinical clue to the interaction between osteoblasts and endothelial cells, older subjects and patients with osteoporosis have been shown to have decreased blood vessels in their skeletal tissue that is accompanied by a parallel decrease in osteoblast number (227, 228). In this cross-talk between endothelial cells and osteoblasts, VEGF, which acts as an angiogenic factor, may play an important role (59). A recent study by Gerber et al. (229) revealed that blocking the action of endogenous VEGF inhibits both bone formation and resorption in juvenile mice.
VEGF, a homodimeric protein with a signal peptide, specifically stimulates endothelial cell proliferation by binding to VEGF receptors (Flt-1 and KDR) that are expressed exclusively on the cells (230, 231). During osteoblast development, both VEGF and VEGF receptor expression increases in parallel with the advancement of the cell differentiation (232). VEGF mRNA expression has been found to be induced by 1,25(OH)2D3 in SaOS-2 osteoblast-like cells (233), human osteoblast-like cells (234), and cocultures of osteoblasts with endothelial cells (235). Stimulation of endothelial proliferation is partly blocked by anti-VEGF antibodies in cocultures of endothelial and osteoblastic cells (59). 1,25(OH)2D3induced VEGF expression in osteoblasts is enhanced if the cells are cocultured with endothelial cells (59). Interestingly, 1,25(OH)2D3 induces the expression of endothelial, but not osteoblastic, VEGF receptor expression only when these cells are cultured together (59). According to a model proposed by Wang et al. (59), activated endothelial cells expressing a greater amount of VEGF receptor genes, in turn, produce osteotropic growth factors, such as IGF-I and endothelin-1, which synergistically stimulate the proliferation of osteoblastic cells and alkaline phosphatase activity (Fig. 7) (59). In view of these observations, it is plausible to state that there is a mutual communication between osteoblasts and endothelial cells in bone. The anabolic effects of 1,25(OH)2D3 on osteoblasts are mediated in a paracrine manner and enhanced by a VEGF/VEGF receptor system between osteoblasts and endothelial cells.
|
,25(OH)2D3 increases VEGF expression in osteoblasts is not currently known. VEGF promoter is a target for 1,25(OH)2D3 despite the lack of classical response elements in it (233). In a manner similar to prostaglandin E2-induced VEGF mRNA expression in rat osteoblastic cells (236), the 1,25(OH)2D3-stimulated expression of VEGF mRNA in human osteoblast-like cells was completely inhibited by H-7, a general inhibitor of protein kinases, but only partially inhibited by staurosporine, a more specific inhibitor of PKC (234). Therefore, although there are several reports showing that the 1,25(OH)2D3 activates PKC at physiological concentrations (237, 238), this activation does not seem to be completely responsible for the induction of VEGF expression in osteoblasts.
e. EGF family.
EGF and related family members have been shown to be important growth factors for a variety of cells including numerous epithelial and epidermal cells (discussed further in Section III.C.1) in vivo and in vitro. Bone cells express EGFRs (239, 240), and EGF has been demonstrated to stimulate DNA synthesis as well as to enhance growth of osteoblast-like cells (241). EGF has also been shown to increase bone resorption (242, 243), decrease collagen synthesis (244), and inhibit the formation of mineralized nodules (245). Previously, PTH has been shown to increase the expression of EGFR in the UMR 106-01 osteoblastic cell line (246). In an early study by Petkovich et al. (247), 1,25(OH)2D3 caused a dose- and time-dependent increase in EGFR expression in RCJ 1.20 rat calvarial cells. The increase in EGFRs potentiated the effect of EGF on anchorage-dependent and anchorage-independent growth of these cells (247). More recently, the stimulatory effect of 1,25(OH)2D3 on EGFR expression was also confirmed in UMR 106-01 osteoblast-like cells (248). In this study, the effects of 1,25(OH)2D3 on EGFRs were completely abolished in the presence of retinoic acid (248). Because EGF and other family members are important factors in the growth and development of osteoblasts, modulation of its receptors by 1,25(OH)2D3 plays an important role in its effects on these cells.
B. Osteoclasts
1. Formation and function of the osteoclast.
As discussed previously, bone resorption is one arm of the bone remodeling process that is usually balanced by bone formation to maintain bone mass. Under normal conditions, bone-resorbing osteoclasts are present in appropriate numbers, and they cycle between active and inactive states in a tightly regulated fashion. When in balance, the remodeling process involves equal and linked participation of both osteoblasts and osteoclasts. Deviation of the balance in favor of resorption leads to bone loss and, therefore, osteoporosis (249). Osteoclasts resorb bone by a complex, multistep process that occurs in an orderly sequence and utilizes cross-reacting intracellular signaling pathways (249, 250). The earliest identifiable osteoclast precursors are derived from colony forming unit-granulocyte/macrophage (CFU-GM) from the granulocyte/macrophage lineage (249). Osteoclast precursor cells are attracted to regions of mineralized bone matrices that are identified in an unknown way to undergo resorption to maintain plasma calcium levels.
The osteoclast life span can be subdivided into distinct stages as follows: 1) homing of the precursor cells (CFU-GM) to bone surfaces and proliferation (formation of preosteoclasts), 2) fusion of mononuclear preosteoclasts into multinucleate immature osteoclasts, 3) development of polarity through adherence to bone and assembly of an acidifying resorptive apparatus (ruffled border) indicative of cellular activation, 4) detachment and final inactivation that ends in apoptosis. As discussed in detail below, endocrine and local control by several factors are effective in osteoclastogenesis and function of the cells.
2. Endocrine and local control of osteoclast activity.
Osteoclast function is regulated by a variety of hormones, including 1,25(OH)2D3. Bone resorption is stimulated by PTH and 1,25(OH)2D3, whereas it is inhibited by sex steroids (discussed earlier) and calcitonin. Most stimulators appear to act indirectly, in contrast to most inhibitors, which act directly. PTH action in bone is a good example for the indirect mechanism. PTH increases bone resorption by increasing the activity and number of osteoclasts (249). However, it has also been noted that isolated osteoclasts do not express PTH receptors and therefore, isolated osteoclasts can not directly respond to PTH (251, 252). The addition of osteoblasts to isolated osteoclasts restores the ability of the osteoclasts to resorb bone in response to PTH (251). This finding suggests that PTH activates osteoclasts indirectly by stimulating osteoblasts and perhaps osteoblast precursors, which then activate osteoclasts.
There are several growth factors and cytokines that are involved in the local control of osteoclastogenesis and osteoclast function (Table 4).
|
, the effects of these factors on osteoclast formation is in accordance with their effects on osteoclast function, except in the case of macrophage-colony stimulating factor (M-CSF). M-CSF stimulates osteoclastogenesis despite being inhibitory with respect to osteoclast activity (262). The effects of TGFß on osteoclastogenesis and function are variable. For example, Chenu et al. (260) have shown that TGFß inhibits both the proliferation and fusion of human osteoclast progenitors. However, TGFß also enhances osteoclast formation and function through its interaction with receptor activator of the nuclear factor 
B (RANK) ligand (RANKL) (269). To evaluate the physiological role of TGFß in bone development and turnover in vivo, Erlebacher and Derynck (261) have used transgenic mice that overexpress TGFß from the osteocalcin promoter. Osteoblast-specific overexpression of TGFß resulted in progressive bone loss associated with increases in osteoblastic matrix deposition, increased density of matrix-embedded osteocytes, and enhanced osteoclastic bone resorption. This phenotype closely resembles the bone abnormalities seen in human hyperparathyroidism and osteoporosis implicating TGFß as a physiological regulator of bone remodeling (261).
3. Regulatory effects of 1,25(OH)2D3 on local factors.
Bone remodeling is a complex process during which previously described osteoblastic bone formation is coupled with osteoclastic bone resorption by systemic and local factors that function together to maintain bone mass as illustrated in Fig. 8. 1,25(OH)2D3 is a potent stimulator of osteoclastic bone resorption and osteoclast formation. It has been shown that 1,25(OH)2D3 acts as a fusigen for committed osteoclast precursors (270). It is unlikely that 1,25(OH)2D3 acts on mature osteoclasts directly, because they do not express VDR (271). Therefore, it seems that 1,25(OH)2D3 exerts these effects indirectly by modulating local control mechanisms as summarized in Table 3 and Fig. 8, and discussed in the following sections.
|
and IL-1 cytokine systems, this newly recognized cytokine system also consists of 1) a ligand, RANKL, that exists in a cell-bound and soluble form; 2) a cell-bound receptor, RANK; and 3) a secreted decoy receptor, OPG (268, 273). OPG is expressed ubiquitously and abundantly by a variety of cell types including the osteoclast precursors (274). Various skeletal and extraskeletal cells such as stromal cells, osteoblasts, osteoclasts, mesenchymal periostal cells, chondrocytes, and endothelial cells express RANKL (275, 276, 277, 278, 279). It has recently been demonst
, ß1 integrin expression; 
, ß4 integrin expression. Note that ß1 and ß4 integrin expressions are at maximum in stem cells, and their expression decreases in transit-amplifying cells. With the advancement of differentiation, the cells no longer express ß1 and ß4 integrins.
, cAMP responsive elements; 
, vitamin D response elements. [Reproduced from Y. Wu et al.: Biochemistry 38:2654–2660, 1999 (
