Biosynthesis of heparan sulfate proteoglycans of developing chick breast skeletal muscle in vitro

Experimental

Cell Research 166 (1986) 327-339

Biosynthesis of Heparan Sulfate Proteoglycans of Developing Chick Breast Skeletal Muscle In Vitro D. M. NOONAN,‘,*

D. J. MALEMUDls2

and R. J. PRZYBYLSKI’

‘Department of Developmental Genetics and Anatomy, and ‘Department of Medicine, Case Western Reserve University, Cleveland, OH 44106, USA

Previous studies have reported an increase in heparan sulfate glycosaminoglycan (HSGAG) during skeletal muscle differentiation in culture. We have investigated this phenomenon further in relation to the heparan sulfate proteoglycans (HSPG) produced by myogenic cultures. Pulse-chase analysis indicated an approx. 3-fold increase in heparan sulfate synthesis in myotube cultures over that in proliferating or aligning myoblast cultures. Muscle ftbroblast culture heparan sulfate synthesis was higher than that of myoblasts but was lower than myotubes. The turnover rates appeared to be the same for all stages of development, with a tlj2 of approx. 5 h. Enrichment for heparan sulfate by Sepharose CL4B and DEAE-Sephacel chromatography indicated an increase in the hydrodynamic size of the proteoglycan produced by myotubes over that from myoblasts, with a shift in K,” from 0.14-O. 19 to 0.07. Fibroblasts synthesized the smallest proteoglycan, with a K,” of 0.22. All of the proteoglycans contained similar sized glycosaminoglycan chains with an estimated molecular weight of 3OOfKklOOOO.Localization of the heparan sulfate proteoglycan in myotube cultures by trypsin sensitivity indicated much of the intact proteoglycan to be closely associated with the cell surface, while internalized material appeared in a degraded form. @ 1986 Academic press, IOC.

Heparan sulfate proteoglycans (HSPG) are found as a major component of basement membranes [12] and on cell surfaces where they appear to be membrane intercalated [18, 291. Heparan sulfate proteoglycans function in renal filtration in the glomerular basement membrane [16] and in cell-substratum adhesion and spreading of cells in culture [30]. There is ample evidence of the involvement of HSPG in morphogenesis; they have been linked to cellular growth control [21] and branching of salivary gland ducts [6]. We have been using an in vitro system of skeletal muscle myogenesis as a developmental model for studying cell surface and matrix changes during muscle development. This culture system offers several practical advantages in the study of myogenesis, in that single cells isolated from embryonic muscle enter stages of proliferation, alignment and fusion with greater synchrony than encountered in vivo. Furthermore, skeletal muscle cells also alter their patterns of extracellular matrix synthesis during myogenesis; myoblasts do not form a basement membrane, whereas newly formed myotubes do. A relative increase in the synthesis of HSPG was found in the time period associated with myotube formation [3,22, 271. * To whom offprint requests should be sent. Address: NIH, NIDR, 30/414, Bethesda, MD 20892, USA. 22468340

Copyright Q 1986 by Academic Press, Inc. AU rights of reproduction in any form reserved OLM4-4827/86 503.00

320 Noonan, Malemud and Przybylski Carrino & Caplan [9] characterized a hydrodynamically large chondroitin-6sulfate proteoglycan in muscle, both in vitro and in vivo, which was replaced by hydrodynamically smaller proteoglycan at a later stage of development. Antibodies which recognize HSPG have been co-localized with acetylcholine receptor plaques and newly forming neuromuscular junctions [4, 5, 71. The rate of synthesis, composition and turnover of the HSPG during myogenesis in vitro is unknown. We report here on an increase in the synthesis rate and hydrodynamic size of the HSPG of skeletal muscle with myotube formation. These changes may be key events in the establishment of a basement membrane by the newly formed myotubes. We also provide preliminary characterizations of the structure and localization of HSPG during in vitro myogenesis. MATERIALS

AND METHODS

Preparation of Cell Cultures Twelve-day chick breast muscle was cultured according to the method of Sandra et al.- [31]. Briefly: Breast muscle was dissected from 12-day embryonic chicks (white leghorn), minced with scissors and dissociated with 0.05% collagenase (Worthington CLS II, lot selected for myotube formation) for 10 min at 37°C with occasional agitation. Dissociated cells were filtered with Nitex to remove myotubes and cell clumps. Filtered cells were pelleted, resuspended and plated in Leibovitz L-15 medium buffered with 10 mM HEPES (N-2-hydroxyethelpiperazine-N’-Zethanesulfonic acid) at pH 7.4, and supplemented with 10% horse serum (K.C. Biological; lot selected for myotube formation), 5 % chick embryo extract and 1% antibiotic-antimycotic Xstreptomycin sulfate, penicillin sulfate, fungizone (Gibco)). Cells were plated in 7 ml medium at 4x lo6 cells per dish in 100x 15 mm tissue culture dishes precoated with 1% gelatin. Cultures were incubated at 37’C in a water-saturated atmosphere of 5 % COz-95% air. Muscle fibroblasts were obtained by serial subculture of primary muscle cultures. Muscle cultures were incubated until the 5th day of culture. Cells were detached with 0.5 % trypsin in Spinners salts, filtered through bolting silk to remove myotubes, pelleted and replated in L-15 medium. Five successive subcultures yielded a nearly pure fibroblast population by morphological criteria. These cultures were radiolabeled and processed as described below for muscle cultures.

Radioisotopic Labeling of Cell Cultures Culture medium was supplemented with ‘?SO:- (20 uCi/ml, Amersham Corp., 1 10&l 200 Gil mmol). The antibiotic-antimycotic was omitted from the medium to reduce the unlabeled sulfate. For continuous radiolabeling the radiolabel was added with the plating medium. Culture media was decanted from cultures and cells were washed three times with phosphatebuffered saline (PBS) supplemented with 1 mM EDTA (ethylenedinitrilo-tetra-acetic acid), 1 mM NEM (N-ethyl maleimide) and 1 mM PMSF (phenylmethylsulfonylfluoride). Cells were removed with a rubber policeman in buffer additionally supplemented with pepstatin (10 @ml) and pelleted by centrifugation.

Solubilization of Proteoglycans Cell pellets from radiolabeled cultures were solubilized at 37°C with l-l .5 ml SDS extraction buffer (0.2% SDS in extraction buffer: 0.05 M Na acetate, pH 7.4, containing 1 mM EDTA, 1 mM NEM, 1 mM PMSF and 10 @ml pepstatin). These samples were then directly applied to CL-6B columns in SDS-Bis buffer as below, or were nitrous acid deaminated as described below, prior to chromatography.

Determination of Heparan Sulfate The amount of heparan sulfate was determined by exhaustive nitrous acid deamination. One-half of the sample was nitrous acid deaminated by addition of glacial acetic acid and 18% NaN& (10 ul each) Erp Cell Res 166 (1986)

Skeletal muscle heparan sulfate proteoglycans

329

in a total volume of 0.1 ml. The other half, used as a control, received glacial acetic acid and water to a total volume of 0.1 ml. The samples were incubated for 80 min at 25”C, and the reaction terminated by addition of 2 M NI&(SOp)NH2 (100 pl) and incubation for 30 min.

Glycosaminoglycan

Chain Size Determination

Samples were reduced with alkaline borohydride. A stock solution of 7.5 % NaBH., was prepared in 0.1 M NaOH and stored at -20°C. An aliquot of the sample was mixed 1: 1 with stock solution and incubated 15-17 h at 37°C. Afterward samples were brought to neutral pH by adding glacial acetic acid.

Gel Filtration

Chromatography

The samples that were obtained were chromatographed on either Sepharose Cl-6B (110x 1.5 cm), Ultrogel A6 (110~1.5 cm), Sepharose CL-4B (110x1.5 cm) or Sepharose Cl-2B (110X0.25 cm). Columns were equilibrated and eluted with 0.2% SDS in 150 mM ‘Iris, pH 7.4. The flow rate for Sepharose CLdB and CL4B and Ultrogel A6 was approx. 0.1 mYmin and 1.5 ml fractions were collected. Sepharose CL-2B column flow rate was approx. 0.025 ml/min and 0.375 ml fractions were collected. Recoveries were approx. 80-90% of radiolabel from these columns.

Pulse-Chase

Analysis

of Proteoglycan

Synthesis

To examine the synthesis and turnover of heparan sulfate proteoglycans, radiolabel pulsexhase experiments were employed. Pulse times were determined by measuring the incorporation of 35SO$radiolabel into macromolecules per w DNA as a function of time for each developmental stage. A pulse of radiolabel was given 8 h prior to the developmental time point (i.e. 24, 42 and 72 h) and samples processed at 2-h intervals thereafter. Proliferating, aligning and fusion cultures all initially incorporated “SO:- at approximately the same rate, indicating there were no overt changes in inorganic sulfate pool sizes in the differing developmental stages (data not shown). Linear incorporation occurred until 6-8 h; a pulse time of 6 h was chosen. The culture medium was removed at the desired developmental time point and replaced with 3 ml/dish of “SO:--radiolabeled media. The medium was removed at the end of the radiolabeling period (6 h), and the cultures rinsed quickly with unlabeled media. Cultures were then fed 7 ml of unlabeled medium which was additionally supplemented with antibiotic;intimycotic mixture (which also increases the pool of unlabeled sulfate). These cultures were returned to the incubator for the chase time (0, 3 or 6 h). The cell layer was removed, rinsed, pelleted, solubilized in SDS extraction buffer and chromatographed as above for total proteoglycans. Aliquots were treated with nitrous acid to determine regions containing heparan sulfate. The DNA content of samples was determined on an Aminco-Bowman spectrofluorimeter by the mithrimycin method of Hill & Whatley [13]. SDS did not interfere with the DNA determination under these conditions by comparison of standards.

Enrichment

Procedures

.

Cell pellets were prepared from cultures radiolabeled with 35SO:- from onset of culture until the desired developmental stage and solubilized in 10 vol SDS-extraction buffer at 20°C for 30 min. Insoluble material was removed by centrifugation at 20000 g for 20 min. Samples were then concentrated by ultrafiltration in an Amicon-stirred cell apparatus with a 10000 NMW cut-off filter (Amicon YM-10 or Millipore PTGC). The samples were chromatographed on Sepharose CL-4B. Regions containing heparan sulfate were determined by nitrous acid deamination of an aliquot of the sample. Fractions containing heparan sulfate were pooled and concentrated by ultrafiltration in an Amicon-stirred cell as above. Proteoglycans were then chromatographed on columns of DEAESephacel. Approx. 2 ml of fresh resin was used for each chromatograph. Columns were equilibrated with 8 M urea and 0.5 % CHAPS (3-[(3lcholamidopropyl) dimethyl-ammoniol-1-propanesulfonate) in extraction buffer (0.5 M Na acetate, pH 7.4, containing 1 mM EDTA, 1 mM NEM, 1 mM PMSF and 10 &ml pepstatin) with 0.15 M NaCl. Samples (l-l.5 ml) were loaded onto columns and columns reequilibrated with 20 ml of 0.15 M NaCl, 8 M urea and 0.5% CHAPS in extraction buffer. Columns were eluted with a linear gradient (0.15 M NaCl to 1 M NaCl) in 8 M urea and 0.5% CHAPS in Exp Cell Res 164 (1986)

330

Noonan, Malemud

and Przybylski

extraction buffer (approx. 70 ml each). A 204 aliquot of each fraction was used to determine radioactivity. Another 20-ul aliquot was taken from every fifth fraction to determine refractive index of the sample. The molar&y of NaCl was determined by comparison to the refractive index of known samples derived from aliquots of the gradient buffers. The heparan sulfate-containing fractions were pooled and concentrated on an Amicon ultrafiltration unit as above.

Determination

of Trypsin Sensitivity

of Proteoglycans

Seventy-two-hour cultures continuously radiolabeled with 35SO$- were sequentially washed with six changes of fresh unradiolabeled L-15 medium over a period of 80 min. The washed cell layer was then treated with 0.075 % trypsin for 10 min at 20°C. The cells were removed with a rubber policeman, pelleted and supematant decanted. The pooled media washes were adjusted to contain 0.2 % SDS and concentrated by ultrafiltration in an Amicon-stirred cell. The trypsin supematant was adjusted to contain 0.2 % SDS, 0.1% soybean trypsin inhibitor, 1 mM PMSF and concentrated. The cell pellet was solubilized with 1 ml of buffer containing 0.2 % SDS, 0.1% soybean trypsin inhibitor and 1 mM PMSF.

Data Analysis Data from the scintillation counter was fed into an Apple II-plus microcomputer. The data for each chromatogram were then normalized by converting cpm in each fraction into a percentage of total cpm recovered in all fractions after background subtraction. This procedure allowed comparisons between samples when total applied radioactivity or counting efficiency between column runs differed. Chromatography data could then be analysed to give percentages of total radioactivity for selected portions of the chromatograms using a program which summed the percentage of total radioactivity for fractions in the desired regions. This program also calculated the percentage of nitrous acid-sensitive material in specified regions where chromatography of nitrous acid-deaminated material was performed. Data was plotted using a Huston Hiplot plotter (Huston Instrument, Austin, Texas).

RESULTS Myogenesis

in Cell Culture

Myogenesis occurs as three progressive phases. Proliferation began after the onset of culture and progressed to the myoblast alignment phase, which peaks at approx. 42 h. Myoblast fusion begins shortly thereafter and continued for the next 36 h. Though many as-yet unfused myoblasts were present at 72 h in culture, the high fusion index (61%) at this time, coupled with minimal fibroblasts, led us to use this time point as representing the myotube stage. Details of the population dynamics and morphological events appeared previously [28]. The use of mitotic inhibitors was’ avoided, as these agents may affect matrix synthesis [I, 141 and non-mitotic cell function [2]. Analysis of Proteoglycans

Proteoglycans radiolabeled with 3sSO$- were resolved into peaks in three general regions by Sepharose CLdB chromatography. Fig. 1 shows the chromatographic profiles of 6-h pulse-radiolabeled cultures of developing muscle cultures (fig. 1A-C) and muscle fibroblasts (tig. ID). In myogenic cultures, there is a sharp void volume peak which is largely nitrous acid-insensitive. In Sepharose Cl-2B chromatography of continuously radiolabeled cultures (data not shown) Exp Cellt Res 166 (1986)

Skeletal muscle heparan sulfate proteoglycans

33 1

2 0 FRACTION

NUMBER

3 CHASE

TIME

6 thoursl

Fig. 1. Sepharose CL-6B chromatographs of developing muscle cultures pulse-radiolabeled 6 h with “SO$-. Samples were solubilized in 0.2% SDS, treated as control (-) or with nitrous acid (---) and applied to columns. (A) 24-h myoblast cultures; (B) 42-h aligning myoblast cultures; (C) 72-h myotube cultures; (0) muscle fibroblast culture. Arrowhead, void volume; arrow, inclusion volume. The sharp peaks at the inclusion volume denote free sulfate. Fig. 2. Analysis of heparan sulfate in developing muscle cultures pulse-chased with 35SO:-. O-O, 24-h myoblast cultures; A-A, 42-h aligning myoblast cultures; 0-Q 72-h myotube cultures; x , muscle fibroblast cultures. (A) The cpm in heparan sulfate per cogDNA of cultures; (B) the data in (A) presented as a percentage of initial value to indicate turnover rate.

this peak resolved as a broad peak of nitrous acid-insensitive material (K,” 0.25430) at each developmental stage, and was assumed to be the large muscle chondroitin-6-sulfate proteoglycan characterized by Carrino & Caplan [9]. A second peak just included from the void volume is present in all cultures. This peak contained the nitrous acid-sensitive HSPG. The amount of nitrous acidsensitive material in this peak was slightly increased in aligning myoblast cultures (fig. 1B) and particularly increased in the newly formed myotube cultures (fig. 1 C) over that of proliferating myoblast cultures (fig. 1A). There was little material in the more included regions in pulse-radiolabeled cultures; however, continuously radiolabeled aligning myoblast and myotube cultures demonstrated a small, partially nitrous acid-sensitive peak (K,” approx. 0.71) in this region. The increase in radiolabeled HSPG during myogenesis was similar to that found Exp Cd Res 166 (1986)

332

Noonan, Malemud

and Przybylski

IO -A a-

wa

H

12

/ /--

9

,/-

6 3

I

/

0 !,h_, 0

, 20 FRACTION

40

Fig. 3. Enrichment for heparan sulfate. (A) Preparative Sepharose CL-4B column chromatograph (-) with an aliquot treated with nitrous acid (---). Samples below bar were pooled and run on DEAESephacel. (B) DEAESephacel column chromatograph of fractions 4268 of chromatograph in (A). Samples below bar were collected for further analysis. ---, Molarity of NaCl elution gradient.

60

NUMBER

with [3H]glucosamine radiolabel glycosaminoglycan analysis [22, 23, 271. This is evidence that changes in sulfation pattern were not responsible for the increase in radioactivity of the HSPG associated with myotube formation in vitro. Pulse-Chase

Analysis

of Radiolabeled

Cultures

The increase in HSPG could have occurred via three possible mechanisms: an increase in HSPG synthesis, a decrease in HSPG turnover, or a decrease in the synthesis of non-heparan sulfate glycosaminoglycans. To investigate these alternatives, the synthesis and turnover rates of HSPG were measured. Analysis of the HSPG cpm as a function of DNA content for each developmental stage is shown in fig. 2. These data showed that 72-h myotubes incorporated 35SOi- into heparan sulfate proteoglycan (HSPG) three times that of myoblast cultures (fig. 2A). The relative turnover times estimated for these molecules by expression of the heparan sulfate cpm per ug DNA with time as a percentage initial value is shown in fig. 2B. All samples appear to have essentially equal turnover rates with a tin of roughly 5 h (fig. 2 B). These data clearly indicated that the increase in HSPG in myotubes was due to an increase in the synthesis of heparan sulfate and not to a decreased turnover rate. More extended (12, 18, 24 and 48 h) chase studies on myotube proteoglycans did not indicate an obvious precursor-product relationship between the larger HSPG and the heparan sulfate eluting with a higher K,, (approx. 0.71) on Sepharose CLdB in continuously radiolabeled cultures. Sepharose CL-6B chromatographs of 6-h pulse-radiolabeled cultures indicated an increase in HSPG at 72 h vs 24 h in culture (fig. 1A-C). Muscle-derived fibroblast cultures were anafysed for comparison. Fibroblast cultures radiolaExp Cell Res 166 (1986)

Skeletal muscle heparan suEfate proteoglycans

333

I

1OrC

FRACTION

NUMBER

Fig. 4. Analysis of enriched heparan sulfate proteoglycans from muscle cultures. Sepharose CL-6B chromatographs of (A-D) proteoglycans; (E-H) glycosaminoglycans from muscle cultures. (A, E) 24-h

myoblasts; (B, F) 42-h aligning myoblasts; (C, C) 72-h myotubes; (0, H) muscle tibroblasts. -, Control; ---, nitrous acid-treated. Arrowhead, void volume; arrow, inclusion volume.

beled for 6 h (fig. 1D) synthesized predominantly a HSPG, and did not synthesize a large nitrous acid-insensitive proteoglycan. Fibroblast cultures contained more HSPG per pg DNA than myoblast cultures did, but less than myotube cultures after the same radiolabeling period (fig 2A). Analysis of Heparan Sulfate Proteoglycans To better characterize the HSPG from developing muscle cultures, an enrichment for the HSPG utilizing SDS extraction, Sepharose CL-4B (fig. 3A) and DEAE-Sephacel chromatography (fig. 3B) was used. This enrichment protocol enabled us to assess the hydrodynamic sizes of the proteoglycans and glycosaminoglycan side chains. However, it did not resolve the proteoglycans as successfully as others have found with similar procedures [12, 361. Heparan sulfate proteoglycan-enriched samples, chromatographed on Sepharose CLdB, with heparan sulfate-containing regions were determined by nitrous acid determination, as shown in fig. 4A-D. The glycosaminoglycan chain length was estimated by the shift in elution after alkaline borohydride reduction (fig. 4&H). There Exp Cell Res 166 (1986)

334

Noonan, Malemud

and Przybylski

Table 1. Properties of heparan sulfate-proteoglycans muscle and muscle fibroblast cultures Culture 24-h myoblast 42-h myoblast 72-h myotube 72-h myotube 35SOi- pulse Muscle fibroblast

Proteoglycan L

Glycosaminoglycan K.&V

0.14 0.19 0.07

0.36 0.45 0.37

0.08 0.22

0.36” 0.41”

of developing

skeletal

K,, on Sepharose CL-6B of heparan sulfate proteoglycans and glycosaminoglycans of developing muscle cultures. The Kav was determined by subtraction of nitrous acid-insensitive radioactivity in each fraction from total radioactivity in that fraction. The remainder represents the heparan sulfate content of the sample. The percentage of total heparan sulfate in each fraction, and the fraction in which approx. 50% of the material was distributed on either side was determined. The K., was calculated from this fraction. The samples were continuously radiolabeled with ‘5SO:- as described in text with the exception of 6-h pulse-radiolabeled sample as noted. Samples were enriched for heparan sulfate as described in text. LI The K,, of sample was determined without subtraction of nitrous acid-insensitive material.

is an apparent shift to a larger hydrodynamic size in the elution of myotube heparan sulfate (fig. 4 C) from that of myoblast heparan sulfate (fig. 4A, B). There was no apparent shift in the elution of the glycosaminoglycan chains of myotubes (fig. 4G) when compared with myoblasts (fig. 4E, F’). Note that the musclederived tibroblasts appeared to make the hydrodynamically smaller heparan sulfate proteoglycan (fig. 40) with a similar sized glycosaminoglycan chain (fig. 4H). To determine more accurately the characteristics of the HSPG, the nitrous acid-insensitive material in each fraction was subtracted from that fraction’s total. The K,, of the nitrous acid material was then determined from these adjusted values (table 1). The HSPG of myoblast cultures eluted as a broad peak with a K,, of between 0.14 (24-h cultures) and 0.19 (48-h cultures). The HSGAG chains of myoblast cultures eluted with a K,, of 0.36X1.45. Myotube culture HSPG eluted with a lower K,, (0.07) and with a narrower distribution than in myoblast samples (fig. 4). The elution position for the myotube culture HSGAG chains was the same (K,, 0.37) as that in myoblast cultures. Fibroblast culture HSPG eluted with a larger K,, than that of myoblasts (0.22), but again with similar glycosaminoglycan chain size (0.41). Although there was considerable overlap in hydrodynamic size between the various HSPG, the HSPG of myotube cultures eluted with a lower K,, than that of myoblasts or fibroblasts. There was no consistent relation between the K,, of regions fractionated on Sepharose CL4B and the K,, of the enriched material on Sepharose CL-6B, indicating that the differences between elution of the HSPG were not a result of fractionation on Sepharose CL-4B. Analysis of the hydrodynamically smaller regions of Sepharose CL-4B chromaExp Cell Res 166 (1986)

Skeletal muscle heparan sulfate proteoglycans

335

Table 2. Distribution of radioactivity in cellular compartments Compartment

Percentage total cpm

Percentage total heparan sulfate

Secreted (media) Surface (trypsin-sensitive) Intracellular (trypsin-insensitive)

49.8 25.4 24.8

19.2 30.8 50.0

The distribution of radioactivity into cellular compartments as defined by their removal from the cells. 72-h myotube cultures which were continuously radiolabeled with “SO:- were washed with several changes of L-15 medium, followed by incubation with 0.075 % trypsin for 10 min, as described in text.

tographs (fractions 71-90) after identical enrichment procedures indicated that this material did not change significantly after treatment with either alkaline borohydride reduction or nitrous acid deamination (data not shown). Localization of Heparan Sulfate Proteoglycan Several studies have reported trypsin-sensitive glycosaminoglycans in many cell types, and a large proportion of the glycosaminoglycans are sensitive to crude heparinase [ll]. The intercellular glycosaminoglycans are usually small alkaline-borohydride-insensitive materials, thought to be degradation products. Yanagishita & Hascall [36] reported that 93 % of the synthesized glycosaminoglycans were at some point located at the cell surface by trypsin sensitivity. As many of the possible functions of the muscle proteoglycans suggest their localization to the external surface of plasma membranes, it was of interest to examine this aspect of the muscle cell proteoglycans. Preliminary experiments indicated that approx. 60% of cetylpyridinum chloride-precipitable “SO$- radioactivity was removed from continuously radiolabeled cell cultures by incubation in L-15 medium alone. An additional 5% was removed with trypsin in L-15 medium. Twenty-two percent of [3H]glucosamine radiolabel was located in medium alone with an additional 8% removed by trypsin. Approx. 10% of [3H]leucine was removed. These data suggested a significant rate of secretion or loss of loosely bound 3sSOi- surface material. In the following experiment, cultures which had been continuously radiolabeled with “SO$- were washed sequentially with L-15 medium until loss of radiolabel reached a basal rate. The appearance of radiolabel in the medium was similar for cultures washed with medium supplemented with 15% horse serum and 5 % chick embryo extract. The medium washes were pooled and concentrated prior to chromatography, The washed cell layer was treated with trypsin to remove accessible proteins, presumably those localized at the surface of the cells. Cells were removed by centrifugation and analysed separately. Gel filtration chromatographs of the cell wash, trypsin sensitive, and cell pellet material are shown in fig. 5. The distribution of radioactivity is summarized in table 2. The cell wash, containing secreted (including loosely bound) material was relatively Exp Cell Res 166 (1986)

336 Noonan, Malemud and Przybylski

10

Fig. 5. Sepharose CLdB chromatographs of proteoglycans removed from 72-h ‘?30:- radiolabeled myotube cultures. (A,B) Material removed by (A) serial washes with medium alone; (B) trypsin treatment (after media washes). (C) Material insensitive to removal by medium or trypsin treatment. -, Control; ---, nitrous acid treatment. Arrowhead, void volume; arrow, inclusion volume.

C

8 6 4i

FRACTION

NUMBER

nitrous acid-insensitive, and eluted with a large hydrodynamic size (fig. 5A), indicating it to be derived mostly from the muscle chondroitin 6 sulfate proteoglycan. The trypsin-sensitive (i.e. surface) material, contained significant amounts of nitrous acid-sensitive material, with a K,” approximately the same as the cell wash (fig. 5B). This indicated that the surface material contained both intact heparan sulfate and chondroitin sulfate proteoglycans. The material associated with the cell pellet was relatively small, and mostly nitrous acid-sensitive (fig. 5 0. There was also a significant peak of radioactivity at the inclusion volume of the column. These data suggestedthat this material may be composed of intracellular degradation products derived mostly from the surface (trypsin-sensitive) HSPG. However, some internalization of surface material could have occurred with a cell wash time of 80 min and may have been enhanced by the trypsin treatment [ 151. DISCUSSION Myotube cultures exhibited a threefold higher incorporation of 3’SOi- radiolabel over a 6-h period into heparan sulfate proteoglycans (HSP) as compared with myoblast cultures. Pulse-chase analysis of muscle cultures demonstrated that the increase in HSP was due to an increase in biosynthesis rate, as the turnover rates were the same for myoblast and myotube cultures. Pacifici $ Molinaro [27] also demonstrated an increase in 35SOi- incorporation into fip

Cell Res 166 (1986)

Skeletal muscle heparan sulfate proteoglycans

337

heparan sulfate glycosaminoglycans (HSGAG) by myotubes during a 12-h period, even in cultures treated with cytosine arabinoside to eliminate mitotic cells. Preliminary characterization of the structure of myogenic HSP was obtained from their elution profiles on Sepharose CL-6B. The glycosaminoglycan chains from the HSPG all appeared to elute in approximately the same position on Sepharose CLdB. Chromatography of 42-h samples enriched in HSPG and HSGAG on Sepharose CL-6B but equilibrated with 4 M guanidine hydrochloride eluted at approximately the same position as columns equilibrated with 0.2% SDS, indicating valid comparisons of SDS equilibrated samples with results of other studies in which 4 M guanidine hydrochloride was used. Using the data of Wasteson [34] for chondroitin sulfate glycosaminoglycans (assuming that HSGAG have similar hydrodynamic properties), a molecular weight range of 30000-40000 was estimated for the HSGAG chains from their elution position (Kav on Sepharose CLdB). While HSPG have been characterized in several mature cell types, HSPG of differentiating tissue have not been studied. Our analysis of developing skeletal muscle shows that there is a change in the synthesis and structure of HSPG when myoblasts differentiate into myotubes. The HSPG of myotube cultures was hydrodynamically smaller than that of the EHS tumor [ 121.The muscle HSPG is more similar in size characteristics to the PYS-2 teratocarcinoma cell line, derived from embryonic mouse parietal endoderm [26]. The HSPG found in muscle fibroblast cultures is smaller than the HSPG reported from other fibroblastic cells [S, 17,331 but larger than those reported for hepatocytes [18, 241. The glycosaminoglycan chains of HSPG reported for other fibroblastic cells were estimated to be 40000 molecular weight, similar to the muscle fibroblast culture HSGAG chains. Muscle fibroblast cell fraction proteoglycans bear a striking resemblance in elution characteristics on Sepharose CL6B to the cell fraction proteoglycans from SV40 transformed BALB 3T3 cells reported by Lark & Culp [20]. The hydrodynamically smaller material on Sepharose CL-6B from continuously sulfate-radiolabeled myotube cultures was insensitive to alkaline borohydride and contained variable amounts of nitrous acid-sensitive material. Since much of the trypsin-insensitive heparan sulfate eluted in this region, it was thought that this material represented internalized degradation products of the surface proteoglycan. Several other studies have reported small, alkaline reduction resistant intracellular heparan sulfate [17, 33, 361. Yanagishita & Hascall [36] demonstrated the intracellular heparan sulfate in ovarian granulosa cells to be a degradation product of the cell surface proteoglycan. A t r12of approx. 4 h was reported for removal of the HSPG from the cell layer. Two possible degradation pathways for internalized HSPG were demonstrated in the ovarian granulosa cells, a rapid one which released free sulfate with a tli2 of 30 min, and a slower one which resulted in small glycosaminoglycan chains with a degradative tll2 of 3-4 h, approximately the same rate as removal from the cell surface. The tl12 of HSPG removal for Exp CellRes 166(1986)

338

Noonan, Malemud

and Przybylski

myogenic cultures correlates well with that from the ovarian granulosa cells. A similar dual degradation pathway as in ovarian granulosa cells could explain the apparent absence of any pulse-chase relationship between the large proteoglycan and the smaller material (K,, 0.71) noted in continuously radiolabeled muscle cultures. The change in biosynthetic rate and structure of HSPG induced by myotube formation in muscle cultures is indicative of changes in the function of these molecules. As myoblasts do not accumulate a basement membrane, the HSP of myoblasts may be important in other functions of differentiating cells. Fibroblasts deposit HSPG in the extracellular matrix associated with the substrate between cells and may be important in cell-cell and cell-matrix adhesion [35]. Both myoblasts and fibroblasts in muscle cultures were found to contain laminin and type IV collagen by Kiihl et al. [193. Myotubes did not appear to contain these components internally but accumulated them in a basement membrane, suggesting that myoblasts and fibroblasts are responsible for the matrix deposited on myotubes. Olwin & Hall [25] have reported an increase in synthesis of laminin upon myotube formation in the mouse skeletal muscle cell line C2. These authors confirm the observation that surface staining of laminin was found only on myotubes. In addition, myotube formation was associated with the formation of laminin into Triton X-100 insoluble, but 4 M guanidine HCl-soluble complexes. The less soluble laminin may be that which has become associated with HSPG and type IV collagen. A change in the biosynthesis rate and possibly the structure of myotube heparan sulfate may be the initial event in accumulation of laminin, type IV collagen and other components as a basement membrane. An increase in the concentration of the cell surface heparan sulfate binding components (laminin, fibronectin or type IV collagen), may stabilize the surface arrangement of these molecules. Stabilization could reduce the degradation rate of HSPG in a similar way to the reduction of HSPG turnover by collagen in mammary epithelial cell cultures [lo]. Immunofluorescence and electron microscopy indicate that the initial site of basement membrane formation is at the neuromuscular junction-or, in the absence of neurons, acetylcholine receptor plaques [32]. Heparan sulfate proteoglycans are localized with acetylcholine receptor plaques [4,7] and neuromuscular junctions [5] early in their formation. This indicates that the HSPG of muscle may be involved in the initial formation of the basement membrane at these sites. Further studies on this system may elucidate the mechanisms of basement membrane formation. We would like to thank Vilma Szgetti for her invaluable assistance with cell cultures. This work was supported by NIH grants AM25202 and AG02205, and the Muscular Dystrophy Association.

REFERENCES 1. Abbott, J, Mayne, R & Holtzer, H, Dev biol 28 (1972) 430. 2. Agiro, V & Johnson, M, Sot neurosci abst 7 (1981) 346. Exp Cd Res I66 (1986)

Skeletal muscle heparan sulfate proteoglycans

339

3. Ahrens, P B, Solursh, M & Meir, S, J exp zoo1 202 (1977) 375. 4. Anderson, M J & Fambrough, D M, J cell bio197 (1983) 13%. 5. Anderson, M J, Klier, F G & Tanguay, K E, J cell bio199 (1984) 1679. 6. Banejee, S D, Cohn, R H & Bernfield, M R, J cell bio173 (1977) 445. 7. Bayne, E K, Anderson, M J & Fambrough, D M, J cell bio199 (1984) 1486. 8. Carlstedt, I, Coster, L & Malstrom, A, Biochem j 197 (1981) 217. 9. Carrino, D A & Caplan, A I, J biol them 257 (1982) 14145. 10. David, G & Bernfield, M, J cell biol91 (1981) 281. 11. Gill, P J, Adler, J, Silbert, C K & Silbert, J E, Biochem j 194 (1981) 299. 12. Hassell, J R, Robey, P G, Barrach, H, Wilczek, J, Rennard, S I & Martin, G R, Proc natl acad sci US 77 (1980) 4494. 13. Hill, B T & Whatley, S, FEBS lett 56 (1975) 20. 14. Ho, P, Levitt, D & Dorfman, A, Dev biol55 (1977) 233. 15. Huet, C H & Hertzberg, M, J ultrastruct res 42 (1973) 186. 16. Kanwar, Y S, Linker, A & Farquhar, M A, J cell bio186 (1980) 688. 17. Kleinman, H K, Silbert, J E & Silbert, C K, Connective tissue res 4 (1975) 17. 18. Kjellen, L, Pettersson, I & Hiiik, M, Proc natl acad sci US 78 (1983) 5371. 19. Kiihl, U, Timpl, R & von der Mark, K, Dev bio193 (1982) 344. 20. Lark, M W & Culp, L A, J biol them 259 (1984) 574. 21. Matuoka, K & Mitsui, Y, Cell struct func 6 (1981) 23. 22. Noonan, D M, Culp, L A & Przybylski, R J, J cell bio195 (1982) 118a. 23. Noonan, D M, Thesis, Case Western Reserve University, Cleveland, Ohio (1985). 24. Oldberg, A, Kjellen, L & Ha(ik, M, J biol them 254 (1979) 8505. 25. Olwin, B B & Hall, Z W, Dev biol 112 (1985) 359. 26. Oohira, A, Wight, T N, McPherson, J & Bornstein, P, J cell bio192 (1982) 357. 27. Pacifici, M & Molinaro, M, Exp cell res 126 (1980) 143. 28. Ptzybylski, R J, Bullaro, J C & MacBride, R G, Am j physio1237 (1972) cl66 29. Rapraeger, A C & Bernfield, M, J biol them 258 (1983) 3632. 30. Rollins, B J, Cathcart, M K & Culp, L A, The glycoconjugates (ed M Horowitz) vol. 3, p. 289. Academic Press, New York (1982). 31. Sandra, A, Leon, M A & Przybylski, R J, J cell sci 28 (1977) 251. 32. Sanes, J R, J cell bio193 (1982) 442. 33. Vogel, K G & Peterson, D W, J biol them 256 (1981) 13235. 34. Wasteson, A, J chrom 59 (1971) 87. 35. Woods, A, Hook, M, Kjellen, L, Smith, C G & Rees, D A, J cell bio199 (1984) 1743. 36. Yanagishita, M & Hascall, V C, J biol them 259 (1984) 10207. Received February 7, 1986 Revised version received April 15, 1986

Printed in Sweden

Exp Cell Res 166 (1986)