a lipid-anchored heparan sulfate proteoglycan is present in the surface of differentiated skeletal...

Eur. J. Biochern. 216, 587-595 (1993) 0 FEBS 1993

A lipid-anchored heparan sulfate proteoglycan is present in the surface of differentiated skeletal muscle cells Isolation and biochemical characterization

Alejandro CAMPOS', Rebeca NUREZ', Cecilia S. KOENIG', David J. CAREY' and Enrique BRANDAN' I Molecular Neurobiology Unit, Department of Cell and Molecular Biology, Faculty of Biological Sciences, Catholic University of Chile,

Santiago, Chile Weis Center for Research, Geisinger Clinic, Danville, USA

(Received March 17/June 15, 1993) - EJB 93 0392/6

We have investigated the presence of hydrophobic membrane-associated heparan sulfate proteoglycans (HSPG) on the cell surface of differentiated skeletal muscle cells. A HSPG releasable by incubation with a phosphatidylinositol-specific phospholipase c (PtdIns-PLC) was obtained. HSPG were also isolated from Triton X-100 extracts of the cells. The hydrodynamic characteristics of the PtdIns-PLC-releasable and detergent-extracted HSPG were indistinguishable. SDSPAGE analysis of the PtdIns-PLC-releasable HSPG indicated a molecular mass of 250 kDa. Analysis of proteins immunoprecipitated by specific antibodies against a HSPG isolated from Schwann cells demonstrated that the antisera precipitated an intact HSPG that was present in the pool of proteins released by PtdIns-PLC and by Triton X-100 from ["S]sulfate labeled cells. Nitrous acid degrada- tion of the immunoprecipitated proteins released by PtdIns-PLC from ["Slmethionine labeled cells produced a single 67-kDa core protein. Analysis of hydrophobicity of the purified HSPG revealed that only the HSPG obtained from the detergent extract were able to be incorporated into the liposomes whereas the PtdIns-PLC-released HSPG was not.

Immunocy tolocalization analysis of the differentiated cells indicated that the PtdIns-PLC-releasable HSPG was located on the cell surface. Prior incubation of the cells with PtdIns-PLC significantly reduced the surface staining. Analysis of skeletal-muscle sections of adult rat skeletal muscles indicated that this HSPG localized exclusively at the endomysium. This localization suggest that these HSPG may be acting as a cell receptor for extracellular-matrix (ECM) components.

Heparan sulfate proteoglycans (HSPG) present on the cell surface can bind a wide variety of ligands, including cell adhesion molecules [l], matrix components [2-41, growth factors [5, 61, enzymes [7, 81 and enzymes inhibitors [9]. This indicates that the presence of HSPG on the cell surface may influence the response of a particular type of cell to changes in the environment. Thus, HSPG are involved in the binding and presentation to the cell of growth factors such as fibroblast-growth factors (FGF) [lo, 111. It has been shown that repression of myogenic differentiation by FGF is dependent on the presence of cellular heparan sulfate, suggesting that the regulation of the expression of these molecules may have important implications for skeletal-muscle development [12]. It has also been suggested that HSPG present at the cell surface are involved in the interaction with extracellular matrix (ECM) and cytoskeleton components [3,

Correspondence to E. Brandan, Department of Cell and Molecu- lar Biology, Faculty of Biological Sciences, Catholic University of Chile, P. 0. Box 114-D, Santiago, Chile

Fax: +56 2 222 5515. Abbreviations. HSPG, heparan sulfate proteoglycan(s) ; FGF, fi-

broblast-growth factors; ECM, extracellular matrix ; glycosyl- Ptdlns, glycosylphosphatidyhnositol; PtdIns-PLC, phosphatidylinositol-specific phospholipase C ; PtdIns, phosphatidylinositol; CPC, cetylpyridinium chloride; GAG, glycosaminoglycans.

131, making these molecules key elements in the interaction of the cell and the extracellular media.

Biochemical and molecular biological studies have pro- vided evidence for various modes of attachment of HSPG to the cell membrane. HSPG can be anchored by a transmem- brane core protein [ 14, 151 or a glycosylphosphatidylinositol (glycosyl-PtdIns) anchor [13, 161. In addition to the tightly membrane-associated form of HSPG, cells also contain pe- ripherally associated HSPG that can be solubilized with salt, heparin, or anionic molecules [17]. These peripheral forms seems to be synthesized as integral component that are pro- cessed to peripheral forms at the plasma membrane [17, 181.

Several studies indicate that the ECM is vital for normal myogenesis [19, 201 but the exact mechanisms through which it modulates this process remain to be established. It has been shown that inhibitors of collagen synthesis, [21] as well as antibodies against integrin receptor, or synthetic peptides which inhibit integrin-receptor function, block the fusion and differentiation of muscle cells [22, 231. The presence of proteoglycans is essential in order to have a func- tional ECM [24], consequently we have been interested in the identification of proteoglycans synthesized by skeletal- muscle cells and an analysis of the possible functions of these molecules on the cell surface. We have demonstrated the presence of several forms of proteoglycans synthesized and secreted by differentiated muscle cells [25, 261. We have

588

found that the level of expression of the proteoglycan decorin, present in the skeletal-muscle ECM, is influenced by the presence of the inervating motor nerve [27,28]. There is, however, no information concerning the structure, function or regulation of the expression of plasma-membrane-associated proteoglycans in skeletal muscles. In this study we de- scribe the purification and biochemical characterization from in vitro differentiated skeletal-muscle cells of a membrane bound HSPG, that can be releasable from the cell surface by PtdIns-PLC. This HSPG localizes in adult skeletal muscles with laminin at the basal lamina.

MATERIALS AND METHODS

Materials The following material were purchased from the suppli-

ers indicated. [35S]Na,S04, carrier free was obtained from New England Nuclear. [3sS]methionine (1 000 Ci/mmol) was obtained from Amersham. [3H]PtdIns was a kind gift of Dr Miguel Bronfman (Catholic University of Chile, Santiago, Chile). Chondroitinase ABC lyase, Protein-A- Sepharose and benzamidine were from Sigma Chemical Co. Hepariti- nase was from Miles Laboratories. PtdIns-PLC purified from Bacillus thuringiensis was a kind gift of Dr Martin Low (Co- lumbia University, New York, USA). C,C,:: cells, a myogenic cell line isolated from mouse skeletal muscle [29], was obtained from the American Type Culture Collection. Other reagents were obtained from commercial sources.

Methods Cell cultures

The C,C,, cells were grown as described previously 1251, except that the fusion medium consisted of 98% Dulbecco's modified Eagle's medium with 1 g/l glucose supplemented with 2% horse serum. The studies were performed with differentiated cells, which were maintained in fusion medium for 6-7 days.

Labeling of cultures and analysis of proteoglycans

35-mm dishes containing contractile differentiated myotubes were radiolabeled in the presence of 50 pCi ["Slsulfate for 18 h in CRC-30 medium with 2% horse serum. The incubation media was removed and the cells were rinsed three times with ice-cold NaClP, (0.15 M NaC'1, 0.05 M sodium phosphate, pH7.5) then extracted for 15 min at 0°C with 10 mM phytic acid, 0.1 M Tris/HCl, pH 7.5, in the presence of the following protease inhibitors: 0.05 M 6-aminohexa- noic acid, 0.01 M EDTA, 0.005 M n-ethylmaleimide, 0.005 M benzamidine and 0.001 M phenylmethylsulfonyl fluoride. Following this treatment the cells were extracted with NaClP, containing 1% Triton X-100 in the presence of the above protease inhibitors, for 5 min at O"C, with gentle rotation. The plates were then extracted with the same detergent buffer in the presence of 0.5 M KC1. Finally the plates were scraped, using a rubber policeman, in the presence of 1 % SDS, 0.05 M Tris/HCl, pH 7.5, or 4 M guanidinium chloride, 0.05 M sodium acetate, pH 5.8. These two solutions removed essentially the same amount of radioactive material. Alternatively, the cells were incubated as above with the phytic acid solution, followed by incubation with 2.5 U/ml PtdIns-PLC or 50 pg/ml trypsin at 37°C for 15 min. [3sS]su1-

fate incorporation into macromolecules was evaluated by precipitation with cold 12% trichloroacetic acid for the phytic-acid extraction and enzymic treatments, or by binding to impregnated cetylpyridinium chloride (CPC) filters for the detergent extractions, as described [30, 311 ; similar and highly reproducible results were found using both methods. The radioactive material obtained after PtdIns-PLC treatment or Triton X-100 extraction was applied to a 3.0 ml DEAE- Sephacel column equilibrated in 0.05 M Tris/HCl, pH 7.4, 0.1 M NaCl, 1% Triton X-100 and protease inhibitors at a flow rate of 5.0 ml/h. The column was washed with 10 column volumes of the same buffer then eluted with a linear NaCl gradient from 0.1 M to 1.0 M NaCl in the same buffer (60 ml total volume). Elution was performed at a flow rate of 5.0 ml/h and fractions of 1 .O ml were collected. The NaCl gradient was monitored by measuring conductivity of the fractions. The different fractions were pooled, dialyzed for 12 h against 10 mM Tris/HCl, pH 7,4, 0.5% Triton X-100 and 0.1 M NaCl containing protease inhibitors and concentrated by dialysis against poly(ethy1ene glycol) (average 8 kDa).

Filtration chromatography

Pooled fractions from the DEAE-Sephacel column or immunoprecipitates were fractionated on an analytical Sepha- rose CL-6B column prepared in 1% SDS, 0.1 M NaC1, 50 mM Tris/HCl, pH 8.0. Samples (0.5 ml) were applied to the column together with previously fractionated blue Dextran (2000) and phenol red, to mark void and total volumes respectively. Columns were eluted at a flow rate of 5.0 ml/h and effluent fractions of 1.0 ml were collected and aliquots counted for radioactivity.

Enzymic treatments and chemical analyses

Chondroitinase ABC treatment of ['5S]proteoglycans was performed exactly as previously described [31]. Heparitinase treatment was carried out as described in [32]. Nitrous acid treatment was as described [33]. Glycosaminoglycan (GAG) chains were removed from the proteoglycans exactly as described previously by us [27].

Immunoprecipitation of Ptdlns-PLC-releasable HSPG

HSPG from Triton X-100 extracts or from the material released by PtdIns-PLC treatment of the skeletal-muscle cells were immunoprecipitated with Protein-A- Sepharose coated with antibodies against a rat PtdIns-PLC-releasable HSPG exactly as described [ 131. The immunoprecipitates were subjected to electrophoresis in SDS/PAGE as described below. The antibodies specifically immunoprecipitate HSPG from Schwann cells that are PtdIns-PLC releasable [ 131. The antibodies do not immunoprecipitate other HSPG such as perli- can, syndecan-1 or syndecan-3. About 80% of the total PtdIns-PLC-releasable HSPG are immunoprecipitated by the antibodies.

SDS/PAGE analysis of proteoglycuns ["SISulfate-labeled or [35S]methionine-labeled samples

obtained by PtdIns-PLC treatment of the cells followed by immunoprecipitation, were analyzed by electrophoresis on 3 - 10% SDS/PAGE and fluorography as described previously [34].

589

Liposowte jormation Samples from the Sepharose CL-6B column were dia-

lyzed against 0.05 M TrismCl, pH 7.4,O.l M NaCl plus protease inhibitors, for 24 h at 4 ° C then applied to a small Q- sepharose column previously equilibrated in the same buffer. The column was eluted with 4 M guanidineMC1, 0.05 M sodium acetate pH 5.8, 0.1 M octylglucoside. Phosphatidylcho- line (from egg yolk) was evaporated to dryness under vac- uum, redissolved in diethyl ether, dried again, then mixed with the 35S-labeled material eluted from the Q-sepbarose column. The mixture was dialyzed against 4 M guanidine/ HQO.05 M sodium acetate, pH 5.4, for 12 h at 4°C [35,361. The liposomes containing proteoglycans were fractionated by gel filtration on a Sepharose CL-2B column (100 cm X 1.5 cm) in the presence of the guanidinium buffer at a flow rate of 5 ml/h. The elution profile of liposomes alone was obtained by analyzing liposomes reconstituted in the presence of ['HIPtdIns. Control samples were subjected to the same treatment but in absence of lipid or in the presence of 0.5% Triton X-100 in the column buffer.

Immunojluorescent staining

Cells to be immunostained were grown on chamber plas- tic plates (Lab-Tek, Nunc). The medium was removed and the plates were rinsed with NaCL/P,. For staining of extracellular PtdIns-HSPG, cells were incubated with the first antibody (diluted 1 : 50 in Blotto) [13] for 1 h at 4" C before fixation. After rinsing, the cells were fixed with 3% paraformaldehyde for 30 min at room temperature. PtdIns- PLC treated cells were obtained by incubation of the cells for 15 min at 37°C with 2.5 unitdm1 of enzyme and then stained as above. For staining of intracellular HSPG cells were fixed with paraformaldehyde and permeabilized with 0.05% Triton X-100 for 2 min. In both cases cells were rinsed with Blotto, then incubated with an affinity-purified fluorescein-conjugated second antibody diluted in Blotto for 1 h at room temperature. After rinsing, the plates and the cover slips were mounted and viewed with a Nikon Diaphot inverted microscope equipped for epifluorescence.

Histology

Leg muscles from rats were cross-sectioned at a thickness of 4 pm in a cryostat. Sections were fixed in 3% paraformaldehyde in NaCl/P, and incubated sequentially with the antiserum against rat PtdIns-PLC-releasable HSPG or rat anti-laminin (Telios Inc.; 1 :50 dilution in Blotto) for 1 h at room temperature then with a fluorescein-conjugated anti-(rabbit IgG) serum for 30 min at room temperature. After rinsing, the sections were mounted on glass slides and viewed with a Nikon Diaphot inverted microscope equipped for epifluorescence.

RESULTS A PtdIns-linked HSPG is present at the cell surface of skeletal-muscle cells

To characterize the presence of HSPG on the surface of skeletal-muscle cells we incubated in vitro differentiated skeletal-muscle cells with ["S]sulfate and sequentially treated the cells with phytic acid, to remove peripheral proteoglycans, and with either a mild trypsinization or the non-

Table 1. Extraction of proteoglycans from the surface of in vitro differentiated myotubes. Differentiated myotubes were incubated with [3sS]sulfate for 18 h and sequentially extracted as indicated. The trypsin and PtdIns-PLC concentrations corresponded to 50 pg/ ml and 2.5 U/ml, respectively. Radioactive proteoglycans were evaluated as described in Materials and Methods. Blank values for phytic acid extraction and PtdIns-PLC incubation were substracted and correspond to less than 1% and 5 % , respectively. The values correspond to the average i standard deviation of six independent experiments.

Treatment Solubilized proteoglycan

Phytic acid Trypin SDS Phytic acid Triton X-100 Triton X-lOO/KCl SDS Phytic acid PtdIns-PLC SDS

90 8.1 i 3.6

76.2 i 1.7 15.3 I 1.8 8.3 5 3.6

56.4 i 7.8 18.0 i 2.5 14.1 i 8.2 14.2 I 3.8 18.2 5 4.5 53.9 5 7.3

30-

[Ptdlns-PLC] (Uirnl)

Fig. 1. Release of sulfated proteoglycan from the surface of differentiated myotubes by PtdIns-PLC. Differentiated myotubes were incubated for 15 min at 37°C at the indicated concentrations. The proteogl ycans released were determined by cold trichloroacetic acid precipitation. 100% corresponds to the total proteoglycans extracted from the cells with SDS. Similar results were found when the released proteglycans were determined by binding to CPC-impregnated filters.

ionic detergent Triton X-100. Table 1 shows that trypsin removed 76% of the total proteoglycans synthesized by the cells. Triton X-100 removed 56% of the total protoglycans. If the cells were then incubated with a Triton X-lOO/KCI buffer, an additional 18% of the total proteoglycans were solubilized. These results indicate that about 85 % of the total proteoglycans synthesized by skeletal-muscle cells are probably associated with the cell surface.

To investigate if some of these surface-associated proteogl ycans are anchored to the membrane through glycosyl- PtdIns anchors, differentiated muscle cell cultures were labeled with [3SS]sulfate then sequentially incubated with phytic acid, PtdIns-PLC and SDS. Table 1 shows that under these conditions about 18% of the total proteoglycan population was solubilized by PtdIns-PLC treatment. Fig. 1 shows that maximal proteoglycan release occurred as a PtdIns-PLC con- centration of 2.5 U/ml. To characterize the glycosyl-PtdIns- linked proteoglycans, the material solubilized by Triton X- 100 and released by PtdIns-PLC treatment was separated by DEAE-Sephacel chromatography by elution with a sodium chloride gradient. Fig. 2A shows the chromatographic profile

590

KAV K A V

Fig. 2. Chromatographic analysis of cell-surface-associated proteoglycans synthesized by differentiated skeletal-muscle cells. DEAE- Sephacel chromatography of proteoglycans obtained from differentiated skeletal-muscle cells extracted with Triton X-100 (A) or released by PtdIns-PLC (B). (C) and (D) correspond to the Sepharose CL-6B profiles of the DEAE-Sephacel bound material. (C) is the proteoglycans from peak I of (A), and (D) corresponds to the DEAE-Sephacel bound material of (B). Radioactivity corresponds to 0.05-ml aliquot from 1.0-ml collected fractions.

of the Triton X-1 00 solubilized material indicating that over 90% of the radioactive material was bound to the column and resolved into two well defined peaks, I and 11, eluting at 0.4 M and 0.6 M NaCl respectively. Further analysis of peaks I and I1 indicated that they corresponded to HSPG and chon- droitin sulfate/dermatan sulfate, respectively (data not shown). Fig. 2B shows the DEAE-Sephacel profile of the proteoglycans solubilized by the PtdIns-PLC treatment. Most of the radioactive material eluted as a single peak at 0.4 M NaC1. Analysis of this material indicate that it contained exclusively HSPG. To determine how many HSPG species were present in each peak, aliquots from each were fractionated on a Sepharose CL-6B column. Fig. 2C shows that the HSPG solubilized by Triton X-1 00 resolved into two species eluting with K,, values of 0.28 and 0.68. The PtdIns-PLC- released HSPG eluted as a single peak with a K,, of 0.28 (Fig. 2D). To characterize further the HSPG solubilized by PtdIns-PLC treatment, aliquots were incubated with heparitinase and chondroitinase ABC and separated on SDS/PAGE followed by fluorography. Fig. 3 shows that the proteoglycans solubilized by PtdIns-PLC migrated as a broad band with a molecular mass of 250 kDa, were totally degraded by heparitinase treatment and were resistant to chondroitinase ABC treatment, indicating their heparan sulfate nature.

Antibodies against a HSPG isolated from cultured Schwann cells recognizes the skeletal-muscle PtdIns-PLC-releasable HSPG

To investigate whether the HSPG with K,, 0.28, solubilized by Triton X-100 and PtdIns-PLC treatment were re- lated, aliquots containing either the total population of proteoglycans solubilized with Triton X-100 or the HSPG solubilized by PtdIns-PLC treatment were incubated with anti- HSPG serum obtained against a HSPG isolated from Schwann cells [13]. The immunoprecipitates produced were analyzed by chromatography on Sepharose CL-6B. As shown in the Fig. 4A, the antiserum only precipitated the

1 2 3

-180

-1 16

- 84 - 58 - 4 0

- 36 - 26

Fig. 3. The proteoglycan released by PtdIns-PLC from skeletal- muscle cells is a HSPG. Skeletal-muscle cells were incubated with 2.5 U/ml PtdIns-PLC as described in Materials and Methods. Ali- quots from the incubation media were treated with: 1, buffer alone; 2, chondroitinase ABC; 3, heparitinase. After the treatments the material was separated on 3-10% SDSRAGE followed by fluorography. Molecular-mass standards are given on the right in kDa.

HSPG with Kav 0.28 from the Triton X-100 extract. A similar result was obtained when the immunoprecipitate from the PtdIns-PLC-released material was analyzed (Fig. 4B). Analysis of the immunoprecipitates obtained from the Triton X-100 extract and the PtdIns-PLC treatment by SDSPAGE indicated that HSPG of the same molecular mass, about 250 kDa, were immunoprecipitated in each case (Fig. 4C). These results indicated that the HSPG solubilized by Triton X-100 from the skeletal-muscle cells (Kav 0.28) and by the PtdIns-PLC treatment were recognized by the same antiserum and had similar hydrodynamics characteristics, suggesting the presence of the same molecule in both fractions. These findings also suggest that this HSPG is attached to the cell membrane by a glycosyl-PtdIns anchor.

591

@ 1

0 0.0 0.2 0.L 0.6 0.8 tD

KAV

L00,-1

", v, ~~~~ 100 0 0.0 0.2 0.4 0.6 0.8 1.0

K A V

1 2

KAV

0 + - + -

180 - 116-

84 - 58 - 48,

T X - 1 0 0 P I - P L C

Fig.4. The HSPG present in the Triton X-100 extracts and in the PtdIns-PLC-released material are recognized by antibodies against a HSPG from Schwann cells. Aliquots of Triton X-100 extracts and material released by PtdIns-PLC from [35S]sulfate-labeled cells were incubated with antibodies against a HSPG from Schwann cells. (A) and (B) correspond to Sepharose CL-6B profiles of the immunoprecipitates of the detergent extracts and PtdIns-PLC- released material, respectively, radioactivity corresponds to 0.1 ml from 1.0-ml collected fractions. (C) is a fluorogram of the immunoprecipitates separated on 3 -1 0% SDSPAGE. (+), In the presence of the antibodies; {-), in the absence of the antibodies. Molecular- mass standards are given on the left in kDa.

Characterization of the PtdIns-HSPG

To further characterize the glycosyl-PtdIns-anchored HSPG, the peak eluting with KAu 0.28 from thc Triton X-100 extracts was isolated and treated with sodium hydroxide in the presence of sodium borohydride to remove the GAG chains. Fig. 5 A shows that the alkaline treatment-released radioactive GAG chains which eluted from a Sepharose CL- 6B column with a Kay of 0.68. The same result was found when the HSPG released by PtdIns-PLC were analyzed (data not shown). Next, we decided to determine the molecular mass of the core protein of the Ptdlns-linked HSPG. In vitro differentiated skeletal-muscle cells were metabolically labeled with r3Imethionine and incubated with PtdIns-PLC. The released material was immunoprecipitated and analyzed by SDS/PAGE. As shown in the Fig. 5 B lane 1, the apparent molecular mass of the precipitated molecules corresponded

-180

-1 16

- 84 - 58

- 40

Fig. 5. Structural analysis of the glycosyl-PtdIns-HSPG from skeletal-muscle cells. (A) Sepharose CL-6B chromatographic profile of GAG chains obtained after alkaline treatment of the HSPG present in the Triton X-100 extracts that eluted with Kav 0.28 from Sepharose CLdB. Radioactivity corresponds to 0.1 ml from 1 .O-ml collected fractions. (B) Differentiated muscle cells were incubated with [ "S]methionine and treated with PtdIns-PLC. The released material was immunoprecipitated as explained in the legend to Fig. 4, and separated on 3- 10% SDSE'AGE. 1, untreated immunoprecipitate; 2, nitrous acid treated immunoprecipitate. Molecular- mass standards are given in kDa.

to a broad band of 250 kDa. This result was consistent with the molecular mass observed for the HSPG immunoprecipitated from cells labeled with [35S]sulfate (Fig. 4C). Nitrous acid digestion of the immunoprecipitated material resulted in the disappearance of the high-molecular-mass smear and the appearance of a single broad radiolabeled band migrating at a relative molecular mass of 67 kDa (Fig. 5B). These results suggested that the core protein of PtdIns-PLC-released HSPG has a molecular mass of 67 kDa. Similar results were found when the material released by Triton X-I00 was analyzed (data not shown), suggesting that the PtdIns-PLC-released immunoreactive HSPG probably possessed the same core protein as the intact membrane-bound HSPG.

The glycosyl-PtdIns-linked HSPG possess hydrophobic properties

As the glycosyl-PtdIns-linked HSPG was released from the surface by PtdIns-PLC treatment, we decided to investigate whether the intact and PtdIns-PLC-released HSPG possessed hydrophobic properties. To achieve this, liposomes

592

Efjq--J I: ?LOO ";;:El - t

200 5 100 + 1

0.0 0.2 O L 0.6 0.8 1.0 KAV W 0.0 0.2 0.L 0.6 0.8 1.0

Z @

In 0

200

0.0 0.2 0.L 0.6 0.6 10 KAV KAV

Fig. 6. Only the glycosyl-PtdIns-HSPG were able to incorporate into liposomes. Chromatographic profiles on Sepharose CL-2B of (A) reconstituted liposomes using ['HIPtdIns as a tracer; (B) liposomes reconstituted in the presence of ["SIHSPG extracted by Triton X-100 and isolated from the peak eluting with Kav 0.28 from Sepharose CL-6B column; (C) the Triton X-100-extracted HSPG was reconstituted into liposomes and chromatographed in buffer containing Triton X-1 00 (0) or chromatographed in buffer containing Triton X-100 without liposome reconstitution (0). (D) As B but using the [3SS]HSPG obtained by PtdIns-PLC treatment.

were reconstituted in the presence of the Triton X-100-released HSPG (& 0.28) and the PtdIns-PLC-released HSPG. Fig. 6A shows the chromatographic profile on a Sepharose CL-2B column of the liposomes, using ['HIPtdIns phospho- lipid as a tracer. Most of the liposomes were excluded from the column. When the solubilized Triton X-100 HSPG (Kab 0.28 in Sepharose CL-6B) were used in the reconstitution and fractionated on Sepharose CL-2B, most of the material was excluded from the column with a profile similar to that with the liposomes alone (Fig. 6B), suggesting that the intact HSPG were incorporated into the liposomes. However, when the same reconstituted material was separated on a Sepharose CL-2B column in the presence of Triton X-100, to dissolve the proteoliposomes, the V5S]HSPG were included in the column and eluted with a Kav of 0.77 (Fig. 6C). When the [TIHSPG which had not been reconstituted were fractionated on the column in the presence of Triton X-100 the ["SIHSPG were included in the column (Fig. 6C). To investigate if the PtdIns-PLC-released HSPG possessed hydrophobic properties, a reconstitution experiment was performed using the phospholipase-released HSPG. Fig. 6D shows that none of the [TIHSPG was incorporated into the liposomes. The HSPG eluted as a single peak from the Sepharose CL- 2B column, with a K,, of 0.77. These results suggested that only the intact HSPG possessed hydrophobic properties and these resided in the glycosyl-PtdIns tail that linked to the HSPG and was cleaved by PtdIns-PLC digestion.

Immunolocalization experiments We used indirect immunofluorescence staining with the

specific glycosyl-PtdIns-linked-HSPG an tiserum to determine where in the skeletal myotubes the glycosyl-PtdIns- linked HSPG was localized. In Fig. 7a-c shows that when intact cultured myotubes were incubated with the antiserum, cell-surface staining was clearly observed. Fig. 7 d shows the localization of the HSPG in permeabilized myotubes The staining was found on the surface and inside the myotubes

in the cytoplasm, around the nuclei. Fig. 7e shows that the extent of the staining decreased significantly when the myotubes were incubated with PtdIns-PLC before staining with the antiserum. Fig. 7f shows a phase-contrast micrograph of differentiated myotubes. The typical elongated shape and multiple nuclei in each myotube can be observed. These observations indicated that the glycosyl-PtdIns-linked HSPG are localized mainly on the cell surface, anchored through glycosyl-PtdIns and are also localized inside the cells.

The anti-HSPG immunoglobulins were also used to stain frozen sections of adult rat leg muscles. As shown in the Fig. 8 a, the antibodies stained the endomysium that surrounds each muscle fiber, which corresponds to a delicate layer of connective tissue composed mainly of basement membrane and reticular fibers. The pattern was indistinguishable from that obtained with anti-laminin antibodies (Fig. 8 b). Laminin is present in the skeletal-muscle basement membrane [37]. Thus, the anti-HSPG immunoglobulins ap- peared to be staining molecules present exclusively in the surface of the skeletal muscle. No staining was observed in the perimysium, a fibrous connective tissue which surround bundles of muscle fibers.

DISCUSSION

The results presented in this study provide experimental evidence for the presence of hydrophobic HSPG on the surface of differentiated myotubes. More specifically we have characterized a HSPG that is linked to the membrane of the muscle cells via a glycosyl-PtdIns anchor. This HSPG could be released in a soluble form from the surface of the cells by treatment with Ptdlns-PLC. About 20% of the total muscle cell proteoglycan population corresponded to this species. This HSPG could also be isolated from the proteoglycan population obtained after Triton X-1 00 solubilization. Biochemi- cal and immunological analysis of the HSPG using specific antibodies against a HSPG purified from Schwann cells [ 131,

593

Fig. 7. The glycosyl-PtdIns-HSPG are located on the surface of differentiated myotubes. (a-e) Immunofluorescence staining with anti- (glycosyl-PtdIns-HSPG) sera. (a-c) In intact cells fine fluorescent lines of staining are seen embracing the myotube border (arrow heads) while the cytoplasm is poorly stained (a, b). At higher magnification (c) the strong reaction present at the cell surface appears as discontinu- ous clumps of fluorescent staining (arrows). (d) In penneabilized cells fluorescent staining is visible in the myotubes and the cytoplasm now appears strongly stained. The nuclei are devoid of staining (arrows). (e) In intact cells previously incubated with PtdIns-PLC, the immunofluorescent staining present on the surface almost completely disappears. (f) Phase-contrast micrographs of the differentiated myotubes. The myotubes are clearly seen (arrow heads) and exhibit nuclei accumulation in single cells (arrows). The bars correspond to 10 pM

indicated that a HSPG present in the Triton X-100 fraction and in the PtdIns-PLC treated fraction were the same, sup- porting the idea that this HSPG is present at the cell surface. However, analysis of hydrophobicity, studied by incorporation to liposomes, revealed that only the HSPG extracted by the detergent possessed hydrophobics properties. These results strongly suggest that the hydrophobic properties of this HSPG reside in the glycosyl-PtdIns tail covalently bound to the carboxy terminus of the core protein [38].

At least two species of [3sS]sulfate labeled macromolecules were isolated from the detergent extract after Sepharose CL-6B chromatography, with K,, values 0.28 and 0.67. Only the higher-molecular-mass form was able to incorporate into liposomes. The lower-molecular-mass form did not possess hydrophobic properties and was unaffected by alkali treatment, suggesting that these molecules are GAG chains released from intact HSPG (unpublished results). The molecular mass determined for the core protein of the PtdIns-PLC- released HSPG was approximately to 67 kDa The same value was obtained for the detergent solubilized HSPG. These results are consistent with the mode of action of the PtdIns- PLC, hydrolyzing the phospholipids and releasing the intact

protein 1381. Interestingly, we have found the same HSPG in the incubation media of the differentiated muscle cells, and analysis of the core protein also indicate the same molecular mass (unpublished results). This suggests the presence of an endogenous processing mechanism, which releases the intact HSPG from the membrane into the incubation media. This activity probably corresponds to a endogenous phospholipase or a protease acting in or very near the carboxy-terminus of these type of anchored proteins.

Immunocytolocalization studies indicated that the glycosyl-PtdIns-linked HSPG are located on the surface of the differentiated myotubes. As expected, the reaction was lost when the cells were previously incubated with PtdIns-PLC. A strong reaction was observed when the myotubes were permeabilized before adding the antibodies. This indicates that an intracellular pool of the HSPG may exist. We have experimental data which indicate that after PtdIns-PLC treatment of the differentiated cells the detergent Triton X-I00 solubilized about 40% of the total glycosyl-PtdIns-HSPG, confirming the existence of the intracellular pool (unpublished results). From our results we can estimate that the PtdIns linked HSPG present on the surface correspond ap-

594

Fig. 8. Immunofluorescence staining with anti-(glycosyl-PtdIns-HSPG) and anti-laminin sera of transverse frozen sections of rat skeletal muscle. The distribution of the immunoreactivity in the muscle fascicles is similar for both antibodies. (a) Using the anti- (glycosyl-PtdIns-HSPG) serum the polygonal outline of each cross-sectioned muscle fiber appears strongly decorated; fluorescence seems to be concentrated in the region of the endomysium adjacent to the fibers (arrow heads). No staining of connective tissue elements of the perimysium is detected (arrows). A positive reaction is observed in the wall of the blood vessels (v). (b) The distribution of anti-laminin- serum staining is shown. Arrow heads point to the basement membrane adjacent to the fibers while the arrows show the perirnysium devoid of reaction. The bar corresponds to 100 pM.

proximately to 60% of the total cell-surface HSPG. 30% of the surface-associated HSPG correspond to HSPG which re- quire additional detergent ionic strength for the solubilization, suggesting an interaction with ECM and/or cytoskeletal elements [41]. We have preliminary data which indicate that syndecan-1 is present in proliferating myoblasts and also in differentiated myotubes, because of the hydrodynamic characteristics of this HSPG [30] it might migrate with the PtdIns-HSPG in the Sepharose CL-6B column. The defin- itive presence and localization of syndecan-1 is under current investigation.

We also investigated the cellular localization of the PtdIns-linked HSPG in adult rat muscles. In this tissue it was found localizing with laminin, surrounding each muscle fiber in the endomysium. No reaction was observed, however, at the level of the perimysium. This is interesting, because we have described previously that the proteoglycan decorin is highly concentrated in the perimysium and is also present in the endomysium 1281. It has been suggested that a glycosyl- PtdIns-HSPG, very similar or identical to the skeletal-muscle HSPG is present in sciatic nerves on the outer Schwann- cell membrane that is in direct contact with the laminin-rich basement membrane that surrounds the individual axon- Schwann-cell units [ 131. These observations provide indirect evidence that this HSPG may be acting as a cell-surface receptor for laminin present in the basement membrane. In this muscle-cell type the expression of laminin is developmen- tally regulated [39], consequently it would be interesting to determine whether the expression of the glycosyl-PtdIns- HSPG is also regulated during cell differentiation. If this is the case, it would be important to design experiments that would interfere with the normal function of this HSPG (antibodies or specific inhibitors) and to study the effect of these

alterations on muscle-cell differentiation. However, it has been shown that other HSPG serve as binding sites for mem- bers of the heparin binding FGF, the best studied of which is basic FGF (bFGF). The experiments indicate that binding of bFGF to these low-affinity sites may be essential for binding to the high-affinity sites that correspond to the signal-trans- ducing receptors [lo, 111. Recently, it has been shown that the repression of myogenic differentiation by bFGF requires cellular heparan sulfate [12] and it has been suggested that the binding sites for bGFG may be the proteoglycan syndecan 1401. To determine the exact function of the HSPG on the surface of the muscle cells, further studies focused on understanding the molecular characteristics, cellular location and level of expression during development of these molecules are required.

We thank Mr Rodrigo Urrea for the tissue culture work. This work was supported by a Fogarty International Research Collabora- tion Award (NIH, R03TW00093) to D. J. C. and E. B., and grants from Fondo Nucional de Ciencie y Tecnologid (565-93) and the International Foundation for Sciences (IFS-A-1407/2) to E. B. and the National Institutes of Health (R01NS21925) to D. J. C.

REFERENCES 1. Cole, G. J. & Burg, M. (1989) Exp. Cell Rex 182, 44-60. 2. Koda, J. E., Rapraeger, A. & Bernfield, M. (1985) J. Bid.

3. Saunders, S. & Bemfield, M. (1988) J. Cell B i d . 106, 423-

4. Sun, X., Mosher, D. F. & Rapraeger, A. (1989) I. Bid . Chern.

5 . Segarini, P. R. & Seyedin, S. M. (1988) J. B i d . Chern. 263,

Chern. 260, 8157-8162.

430.

264, 2885-2889.

8366-8370.

595

6. Cheifetz, S., Andres, J. L. & MassaguC, J. (1 988) J. Biol. Chem.

7. Brandan, E. & Inestrosa, N. C. (1986) J. Neurosci. Res. 15,

8. Cisar, L. A., Hoogewerf, A. J., Cupp, M., Rapport, C. A. & Besandaun, A. (1 989) J. Biol. Chem. 264, 1767 - 1774.

9. Marcum, J. A. & Rosemberg, R. D. (1987) Semdn. Tlzromb. Hemostosis. 13, 464-474.

10. Rapraeger, A,, Krufia, A. & Olwin, B. B. (1991) Science 252, 1705 - 1708.

11. Yayon, A,, Klagsbrun, M., Esko, J. D., Leder, P. & Omitz, D. M. (1991) Cell 64, 841-848.

12. Olwin, B. B. & Rapraeger, A. (1992) J. Cell Biol. 118, 631- 639.

13. Carey, D. J. & Stahl, R. C. (1990) J. Cell Biol. I l l , 2053- 2062.

14. Marynen, P., Zhang, J., Cassiman, J. J., Van den Berghe, H. & David, G. (1989) J. Biol. Chem. 264, 7017-7024.

15. Saunders, S., Jalkanen, M., O’Farrel, S. & Bemfield, M. (1989) J. Cell Bid. 108, 1547-1556.

16. David, G., Lories, V., Decock, B., Marynen, P., Cassiman, J. J. &Van den Berghe, H. (1990) J. Cell Biol. 111, 3165-3176.

17. Carey, D. J. & Evans, D. M. (1989) J. Cell Biol. 108, 1891- 1897.

18. Brandan, E. & Hirschberg, C. B. (1989) J. Biol. Chem. 264,

19. Hauschka, S . D. & Konigsberg, I. R. (1966) Proc. Nut1 Acud.

20. Saitoh, O., Periasamy, M., Kan, M. & Matsuda, R. (1992) Exp.

21. Nandan, E., Clarke, E. P., Ball, E. H. & Sanwall, B. D. (1990)

22. Ruoslahti, E. & Pierschbacher, M. D. (1987) Science 238,491 -

263, 16984-16991.

185-196.

10520- 10 526.

SC~. USA 55, 11 9 - 126.

Cell Rex 200, 70-76.

J. Cell Biol. 110, 1673-1679.

497.

23. Buck, C. A. & Horwitz, A. F. (1987) Annu. Rev. Cell Biol. 3,

24. Hamati, F. H., Britton, E. L. & Carey, D. J. (1989) J. Cell Bid.

25. Brandan, E., Fuentes, M. E. & Andrade. W. (1991) ELK J. Cell

26. Andrade, W. & Brandan, E. (1991) Comp. Biochem. Physiol. B

27. Fadic, R., Brandan, E. & Inestrosa, N. C. (1990) J. Neurosci.

28. Brandan, E., Fuentes, M. E. & Andrade, W. (1992) J. Neurosci.

29. Yaffe, D., Saxel, 0. (1977) Nuture 270, 725-727. 30. Rapraeger, A. & Bemfield, M. (1985) J. Bid. Chem. 260,

31. Brandan, E. & Inestrosa, N. C. (1987) J. Neurobiol. 18, 271 -

32. Yanagishita, M. & McQuillan, D. J. (1989) J. Biol. Chem. 264,

33. Shively, J. E. & Conrad, H. E. (1976) Biochemistry 15, 3932-

34. Carlson, S. S. &Wright, T. N. (1987) J. Cell Bid. 205, 3075-

35. Lories, V., David, G., Cassiman, J. J. & Van den Berghe, H.

36. Andres, J. L., Stanley, K., Cheifetz, S. & MassaguC, J. (1989)

37. Sanes, J. R. (1989) Annu. Rev. Neurosci. 12, 521 -546. 38. Low, M. & Saltiel, A. (1988) Science 239, 268-275. 39. Olwin, B. B. & Hall, Z. (1985) Dev. Biol. 112, 359-367. 40. Salmivirta, M., Heino, J. & Jalkanen, M. (1992) J. Biol. Chem.

41. Rapraeger, A., Jalkanene, M. & Bernfield, M. (1986) J. Cell

179-206.

108, 249552505,

Biol. 55, 209-216.

Comp. Biochem. 100, 565-570.

10, 3516-3523.

Res. 32, 51 -59.

4103-4109.

282.

17551 -17558.

3942.

3086.

(1986) Eur. J. Biochem. 158, 351-360.

J. Cell Biol. 109, 3137-3145.

267, 17606-17610.

Biol. 103, 2663-2696.

a lipid-anchored heparan sulfate proteoglycan is present in the surface of differentiated skeletal...

Documents