Parathyroid Hormone Regulates Transforming Growth Factor β1 and β2 Synthesis in Osteoblasts Via Divergent Signaling Pathways

Authors

  • Yanhong Wu,

    1. Nephrology Research Unit, Departments of Medicine, Biochemistry and Molecular Biology, Mayo Clinic/Foundation, Rochester, Minnesota, U.S.A.
    Search for more papers by this author
  • Rajiv Kumar

    Corresponding author
    1. Nephrology Research Unit, Departments of Medicine, Biochemistry and Molecular Biology, Mayo Clinic/Foundation, Rochester, Minnesota, U.S.A.
    • Dr. Rajiv Kumar, Mayo Clinic 200 First Street SW, 911A Guggenheim Building Rochester, MN 55905, U.S.A.
    Search for more papers by this author

Abstract

Parathyroid hormone 1–34 [PTH(1–34)] was shown to increase transforming growth factor β1 (TGF-β1) and TGF-β2 concentrations in supernatants of cultured human osteoblasts and to increase TGF-β1 and TGF-β2 messenger RNA (mRNA) concentrations and gene transcription in these cells. Because PTH(1–34) activates both protein kinase C (PKC) and protein kinase A (PKA) pathways in osteoblasts, we investigated the role of each kinase pathway in activation of TGF-β3 isoforms. PTH(29–32), which activates the PKC pathway in rat osteoblasts, increased TGF-β1 but not TGF-β2 concentrations in supernatants of osteoblasts. Phorbol myristate acetate (PMA), a PKC agonist, increased TGF-β1 but not TGF-β2 concentrations. Specific PKC antagonists safingol and Gö6976 attenuated PTH(1–34)-mediated increases in TGF-β1 but not TGF-β2 synthesis. PTH(1–31), which increases PKA activity in several cell culture systems, increased TGF-β2 but not TGF-β1 concentrations in human osteoblast supernatants. Forskolin, a PKA agonist, increased TGF-β2 but not TGF-β1 concentrations in supernatants of human osteoblasts. The PKA antagonist H-89 blunted PTH(1–34)-mediated increases in TGF-β2 but not TGF-β1 synthesis. Our results are consistent with the concept that PTH increases TGF-β1 expression and secretion by pathways that involve the PKC pathway, whereas it increases TGF-β2 expression and secretion via the PKA pathway. (J Bone Miner Res 2000;15:879–884)

INTRODUCTION

Bonedensity is regulated by changes in the rate of bone formation or bone resorption. These processes are regulated by systemic hormones such as parathyroid hormone (PTH), 1α,25-dihydroxyvitamin D3 [1α,25(OH)2D3], and estradiol, by local bone cytokines, prostaglandin E2, and growth factors such as transforming growth factor β (TGF-β) and insulin-like growth factor.(1–11) The regulation of local growth factors and cytokines in bone and bone cells by hormones is an important mechanism by which bone turnover is regulated. PTH, 1α,25(OH)2D3, and estrogens modulate the concentrations of cytokines, prostaglandins, and growth factors in bone.(11–21) However, the cellular mechanisms by which calciotropic systemic hormones control growth factor synthesis, have not been examined systematically.

TGF-β is one of the most abundant growth factors secreted by bone cells, and it affects bone growth and development, the proliferation and differentiation of osteoblasts, and the activity of osteoclasts in bone.(17,18,22–24) Some information is available on the mechanisms by which estrogens and 1α,25(OH)2D3 increase the synthesis of TGF-β (15,16,25) For example, vitamin D response elements have been characterized in the promoter of the TGF-β2 gene that mediate the effects of 1α,25(OH)2D3 on TGF-β2 gene transcription.(16) Similarly, novel estrogen response elements have been detected in the TGF-β3 promoter.(25) Although it has been shown that PTH alters TGF-β expression in bone, information concerning the mechanisms by which PTH controls TGF-β synthesis is not available.(14) Thus, it is not known as to which isoforms of TGF-β are altered by PTH. The signaling pathways by which PTH modulates TGF-β synthesis have not been examined. Because PTH increases both protein kinase A (PKA) and protein kinase C (PKC) activities in osteoblasts, either one or both of the these kinase pathways could potentially be responsible for the effects of PTH on TGF-β secretion and synthesis. We now report data that are consistent with the concept that PTH-mediated changes in TGF-β1 synthesis in human osteoblasts are mediated by the PKC pathway, whereas changes in TGF-β2 synthesis are mediated by the PKA pathway. The PTH-induced changes in TGF-β expression are caused by, in part, changes in the rate of transcription of the TGF-β genes.

MATERIALS AND METHODS

General

Human PTH(1–34), forskolin, 3-isobutyl-methylxanthine (IBMX), and H-89 [(N-[2-(p-bromocinnamylamino)ethyl]-5 isoquinoline sulfonamide)] were obtained from Sigma (St. Louis, MO, U.S.A.). Phorbol myristate acetate (PMA), safingol, Gö6976, and H-89 dihydrochloride were obtained from Calbiochem (San Diego, CA, U.S.A.). Human PTH(1–31) and PTH(29–32) were obtained from Bachem (Torrance, CA, U.S.A.).

Cell culture

Human fetal osteoblast cells (hFOB) were obtained from Dr. T.C Spelsberg, and maintained in Dulbecco's modified Eagle medium (DMEM)/F12 medium supplemented with 10% fetal bovine serum (FBS) or 1% ITS + (Becton Dickinson, Bedford, MA, U.S.A.; ITS + is an aqueous solution containing insulin, transferrin [12.5 mg], selenous acid [12.5 μg], bovine serum albumin [BSA; 2.5G], and linoleic acid [10.7 mg]) and 300 μg/ml G418.(15) Cells were grown at 37°C for 48 h to ∼80% confluence in DMEM/F12, 10% FBS. Serum-containing medium was removed and cells were grown in DMEM/F12/ITS + for another 4 h. Fresh DMEM/F12/ITS + with PTH peptides was then added to the cells for times indicated in the figures. Antagonists were added to the cells 1 h before addition of peptides and were maintained in the medium throughout the rest of the experiment. Supernatant medium was collected for measurement of TGF-β1 and TGF- β 2 with Quantikine Kits (R&D Systems, Minneapolis, MN, U.S.A.).

RNA-based polymerase chain reaction for quantitation of TGF-β1 and TGF-β2 messenger RNA

Human fetal osteoblasts were treated with 10–8 M PTH(1–34) and were harvested at 8, 16, 24 h, respectively. Total RNA was isolated and equal amounts of total RNA (1 μg) were reverse-transcribed into complementary DNA (cDNA) using an oligo(dT)16 primer and avian myeloblastosis virus (AMV) reverse transcriptase.(15) The RNA-cDNA duplex was amplified with Taq polymerase for 30 cycles with oligonucleotide primer pairs specific for human TGF-β1 and TGF-β2. Human glyceraldehyde 3-phosphate dehydrogenase was amplified under the same conditions as an internal control. Primer pairs used for amplification and the predicted sizes of the products (in parentheses) were TGF-β1 (767 base pairs [bp]):

  • forward primer, 5′-CTATCGACATGGAGCTGGTGAAG-3′; reverse primer, 5′-CGTGGAGCTGAAGCAATAGTTGG-3′ TGF-β2 (484 bp):

  • forward primer, 5′-CTTACTCGCCAAAGTCAGGGTTC-3′; reverse primer, 5′-GCTGTTGTAGATGGAAATCACCTCC-3′ GPDH (984 bp):

  • forward primer, 5′-TGAAGGTCGGAGTCAACGGATTT-GGT-3′;

  • reverse primer, 5′-CATGTGGGCCATGAGGTCCACCAC-3′

PCR products were electrophoresed on agarose gels and the amount of product was quantitated using an image analysis system as described earlier and available from the National Institutes of Health (URL for Image Program: http://rsb.info.nih.gov/nih-image/).(15) Densities of TGF- β 1 and TGF- β 2 messenger RNA (mRNA) bands were normalized against those for glyceraldehyde 3-phosphate.

Analysis of transcription

Nuclear run-on studies were performed with nuclei from PTH (10–8 M) treated or untreated cells (1 × 107 cells) collected at 0.5, 1, 2, 4, 8, 16, and 24 h respectively and stored at −70°C in glycerol storage buffer.(15) Thawed nuclei (100 μl) were added to a reaction mixture containing 10 × transcription buffer containing 0.3 mM digoxigenin (DIG)-11-uridine triphosphate (UTP) (described in Ref 15). Extension of RNA chains that had been initiated at the time of isolation was continued (run-on) at 37°C for 1 h. After DNAse I and proteinase K treatment, the solution was extracted with phenol/chloroform. RNA was precipitated using ammonium acetate and isopropyl alcohol and quantitated. For hybridization, 5 μg of linearized and denatured plasmid DNA containing TGF-β1, TGF-β2, or glyceraldehyde 3-phosphate dehydrogenase complementary DNAs (cDNAs) were loaded onto nitrocellulose filters. DNA was fixed to the filters by UV cross-linking. Filters were probed with equal amounts of labeled RNA probes (150 ng/ml). Hybridization, washing, and detection were performed as described in Ref. 15.

RESULTS

As shown in Fig. 1A, PTH(1–34) (a PTH analog that increases both PKC and PKA activities in osteoblasts), in a dose-dependent manner, increased TGF-β1 concentrations in supernatants of human osteoblasts maintained in culture.(26,27) Similarly, as shown in Fig. 1B, PTH(1–34), in a dose-dependent fashion, increased TGF-β2 concentrations in supernatants of human osteoblasts maintained in culture. The concentrations of TGF-β1 or TGF-β2 increased with time after the addition of PTH(1–34). As shown in Figs. 2A and 2B, PTH(1–34) increased TGF-β1 and β2 mRNA concentrations. Shown in the inset are the results of a representative experiment showing that TGF-β1 and TGF-β2 mRNA amounts increased whereas those for glyceraldehyde 3-phosphate did not. The transcription rates of the TGF-β1 and TGF-β2 genes also were increased (Fig. 3). The increase in transcription occurred within 1 h of the addition of PTH(1–34). Thus, PTH(1–34) causes an increase in both TGF-β1 and TGF-β2 peptide secretion and gene transcription.

Figure Fig. 1..

Effect of PTH(1–34) on (A) TGF-β1 and (B) TGF-β2 secretion by hFOB cell cultures. Cells were treated with various concentrations of PTH(1–34) and cell supernatants were collected at 24, 48, and 72 h, respectively. The concentrations of TGF-β1 and TGF-β2 in supernatants were measured by ELISA. Results are mean ± SE (n = 9). In the presence of PTH(1–34) concentrations of 10–10 M or greater, TGF-β1 concentrations at 48 h and 72 h and TGF-β2 concentrations at 24, 48, and 72 h were statistically greater than control (p < 0.05 or less).

Figure Fig. 2..

Effect of PTH(1–34) on (A) TGF-β1 mRNA and (B) TGF-β2 mRNA levels in hFOB cells. Human FOB cells were treated by PTH(1–34) (10–8 M); total RNA was collected at 0, 8, 16, and 24 h, respectively. Equal amounts of total RNA (1 μg) were used for reverse-transcription polymerase chain reaction (RT-PCR) with oligonucleotide primer pairs specific for human TGF-β1, TGF-β2, and glyceraldehyde 3-phosphate dehydrogenase. Results for TGF-β1 or TGF-β2 mRNA were normalized against mRNA concentrations for glyceraldehyde 3-phosphate dehydrogenase (representative example shown in inset). Results are mean ± SE from three experiments (n = 3). Values at 8, 16, and 24 h are significantly greater than control (p < 0.05).

Figure Fig. 3..

Effect of PTH(1–34) on TGF-β1 or TGF-β2 gene transcription assessed by nuclear run-on. The amounts of TGF-β1 or TGF-β2 transcripts were normalized against transcript amounts for glyceraldehyde 3-phosphate dehydrogenase.

To examine the signaling mechanisms by which PTH alters TGF-β1 and TGF-β2 concentrations, we treated human osteoblasts with PTH analogs that activate PKA or PKC or both pathways in osteoblasts.(26,27) As shown in Fig. 4 (top, left panel), 10–8 M PTH(1–34), which activates both PKC and PKA, caused a 2-fold increase in TGF-β1 concentrations in human osteoblast supernatants within 48 h. PTH(29–32) (10–8 M), a PTH agonist that only increases PKC activity in ROS 17/2.8 rat osteosarcoma cells, increased TGF-β1 concentrations equivalent to those seen in response to an equimolar amount of PTH(1–34).(27) PTH(1–31), 10–8 M, a PTH agonist that increases PKA activity in several cell types, did not increase TGF-β1 concentrations in supernatants of human osteoblasts.(26,27) We next treated human osteoblasts with PMA, 10–6 M, a PKC activator. PMA increased TGF-β1 concentrations (Fig. 4, bottom, left panel). When human bone cells were treated with PTH(1–34) in the presence of either safingol, a selective PKC inhibitor, or Gö6976, a selective PKC α-isoenzyme and β1-isoenzyme inhibitor, the activity of PTH(1–34) was significantly attenuated, suggesting that the PKC pathway plays a role in the secretion of TGF-β1 in human bone cells.(28) Of note, a PKA inhibitor H-89 (N-[2-(p-bromocinnamylamino)ethyl]-5 isoquinoline sulfonamide) did not block increases in TGF-β1 concentrations after treatment of osteoblasts with PTH(1–34) (data not shown).(29) An adenylyl cyclase stimulator, forskolin, likewise had no effects on TGF-β1 secretion (data not shown).

Figure Fig. 4..

(Top, Left Panel) Effect of PTH analogs on TGF-β1 secretion in hFOB cell cultures. The hFOB cells were treated with 10–8 M PTH analogs, including PTH(1–34), PTH(1–31), and PTH(29–32) for 48 h. (Bottom, Left Panel) Effect of PKC activator, PMA (10–6 M), PKC inhibitor, safingol (10–6 M) + PTH(1–34) (10–8 M), and a selective PKC a-isoenzyme and β1-isoenzyme inhibitor, Gö6976 (10–6 M) + PTH(1–34) (10–8 M) on TGF-β1 secretion in hFOB cell culture supernatants. Cells were treated for 48 h. Cell culture supernatants were collected and ELISA was performed to measure TGF-β1 concentrations. Results are mean ± SE (n = 9). TGF-β1 concentrations in PTH(1–34), PTH(29–32), and PMA-treated cells were significantly greater than in control cells (p < 0.05 or less). No differences were noted between control and PTH(1–31)-treated cells. In cells treated with PTH(1–34) plus safingol or PTH(1–34) plus Gö6976, TGF-β1 concentrations were intermediate between those found in control cells and cells treated with PTH(1–34) and PTH(29–32), respectively. (Top, Right Panel) Effect of PTH analogs on TGF-β2 secretion in hFOB cell cultures. The hFOB cells were treated with 10–8 M PTH analogs, including PTH(1–34), PTH(1–31), and PTH(29–32) for 48 h. (Bottom, Right Panel) Effect of forskolin (10–5 M), PTH(1–34) (10–8M) + IBMX (50 μM), and the PKA inhibitor H-89 (10–5 M) + PTH(1–34) (10–8 M) on TGF-β2 secretion in hFOB cell culture supernatants. Cells were treated for 48 h. Cell culture supernatants were collected and ELISA was performed to measure TGF-β2 concentrations. Results are mean ± SE (n = 9). TGF-β2 concentrations in PTH(1–34), PTH(1–31), forskolin, and PTH(1–34) + IBMX-treated cells were significantly greater than in control cells (p < 0.05 or less). No differences were noted between control and PTH(29–32)-treated cells. In cells treated with PTH(1–34) + H-89, TGF-β2 concentrations were intermediate between those found in control cells and cells treated with PTH(1–34) and PTH(1–31), respectively.

PTH analogs and various protein kinase antagonists (PKA) altered TGF-β2 secretion in a different manner compared with TGF-β1 secretion. As shown in Fig. 4 (top, right panel), PTH(1–34) increased TGF-β2 concentrations in supernatants of human osteoblasts maintained in culture. In contrast to the effects seen in the case of TGF-β1, PTH(1–31) (10–8 M), a PTH analog that stimulates PKA, also increased TGF-β2 concentrations in an amount equivalent to PTH(1–34). PTH(29–32) (10–8 M), a PTH agonist that increases only PKC activity in ROS 17/2.8 rat osteosarcoma cells, did not increase TGF-β2 concentrations. As shown in Fig. 4 (bottom, right panel), forskolin increased TGF-β2 concentrations. In the presence of IBMX, a phosphodiesterase inhibitor, PTH(1–34) increased TGF-β2 concentrations equivalent to those seen with PTH(1–34) alone. H-89 (10–5 M) attenuated the effects of PTH(1–34) with respect to TGF-β2 secretion in human osteoblasts. A PKC activator, PMA had no effect on TGF-β2 secretion and safingol or Gö6976 did not block PTH-mediated increases in TGF-β2 secretion (data not shown).

DISCUSSION

PTH has a significant impact on bone formation and resorption but the mechanism of action of PTH in bone is incompletely understood.(1,26) Previous studies have proposed that PTH alters local concentrations of growth factors in bone and that growth factors may act as mediators of the physiological effects of PTH in bone.(12–14) bib12 bib13 bib14The effects of PTH in osteoblasts are mediated by the activation of adenylyl cyclase and the generation of cyclic adenosine monophosphate (cAMP), by the activation of PKC after the generation of diacylglycerol and inositol tris-phosphate by phospholipase C and by a rise in intracellular calcium.(26,27,30) The growth factor TGF-β1 also alters PTH receptors and responsiveness in osteoblasts.(31)

The mechanisms by which PTH alters TGF-β concentrations are unclear. Our data clearly show that both TGF-β1 and TGF-β2 isoform secretion and synthesis in human osteoblasts are increased by PTH. The effects of the PTH on TGF-β1 and TGF-β2 peptide concentrations are associated with changes in TGF-β1 and TGF-β2 mRNA concentrations and changes in TGF-β1 and TGF-β2 gene transcription as assessed by nuclear run-on studies, suggesting that PTH effects are occurring at the level of transcription. This does not rule out the possibility of PTH-induced changes in mRNA stability, protein synthesis, or protein processing.

We show that different PTH analogs have different effects on the synthesis of TGF-β1 and TGF-β2. The complete PTH agonist, PTH(1–34), increases both TGF-β1 and TGF-β2 secretion and transcription. However, PTH(29–32), which increases only PKC activity in ROS 17/2.8 osteosarcoma cells, increases TGF-β1 but not TGF-β2 concentrations in human osteoblasts. PTH(1–31), which increases PKA activity in several cell types, on the other hand, does not increase TGF-β1 secretion but increases TGF-β2 secretion. The PKC activator PMA increases TGF-β1 concentrations but has no effects on TGF-β2 secretion. Consistent with these observations are attenuating effects of the PKC inhibitors saphingol and Gö6976 on PTH(1–34)-mediated increases in TGF-β1 synthesis. Taken together, these data are consistent with the idea that PTH increases TGF-β1 synthesis by pathways that involve PKC activation.

Our data also show that TGF-β2 secretion in human osteoblasts is altered predominantly by PTH(1–34) and PTH(1–31), which stimulate PKA, but not by PTH(29–32), which stimulates PKC alone. Forskolin, a PKA stimulator, increases TGF-β2 but does not alter TGF-β1 secretion in these cells. In agreement with this observation is the attenuation of PTH(1–34) stimulation of TGF-β2 secretion by H89, a PKA inhibitor. Collectively, the data show that the effects of PTH on TGF-β2 secretion are probably mediated via the PKA pathway. It is interesting to note that not all PTH(1–34)-mediated increases in TGF-β1 and TGF-β2 secretion are blocked by the PKC or PKA inhibitors used in our experiments. This suggests that other mechanisms might be operative with respect to stimulation of TGF-β isoforms by PTH.

How changes in PTH-mediated PKA and PKC activation alter TGF-β1 and TGF-β2 gene transcription is not known. The differences in signaling pathways used could relate to the structure of the TGF-β1 and TGF-β2 promoters. The TGF-β1 promoter has several AP-1 sites, which bind c-jun and c-fos as heteromeric complexes in a sequence-specific manner but no cAMP response elements (CREs).(32) AP-1 binding sequences have been shown to be targets for regulation by steroids and 1α,25(OH)2D3.(33,34) PTH has been shown to increase intracellular calcium, translocate PKC to the cell surface, stimulate PKC activity, and increase c-fos and c-jun expression in osteoblastic cell lines.(35–37) It is possible that PTH might influence TGF-β1 transcription via AP-1 sites in the TGF-β1 promoter. The TGF-β2 promoter has several CREs and AP-2 binding sites. Prior experiments have shown that TGF-β2 gene transcription is up-regulated by forskolin.(38) The cAMP-dependent PKA phosphorylates CRE-binding protein, increasing its transcriptional activity and the c-fos and 25-hydroxyvitamin D3-1α-hydroxylase genes are regulated by PTH through CREs in their promoters.(39,40) It is possible that the effect of PTH on TGF-β2 gene transcription is mediated through CREs in the TGF-β2 promoter. In conclusion, our results show that PTH regulates TGF-β1 and TGF-β2 synthesis in human osteoblasts by different mechanisms. Our data are thus consistent with the notion that PTH alters TGF-β1 secretion and synthesis through the PKC signaling pathway, whereas, it regulates TGF-β2 secretion and synthesis via the PKA/cAMP-dependent pathway.

Acknowledgements

This work was supported by the National Institutes of Health (NIH) grant DK25409 (R.K.).

Ancillary