JOURNAL OF CELLULAR PHYSIOLOGY 149:43-43
(1991)
Stimulation of Large Proteoglycan Synthesis in Cultured Smooth Muscle Cells From Pig Aorta by Endothelial Cell-Conditioned Medium ELIANE BERROU, MONIQUE BRETON, ELISABETH DEUDON, AND JACQUESPICARD* Laboratoire de Biochimie, lnserm U.187, Faculte de Medecine Saint-Antoine, 75571, Paris Cedex 12 (E.B., €,D.,/.P.) and lnserm U.326, HBpital Purpan, 3 1059, Toulouse Cedex, (M.B.) France We have previously shown (Berrou et al., J. Cell. Phys., 137:430-438, 1988)that porcine endothelial cell-conditioned medium (ECCM) stimulates proteoglycan synthesis by smooth muscle cells from pig aorta. ECCM stimulation requires protein cores for glycosaminoglycan chain initiation and is accompanied by an increase in the hydrodynamic size of proteoglycans secreted into the medium. This work investigates the mechanisms involved in the ECCM effect. 1) Control and ECCM stimulated proteoglycan synthesis (measured by a 20 min [3'S]-sulfate labeling assay) was not inhibited by cycloheximide, indicating that the proteoglycans were composed of preexisting protein cores and that ECCM stimulates glycosylation of these protein cores. 2) Whereas ECCM stimulation of [ 3 5 S ] methionine incorporation into secreted proteins only occurred after a 6 h incubation, the increase in ["S] methionine-labeled proteoglycans was observed after 1 h, and the increase was stable for at least 16 h. 3) As analysed by electrophoresis in SDS, chondroitinase digestion generated from [14C] serine-labeled proteoglycans 7 protein cores of high apparent molecular mass (550-200 kDa) and one of 47 kDa. The two protein cores of highest apparent molecular masses (550 and 460 kDa), but not the 47 kDa protein cores, showed increased ['4C]-serine incorporation in response to ECCM (51 %, as measured by Sepharose CL-6B chromatography). 4) Finally, incorporation of [35S]-sulfate into chondroitinase-generated glycosaminoglycan linkage stubs on protein cores was determined by Sepharose CL-66 chromatography: ECCM did not modify the ratio [35S]/['4C1in stimulated protein cores, indicating that ECCM did not affect the number of glycosaminoglycan chains. The results of these studies reveal that 1) endothelial cells secrete factor(s) that preferentially stimulate synthesis of the largest smooth muscle cell proteoglycans without structural modifications and 2)the stimulation proceeds via increased glycosylation of protein core through enhancement of xylosylated protein core, followed by enhanced protein synthesis.
Proteoglycans are complex macromolecules com- the functions of each population. By their multiple posed of a protein core carrying glycosaminoglycan interactions, proteoglycans are involved in cell adhechains of one or several types. Proteoglycans in the sion, cell migration, and presumably in growth control arterial wall mainly contain chondroitinidermatan sul- (Ruoslahti, 1989), and all of these cellular events are fate (CSIDS)and their relative proportions vary accord- associated with the development of atherosclerotic leing to the species (Radhakrishnamurthy et al., 1990). sion. Moreover, focal accumulation of proteoglycans Different families of CSiDS proteoglycans have been that accompanies intimal smooth muscle cell proliferrecovered from different locations in the arterial wall ation may predispose the arterial wall to lipid accumu(Wight, 1989): large proteoglycans are present princi- lation, calcification, and thrombosis, because of the pally in the interstitial matrix, while small proteogly- ability of arterial wall proteoglycans to interact with cans are regularly aligned along collagen fibrils. molecules involved in these processes. Heparan sulfate proteoglycans appear to be present in Under normal conditions, proteoglycan synthesis inarterial basement membranes and interact with elastic creases when quiescent smooth muscle cells are stimfibers. From cell culture studies i t appears that certain ulated to divide (Breton e t al., 1986; Wight, 1989). classes of proteoglycans which are hydrophobic may be However, during the development of atherosclerotic associated with the plasma membranes of vascular lesion, the accumulation of proteoglycans may also cells (Wight, 1989; Schmidt and Buddecke, 1988). The variety of biological functions attributed to proteoglycans reflects their structural diversity, and Received March 13, 1991; accepted June 25, 1991 their different locations might indicate differences in *To whom reprint requestsicorrespondence should be addressed. 0 1991 WILEY-LISS, INC
AORTIC CELLS AND LARGE PROTEOGLYCAN SYNTHESIS
result from other processes. Studies of injured aortas revealed a stimulation of glycosaminoglycan synthesis during regeneration of the endotheliuni which did not occur in deendothelialized and in uninjured regions of aorta (Alavi and Moore, 1987; Wight et al., 1983). The role of endothelial cells in stimulating glycosaminoglycan synthesis by smooth muscle cells was confirmed by studies with cell cultures (Scott and Merrilees, 1987; Merrilees et al., 1990). Recently, Meeriles and Scott (19901 reported that stimulation of glycosaminoglycan synthesis by smooth muscle cells might be mediated by transforming growth factor-p. However, all of these studies mainly focused on the level of synthesis and distribution of arterial glycosaminoglycans and did not attempt to identify which of the proteoglycan populations were increased, or to characterize the mechanisms of stimulation of proteoglycan synthesis. We have previously shown (Berrou et al., 1988) that endothelial cells secrete factor(s) that stimulate proteoglycan synthesis by quiescent smooth muscle cells. This stimulation requires a protein acceptor for glycosaminoglycan chain initiation and is accompanied by a n increase in proteoglycan hydrodynamic size without modifications of the structural characteristics of their glycosaminoglycan chains. Thus, studies were pursued 1)to investigate whether the mechanisms implicated in the stimulation of smooth muscle cell proteoglycan synthesis involve either glycosylation of protein cores or protein synthesis or both and 2) to determine whether the increase in the hydrodynamic size of proteoglycans might be attributed to a selective increase in large proteoglycans andlor to the incorporation of additional glycosaminoglycan chains.
MATERIALS Plastic culture flasks were from Costar (Cambridge, USA); Dulbecco's Modified Eagle Minimum Medium (DMEM) containing 1gil glucose and 2 mM glutamine; Minimum Essential Medium with Earle's salts (MEM) containing 2 mM glutamine; penicillin; streptomycin and newborn calf serum were purchased from Boehringer (Mannheim, F.R.G.). Fetal bovine serum was obtained from Biological Industries (Beth Haemek Israel). Chondroitin ABC lyase (Proteus vulgaris EC 4.2.2.4) was from Miles Scientific (Naperville, USA). DEAE-Sephacel, Sepharose CL-6B, and Sephadex G-25 M, PD-10 columns were purchased from Pharmacia Fine Chemicals (Uppsala, Sweden). Reagents for polyacrylamide-gel electrophoresis were from Biorad (Richmond, USA). Na,135S041 (500 mciimmole) and Enlightning were obtained from New England Nuclear (Dreieich, F.R.G.); L-[U-'4C] serine (150 mciimmole) and L-[35S]-methionine (1075 Ci/mmole) from C.E.A. (GifiYvette, France). Dialysis was performed with Hollow Fiber Bundles, molecular mass cut off 6,000 Da, purchased from Spectrum (Los Angeles, CAI. All other chemicals were of analytical grade and were purchased from Fluka (Buchs, Switzerland) and Sigma Chemical Co. (Saint Louis, MO). METHODS Endothelial cells culture Endothelial cells were isolated from pig thoracic aorta and grown in flasks coated with smooth muscle
437
cell extracellular matrix as previously described (Berrou et al., 1988). The culture medium used was DMEM supplemented with 10% newborn calf serum and 5% fetal bovine serum.
Smooth muscle cell cultures Primary cultures of smooth muscle cells from media explants of pig thoracic aorta were established according to the method of Ross (1971). Cells were cultured in MEM with 10% fetal bovine serum and blocked in Go phase as previously described (Breton e t al., 1986). Smooth muscle cells in the second to the seven passage were used for all experiments. Experiments were initiated by incubation of quiescent smooth muscle cells in MEM supplemented with 0.1% (wiv) albumin (control medium) or in 48 h endothelial-cell-conditioned medium, diluted (1:1) with MEM, and adjusted to 0.1% (wiv) albumin (ECCM) (Berrou et al., 1988). Effect of cycloheximide on the synthesis of [35S]sulfate-labeled proteoglycans secreted into the medium After 4 h incubation with fresh control medium, quiescent smooth muscle cells were placed in control medium or ECCM, both containing 100 $.Xml Na,135S041, in the presence or absence of 100 pgiml cycloheximide; 135S]-sulfatehas been shown to be a specific marker for proteoglycans (Breton et al., 1986). After 20 min incubation, the medium was collected and the radioactivity incorporated into proteoglycans was determined by chromatography on Sephadex G-25 M, PD-10 columns (Berrou et al., 1988). Measurement of the incorporation of ["Sl-methionine into proteins secreted into the medium Quiescent smooth muscle cells were placed in control medium or ECCM each containing 15 pCiiml ["S1methionine. At the end of different labeling periods the radioactive medium was collected, and the radioactivity incorporated into the secreted proteins in aliquots of the medium was determined as described above. Isolation and purification of radiolabeled proteoglycans Quiescent smooth muscle cells were placed in control medium or ECCM each containing either 15 FCiiml ["SJ-methionine, or 17 FCiiml ['4Cl-serine, or 200 pCiiml Na,[35S041. At the end of different labeling periods the radioactive media of four 25 cm2 flasks were pooled, and solid urea was added to a final concentration of 8 M. CHAPS (3- [ (3- Cholamidopropyl) - dimethylammoniol I - 1-propanesulfonate) was added to a final concentration of 0.5% (wiv), proteinase inhibitors (10 mM EDTA, 0.1 M 6-amino-hexanoic acid, 5 mM benzamidine-HCI) were added, and the pH was adjusted to 7.6. Proteoglycans were purified by ion-exchange chromatography on small DEAE-Sephacel columns (0.8 ml). The columns were washed with buffer containing 8 M urea, 0.5% CHAPS, 50 mM Tris-HC1, and 0.25 M NaCl, pH 7.6, until no radioactivity eluted; proteoglycans were then eluted with 4 M guanidine HC1, 0.5% CHAPS. The
438
BERROU ET AL.
recovery of radioactivity from DEAE-Sephacel columns was 90% o r better. All purification steps were perECCM formed in the presence of proteinase inhibitors. Isolation and purification of proteoglycan 2.0 protein cores [ 14C]-serine or i3% I-sulfate labeled proteoglycans isolated on DEAE-Sephacel columns were dialyzed against distilled water in the presence of 0.2% CHAPS and proteinase inhibitors a t pH 7.6. To remove gly1 .o cosaminoglycan chains, the labeled proteoglycans were incubated in 83 mU1ml chondroitin ABC lyase in 33 mM Tris-HC1, 33 mM sodium acetate, and 3 pgiml bovine serum albumin, pH 8.0, a t 25°C for 30 min in the 0 CHX presence of proteinase inhibitors. These incubation - + - + conditions resulted in the liberation of intact protein cores by hydrolysis of glycosaminoglycan chains. Fig. 1. Effect of cycloheximide on the stimulation of L3'Sl sulfatelabeled proteoglycan synthesis by ECCM. Quiescent smooth muscle SDS/polyacrylamide-gel electrophoresis cells from pig aorta were placed in fresh control medium or ECCM Intact or digested proteoglycans were subjected to with ( + I or without ( - ) 100 pgiml cycloheximide (CHX).Each medium 100 FCiiml I"S1-sulfate. After 20 min incubation the electrophoresis on 3-12% polyacrylamide gels (TIC = contained amount of radioactivity incorporated into proteoglycans was deter3012.671, with a 3% stacking gel in the buffer system of mined. Each point represents the mean of five determinations ? SE. Laemmli (1970). Samples were precipitated with a The values obtained for ECCM and control medium are significantly 9-fold excess of 95% ethanol a t 4°C overnight and then different with P
(1991)
Stimulation of Large Proteoglycan Synthesis in Cultured Smooth Muscle Cells From Pig Aorta by Endothelial Cell-Conditioned Medium ELIANE BERROU, MONIQUE BRETON, ELISABETH DEUDON, AND JACQUESPICARD* Laboratoire de Biochimie, lnserm U.187, Faculte de Medecine Saint-Antoine, 75571, Paris Cedex 12 (E.B., €,D.,/.P.) and lnserm U.326, HBpital Purpan, 3 1059, Toulouse Cedex, (M.B.) France We have previously shown (Berrou et al., J. Cell. Phys., 137:430-438, 1988)that porcine endothelial cell-conditioned medium (ECCM) stimulates proteoglycan synthesis by smooth muscle cells from pig aorta. ECCM stimulation requires protein cores for glycosaminoglycan chain initiation and is accompanied by an increase in the hydrodynamic size of proteoglycans secreted into the medium. This work investigates the mechanisms involved in the ECCM effect. 1) Control and ECCM stimulated proteoglycan synthesis (measured by a 20 min [3'S]-sulfate labeling assay) was not inhibited by cycloheximide, indicating that the proteoglycans were composed of preexisting protein cores and that ECCM stimulates glycosylation of these protein cores. 2) Whereas ECCM stimulation of [ 3 5 S ] methionine incorporation into secreted proteins only occurred after a 6 h incubation, the increase in ["S] methionine-labeled proteoglycans was observed after 1 h, and the increase was stable for at least 16 h. 3) As analysed by electrophoresis in SDS, chondroitinase digestion generated from [14C] serine-labeled proteoglycans 7 protein cores of high apparent molecular mass (550-200 kDa) and one of 47 kDa. The two protein cores of highest apparent molecular masses (550 and 460 kDa), but not the 47 kDa protein cores, showed increased ['4C]-serine incorporation in response to ECCM (51 %, as measured by Sepharose CL-6B chromatography). 4) Finally, incorporation of [35S]-sulfate into chondroitinase-generated glycosaminoglycan linkage stubs on protein cores was determined by Sepharose CL-66 chromatography: ECCM did not modify the ratio [35S]/['4C1in stimulated protein cores, indicating that ECCM did not affect the number of glycosaminoglycan chains. The results of these studies reveal that 1) endothelial cells secrete factor(s) that preferentially stimulate synthesis of the largest smooth muscle cell proteoglycans without structural modifications and 2)the stimulation proceeds via increased glycosylation of protein core through enhancement of xylosylated protein core, followed by enhanced protein synthesis.
Proteoglycans are complex macromolecules com- the functions of each population. By their multiple posed of a protein core carrying glycosaminoglycan interactions, proteoglycans are involved in cell adhechains of one or several types. Proteoglycans in the sion, cell migration, and presumably in growth control arterial wall mainly contain chondroitinidermatan sul- (Ruoslahti, 1989), and all of these cellular events are fate (CSIDS)and their relative proportions vary accord- associated with the development of atherosclerotic leing to the species (Radhakrishnamurthy et al., 1990). sion. Moreover, focal accumulation of proteoglycans Different families of CSiDS proteoglycans have been that accompanies intimal smooth muscle cell proliferrecovered from different locations in the arterial wall ation may predispose the arterial wall to lipid accumu(Wight, 1989): large proteoglycans are present princi- lation, calcification, and thrombosis, because of the pally in the interstitial matrix, while small proteogly- ability of arterial wall proteoglycans to interact with cans are regularly aligned along collagen fibrils. molecules involved in these processes. Heparan sulfate proteoglycans appear to be present in Under normal conditions, proteoglycan synthesis inarterial basement membranes and interact with elastic creases when quiescent smooth muscle cells are stimfibers. From cell culture studies i t appears that certain ulated to divide (Breton e t al., 1986; Wight, 1989). classes of proteoglycans which are hydrophobic may be However, during the development of atherosclerotic associated with the plasma membranes of vascular lesion, the accumulation of proteoglycans may also cells (Wight, 1989; Schmidt and Buddecke, 1988). The variety of biological functions attributed to proteoglycans reflects their structural diversity, and Received March 13, 1991; accepted June 25, 1991 their different locations might indicate differences in *To whom reprint requestsicorrespondence should be addressed. 0 1991 WILEY-LISS, INC
AORTIC CELLS AND LARGE PROTEOGLYCAN SYNTHESIS
result from other processes. Studies of injured aortas revealed a stimulation of glycosaminoglycan synthesis during regeneration of the endotheliuni which did not occur in deendothelialized and in uninjured regions of aorta (Alavi and Moore, 1987; Wight et al., 1983). The role of endothelial cells in stimulating glycosaminoglycan synthesis by smooth muscle cells was confirmed by studies with cell cultures (Scott and Merrilees, 1987; Merrilees et al., 1990). Recently, Meeriles and Scott (19901 reported that stimulation of glycosaminoglycan synthesis by smooth muscle cells might be mediated by transforming growth factor-p. However, all of these studies mainly focused on the level of synthesis and distribution of arterial glycosaminoglycans and did not attempt to identify which of the proteoglycan populations were increased, or to characterize the mechanisms of stimulation of proteoglycan synthesis. We have previously shown (Berrou et al., 1988) that endothelial cells secrete factor(s) that stimulate proteoglycan synthesis by quiescent smooth muscle cells. This stimulation requires a protein acceptor for glycosaminoglycan chain initiation and is accompanied by a n increase in proteoglycan hydrodynamic size without modifications of the structural characteristics of their glycosaminoglycan chains. Thus, studies were pursued 1)to investigate whether the mechanisms implicated in the stimulation of smooth muscle cell proteoglycan synthesis involve either glycosylation of protein cores or protein synthesis or both and 2) to determine whether the increase in the hydrodynamic size of proteoglycans might be attributed to a selective increase in large proteoglycans andlor to the incorporation of additional glycosaminoglycan chains.
MATERIALS Plastic culture flasks were from Costar (Cambridge, USA); Dulbecco's Modified Eagle Minimum Medium (DMEM) containing 1gil glucose and 2 mM glutamine; Minimum Essential Medium with Earle's salts (MEM) containing 2 mM glutamine; penicillin; streptomycin and newborn calf serum were purchased from Boehringer (Mannheim, F.R.G.). Fetal bovine serum was obtained from Biological Industries (Beth Haemek Israel). Chondroitin ABC lyase (Proteus vulgaris EC 4.2.2.4) was from Miles Scientific (Naperville, USA). DEAE-Sephacel, Sepharose CL-6B, and Sephadex G-25 M, PD-10 columns were purchased from Pharmacia Fine Chemicals (Uppsala, Sweden). Reagents for polyacrylamide-gel electrophoresis were from Biorad (Richmond, USA). Na,135S041 (500 mciimmole) and Enlightning were obtained from New England Nuclear (Dreieich, F.R.G.); L-[U-'4C] serine (150 mciimmole) and L-[35S]-methionine (1075 Ci/mmole) from C.E.A. (GifiYvette, France). Dialysis was performed with Hollow Fiber Bundles, molecular mass cut off 6,000 Da, purchased from Spectrum (Los Angeles, CAI. All other chemicals were of analytical grade and were purchased from Fluka (Buchs, Switzerland) and Sigma Chemical Co. (Saint Louis, MO). METHODS Endothelial cells culture Endothelial cells were isolated from pig thoracic aorta and grown in flasks coated with smooth muscle
437
cell extracellular matrix as previously described (Berrou et al., 1988). The culture medium used was DMEM supplemented with 10% newborn calf serum and 5% fetal bovine serum.
Smooth muscle cell cultures Primary cultures of smooth muscle cells from media explants of pig thoracic aorta were established according to the method of Ross (1971). Cells were cultured in MEM with 10% fetal bovine serum and blocked in Go phase as previously described (Breton e t al., 1986). Smooth muscle cells in the second to the seven passage were used for all experiments. Experiments were initiated by incubation of quiescent smooth muscle cells in MEM supplemented with 0.1% (wiv) albumin (control medium) or in 48 h endothelial-cell-conditioned medium, diluted (1:1) with MEM, and adjusted to 0.1% (wiv) albumin (ECCM) (Berrou et al., 1988). Effect of cycloheximide on the synthesis of [35S]sulfate-labeled proteoglycans secreted into the medium After 4 h incubation with fresh control medium, quiescent smooth muscle cells were placed in control medium or ECCM, both containing 100 $.Xml Na,135S041, in the presence or absence of 100 pgiml cycloheximide; 135S]-sulfatehas been shown to be a specific marker for proteoglycans (Breton et al., 1986). After 20 min incubation, the medium was collected and the radioactivity incorporated into proteoglycans was determined by chromatography on Sephadex G-25 M, PD-10 columns (Berrou et al., 1988). Measurement of the incorporation of ["Sl-methionine into proteins secreted into the medium Quiescent smooth muscle cells were placed in control medium or ECCM each containing 15 pCiiml ["S1methionine. At the end of different labeling periods the radioactive medium was collected, and the radioactivity incorporated into the secreted proteins in aliquots of the medium was determined as described above. Isolation and purification of radiolabeled proteoglycans Quiescent smooth muscle cells were placed in control medium or ECCM each containing either 15 FCiiml ["SJ-methionine, or 17 FCiiml ['4Cl-serine, or 200 pCiiml Na,[35S041. At the end of different labeling periods the radioactive media of four 25 cm2 flasks were pooled, and solid urea was added to a final concentration of 8 M. CHAPS (3- [ (3- Cholamidopropyl) - dimethylammoniol I - 1-propanesulfonate) was added to a final concentration of 0.5% (wiv), proteinase inhibitors (10 mM EDTA, 0.1 M 6-amino-hexanoic acid, 5 mM benzamidine-HCI) were added, and the pH was adjusted to 7.6. Proteoglycans were purified by ion-exchange chromatography on small DEAE-Sephacel columns (0.8 ml). The columns were washed with buffer containing 8 M urea, 0.5% CHAPS, 50 mM Tris-HC1, and 0.25 M NaCl, pH 7.6, until no radioactivity eluted; proteoglycans were then eluted with 4 M guanidine HC1, 0.5% CHAPS. The
438
BERROU ET AL.
recovery of radioactivity from DEAE-Sephacel columns was 90% o r better. All purification steps were perECCM formed in the presence of proteinase inhibitors. Isolation and purification of proteoglycan 2.0 protein cores [ 14C]-serine or i3% I-sulfate labeled proteoglycans isolated on DEAE-Sephacel columns were dialyzed against distilled water in the presence of 0.2% CHAPS and proteinase inhibitors a t pH 7.6. To remove gly1 .o cosaminoglycan chains, the labeled proteoglycans were incubated in 83 mU1ml chondroitin ABC lyase in 33 mM Tris-HC1, 33 mM sodium acetate, and 3 pgiml bovine serum albumin, pH 8.0, a t 25°C for 30 min in the 0 CHX presence of proteinase inhibitors. These incubation - + - + conditions resulted in the liberation of intact protein cores by hydrolysis of glycosaminoglycan chains. Fig. 1. Effect of cycloheximide on the stimulation of L3'Sl sulfatelabeled proteoglycan synthesis by ECCM. Quiescent smooth muscle SDS/polyacrylamide-gel electrophoresis cells from pig aorta were placed in fresh control medium or ECCM Intact or digested proteoglycans were subjected to with ( + I or without ( - ) 100 pgiml cycloheximide (CHX).Each medium 100 FCiiml I"S1-sulfate. After 20 min incubation the electrophoresis on 3-12% polyacrylamide gels (TIC = contained amount of radioactivity incorporated into proteoglycans was deter3012.671, with a 3% stacking gel in the buffer system of mined. Each point represents the mean of five determinations ? SE. Laemmli (1970). Samples were precipitated with a The values obtained for ECCM and control medium are significantly 9-fold excess of 95% ethanol a t 4°C overnight and then different with P
