514
Biochimica et Biophysica A cta, 544 (1978) 514--528 © Elsevier/North-HollandBiomedicalPress
BBA 28734 METABOLISM OF SULFATED GLYCOSAMINOGLYCANS IN CULTURED ENDOTHELIAL CELLS AND SMOOTH MUSCLE CELLS FROM BOVINE AORTA
GERTRAUD
GAMSE
a, H A N S G E O R G
FROMME
b
and H A N S K R E S S E a
a Institute of Medical Chemistry and Pregl-Laboratory, University of Graz (Austria) and b Institute of Medical Physics, University ofMiinster (F.R.G.)
(Received May 16th, 1978)
Summary The glycosaminoglycan metabolism of cultured endothelial cells and of cells grown from the intima and from the media layer of bovine aorta thoracica was investigated in a comparative study. The following results were obtained: 1. Endothelial cells have in common with intima and media cells the distribution of newly formed sulfated glycosaminoglycans into extracellular, pericellular and intracellular compartments. Endothelial cells, however, synthesize lower amounts of glycosaminoglycans and distribute them in a different ratio into the three pools. 2. Though all the various cell lines synthesize chondroitin 4-sulfate, chondroitin 6~sulfate, dermatan sulfate, heparan sulfate and small amounts of keratan sulfate, endothelial cells exhibit a unique distribution pattern of sulfated glycosaminoglycans in each of the three compartments. Generally, a high proportion of heparan sulfate and chondroitin 6-sulfate and a very low dermatan sulfate content was detected. 3. Heparan sulfate produced by endothelial cells has a higher N-sulfonate content when compared with that from other sources. The cell membrane-associated heparan sulfate, especially, exhibits some heparin-like features as judged by nitrous acid degradation and susceptibility towards heparitinase.
Introduction Endothelial cells which line the lumen of arteries are thought to play important roles in controlling hemostasis and blood vessel permeability [1--3]. Correspondingly, altered strticture and function of endothelial cells might contribute significantly to the pathogenesis of disorders or the arterial wall. In arteriosclerosis, especially, local endothelial injury seems to be a prerequisite
515 for the proliferation of smooth muscle cells in the intima [4]. Detailed knowledge concerning the metabolism of endothelial cells under normal and pathological conditions would therefore be valuable for a better understanding of the pathophysiology of vascular diseases. The cultivation of endothelial cells has greatly facilitated studies on metabolic functions of these cells. Bovine and human endothelial cells have been shown to synthesize material which by morphological, immunological and biochemical criteria behaves like basement membrane collagen [5,6]. With respect to other mesenchymal tissue components, Buonassisi [7] demonstrated that rabbit endothelial cells secrete into the extracellular space sulfated glycosaminoglycans most of which are degradable by chondroitinase ABC. The majority of the intracellular glycosaminoglycans was resistant to degradation by chondroitinase ABC. In a second report the occurrence of heparan sulfate associated with the cell membrane has been described [8]. In this paper we present the results of a more thorough analysis of the formation and composition of topographically different pools of sulfated glycosaminoglycans synthesized by bovine endothelial cells. It will be shown that these cells exhibit a distinct distribution pattern of sulfated glycosaminoglycans when compared with that of cells grown from the intima and the media layer of the aorta of the same animal. Furthermore, heparan sulfate of endothelial cells is characterized by a more "heparin-like" structure than that of the other cell types. Materials and Methods Materials. Tissue culture media and culture flasks were obtained from LSLabor-Service, Munich, except when otherwise stated. Chondroitin ABC lyase (EC 4.2.2.4) and chondroitin AC lyase (EC 4.2.2.5) were purchased from Sigma, St. Louis, Heparitinase (a heparan sulfate degrading lyase, EC 4.2.2.8) and a keratan sulfate degrading endo-~-galactosidase (not classified) were generously provided by Drs. E.A. Davidson (Hershey, Pa., U.S.A.) and S. Suzuki (Tokyo, Japan), respectively. Chondroitin 4-sulfate was prepared from calf nasal cartilage as described [9]. Tissue culture. All cells were grown from the aorta thoracica of approx. 1-year-old oxen. Endothelial cells were obtained according to Eisenstein et al. [10]. This method avoids enzymatic treatment of the tissue since endothelial cells can be collected after slight distension of the aorta filled with culture medium. Cells were cultured in Eagle's Minimum Essential Medium modified as described [11], and supplemented with nonessential amino acids, 10% fetal calf serum, penicillin, streptomycin, and 10 mM N-2-hydroxyethylpiperazine-N-2~thanesulfonic acid (HEPES). The gas phase was 5% CO2 in air. The cultures required approx. 14 days to be ready for subcultivation which was achieved using 30 units trypsin (Boehringer Mannheim) and 0.2 mg EDTA/ml calcium- and magnesium-free Hank's Balanced Salt Solution, 5 min at 37°C. Cultures from the intima and from the media layer of the aorta thoracica of the same animal were initiated according to a procedure published previously [12] and were propagated as described above. For all experiments, confluent cultures from the fourth or fifth passage were used. Phase contrast and electron microscopy. For electron microscopic studies
516 the cultures were washed with Hanks' Balanced Salt Solution and fixed in a phosphate buffered 1% glutaraldehyde solution, pH 7.2, for 2 h at room temperature. After washing the cells with phosphate buffer for 12 h, fixation was continued for 1 h with 3% OsO4 in K2Cr2OT/KOH buffer. Dehydration was carried out in a graded ethanol series. During substitution with 70% ethanol, block-staining with uranyl acetate and phosphotungstic acid was performed. The cells were directly embedded in Epon 812. Thin sections (interference colour: silver) were prepared by the aid of a LKB Ultrotome ancl stained with lead citrate. Electron micrographs were recorded using a Hitachi-H-500 transmission electron microscope with an accelerating voltage of 75 kV. Phase contrast photographs were taken from native cultures using a Zeiss Axiomat IDC inverted microscope. Isolation of radioactive glycosarninoglycans. Cultures of each cell type were processed identically. Confluent cultures were prepared in 75 cm 2 Falcon plastic flasks and incubated with 15 ml culture medium in the presence of 1.8 • 105 Bq (5 ~Ci)/ml [aSS]sulfate for 3 days. Radioactive glycosaminoglycans present in the culture medium (extracellular pool), in the supernatant obtained after trypsinization of the cells (pericellular pool) and in the cell pellet (intracellular pool) were prepared as described previously except that pronase P treatment of the medium was omitted [12]. For investigation of the kinetics of radiosulfate incorporation into sulfated glycosaminoglycans, confluent cultures in 25 cm 2 Falcon plastic flasks were incubated with 5 ml of the radioactive medium mentioned above. The amount of labeled glycosaminoglycans in the extracellular and pericellular pool was determined after dialysis for 4 days against eight changes, 5 1 each, of 0.1 M (NH4)2SO4. Intracellular glycosaminoglycans were quantitated as previously described [9]. In prelabeling experiments the cultures were kept in the presence of radiosulfate for 3 days. The medium was then removed and the cells were washed three times with unlabeled medium. Half of the cultures were trypsinized, centrifuged and replaced in the same volume of medium which contained additionally 10 mM Na:SO4 (chase medium) for dilution of remaining intracellular and membrane-associated inorganic laSS]sulfate. Medium was changed to chase medium in the other cultures. After 2 h of incubation the culture medium was again replaced by chase medium and that point was considered as the beginning of the chase experiment. Distribution pattern of sulfated glycosarninoglycans. The distribution of sulfated glycosaminoglycans was obtained by two methods complementary, in part, to each other. In the first place, parallel degradations of approx. 10 000 cpm each were performed with chondroitin AC lyase and chondroitin ABC lyase, respectively [13] (0.05 units enzyme, total volume 20 ~l, 4 h at 37°C). The digests were spotted on 2043 a Mgl paper (Schleicher and Schi~ll, Dassel), desalted by descending paper chromatography in n-butanol~ethanol~water, 52 : 32 : 16 (v/v, ref. 14) and separated in n-butanol/1 M NH3/glacial acetic acid, 2 : 1 : 3 (v/v). The paper was cut in 1-cm segments for radioactivity determinations. By this method the relative proportion of chondroitin 4- and 6-sulfate and of dermatan sulfate was calculated. In the second place approx. 500 000 cpm sulfated glycosaminoglycans were degraded with 0.1 unit chondroitin ABC lyase as described above. The digest
517 was then made 1% with respect to sodium dodecyl sulfate (SDS), chromatographed on a 1 × 69 cm Sephadex G-50 column equilibrated and eluted with 1% SDS in 0.1 mM EDTA/1 mM mercaptoethanol/1 mM NAN3/0.01 M Tris, pH 7.2, at a flow rate of 5.6 ml/h, fraction volume 1.8 ml. Fractions containing material from the exclusion volume of the column were pooled. 0.5 mg/ml chondroitin 4-sulfate were added as a carrier, and the polysaccharides were precipitated by addition of 3 vols. ethanol containing 1.3% (w/v) potassium acetate. The precipitate was collected by centrifugation, washed twice with ethanol and dissolved in 236/~1 0.085 M sodium acetate, pH 7.0, containing 1.5 mM calcium acetate. This material was digested with 0.8 ag heparitinase for 24 h at 43°C (the amount of enzyme was sufficient to degrade 40 ~g heparan sulfate according to the procedure of Bhavanandan and Davidson, personal communication), treated with SDS and chromatographed on Sephadex G-50 as mentioned above. Again, fractions representing the void volume were pooled and treated with ethanol and potassium acetate. The ethanol-washed precipitate was dissolved in 50 ~1 0.05 M Tris, pH 7.2, subjected to digestion for 24 h at 37°C with an amount of keratan sulfate~legrading endo-~-galactosidase [15] sufficient to liberate 10 ~mol reducing sugar per h and then rechromatographed on the same Sephadex G-50 column. In pilot studies for each enzyme, completeness of digestion was assured by redigesting the respective enzyme resistant material. The proportion of material not susceptible to digestion with chondroitin ABC lyase agreed well, regardless of whether paper or column chromatography was employed. The relative content of heparan sulfate was calculated by subtracting the proportion of endo-~galactosidase-susceptible material from the chondroitin ABC lyase-resistant fraction. Characterization of [3SS]heparan sulfate. Heparan sulfate was obtained from the secretions and from the pericellular pool of cultured cells by chondroitin ABC lyase treatment of [3SS]glycosaminoglycan preparations followed by gel chromatography as described above. [3SS]Heparan sulfate was applied with carrier, approx. 1 ~g each of heparan sulfate and dermatan sulfate, to cellulose acetate strips (Boskamp, Hersel). Electrophoresis was carried out in barbital/ acetate buffer as described [16]. The strips were stained with gentiana violet and cut in l-ram segments for radioactivity determinations according to ref. 15. For the determination of number and sequence of N-sulfated glucosamine residues a minor modification of the procedure of Shively and Conrad was used [17]. Sodium nitrite was employed instead of barium nitrite, and the reaction was terminated by the addition of a molar excess of Na2CO3. The reaction mixture was chromatographed on Sephadex G-50 as described above. Molecular weight determinations of heparan sulfate chains were performed by gel chromatography on a Sephadex G-200 column (1 × 96 cm) equilibrated with 1.0 M NaC1, fraction volume 1.9 ml, flow rate 2 ml/h. The calibration curve of Wasteson [18] was used. Other methods. Protein was determined according to Lowry et al. [19]. The cell number was obtained after counting the cells in a Biirker-Tiirk chamber. Radioactivity determinations were performed in a Beckman LS 8-100 liquid scintillation spectrometer using 0.21% PPO (w/v) and 0.003% dimethyl-POPOP (w/v) in toluene/ethyleneglycol monomethylether (1 : 1, by volume) as scin-
518 tiUation cocktail in case of liquid samples and Unisolve 100 (Zinsser, Frankfurt) in case of paper chromatograms and electropherograms. Results
Morphology of cultured endothelial cells The morphology o f all cultured cell types was routinely monitored by phase contrast microscopy. In primary cultures endothelial cells grew as "island-like" monolayer clusters of closely apposed polygonal cells which were easily distinguished from smooth muscle cells and fibroblasts. After trypsinization at first cells had a fibroblastic appearance b u t became polygonal with increasing cell to cell contact. Confluency of primary cultures was reached after approx. 2 weeks. Electron microscopic observations of endothelial cells after the fourth passage revealed that the cells possessed an even cell membrane without the attachment pods which were found in other types of vascular cells (Fig. 1). Numerous pinosomes with a diameter between 50 and 75 nm were seen, b u t phagosomes and "myelin-like b o d i e s " were rarely observed. Frequently we f o u n d intracytoplasmic fibers, 5--9 nm thick, and microtubuli, diameter 20-30 nm. Most characteristic was the morphology of the cell to cell connection. Special contact zones appeared as 100--200 nm large zonulae adhaerentes. Weibel-Palade bodies have n o t been identified with certainty. According to morphological criteria, smooth muscle cells were obtained from explants of the media layer and "fibroblast-like" cells from the intima layer of bovine aorta thoracica.
Kinetics of radiosulfate incorporation into sulfated glycosaminoglycans Like all types o f cultured cells so far investigated, endothelial cells distrib u t e d newly synthesized sulfated glycosaminoglycans into three main compartments. Some o f the material remained within the cells, some became associated with the outer part of the cell membrane where it could be removed by trypsin treatment. The third part was secreted into the culture medium. From the kinetics of [3SS]sulfate incorporation into the three glycosaminoglycan pools the following results were obtained: 1. Compared with s m o o t h muscle cells, endothelial cells incorporated only a b o u t one third o f radiosulfate into sulfated glycosaminoglycans on the basis of the cell number. This was calculated from the sum of radiosulfate incorporation into the three compartments in short-term incubation experiments (Fig. 2). Since 1 mg cell protein corresponds to 1 . 6 . 1 0 6 endothelial cells b u t to 0 . 6 3 . 1 0 6 smooth muscle cells, the respective values of Fig. 2 must be corrected for the cell number to get an estimation of the rate of synthesis o f these sulfated polymers. 2. The incorporation of the radioactive tracer into the intracellular and pericellular glycosaminoglycans increased with time in a hyperbolic manner in endothelial cells as well as in s m o o t h muscle cells. The time required for constant labeling of the two pools was similar in the t w o cell types. 3. There was a difference between s m o o t h muscle cells and endothelial cells in the relative proportion of [3SS]glycosaminoglycans in the three pools. In several experiments, endothelial cells contained an approximately 5-fold greater
519
Fig. I . Elect~on micrographs of cultured e n d o t h e l i a l cells, cf, c y t o p l a s m i c fibers; g, Golgi apparatus; m, m i t o c h o n d r l a ; mt, m i c r o t u b u l i ; p, p l n o c y t o t i c vesicles; rer, rough endoplasmic refleulum; za, zonulae adhaerentes.
520
o_
9 8:, A
/
>
Biochimica et Biophysica A cta, 544 (1978) 514--528 © Elsevier/North-HollandBiomedicalPress
BBA 28734 METABOLISM OF SULFATED GLYCOSAMINOGLYCANS IN CULTURED ENDOTHELIAL CELLS AND SMOOTH MUSCLE CELLS FROM BOVINE AORTA
GERTRAUD
GAMSE
a, H A N S G E O R G
FROMME
b
and H A N S K R E S S E a
a Institute of Medical Chemistry and Pregl-Laboratory, University of Graz (Austria) and b Institute of Medical Physics, University ofMiinster (F.R.G.)
(Received May 16th, 1978)
Summary The glycosaminoglycan metabolism of cultured endothelial cells and of cells grown from the intima and from the media layer of bovine aorta thoracica was investigated in a comparative study. The following results were obtained: 1. Endothelial cells have in common with intima and media cells the distribution of newly formed sulfated glycosaminoglycans into extracellular, pericellular and intracellular compartments. Endothelial cells, however, synthesize lower amounts of glycosaminoglycans and distribute them in a different ratio into the three pools. 2. Though all the various cell lines synthesize chondroitin 4-sulfate, chondroitin 6~sulfate, dermatan sulfate, heparan sulfate and small amounts of keratan sulfate, endothelial cells exhibit a unique distribution pattern of sulfated glycosaminoglycans in each of the three compartments. Generally, a high proportion of heparan sulfate and chondroitin 6-sulfate and a very low dermatan sulfate content was detected. 3. Heparan sulfate produced by endothelial cells has a higher N-sulfonate content when compared with that from other sources. The cell membrane-associated heparan sulfate, especially, exhibits some heparin-like features as judged by nitrous acid degradation and susceptibility towards heparitinase.
Introduction Endothelial cells which line the lumen of arteries are thought to play important roles in controlling hemostasis and blood vessel permeability [1--3]. Correspondingly, altered strticture and function of endothelial cells might contribute significantly to the pathogenesis of disorders or the arterial wall. In arteriosclerosis, especially, local endothelial injury seems to be a prerequisite
515 for the proliferation of smooth muscle cells in the intima [4]. Detailed knowledge concerning the metabolism of endothelial cells under normal and pathological conditions would therefore be valuable for a better understanding of the pathophysiology of vascular diseases. The cultivation of endothelial cells has greatly facilitated studies on metabolic functions of these cells. Bovine and human endothelial cells have been shown to synthesize material which by morphological, immunological and biochemical criteria behaves like basement membrane collagen [5,6]. With respect to other mesenchymal tissue components, Buonassisi [7] demonstrated that rabbit endothelial cells secrete into the extracellular space sulfated glycosaminoglycans most of which are degradable by chondroitinase ABC. The majority of the intracellular glycosaminoglycans was resistant to degradation by chondroitinase ABC. In a second report the occurrence of heparan sulfate associated with the cell membrane has been described [8]. In this paper we present the results of a more thorough analysis of the formation and composition of topographically different pools of sulfated glycosaminoglycans synthesized by bovine endothelial cells. It will be shown that these cells exhibit a distinct distribution pattern of sulfated glycosaminoglycans when compared with that of cells grown from the intima and the media layer of the aorta of the same animal. Furthermore, heparan sulfate of endothelial cells is characterized by a more "heparin-like" structure than that of the other cell types. Materials and Methods Materials. Tissue culture media and culture flasks were obtained from LSLabor-Service, Munich, except when otherwise stated. Chondroitin ABC lyase (EC 4.2.2.4) and chondroitin AC lyase (EC 4.2.2.5) were purchased from Sigma, St. Louis, Heparitinase (a heparan sulfate degrading lyase, EC 4.2.2.8) and a keratan sulfate degrading endo-~-galactosidase (not classified) were generously provided by Drs. E.A. Davidson (Hershey, Pa., U.S.A.) and S. Suzuki (Tokyo, Japan), respectively. Chondroitin 4-sulfate was prepared from calf nasal cartilage as described [9]. Tissue culture. All cells were grown from the aorta thoracica of approx. 1-year-old oxen. Endothelial cells were obtained according to Eisenstein et al. [10]. This method avoids enzymatic treatment of the tissue since endothelial cells can be collected after slight distension of the aorta filled with culture medium. Cells were cultured in Eagle's Minimum Essential Medium modified as described [11], and supplemented with nonessential amino acids, 10% fetal calf serum, penicillin, streptomycin, and 10 mM N-2-hydroxyethylpiperazine-N-2~thanesulfonic acid (HEPES). The gas phase was 5% CO2 in air. The cultures required approx. 14 days to be ready for subcultivation which was achieved using 30 units trypsin (Boehringer Mannheim) and 0.2 mg EDTA/ml calcium- and magnesium-free Hank's Balanced Salt Solution, 5 min at 37°C. Cultures from the intima and from the media layer of the aorta thoracica of the same animal were initiated according to a procedure published previously [12] and were propagated as described above. For all experiments, confluent cultures from the fourth or fifth passage were used. Phase contrast and electron microscopy. For electron microscopic studies
516 the cultures were washed with Hanks' Balanced Salt Solution and fixed in a phosphate buffered 1% glutaraldehyde solution, pH 7.2, for 2 h at room temperature. After washing the cells with phosphate buffer for 12 h, fixation was continued for 1 h with 3% OsO4 in K2Cr2OT/KOH buffer. Dehydration was carried out in a graded ethanol series. During substitution with 70% ethanol, block-staining with uranyl acetate and phosphotungstic acid was performed. The cells were directly embedded in Epon 812. Thin sections (interference colour: silver) were prepared by the aid of a LKB Ultrotome ancl stained with lead citrate. Electron micrographs were recorded using a Hitachi-H-500 transmission electron microscope with an accelerating voltage of 75 kV. Phase contrast photographs were taken from native cultures using a Zeiss Axiomat IDC inverted microscope. Isolation of radioactive glycosarninoglycans. Cultures of each cell type were processed identically. Confluent cultures were prepared in 75 cm 2 Falcon plastic flasks and incubated with 15 ml culture medium in the presence of 1.8 • 105 Bq (5 ~Ci)/ml [aSS]sulfate for 3 days. Radioactive glycosaminoglycans present in the culture medium (extracellular pool), in the supernatant obtained after trypsinization of the cells (pericellular pool) and in the cell pellet (intracellular pool) were prepared as described previously except that pronase P treatment of the medium was omitted [12]. For investigation of the kinetics of radiosulfate incorporation into sulfated glycosaminoglycans, confluent cultures in 25 cm 2 Falcon plastic flasks were incubated with 5 ml of the radioactive medium mentioned above. The amount of labeled glycosaminoglycans in the extracellular and pericellular pool was determined after dialysis for 4 days against eight changes, 5 1 each, of 0.1 M (NH4)2SO4. Intracellular glycosaminoglycans were quantitated as previously described [9]. In prelabeling experiments the cultures were kept in the presence of radiosulfate for 3 days. The medium was then removed and the cells were washed three times with unlabeled medium. Half of the cultures were trypsinized, centrifuged and replaced in the same volume of medium which contained additionally 10 mM Na:SO4 (chase medium) for dilution of remaining intracellular and membrane-associated inorganic laSS]sulfate. Medium was changed to chase medium in the other cultures. After 2 h of incubation the culture medium was again replaced by chase medium and that point was considered as the beginning of the chase experiment. Distribution pattern of sulfated glycosarninoglycans. The distribution of sulfated glycosaminoglycans was obtained by two methods complementary, in part, to each other. In the first place, parallel degradations of approx. 10 000 cpm each were performed with chondroitin AC lyase and chondroitin ABC lyase, respectively [13] (0.05 units enzyme, total volume 20 ~l, 4 h at 37°C). The digests were spotted on 2043 a Mgl paper (Schleicher and Schi~ll, Dassel), desalted by descending paper chromatography in n-butanol~ethanol~water, 52 : 32 : 16 (v/v, ref. 14) and separated in n-butanol/1 M NH3/glacial acetic acid, 2 : 1 : 3 (v/v). The paper was cut in 1-cm segments for radioactivity determinations. By this method the relative proportion of chondroitin 4- and 6-sulfate and of dermatan sulfate was calculated. In the second place approx. 500 000 cpm sulfated glycosaminoglycans were degraded with 0.1 unit chondroitin ABC lyase as described above. The digest
517 was then made 1% with respect to sodium dodecyl sulfate (SDS), chromatographed on a 1 × 69 cm Sephadex G-50 column equilibrated and eluted with 1% SDS in 0.1 mM EDTA/1 mM mercaptoethanol/1 mM NAN3/0.01 M Tris, pH 7.2, at a flow rate of 5.6 ml/h, fraction volume 1.8 ml. Fractions containing material from the exclusion volume of the column were pooled. 0.5 mg/ml chondroitin 4-sulfate were added as a carrier, and the polysaccharides were precipitated by addition of 3 vols. ethanol containing 1.3% (w/v) potassium acetate. The precipitate was collected by centrifugation, washed twice with ethanol and dissolved in 236/~1 0.085 M sodium acetate, pH 7.0, containing 1.5 mM calcium acetate. This material was digested with 0.8 ag heparitinase for 24 h at 43°C (the amount of enzyme was sufficient to degrade 40 ~g heparan sulfate according to the procedure of Bhavanandan and Davidson, personal communication), treated with SDS and chromatographed on Sephadex G-50 as mentioned above. Again, fractions representing the void volume were pooled and treated with ethanol and potassium acetate. The ethanol-washed precipitate was dissolved in 50 ~1 0.05 M Tris, pH 7.2, subjected to digestion for 24 h at 37°C with an amount of keratan sulfate~legrading endo-~-galactosidase [15] sufficient to liberate 10 ~mol reducing sugar per h and then rechromatographed on the same Sephadex G-50 column. In pilot studies for each enzyme, completeness of digestion was assured by redigesting the respective enzyme resistant material. The proportion of material not susceptible to digestion with chondroitin ABC lyase agreed well, regardless of whether paper or column chromatography was employed. The relative content of heparan sulfate was calculated by subtracting the proportion of endo-~galactosidase-susceptible material from the chondroitin ABC lyase-resistant fraction. Characterization of [3SS]heparan sulfate. Heparan sulfate was obtained from the secretions and from the pericellular pool of cultured cells by chondroitin ABC lyase treatment of [3SS]glycosaminoglycan preparations followed by gel chromatography as described above. [3SS]Heparan sulfate was applied with carrier, approx. 1 ~g each of heparan sulfate and dermatan sulfate, to cellulose acetate strips (Boskamp, Hersel). Electrophoresis was carried out in barbital/ acetate buffer as described [16]. The strips were stained with gentiana violet and cut in l-ram segments for radioactivity determinations according to ref. 15. For the determination of number and sequence of N-sulfated glucosamine residues a minor modification of the procedure of Shively and Conrad was used [17]. Sodium nitrite was employed instead of barium nitrite, and the reaction was terminated by the addition of a molar excess of Na2CO3. The reaction mixture was chromatographed on Sephadex G-50 as described above. Molecular weight determinations of heparan sulfate chains were performed by gel chromatography on a Sephadex G-200 column (1 × 96 cm) equilibrated with 1.0 M NaC1, fraction volume 1.9 ml, flow rate 2 ml/h. The calibration curve of Wasteson [18] was used. Other methods. Protein was determined according to Lowry et al. [19]. The cell number was obtained after counting the cells in a Biirker-Tiirk chamber. Radioactivity determinations were performed in a Beckman LS 8-100 liquid scintillation spectrometer using 0.21% PPO (w/v) and 0.003% dimethyl-POPOP (w/v) in toluene/ethyleneglycol monomethylether (1 : 1, by volume) as scin-
518 tiUation cocktail in case of liquid samples and Unisolve 100 (Zinsser, Frankfurt) in case of paper chromatograms and electropherograms. Results
Morphology of cultured endothelial cells The morphology o f all cultured cell types was routinely monitored by phase contrast microscopy. In primary cultures endothelial cells grew as "island-like" monolayer clusters of closely apposed polygonal cells which were easily distinguished from smooth muscle cells and fibroblasts. After trypsinization at first cells had a fibroblastic appearance b u t became polygonal with increasing cell to cell contact. Confluency of primary cultures was reached after approx. 2 weeks. Electron microscopic observations of endothelial cells after the fourth passage revealed that the cells possessed an even cell membrane without the attachment pods which were found in other types of vascular cells (Fig. 1). Numerous pinosomes with a diameter between 50 and 75 nm were seen, b u t phagosomes and "myelin-like b o d i e s " were rarely observed. Frequently we f o u n d intracytoplasmic fibers, 5--9 nm thick, and microtubuli, diameter 20-30 nm. Most characteristic was the morphology of the cell to cell connection. Special contact zones appeared as 100--200 nm large zonulae adhaerentes. Weibel-Palade bodies have n o t been identified with certainty. According to morphological criteria, smooth muscle cells were obtained from explants of the media layer and "fibroblast-like" cells from the intima layer of bovine aorta thoracica.
Kinetics of radiosulfate incorporation into sulfated glycosaminoglycans Like all types o f cultured cells so far investigated, endothelial cells distrib u t e d newly synthesized sulfated glycosaminoglycans into three main compartments. Some o f the material remained within the cells, some became associated with the outer part of the cell membrane where it could be removed by trypsin treatment. The third part was secreted into the culture medium. From the kinetics of [3SS]sulfate incorporation into the three glycosaminoglycan pools the following results were obtained: 1. Compared with s m o o t h muscle cells, endothelial cells incorporated only a b o u t one third o f radiosulfate into sulfated glycosaminoglycans on the basis of the cell number. This was calculated from the sum of radiosulfate incorporation into the three compartments in short-term incubation experiments (Fig. 2). Since 1 mg cell protein corresponds to 1 . 6 . 1 0 6 endothelial cells b u t to 0 . 6 3 . 1 0 6 smooth muscle cells, the respective values of Fig. 2 must be corrected for the cell number to get an estimation of the rate of synthesis o f these sulfated polymers. 2. The incorporation of the radioactive tracer into the intracellular and pericellular glycosaminoglycans increased with time in a hyperbolic manner in endothelial cells as well as in s m o o t h muscle cells. The time required for constant labeling of the two pools was similar in the t w o cell types. 3. There was a difference between s m o o t h muscle cells and endothelial cells in the relative proportion of [3SS]glycosaminoglycans in the three pools. In several experiments, endothelial cells contained an approximately 5-fold greater
519
Fig. I . Elect~on micrographs of cultured e n d o t h e l i a l cells, cf, c y t o p l a s m i c fibers; g, Golgi apparatus; m, m i t o c h o n d r l a ; mt, m i c r o t u b u l i ; p, p l n o c y t o t i c vesicles; rer, rough endoplasmic refleulum; za, zonulae adhaerentes.
520
o_
9 8:, A
/
>
