EXPERIMENTAL
CELL
RESEARCH
203, 344-353 (1992)
Vascular Smooth Muscle Cell Detachment from Elastin and Migration through Elastic Laminae Is Promoted by Chondroitin Sulfate-Induced “Shedding” of the 67-kDa Cell Surface Elastin Binding Protein’ ALEKSANDER The Division
HINEK, JENNIFER BOYLE, AND MARLENE RABINOVITCH~
of Cardiovascular
Research, Research Institute, The Hospital for Sick Children, and Departments and Pediatrics, University of Toronto, Toronto, Ontario, Canada M5G 1X8
was necessary to replace the EBP molecules released from the cell surface by CS treatment. In the migration assay, both DA and Ao SMC attached to the top of an elastin membrane, but only DA SMC deficient in EBP migrated through the laminae. Addition of CS, which induced shedding of EBP, resulted in Ao SMC migration associated with increased synthesis of fibronectin. We postulate that C&induced release of EBP from SMC surfaces causes cell detachment from elastin and an increase in fibronectin synthesis, processes which may be critical in promoting SMC migration associated with intimal thickening developmentally in the DA and perhaps also in vascular disease. er law Academic PMSS, IW.
Impaired elastin fiber assembly is observed in the fetal ductus arteriosus (DA), associated with a reduced concentration of elastin binding protein (EBP), a 67kDa galactolectin. It is also seen in cultured aortic (Ao) smooth muscle cells (SMC) following the release of the EBP by glycosaminoglycans rich in N-acetylgalactosamine, such as chondroitin sulfate (CS). In the DA, impaired elastin fiber assembly is observed in conjunction with intimal thickening associated with increased migration of SMC into the subendothelium, a feature we previously related to increased production of fibronectin. In this report, we determined whether SMC use the EBP to attach to an elastin substrate, whether shedding of the EBP promotes SMC migration through a threedimensional network of pure elastic laminae prepared from sheep aorta, and whether the latter is associated with increased production of fibronectin. We observed reduced attachment to elastin-coated surfaces of DA SMC deficient in EBP compared to Ao SMC. Addition of CS but not heparan sulfate (a glycosaminoglycan which does not induce EBP shedding) decreased Ao SMC atas did preincubation. with tachment to elastin, VGVAPG elastin-derived peptides which saturate the EBP. The immunolocalization of cell surface EBP suggested that cells can quickly replace EBP released from their surfaces by CS treatment. The magnitude of CSinduced impaired attachment of SMC to elastin was dose dependent and could be further increased by the administration of cyclohexamide and sodium azide. Also, the reversibility of CS-induced detachment was prevented by monensin. This suggests that a process of new synthesis and intracellular transport of the EBP
INTRODUCTION
’ Supported by a grant from the Medical Research Council of Canada, MT 8546, and a grant from the Heart and Stroke Foundation of Ontario, B-2040. This study was presented in part at the 31st Annual Meeting of American Society of Cell Biology, Boston, MA, December, 1991, and was published in abstract form (J. Cell. Biol. 115:441a, 1991). ’ To whom reprint requests should be addressed at Division of Cardiovascular Research, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, Canada M5G 1X8. Fax: (416) 813-7480. 0014.4827192 $5.00 Copynght 0 1992 by Academic Press, All rights of reproduction in any form
344 Inc. reserved
of Pathology
The process of intimal thickening is associated with atherosclerosis and advanced pulmonary vascular disease. “Intimal cushions” are also formed in the late gestation ductus arteriosus (DA) [I-3]. Under all these conditions, migration of smooth muscle cells (SMC) into the subendothelium is accompanied by certain common changes in the composition and organization of the extracellular matrix. Our attention has been focused on the fact that in atherosclerotic [4-121 or wounded vessels [ 13-161, in systemic hypertension [ 171, in experimental pulmonary vascular disease [ 18-191, and in the DA [20-241, SMC migration into the subendothelium is accompanied by impaired assembly of newly synthesized elastin and increased accumulation of N-acetylgalactosamine-containing glycosaminoglycans (GAGS), especially chondroitin sulfate (CS) and dermatan sulfate (DS) in the vessel wall. It has been postulated that CS- and DS-containing proteoglycans inhibit the attachment of a variety of cells to numerous components of the extracellular matrix such as collagen type I, vitronectin, fibronectin, and laminin [25-291. Our previous in vitro studies have shown a direct association between an excess of free N-acetylgalactosamine-containing GAGS, such as CS
CHONDROITIN
SIJLFATE
AND
and DS. release of the 67-kDa elastin binding protein (EBP) from vascular smooth muscle cell (SMC) surfaces and impaired assembly of elastic fibers 1301. This 67-kDa peripheral membrane protein with both elastin binding and /3-galactosugar binding properties forms a receptor complex with two integral membrane proteins (55 and 61 kDa) which allow a transmembrane link between the extracellular ligand and the cytoskeleton [31]. Of particular interest is the observation that binding P-galactosugars to the lectin site of the 67-kDa protein decreases its affinity for both elastin and cell membrane components of the receptor complex resulting in the release of bound tropoelastin and in the dissociation of the 67-kDa protein from the cell membrane [30-331. The 67-kDa EBP binds the VGVAPG hydrophobic sequence on tropoelastin and appears to guide its assembly through a highly coordinated mechanism in which binding of galactosylated microfibrillar proteins to the lectin site of the 67-kDa EBP releases bound tropoelastin onto the microfibrillar scaffold of the growing elastic fiber [30,31]. It has not been determined whether SMC use this EBP to attach to insoluble elastin, i.e., whether increased release of the EBP, in addition to preventing normal assembly of elastin, will detach the SMC from elastin, nor has it been determined whether release of the EBP will facilitate SMC migration through fenestrations in the elastic laminae. In this report we used CS to promote the release of the EBP from Ao SMC surfaces and observed that its abnormal loss prevented Ao SMC attachment to an elastin substratum and promoted their migration through a three-dimensional network of nure elastic laminae. To confirm a direct cause and effect relationship, we demonstrated loss of Ao SMC attachment to immobilized elastin by saturating the EBP with free VGVAPG elastin peptides and we also showed that CS could induce the detachment of previously attached cells. This process was reversible, but the reversibility required replacement of the EBP, a function of resynthesis and transcellular transport to cell surfaces. Ductus arteriosus but not aortic SMC migrated through elastic laminae, unless the latter were treated with CS. Finally, the CS-induced release of the EBP, in addition to preventing attachment of SMC to elastin, also increased SMC production of fibronectin, a feature which we have related to SMC migration associated with intimal thickening [35] in the DA. METHODS Materials. Chemicals and reagents were obtained as follows. Medium 199, PBS, fetal bovine serum (FBS), and other tissue culture reagents were obtained from GIBCO (Burlington, Ontario). Chondroitin sulfate A from bovine trachea, heparan sulfate from bovine kidney, monensin, cycloheximide, sodium azide, trypsin, soybean trypsin inhibitor, VGVAPG elastin peptides, and GRGDS synthetic peptides, and all reagent grade chemicals, were purchased from
SMOOTH
MUSCLE
MIGRATION
345
Sigma (St. Louis, MO). Monoclonal antibody to chondroitin sulfate clone CS-56 was also obtained from Sigma and “BCZ” monoclonal antibody to the elastin binding protein [36] was a gift from Dr. R. P. Mecham, Washington University (St. Louis, MO). Monoclonal antibody to the proliferating cell nuclear antigen (anti-PCNA) was purchased from Coulter Immunology (Hialeah, FL). Species and typespecific fluorescein (FITC)-conjugated goat anti-mouse (GAM) secondary antibody was supplied by ZYMED Laboratories, Inc. (San Francisco, CA). The peroxidase/antiperoxidase staining kit was obtained from Dako (Santa Barbara, CA). Radioactive [i4C]methionine was supplied by New England Nuclear (Boston, MA) and K-elastin from Elastin Products (Pacific, MO). Gelatin 4BSepharose was purchased from Pharmacia (Uppsala, Sweden). Cell cultures. Smooth muscle cells (SMC) were harvested from the DA and Ao of 138-day gestation fetal Rambouillet lambs by a protocol previously described [37]. The time point chosen was concurrent with the formation of intimal cushions in uiuo. Smooth muscle cells were grown initially from explants of the media as described by Ross [38] and then further passaged by trypsinization. All cultures were maintained in Medium 199 supplemented with 20 mA4 Hepes, 1% antibiotics/antimycotics and 5% FBS. The cells from the second passage which were characterized as smooth muscle using a monoclonal antibody specific to smooth muscle actin [39] were used in all exneriments described below. ;mmunohistochemistry, To confirm the increase in CS synthesis previously reported in DA SMC in uitro [24], light microscopic localization of CS was carried out on DA and Ao paraffin-embedded tissue from three different lambs using a monoclonal antibody to CS (5 mg/ml initial concentration, diluted 1:500). The reaction was developed with the peroxidase/anti-peroxidase (PAP) kit according to the manufacturer’s instructions. We also used immunofluorescence to assess whether CS-induced release of the 67.kDa EBP from Ao SMC surfaces [30] was reversible. Cultured Ao SMC were either incubated for 15 min in the presence of CS (200 @g/ml) added to Medium 199 +lO% FBS and fixed or fresh medium without CS was replaced and incubation was continued for an additional 30 min or 1 h prior to fixation. For immunofluorescent localization of the EBP, the nonpermeabilized cells were fixed for 2 min with 0.5% paraformaldehyde, washed in PBS, blocked with 0.5 M glycine and 1% BSA in PBS, and incubated with the monoclonal antibody to the 67.kDa EBP, 5 mg/ml, diluted 1:200 1361. The cells were then stained with fluorescein-labeled goat anti-mouse immunoglobulin (GAM-FITC) diluted 1:lOO. As a control for studies using the monoclonal antibody, normal ascitic fluid was substituted for the primary antibody. Additional controls included secondary antibody alone. Smooth muscle cell adhesion to elastin andglycosaminoglycans. To document whether the presence of the 67.kDa EBP might be of importance in cell attachment to elastin, we compared EBP-deficient DA SMC with Ao SMC. We also assessed Ao SMC in the presence of CS which induces shedding of the EBP and heparan sulfate (HS), an N-acetylglucosamine glycosaminoglycan which does not induce EBP shedding. Ductus arteriosus or Ao SMC were harvested by 5 min exposure to 0.25% trypsin and 0.02% EDTA in Ca*+ and Mg2+ free phosphate-buffered saline (PBS), sedimented by centrifugation, washed twice in PBS with 0.1% soybean trypsin inhibitor, counted in a Coulter counter, and further incubated for 1 h before plating. The cells were then resuspended in Hanks’ balanced salt solution supplemented with 20 mM Hepes, pH 7.4, and plated (1 X lo5 cells/ml) on 24-well culture dishes coated with K-elastin and on bacteriological petri dishes (Costar) which do not support cell adhesion in the absence of coating. To coat the dishes, K-elastin was dissolved in 0.1 M sodium bicarbonate buffer, added to the 35.mm dishes in a final concentration of 500 rgldish, then air-dried at room temperature for 18 h. Aorta and DA cells plated on elastin-coated dishes were incubated at 37°C for different periods of time between 2 and 10 h. At the end of each time
346
HINEK,
BOYLE,
interval, the medium containing unattached cells was removed and the dishes were washed twice with Hanks’ balanced salt solution. The attached cells were then trypsinized and counted in a cell counter (Coulter Electronics, Hialeah, FL). In addition, in separate dishes, the number of adherent or spread cells was assessed by light microscopy after fixation in 3% formaldehyde and staining with hematoxylin. All determinations were carried out in triplicate and each experiment was repeated twice. The final results describing SMC attachment were expressed as a percentage of the total number of initially plated cells. To assure the specificity of the EBP in promoting SMC attachment to elastin, we preincubated the Ao SMC for 2 h prior to plating the cells on the elastin-coated dishes with VGVAPG (50 pg/ ml), the synthetic elastin peptide that binds the EBP [33,40] or with GRGDS (50 pg/ml), a synthetic peptide that occupies integrin receptors and prevents binding of fibronectin [41, 421. In experiments studying the influence of GAGS on SMC attachment to the elastin substrate, Ao SMC were suspended in control medium alone or with added CS (200-800 pg/ml) which detaches EBP or HS (200-800 pg/ml) which does not detach EBP. Since it was clear that over time, with the lower dose of CS (200 pg/ml), cells began to reattach, additional studies were carried out to determine if this might reflect resynthesis of the EBP as well as transcellular transport. The attachment studies were therefore repeated with cycloheximide (10 @g/ml) to prevent protein synthesis or with sodium azide (0.01%) to prevent exocytosis. In some studies, CS or HS (200 pg/ml) was added to the medium 4 h after plating the Ao SMC to observe whether the attached cells could be detached following release of the EBP. To prevent reattachment, we gave repeated doses of CS at 6 and 8 h. To address whether trans-Golgi transport of the EBP might be involved in its cell surface replacement and in the reattachment of SMC to elastin, CS (200 pg/ml) was added to Ao SMC after 4 h of attachment together with monensin (5 X lo-* M). Smooth muscle cell migration through elastic membranes. Using elastin membranes prepared from sheep aorta in a novel in uitro assay, we tested the functional relationship between Ao SMC migration and attachment to elastin. We expected that SMC migration into elastin membranes might provide a more relevant biological model, vis B vis intimal proliferation. Elastic membranes were prepared by boiling adult female ovine aortas for 45 min in 0.1 N NaOH. This procedure removes all cellular and extracellular components of the vessel wall except insoluble elastin. After an additional 1 h extraction with 1 M NaCI, extensive washing in water, PBS, and Medium 199, the membranes were cut into round pieces to fit snugly in the bottom of 24-well culture dishes, then stored at 4°C in Medium 199. DA and Ao SMC were trypsinized as described above, extensively washed in Medium 199 containingO.l% soybean trypsin inhibitor and 5% FBS, incubated for 3 h with the same medium, and then resuspended in a concentration of lo5 cells/ml in fresh control medium or medium to which GAGS CS or HS (400 pg/ml) or GRGDS synthetic peptides (50 pg/ml) were added. The cell suspensions were then plated on the top of elastic membranes in the tissue culture wells and incubated for 3 or 7 days after the medium had been changed. At the end of each experiment, the membranes were fixed in 2% paraformaldehyde and 2% formaldehyde in PBS and transverse histological sections were prepared. The number of cells which were present on the top of the membrane versus the number of cells that had migrated inside and the depth to which they had migrated, i.e., ~200 srn, 200400 pm, >400 pm, were assessed in three randomly selected fields from each of three sections taken from a single membrane and mean values from three different membranes were compared. To determine whether an increased number of cells in the membranes represented migration or proliferation, cells were also stained with the anti-proliferating cell nuclear antigen (PCNA-cyclin) antibody which detects proliferating cells [43, 441. Sections from intraductal human breast carcinoma were used as a positive control. sulfate and smooth muscle fibronectin production. Chondroitin Since we observed that CS increases SMC migration through elastic
AND
RABINOVITCH
membranes and since increased fibronectin synthesis is important in stimulating SMC migration [24], we also addressed the possibility that the two factors might be related. We assessed the influence of GAGS on fibronectin production in confluent cultures of Ao SMC. Cells (1 X lo6 cells/dish) were initially cultured for 24 h in Medium 199 supplemented with 20 mM Hepes, 1% antibiotics/antimycotics, and 10% FBS with and without exogenous GAGS added. These included chondroitin sulfate and heparan sulfate, each at a concentration of 400 fig/ml. The cells were then radiolabeled with [35S]methionine (10 gCi/ml) for the following 24 h without changing the medium. At the end of this incubation, the conditioned media were colIected, proteinase inhibitors were added, and fibronectin extraction using gelatin 4B-Sepharose micro-columns was carried out as previously described [45]. The proteins bound to the gelatin-sepharose beads were suspended in SDS-sample buffer, boiled for 3 min, then resolved on 5% SDS-PAGE gels and autoradiographed. Fibronectin appeared as a doublet at molecular weight 220 kDa. Using the autoradiograph as a template, these bands were cut from the gel and radioactivity was determined by liquid scintillation spectrometry. The counts were normalized for DNA content isolated from the cell layers of the same cultures. DNA was determined in triplicate from each culture using bisbenamide (Hoechst Reagent H3313, Calbiochem) as previously described [46]. Analysis of data In all studies, means and standard deviations were calculated for each group and statistical analyses of the results were carried out by ANOVA followed by the Duncan test of multiple comparisons to establish which groups were different. RESULTS
Immunostaining for Chondroitin Ductus Arteriosus
Sulfate in Aorta and
Immunohistochemistry using the monoclonal antibody to CS applied to Ao and DA tissue sections from 138-day gestation lamb fetuses showed a strong positive reaction in the DA intimal cushion with some staining also apparent in the media (Fig. la). In the Ao, there was only weak immunostaining localized primarily to the subendothelial region (Fig. lb). Glycosaminoglycans and Elastin Smooth Muscle Cells
Binding
Protein
on
We established that Ao SMC exposed for only 15 min to CS showed scant, punctate, and focal distribution of the EBP (Fig. 2b), while Ao SMC kept in control medium demonstrated strong fibrillar immunostaining over the entire cell surface (Fig. 2a). The normal pattern of EBP distribution could be restored following 30 min reincubation in normal (CS-free) medium (Fig. 2~). These results were also apparent after a l-h reincubation period (data not shown). This suggested that SMC can quickly replace the EBP released from their surfaces by CS treatment. Glycosaminoglycans and Smooth Muscle Cell Attachment to Elastin There was reduced attachment to elastin-coated dishes observed in EBP-deficient DA SMC compared to Ao cells observed at each 2 h time point over the 10-h
CHONDROITIN
SULFATE
AND
SMOOTH
MUSCLE
MIGRATION
347
FIG. 1. Representative peroxidaselantiperoxidase immunostaining of ductus arteriosus (a) and aortic tissue (b) sections from 13%dayold lamb fetus using a monoclonal antibody to CS. The region of the intimal cushion observed in the late gestation ductus arteriosus shows a highly positive reaction and there is also some staining throughout the media (a), while in the aorta a weakly positive reaction is observed primarily in the subendothelial region (b). Magnification X40.
incubation period studied (Fig. 3). In addition, spreading of all Ao SMC on elastin was evident by 4 or 6 h, whereas in DA SMC this was not observed for at least 10 h (data not shown). Chondroitin sulfate-induced release of the EBP from the SMC surfaces decreased Ao SMC
attachment to elastin in a dose-dependent manner (Fig. 4) and also delayed spreading of the attached cells (data not shown). When the cells were incubated with 200 pg/ml of CS, there was an initial inhibition of cell attachment to elastin, then a gradual increase toward
FIG. 2. Representative cultured Ao SMC immunostained with the monoclonal antibody to the 67-kDa EBP using FITC. Ao SMC kept in control medium demonstrate strong fibrillar immunostaining over the entire cell surface (a). Aorta SMC treated for 15 min with CS (400 pg/ml) demonstrate only sparse and punctate immunostaining (b). The fibrillar distribution of the 67-kDa EBP is restored on the cell surface after Ao SMC previously treated with CS were kept in normal medium for 30 min (c). Magnification X400.
348
HINEK, 140
-D-
AoSMC
--A--
DASMC
1
BOYLE,
AND
RABINOVITCH 140 T 6 t
,20
u
Control
--k-
VGVAPG
+
!
GRGDS *
p
CELL
RESEARCH
203, 344-353 (1992)
Vascular Smooth Muscle Cell Detachment from Elastin and Migration through Elastic Laminae Is Promoted by Chondroitin Sulfate-Induced “Shedding” of the 67-kDa Cell Surface Elastin Binding Protein’ ALEKSANDER The Division
HINEK, JENNIFER BOYLE, AND MARLENE RABINOVITCH~
of Cardiovascular
Research, Research Institute, The Hospital for Sick Children, and Departments and Pediatrics, University of Toronto, Toronto, Ontario, Canada M5G 1X8
was necessary to replace the EBP molecules released from the cell surface by CS treatment. In the migration assay, both DA and Ao SMC attached to the top of an elastin membrane, but only DA SMC deficient in EBP migrated through the laminae. Addition of CS, which induced shedding of EBP, resulted in Ao SMC migration associated with increased synthesis of fibronectin. We postulate that C&induced release of EBP from SMC surfaces causes cell detachment from elastin and an increase in fibronectin synthesis, processes which may be critical in promoting SMC migration associated with intimal thickening developmentally in the DA and perhaps also in vascular disease. er law Academic PMSS, IW.
Impaired elastin fiber assembly is observed in the fetal ductus arteriosus (DA), associated with a reduced concentration of elastin binding protein (EBP), a 67kDa galactolectin. It is also seen in cultured aortic (Ao) smooth muscle cells (SMC) following the release of the EBP by glycosaminoglycans rich in N-acetylgalactosamine, such as chondroitin sulfate (CS). In the DA, impaired elastin fiber assembly is observed in conjunction with intimal thickening associated with increased migration of SMC into the subendothelium, a feature we previously related to increased production of fibronectin. In this report, we determined whether SMC use the EBP to attach to an elastin substrate, whether shedding of the EBP promotes SMC migration through a threedimensional network of pure elastic laminae prepared from sheep aorta, and whether the latter is associated with increased production of fibronectin. We observed reduced attachment to elastin-coated surfaces of DA SMC deficient in EBP compared to Ao SMC. Addition of CS but not heparan sulfate (a glycosaminoglycan which does not induce EBP shedding) decreased Ao SMC atas did preincubation. with tachment to elastin, VGVAPG elastin-derived peptides which saturate the EBP. The immunolocalization of cell surface EBP suggested that cells can quickly replace EBP released from their surfaces by CS treatment. The magnitude of CSinduced impaired attachment of SMC to elastin was dose dependent and could be further increased by the administration of cyclohexamide and sodium azide. Also, the reversibility of CS-induced detachment was prevented by monensin. This suggests that a process of new synthesis and intracellular transport of the EBP
INTRODUCTION
’ Supported by a grant from the Medical Research Council of Canada, MT 8546, and a grant from the Heart and Stroke Foundation of Ontario, B-2040. This study was presented in part at the 31st Annual Meeting of American Society of Cell Biology, Boston, MA, December, 1991, and was published in abstract form (J. Cell. Biol. 115:441a, 1991). ’ To whom reprint requests should be addressed at Division of Cardiovascular Research, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, Canada M5G 1X8. Fax: (416) 813-7480. 0014.4827192 $5.00 Copynght 0 1992 by Academic Press, All rights of reproduction in any form
344 Inc. reserved
of Pathology
The process of intimal thickening is associated with atherosclerosis and advanced pulmonary vascular disease. “Intimal cushions” are also formed in the late gestation ductus arteriosus (DA) [I-3]. Under all these conditions, migration of smooth muscle cells (SMC) into the subendothelium is accompanied by certain common changes in the composition and organization of the extracellular matrix. Our attention has been focused on the fact that in atherosclerotic [4-121 or wounded vessels [ 13-161, in systemic hypertension [ 171, in experimental pulmonary vascular disease [ 18-191, and in the DA [20-241, SMC migration into the subendothelium is accompanied by impaired assembly of newly synthesized elastin and increased accumulation of N-acetylgalactosamine-containing glycosaminoglycans (GAGS), especially chondroitin sulfate (CS) and dermatan sulfate (DS) in the vessel wall. It has been postulated that CS- and DS-containing proteoglycans inhibit the attachment of a variety of cells to numerous components of the extracellular matrix such as collagen type I, vitronectin, fibronectin, and laminin [25-291. Our previous in vitro studies have shown a direct association between an excess of free N-acetylgalactosamine-containing GAGS, such as CS
CHONDROITIN
SIJLFATE
AND
and DS. release of the 67-kDa elastin binding protein (EBP) from vascular smooth muscle cell (SMC) surfaces and impaired assembly of elastic fibers 1301. This 67-kDa peripheral membrane protein with both elastin binding and /3-galactosugar binding properties forms a receptor complex with two integral membrane proteins (55 and 61 kDa) which allow a transmembrane link between the extracellular ligand and the cytoskeleton [31]. Of particular interest is the observation that binding P-galactosugars to the lectin site of the 67-kDa protein decreases its affinity for both elastin and cell membrane components of the receptor complex resulting in the release of bound tropoelastin and in the dissociation of the 67-kDa protein from the cell membrane [30-331. The 67-kDa EBP binds the VGVAPG hydrophobic sequence on tropoelastin and appears to guide its assembly through a highly coordinated mechanism in which binding of galactosylated microfibrillar proteins to the lectin site of the 67-kDa EBP releases bound tropoelastin onto the microfibrillar scaffold of the growing elastic fiber [30,31]. It has not been determined whether SMC use this EBP to attach to insoluble elastin, i.e., whether increased release of the EBP, in addition to preventing normal assembly of elastin, will detach the SMC from elastin, nor has it been determined whether release of the EBP will facilitate SMC migration through fenestrations in the elastic laminae. In this report we used CS to promote the release of the EBP from Ao SMC surfaces and observed that its abnormal loss prevented Ao SMC attachment to an elastin substratum and promoted their migration through a three-dimensional network of nure elastic laminae. To confirm a direct cause and effect relationship, we demonstrated loss of Ao SMC attachment to immobilized elastin by saturating the EBP with free VGVAPG elastin peptides and we also showed that CS could induce the detachment of previously attached cells. This process was reversible, but the reversibility required replacement of the EBP, a function of resynthesis and transcellular transport to cell surfaces. Ductus arteriosus but not aortic SMC migrated through elastic laminae, unless the latter were treated with CS. Finally, the CS-induced release of the EBP, in addition to preventing attachment of SMC to elastin, also increased SMC production of fibronectin, a feature which we have related to SMC migration associated with intimal thickening [35] in the DA. METHODS Materials. Chemicals and reagents were obtained as follows. Medium 199, PBS, fetal bovine serum (FBS), and other tissue culture reagents were obtained from GIBCO (Burlington, Ontario). Chondroitin sulfate A from bovine trachea, heparan sulfate from bovine kidney, monensin, cycloheximide, sodium azide, trypsin, soybean trypsin inhibitor, VGVAPG elastin peptides, and GRGDS synthetic peptides, and all reagent grade chemicals, were purchased from
SMOOTH
MUSCLE
MIGRATION
345
Sigma (St. Louis, MO). Monoclonal antibody to chondroitin sulfate clone CS-56 was also obtained from Sigma and “BCZ” monoclonal antibody to the elastin binding protein [36] was a gift from Dr. R. P. Mecham, Washington University (St. Louis, MO). Monoclonal antibody to the proliferating cell nuclear antigen (anti-PCNA) was purchased from Coulter Immunology (Hialeah, FL). Species and typespecific fluorescein (FITC)-conjugated goat anti-mouse (GAM) secondary antibody was supplied by ZYMED Laboratories, Inc. (San Francisco, CA). The peroxidase/antiperoxidase staining kit was obtained from Dako (Santa Barbara, CA). Radioactive [i4C]methionine was supplied by New England Nuclear (Boston, MA) and K-elastin from Elastin Products (Pacific, MO). Gelatin 4BSepharose was purchased from Pharmacia (Uppsala, Sweden). Cell cultures. Smooth muscle cells (SMC) were harvested from the DA and Ao of 138-day gestation fetal Rambouillet lambs by a protocol previously described [37]. The time point chosen was concurrent with the formation of intimal cushions in uiuo. Smooth muscle cells were grown initially from explants of the media as described by Ross [38] and then further passaged by trypsinization. All cultures were maintained in Medium 199 supplemented with 20 mA4 Hepes, 1% antibiotics/antimycotics and 5% FBS. The cells from the second passage which were characterized as smooth muscle using a monoclonal antibody specific to smooth muscle actin [39] were used in all exneriments described below. ;mmunohistochemistry, To confirm the increase in CS synthesis previously reported in DA SMC in uitro [24], light microscopic localization of CS was carried out on DA and Ao paraffin-embedded tissue from three different lambs using a monoclonal antibody to CS (5 mg/ml initial concentration, diluted 1:500). The reaction was developed with the peroxidase/anti-peroxidase (PAP) kit according to the manufacturer’s instructions. We also used immunofluorescence to assess whether CS-induced release of the 67.kDa EBP from Ao SMC surfaces [30] was reversible. Cultured Ao SMC were either incubated for 15 min in the presence of CS (200 @g/ml) added to Medium 199 +lO% FBS and fixed or fresh medium without CS was replaced and incubation was continued for an additional 30 min or 1 h prior to fixation. For immunofluorescent localization of the EBP, the nonpermeabilized cells were fixed for 2 min with 0.5% paraformaldehyde, washed in PBS, blocked with 0.5 M glycine and 1% BSA in PBS, and incubated with the monoclonal antibody to the 67.kDa EBP, 5 mg/ml, diluted 1:200 1361. The cells were then stained with fluorescein-labeled goat anti-mouse immunoglobulin (GAM-FITC) diluted 1:lOO. As a control for studies using the monoclonal antibody, normal ascitic fluid was substituted for the primary antibody. Additional controls included secondary antibody alone. Smooth muscle cell adhesion to elastin andglycosaminoglycans. To document whether the presence of the 67.kDa EBP might be of importance in cell attachment to elastin, we compared EBP-deficient DA SMC with Ao SMC. We also assessed Ao SMC in the presence of CS which induces shedding of the EBP and heparan sulfate (HS), an N-acetylglucosamine glycosaminoglycan which does not induce EBP shedding. Ductus arteriosus or Ao SMC were harvested by 5 min exposure to 0.25% trypsin and 0.02% EDTA in Ca*+ and Mg2+ free phosphate-buffered saline (PBS), sedimented by centrifugation, washed twice in PBS with 0.1% soybean trypsin inhibitor, counted in a Coulter counter, and further incubated for 1 h before plating. The cells were then resuspended in Hanks’ balanced salt solution supplemented with 20 mM Hepes, pH 7.4, and plated (1 X lo5 cells/ml) on 24-well culture dishes coated with K-elastin and on bacteriological petri dishes (Costar) which do not support cell adhesion in the absence of coating. To coat the dishes, K-elastin was dissolved in 0.1 M sodium bicarbonate buffer, added to the 35.mm dishes in a final concentration of 500 rgldish, then air-dried at room temperature for 18 h. Aorta and DA cells plated on elastin-coated dishes were incubated at 37°C for different periods of time between 2 and 10 h. At the end of each time
346
HINEK,
BOYLE,
interval, the medium containing unattached cells was removed and the dishes were washed twice with Hanks’ balanced salt solution. The attached cells were then trypsinized and counted in a cell counter (Coulter Electronics, Hialeah, FL). In addition, in separate dishes, the number of adherent or spread cells was assessed by light microscopy after fixation in 3% formaldehyde and staining with hematoxylin. All determinations were carried out in triplicate and each experiment was repeated twice. The final results describing SMC attachment were expressed as a percentage of the total number of initially plated cells. To assure the specificity of the EBP in promoting SMC attachment to elastin, we preincubated the Ao SMC for 2 h prior to plating the cells on the elastin-coated dishes with VGVAPG (50 pg/ ml), the synthetic elastin peptide that binds the EBP [33,40] or with GRGDS (50 pg/ml), a synthetic peptide that occupies integrin receptors and prevents binding of fibronectin [41, 421. In experiments studying the influence of GAGS on SMC attachment to the elastin substrate, Ao SMC were suspended in control medium alone or with added CS (200-800 pg/ml) which detaches EBP or HS (200-800 pg/ml) which does not detach EBP. Since it was clear that over time, with the lower dose of CS (200 pg/ml), cells began to reattach, additional studies were carried out to determine if this might reflect resynthesis of the EBP as well as transcellular transport. The attachment studies were therefore repeated with cycloheximide (10 @g/ml) to prevent protein synthesis or with sodium azide (0.01%) to prevent exocytosis. In some studies, CS or HS (200 pg/ml) was added to the medium 4 h after plating the Ao SMC to observe whether the attached cells could be detached following release of the EBP. To prevent reattachment, we gave repeated doses of CS at 6 and 8 h. To address whether trans-Golgi transport of the EBP might be involved in its cell surface replacement and in the reattachment of SMC to elastin, CS (200 pg/ml) was added to Ao SMC after 4 h of attachment together with monensin (5 X lo-* M). Smooth muscle cell migration through elastic membranes. Using elastin membranes prepared from sheep aorta in a novel in uitro assay, we tested the functional relationship between Ao SMC migration and attachment to elastin. We expected that SMC migration into elastin membranes might provide a more relevant biological model, vis B vis intimal proliferation. Elastic membranes were prepared by boiling adult female ovine aortas for 45 min in 0.1 N NaOH. This procedure removes all cellular and extracellular components of the vessel wall except insoluble elastin. After an additional 1 h extraction with 1 M NaCI, extensive washing in water, PBS, and Medium 199, the membranes were cut into round pieces to fit snugly in the bottom of 24-well culture dishes, then stored at 4°C in Medium 199. DA and Ao SMC were trypsinized as described above, extensively washed in Medium 199 containingO.l% soybean trypsin inhibitor and 5% FBS, incubated for 3 h with the same medium, and then resuspended in a concentration of lo5 cells/ml in fresh control medium or medium to which GAGS CS or HS (400 pg/ml) or GRGDS synthetic peptides (50 pg/ml) were added. The cell suspensions were then plated on the top of elastic membranes in the tissue culture wells and incubated for 3 or 7 days after the medium had been changed. At the end of each experiment, the membranes were fixed in 2% paraformaldehyde and 2% formaldehyde in PBS and transverse histological sections were prepared. The number of cells which were present on the top of the membrane versus the number of cells that had migrated inside and the depth to which they had migrated, i.e., ~200 srn, 200400 pm, >400 pm, were assessed in three randomly selected fields from each of three sections taken from a single membrane and mean values from three different membranes were compared. To determine whether an increased number of cells in the membranes represented migration or proliferation, cells were also stained with the anti-proliferating cell nuclear antigen (PCNA-cyclin) antibody which detects proliferating cells [43, 441. Sections from intraductal human breast carcinoma were used as a positive control. sulfate and smooth muscle fibronectin production. Chondroitin Since we observed that CS increases SMC migration through elastic
AND
RABINOVITCH
membranes and since increased fibronectin synthesis is important in stimulating SMC migration [24], we also addressed the possibility that the two factors might be related. We assessed the influence of GAGS on fibronectin production in confluent cultures of Ao SMC. Cells (1 X lo6 cells/dish) were initially cultured for 24 h in Medium 199 supplemented with 20 mM Hepes, 1% antibiotics/antimycotics, and 10% FBS with and without exogenous GAGS added. These included chondroitin sulfate and heparan sulfate, each at a concentration of 400 fig/ml. The cells were then radiolabeled with [35S]methionine (10 gCi/ml) for the following 24 h without changing the medium. At the end of this incubation, the conditioned media were colIected, proteinase inhibitors were added, and fibronectin extraction using gelatin 4B-Sepharose micro-columns was carried out as previously described [45]. The proteins bound to the gelatin-sepharose beads were suspended in SDS-sample buffer, boiled for 3 min, then resolved on 5% SDS-PAGE gels and autoradiographed. Fibronectin appeared as a doublet at molecular weight 220 kDa. Using the autoradiograph as a template, these bands were cut from the gel and radioactivity was determined by liquid scintillation spectrometry. The counts were normalized for DNA content isolated from the cell layers of the same cultures. DNA was determined in triplicate from each culture using bisbenamide (Hoechst Reagent H3313, Calbiochem) as previously described [46]. Analysis of data In all studies, means and standard deviations were calculated for each group and statistical analyses of the results were carried out by ANOVA followed by the Duncan test of multiple comparisons to establish which groups were different. RESULTS
Immunostaining for Chondroitin Ductus Arteriosus
Sulfate in Aorta and
Immunohistochemistry using the monoclonal antibody to CS applied to Ao and DA tissue sections from 138-day gestation lamb fetuses showed a strong positive reaction in the DA intimal cushion with some staining also apparent in the media (Fig. la). In the Ao, there was only weak immunostaining localized primarily to the subendothelial region (Fig. lb). Glycosaminoglycans and Elastin Smooth Muscle Cells
Binding
Protein
on
We established that Ao SMC exposed for only 15 min to CS showed scant, punctate, and focal distribution of the EBP (Fig. 2b), while Ao SMC kept in control medium demonstrated strong fibrillar immunostaining over the entire cell surface (Fig. 2a). The normal pattern of EBP distribution could be restored following 30 min reincubation in normal (CS-free) medium (Fig. 2~). These results were also apparent after a l-h reincubation period (data not shown). This suggested that SMC can quickly replace the EBP released from their surfaces by CS treatment. Glycosaminoglycans and Smooth Muscle Cell Attachment to Elastin There was reduced attachment to elastin-coated dishes observed in EBP-deficient DA SMC compared to Ao cells observed at each 2 h time point over the 10-h
CHONDROITIN
SULFATE
AND
SMOOTH
MUSCLE
MIGRATION
347
FIG. 1. Representative peroxidaselantiperoxidase immunostaining of ductus arteriosus (a) and aortic tissue (b) sections from 13%dayold lamb fetus using a monoclonal antibody to CS. The region of the intimal cushion observed in the late gestation ductus arteriosus shows a highly positive reaction and there is also some staining throughout the media (a), while in the aorta a weakly positive reaction is observed primarily in the subendothelial region (b). Magnification X40.
incubation period studied (Fig. 3). In addition, spreading of all Ao SMC on elastin was evident by 4 or 6 h, whereas in DA SMC this was not observed for at least 10 h (data not shown). Chondroitin sulfate-induced release of the EBP from the SMC surfaces decreased Ao SMC
attachment to elastin in a dose-dependent manner (Fig. 4) and also delayed spreading of the attached cells (data not shown). When the cells were incubated with 200 pg/ml of CS, there was an initial inhibition of cell attachment to elastin, then a gradual increase toward
FIG. 2. Representative cultured Ao SMC immunostained with the monoclonal antibody to the 67-kDa EBP using FITC. Ao SMC kept in control medium demonstrate strong fibrillar immunostaining over the entire cell surface (a). Aorta SMC treated for 15 min with CS (400 pg/ml) demonstrate only sparse and punctate immunostaining (b). The fibrillar distribution of the 67-kDa EBP is restored on the cell surface after Ao SMC previously treated with CS were kept in normal medium for 30 min (c). Magnification X400.
348
HINEK, 140
-D-
AoSMC
--A--
DASMC
1
BOYLE,
AND
RABINOVITCH 140 T 6 t
,20
u
Control
--k-
VGVAPG
+
!
GRGDS *
p
