1,25-Dihydroxyvitamin D3 Increases Synthesis of the Vitamin K-dependent Bone Protein by Osteosarcoma Cells*

Rat osteosarcoma cells respond to 1,254hydroxyvi- tamin Da with a 6-fold increase in intracellular and secreted levels of the vitamin K-dependent protein of bone (BGP). The rise in intracellular BGP levels is half-maximal at 6.6 h and precedes the rise in medium BGP levels by 6 h, a time course which is consistent with the postulated steroid hormone action of 1,25-dihydroxy-vitamin Ds. This effect is achieved by physiological levels of 1,25-dihydroxyvitamin DI, with half of the maximal response at a vitamin concentration of 0.04 ng/ml. The specificity of this effect for BGP is demon-strated by the absence of a 1,25-dihydroxyvitamin D3 effect on total protein synthesis by these cells. To our knowledge, BGP is the first example of a bone protein whose rate of synthesis is dramatically and specifically increased by physiological levels of 1,25-dihydroxyvitamin D3. "he possible functions of BGP in the biological actions of 1,25-dihydroxyvitamin DS on bone are discussed.

The active metabolite of vitamin D, 1,25-dihydroxyvitamin D3, stimulates bone mineral mobilization (1) in a process which requires parathyroid hormone (2). This response is blocked by prior administration of actinomycin D (3) suggesting that protein synthesis is required for the action of 1,25(OH)2D3' on bone. The initial event in 1,25(OH)nD3-induced bone calcium mobilization is probably the association of 1,25(OH)2D3 with the cytosolic receptor of bone cells (4)(5)(6)(7). The probable subsequent action of the receptor complex with 1,25(OH)zD3 in bone cells is best illustrated by the action of 1,25(OH)zD3 on intestinal c e h . Here the cytoplasmic receptor plus 1,25(OH)2D3 moves to the nucleus and saturates chromatin receptor binding sites (8) which then alters transcriptional events (9)(10)(11)(12) and leads to de novo synthesis of the vitamin D-dependent cytosolic calcium binding protein (13,14). This evidence suggests that the 1,25(0H)tD3-induced bone calcium mobilization response is mediated by a protein or proteins whose synthesis rate is increased by the action of 1,25(OH)*Ds on bone cells. The putative proteins synthesized * This work was supported in part by United States Public Health Service Grants GM 17702 and AM 25921. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
To whom correspondence should be addressed. by bone cells in response to 1,25(OH)?D3 have yet to be identified, however, and the mechanism by which such proteins might mobilize bone calcium is not understood.
In the present study, we have examined the effect of 1,25(OH)zD3 on the synthesis of the vitamin K-dependent bone protein. This 49-residue protein of known structure (15)(16)(17) contains y-carboxyglutamic acid and has been termed bone Gla protein or BGP. BGP is one of the most abundant noncollagenous proteins in the extracellular bone matrix (18,19). It is synthesized in bone (20) and binds strongly to bone mineral in an association which is mediated by y-carboxyglutamate residues (21, 22). BGP is also found in plasma (23)(24)(25), and clinical studies demonstrate that plasma BGP levels are elevated in patients with metabolic bone diseases characterized by increased rates of bone turnover (23,26). Although the function of BGP is presently unknown, there is some evidence to suggest that it may play a role in calcium or skeletal homeostasis (22,26).
The clonal osteosarcoma cell line used for these studies (ROS 17/2) secretes BGP into culture medium (27) and has receptors for 1,25(OH)2D3 (28). It also displays in cell culture such features of the osteoblastic phenotype as high parathyroid hormone responsiveness and alkaline phosphatase activity (29) and forms a mineralized osteosarcoma when implanted into a rat (29).

EXPERIMENTAL PROCEDURES
Materials-Bone Gla protein was purified from the proteins released by demineralization of rat bone by gel filtration over Sephadex G-100 and subsequent gradient elution from DEAE-Sephadex A-25 as described previously (19). ~- [4,5- Radioimmunoassay-The procedures used for the radioimmunoassay of rat BGP in osteosarcoma cell culture media and in guanidine HCI extracts of osteosarcoma cells have been described (27). Osteosarcoma Cell Culture-The procedures used to maintain osteosarcoma cell lines in culture have been described (27). For measurement of the effect of 1,25(OH)zDa on BGP secretion, each cell type was grown to confluency in 60-mm culture plates in Coons F12 medium supplemented with 10% fetal calf serum. The medium was then changed and the cells were cultured another 24 h. At this time, the experiment was initiated by changing the medium to 5 ml of Coons F12 medium supplemented with 2% fetal calf serum and the desired amount of 1,25(OH)zD3 in 20 p1 of ethanol. Twenty-four and forty-eight hours after the start of the experiment, the medium was exchanged for fresh medium at the same concentration of fetal calf serum and 1,25(OH)& The appearance of BGP in the medium was determined by triplicate radioimmunoassay of aliquots removed at appropriate times. The effect of 1,25(OH)~Do on the general rate of protein synthesis was assessed by incubating cultures with 10 pCi/ml of [3H]leucine. Intracellular BGP levels were measured in trypsindissociated cells after one wash with culture medium and one wash with phosphate-buffered saline. The final cell pellet was lysed by freeze-thawing after addition of 0.15 ml of 6 M guanidine HC1 in 0.1 M Tris buffer at pH 8 per 60-mm culture pellet and assayed for BGP.

1,25-Dihydroxyvitamin D3 Increases Bone Gla Protein Synthesis
above the control level of 15 ng of BGP per 10" cells. Some increase in intracellular BGP levels is seen at 1 h, and half of the total increase is achieved by 6.6 h. 1,25(OH)z& also dramatically elevates the level of BGP in culture media (Fig.  1). The increase in media BGP level occurs about 6 h after the increase in intracellular levels. Fig. 2 demonstrates that the kinetics of BGP appearance in media is identical for cultures pretreated with 1,25(OH)2D3 for 24 and 48 h. Thus, the response of rat osteosarcoma cells to 1,25(OH)2D3 is complete by 24 h.
The dependence of BGP secretion on the concentration of 1,25(OH)zD3 was determined after pretreatment with the vitamin at the test concentration to ensure that the response was maximal. Table I shows that the dose response is essentially the same in media sampled at various times after a 24or 48-h pretreatment. The average increase in media BGP at each 1,25(OH)*D3 concentration is graphed in Fig. 3 to illustrate that the response is centered at physiological levels of 1,25(OH)zD3 in plasma.
The response of ROS 17/2 cells to vitamin D3 was also  Table I. Cells treated with vitamin D3 at concentrations from 0.1 to 100 ng/ml of culture medium did not increase BGP secretion above control after either a 24-or 48-h pretreatment with vitamin D3. Higher vitamin D3 levels did stimulate BGP secretion and 40% of the maximal 6-fold increase in BGP secretion was achieved at a vitamin D3 concentration of 1 pg/ml. This vitamin D3 concentration is over 104-fold greater than the concentration of 1,25(OH)zD3 which gives the same level of BGP response.
The effect. of 1,25(OH)*D3 on total protein synthesis by ROS 17/2 osteosarcoma cells was evaluated by comparing the levels of leucine label incorporated into protein at different 1,25(OH)zD3 concentrations. As can be seen in Table 11, at no concentration of 1,25(OH)zD3 is there a significant increase in net protein synthesis. This result is compatible with a 6-fold stimulation in BGP synthesis since BGP synthesis accounts for only about 0.2% of total protein synthesis by osteosarcoma cells (27).
Several other clonal rat osteosarcoma cell lines were also tested for an effect of 1,25(OH)zD3 on BGP synthesis and secretion. The ROS 2/3 cell line, a low BGP producer (27)

DISCUSSION
The present study demonstrates that physiological 1,25(OH)2D3 concentrations dramatically and specifically increase the rate of biosynthesis of bone Gla protein by osteosarcoma cells. The kinetics of this effect suggest that 1,25(OH)2D:3 may act by activating the gene for BGP in the manner postulated for the 1,25(0H)zD~-dependent regulation of the cytosolic calcium binding protein of intestine. To our knowledge, BGP is the fiist bone protein whose synthesis rate increases in response to 1,25(OH)ZDS. Given the novelty of this regulation, it seems justified to review what is known about the bone Gla protein and to discuss the roles it may play in mediating the action of 1,25(OH)& on bone metabolism.
Although the bone Gla protein is a numerically abundant constituent of the extracellular bone matrix, the evidence to date indicates that it has no function in the formation of this matrix. Bones from warfarin-treated rabbits have less than 5% of the normal level of BGP yet are indistinguishable from bones of control animals in morphology, in mineral and protein content, and in strength (22). In addition, developmental studies show that BGP is virtually absent from the bones of neonatal rats (25) and so cannot play a role in initial bone formation. The present evidence that 1,25(OH)2D:3 increases BGP synthesis supports the conclusion that BGP does not function in bone matrix formation since 1,25(OH)zDn inhibits rather than stimulates the synthesis of matrix proteins (30).
Several lines of evidence support an informational function for BGP. The protein is only 49 residues long at the point of secretion from cells (27), a size compatible with hormones. It also has sequence features (15) such as two pairs of basic residues, which are proteolytic cleavage sites in the activation of hormones such as proinsulin (31), and a Pro-Lys unit, which is an effector site in many informational proteins (32). Finally, BGP is present in plasma as well as bone and has a uniquely high affinity for hydroxyapatite which could direct it to specific bone domains.
It seems to us most likely that the active informational species is plasma BGP and that its site of action is the exposed mineral surface of bone. Developmental studies indicate that plasma BGP may arise from new synthesis rather than from release of matrix BGP by resorption (25). In addition, calculations based on the affinity of BGP for hydroxyapatite show that the BGP found in bone could come from equilibrium binding of plasma BGP to bone hydroxyapatite (24). If BGP is synthesized and secreted into plasma in response to 1,25(OH)2Dij, what kind of biological response is likely? One clue is provided by the results of clinical studies which demonstrate that BGP is elevated in plasma from patients with metabolic bone diseases characterized by increased rates of bone turnover, such as Paget's disease of bone and primary and secondary hyperparathyroidism. The elevated plasma BGP levels may target exposed mineral surfaces for the increased levels of osteoblastic and osteoclastic activity observed in those metabolic bone diseases. Experiments are in progress to test these and other possible roles for BGP as mediator of the action of vitamin D on bone.