0013-7227/91/1296-3313$03.00/0 Endocrinology Copyright P 1991 by The Endocrine Society
Vol. 129, No. 6 Printed in U.S.A.
Modulation of Transforming Growth Factor-/? Production in Normal Human Osteoblast-Like Cells by 17/?-Estradiol and Parathyroid Hormone MERRY JO OURSLER, CASE CORTESE, PHILIP KEETING, MARLYS A. ANDERSON, SUSAN K. BONDE, B. LAWRENCE RIGGS, AND THOMAS C. SPELSBERG Department of Biochemistry and Molecular Biology and the Endocrine Research Unit, Mayo Clinic and Mayo Foundation, Rochester, Minnesota 55905
ABSTRACT. Although our laboratory has reported that normal human osteoblast-like (hOB) cells contain estrogen receptors, we have failed to find major effects of 17/3-estradiol (E2) on modulation of proliferation of bone matrix protein production by hOB cells. Because the major effect of E2 in vivo is to decrease bone resorption and because transforming growth factor-/? (TGF-/3) has been reported to decrease osteoclast-mediated bone resorption, we have tested the hypothesis that the effect of E2 on osteoclast activity is, at least in part, indirectly mediated by enhancing production of TGF-j8 by osteoblasts. We therefore have extended our studies to examine possible TGF-/3 gene expression including the modulation of the release of TGF-/? by E2 in near homogenous populations of hOB cells. TGF-/3 protein production was measured using growth inhibition of CCL-64 cells and verified by blocking effects with anti-TGF-/3 antibodies.
P
OSTMENOPAUSAL osteoporosis is a well recognized national health problem that affects 1.5 million people annually (1). Although 17/3-estradiol (E2) has long been recognized as an effective treatment for osteoporosis, the exact role that this hormone plays in bone metabolism remains unclear (2). Histological evidence suggests that the influence of E 2 on bone metabolism in vivo is principally through decreasing bone resorption (3). Since recent studies have demonstrated E2 receptors in normal human osteoblast-like cell cultures and rat osteosarcoma cells, one possible mechanism for E2 action in bone could be osteoblast-mediated (4, 5). Our recent data suggest that normal human osteoblast-like cells do not respond to E2 by modulation of mitogenic activity or bone matrix protein production (6). We have therefore extended our studies of E2 effects on osteoblast activities to examine modulation of growth factor production by E2 in normal human osteoblast-like cells. Received June 22,1991. Address all correspondence and requests for reprints to: Dr. Thomas 0. Spelsberg, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905.
TGF-j8i messenger RNA (mRNA) steady state levels were assessed by northern blot analysis and quantitated by densitometric measurement using 18S ribosomal RNA as a reference. There was an E2 dose-dependent increase in TGF-/3 protein production within 24 h of challenge with E2. Northern blots from these cells demonstrated a dose-dependent increase in steady state mRNA levels of TGF-/3i within 6 h of treatment. PTH was also a potent stimulator of TGF-/3 protein and message levels in a dose-dependent manner. Interestingly, coincubation of equimolar concentrations of E2 and PTH (10~8 M) abrogated the stimulation of TGF-/3i mRNA and protein. Decreasing the relative concentration of PTH in this coincubation with E2 increased TGF-/3i mRNA and protein levels. These data support the fact that E2 modulates TGF-/3 production in osteoblasts. In this manner TGF-/3 may mediate E2 inhibition of osteoclast activity. (Endocrinology 129: 3313-3320, 1991)
Transforming growth factor type /3s (TGF-/3s) constitute a family of multifunctional proteins with a wide range of biological activities (reviewed in Ref. 7). TGF-/3 is a potent growth factor in bone, produced by osteoblasts as well as modulating osteoblast activities. In vivo, TGF/? appears to stimulate bone formation (8, 9), whereas in. vitro studies have supplied conflicting information on the actions of TGF-/3 on osteoblasts. In some studies of cultures of mixed cells from rat calvarial digests, TGF-/3 was reported to be nonmitogenic and caused a decrease in alkaline phosphatase activity (an indicator of osteoblast differentiation) (10). In contrast, other studies have observed that TGF-/3 is a potent mitogen and stimulator of alkaline phosphatase for these cells (11). The reasons for these differences remain obscure, but differences in culture conditions could be the cause. Three well characterized transformed cell lines, ROS 17/2.8, UMR 106, and MG-63, all respond to TGF-/3 with decreased mitogenic activity, increased collagen production, and increased alkaline phosphatase activity (12, 13). In contrast to this, normal human osteoblast-like cells and the
3313
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 July 2015. at 20:22 For personal use only. No other uses without permission. . All rights reserved.
3314
E2 AND PTH REGULATION OF TGF-0 IN HUMAN OSTEOBLASTS
murine nontransformed MC3T3 El cell line respond with a decrease in alkaline phosphatase activity (14). In addition to alkaline phosphatase and collagen, TGF-/3 also influences the production of several osteoblast proteins including osteonectin, osteopontin, and other matrix components (reviewed in Ref. 12). All of the evidence to date has demonstrated that TGF-/? is usually secreted by cells in a latent form which must be activated in order to influence cellular events (15, 16). Platelet TGF-/3 is released as a complex comprised of: 1) a mature disulfidebonded protein; 2) its cleaved leader peptide; and 3) a proteoglycan of 135,000 Daltons (17). Although platelets produce the most TGF-/3 per cell, bone is the largest tissue source of TGF-/? in the body (18). Bone-forming osteoblasts produce TGF-/3 as a significantly smaller latent form of 100,000 Daltons which lacks the 135,000Dalton proteoglycan (12). The significance of the differences in the latent complexes remains unclear. In light of the observations that osteoblasts produce TGF-0 in significant amounts (19), TGF-/? may function as an autocrine factor in osteoblasts. Bone-resorbing osteoclasts also appear to be influenced by TGF-/3, thus TGF-/3 may alter resorption as a paracrine factor in bone. Reports of the effects of TGF-/3 on osteoclasts suggest that TGF-/5 is involved both in the direct inhibition of osteoclast resorption activity (20, 21) as well as the inhibition of osteoclast precursor recruitment (9, 21, 22). The abundance of data concerning the roles of TGF-0 within the bone environment indicates that this factor plays a major role in bone metabolism. We have therefore examined normal human osteoblast-like cells for E2 modulation of TGF-j8 production. Materials and Methods Cell culture medium, fetal calf serum, crude bacterial collagenase, and culture flasks and dishes were purchased from GIBCO (Grand Island, NY). [3H]Thymidine and random hexamer DNA labeling kits were purchased from New England Nuclear (Danvers, MA). Magna 66 nylon membranes were purchased from MSI (Pittsburgh, PA). TGF-ft complementary DNA probe was the generous gift of Dr. Rid Derynck (Genentech, South San Francisco, CA). Purified human TGF-/3j was purchased from R&D Systems (Minneapolis, MN). P T H ^ was purchased from Peninsula Laboratories (Belmont, CA). XOMAT film was purchased from Eastman Kodak (Rochester, NY). Bone cell cultures Normal human osteoblast-like (hOB) cells were cultured from trabecular bone explants obtained at the time of orthopedic procedures on patients who had no evidence of metabolic bone disease. The bone cells were cultured by a procedure modified (5) from that developed by Robey and Termine (23). The hOB cells were grown in a phenol red-free, calcium-free medium approximately equivalent to a 1:1 (vol/vol) mixture of
Endo • 1991 Voll29«No6
Dulbecco's modified Eagle's medium and Ham's F-12 medium (culture medium) with 10% (vol/vol) fetal calf serum and penicillin and streptomycin (100 U/ml and 100 Mg/ml) added. These cells display a mature osteoblast phenotype (5, 24), and immunocytochemical analysis of the specific marker protein osteocalcin, a protein synthesized only by osteoblasts and odontoblasts (25), indicated that the cultures were at least 95% homogeneous (5, 24). All experiments were performed with cells subcultured at the end of their first passage. The medium was supplemented to 1 mM with calcium when cells were subcultured for experiments and was maintained at that concentration throughout all studies. TGF-p assay Hormone treatment of hOBs. After hOBs reached confluence, media was removed, and the cells rinsed twice with PBS. Test media (serum-free culture medium supplemented with 0.25% (wt/wt) BSA, 1 mM calcium, and 50 fig/ral vitamin C) and hormone or vehicle were added at the concentration and for the time period indicated in the figure legends. Sample preparation. Conditioned media was harvested and centrifuged at 10,000 X g for 5 min at 4 C to remove cell debris. Supernatants were divided into 4-ml aliquots and stored at -70 C until assayed. For activation of latent TGF-/J, media was dialyzed against McCoy's 5A, pH 1.5, for 18 h at 4 C, then against fresh McCoy's 5A, pH 7.4, for 18 h. Neutralizing antibody (R and D) at 10"8 M concentration was mixed with activated samples for 60 min at 37 C, microcentrifuged for 30 min at 4 C, and the resultant supernatant assayed for TGF-/3 activity. TGF-@ bioassay. DNA synthesis inhibition assay was carried out as described by Danielpour et al. (26) with minor modifications. Mink lung epithelial (CCL-64) cells were kindly provided by Dr. E. Leof (Vanderbilt University, Nashville, TN) and were maintained in McCoy's 5A supplemented with 10% (vol/vol) fetal bovine serum (culture media). Briefly, CCL-64 cells in logarithmic growth phase were trypsinized, resuspended in 10% (vol/vol) fetal bovine serum, pelleted, resuspended in culture media, plated at 2 X 104 cells/cm2 in 24-well plated (Falcon, Oxnard, CA) and cultured for 48 h at 37 C in 95% (vol/vol) air, 5% (vol/vol) CO2. Media was aspirated off and replaced with filter sterilized (0.22 micron pore size, Millex, Millipore Corp., Bedford, MA) 50% conditioned media in culture media or purified TGF-/3 in culture media. Plates were cultured for 18 h as above, followed by pulse labeling with 10 lA 100 jiCi/ml [3H]thymidine (25 Ci/mmol, Amersham Corp., Arlington Heights, IL) for 6 h. Media was removed and the cells fixed with 3 20-min applications of 4 C 10% (vol/vol) trichloroacetic acid. Samples were solubilized in 0.2% (vol/vol) NaOH with 40 Mg/ml salmon sperm DNA. Specificity was validated by neutralizing antibodies using anti-TGF-/? antibodies (see Sample preparation). Radioactivity was measured by liquid scintillation counting. Isolation and Northern blot analysis Total RNA was from cultured hOBs treated as described above for TGF-/3 protein production studies by the method of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 July 2015. at 20:22 For personal use only. No other uses without permission. . All rights reserved.
E2 AND PTH REGULATION OF TGF-/3 IN HUMAN OSTEOBLASTS Chomcynske and Sacci (27). Ten micrograms of each RNA were separated on a 1% (wt/vol) agarose-glyoxal gel and transferred to a nylon membrane (MSI). The RNA blots were hybridized for 18 h at 43 C with a complementary DNA to human TGF-/3 followed by an 18S ribosomal probe (relative sample loading efficiency) labeled with 32P by random primer extension with a multiprime kit (Amersham). The blots were washed twice with 15 mM NaCl/1.5 mM sodium citrate (pH 7), 0.1% (wt/wt) sodium dodecyl sulfate for 15 min at 60 C, and analyzed by autoradiography followed by densitometric analysis using a Shimadzu (Kyoto, Japan) Dual-Wavelength Flying Spot Scanning Densitometer. Statistical analysis The effect of treatment with each dose of steroid was compared with controls by the Student two-tailed paired t test. Values are expressed as means ± SD.
Results Effect of E
Vol. 129, No. 6 Printed in U.S.A.
Modulation of Transforming Growth Factor-/? Production in Normal Human Osteoblast-Like Cells by 17/?-Estradiol and Parathyroid Hormone MERRY JO OURSLER, CASE CORTESE, PHILIP KEETING, MARLYS A. ANDERSON, SUSAN K. BONDE, B. LAWRENCE RIGGS, AND THOMAS C. SPELSBERG Department of Biochemistry and Molecular Biology and the Endocrine Research Unit, Mayo Clinic and Mayo Foundation, Rochester, Minnesota 55905
ABSTRACT. Although our laboratory has reported that normal human osteoblast-like (hOB) cells contain estrogen receptors, we have failed to find major effects of 17/3-estradiol (E2) on modulation of proliferation of bone matrix protein production by hOB cells. Because the major effect of E2 in vivo is to decrease bone resorption and because transforming growth factor-/? (TGF-/3) has been reported to decrease osteoclast-mediated bone resorption, we have tested the hypothesis that the effect of E2 on osteoclast activity is, at least in part, indirectly mediated by enhancing production of TGF-j8 by osteoblasts. We therefore have extended our studies to examine possible TGF-/3 gene expression including the modulation of the release of TGF-/? by E2 in near homogenous populations of hOB cells. TGF-/3 protein production was measured using growth inhibition of CCL-64 cells and verified by blocking effects with anti-TGF-/3 antibodies.
P
OSTMENOPAUSAL osteoporosis is a well recognized national health problem that affects 1.5 million people annually (1). Although 17/3-estradiol (E2) has long been recognized as an effective treatment for osteoporosis, the exact role that this hormone plays in bone metabolism remains unclear (2). Histological evidence suggests that the influence of E 2 on bone metabolism in vivo is principally through decreasing bone resorption (3). Since recent studies have demonstrated E2 receptors in normal human osteoblast-like cell cultures and rat osteosarcoma cells, one possible mechanism for E2 action in bone could be osteoblast-mediated (4, 5). Our recent data suggest that normal human osteoblast-like cells do not respond to E2 by modulation of mitogenic activity or bone matrix protein production (6). We have therefore extended our studies of E2 effects on osteoblast activities to examine modulation of growth factor production by E2 in normal human osteoblast-like cells. Received June 22,1991. Address all correspondence and requests for reprints to: Dr. Thomas 0. Spelsberg, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905.
TGF-j8i messenger RNA (mRNA) steady state levels were assessed by northern blot analysis and quantitated by densitometric measurement using 18S ribosomal RNA as a reference. There was an E2 dose-dependent increase in TGF-/3 protein production within 24 h of challenge with E2. Northern blots from these cells demonstrated a dose-dependent increase in steady state mRNA levels of TGF-/3i within 6 h of treatment. PTH was also a potent stimulator of TGF-/3 protein and message levels in a dose-dependent manner. Interestingly, coincubation of equimolar concentrations of E2 and PTH (10~8 M) abrogated the stimulation of TGF-/3i mRNA and protein. Decreasing the relative concentration of PTH in this coincubation with E2 increased TGF-/3i mRNA and protein levels. These data support the fact that E2 modulates TGF-/3 production in osteoblasts. In this manner TGF-/3 may mediate E2 inhibition of osteoclast activity. (Endocrinology 129: 3313-3320, 1991)
Transforming growth factor type /3s (TGF-/3s) constitute a family of multifunctional proteins with a wide range of biological activities (reviewed in Ref. 7). TGF-/3 is a potent growth factor in bone, produced by osteoblasts as well as modulating osteoblast activities. In vivo, TGF/? appears to stimulate bone formation (8, 9), whereas in. vitro studies have supplied conflicting information on the actions of TGF-/3 on osteoblasts. In some studies of cultures of mixed cells from rat calvarial digests, TGF-/3 was reported to be nonmitogenic and caused a decrease in alkaline phosphatase activity (an indicator of osteoblast differentiation) (10). In contrast, other studies have observed that TGF-/3 is a potent mitogen and stimulator of alkaline phosphatase for these cells (11). The reasons for these differences remain obscure, but differences in culture conditions could be the cause. Three well characterized transformed cell lines, ROS 17/2.8, UMR 106, and MG-63, all respond to TGF-/3 with decreased mitogenic activity, increased collagen production, and increased alkaline phosphatase activity (12, 13). In contrast to this, normal human osteoblast-like cells and the
3313
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 July 2015. at 20:22 For personal use only. No other uses without permission. . All rights reserved.
3314
E2 AND PTH REGULATION OF TGF-0 IN HUMAN OSTEOBLASTS
murine nontransformed MC3T3 El cell line respond with a decrease in alkaline phosphatase activity (14). In addition to alkaline phosphatase and collagen, TGF-/3 also influences the production of several osteoblast proteins including osteonectin, osteopontin, and other matrix components (reviewed in Ref. 12). All of the evidence to date has demonstrated that TGF-/? is usually secreted by cells in a latent form which must be activated in order to influence cellular events (15, 16). Platelet TGF-/3 is released as a complex comprised of: 1) a mature disulfidebonded protein; 2) its cleaved leader peptide; and 3) a proteoglycan of 135,000 Daltons (17). Although platelets produce the most TGF-/3 per cell, bone is the largest tissue source of TGF-/? in the body (18). Bone-forming osteoblasts produce TGF-/3 as a significantly smaller latent form of 100,000 Daltons which lacks the 135,000Dalton proteoglycan (12). The significance of the differences in the latent complexes remains unclear. In light of the observations that osteoblasts produce TGF-0 in significant amounts (19), TGF-/? may function as an autocrine factor in osteoblasts. Bone-resorbing osteoclasts also appear to be influenced by TGF-/3, thus TGF-/3 may alter resorption as a paracrine factor in bone. Reports of the effects of TGF-/3 on osteoclasts suggest that TGF-/5 is involved both in the direct inhibition of osteoclast resorption activity (20, 21) as well as the inhibition of osteoclast precursor recruitment (9, 21, 22). The abundance of data concerning the roles of TGF-0 within the bone environment indicates that this factor plays a major role in bone metabolism. We have therefore examined normal human osteoblast-like cells for E2 modulation of TGF-j8 production. Materials and Methods Cell culture medium, fetal calf serum, crude bacterial collagenase, and culture flasks and dishes were purchased from GIBCO (Grand Island, NY). [3H]Thymidine and random hexamer DNA labeling kits were purchased from New England Nuclear (Danvers, MA). Magna 66 nylon membranes were purchased from MSI (Pittsburgh, PA). TGF-ft complementary DNA probe was the generous gift of Dr. Rid Derynck (Genentech, South San Francisco, CA). Purified human TGF-/3j was purchased from R&D Systems (Minneapolis, MN). P T H ^ was purchased from Peninsula Laboratories (Belmont, CA). XOMAT film was purchased from Eastman Kodak (Rochester, NY). Bone cell cultures Normal human osteoblast-like (hOB) cells were cultured from trabecular bone explants obtained at the time of orthopedic procedures on patients who had no evidence of metabolic bone disease. The bone cells were cultured by a procedure modified (5) from that developed by Robey and Termine (23). The hOB cells were grown in a phenol red-free, calcium-free medium approximately equivalent to a 1:1 (vol/vol) mixture of
Endo • 1991 Voll29«No6
Dulbecco's modified Eagle's medium and Ham's F-12 medium (culture medium) with 10% (vol/vol) fetal calf serum and penicillin and streptomycin (100 U/ml and 100 Mg/ml) added. These cells display a mature osteoblast phenotype (5, 24), and immunocytochemical analysis of the specific marker protein osteocalcin, a protein synthesized only by osteoblasts and odontoblasts (25), indicated that the cultures were at least 95% homogeneous (5, 24). All experiments were performed with cells subcultured at the end of their first passage. The medium was supplemented to 1 mM with calcium when cells were subcultured for experiments and was maintained at that concentration throughout all studies. TGF-p assay Hormone treatment of hOBs. After hOBs reached confluence, media was removed, and the cells rinsed twice with PBS. Test media (serum-free culture medium supplemented with 0.25% (wt/wt) BSA, 1 mM calcium, and 50 fig/ral vitamin C) and hormone or vehicle were added at the concentration and for the time period indicated in the figure legends. Sample preparation. Conditioned media was harvested and centrifuged at 10,000 X g for 5 min at 4 C to remove cell debris. Supernatants were divided into 4-ml aliquots and stored at -70 C until assayed. For activation of latent TGF-/J, media was dialyzed against McCoy's 5A, pH 1.5, for 18 h at 4 C, then against fresh McCoy's 5A, pH 7.4, for 18 h. Neutralizing antibody (R and D) at 10"8 M concentration was mixed with activated samples for 60 min at 37 C, microcentrifuged for 30 min at 4 C, and the resultant supernatant assayed for TGF-/3 activity. TGF-@ bioassay. DNA synthesis inhibition assay was carried out as described by Danielpour et al. (26) with minor modifications. Mink lung epithelial (CCL-64) cells were kindly provided by Dr. E. Leof (Vanderbilt University, Nashville, TN) and were maintained in McCoy's 5A supplemented with 10% (vol/vol) fetal bovine serum (culture media). Briefly, CCL-64 cells in logarithmic growth phase were trypsinized, resuspended in 10% (vol/vol) fetal bovine serum, pelleted, resuspended in culture media, plated at 2 X 104 cells/cm2 in 24-well plated (Falcon, Oxnard, CA) and cultured for 48 h at 37 C in 95% (vol/vol) air, 5% (vol/vol) CO2. Media was aspirated off and replaced with filter sterilized (0.22 micron pore size, Millex, Millipore Corp., Bedford, MA) 50% conditioned media in culture media or purified TGF-/3 in culture media. Plates were cultured for 18 h as above, followed by pulse labeling with 10 lA 100 jiCi/ml [3H]thymidine (25 Ci/mmol, Amersham Corp., Arlington Heights, IL) for 6 h. Media was removed and the cells fixed with 3 20-min applications of 4 C 10% (vol/vol) trichloroacetic acid. Samples were solubilized in 0.2% (vol/vol) NaOH with 40 Mg/ml salmon sperm DNA. Specificity was validated by neutralizing antibodies using anti-TGF-/? antibodies (see Sample preparation). Radioactivity was measured by liquid scintillation counting. Isolation and Northern blot analysis Total RNA was from cultured hOBs treated as described above for TGF-/3 protein production studies by the method of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 July 2015. at 20:22 For personal use only. No other uses without permission. . All rights reserved.
E2 AND PTH REGULATION OF TGF-/3 IN HUMAN OSTEOBLASTS Chomcynske and Sacci (27). Ten micrograms of each RNA were separated on a 1% (wt/vol) agarose-glyoxal gel and transferred to a nylon membrane (MSI). The RNA blots were hybridized for 18 h at 43 C with a complementary DNA to human TGF-/3 followed by an 18S ribosomal probe (relative sample loading efficiency) labeled with 32P by random primer extension with a multiprime kit (Amersham). The blots were washed twice with 15 mM NaCl/1.5 mM sodium citrate (pH 7), 0.1% (wt/wt) sodium dodecyl sulfate for 15 min at 60 C, and analyzed by autoradiography followed by densitometric analysis using a Shimadzu (Kyoto, Japan) Dual-Wavelength Flying Spot Scanning Densitometer. Statistical analysis The effect of treatment with each dose of steroid was compared with controls by the Student two-tailed paired t test. Values are expressed as means ± SD.
Results Effect of E
