Transforming growth factor-L1 regulation of resting zone ...Transforming growth factor-L1 regulation...

14
Transforming growth factor-L1 regulation of resting zone chondrocytes is mediated by two separate but interacting pathways V.L. Sylvia a , Z. Schwartz a;b;c , D.D. Dean a , B.D. Boyan a;b;d; * a Department of Orthopaedics, The University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, TX 78229-3900, USA b Department of Periodontics, The University of Texas Health Science Center at San Antonio, San Antonio, TX 78229-3900, USA c Department of Periodontics, Hebrew University Hadassah Faculty of Dental Medicine, Jerusalem 91010, Israel d Department of Biochemistry, The University of Texas Health Science Center at San Antonio, San Antonio, TX 78229-3900, USA Received 29 July 1999; received in revised form 3 February 2000; accepted 17 February 2000 Abstract Previous studies have shown that transforming growth factor-L1 (TGF-L1) stimulates protein kinase C (PKC) via a mechanism that is independent of phospholipase C or tyrosine kinase, but involves a pertussis toxin-sensitive G-protein. Maximal activation occurs at 12 h and requires new gene expression. To understand the signaling pathways involved, resting zone chondrocytes were incubated with TGF-L1 and PKC activity was inhibited with chelerythrine, staurosporine or H-7. [ 35 S]Sulfate incorporation was inhibited, indicating that PKC mediates the effects of TGF-L1 on matrix production. However, there was little, if any, effect on TGF-L1-dependent increases in [ 3 H]thymidine incorporation, and TGF-L1- stimulated alkaline phosphatase was unaffected, indicating that these responses to the growth factor are not regulated via PKC. TGF-L1 caused a dose-dependent increase in prostaglandin E 2 (PGE 2 ) production which was further increased by PKC inhibition. The increase was regulated by TGF-L1-dependent effects on phospholipase A 2 (PLA 2 ). Activation of PLA 2 inhibited TGF-L1 effects on PKC, and inhibition of PLA 2 activated TGF-L1-dependent PKC. Exogenous arachidonic acid also inhibited TGF-L1-dependent increases in PKC. The effects of TGF-L1 on PKC involve genomic mechanisms, but not regulation of existing membrane-associated enzyme, since no direct effect of the growth factor on plasma membrane or matrix vesicle PKC was observed. These results support the hypothesis that TGF-L1 modulates its effects on matrix production through PKC, but its effects on alkaline phosphatase are mediated by production of PGE 2 and protein kinase A (PKA). Inhibition of PKA also decreases TGF-L1-dependent proliferation. We have previously shown that PGE 2 stimulates alkaline phosphatase through its EP2 receptor, whereas EP1 signaling causes a decrease in PKC. Thus, there is cross-talk between the two pathways. ß 2000 Elsevier Science B.V. All rights reserved. Keywords : Chondrocyte culture ; Transforming growth factor-L1; 24,25-(OH) 2 D 3 ; Protein kinase C; Alkaline phosphatase; Signal transduction 1. Introduction Transforming growth factor L-1 (TGF-L1) is an important regulator of chondrocyte proliferation and di¡erentiation that was originally puri¢ed from demineralized bone based on its ability to stimulate 0167-4889 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved. PII:S0167-4889(00)00030-6 * Corresponding author. Department of Orthopaedics, The University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, TX 78229-3900, USA. Fax : +1-210-567-6295 ; E-mail : [email protected] Biochimica et Biophysica Acta 1496 (2000) 311^324 www.elsevier.com/locate/bba

Transcript of Transforming growth factor-L1 regulation of resting zone ...Transforming growth factor-L1 regulation...

Page 1: Transforming growth factor-L1 regulation of resting zone ...Transforming growth factor-L1 regulation of resting zone chondrocytes is mediated by two separate but interacting pathways

Transforming growth factor-L1 regulation of resting zone chondrocytes ismediated by two separate but interacting pathways

V.L. Sylvia a, Z. Schwartz a;b;c, D.D. Dean a, B.D. Boyan a;b;d;*a Department of Orthopaedics, The University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive,

San Antonio, TX 78229-3900, USAb Department of Periodontics, The University of Texas Health Science Center at San Antonio, San Antonio, TX 78229-3900, USA

c Department of Periodontics, Hebrew University Hadassah Faculty of Dental Medicine, Jerusalem 91010, Israeld Department of Biochemistry, The University of Texas Health Science Center at San Antonio, San Antonio, TX 78229-3900, USA

Received 29 July 1999; received in revised form 3 February 2000; accepted 17 February 2000

Abstract

Previous studies have shown that transforming growth factor-L1 (TGF-L1) stimulates protein kinase C (PKC) via amechanism that is independent of phospholipase C or tyrosine kinase, but involves a pertussis toxin-sensitive G-protein.Maximal activation occurs at 12 h and requires new gene expression. To understand the signaling pathways involved, restingzone chondrocytes were incubated with TGF-L1 and PKC activity was inhibited with chelerythrine, staurosporine or H-7.[35S]Sulfate incorporation was inhibited, indicating that PKC mediates the effects of TGF-L1 on matrix production.However, there was little, if any, effect on TGF-L1-dependent increases in [3H]thymidine incorporation, and TGF-L1-stimulated alkaline phosphatase was unaffected, indicating that these responses to the growth factor are not regulated viaPKC. TGF-L1 caused a dose-dependent increase in prostaglandin E2 (PGE2) production which was further increased byPKC inhibition. The increase was regulated by TGF-L1-dependent effects on phospholipase A2 (PLA2). Activation of PLA2

inhibited TGF-L1 effects on PKC, and inhibition of PLA2 activated TGF-L1-dependent PKC. Exogenous arachidonic acidalso inhibited TGF-L1-dependent increases in PKC. The effects of TGF-L1 on PKC involve genomic mechanisms, but notregulation of existing membrane-associated enzyme, since no direct effect of the growth factor on plasma membrane ormatrix vesicle PKC was observed. These results support the hypothesis that TGF-L1 modulates its effects on matrixproduction through PKC, but its effects on alkaline phosphatase are mediated by production of PGE2 and protein kinase A(PKA). Inhibition of PKA also decreases TGF-L1-dependent proliferation. We have previously shown that PGE2 stimulatesalkaline phosphatase through its EP2 receptor, whereas EP1 signaling causes a decrease in PKC. Thus, there is cross-talkbetween the two pathways. ß 2000 Elsevier Science B.V. All rights reserved.

Keywords: Chondrocyte culture; Transforming growth factor-L1; 24,25-(OH)2D3 ; Protein kinase C; Alkaline phosphatase;Signal transduction

1. Introduction

Transforming growth factor L-1 (TGF-L1) is animportant regulator of chondrocyte proliferationand di¡erentiation that was originally puri¢ed fromdemineralized bone based on its ability to stimulate

0167-4889 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved.PII: S 0 1 6 7 - 4 8 8 9 ( 0 0 ) 0 0 0 3 0 - 6

* Corresponding author. Department of Orthopaedics, TheUniversity of Texas Health Science Center at San Antonio,7703 Floyd Curl Drive, San Antonio, TX 78229-3900, USA.Fax: +1-210-567-6295; E-mail : [email protected]

BBAMCR 14618 5-4-00

Biochimica et Biophysica Acta 1496 (2000) 311^324www.elsevier.com/locate/bba

Page 2: Transforming growth factor-L1 regulation of resting zone ...Transforming growth factor-L1 regulation of resting zone chondrocytes is mediated by two separate but interacting pathways

the di¡erentiation of neonatal rat muscle cells intochondrocytes [1^3]. TGF-L1 also regulates commit-ted chondrocytes, particularly those in the endochon-dral pathway [4,5]. Both alkaline phosphatase specif-ic activity and proteoglycan production [5,6] areincreased by this growth factor. In addition to itsdirect e¡ects on cells, TGF-L1 works in conjunctionwith other cytokines to modulate their e¡ects, furtheramplifying its potential biological functions [7].

Cells in the resting zone of costochondral growthplate cartilage are particularly sensitive to TGF-L1[5,8,9]. These cells exhibit maximal responses todoses of 0.22 ng/ml, whereas higher concentrationsare required for osteoblasts to exhibit the same e¡ect[10,11]. TGF-L1 regulates proliferation of restingzone chondrocytes in a time-dependent manner,causing increased proliferation following a 24 h ex-posure, but decreased proliferation following longerexposure [5]. Alkaline phosphatase is stimulated in abiphasic manner [5], indicating increased di¡erentia-tion. Indeed, exposure of resting zone chondrocytesto TGF-L1 for more than 96 h causes the cells toassume a hypertrophic chondrocyte phenotype [12].The e¡ects of TGF-L1 are synergistic with those of24,25-(OH)2D3, further supporting the role of thisgrowth factor in regulating the di¡erentiation andendochondral maturation of these cells [13].

Resting zone chondrocytes synthesize TGF-L1 inlatent form [4,14] and store it in their extracellularmatrix as a 290 kDa complex consisting of latentTGF-L1, latent TGF-L1 binding protein-1 and thelatency-associated peptide [15]. Extracellular matrixvesicles produced by these cells can activate latentTGF-L1 when they are exposed to 1,25-(OH)2D3

[14]. The interrelationship of TGF-L1 action and vi-tamin D metabolites is also demonstrated by the factthat TGF-L1 causes resting zone cells to produceincreased 1,25-(OH)2D3 within 1 h and increased24,25-(OH)2D3 at 24 h [16], which is correlatedwith TGF-L1-dependent downregulation of the 1K-hydroxylase and upregulation of the 24-hydroxylasein these cells [9].

TGF-L1 causes increased protein kinase C (PKC)activity in resting zone cells [8,17], suggesting thatPKC may be involved in mediating its actions onthese cells. To determine if this was the case, wetreated the cells with TGF-L1 and examined the ef-fects of PKC inhibition on [3H]thymidine incorpora-

tion, alkaline phosphatase speci¢c activity and pro-teoglycan sulfation. Since prostaglandin E2 (PGE2) isalso known to modulate PKC activity in these cells[18], we also examined if TGF-L1 exerts its e¡ectsthrough prostaglandin production. The role of pro-tein kinase A (PKA) was examined as a function ofcell proliferation. We then examined the mechanismsby which TGF-L1 stimulates PKC, focusing on therole of phospholipase A2 (PLA2) and the product ofits action, arachidonic acid. Finally, we determined ifTGF-L1 exerts a direct action on membrane PKC.

2. Materials and methods

2.1. Reagents

The following chemicals were purchased from Sig-ma Chemical (St. Louis, MO, USA): palmitic acid,indomethacin (a general cyclooxygenase inhibitor),quinacrine (PLA2 inhibitor) and reagents for the al-kaline phosphatase assay. The following chemicalswere purchased from Calbiochem (San Diego, CA,USA): 1,2-dioctanoyl-sn-glycerol (DOG), cheleryth-rine (PKC inhibitor), staurosporine, H-7, H-8, arachi-donic acid and U73122 (phospholipase C inhibitor).The following chemicals were purchased from BIO-MOL Research Laboratories (Plymouth Meeting,PA, USA): selective PLA2 inhibitors oleyloxyethyl-phosphorylcholine (OEPC) and arachidonyltri£uoro-methyl-ketone (AACOCF3), and selective PLA2 ac-tivators mastoparan and melittin. PKC assayreagents and Dulbecco's modi¢ed Eagle medium(DMEM) were obtained from Gibco BRL (Gaithers-burg, MD, USA). The protein content of each sam-ple was determined using the bicinchoninic acid(BCA) protein assay reagent [19] obtained fromPierce Chemical (Rockford, IL, USA). Recombinanthuman TGF-L1 was obtained from RpD Systems(Minneapolis, MN, USA). Fetal bovine serum(FBS) was purchased from Atlanta Biologicals (Nor-cross, GA, USA). The PGE2 radioimmunoassay kit,[3H]thymidine, [35S]sulfate and [32P]ATP were ob-tained from NEN DuPont (Boston, MA, USA).

2.2. Chondrocyte cultures

The rat costochondral chondrocyte culture system

BBAMCR 14618 5-4-00

V.L. Sylvia et al. / Biochimica et Biophysica Acta 1496 (2000) 311^324312

Page 3: Transforming growth factor-L1 regulation of resting zone ...Transforming growth factor-L1 regulation of resting zone chondrocytes is mediated by two separate but interacting pathways

used in this study has been described in detail pre-viously [20]. Cells were derived from the resting zoneand growth zone of costochondral cartilage from 125g male Sprague-Dawley rats (Harlan, Indianapolis,IN, USA) and cultured in DMEM containing 10%FBS, 50 Wg/ml vitamin C and antibiotics in an at-mosphere of 5% CO2 and 100% humidity at 37³C.Fourth passage cells were used for all experimentssince prior studies have shown that these cells retaintheir chondrogenic phenotype, including synthesis oftype II collagen [21], ability to form cartilage noduleswhen implanted into nude mouse thigh muscle [22],and di¡erential responsiveness to vitamin D metab-olites and a number of other factors [5,14,20,21,23^28].

When the cells reached con£uence, they weretreated with TGF-L1 for 24 h to assess physiologicalresponses and for 12 h to measure PKC speci¢c ac-tivity. Recombinant human TGF-L1 was dissolved inphosphate-bu¡ered saline and diluted to the appro-priate concentration with DMEM before addition tothe cultures. Control cultures contained vehicle orinhibitors or activators.

2.3. Role of PKC in physiologic response of restingzone cells to TGF-L1

To determine whether the response of resting zonecells to TGF-L1 involves PKC, con£uent fourth pas-sage cultures were incubated with TGF-L1 þ PKCinhibitors as described below. We used chelerythrineas an inhibitor of PKC for these studies because ofits relative speci¢city for the enzyme. Chelerythrinehas an IC50 of 0.05 WM for PKC, whereas its IC50 forPKA is s 50 WM [29]. H-7, H-8 and staurosporinealso inhibit a casein kinase and a myosin light chainkinase [30]. Staurosporine has also been reported toinhibit certain tyrosine kinases as well [31]. PKC in-hibitor peptide 19-36, which is speci¢c for PKC,could not be used for these studies because it is toobig to penetrate the cell membrane [32]. We usedstaurosporine to con¢rm our observations with chel-erythrine, since staurosporine is an e¡ective inhibitorof most isoforms of PKC, albeit not as speci¢c aschelerythrine. In addition, cultures were treated withH-7 [33].

2.3.1. [3H]Thymidine incorporationDNA synthesis was assessed by measuring

[3H]thymidine incorporation into trichloroaceticacid (TCA) insoluble cell precipitates as describedpreviously [21]. Quiescence was induced by incubat-ing con£uent cultures for 48 h in DMEM containing1% FBS. The medium was then replaced for 24 hwith DMEM containing 1% FBS, vehicle þ TGF-L1and the appropriate concentration of PKC inhibitor.PKC activity in cultures treated with 0.22 ng/mlTGF-L1 was inhibited with either 0.01 or 0.10 WMchelerythrine [29], or 1 or 5 WM staurosporine [34]for 24 h. In addition, cultures were treated with 0.88ng/ml TGF-L1 þ PKC inhibitor H-7 (1, 5 or 10 WM)[33]. Two hours prior to harvest, [3H]thymidine wasadded. Radioactivity in TCA-precipitable materialwas measured by liquid scintillation spectroscopy.

2.3.2. Alkaline phosphatase activityAlkaline phosphatase (orthophosphoric monoester

phosphohydrolase, alkaline (EC 3.1.3.1)) was mea-sured as a function of release of para-nitrophenolfrom para-nitrophenylphosphate at pH 10.2 [35]. En-zyme activity was assayed in lysates of the cell layer,as well as in isolated matrix vesicle and plasma mem-branes as previously described [36]. Cultures weretreated with 0.11, 0.22 or 0.88 ng/ml TGF-L1 þ 1 or10 WM chelerythrine.

2.3.3. [35S]Sulfate incorporationProteoglycan synthesis was assessed by measuring

[35S]sulfate incorporation by con£uent cultures offourth passage resting zone chondrocytes using amodi¢cation of the method of O'Keefe et al. [37].In prior studies, we found that the amount of radio-labeled proteoglycan secreted by the chondrocytesinto the medium was less than 15% of the total ra-diolabeled proteoglycan (medium and cell layer) syn-thesized [38,39]. Because of this, we only examinedthe e¡ects of hormone and growth factor treatmenton [35S]sulfate incorporation in the cell layer. At con-£uence, fresh medium containing vehicle þ 0.11, 0.22or 0.88 WM TGF-L1 þ 1 or 10 WM chelerythrine wasadded to the cells and the incubation continued foran additional 24 h. Four hours prior to harvest, 50 WlDMEM containing 18 WCi/ml [35S]sulfate and 0.814mM carrier sulfate was added to each culture. Atharvest, the conditioned media were removed, the

BBAMCR 14618 5-4-00

V.L. Sylvia et al. / Biochimica et Biophysica Acta 1496 (2000) 311^324 313

Page 4: Transforming growth factor-L1 regulation of resting zone ...Transforming growth factor-L1 regulation of resting zone chondrocytes is mediated by two separate but interacting pathways

cell layers (cells and matrix) collected and theamount of [35S]sulfate incorporated determined byliquid scintillation spectrometry. The protein contentwas determined using the BCA protein assay and thedata expressed as disintegrations per minute per mgprotein in the cell layer.

2.3.4. PGE2 productionPGE2 production by resting zone chondrocyte cul-

tures treated with TGF-L1 was assessed by radioim-munoassay. Cultures were treated for 24 h with 0.11or 0.22 ng/ml TGF-L1 in the presence or absence of0.01, 0.1 or 1 WM chelerythrine or H-7 and thenassayed for PGE2 production using a radioimmuno-assay kit (NEN DuPont, Boston, MA, USA).

2.4. Role of PKA

To assess the role of PKA in cell response to TGF-L1, the resting zone chondrocytes were incubatedwith 0.88 ng/ml TGF-L1 þ 1, 5 or 10 WM H-8 [33].[3H]Thymidine incorporation was measured as de-scribed above.

2.5. Role of PLA2 in TGF-L1-dependent regulation ofPKC

2.5.1. PKC speci¢c activityCell layer lysates containing equivalent amounts of

protein were mixed for 20 min with a lipid prep-aration containing phorbol-12-myristate-13-acetate,phosphatidylserine and Triton X-100 mixed micellesto provide the necessary cofactors and conditions foroptimal enzyme activity [40]. To this mixture, a high-a¤nity myelin basic protein peptide and [32P]ATP(25 W Ci/ml) were added to a ¢nal assay volume of50 Wl. Following a 10 min incubation in a 30³Cwaterbath, samples were spotted onto phosphocellu-lose discs, which were then washed twice with 1%phosphoric acid and once with distilled water to re-move unincorporated label prior to placement in ascintillation counter.

2.5.2. PLA2 activationWe examined the role of PLA2 in TGF-L1-depen-

dent activation of PKC by selectively activating orinhibiting the enzyme. To examine the e¡ect of PLA2

activation on PKC activity in control and TGF-L1-

treated cultures, the PLA2 activators mastoparan(wasp venom peptide) [41] and melittin (bee venompeptide) [42] were used. Cultures were treated for 12h with vehicle alone; 0.22 ng/ml TGF-L1; 0.4, 2 or20 Wg/ml mastoparan; or 0.22 ng/ml TGF-L1+0.4,2 or 20 Wg/ml mastoparan. Melittin was used ina similar fashion, except the concentrations were0.03, 0.3 and 3 Wg/ml.

2.5.3. PLA2 inhibitionQuinacrine was used as a general inhibitor of

PLA2 activity [43]. Cultures were treated for 12 hwith DMEM containing 10% FBS plus 0.22 ng/mlTGF-L1 þ 0, 0.1, 1.0 or 10 WM quinacrine as de-scribed previously [18]. Since quinacrine is a generalinhibitor of PLA2, we also examined the e¡ects of0.1, 1 and 10 WM OEPC, which selectively inhibitssecretory PLA2 [44], and 0.1, 1 and 10 WM AA-COCF3, which is speci¢c for cytosolic forms [45].In each instance, cultures were treated with 0.22ng/ml TGF-L1+inhibitor for 12 h. Control cultureswere treated with TGF-L1+the inhibitor vehicle(0.02% ethanol in DMEM). PKC activity was mea-sured at 12 h.

2.5.4. E¡ect of arachidonic acidTo assess the role of the PGE2 pathway in TGF-

L1-dependent PKC stimulation, we examined the ef-fect of arachidonic acid, which is produced by PLA2,the rate-limiting enzyme of prostaglandin produc-tion. Cultures were treated for 12 h with vehiclealone or with 0.11 or 0.22 ng/ml TGF-L1 in the pres-ence or absence of 1, 10 or 100 WM arachidonic acid.Control cultures were treated with TGF-L1+the ara-chidonic acid vehicle (0.02% ethanol in DMEM).PKC activity was measured as described above.

2.5.5. E¡ect of diacylglycerol and palmitic acidTo assess the role of other phospholipids in TGF-

L1 action, we examined the e¡ects of diacylglycerol,which is the product of phospholipase C action, andthe e¡ects of the saturated fatty acid, palmitic acid,on PKC activity. Cultures were treated with vehiclealone, 0.11 ng/ml TGF-L1 or 0.22 ng/ml TGF-L1 inthe presence or absence of 1, 10 or 100 WM DOG orpalmitic acid. Control cultures were treated withTGF-L1+vehicle (0.02% ethanol in DMEM). PKCactivity was measured at 12 h.

BBAMCR 14618 5-4-00

V.L. Sylvia et al. / Biochimica et Biophysica Acta 1496 (2000) 311^324314

Page 5: Transforming growth factor-L1 regulation of resting zone ...Transforming growth factor-L1 regulation of resting zone chondrocytes is mediated by two separate but interacting pathways

2.5.6. Direct e¡ect of TGF-L1 on membrane-associated PKC

To determine whether TGF-L1 regulates PKC ac-tivity by direct action on membrane-associated en-zyme, isolated plasma membranes and matrixvesicles were incubated directly with TGF-L1. Bothmembrane fractions have been shown previously tocontain PKCK and PKCj [46]; PKCK predominatesin plasma membranes, whereas PKCj predominatesin matrix vesicles. Because we previously demon-strated that the e¡ect of TGF-L1 is on PKCK [8],our assay conditions were speci¢cally designed todetect this isoform.

For these experiments, plasma membranes wereisolated by di¡erential and sucrose density gradientcentrifugation of homogenized cells derived fromcon£uent, fourth passage cells [36,47]. Matrix vesicleswere isolated by di¡erential centrifugation of thetrypsin-digested matrix [48]. Following assay for pro-tein content [19], plasma membranes or matrixvesicles were suspended in 0.9% NaCl and frozenat 370³C. Matrix vesicles isolated in this mannertypically exhibit greater than 2-fold enrichment ofalkaline phosphatase speci¢c activity when comparedwith the plasma membranes [23,26,48] and have atransmission electron microscopic appearance consis-tent with matrix vesicles in vivo [48]. Matrix vesiclesor plasma membranes (1 mg protein/ml in 0.9%NaCl containing 10% FBS) were incubated in theabsence (vehicle only) or presence of a ¢nal concen-tration of 0.11, 0.22 or 0.88 ng/ml TGF-L1 for either9, 90 or 270 min at 37³C as described previously forour vitamin D studies [46]. Following incubation,samples were assayed for PKC activity.

2.6. Statistical analysis

Unless otherwise noted, the data presented beloware from one of three or more independent experi-ments, all showing comparable results. Each datapoint represents the mean þ S.E.M. for six cultures(n = 6). For membrane PKC assays, each data pointis the mean þ S.E.M. for six membrane preparations,where each preparation was derived from one T-150£ask of cells. Data were analyzed by ANOVA. Stat-istical signi¢cance was determined using Bonferroni'st-test, with P6 0.05 being considered signi¢cant.Treatment/control ratios compare the means for

each variable from ¢ve or more independent experi-ments as indicated in the ¢gure legend (nv 5). Stat-istical comparisons were performed using the Wil-coxon matched pair rank sum test. P6 0.05 wasconsidered signi¢cant.

3. Results

3.1. Role of PKC in cell response to TGF-L1

3.1.1. Cell proliferationThe results of this series of experiments showed

that cell proliferation involves a PKC-mediated path-way, since all of the PKC inhibitors caused a dose-dependent decrease in [3H]thymidine incorporation.However, we were unable to demonstrate in an un-equivocal manner that the TGF-L1-stimulated in-crease in [3H]thymidine incorporation was also medi-ated by a PKC-dependent pathway. This is becausemuch of the e¡ect of the PKC inhibitors could beaccounted for by the decrease in basal levels of DNAsynthesis. The PKC inhibitor chelerythrine signi¢-cantly reduced [3H]thymidine incorporation in bothcontrol and TGF-L1-treated cultures in a dose-dependent manner (Fig. 1A). Treatment/control ra-

Table 1The e¡ect of TGF-L1 and protein kinase inhibitors on[3H]thymidine incorporation by resting zone chondrocytes

E¡ector DPM/well (U104)

Control 4.12 þ 0.04PKC inhibitor (H-7)

1 WM 3.72 þ 0.11*5 WM 3.10 þ 0.07*10 WM 2.29 þ 0.20*�

PKA inhibitor (H-8)1 WM 3.90 þ 0.06*5 WM 3.25 þ 0.07*10 WM 2.01 þ 0.42*�

TGF-L1 (0.88 ng/ml) 13.53 þ 0.34*TGF-L1+1 WM H-7 12.09 þ 0.12*TGF-L1+5 WM H-7 8.18 þ 0.11*#

TGF-L1+10 WM H-7 5.03 þ 0.31*#�

TGF-L1+1 WM H-8 13.51 þ 0.40*TGF-L1+5 WM H-8 12.08 þ 0.22*TGF-L1+10 WM H-8 9.64 þ 0.46*#�

Values represent the mean þ S.E.M. for six independent cul-tures. *P6 0.05, treatment vs. control; #P6 0.05, TGF-L1+in-hibitor vs. TGF-L1 alone; �P6 0.05, vs. 1 WM inhibitor.

BBAMCR 14618 5-4-00

V.L. Sylvia et al. / Biochimica et Biophysica Acta 1496 (2000) 311^324 315

Page 6: Transforming growth factor-L1 regulation of resting zone ...Transforming growth factor-L1 regulation of resting zone chondrocytes is mediated by two separate but interacting pathways

tios from nine independent experiments demonstratethat the PKC inhibitor consistently reduced[3H]thymidine incorporation in control and TGF-L1-stimulated cultures in a dose-dependent manner

(Fig. 1B). Similarly, another PKC inhibitor, stauro-sporine, also blocked [3H]thymidine incorporation inboth control and TGF-L1-treated cultures (Fig. 2A).The e¡ect was dose-dependent and was consistently

Fig. 2. The e¡ect of the PKC inhibitor staurosporine on TGF-L1-induced [3H]thymidine incorporation by resting zone chon-drocytes. Con£uent, fourth passage cells were treated with con-trol media or 0.22 ng/ml TGF-L1 in the absence or presence of1 or 5 WM staurosporine for 24 h. Two hours prior to harvest,[3H]thymidine was added to the cultures. At harvest, the celllayers were washed, precipitated with TCA as described in Sec-tion 2 and counted in a scintillation counter. Data are fromone of nine independent experiments yielding similar results (A)or combined and expressed as treatment/control ratios (B). Forthe representative experiment shown, values are the mean þS.E.M. of six cultures. For treatment/control ratios, values arethe mean þ S.E.M. of nine experiments. *P6 0.05, vs. no stau-rosporine; �P6 0.05, 1 vs. 5 WM staurosporine; #P6 0.05, con-trol vs. TGF-L1.

Fig. 1. The e¡ect of the PKC inhibitor chelerythrine on TGF-L1-induced [3H]thymidine incorporation by resting zone chon-drocytes. Con£uent, fourth passage cells were treated with con-trol media or 0.22 ng/ml TGF-L1 in the absence or presence of0.01 or 0.1 WM chelerythrine for 24 h. Two hours prior to har-vest, [3H]thymidine was added to the cultures. At harvest, thecell layers were washed, precipitated with TCA as described inSection 2 and counted in a scintillation counter. Data are fromone of nine independent experiments yielding similar results (A)or combined and expressed as treatment/control ratios (B). Forthe representative experiment, values are the mean þ S.E.M. ofsix cultures. For treatment/control ratios, values are themean þ S.E.M. of nine experiments. *P6 0.05, vs. no cheleryth-rine; �P6 0.05, 0.01 vs. 0.1 WM chelerythrine; #P6 0.05, con-trol vs. TGF-L1.

BBAMCR 14618 5-4-00

V.L. Sylvia et al. / Biochimica et Biophysica Acta 1496 (2000) 311^324316

Page 7: Transforming growth factor-L1 regulation of resting zone ...Transforming growth factor-L1 regulation of resting zone chondrocytes is mediated by two separate but interacting pathways

observed (Fig. 2B). In cultures treated with TGF-L1and 5 WM staurosporine, [3H]thymidine incorpora-tion was reduced to levels observed in unstimulatedcultures. Finally, the PKC inhibitor H-7 also causeda dose-dependent decrease in [3H]thymidine incorpo-ration in control cultures and in cultures treated with0.88 ng/ml TGF-L1 (Table 1). At the highest concen-tration of H-7, [3H]thymidine incorporation in theTGF-L1-stimulated cultures was reduced to controllevels.

3.1.2. Alkaline phosphatase speci¢c activityThe TGF-L1-stimulated increase in alkaline phos-

Fig. 3. E¡ect of PKC inhibition on TGF-L1-induced proteogly-can sulfation of resting zone chondrocyte cell layers. Cells weretreated for 24 h with varying concentrations of TGF-L1 in thepresence or absence of 1 or 10 WM chelerythrine and then as-sayed for proteoglycan sulfation as described in Section 2. Eachdata point is the mean þ S.E.M. for six cultures from one ex-periment; the experiment was repeated two more times withcomparable results. *P6 0.05, vs. cultures not treated withTGF-L1; �P6 0.05, 1 WM vs. 10 WM chelerythrine; #P6 0.05,chelerythrine vs. untreated control; �P6 0.05, 0.88 ng/ml TGF-L1 vs. 0.22 ng/ml TGF-L1.

Table 2E¡ect of the PKC inhibitor, chelerythrine, on alkaline phosphatase speci¢c activity of resting zone chondrocytes

Treatment Alkaline phosphatase speci¢c activity (Wmol Pi/mg protein/min)

Control 1 WM chelerythrine 10 WM chelerythrine

Control 0.99 þ 0.11 0.75 þ 0.07# 0.59 þ 0.04#

0.11 ng/ml TGF-L1 1.42 þ 0.06* 1.57 þ 0.10* 1.61 þ 0.10*0.22 ng/ml TGF-L1 1.62 þ 0.08* 1.64 þ 0.08* 1.59 þ 0.07*0.88 ng/ml TGF-L1 1.00 þ 0.16 0.89 þ 0.13 0.80 þ 0.10

Values represent the mean þ S.E.M. for six independent cultures. *P6 0.05, vs. cultures not treated with TGF-L1; #P6 0.05, vs. nottreated with chelerythrine.

Fig. 4. E¡ect of PKC inhibition on TGF-L1-induced PGE2 pro-duction by resting zone chondrocyte cell layers. Cells weretreated for 24 h with varying concentrations of TGF-L1 in thepresence or absence of 0.01, 0.1 or 1 WM chelerythrine (A) orH-7 (B) and then assayed for PGE2 production as described inSection 2. Each data point is the mean þ S.E.M. for six culturesfrom one experiment; the experiment was repeated two moretimes with comparable results. *P6 0.05, vs. cultures nottreated with chelerythrine or H-7; �P6 0.05, 0.11 ng/ml TGF-L1 vs. 0.22 ng/ml TGF-L1; #P6 0.05, vs. cultures not treatedwith TGF-L1; �P6 0.05, 0.1 WM H-7 vs. 1.0 WM H-7.

BBAMCR 14618 5-4-00

V.L. Sylvia et al. / Biochimica et Biophysica Acta 1496 (2000) 311^324 317

Page 8: Transforming growth factor-L1 regulation of resting zone ...Transforming growth factor-L1 regulation of resting zone chondrocytes is mediated by two separate but interacting pathways

phatase speci¢c activity is not regulated by a PKC-dependent mechanism (Table 2). Even though inhi-bition of PKC activity caused a dose-dependent de-crease in basal enzyme activity, no e¡ect was notedin the TGF-L1-treated cultures.

3.1.3. Proteoglycan productionTGF-L1 stimulates proteoglycan production by a

PKC-mediated pathway (Fig. 3). Inhibition of PKCactivity with chelerythrine had no e¡ect on proteo-glycan sulfation in control cultures, but had a con-centration-dependent inhibitory e¡ect in culturestreated with TGF-L1. In cultures treated with 0.22ng/ml TGF-L1, 1 WM chelerythrine reduced[35S]sulfate incorporation by approximately 50%,and 10 WM chelerythrine reduced proteoglycan sulfa-tion to control levels. In cultures treated with 0.88ng/ml TGF-L1, 10 WM chelerythrine blocked s 80%of the TGF-L1-stimulated increase in [35S]sulfate in-corporation.

3.1.4. PGE2 productionThe production of PGE2 by resting zone chondro-

cyte cultures was increased by TGF-L1 in a dose-de-pendent manner (Fig. 4). Inhibition of PKC witheither chelerythrine (A) or H-7 (B) caused a small,dose-dependent increase in PGE2 production in thecontrol cultures. Chelerythrine, at 0.1 and 1.0 WM,caused a small, dose-dependent increase in PGE2

production in cultures treated with TGF-L1. In cul-tures treated with 1 WM chelerythrine, PGE2 produc-tion in response to 0.11 ng/ml TGF-L1 was compa-rable to that of cultures treated with 0.22 ng/ml. The

stimulatory e¡ect of H-7 was only seen in culturesexposed to 1 WM inhibitor.

3.2. Inhibition of PKA with H-8

Modulation of cell proliferation under basal con-ditions involves PKA, since inhibition of this signal-ing pathway with H-8 caused a decrease in[3H]thymidine incorporation (Table 1). H-8 also de-

Table 3E¡ect of DOG or palmitic acid on TGF-L1-induced PKC activity of resting zone chondrocyte cell layers

Treatment PKC speci¢c activity (pmol PO4/Wg protein/min)

Control 0.11 ng/ml TGF-L1 0.22 ng/ml TGF-L1

Control 0.42 þ 0.02 0.90 þ 0.03# 1.54 þ 0.04�#

1 WM DOG 0.46 þ 0.02 0.95 þ 0.04# 1.50 þ 0.08�#

10 WM DOG 0.40 þ 0.02 0.93 þ 0.02# 1.53 þ 0.06�#

100 WM DOG 0.58 þ 0.07 0.89 þ 0.04# 1.55 þ 0.06�#

Control 0.56 þ 0.03 1.90 þ 0.10# 3.24 þ 0.17�#

1 WM palmitic acid 0.59 þ 0.04 1.86 þ 0.11# 3.21 þ 0.14�#

1 WM palmitic acid 0.60 þ 0.01 1.80 þ 0.10# 3.01 þ 0.09�#

1 WM palmitic acid 0.61 þ 0.04 1.77 þ 0.10 3.07 þ 0.14�#

Values represent the mean þ S.E.M. for six independent cultures. �P6 0.05, 0.11 ng/ml TGF-L1 vs. 0.22 ng/ml TGF-L1; #P6 0.05, vs.cultures not treated with TGF-L1.

Fig. 5. E¡ect of arachidonic acid on TGF-L1-induced PKC ac-tivity of resting zone chondrocyte cell layers. Cells were treatedfor 12 h with varying concentrations of TGF-L1 in the presenceor absence of 1, 10 or 100 WM arachidonic acid and then as-sayed for PKC activity as described in Section 2. Each datapoint is the mean þ S.E.M. for six cultures from one experi-ment; the experiment was repeated two more times with com-parable results. �P6 0.05, 0.11 ng/ml TGF-L1 vs. 0.22 ng/mlTGF-L1; *P6 0.05, vs. cultures not treated with arachidonicacid; #P6 0.05, vs. cultures not treated with TGF-L1.

BBAMCR 14618 5-4-00

V.L. Sylvia et al. / Biochimica et Biophysica Acta 1496 (2000) 311^324318

Page 9: Transforming growth factor-L1 regulation of resting zone ...Transforming growth factor-L1 regulation of resting zone chondrocytes is mediated by two separate but interacting pathways

creased the stimulatory e¡ect of TGF-L1 on[3H]thymidine incorporation by 50%, implicatingPKA in the signaling pathway.

3.3. The role of arachidonic acid as mediator ofTGF-L1 action

As shown previously [49], arachidonic acid is anegative regulator of PKC activity in resting zonechondrocytes (Fig. 5). The results of the present

study indicate that the TGF-L1-dependent increasein PKC may also be mediated by arachidonic acid.In cultures treated with 0.11 ng/ml TGF-L1, 1 WMarachidonic acid reduced PKC to levels below thosein unstimulated cultures, and at 10 and 100 WM, ar-achidonic acid blocked the TGF-L1-dependent in-crease in PKC. In cultures treated with 0.22 ng/mlTGF-L1, arachidonic acid caused a dose-dependentdecrease in PKC. However, even at concentrations of100 WM, only partial inhibition was achieved (55%).DOG and palmitic acid had no signi¢cant e¡ectupon basal PKC activity or upon the TGF-L1-depen-dent increase in PKC activity (Table 3).

3.4. The role of PLA2 in the TGF-L1-dependentregulation of PKC activity

Further evidence that arachidonic acid plays a rolein mediating the e¡ects of TGF-L1 is shown in Figs.6 and 7. Stimulation of arachidonic acid release byeither mastoparan (Fig. 6A) or melittin (Fig. 6B) hadno e¡ect on basal PKC activity at 12 h, but causeddose-dependent decreases in TGF-L1-stimulated cul-tures.

Fig. 6. E¡ect of activating PLA2 on TGF-L1-induced PKC ac-tivity of resting zone chondrocyte cell layers. Cells were treatedfor 12 h with 0.22 ng/ml TGF-L1 in the presence or absence of0.4, 2 or 20 Wg/ml mastoparan (A), or 0.03, 0.3 or 3 Wg/ml me-littin (B) and then assayed for PKC activity as described in Sec-tion 2. Each data point is the mean þ S.E.M. for six culturesfrom one experiment; the experiment was repeated two moretimes with comparable results. *P6 0.05, vs. cultures nottreated with PLA2 activator; #P6 0.05, vs. cultures not treatedwith TGF-L1.

Fig. 7. E¡ect of inhibiting PLA2 on TGF-L1-induced PKC ac-tivity of resting zone chondrocyte cell layers. Cells were treatedfor 12 h with 0.22 ng/ml TGF-L1 in the presence or absence of0.1, 1 or 10 WM quinacrine and then assayed for PKC activityas described in Section 2. Each data point is the mean þ S.E.M.for six cultures from one experiment; the experiment was re-peated two more times with comparable results. *P6 0.05, vs.cultures not treated with PLA2 inhibitor; #P6 0.05, vs. culturesnot treated with TGF-L1.

BBAMCR 14618 5-4-00

V.L. Sylvia et al. / Biochimica et Biophysica Acta 1496 (2000) 311^324 319

Page 10: Transforming growth factor-L1 regulation of resting zone ...Transforming growth factor-L1 regulation of resting zone chondrocytes is mediated by two separate but interacting pathways

Conversely, inhibition of PLA2 activity by quina-crine increased PKC activity in both untreated andTGF-L1-treated cultures at all concentrations tested(Fig. 7). At 0.1 and 1 WM quinacrine, the e¡ect wasadditive, but at 10 WM quinacrine, inhibition ofPLA2 was synergistic with the e¡ect of TGF-L1 onPKC. OEPC, an inhibitor of secretory PLA2, in-creased basal and TGF-L1-stimulated PKC activity(Fig. 8A) to a greater extent than did AACOCF3, aninhibitor of cytosolic PLA2 (Fig. 8B). At 10 WM, the

e¡ects of both OEPC and AACOCF3 were synergis-tic with the e¡ects of TGF-L1.

3.5. Direct e¡ect of TGF-L1 on membrane PKC

The e¡ects of TGF-L1 on PKC are not mediatedby a direct e¡ect on pre-existing membrane-associ-ated PKC (Table 4). There was no change in PKCspeci¢c activity of isolated matrix vesicles or plasmamembranes that were incubated directly with TGF-L1. In addition, TGF-L1 did not elicit a series ofmembrane-mediated events leading to an increasein pre-existing membrane-associated PKC, since spe-ci¢c activity did not change as a function of time.

4. Discussion

This study demonstrates that some, but not all, ofthe e¡ects of TGF-L1 on resting zone chondrocytesare mediated by PKC. Incorporation of [35S]sulfatewas blocked by PKC inhibitors, indicating thatTGF-L1-dependent proteoglycan production involvesa PKC signaling pathway. In contrast, inhibition ofPKC had no e¡ect on TGF-L1-stimulated alkalinephosphatase speci¢c activity and little, if any, e¡ecton [3H]thymidine incorporation, suggesting thatPKC is not involved. This is in contrast to 24,25-(OH)2D3-dependent regulation of alkaline phospha-tase and cell proliferation, in which inhibition ofPKC also inhibits the e¡ects of the vitamin D me-tabolite (unpublished data).

Here, we show for the ¢rst time that production ofthe local factor PGE2 is stimulated by TGF-L1.Moreover, this stimulation is enhanced by inhibitionof PKC. This suggests that PKC is a negative regu-lator of PGE2 production in resting zone chondro-cytes [36]. Since release of arachidonic acid by PLA2

is the rate-limiting step in prostaglandin synthesis[50], the observation that PKC inhibition enhancesPGE2 production suggests a feedback loop betweenthe PLA2 and PKC pathways. Recent studies sup-port this hypothesis since changes in PLA2 activitya¡ect PKC activity in these cells [51].

PGE2 has multiple e¡ects on the chondrocytes,promoting di¡erentiation and anabolic responsesvia cAMP production and PKC activity [52]. Thisprostaglandin acts through its own membrane recep-

Fig. 8. E¡ect of inhibiting PLA2 on TGF-L1-induced PKC ac-tivity of resting zone chondrocyte cell layers. Cells were treatedfor 12 h with 0.22 ng/ml TGF-L1 in the presence or absence of0.1, 1 or 10 WM OEPC (A), or 0.1, 1 or 10 WM AACOCF3 (B)and then assayed for PKC activity as described in Section 2.Each data point is the mean þ S.E.M. for six cultures from oneexperiment; the experiment was repeated two more times withcomparable results. *P6 0.05, vs. cultures not treated withPLA2 inhibitor; #P6 0.05, vs. cultures not treated with TGF-L1.

BBAMCR 14618 5-4-00

V.L. Sylvia et al. / Biochimica et Biophysica Acta 1496 (2000) 311^324320

Page 11: Transforming growth factor-L1 regulation of resting zone ...Transforming growth factor-L1 regulation of resting zone chondrocytes is mediated by two separate but interacting pathways

tors to exert its e¡ects on cells. Resting zone cellshave both EP1 and EP2 receptors, as well as anEP1 variant, EP1v [53]. The increase in cAMP leadsto increased PKA activity. The importance of thispathway in the response to TGF-L1 is evident inthe decrease in proliferation following treatment ofthe cells with TGF-L1 and the PKA inhibitor, H-8.Our results also con¢rm our previous observationthat TGF-L1 stimulation of PKC was dependenton a pertussis toxin-sensitive G-protein [8]. Othershave also reported a role for a pertussis toxin-sensi-tive G-protein-dependent pathway in the action ofTGF-L1 on renal tubular epithelial cells [54].

There is clearly cross-talk between the PGE2 andPKC pathways, since inhibition of PGE2 productionincreases PKC activity [18], inhibition of PKC in-creases PGE2 production [55] and PGE2 inhibitsPKC activity [18]. The results of the present studysuggest that TGF-L1 may exert its e¡ects on PGE2

production in part via regulation of PLA2 and, there-fore, on the amount of arachidonic acid. Inhibitionof PLA2 by the general inhibitor quinacrine, the se-cretory PLA2 inhibitor OEPC, or the cytosolic PLA2

inhibitor AACOCF3, all enhanced the stimulatorye¡ects of TGF-L1 on PKC. While this e¡ect wasadditive with the increase in basal levels of PKC incontrol cultures, at the highest concentration ofOEPC, which inhibits secretory PLA2, the e¡ectswere synergistic with TGF-L1, suggesting that thesecretory isozyme is responsible. In contrast, activa-tion of PLA2 with either mastoparan or melittin de-creased the e¡ects of TGF-L1 on PKC, but had no

e¡ect on basal enzyme activity. Finally, arachidonicacid itself blocked the e¡ect of TGF-L1 on PKC atconcentrations of the fatty acid that had no e¡ect onbasal PKC activity. This e¡ect of arachidonic acidwas speci¢c since palmitic acid, a precursor of ara-chidonic acid, had no e¡ect. Moreover, the diacylgly-cerol, DOG, had no e¡ect, indicating that phospho-lipase C and/or phospholipase D [56] were notinvolved. This con¢rms our previous observationthat TGF-L1 stimulates PKC via a mechanism thatis independent of phospholipase C [8] based on inhi-bition of enzymes with speci¢c inhibitors.

The results of this study help explain the synergis-tic e¡ects of TGF-L1 and 24,25-(OH)2D3 on restingzone chondrocytes. Both agents cause an increase inPKC activity and PGE2 production [8,17,57], yetthey exert disparate e¡ects on the cell. Whereas24,25-(OH)2D3 stimulates alkaline phosphatase ac-tivity via PKC, TGF-L1 does not; whereas 24,25-(OH)2D3 modulates proliferation via PKC, TGF-L1does not. Thus, for these parameters at least, syner-gistic responses may occur because more than onemechanism is involved.

Time course of action may be another factor.24,25-(OH)2D3 elicits rapid responses on existingPKC associated with the plasma membrane and ma-trix vesicles produced by the chondrocytes [46].When plasma membranes are incubated directlywith 24,25-(OH)2D3, PKCK activity is increased. Incontrast, when matrix vesicles are incubated directlywith the vitamin D metabolite, PKCK activity is un-a¡ected, but PKCj activity is decreased. TGF-L1

Table 4Direct e¡ect of TGF-L1 on PKC activity of resting zone chondrocyte matrix vesicles and plasma membranes

Treatment PKC speci¢c activity (pmol PO4/Wg protein/min)

9 min 90 min 270 min

Matrix vesiclesControl 0.67 þ 0.08 0.74 þ 0.08 0.68 þ 0.050.11 ng/ml TGF-L1 0.60 þ 0.03 0.68 þ 0.04 0.67 þ 0.040.22 ng/ml TGF-L1 0.62 þ 0.12 0.72 þ 0.13 0.69 þ 0.040.88 ng/ml TGF-L1 0.64 þ 0.02 0.70 þ 0.02 0.70 þ 0.03Plasma membranesControl 0.24 þ 0.05 0.24 þ 0.01 0.32 þ 0.020.11 ng/ml TGF-L1 0.26 þ 0.02 0.22 þ 0.01 0.29 þ 0.010.22 ng/ml TGF-L1 0.24 þ 0.01 0.25 þ 0.01 0.33 þ 0.020.88 ng/ml TGF-L1 0.28 þ 0.03 0.26 þ 0.03 0.35 þ 0.02

Values represent the mean þ S.E.M. for six independent samples; each sample is the membranes isolated from one T-150 £ask of cells.

BBAMCR 14618 5-4-00

V.L. Sylvia et al. / Biochimica et Biophysica Acta 1496 (2000) 311^324 321

Page 12: Transforming growth factor-L1 regulation of resting zone ...Transforming growth factor-L1 regulation of resting zone chondrocytes is mediated by two separate but interacting pathways

does not exert a similar rapid e¡ect, however, sincethere was no change in PKC activity of either theplasma membranes or the matrix vesicles when theywere incubated directly with the growth factor.Moreover, TGF-L1 does not elicit translocation ofPKC from the cytosol to the nucleus [8], whereas24,25-(OH)2D3 does cause translocation to occur[17].

Even for the genomic e¡ects on PKC, there aredi¡erences in the time course of action. Whereas24,25-(OH)2D3 exerts its maximal e¡ects on PKCat 90 min [17], maximal stimulation in response toTGF-L1 is not observed until 12 h [8]. This impliesthe existence of a cascade of regulatory events, ini-tiated by 24,25-(OH)2D3, that enhance or comple-ment the events initiated by TGF-L1. Neither24,25-(OH)2D3 nor TGF-L1 require phospholipaseC, but 24,25-(OH)2D3 does exert its stimulatory ef-fects via phospholipase D, resulting in diacylglycerolproduction [18]. Diacylglycerol has no e¡ect onthe TGF-L1-dependent response, however. 24,25-(OH)2D3 e¡ects require neither Gi nor Gs [18], butTGF-L1-dependent PKC requires a pertussis toxin-

sensitive G-protein. Whereas 24,25-(OH)2D3 de-creases PGE2 production and PKA is not requiredfor its action on resting zone chondrocytes, TGF-L1causes an increase in PGE2 production and mediatesat least some of its e¡ects through PKA. For both24,25-(OH)2D3 and TGF-L1, modulation of PLA2 ispivotal. Inhibition of this enzyme activity by 24,25-(OH)2D3 stimulates PKC, and this is true for TGF-L1 as well. These observations suggest that the tworegulatory factors exert their e¡ects through separatepathways and support the hypothesis that e¡ects of24,25-(OH)2D3 on the cell may mediate the e¡ects ofTGF-L1.

Part of the biological signal of TGF-L1 is con-veyed through Smad proteins [58,59]. Upon bindingto TGF-L1, the type II receptor associates with andactivates the type I receptor. The type I receptorphosphorylates Smads 2 and 3, which translocateinto the nucleus and act as transcription factors[60]. The actions of Smads in TGF-L1-dependentgene expression are believed to be responsible forpart of the biological e¡ect of TGF-L1, albeit notthe entire e¡ect. Another important pathway known

Fig. 9. Mechanisms of TGF-L1 action in resting zone chondrocytes. TGF-L ; PLA2 ; arachidonic acid (AA); PGE2 ; alkaline phospha-tase (ALPase); PKA; PKC; 3= inhibition, + = stimulation, ( þ ) = inhibition or stimulation was time-dependent.

BBAMCR 14618 5-4-00

V.L. Sylvia et al. / Biochimica et Biophysica Acta 1496 (2000) 311^324322

Page 13: Transforming growth factor-L1 regulation of resting zone ...Transforming growth factor-L1 regulation of resting zone chondrocytes is mediated by two separate but interacting pathways

to be activated by TGF-L1 is the Ras^Raf^MEK^MAPK pathway [61,62]. PKC impinges upon thispathway as it has been recently demonstrated thatTGF-L1 activation of MAPK is dependent uponPKC [63].

In summary, the results of this study con¢rm thatthe physiological e¡ects of TGF-L1 on resting zonechondrocytes are mediated by two separate path-ways, the PKC pathway and the PGE2 pathway.Additionally, there is cross-talk between these path-ways (see Fig. 9). In this schematic diagram, we seethat TGF-L1 binds to its type II receptor, whichbinds to the type I receptor. TGF-L1 type I receptoractivates Smad proteins and also activates a G-pro-tein-dependent cascade, altering the activity of PLA2,which produces arachidonic acid. Depending uponthe time it is measured, PLA2 activity may eitherbe activated or inhibited, as indicated by the ` þ '.Previously, we have reported that TGF-L1 inhibitsPLA2 in plasma membranes and matrix vesicles [5].It is clear from the present study, however, thatTGF-L1 increases PGE2 production. This suggeststhat TGF-L1-dependent PGE2 production is derivedfrom arachidonic acid released from phospholipidsother than phosphatidylethanolamine (PE) since14C-labeled PE was used as a substrate for measuringPLA2 activity.

Arachidonic acid is metabolized by cyclooxygenaseto PGE2, which binds to EP1 and EP2 PGE2 recep-tors on the resting zone chondrocyte cell surface [53].EP1 signaling inhibits PKC activity, while EP2 sig-naling increases cAMP, which activates PKA. This isconsistent with our previous ¢nding that PGE2 treat-ment inhibits PKC activity in resting zone chondro-cytes [18], but stimulates alkaline phosphatase activ-ity. This suggests that TGF-L1 signals through PKAto alkaline phosphatase activity and through PKC tostimulate proliferation and proteoglycan production.Therefore, the regulation of both the PKC pathwayand the PGE2 pathway by TGF-L1 is essential for itsrole in the proliferation and di¡erentiation of growthplate chondrocytes.

Acknowledgements

The authors wish to thank Ms. Sandra Messier forher help in preparing the manuscript and Ms. Mon-

ica Luna and Dr. Zhi Chang for their technical as-sistance. This study was supported by US PHSGrants DE-08603 and DE-05937 and the Centerfor the Enhancement of the Biology/Biomaterials In-terface (CEBBI) at the University of Texas HealthScience Center at San Antonio.

References

[1] D.M. Rosen, S.A. Stempien, A.Y. Thompson, J.E. Brennan,L.R. Ellingworth, S.M. Seyedin, Exp. Cell. Res. 165 (1986)127^138.

[2] D.M. Rosen, S.A. Stempien, A.Y. Thompson, S.M. Seyedin,J. Cell. Physiol. 134 (1988) 337^346.

[3] S. Seyedin, A.Y. Thompson, H. Bentz, D.M. Rosen, J.McPherson, A. Conti, N. Siegel, G. Galluppi, K.A. Piez,J. Biol. Chem. 261 (1986) 5693^5695.

[4] R.N. Rosier, R.J. O'Keefe, I.D. Crabb, J.E. Puzas, Connect.Tissue Res. 20 (1989) 295^301.

[5] Z. Schwartz, L.F. Bonewald, K. Caul¢eld, B.P. Brooks, B.D.Boyan, Endocrinology 132 (1993) 1544^1552.

[6] T.I. Morales, A.B. Roberts, J. Biol. Chem. 263 (1988)12828^12831.

[7] M.B. Sporn, A.B. Roberts, L.M. Wake¢eld, R.K. Assoian,Science 233 (1986) 532^534.

[8] V.L. Sylvia, S. Mackey, Z. Schwartz, L. Shuman, R. Gomez,B.D. Boyan, J. Bone Miner. Res. 9 (1994) 1477^1487.

[9] H.A. Pedrozo, B.D. Boyan, J. Mazock, D.D. Dean, R. Go-mez, Z. Schwartz, Calcif. Tissue Int. 64 (1999) 50^56.

[10] L.F. Bonewald, M.B. Kester, Z. Schwartz, L.D. Swain, A.G.Khare, T.L. Johnson, R.J. Leach, B.D. Boyan, J. Biol.Chem. 267 (1992) 8943^8949.

[11] L.F. Bonewald, Z. Schwartz, L.D. Swain, B.D. Boyan, BoneMiner. 17 (1992) 139^144.

[12] Z. Schwartz, V.L. Sylvia, Y. Liu, D.D. Dean, B.D. Boyan,Bone 23 (1998) 465^470.

[13] Z. Schwartz, V.L. Sylvia, D.D. Dean, B.D. Boyan, Endocri-nology 139 (1998) 534^545.

[14] B.D. Boyan, Z. Schwartz, S. Park-Snyder, D.D. Dean, F.Yang, D. Twardzik, L.F. Bonewald, J. Biol. Chem. 269(1994) 28374^28381.

[15] H.A. Pedrozo, Z. Schwartz, R. Gomez, A. Ornoy, W. Xin-Sheng, S.L. Dallas, L.F. Bonewald, D.D. Dean, B.D. Boyan,J. Cell. Physiol. 177 (1998) 343^354.

[16] Z. Schwartz, B.P. Brooks, L.D. Swain, F. Del Toro, A.W.Norman, B.D. Boyan, Endocrinology 130 (1992) 2495^2504.

[17] V.L. Sylvia, Z. Schwartz, L. Schuman, R.T. Morgan, S.Mackey, R. Gomez, B.D. Boyan, J. Cell. Physiol. 157(1993) 271^278.

[18] S.H. Helm, V.L. Sylvia, T. Harmon, D.D. Dean, B.D. Boy-an, Z. Schwartz, J. Cell. Physiol. 169 (1996) 509^521.

[19] P.K. Smith, R.I. Krohn, G.T. Hermanson, A.K. Mallia,F.H. Gartner, M.D. Provenzano, E.K. Fujimoto, N.M.

BBAMCR 14618 5-4-00

V.L. Sylvia et al. / Biochimica et Biophysica Acta 1496 (2000) 311^324 323

Page 14: Transforming growth factor-L1 regulation of resting zone ...Transforming growth factor-L1 regulation of resting zone chondrocytes is mediated by two separate but interacting pathways

Goeke, B.J. Olson, D.C. Klenk, Anal. Biochem. 150 (1985)76^85.

[20] B.D. Boyan, Z. Schwartz, L.D. Swain, D.L. Carnes Jr., T.Zislis, Bone 9 (1988) 185^194.

[21] Z. Schwartz, D.L. Schlader, V. Ramirez, M.B. Kennedy,B.D. Boyan, J. Bone Miner. Res. 4 (1989) 199^207.

[22] B.D. Boyan, L.D. Swain, Z. Schwartz, V. Ramirez, D.L.Carnes Jr., Clin. Orthop. Rel. Res. 277 (1992) 266^276.

[23] B.D. Boyan, Z. Schwartz, D.L. Carnes Jr., V. Ramirez, En-docrinology 122 (1988) 2851^2860.

[24] B.D. Boyan, Z. Schwartz, L.D. Swain, Crit. Rev. Oral Biol.Med. 3 (1992) 15^30.

[25] Z. Schwartz, B.D. Boyan, Endocrinology 122 (1988) 2191^2198.

[26] Z. Schwartz, G. Knight, L.D. Swain, B.D. Boyan, J. Biol.Chem. 263 (1988) 6023^6026.

[27] Z. Schwartz, D.D. Dean, J.K. Walton, B.P. Brooks, B.D.Boyan, Endocrinology 136 (1995) 402^411.

[28] Z. Schwartz, R.H. Hancock, D.D. Dean, B.P. Brooks, R.Gomez, A.L. Boskey, G. Balian, B.D. Boyan, Endocrine 3(1995) 351^360.

[29] J.M. Herbert, J.M. Augereau, J. Gleye, J.P. Ma¡rand, Bio-chem. Biophys. Res. Commun. 172 (1990) 993^999.

[30] M. Nagano, H. Suzuki, K. Ui-Tei, S. Sato, T. Miyake, Y.Miyata, Neurosci. Res. 31 (1998) 113^121.

[31] Y. Zilberman, Y. Gutman, Biochem. Pharmacol. 44 (1992)1563^1568.

[32] N.E. Ward, C.A. O'Brian, Biochemistry 32 (1993) 11903^11909.

[33] S. Pugazhenthi, S.V. Mantha, R.L. Khandelwal, Biochem.Int. 20 (1990) 941^948.

[34] T. Tamaoki, H. Nomoto, I. Takahashi, Y. Kato, M. Mor-imoto, F. Tomita, Biochem. Biophys. Res. Commun. 135(1986) 397^402.

[35] J.P. Bretaudiere and T. Spillman, in: H.U. Bergmeyer (Ed.),Methods of Enzymatic Analysis, Verlag Chemica, Wein-heim, 1984, pp. 75^92.

[36] Z. Schwartz, D.L. Schlader, L.D. Swain, B.D. Boyan, En-docrinology 123 (1988) 2878^2884.

[37] R.J. O'Keefe, J.E. Puzas, J.S. Brand, R.N. Rosier, Endocri-nology 122 (1988) 2953^2961.

[38] E. Nasatzky, Z. Schwartz, B.D. Boyan, W.A. Soskolne, A.Ornoy, J. Cell Physiol. 154 (1993) 359^367.

[39] E. Nasatzky, Z. Schwartz, W.A. Soskolne, B.P. Brooks,D.D. Dean, B.D. Boyan, A. Ornoy, Endocr. J. 2 (1994)207^215.

[40] R.M. Bell, Y. Hannun, C. Loomis, Methods Enzymol. 124(1986) 353^359.

[41] A. Argiolas, J.J. Pisano, J. Biol. Chem. 258 (1983) 13697^13702.

[42] E. Habermann, Science 177 (1972) 314^322.[43] D. Church, S. Braconi, M. Vallotton, U. Lang, Biochem. J.

290 (1993) 477^482.[44] R.L. Magolda, W. Galbraith, J. Cell Biochem. 40 (1989)

371^386.[45] I.P. Street, H.K. Lin, F. Laliberte, F. Ghomaschi, Z. Wang,

H. Perrier, N.M. Tremblay, Z. Huang, P.K. Weech, M.H.Gelb, Biochemistry 32 (1993) 5935^5940.

[46] V.L. Sylvia, Z. Schwartz, E.B. Ellis, S.H. Helm, R. Gomez,D.D. Dean, B.D. Boyan, J. Cell Physiol. 167 (1996) 380^393.

[47] D.F. Fitzpatrick, G.R. Davenport, L. Forte, E. Landon,J. Biol. Chem. 244 (1969) 3561^3569.

[48] S.Y. Ali, S.W. Sajdera, H.C. Anderson, Proc. Natl. Acad.Sci. USA 67 (1970) 1513^1520.

[49] B.D. Boyan, V.L. Sylvia, D. Curry, Z. Chang, D.D. Dean,Z. Schwartz, J. Cell Physiol. 176 (1998) 516^524.

[50] J.R. Fletcher, Arch. Surg. 128 (1993) 1192^1196.[51] B.D. Boyan, V.L. Sylvia, D.D. Dean and Z. Schwartz, in:

A.W. Norman, R. Bouillon and M. Thomasset (Eds.), Vita-min D: Chemistry, Biology, and Clinical Applications of theSteroid Hormone, University of California at Riverside, Riv-erside, CA, 1997, pp. 353^360.

[52] Z. Schwartz, R.M. Gilley, V.L. Sylvia, D.D. Dean, B.D.Boyan, Endocrinology 139 (1998) 1825^1834.

[53] F. Del Toro, V.L. Sylvia, S.R. Schubkegel, R. Campos,D.D. Dean, B.D. Boyan, Z. Schwartz, Trans. Orthop. Res.Soc. 24 (1999) 719.

[54] H. Wu, P. Nambi, Pharmacology 57 (1997) 8^12.[55] Z. Schwartz, R.M. Gilley, V.L. Sylvia, D.D. Dean, B.D.

Boyan, Bone 24 (1999) 475^484.[56] E.G. Lapetina, B. Reep, B.R. Ganong, R.M. Bell, J. Biol.

Chem. 260 (1985) 1358^1361.[57] Z. Schwartz, L.D. Swain, D.W. Kelly, B.P. Brooks, B.D.

Boyan, Bone 13 (1992) 395^401.[58] P. Hu, M. Datto, X. Wang, Endocr. Rev. 19 (1998) 349^363.[59] R. Padgett, S. Cho, C. Evangelista, Pharmacol. Ther. 78

(1998) 47^52.[60] Y. Zhang, X. Feng, R. Wu, R. Derynck, Nature 383 (1996)

168^172.[61] Z. Yan, S. Winawer, E. Friedman, J. Biol. Chem. 269 (1994)

13231^13237.[62] K. Yamaguchi, K. Ogita, S. Nakamura, Y. Nishizuka, Bio-

chem. Biophys. Res. Commun. 210 (1995) 639^647.[63] A. Axmann, D. Seidel, T. Reimann, U. Hempel, K. Wenzel,

Biochem. Biophys. Res. Commun. 249 (1998) 456^460.

BBAMCR 14618 5-4-00

V.L. Sylvia et al. / Biochimica et Biophysica Acta 1496 (2000) 311^324324