Effect of p-Nitrophenyl-P-D-xyloside on Proteoglycan and Glycosaminoglycan Biosynthesis in Rat Serosal Mast Cell Cultures*

Rat serosal mast cells cultured in the presence of heat-inactivated fetal calf serum incorporated (36S) sulfate into heparin proteoglycan of approximately M, = 750,000 after a 3-h pulse and a 2-h chase. P-D- Xyloside (0.1 mm) treatments of cultures of rat mast cells resulted in an insignificant increase in total (36S) sulfate incorporation and the appearance of free gly- cosaminoglycans without a change in proteoglycan size. At higher B-D-xyloside concentrations, total (3SS)sulfate incorporation was inhibited and an increase in the relative glycosaminoglycan content was ob- served concomitant with a reduction in proteoglycan amount and size. As assessed by susceptibility to diges- tion by chondroitinase ABC, hydrolysis by nitrous acid, 13HJhexosamine content, and electrophoretic mobility, only heparin chains were polymerized onto the proteo- glycan core in all cultures. In contrast, individual glycosaminoglycans which appeared only after P-D-xJ~o- side treatment were predominantly chondroitin sulfate rather than heparin, indicating that the fi-D-xyloside acceptor supported polymerization of chondroitin sulfate but not of heparin glycosaminoglycan. Thus, the peptide core is an important determinant for the syn- thesis of heparin glycosaminoglycan by rat peritoneal mast cells. macromolecular fractions were each mixed with 4 M GdnHCl containing protease inhibitors and CsC12 (density 1.53 to 1.55 g/ml), and centrifuged for 48 h at 85,000 X g (28). The resulting dissociative CsCIZ gradients were divided into four approximately equal fractions, differing in their buoyant densities. The bottom fractions which contained proteoglycans and glycosaminoglycans were dialyzed (M, = 3000 cut-off) sequentially against water, 1 M sodium acetate, and water to remove salt and unincorporated radioactivity. The proteoglycans and glycosaminoglycans from [3H]glucosamine-labeled mast cell cultures were subjected to cetylpyridinium chloride precipitation (29) to remove any carbohydrate-rich glycoproteins also present at the bottom of the dissociative CsClz gradient. The fractions were mixed with 0.5 ml of 0.15 M sodium acetate, pH 6.0, containing 250 pg of chondrosarcoma proteoglycan carrier, and ["Hlproteogly- cans and [3H]glycosaminoglycans were precipitated by the addition of 0.5 ml of 2% (w/v) cetylpyridinium chloride. The macromolecules were washed twice with 1% cetylpyridinium chloride in 0.15 M sodium acetate by sedimentation at 10,OOO X g for 10 min, redissolved in 100 pl of 80% (v/v) propanol, converted to their sodium salts by the addition of 50 pl of 10% (w/v) sodium acetate, reprecipitated by the addition of 850 p1 of 4 "C absolute ethanol and 250 p1 of heparin carrier, and recovered by centrifugation.


Rat serosal mast cells cultured in
(3SS)sulfate incorporation was inhibited and an increase in the relative glycosaminoglycan content was observed concomitant with a reduction in proteoglycan amount and size. As assessed by susceptibility to digestion by chondroitinase ABC, hydrolysis by nitrous acid, 13HJhexosamine content, and electrophoretic mobility, only heparin chains were polymerized onto the proteoglycan core in all cultures. In contrast, individual glycosaminoglycans which appeared only after P-D-xJ~oside treatment were predominantly chondroitin sulfate rather than heparin, indicating that the fi-D-xyloside acceptor supported polymerization of chondroitin sulfate but not of heparin glycosaminoglycan. Thus, the peptide core is an important determinant for the synthesis of heparin glycosaminoglycan by rat peritoneal mast cells.
In contrast to the continuously secreted chondrocyte proteoglycan (1,2), rat mast cell proteoglycan resides intracellularly ( 3 , 4 ) and is secreted only upon specific activation of the cell (5). Rat mast cell proteoglycan is approximately M, = 750,000 (3) and possesses a peptide core, estimated a t M, = 12,000, containing h o s t exclusively serine and glycine (6). Heparin glycosaminoglycans are attached 0-glycosidically to serine, a common linkage site for all 0-glycosidically-linked glycosaminoglycans with the possible exception of cartilage keratan sulfate (7, 8). The carbohydrate linkage region of heparin, GlcUA -+ Gal + Gal + Xyl + serine (9),' is reported This work was supported by Grants AI-07722 and RR-05669 from the National Institutes of Health, and by a grant from The Kroc Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.  to be identical to that of chondroitin sulfate (10,11). The addition of the fifth monosaccharide, GlcNAc or GalNAc, is the fEst biosynthetic difference between heparin and the chondroitin sulfate precursor glycosaminoglycan chains. The requirement of a serine-glycine sequence to signal glycosaminoglycan initiation is consistent with other types of glycosylation of proteins requiring specific peptide sequences, such as the transfer of mannose-type oligosaccharides to glycoproteins via dolichol, which requires the peptide sequence asparagine-X-serine or threonine (12). P-D-Xyloside treatment of chondrocytes in vitro results in the polymerization of large amounts of chondroitin sulfate onto the drug (1, 13-17); whereas /?-D-xyloside-treated SV40-transformed Swiss mouse 3T3 fibroblasts (18), mouse mastocytoma tissue (19), and rat hepatocytes (20) polymerize both heparan sulfate-related and chondroitin sulfate-related glycosaminoglycans onto this compound. We now report that P-D-xyloside acts on normal rat serosal mast cells to decrease the size of newly synthesized heparin proteoglycan and to uncover a latent capacity of the cell to polymerize chondroitin sulfate onto the exogenous acceptor. We conclude that more specific initiation requirements than those offered by the /?-D-xyloside are needed for heparin biosynthesis, since relatively few, if any, heparin chains are synthesized onto the drug.
Isolation, Culture and Radiolabeling of Rat Mast Cells-Serosal cells from 40-50 male or female Sprague-Dawley rats, each weighing 150-300 g, were collected by lavage of the peritoneal cavity of each rat with 20 ml of Tyrode's buffer containing 0.1% (w/v) gelatin and 0.005% (w/v) pig mucosa heparin glycosaminoglycan (3). The mast cells were concentrated by isopyknic and isokinetic sedimentations

Mast Cell Proteoglycan Synthesis
blue staining and incorporation of (3sS)sulfate exclusively into heparin. Approximately 0.5 to 1 X lo6 mast cells were obtained per rat.
Mast cells (-1 X IO6) were resuspended in 1 ml of Dulbecco's modified Eagle's medium containing 4 g/Iiter of glucose and supplemented with 15 m~ 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid, 10 m~ N-[tris(hydroxymethyl)methyl-2-amino]ethanesulfonic acid, 10 m~ 2-[bis(2-hydroxyethyl)amino]ethanesulfonic acid (pH 7.21, 100 units/ml of penicillin, 100 units/& of streptomycin, and 15% (v/v) heat-inactivated fetal calf serum (2,23). Mast cells were incubated, generally for 3 h, at 37 "C in a humidified atmosphere of 5% COZ and 95% air in Medium A containing 25-200 pCi of (35S)sulfate/ ml, 50-200 pCi of [3H]serine/ml, or 100 pCi of [3H]glucosamine/ml. P-D-Xyloside, dissolved in dimethyl sulfoxide at a concentration of 80 mg/ml, was added to mast cell cultures to make final solutions of 0.1 mM to 10 mM xyloside. At the end of the labeling period, cells were centrifuged at 400 X g for 4 min and the culture medium was removed. Cell-associated proteoglycans were liberated at 4 "C by adding 100 pi of 1% (w/v) zwittergent 3-12 containing 0.1 M 6-aminohexanoic acid, 0.01 M sodium-EDTA, 0.005 M benzamidine HCI, 0.001 M sodium iodoacetamide, and 0.1 M sodium acetate, followed 30-60 s later by 1 ml of 4 M guanidine HCI containing the same protease inhibitors (24,25). In some experiments, radiolabeled mast cells were washed with I ml of 1 M NaCl in Tyrode's buffer before the zwittergent/GdnHCl disruption step, in order to dissociate any bound pericellular radiolabeled macromolecules (26). Radiolabeled fractions were quantitated directly or were subjected to CsC12 density gradient centrifugation to remove mast cell-derived proteases and glycosidases before characterization of the molecules. A 250-p1 sample from each fraction was mixed with 50 pl of chondrosarcoma proteoglycan or heparin glycosaminoglycan carrier (10 mg/ml), made 4 M in GdnHCl, pH 7.5, individually applied to Sephadex G-25 PD-10 columns (0.8 X 29 cm) and eluted in 0.5 ml fractions with 4 M GdnHC1,O.l M sodium sulfate, 0.1 M Tris-HC1, pH 7.5 (27). Each eluted fraction was mixed with an equal volume of 70% (v/v) ethanol and 12.5 ml of Hydrofluor, and radioactivity was measured with a liquid scintillation counter (Mark I11 liquid scintillation system, Searle Radiographics Inc., Des Plaines, IL). The total radioactivity in the excluded volume, VO, was considered to be incorporated into macromolecules (27). incubated in Medium A with 50 pCi of (35S)sulfate, with or without For pulse-chase experiments, duplicate mast cell cultures were p-D-xyloside. After a 3-h pulse, each mast cell suspension was centrifuged at 400 X g and the radioactive medium removed and replaced with fresh, nonradioactive Medium A. Incubation was continued at 37 "C for various chase times up to 2 h, and the macromolecular j5Sradioactivity appearing in the chase medium and remaining cellassociated was determined. Isolation of Radiolabeled Mast Cell Proteoglycans and Glycosaminoglycans-(35S)sulfate-labeled, [3H]serine-labeled, and [3H]glucosamine-labeled macromolecular fractions were each mixed with 4 M GdnHCl containing protease inhibitors and CsC12 (density 1.53 to 1.55 g/ml), and centrifuged for 48 h at 85,000 X g (28). The resulting dissociative CsCIZ gradients were divided into four approximately equal fractions, differing in their buoyant densities. The bottom fractions which contained proteoglycans and glycosaminoglycans were dialyzed ( M , = 3000 cut-off) sequentially against water, 1 M sodium acetate, and water to remove salt and unincorporated radioactivity.
The proteoglycans and glycosaminoglycans from [3H]glucosaminelabeled mast cell cultures were subjected to cetylpyridinium chloride precipitation (29) to remove any carbohydrate-rich glycoproteins also present at the bottom of the dissociative CsClz gradient. The fractions were mixed with 0.5 ml of 0.15 M sodium acetate, pH 6.0, containing 250 pg of chondrosarcoma proteoglycan carrier, and ["Hlproteoglycans and [3H]glycosaminoglycans were precipitated by the addition of 0.5 ml of 2% (w/v) cetylpyridinium chloride. The macromolecules were washed twice with 1% cetylpyridinium chloride in 0.15 M sodium acetate by sedimentation at 10, OOO X g for 10 min, redissolved in 100 pl of 80% (v/v) propanol, converted to their sodium salts by the addition of 50 pl of 10% (w/v) sodium acetate, reprecipitated by the addition of 850 p1 of 4 "C absolute ethanol and 250 p1 of heparin carrier, and recovered by centrifugation.
Physicochemical and Biochemical Characterization of Mast Cell Proteoglycans and Glycosaminoglycans-Sepharose CL-IB filtration under dissociative conditions was employed to separate glycosaminoglycans polymerized onto the P-D-xyloside from proteoglycan and to determine the hydrodynamic size of radiolabeled proteoglycans. A 250 pl-sample of each detergent/GdnHCI cell extract fraction was applied directly, whereas medium and 1 M NaCl wash fractions were made 4 M in GdnHCl before application to replicate Sepharose rate of 1.5 ml/h with 4 M GdnHCl containing 50 pg/ml of pig mucosa CL-4B columns (0.6 X 120 cm). Each column was eluted at a flow heparin glycosaminoglycan, 0.1 M sodium sulfate and 0.1 M Tris-HCI, pH 7.0. Albumin, 0.25% (w/v), was included in the elution buffer for [3H]glucosamine-labeled samples. The 0.5 ml-column fractions were mixed with an equal volume of 70% ethanol and, after the addition of 12.5 ml of Hydrofluor, were analyzed for radioactivity. For a preparative purification, only one-tenth of each fraction was analyzed for radioactivity, and the proteoglycan-and glycosaminoglycan-containing fractions were pooled separately, dialyzed exhaustively against water, and lyophilized. acid under identical conditions and subjected to gel filtration chromatography, were assessed for changes in size by measurement of uronic acid (31) after the dimethoxyethane was removed. Lyophilized [3H]glucosamine-labeIed and (35S)sulfate-labeled proteoglycans and glycosaminoglycans were treated with chondroitinase ABC (32) and degradation products were determined by cellulose thin layer chromatography. The relative percentage of the total incorporated radioactivity which localized with ADi-4S and ADi-6S carrier indicated the amount of (*S)sulfate or [3H]glucosamine incorporated into chondroitin sulfates. A 5-fold (w/w) excess of heparin glycosaminoglycan over chondrosarcoma proteoglycan did not alter chondroitinase ABC digestion of macromolecular chondroitin sulfate.
[3H]Glucosamine-labeled proteoglycans and glycosaminoglycans were acid hydrolyzed in 8 M HCI for 3 h at 95 "C, and analyzed by automated ion exchange chromatography for distribution of radioactivity into glucosamine and galactosamine (33).
Autoradiography of Labeled Glycosaminoglycans-The (%) sulfate-labeled intracellular proteoglycans and glycosaminoglycans formed in the presence and absence of 13-D-xyloside were purified by density gradient centrifugation. Dialyzed and lyophilized bottom fractions were dissolved in 0.5 ml of 0.05 M NaOH, 1.0 M NaBH4, and heated at 45 "C for 48 h (34). After neutralization with 2 M acetic acid, the liberated 0-glycosidically linked glycosaminoglycans were dialyzed for 6 h against 0.001 M sodium acetate. A 10 pl-sample was electrophoresed on cellulose acetate for 75 min at 185 v in 0.1 M pyridine containing 0.46 M formic acid, pH 3.0 (35). The dried electrophoresis sheet was then sprayed with EN3HANCE and exposed to Kodak XR-5 fdm for 48 h at -80 "C.

The in vitro conditions developed by Handley et al. (23)
and Kimura et al. (2) for cultured chondrocytes were found to be suitable for short term primary cultures of rat serosal mast cells. (35S)Sulfate incorporation into mast cell proteoglycan was dependent on the presence in the culture medium of fetal calf serum (Fig. l), and its effect was heat-stable at 60 "C for 6 h. As assessed by quantitative PD-10 chromatography after 3 h in Medium A, approximately 10-15% of the total radiolabeled macromolecules were released into the culture medium and another 10-20% were displaced from the cell pellet by a 1 M NaCl wash. Treatment of the cell pellet with 4 M GdnHCl solubilized less than 15% of the cell-associated (35S)sulfate macromolecules, whereas a 60-s exposure to zwittergent before 4 M GdnHCl extraction liberated the intracellular mast cell radiolabeled macromolecules, and was routinely used rather than sonication in 1 M NaCl (3) or digestion with pronase ( 6 ) . The zwittergent/GdnHCl extraction procedure also permitted immediate density gradient centrifugation under dissociative conditions so as to remove endogenous mast cell proteases and glycosidases. More than 90% of the (35S)sulfate-labeled macromolecules and all of the chemically determined heparin  serine-labeled macromolecules in the mast cell extract fraction Cells (10') were exposed to 200 pCi/ml of either (%)sulfate ( n = 3) or r3Hlserine ( n = 5). extracted with zwittergent and GdnHCl and were isolated by density gradient centrifugation.
Macromolecular e5S)Sulfate Incorporation in Response to P-~-Xyloside Treatment-Three sets of six mast cell cultures each were prepared in Medium A containing ("S)sulfate. Five members of each set received xyloside at a concentration ranging from 0.1 mM to 10 mM; the sets were then incubated for 3 , 5 , or 20 h. The presence of dimethyl sulfoxide in Medium A, at the concentrations used in P-D-xyloside-treated rat mast cell cultures, had no effect on ("S)sulfate incorporation. At a P-D-xyloside concentration of 0.1 mM, there was an insignificant increase (p > 0.1) of total (35S)sulfate incorporation (Fig.   2), whereas a progressive inhibition of incorporation occurred at higher P-D-xyloside concentrations. Inhibition was significant a t 3 mM (p < 0.005) and at 10 mM fi-D-xyloside, (35S)sulfate incorporation was suppressed by 90% (Fig. 2). The relative percentage of "S-macromolecular radioactivity released into the culture medium (19 f 9%; R = 6)) uersus that associated with the cell pellet did not change appreciably during 20 h of P-D-xyloside treatment under these in vitro conditions.
In control mast cell cultures fortified with heat-inactivated fetal calf serum (Fig. 3B) The carbazole reaction of fraction D4 produced a nonspecific brown color, and only fraction D l was positive for heparin by the Azure A test. were recovered at the bottom of the centrifuge tube with 2-4% of the labeled [3H]serine macromolecules ( Table I). Mast cell proteins were concentrated at the top of the gradient. When cultures were pulsed for 3-h time intervals, the rate of (35S)sulfate incorporation was linear up to approximately 20 h. The age and the sue of rats within the range used in this study, 150-300 g, did not influence overall proteoglycan synthesis. Studies on mast cell proteoglycan biosynthesis were routinely conducted for less than 24 h and in the presence of heat-inactivated fetal calf serum; [35S]macromolecules were xyloside-treated cultures (B-E). All radioactivity eluted in fractions 1-6 before treatment, and the shaded area is the estimated radioactivity resistant to hydrolysis. Recovery of radioactivity was 100% when heparin carrier was added, while recovery was less than 10% without carrier.

A. Medlum
comparable in size to those produced by control cultures (Fig.   3 C ) , and some radioactivity was shifted to a low molecular weight glycosaminoglycan fraction. In the presence of p-Dxyloside concentrations which decrease total (35S)sulfate incorporation, mast cell proteoglycan was progressively smaller in size and the relative percentage of incorporated radioactivity in free glycosaminoglycans increased (Fig. 3, D and E ) .
Sepharose CL-4B chromatography was performed on the macromolecular 35S-radioactivity in the culture medium, in the 1 M NaCl wash of cells, and in the detergent/GdnHCl extract of cells to compare the relative quantities of proteoglycan and free glycosaminoglycans released and retained by rat mast cell cultures when exposed to 1 mM P-D-xyloside. In the culture medium, the ratio of [35S]glycosaminoglycans to [35S]proteoglycans was approximately 4:l as compared to the ratios of 1:4 and 1:2, respectively, for the pericellular and intracellular compartments (Fig. 4). However, about 85% of the total [35S]glycosaminoglycans after a 3-h pulse with radioactivity still remained cell-associated.
Mast cell cultures that were exposed 3 h to 3 mM p-D- did not change in molecular weight (Fig. 5, A-D). In P-Dxyloside-treated cultures there was a similarity of behavior of intracellular glycosaminoglycan and proteoglycan (Fig. 5, E-H), which contrasts to the short retention time of glycosaminoglycans in cell types such as chondrocytes (16,17) and transformed fibroblasts (18), and suggests that the [35S]glycosaminoglycan did not originate in a cell contaminant which continuously secretes proteoglycan.
Characterization of Radiolabeled Proteoglycan and Glycosaminoglycan Fractions in Control a n d P-~-Xyloside-Treated Cultures-("S)Sulfate-labeled intracellular macromolecules from mast cells cultured without and with 0.3 mM P-D-xyloside or 3 mM P-D-xyloside were separated into proteoglycan-enriched and glycosaminoglycan-enriched fractions by Sepharose CL-4B gel filtration and subjected to nitrous acid degradation. The [35S]proteoglycan fraction obtained from the cultures of control rat mast cells were degraded >96% (Fig.  6 A ) as indicated by conversion of macromolecular "S-radioactivity to oligosaccharides which eluted in the included volume from Sephadex G-25 columns. With the addition of 0.3 mM P-D-xyloside to Medium A, h o s t the entire [35S]proteoglycan fraction was again hydrolyzed by nitrous acid to smaller oligosaccharides (Fig. 6 B ) . At a 3 mM P-D-xyloside concentration which further reduced the size and amount of proteoglycan produced, approximately 82% of the "S-macromolecular proteoglycan fraction was degraded (Fig. 6C) indicating that the cultures continued to incorporate ("S)sulfate into heparin chains which were polymerized onto the mast cell proteoglycan core. In contrast, the (35S)sulfate-labeled glycosaminoglycans appearing in the /3-D-xyloside-treated cultures were more than 85% resistant to nitrous acid hydrolysis (Figs. 6, D and E ) . Under the conditions employed, approximately 90% of standard pig mucosa heparin carrier was hydrolyzed to oligosaccharides, while commercial chondroitin sulfate carrier was not susceptible to nitrous acid degradation. When the proteoglycans isolated by density gradient centrifugation and Sepharose CL-4B fitration were incubated with chondroitinase ABC and the liberated chondroitin sulfate disaccharides analyzed by thin layer chromatography, there was insignificant degradation of proteoglycan from control or from 0.3 mM P-D-xyloside-treated cultures and only 10% degradation of the proteoglycan from 3 rn P-D-xyloside-treated cultures (Table 11). On the other hand, the individual glycosaminoglycans induced by P-D-xyloside treatment were more  '' A change of +-5% in the distribution of radioactivity was considered to be within the error of the procedure because there was some overlapping of proteoglycans and glycosaminoglycans on gel filtration (Figs. 3 and 7).
ADi-dS (disulfated disaccharide) refers to the position of the marker and does not necessarily mean that the hydrolysate product is a disaccharide. than 70% digested, and the radioactivity comigrated with the ADi-6S and ADi-4S carriers. In three consecutive experiments with mast cells exposed to 0.1 or 0.3 mM P-D-xyloside, but not to 3.0 m, the chondroitinase ABC-treated glycosaminoglycans contained hydrolysate products which chromatographed in the position of disulfated disaccharides. This material was not further characterized in terms of the number of sulfate residues per disaccharide or the presence of oligosaccharides resistant to chondroitinase ABC. After a 3-h pulse with ['H]glucosamine and subsequent density gradient centrifugation, 62%, 49%, and 42% of the total radioactivity in the cell extract fractions of non-P-D-xyloside, 0.3 mM P-D-xyloside, and 3 mM P-D-xyloside-treated mast cells, respectively, were found in the high buoyant density macromolecules.' As analyzed by Sepharose CL-4B chromatography, the only substantial difference of P-D-xylosidetreated mast cells exposed to [3H]glucosamine (Fig. 7) uersus ("@sulfate (Fig. 3) was a higher relative predominance of "Hradioactivity incorporated into the glycosaminoglycan fraction. Cetylpyridinium chloride precipitation of the density gradient fraction before Sepharose CL-4B chromatography did not change the chromatographic profile, indicating that the 3H-radioactivity was not in glycoproteins. After 8 M HCI hydrolysis, approximately 85% of the total radioactivity in the proteoglycan fraction from the control cultures, or from those maintained in 0.3 nm p-D-xyloside, eluted from ion exchange chromatography with glucosamine carrier, whereas about 85% of the total radioactivity in the glycosaminoglycan fraction from 0.3 mM,!?-D-xyloside-treated mast cells chromatographed as [3H]galactosamine. Thus, the resistance to degradation by Proteoglycan Synthesis nitrous acid and the susceptibility to chondroitinase ABC of the glycosaminoglycan-enriched fraction appearing after p-Dxyloside treatment was not due to an incomplete processing of heparin precursor chains, but to polymerization of chondroitin sulfate glycosaminoglycan onto the exogenous drug.
The total (35S)sulfate-labeled pools of macromolecules from drug-and nondrug-treated rat mast cells were each purified by density gradient centrifugation and subjected to /3-elimination. The glycosaminoglycans were separated by cellulose acetate electrophoresis and analyzed by autoradiography. Glycosaminoglycans from mast cells maintained in 'Medium A were polydispersed in negative charge and co-migrated with pig mucosa heparin carrier (data not shown), whereas those from 3 mM P-D-xyloside-treated cultures contained additional glycosaminoglycans which co-electrophoresed with whale cartilage chondroitin-4-sulfate. DISCUSSION P-D-Xyloside, an acceptor for glycosaminoglycan polymerization, has previously been shown to uncouple glycosaminoglycan synthesis from proteoglycan core synthesis in chondrocyte and fibroblast cultures (1, 13-18) which normally continuously secrete proteoglycan. ,8-D-Xyloside at 0.1-1 m~ concentrations reportedly augmented total (35S)sulfate incorporation by polymerizing chondroitin sulfate onto the drug acceptor without altering the size of the proteoglycan molecule. At higher concentrations (3-10 mM P-D-xyloside) (35S)sulfate incorporation was suppressed, proteoglycans were smaller in size due to the addition of fewer and smaller glycosaminoglycans to the core, and the relative amount of chondroitin sulfate polymerized onto the fl-D-xyloside acceptor was increased.
Rat mast cells, which secrete stored proteoglycan only in response to specific activation ( 5 ) , can be concentrated to >97% purity and contain more than 90% heparin proteoglycan of defined hydrodynamic size and peptide core composition (3,6). The introduction of heat-inactivated fetal calf serum to cultures of rat mast cells accelerated the sulfation and formation of macromolecular heparin proteoglycan such that 3h cultures (Figs. 1 and 3B), rather than 14-h cultures (3), were sufficient to yield labeled completed macromolecular products. Zwittergent detergent, GdnHC1, and protease inhibitors were used to liberate mast cell proteoglycan in order to avoid the necessity of adding exogenous proteases and to minimize the effects of endogenous proteases and glycosidases during cell disruption. In addition, liberated macromolecules were generally subjected to immediate density gradient centrifugation to remove proteases, glycosidases, and glycoproteins (Table I).
P-D-Xyloside produced a dose-related suppression of total (35S)sulfate incorporation (medium and cell pellet) from 0.3 mM to 3 mM concentration, with total suppression a t 10 mM (Fig. 2). There was no net release in cytoplasmic lactate dehydrogenase or in secretory granule &hexosaminidase into the culture medium or apparent alteration in granule density as assessed by phase microscopy or metachromatic staining with up to 3 mM P-D-xyloside. Mast cell cultures exposed to 0.1 and 0.3 mM P-D-xyloside contained a cell-associated proteoglycan with a hydrodynamic size and polydispersity comparable to that of the control cultures, but in addition formed radiolabeled macromolecules which filtered with the size of glycosaminoglycans (Fig. 3). With 1 and 3 mM P-D-xyloside there was a decrease in quantity and sue of cell-associated radiolabeled proteoglycan and a relative increase in the quantity of radiolabeled glycosaminoglycan. When the effect of 1 mM P-D-xyloside was examined for the distribution of proteoglycan relative to glycosaminoglycan, the ratio was approxi-mately 4:l for cell-associated macromolecules (Fig. 4). Approximately 75% of the total macromolecules were intracellular and 15% were cell surface-associated as indicated by 1 M NaCl salt elution before cell disruption. Under conditions of P-D-xyloside treatment, the mast cell glycosaminoglycans had an apparent one-half intracellular retention time of >2 h (Fig. 5), which is substantially longer than chondroitin sulfate polymerized onto proteoglycan or b-D-xyloside by chondrocytes. The lack of extracellular proteoglycan in P-D-xylosidetreated cultures suggests that the glycosaminoglycans attached to the drug are released because either they are not bound to cationic protein in a manner similar to heparin, or because they are not in the same subcellular compartment. Thus, P-D-xyloside treatment of normal rat mast cell cultures decreased (35S)sulfate incorporation into macromolecules, uncoupled glycosaminoglycan synthesis from proteoglycan formation, and caused the preferential and spontaneous release of glycosaminoglycans into the extracellular fluid.
That the glycosaminoglycan polymerized onto the P-D-XYloside acceptor is chondroitin sulfate, whereas that forming on proteoglycan is heparin, was established by chemical and biochemical analyses of separated macromolecules. The [35S] proteoglycan obtained from control cultures by extraction and gel fitration chromatography was more than 95% degraded by nitrous acid (Fig. 6), indicating a glycosaminoglycan composition containing the characteristic N-sulfated glucosamine monosaccharide of heparin and heparan sulfate (30). The proteoglycan synthesized in the presence of 0.3 and 3 mM p-D-xyloside was also heparin by this criterion. The small amount of proteoglycan apparently resistant to degradation by nitrous acid is attributed to a reduced hydrodynamic size, which results in incomplete chromatographic resolution from glycosaminoglycans. The glycosaminoglycans formed in the presence of either 0.3 mhl or 3 mM P-D-xyloside were more than 85% resistant to nitrous acid hydrolysis, indicating that they were not heparin or heparan sulfate. Further evidence that the proteoglycan formed in the presence of 0.3 or even 3 mM p-D-xyloside contained heparin was derived from the electrophoretic mobility of its liberated glycosaminoglycans and from its resistance to degradation by chondroitinase ABC (Table 11). In contrast, the glycosaminoglycans derived from the same P-D-xyloside-treated cultures were 70% degraded by chondroitinase ABC to form chondroitin-6-sulfate and chondroitin-4-sulfate disaccharides. That so much radioactivity was recovered in these individual disaccharide fractions indicated a predominant "all or none" polymerization of chondroitin sulfate onto the exogenous drug. When intracellular [3H]proteoglycans and ["H]glycosaminoglycans were purified by density gradient centrifugation, cetylpyridinium chloride precipitation, and Sepharose CL-4B chromatography (Fig. 7), 85% of the total radioactivity in the glycosaminoglycan fraction from 0.3 and 3.0 mM P-D-xyloside-treated mast cell cultures were identified as E3H]galactosamine. In contrast, 85% of the [3H]hexosamine in the proteoglycan fractions from all cultures was [3H]glucosamine; thus, heparin was polymerized onto proteoglycan, and chondroitin sulfate predominantly onto the P-D-xyloside acceptor. The peptide sequence of serine-glycine at glycosaminoglycan initiation sites and the carbohydrate sequence of GlcUA -Gal Gal -Xyl serine have been reported to be identical for chondroitin sulfate and heparin-rich proteoglycans (10). That the proteoglycan core affects in some manner the choice for addition of the fifth monosaccharide in glycosaminoglycan biosynthesis, GlcNAc or GalNAc, is indicated by the present study in which rat serosal mast cells make heparin proteoglycan but polym-erize chondroitin sulfate rather than heparin onto the competing b-D-xyloside acceptor. The dual capacity to generate both chondroitin sulfate and heparin glycosaminoglycan has long been recognized in mouse mastocytoma cells (36), and the balance is shifted in favor of chondroitin sulfate by b-Dxyloside (19). The finding that the normal rat mast cell consistently synthesizes heparin proteoglycan, despite a proven capacity to polymerize chondroitin sulfate, could indicate that chondroitin sulfate biosynthesis requires a more complex core than the serine-glycine copolymer, that cores are directed to different subcellular compartments specializing in synthesis of different glycosaminoglycans, or that there is a steric inhibition for the chondroitin sulfate polymerizing enzymes by the dense number of heparin initiation sites on heparin proteoglycan core.