Analysis of cartilage differentiation from skeletal muscle grown on bone matrix

DEVELOPMENTAL

BIOLOGY

78,332-351 (1980)

Analysis of Cartilage

II. Chondroitin

Differentiation from Skeletal Muscle Grown on Bone Matrix

Sulfate Synthesis

and Reaction

MARK A. NATHANSON' Department of Anatomy, Harvard Medical

to Exogenous

AND ELIZABETH

Glycosaminoglycans

D. HAY

School, 25 Shattuck Street, Boston, Massachusetts 02115

Received August 1, 1979; accepted in revised form January 3, 1980 In the first paper in this series (Nathanson, M. A., and Hay, E. D. (1980). Develop. Biol. 78, 301-331), we described the ultrastructural alterations that take place when embryonic skeletal muscle is induced to form hyaline cartilage by demineralized bone matrix in vitro. In this paper, we analyze the pattern of appearance of chondroitin sulfates and dermatan sulfate in injured muscle in situ and in explants of muscle cultured either on bone matrix or on collagen gel. We also investigate the effects of exogenous glycosaminoglycans on the cultures to determine whether chondroitin sulfate @h-S) and hyaluronic acid (HA) can enhance or inhibit the biochemical differentiation of cartilage under these conditions. Our results indicate that during the first morphological phase, 1-3 days in vitro, there is an increased sulfate uptake, a shift in the relative abundance of Ch-S, and an increase in the ratio of chondroitin-4-sulfate (Ch-4-S) to chondroitinB-sulfate (Ch-6-S); this change is correlated with the transformation of myoblasts to fibroblast-like cells in both types of cultures. A similar increase in the Ch-4-S/Ch-6-S ratio occurs in injured muscle in situ, suggesting that phase I is a regenerative response. Explants on bone matrix sustain Ch-4-S levels between 4 and 5 days (phase II) and show a large increase in Ch-4-S and sulfate incorporation when they form cartilage at 6-10 days (phase III). Explants on collagen gels regenerate muscle at 4-10 days with decreasing Ch-4-S/Ch-6-S ratios and decreasing sulfate incorporation. The data demonstrate that an environmental influence, such as trauma, is sufficient to alter the biosynthetic expression of skeletal muscle and that under appropriate conditions (such as the presence of bone matrix) this response may be augmented, leading to the synthesis of extracellular matrix components at ratios characteristic of cartilage. Exogenous Ch-S and HA did not significantly effect this overall pattern. These results are discussed in relation to the morphological observations presented in the preceding paper. INTRODUCTION

When embryonic skeletal muscle is cultured on demineralized bone (bone matrix) in vitro, it gives rise to cartilage (Nogami and Urist, 1970, 1974a,b; Anderson and Griner, 1977; Nathanson et al., 1978; Nathanson and Hay, 1980). The ability to differentiate into cartilage is not restricted to a single precursor cell type, but is shared by cells of several mesenchymal origins (Urist et al., 1969; Nathanson et al., 1978) The cells which are challenged to form cartilage ’ Present address: Department Jersey Medical School, 100 Bergen Jersey 07103.

of Anatomy, St., Newark,

0 1980 by Academic

of reproduction

in any

chondrogenic lineage. While nonchondrogenie cells can be induced to form cartilage only by culturing them on bone matrix, the ultrastructural aspects of their subsequent differentiation closely resemble those of normal chondrogenesis (Anderson and Griner, 1977; Nathanson and Hay, 1980). Skeletal muscle cultured on bone matrix gives rise to a population of fibroblast-like cells which derive from the myoblasts and fibroblasts that were originally explanted (phase I, Table 1). After several days on bone matrix (phase II, Table l), these fibroblast-

332

0012-1606/80/100332-20$02.00/O Copyright All rights

New New

in vitro would not normally have done so in vivo and thus lie outside of a direct

Press, form

Inc. reserved.

NATHANSON

tally low in skeletal muscle and high in cartilage. Ten-day-old avian vertebral cartilage, for example, contains increased proportions of Ch-4-S relative to Ch-6-S (Shulman and Meyer, 1968). Variability in chondroitin sulfate isomers, however, has been reported in somite, epiphyseal, articular, costal, cricoid, thyroid, and tracheal cartilages in the chick embryo and in other species (Marzullo and Lash, 1967; Robinson and Dorfman, 1969; Ogston, 1970;Kosher et al., 1973; Mathews, 1975). The proportions of the chondroitin sulfate isomers have also been said to vary among different adult nonchondrogenic cells and tissues (Conrad et al., 1977; Dietrich et al., 1976) and in relation to hyaluronic acid at different embryonic stages (Solursh and Morriss, 1977). The pattern we observed in rat limbs, however, is quite consistent during chondrogenesis

like cells alter their phenotype to that of chondrocytes (phase III, Table 1). Embryonic chondrocytes are not present as contaminants and, moreover, comparable cultures consisting of explants on collagen gels produce only skeletal muscle (Nathanson and Hay, 1980). In this communication, we report on the types and amount of the principal sulfated glycosaminoglycans (GAG) synthesized by skeletal muscle explants transforming into cartilage on bone matrix and by cultures regenerating muscle on collagen gels. The two isomers of chondroitin sulfate (chondroitin-4-sulfate and chondroitin-6-sulfate) and an epimer (dermatan sulfate, formerly known as chondroitin sulfate B) are analyzed throughout the culture period. Particular attention is given to the ratio of chondroitin-4-sulfate (Ch-4-S) to chondroitin-6sulfate (Ch-6-S), because this ratio is typiTABLE TIME

COURSE

OF CARTILAGE

Days in culture

1

FORMATION

ON BONE

Morphological Bone

0

Skeletal sis.

muscle

contains

IN VITRO *

Collagen

heterochromatic

myoblasts

and fibroblasts;

gel

myotubes

undergoing

necro-

I

Myoblasts and tibroblasts become euchromatic; myoblasts develop rough endoplasmic reticulum. Myoblast transition to fibroblast-like cells nearly complete.

2-3

MATRIX

alteration

matrix

Phase 1

333

In Vitro Chondrogenesis

AND HAY

Myoblasts and fibroblasts become euchromatic; myoblasts develop rough endoplasmic reticulum. Myoblasts regain myoblast-like morphology.

Phase II 4-5

Mononucleate cells entirely fibroblast-like. broblast-like cells large and contain endoplasmic reticulum. Myofibrillar occurs throughout explant.

Firough debris

Phase 6

myoblasts, myotubes apparatus.

Newly formed myotubes contractile apparatus. domly through explant. disappearing.

extensive course randebris is

III

First appearance of chondrocytes, surrounded by cartilage type collagen fibrils. Fibroblastlike cells outnumber chondrocytes. Chondrocytes increase in number; chondrocytes and fibroblast-like cells intermingle.

8 10

n Data

Explant consists of skeletal muscle, and few tibroblasta. Newly formed contain well-formed contractile Myofibrillar debris is present.

Discrete regions of explant contain chondrocytes; other regions contain fibroblast-like cells. Myofibrillar debris is disappearing. from

Nathanson

and Hay

(1980).

developing Myotubes Myofibrillar

334

DEVELOPMENTAL

BIOLOGY

in vitro; the direction is always toward that characteristic of rat embryonic limb cartilage in situ, that is, toward an increased Ch-4-S level relative to that of Ch8-S. In this paper, we also explore the effect of exogenous GAG on sulfated GAG synthesis in the two types of cultures. Changes in proportions of chondroitin sulfates and dermatan sulfate should reflect different functional states, and local factors within the embryo may control their expression. Indeed, exogenous chondroitin sulfate and chondromucoprotein have been reported to enhance chondroitin sulfate and cartilagespecific proteoglycan synthesis (Kosher et al., 1973; Kosher and Lash, 1975; Meier and Hay, 1974; Nevo and Dorfman, 1972; Schwartz and Dorfman, 1975; Lash and Vasan, 1977) and exogenous hyaluronic acid (HA) to inhibit chondroitin sulfate synthesis (Solursh et al., 1974; Toole, 1973b; Toole et al., 1972). The onset of cartilage differentiation in regenerating amphibian limbs coincides with the appearance of a hyaluronidase (Toole, 1973a; Toole and Gross, 1971; Smith et al., 1975). A similar sequence has been noted in the chick limb and cornea (Toole, 1972; Toole and Trelstad, 1971). These observations have led to the hypothesis that HA acts to inhibit the differentiation of mesenchymal cells, possibly by preventing their aggregation, and that the time-dependent appearance of hyaluronidase removes the inhibitory agent (Toole, 1972). Since cartilage appears at a precise time in bone matrix cultures, this system provides a good subject for analysis of the effect of such factors on sulfated GAG synthesis by differentiating mesenchymal cells in organ culture. MATERIALS

Preparation

AND

METHODS

of Organ Cultures

Organ cultures were prepared as described previously (Nathanson and Hay, 1980). Briefly, skeletal muscle was isolated from the upper arms and thighs of embryonic rat fetuses at 19 days of gestation

VOLUME

%I980

(timed pregnant rata from Charles River Laboratories, Wilmington, Mass.), cleaned of contaminating tissue, and minced in culture medium. During isolation the embryonic muscle was held in ice-cold Hanks’ balanced salt solution, pH 7.4. Cleaned and minced skeletal muscle was explanted onto hemicylinders of bone matrix or, in control cultures, onto gels of type I collagen. Bone matrix was prepared and used as described by Urist (see Nathanson et al., 1978). After hydration in culture medium, diaphyseal bone matrix cylinders were split longitudinally to form hemicylinders and narrow strips of bone matrix were removed from the edges of the hemicylinders for use as tissue overlays. Hemicylinders and tissue overlays were then coated with chicken plasma (Grand Island Biological Co., Grand Island, N. Y.). Cellulose ester filters (pore size 0.45 pm; Millipore Corp., Bedford, Mass.) were coated with at least three applications of a gel of type I collagen (Nathanson and Hay, 1980). Coated filters were air dried, sterilized in 70% ethanol, and either air dried and stored or rehydrated in culture medium. Stainless-steel organ culture grids were placed over the center well of organ culture dishes (Falcon Plastics, Los Angeles, Calif.), in an inverted position, and the bone matrix and collagen gel substrata were placed onto the grids. Aliquots of the minced muscle were then placed onto the substrata. Each explant onto bone matrix was overlaid with the narrow strip of bone matrix; explants onto collagen gels remained uncovered, except for the meniscus of culture medium. Special attention was paid to the amount of tissue explanted such that bone matrices and collagen gels received similar amounts of tissue. The culture medium consisted of medium CMRL-1066 (prepared from 10X concentrate; containing L-glutamine) containing 15% fetal calf serum and 1% antibiotics (penicillin-streptomycin, from a stock concentration of 10,000 units/ml and 10,000 pg/ml, respectively; final concentration, 100

NATHANSON

AND HAY

units/ml-100 pg/ml; Grand Island Biological Co., N.Y.). Prior to use fetal calf serum was heat inactivated by heating it to 56°C for 45 min. The culture medium was sterilized by passing it through a sterile, graded filter series, under pressure, with a final pore size of 0.22 pm (Millipore Corp.). Explants were cultured in a water-jacketed incubator (National Appliance Co., Portland, Ore.), at 37”C, for varying periods of time as described below. The atmosphere was maintained at 5% COn in air, at a gas flow rate equal to the volume of the incubator chamber per hour. Feeding was performed on alternate days by replacing onehalf to three-fourths of the medium. Humidity was provided by saturating an absorbent pad in the outer well of the organ culture dishes with sterile water and by placing two trays of deionized water directly into the incubator chamber.

In Vitro Chondrogenesis

Analytical

335

Methods

Uranic acid was determined by a modified carbazole method (Bitter and Muir, 1962). Chondroitin sulfate isomers were quantitated by the chondroitinase method (Saito et al., 1968; Yamagata et al., 1968), using the buffer system described below. Total DNA was determined by a modification of the fluorometric method of Kissane and Robbins (Santoianni and Ayala, 1965), modified further by measuring the fluorescent product in HCl rather than the original perchloric acid (Hinegardner, 1971). For comparison all of the data have been normalized to represent measured incorporation per lo6 cells, using rat cardiac muscle (6.5 pg DNA/nucleus) as representative diploid nuclei. Purified glycosaminoglycan standards were the generous gift of Dr. J. A. Cifonelli. 3,5-Diaminobenzoic acid dihydrochloride was purchased from EastAddition of Exogenous Glycosaminoglyman Kodak Co., Rochester, N.Y. Chondrocans itinase ABC, chondroitinase AC-II, and To test the effect of exogenous glycosa- purified disaccharide standards, A-di-6-S(2minoglycans on the biosynthesis of chon- acetamido-2-deoxy-3-0-(/3-n-gluco-4-enedroitin sulfate, chondroitin sulfate (grade pyranosyluronic acid) - D - galactose - 6 - sulIII, mixed isomers, lot 52C-3250, Sigma fate) and A-di-4-S(2-acetamido-2-deoxy-3Chemical Co., St. Louis, MO.) and hyaluO(p-D-gluco-4-enepyranosyluronic acid)-Dronic acid (grade IIIS, lot 34C-2920, Sigma galactose-4-sulfate), were purchased from Chemical Co.) were prepared as stock so- Miles Laboratories, Elkhart, Ind. All other lutions of 16 and 8 mg/ml, respectively, and components were reagent grade. added to stock aliquots of culture medium The synthesis of chondroitin sulfates and to a final concentration of 200 pg/ml. The dermatan sulfate was measured at 1,2,3,4, experimental cultures were fed with culture and 10 days in vitro. The first 4 days commedium containing these glycosaminoglyprises the prechondrogenic phase, and 10 cans throughout the entire culture interval, days is within the chondrogenic phase. At whereas controls received the medium only. the termination of the 24-hr labeling period The glycosaminoglycan-containing culture the cultures were frozen at -20°C until medium was initially added to the dishes, assayed. The culture medium was disvia pipet, directly over the explanted skel- carded, as we wanted to measure chondroietal muscle to dilute out the original me- tin sulfates contained within actual extradium. Glycosaminoglycan synthesis was cellular matrix. When sufficient samples monitored by feeding the cultures with a had accumulated they were thawed on ice, complete change of the respective culture homogenized completely in 0.5 ml of 0.2 M medium containing 10 &i/ml Na235S04 Tris-HCl (adjusted to a pH of 8.0 at 5O”C), (carrier free, New England Nuclear, Bos- and maintained on ice. Homogenization ton, Mass.) for the final 24 hr prior to was performed in a 1.0~ml ground-glass hoharvesting. mogenizer with a conical base (Bellco Bio-

336

DEVELOPMENTAL BIOLOGY

logical Glassware, Vineland, N.J.). Cultures on bone matrix were homogenized with the matrix. Cultures on collagen gels were manually lifted from the supporting filter under dark-field illumination, and the explant plus the collagen gel were homogenized together. Following homogenization a 50-~1 aliquot was transferred to a siliconized 3.0ml centrifuge tube and frozen for subsequent DNA assay. The remainder, plus several washes of the homogenizer, was transferred to a siliconized 15-ml Corex centrifuge tube (Corning Glass Works, Corning, N.Y.) and held on ice until all of the samples had been processed. Endogenous enzymes were destroyed by boiling for 5 min and the glycosaminoglycans were removed from the proteoglycan by digestion with predigested Pronase in the presence of calcium chloride and ethanol (de la Haba and Holtzer, 1965). The Pronase digestion was carried out at 50°C for 48 hr, with 2.4 mg Pronase added initially and an equal amount added after 24 hr. Digestion was terminated by placing the homogenates into a boiling water bath for 10 min. High molecular weight sulfated material was separated from unincorporated label by gel filtration (Sephadex G-25, Pharmacia, Uppsala, Sweden; 2.5 X 28.5 cm) in 20% ethanol containing 0.1 M ammonium acetate. The Sephadex columns were calibrated with standard glycosaminoglycans, blue dextran-2000 (molecular weight, 2 x 106;Pharmacia), phenol red (0.5% aqueous solution; Grand Island Biological Co., Grand Island, N.Y.), Na235S04, and 3H20. On one representative column both blue dextran and the standard glycosaminoglycans eluted at the void volume (Vi = 58 ml), while Na235S04eluted in a broad zone at 90-140 ml. The column was deliberately overloaded with Na235S04(5 $i) to verify that it would not overlap the glycosaminoglycan peak. Tritiated water yielded a total running volume for the column ( Vs) of 11% 120 ml. Phenol red was retarded and eluted as a broad zone at 140-200+ ml. Insofar as

VOLUME 78,198O

culture medium components may be similarly retarded, the columns were extensively washed between each run. The high molecular weight material was located by counting 100 ~1 of each fraction (2.0 ml) in Aquasol (New England Nuclear, Boston, Mass.) and plotting the location of the radioactive peak. No degradation was found and the peak was consistently symmetrical at the VO.Fractions containing high molecular weight sulfated material were pooled in siliconized 50-ml Erlenmeyer flasks, frozen in liquid nitrogen, and lyophilized to dryness. Total incorporation was determined by dissolving the lyophilized material in 1.0 ml deionized water and counting 0.5 ml in Aquasol. Corrections were made for background and the final counts were corrected to refer to the entire sample. Counting was performed in a Beckman LS100 liquid scintillation counter with a window of O-590. Counting efficiency was determined to be 7580% by the external standard-channels ratio method. External standard ratios demonstrated that similar quench was present in all samples. A remaining 0.4 ml of the high molecular weight sulfated material was transferred to a siliconized Corex centrifuge tube, frozen in liquid nitrogen, and again lyophilized to dryness. For the determination of the rate of sulfate incorporation by the explanted cells, we assume that sulfate is taken up at a constant rate during the 24-hr labeling period. “Rate” is thus defined for present purposes as the amount of sulfate incorporated into high molecular weight material per lo6 cells per hour over each 24-hr period. “Change in rate” or “acceleration” is defined as r2 - rl/I, where r2 is one measured rate, rl is the rate immediately preceding, and I is the interval in hours. Standard glycosaminoglycans were treated with several concentrations of chondroitinase to determine optimal conditions for the assay (Method 3 of Yamagata et al., 1968) (Fig. 1). Reactions were

NATHANSON

AND

HAY

carried out in a buffer consisting of 0.05 M Tris-HCl-0.05 M NaCl. Chondroitinase ABC and chondroitinase AC-II were used at pH 8.0 and pH 7.3, respectively; the buffers were adjusted to yield these pH values at 37°C. Control experiments demonstrated that under the present conditions 300 pg of the standard glycosaminoglycans (chondroitin4-sulfate, dermatan sulfate, chondroitin-6-

IO

337

In Vitro Chondrogenesis

sulfate, hyaluronic acid) was degraded by 0.05 and 0.25 unit of chondroitinase ABC and AC-II, respectively (Fig. 1). The kinetics were determined by preparing a duplicate series containing 300 pg of each standard glycosaminoglycan (as uranic acid) in the appropirate buffer. After preincubation at 37°C the enzyme was added. Final volume was 100 ~1. Aliquots of 20 ~1 were removed at 5, 10, 20, 40, and 60 min and

CHONOROITINASE

ABC

CHONOROITI

AC3

20 30 40 MINUTESAT 37” C

NASE

50

60

FIG. 1. Enzymatic depolymerization of purified glycosaminoglycan standards by chondroitinases ABC and AC-II. Each standard was present at an initial concentration of 300 pg in a buffer consisting of 0.05 M Tris-HCI containing 0.05 M NaCl, pH 8.0 for chondroitinase ABC and pH 7.3 for chondroitinase AC-II. The reaction was carried out at 37°C. Aliquots corresponding to 60 pg each were taken at the indicated intervals after adding 0.05 unit chondroitinase ABC or 0.25 unit chondroitinase AC-II, and the extent of depolymerization was measured by uv absorbance. 100% depolymerization corresponds to 0.12 fl chondroitin sulfate and 0.14 fl hyaluronic acid.

338

DEVELOPMENTAL

BIOLOGY

added to 0.9 ml of 0.05 M KCl-HCl buffer, pH 1.8, to stop the reaction. Each aliquot thus contained 60 pg, corresponding to either 0.12 fl chondroitin sulfate or 0.14 @f hyaluronic acid. The time at which the entire 60 pg aliquot was completely degraded to disaccharides corresponds to complete degradation of the 300~pg sample. Both chondroitinase ABC and chondroitinase AC-II degraded the 6O+g aliquot within 60 min. With chondroitinase ABC, hyaluronic acid remained essentially undegraded, however, some batches of this enzyme degraded hyaluronic acid to a slightly greater extent, and 0.1 unit of the enzyme was capable of degrading 120 pg of hyaluronic acid within 120 min. Chondroitinase AC-II was active toward hyaluronic acid, but inactive toward dermatan sulfate. In each case the enzymes did not display detectable disaccharidase activity. Standard conditions for the enzymatic depolymerization of chondroitin sulfate isomers were as follows: The lyophilized samples were dissolved in 100 ~1 of deionized water and 50 ~1 was transferred to a second set of test tubes. Thirty microliters of 2x buffer, pH 8.0, was added to one set, followed by 0.05 unit (in 10 ~1) of chondroitinase ABC. The second set received the same amount of buffer, pH 7.3, followed by 0.25 unit (in 10 ~1) of chondroitinase AC-II. The reaction was allowed to proceed for 40 min at 37’C, at which time additional enzyme (as before) was added. The reaction was allowed to proceed for a total of 120 min. The tubes were then placed on ice and immediately spotted, along with standard disaccharides, on prescored Whatman 3MM chromatography paper and chromatographed for 20-24 hr in Solvent A (Saito et al., 1968), consisting of N-butanol, acetic acid, ammonium hydroxide (2:3:1). After drying, the positions of standard disaccharides were visualized with 252-nm uv light and the chromatograms were cut into 0.5in. segments. Individual segments were eluted with 1.0 ml of deionized water, in

VOLUME

78,198O

glass scintillation vials, and counted in Aquasol. The position of the radioactive disaccharides was plotted along with the position of the standard disaccharides. Sulfate incorporation into disaccharides was determined by summing the counts in each peak which migrated with an R, similar to that of standard disaccharides. Correction was made for background and the final counts were corrected to refer to the entire sample. RESULTS

Sulfate Incorporation Cartilage in Situ

into

Muscle

and

Since our working premise was that the biosynthesis of chondroitin sulfates and dermatan sulfate would provide a useful index of cartilage differentiation (Introduction), it was necessary first to determine whether these are the principal sulfated glycosaminoglycans (GAG) of embryonic rat cartilage. For this determination a pregnant rat was injected with 100 @i of Na235S04 at 18 days of gestation and sacrificed 24 hr later. Sternal cartilage and thigh muscle was isolated from the 19-day fetuses and assayed for (1) total incorporation into high molecular weight sulfated material and (2) the percentage of this material in chondroitin sulfates and dermatan sulfate.

INCORPORATION CARTILAGE Tissue

Sternal Thigh

cartilage muscle

TABLE 2 OF SULFATE BY EMBRYONIC AND SKELETAL MUSCLE” Total incorporation (cpm/ 10” nuclei)

Percentage of total incorporation into chondroitin sulfates and dermatan sulfate

64,289

94.7

2,713

100.0

0 Embryonic sternal cartilage and thigh muscle were labeled by injecting a pregnant rat (at 21 days of gestation) with 1 mCi Nan ‘“SO4 and isolating the fetal tissues after a 24-hr labeling period. The data represent the incorporation of labeled sulfate into high molecular weight material from pooled sibling tissues expressed per lob; nuclei.

NATHANSON

AND HAY

The results demonstrate that sternal cartilage incorporates 24 times as much sulfate as skeletal muscle into GAG and that both tissues incorporate the isotope essentially only into chondroitin sulfates and dermatan sulfate (Table 2; see also Ebert and Prockop, 1967). Since this extracellular product is preferentially synthesized by chondrocytes, chondroitin and dermatan sulfate synthesis can be used to monitor

In Vitro

339

Chondrogenesis

differentiation of cartilage skeletal muscle.

from embryonic

Sulfate Incorporation by Explants Bone Matrix and Collagen Gels

onto

The levels of incorporation of Na235S04 by cultures explanted onto bone matrix (open circles, Fig. 2) can be related to the three phases (l-3 days, 4-5 days, 6-10 days) of morphological alterations previously de-

150I IO70-

IO6-

‘4------T 2

3 DAYS IN CULTURE

4

FIG. 2. Incorporation of labeled sulfate into high molecular weight material by explants of minced skeletal muscle grown in vitro on bone matrix or as a control on gels of type I collagen. Explants were labeled by a complete change of medium containing 10 @i/ml Nan”“S04 for 24 hr prior to harvesting at the indicated intervals. Each time point corresponds to the mean sulfate incorporation (&standard errors) of six cultures. The standard error bars indicate a 95% confidence level. Explants onto bone matrix give rise to fibroblast-like cells after 2 days and chondrocytes after 6 days. Explants onto collagen gels regenerated skeletal muscle by 4 days (Nathanson and Hay, 1980).

340

DEVELOPMENTAL

BIOLOGY

scribed (Table 1). The first morphological phase (phase I) is characterized by a slight initial rise in sulfate incorporation to a level somewhat above that of the control cultures onto collagen gels (Fig. 2, closed circles, l-3 days). The second morphological phase (4-5 days in vitro) occurs prior to the appearance of cartilage in the bone matrix cultures, when the mononucleate cells are entirely fibroblast-like in morphology; on collagen gels muscle is forming 4-5 days in vitro (Table 1). At this time, explants onto bone matrix increase their sulfate incorporation to 4.3 times that of the control cultures, while explants on collagen gels remain at the baseline level (Fig 2). During the third phase, 6 to 10 days in vitro, sulfate incorporation by explants on bone matrix increases to 22 times that of the cultures on collagen gels (Fig. 2). During this time, overt cartilage appears in the explants on bone matrix, while those on collagen gels continue myogenesis. The three phases (Table 1) are also characterized by changes in the rate of sulfate uptake. Cells in the first phase incorporate 140 cpm 35S04/106 cells/hr from day 1 to day 2, followed by 30 cpm 35S04/10” cells/ hr from day 2 to day 3. Similar cultures between 3 and 4 days incorporate 590 cpm 35S04/106 ceIls/hr, or 4.2 times greater than at the beginning of the first phase. The extended interval of the third phase (6-10 days) and the increased level of sulfate incorporation necessitated making the abscissa and ordinate in Fig. 2 nonlinear. The change in rate (acceleration) of sulfate uptake, however, increases only slightly, to 666 cpm 35S04/106 cells/hr. The increase in sulfate incorporation into cultures on bone matrix at 4 days in culture occurs at a time when the total DNA in the cultures is declining (Fig. 3). The decline in DNA is due partly to the disappearance of necrotic myotube nuclei; no obvious cell death was observed among the mononucleate cells (Nathanson and Hay, 1980). If the decline is due mainly to the removal of

VOLUME

78,1%0

necrotic DNA, the decrease observed at 4 days on bone matrix might contribute to the observed increase in sulfate incorporation as expressed on a per-cell basis (Fig. 2). However, control cultures which exhibit a consistent decline in total DNA do not increase their incorporation of sulfate on a per-cell basis. It seems reasonable to conclude that the surviving mononucleate cells are responsible for the observed increase in sulfate uptake on bone matrix. Effect of Exogenous GJycosaminoglycans on Sulfate Incorporation Exogenous GAG (chondroitin sulfate, Ch-S; hyaluronic acid, HA) added to the cultures were present in the medium throughout the entire culture interval at an uncorrected concentration of 200 pg/ml. This concentration was selected because of its reported positive effect on chondrogenesis (Nevo and Dorfman, 1972). The ChS preparation contained 79% Ch-S (as uranic acid) and consisted mainly of chondroitin-6-sulfate. The hyaluronic acid preparation contained 92% hyaluronic acid (as uranic acid). Of the 200 pg/ml, Ch-S and HA were present at concentrations of 158 and 184 pg/ml, respectively. Neither GAG had any effect on the rate or absolute levels of sulfate uptake by explants onto bone matrix (Fig. 4; Ch-S treated, open squares; HA treated, open triangles; untreated cultures, open circles). Similarly, exogenous Ch-S and HA did not alter the pattern of sulfate incorporation by control explants onto collagen gels (data not shown). The similarity of sulfate incorporation values (per cell) among the treated and untreated cultures demonstrates that, for all practical purposes, they are identical. There was some variation in total DNA (Fig. 3); this point requires further study. Synthesis of Chondroitin matan Sulfate

Sulfates

and Der-

A portion of the high molecular weight sulfated material isolated from 19-day em-

NATHANSON

AND

HAY

bryonic rat thigh muscle and sternal chondrocytes was separated into chondroitin sulfate isomers and dermatan sulfate by degradation with the enzymes chondroitinase ABC and chondroitinase AC-II (Table 3). Nineteen-day embryonic rat thigh muscle incorporates sulfate mainly into chondroitin-8sulfate (Ch-6-S) (62.5%), with the remainder being incorporated almost equally into chondroitin-4-sulfate (Ch-4-S) and dermatan sulfate (DS). In contrast, embryonic cartilage from these same embryos incorporates most of the sulfate into Ch-4-S, with 6.7% entering Ch-6-S and none entering DS. Whereas 100% of the sulfate incorporated by thigh muscle was incorporated into Ch-S and DS, only 97.4% was similarly incorporated by sternal chondro-

341

In Vitro Chondrogenesis

cytes. Cartilage thus differs from skeletal muscle not only in the amount of sulfate incorporated (Table 2), but also in the sulfated GAG types synthesized. For convenience, these biochemical parameters will be compared by Ch-4-S/Ch-6-S ratios (Table 3, ratio = 13 for cartilage and 0.3 for skeletal muscle). A portion of the high molecular weight sulfated material from both untreated and GAG-treated cultures was similarly analyzed. Some variation in terms of percentage values was found from experiment to experiment; however, the pattern of appearance of the individual chondroitin sulfate isomers was consistent within each experiment. These data are shown in Fig. 5 as sulfate incorporation of single representa-

BONE MATRIX

COLLAGEN

GEL

B.

DAYS IN CULTURE FIG. 3. DNA content, expressed as total nuclei, of explants of minced skeletal muscle onto bone matrix (left) or onto collagen gels (right) without exogenous GAG (dotted lines) and treated with exogenous GAG (solid lines: Ch-S, squares; HA, triangles). After homogenization, an aliquot of the homogenate from each culture was assayed for total DNA (ng) and converted into total nuclei using the DNA content of a single rat cardiac muscle cell as a representative standard. Each point represents the average number of nuclei (*standard errors) of duplicate cultures.

342

DEVELOPMENTAL

BIOLO~GY

tive cultures. As early as 24 hr after explantation, cultures on bone matrix increase their synthesis of Ch4-S from an initial level of 18% (thigh muscle, Table 3) to 52% of the total sulfated material, whereas Ch6-S declines from a level of 63% (skeletal muscle) to 17% (Fig. 5A). DS displays no consistent pattern, averaging about 20% of the total sulfate incorporated by explants

VOLUME 78,198O

onto bone matrix. These levels of sulfate incorporation are maintained through 4 days in culture (phase III), even though total sulfate incorporation begins to rise at 4 days. Subsequently, explants onto bone matrix show a rise in the synthesis of Ch-4S and Ch-6-S declines. By the termination of the experiment, Ch-4-S comprises 75% and Ch-6-S comprises 5% of the total sul-

190 150 110

6

2 1 I

I

2

I

3 DAYS IN CULTURE

I

4

\ 1’6

FIG. 4. Incorporation of labeled sulfate into high molecular weight sulfated material by explants of minced skeletal muscle grown on bone matrix and either left untreated (circles) or treated with Ch-S (squares) or HA (triangles). Each point represents the mean sulfate incorporation of duplicate cultures (*standard errors, the mean of indicated by asterisks when overlapping). The curve labeled “bone matrix . ” in Fig. 2 represents these points at the indicated times and summarizes the standard errors, demonstrating the high degree of similarity among these cultures.

NATHANSON

AND

HAY

fated material. It is noteworthy that even though the explanted cells incorporate sulfate at greatly increased levels and synthesize increased amounts of Ch-4-S, the synthesis of DS is maintained. The synthesis of DS correlates well with our finding that not all of the explanted cells form chondrocytes at the same time; in each experiment there remains a population of fibroblastlike cells at these time intervals (Nathanson and Hay, 1980). The cultures onto collagen gels are also found to alter their synthesis of chondroitin sulfate isomers, presumably as a response to mincing (see last section under Results). After 24 hr in culture, Ch-4-S comprises 50% and Ch-6-S comprises 20% of the total sulfated material. The cultures do not sustain this shift, however, and the level of Ch4-S declines to approximately 33% of the total sulfated material at 2 days, where it remains thereafter (Fig. 5D). Ch-6-S is maintained between 14 and 20% of the total sulfated material. While the levels of Ch-4S and Ch-6-S are altered by the in vitro procedures, they do not completely revert to levels characteristic of skeletal muscle, even though skeletal muscle regenerates 410 days after explantation (Table I). Explants onto bone matrix incorporate more sulfate into Ch-S and DS than control cultures on collagen gels, but these values (Fig. 5) are somewhat less than the 95% level of cartilage (Table 2), even during the phase of chondrogenesis.

Synthesis of Chondroitin sponse to Exogenous cans

Sulfates in ReGlycosaminogly-

The synthesis of Ch-S and DS by skeletal muscle explanted onto bone matrix was found to be nearly identical in the presence or absence of exogenous Ch-S and HA (Figs. 5B and C). At 4 days, Ch-S-treated cultures on bone matrix synthesize Ch4-S at a level of 55-58%, and HA-treated, 64% of the total sulfated material. By the termination of the experiment, however, all of

343

In Vitro Chondrogenesis TABLE

3

INCORPORATION OF SULFATE INTO CHONDROITIN SULFATES AND DERMATAN SULFATE BY EMBRYONIC CARTILAGE AND SKELETAL MUSCLES Tissue

Percentage Chondroitin-6.sulfate

Sternal

6.7

of total corporation Dermatan sulfate

sulfate Chondroitin-4-sulfate

in-

Chondroitin-lsulfate/chondroitin-6-z&fate

0

88.0

13.09

19.2

17.7

0.28

cart-

ilage Thigh muscle

62.5

o A portion of the high molecular weight sulfated material used for the determination shown in Table 2 was degraded into the constituent chondroitin sulfate isomers by chondroitinases ABC and AC-II. The data are corrected to refer to the entire pooled sample shown in Fig. 2.

the cultures synthesize Ch-4-S at nearly equivalent levels (untreated = 75%, Ch-S treated = 80%, HA treated = 79%). The synthesis of Ch-6-S accounts for 14-19% of the total sulfated material at the beginning of the experiment and falls continuously to a level of 4-5.5% by the termination of the experiment. Explants cultured on collagen gels synthesize Ch-S isomers at variable levels in the presence of exogenous GAG (Figs. 5E and F). In the presence of exogenous Ch-S, the synthesis of Ch-4-S declines from 1 to 2 days, to reach a value of 23% of the total sulfate incorporation (Fig. 5E), Ch-4-S increasing at 10 days to a level of 42% of the total sulfate incorporation. HA-treated control cultures are even more variable in the synthesis of Ch-4-S, in that the initial level is only 27% of the total sulfate incorporation, while the final value is 40%. The synthesis of Ch-6-S by treated cultures is similar in value to that by the untreated controls. While the treated cultures on collagen gels subsequently increase their biosynthesis of Ch-4-S to values approaching the initial levels, these values are far short of those achieved in the cultures on bone matrix.

DEVELOPMENTAL BIOLOGY

344

VOLUME 78,198O

BONE MATRIX 80

COLLAGEN GEL

UNTREATED

UNTREATED

A.

,/

60

_m-4-s 1.

8%

--

+Ch-S

tCh-S

60-

DAYS IN CULTURE FIG. 5. Incorporation of labeled sulfate into Ch-4-S, ChS-S, and DS by minced skeletal muscle explanted onto bone matrix or onto collagen gels. Cultures either were untreated or were treated with Ch-S or HA. An aliquot of the high molecular weight sulfated material shown in Fig. 4 was degraded into its constituent disaccharides by chondroitinases ABC and AC-II; results from single representative cultures are shown. The data are corrected to refer to the same amount of sample and are plotted as the percentage of total sulfate incorporation.

Biosynthesis of Chondroitin Sulfates Relatiue to Other Sulfated Species Both authentic embryonic cartilage and embryonic skeletal muscle incorporate es-

sentially all of the sulfate label into Ch-S and DS (Table 2), but when skeletal muscle was cultured in vitro on bone matrix or on collagen gels, not all of the label found in high molecular weight sulfated material

NATHANSON

In Vitro Chondrogenesis

AND HAY

could be accounted for as Ch-S and DS (Table 4). In each case, the origin of the chromatograms contained labeled material which was undegraded by the chondroitinases (data not shown). This residual material represents the sulfated glycosaminoglycans which are not related to Ch-S (keratan sulfate, heparan sulfate, or heparin). Even though explants onto bone matrix do not reach the level of Ch-S synthesis of embryonic cartilage in situ (95% of sulfated GAG), they approach this level from day 3 onward in untreated and Ch-S treated cultures (Table 4). The values for HA-treated cultures on bone matrix are somewhat lower at l-2 days in vitro, but are in the same range thereafter. At 10 days, all the explants onto bone matrix synthesize mainly Ch-S, with only 10-15s of the label entering other sulfated species. Control explants onto collagen gels, in contrast, synthesize less Ch-S relative to other sulfated species and do so at decreasing levels by the termination of the experiment. Our morphological data show that skeletal muscle was regenerating by 4 days in vitro. The cultures on gels, however, synthesize a lower proportion of Ch-S to total sulfated GAG (Table 4) than embryonic skeletal muscle (Table 2). HA-treated cultures show more variability in the early period in vitro, but at 10 days they stabilize at levels similar to those of the Ch-S and untreated cultures (-60% Ch-S). The results just presented are consistent with our PERCENTAGE

OF TOTAL

Days in culture

Explants Untreated

1 2 3 4 10

SULFATE

85.0 75.5 SO.6 83.0 89.0

finding that bone matrix supports the synthesis of Ch-S isomers and that control cultures on collagen gels cannot sustain ChS synthesis, but rather produce a different pattern of sulfated GAG.

Alteration of Chondroitim4Sulfate in Vivo

to bone matrix

80.3 75.2 92.2 82.1 86.2

SULFATES

Explants Hyaluronic acid treated 67.8 73.7 89.7 85.1 83.4

Levels

In 24-l-n cultures of minced muscle, Ch4-S represents 50% of the chondroitin sulfate isomers being synthesized, even though chondrogenic cell types have not yet appeared. Furthermore, Ch-4-S is produced in larger amounts in cultures regenerating muscle (25-50s) than in the muscle of origin (18%, Table 3). These data suggest that the experimental conditions might be sufficient to step up the production of Ch-4-S. To test the possibility that the trauma of mincing the muscle plays a role, operations were performed in which the gastrocnemius muscle of newborn rats was excised under ether anesthesia, then minced in isotonic saline, and the mince was reimplanted in situ (Carlson, 1968). In each case the unminced, contralateral gastrocnemius served as a control. Immediately following the operation each rat was given a single intraperitoneal injection of 100 $!i of Na235S04 and left undisturbed with its mother for a period of 3 days. Total high molecular weight sulfated material of each leg was isolated, measured, and separated into ChS and DS. The operated and control gastrocnemius

TABLE 4 INCORPORATION DUE TO CHONDROITIN SULFATE”

Chondroitm sulfate treated

345

Untreated 77.0 67.5 79.3 70.3 60.1

AND DERMATAN

to collagen

Chondroitin sulfate treated 86.8 64.5 70.3 61.3 62.1

gels Hyaluronic acid treated 55.0 74.4 100.0 58.5 62.1

” The data represent the sum of the percentage distribution into chondroitin sulfate isomers shown in Fig. 6. The l-day value for untreated cultures on bone matrix and 3-day value for hyaluronic acid-treated cultures on collagen gels are inconsistent and probably represent individual variations within this experiment.

346

DEVELOPMENTAL

BIOLOGY TABLE

Muscle

INCORPORATION OF SULFATE BY REIMPLANTED cpm/ld nuclei % CH-6-S %DS

VOLUME

78,1980

5 MINCES OF RAT GASTROCNEMIUS I CH-4-S CH-CS/CH-6-S

MUSCLE” % Tota;c;~n/lO”

Operated Control

834 1116

11.1 29.2

17.5 27.2

36.7 11.0

3.31 0.36

65.3 68.0

Operated Control

1707 1624

10.4 20.4

24.6 27.4

26.4 9.1

2.73 0.45

63.4 56.9

n The experiment consisted of a duplicate series of samples and all values are shown. High molecular weight sulfated material was labeled for 24 hr prior to isolation on day 3 after operation. The data indicate a shift toward increased synthesis of Ch-4-S, but insofar as they represent a day 3 sample in viva they cannot be interpreted relative to the data shown in Table 3 or Fig. 5. Abbreviations: CH-6-S, chondroitin-6-sulfate; DS, dermatan sulfate; Ch-4-S, chondroitin 4-sulfate.

incorporate similar total amounts of Na235S04 (Table 5). However, in each case the operated gastrocnemius increased its production of Ch-4-S relative to Ch-6-S and DS (Table 5). The effect of the operation therefore was to cause a shift in the Ch-4S/Ch-6-S ratio from an average of 0.42 (contralateral side) to one of 3.02 (operated side). These data suggest that the operative trauma alone contributes to the alteration in levels of the chondroitin sulfate isomers observed at the first culture point.

I), explanted myoblasts and fibroblasts transform to a population of fibroblast-like cells. This phase, which can be observed in explants grown on either bone matrix or collagen gels, begins as early as 24 hr in uitro and lasts through 2 days. The biochemical phase I (l-3 days in vitro) is characterized by (1) a 50% increase in total sulfate uptake as compared with that of the same muscle in situ and (2) a shift in the relative abundance of the chondroitin sulfate isomers Ch-4-S and Ch-6-S. Insofar as this phase (I) is one in which myoblasts DISCUSSION begin to assume the morphology of the The tranformation of embryonic rat skel- fibroblast-like cells, the increase in sulfate etal muscle to cartilage on bone matrix in uptake most likely reflects the transformavitro has been shown previously to involve tion of myoblasts to secretory cells. a three phase alteration in morphological Whereas the starting material synthesized chondroitin sulfate isomers at a ratio of Chphenotype, culminating in the differentiation of chondrocytes from the mononucle- 4-S to Ch-6-S of 0.3, after 24 hr in vitro the ate cells released from the minced muscle explants onto bone matrix synthesized (Table 1). The results presented in this these isomers at a ratio of 2.6-3.8 and those communication demonstrate that the bio- on collagen gels at a ratio of 1.5-2.6. These chemical differentiation of these chondro- ratios represent an actual increase in the cytes also involves several phases. In the synthesis of Ch-4-S and a concurrent deDiscussion, we shall first examine the cor- crease in the synthesis of Ch-6-S, rather relations between the phases of morpholog- than a simple decrease in the synthesis of ical differentiation and the phases of bio- Ch-6-S, because total Ch-S synthesis inchemical differentiation. We also examined creases at the same time. These values difsulfated GAG synthesis in cultures on bone fer from those obtained by other investimatrix and collagen gels in the presence of gators for authentic fibroblasts grown for exogenous GAG; we will consider the sig- periods of time in vitro (Ch-4-S/Ch-6nificance of these results in the latter part S-0.6-1.1, calculated from Conrad et al., 1977). We believe that phase I can be likof the Discussion. In the first morphological phase (phase ened to a regenerative response on the part

NATHANSON

AND HAY

of the minced muscle (Nathanson and Hay, 1980). To explore the possibility that the changes observed in phase I reflect a regenerative response to injury, we analyzed the relative biosynthesis of chondroitin sulfate isomers in excised and minced skeletal muscle reinserted into the limb. Enhanced biosynthesis of Ch-4-S occurs in uiuo which is similar to the biosynthetic alteration seen in 24-hr cultures of skeletal muscle on bone matrix and collagen gels. It appears, thus, that this early increase in Ch-4-S biosynthesis is not due to an effect of either the bone matrix, the collagen gel, or the in vitro environment, but is due to the operational trauma. It is unclear why such trauma should alter the Ch-4-S/Ch-6-S ratio, but not affect the total sulfate incorporation in vivo. The increased biosynthesis of Ch-4-S presumably reflects a regnerative response which is related to glycosaminoglycan production in a complex fashion. Explants onto collagen gels do not maintain the initial levels of Ch-4-S when they begin to redifferentiate into skeletal muscle. Since explants onto bone matrix sustain these levels (phase II) and form cartilage (phase III), the effect in phase II is one of sustaining a biosynthetic shift already in progress. Phase II can be defined morphologically as a period (4-5 days) during which the explanted cells on bone matrix remain fibroblast-like and acquire increased amounts of granular endoplasmic reticulum, composed of dilated cisternae containing electron-dense material (Nathanson and Hay, 1980). This morphology is often correlated with increased secretory activity, and in this regard it is interesting to note that the explants on bone matrix at 4 days increase their incorporation of sulfate to five times greater than that observed in cultures on collagen gels. In spite of this increased sulfate uptake, the explanted cells do not markedly alter the relative amount of Ch-S in total GAG produced at 4 days. Other investigators report that the

In Vitro

Chondrogenesis

347

levels of Ch-S synthesis can be independent of the level of sulfate uptake (Schwartz and Dorfman, 1975). Phase II can be considered an interim period, during which bone matrix enhances the regenerative response and the cells prepare to become chondrocytes. The formation of chondrocytes occurs during phase III (6-10 days) and is accompanied by a large increase in (1) the total sulfate uptake and (2) the proportion of Ch4-S relative to Ch-6-S. Both the shift in Ch4-S/Ch-6-S ratio and the increase in total sulfate incorporation are correlated with the increase in the chondrocyte population and probably will be found to be accompanied by changes in proteoglycan composition as well. We found the ratio of Ch-4-S/ Ch-6-S in embryonic rat cartilage in situ to be 13.1, whereas the ratio in lo-day-old bone matrix cultures is 16.3. Moreover, the chondrogenic cultures even at 10 days produced significantly more non-chondroitin sulfate GAG than cartilage in situ. As noted above, the cultures on collagen gels show an initial “regenerative” response (Phase I) with an increase in Ch-4-S and a decrease in Ch-6-S synthesis. However, in contrast with that of the cultures on bone matrix, this increase is short-lived. As these cultures begin to regenerate skeletal muscle the Ch-4-S/Ch-6-S ratios fall to levels more characteristic of cultured fibroblasts (Conrad et al., 1977). It is noteworthy that these ratios are somewhat higher than for embryonic rat skeletal muscle in situ and are found to rise slightly by the termination of the experiment. But even at the greatest level, the synthesis of chondroitin sulfate isomers does not surpass the earliest values obtained for cultures on bone matrix. Whereas the appearance of chondrocytic morphology and greatly enhanced sulfate incorporation appear to be solely dependent upon the presence of bone matrix, the observed shift in ratios of Ch-S isomers is not. Both cultures on collagen gels and those on bone matrix synthesize increased amounts of Ch-4-S (and decreased amounts

348

DEVELOPMENTALBIOLOGY

of Ch-6-S). The early shift to Ch4-S biosynthesis differs from certain other systems in which precartilaginous tissue synthesizes predominantly Ch-6-S (Robinson and Dorfman, 1969; Kosher et al., 1973; Marzullo and Lash, 1967). Similarly, as noted in the Introduction, not all cartilage contains high levels of Ch-4-S, suggesting that the Ch-S isomers may be multifunctional. We have shown that Ch4-S appears rapidly and is synthesized at the expense of Ch-6-S in our cultures, which gives further support to the idea that these isomers are multifunctional. It is quite surprising that Ch-4-S is so easy to induce, for skeletal muscle seems to exhibit low levels of 4’-sulfotranseferase, as evidenced by its low level of Ch-4-S biosynthesis. The increased levels of Ch-4-S could be accomplished via the induction of de novo sulfotransferase activity or by the unmasking of an inhibited sulfotransferase. While the actual mechanism awaits further investigation, it is sufficient for the present discussion to point out that skeletal muscle cells and fibroblasts have a biosynthetic potential similar to that of chondrogenic cells and this potential is not necessarily dependent upon the presence of bone matrix. These observations support our working hypothesis that skeletal muscle and cartilage do not comprise different embryonic lineages, but represent variations in the differentiation of the same mesenchymal cell population. The variation in the DNA content of the cultures deserves further comment. The cultures on bone matrix and collagen gels were set up with approximately the same initial tissue mass, and from the first to the tenth days, they all declined in DNA content (from an equivalent of 4 X 10” cells on day 1 to 1 x lo6 on day 10). Since the minced muscle contains myotubes that degenerate in response to the injury, a certain amount of this loss in DNA is due to the disappearance of pycnotic myotube nuclei (Nathanson and Hay, 1980). Such a loss in inert DNA after day 3 could partly account

VOLUME X(1980

for the enhanced sulfate incorporation on a per-DNA basis observed in the cultures on bone matrix at 4 days, but is insufficient to explain the large increase at 10 days. Urist et al. (1978) report uptake of r3H]thymidine in muscle explants on bone matrix and there very well may be an initial proliferation of the mononucleate cells. We have not a yet investigated this point. The results reported here demonstrate that exogenous Ch-S and HA do not enhance or diminish sulfate incorporation into high molecular weight sulfated material, or the proportion of this material in Ch-S isomers in cultures on bone matrix or collagen gels. On the contrary, exogenous Ch-S at the same concentration we used has been reported to stimulate sulfate incorporation into Ch-S in chondrocytes grown in suspension (Nevo and Dorfman, 1972) and monolayer culture (Huang, 1974; Schwartz and Dorfman, 1975). Similarly, isolated cornea1 epithelium responds to exogenous Ch-S in vitro by an increased synthesis of sulfated GAG (Meier and Hay, 1974). These and other observations (see Introduction) have led to a widely accepted hypothesis that Ch-S has the ability to enhance its own synthesis, whereas hyaluronic acid seems to have an inhibitory effect on differentiation. The conflicting results discussed above may have several explanations. Enzymes were used to isolate cells in the studies of somites, cornea, and chondrocytes mentioned above, but not in the present study. Perhaps it is necessary to remove cell-associated GAG in order to demonstrate a stimulatory effect of exogenous Ch-S on cultured cells. Ch-S and other GAG have been found associated with cell surfaces (Kramer, 1971; Kojima and Yamagata, 1971; Lippman, 1968; Kelley et al., 1977) and embryonic basement membranes (Cohn et al., 1977; Hay and Meier, 1974). It may be the restoration of removed extracellular coats that “enhances” cell differentiation in such in vitro systems, and the

NATHANSON

AND HAY

“enhanced” level might even be equivalent to the normal in uiuo level (see Hay, 1977). In our organ cultures, the differentiating cells are surrounded by connective tissue that is not disrupted at the time exogenous GAG is added and enzymes were not used for the initial isolation. There could be other reasons for the differences in reported effects of GAG on cell differentiation. Exogenous HA seems not to affect chondrogenesis when used in conjunction with some batches of fetal calf serum (Solursh et al., 1974; seealso Solursh and Meier, 1973). Serum contains variable amounts of substances like thyroid hormone (Audhya and Gibson, 1976), which may antagonize HA-induced inhibition of chondrogenesis (Toole, 1973b). The synthesis of chondroitin sulfate appears to be modulated in vitro by conditioned medium (Solursh and Meier, 1973a), vitamin A (Solursh and Meier, 1973b), retinoic acid (Shapiro and Poon, 1976), and nicotinamide (Overman et al. 1972). The effect of exogenous glycosaminoglycans on tissues in situ and in combination with various culture media remains relatively unexplored, but the studies cited here point out the complexity of the in viva and in vitro environment and the care with which such investigations should be interpreted. The authors wish to thank Mr. Wayne B. Colin for assistance with the chromatographic and biochemical procedures, Dr. J. E. Silbert and the members of his laboratory for many profitable discussions, and Dr. J. A. Cifonelli for his generous gift of purified glycosaminoglycan standards produced under Contract AM52205 from the National Institutes of Health. This research was supported by United States Public Health Service Fellowship F32-AM-05481 to M.A.N. and N.I.H. Grant HD-00143 to E.D.H. REFERENCES ANDERSON, H. C., and GRINER, S. A. (1977). Cartilage induction in vitro. Ultrastructural studies. Deuelop. Biol. 60,351-358. AUDHYA, T. K., and GIBSON, K. D. (1976). Effects of medium composition and metabolic inhibitors on glycosaminoglycan synthesis in chick embryo cartilage and its stimulation by serum and triiodothyro-

In Vitro

349

Chondrogenesis

nine. Biochim. BITTER, T., and acid carbazole CARLSON, B. M. pletely excised rat from minced

Biophys. Acta 437,364-376. MUIR, H. (1972). A modified uranic reaction. Anal. Biochem. 4.330-334. (1968). Regeneration of the comgastrocnemius muscle in the frog and muscle fragments. J. Morphol. 125,

447-472. COHN, R. H., BANERJEE, S. D., and BERNFIELD, M. R. (1977). Basal lamina of embryonic salivary epithelia. Nature of glycosaminoglycan and organization of extracellular materials. J. Cell Biol. 73,464-

47a. CONRAD, G. W., HAMILTON, C., and HAYNES, E. (1977). Differences in glycosaminoglycans synthesized by fibroblast-like cells from chick cornea, heart and skin. J. Biol. Chem. 252,6861-6870. DIETRICH, C. P., SAMPAIO, L. O., and TOLEDO, 0. M. S. (1976). Characteristic distribution of sulfated mucopolysaccharides in different tissues and in their respective mitochondria. Biochem. Biophys. Res. Commun. 71, l-10. EBERT, P. S., and PROCKOP, D. J. (1967). Influence of cortisol on the synthesis of sulfated mucopolysaccharides and collagen in chick embryos. Biochim. Biophys. Acta 135,45-55. DE LA HABA, G., and HOLTZER, H. (1965). Chondroitin sulfate: Inhibition of synthesis by puromycin. Science 149, 1263-1265. HAY, E. D. (1977). Cell-matrix interaction in embryonic induction. In “International Cell Biology 19761977” (B. R. Brinkley and K. R. Porter, eds.), pp. 50-57. Rockefeller Press, New York. HAY, E. D. and MEIER, S. (1974). Glycosaminoglycan synthesis by embryonic inductors: Neural tube, notochord, and lens. J. Cell Biol. 62, 889-898. HINEGARDNER, R. T. (1971). An improved fluorometric assay for DNA. Anal. Biochem. 39,197-201. HUANG, P. (1974). Effect of extracellular chondroitin sulfate on cultured chondrocytes. J. Cell Biol. 62,

861-886. KELLEY, R. O., PALMER, G. C., CRISSMAN, H. A., and NILSON, J. H. (1977). Interaction of glycosaminoglycan and adenylate cyclase at the surface of cultured human diploid fibroblasts (HLM 15). J. Cell Sci. 28,

237-250. KOJIMA, K., and YAMAGATA, T. (1971). Glycosaminoglycans and electrokinetic behavior of rat ascites hepatoma cells. Exp. Cell Res. 67, 142-146. KOSHER, R. A., and LASH, J. W. (1975). Notochordal stimulation of in vitro somite chondrogenesis before and after enzymatic removal of perinotochordal materials. Develop. Biol. 42, 362-378. KOSHER, R. A., LASH, J. W., and MINOR, R. R. (1973). Environmental enhancement of in vitro chondrogenesis. IV. Stimulation of somite chondrogenesis by exogenous chondromucoprotein. Develop. Biol.

35,210-220. KRAMER,

P. M.

(1971).

Heparan

sulfates

of cultured

350

DEVELOPMENTAL

BIOI

cells. I. Membrane associated and cell-sap species in Chinese hamster cells. Biochemistry 10, 1437-1445. LASH, J. W., and VASAN, N. S. (1977). Tissue interactions and extracellular matrix components. Sot. Gen. Physiol. Ser. 32, 101-113. LIPPMAN, M. (1968). Glycosaminoglycans and cell division. In “Epithelial-Mesenchymal Interactions” (R. Fleischmajer and R. E. Billingham, eds.), pp. 208-229. Williams and Wilkins, Baltimore. MARZULLO, G., and LASH, J. W. (1967). Acquisition of the chondrogenic phenotype. In “Morphological and Biochemical Analysis of Cytodifferentiation” (E. Hagen, W. Wechsler, and P. Zilliken, eds.), Experimental Biology and Medicine, Vol. 1, pp. 213218. Karger, Basel. MATHEWS, M. B. (1975). Polyanionic glycans in development and aging of vertebrate cartilage. In “Connective Tissue: Macromolecular Structure and Evolution,” Molecular Biology and Biophysics, Vol. 19, Chap. 8, pp. 156-171. Springer Verlag, New York. MEIER, S., and HAY, E. D. (1974). Stimulation of extracellular matrix synthesis in the developing cornea by glycosaminoglycans. Proc. Nat. Acad. Sci. USA 71,2310-2313. NATHANSON, M. A., and HAY, E. D. (1980). Analysis of cartilage differentiation from skeletal muscle grown on bone matrix. I. Ultrastructural aspects.

Develop. Biol. 78,301-331. NATHANSON, M. A., HILFER, S. R., and SEARLS, R. L. (1978). Formation of cartilage by non-chondrogenic cell types. Develop. Biol. 64,99-117. NEVO, Z., and DORFMAN, A. (1972). Stimulation of chondromucoprotein synthesis in chondrocytes by extracellular chondromucoprotein. Proc. Nat. Acad.

Sci. USA 69,2069-2072. NOCAMI, H., and URIST, M. R. (19701. A substratum of bone matrix for differentiation of mesenchymal cells into chondro-osseous tissues in vitro. Exp. Cell

Res. 63.404-410. NOGAMI, H., and URIST, M. R. (1974a). Explants, transplants, and implants of a cartilage and bone morphogenetic matrix. Clin. Orthopaed. 103, 235251. NOGAMI, H., and URIST, M. R. (1974b). Substrata prepared from bone matrix for chondrogenesis in tissue culture. J. Cell Biol. 62,510-519. OGSTON, A. G. (1970). The biological functions of the glycosaminoglycans. In “Chemistry and Molecular Biology of the Intercellular Matrix” (E. A. Balazs, ed.), pp. 1231-1240. Academic Press, New York. OVERMAN, D. O., SEEGMILLER, R. E., and RUNNER, M. N. (1972). Coenzyme competition and precursor specificity during teratogenesis induced by6-aminonicotinamide. Develop. Biol. 28,573-582. ROBINSON, H. C. and DORFMAN, A. (1969). The sulfation of chondroitin sulfate in embryonic chick cartilage epiphyses. J. Biol. Chem. 244,348-352. SAITO, H., YAMAGATA, T., and SUZUKI, S. (1968).

.OGY

VOLUME

78,1986

Enzymatic quantities

methods for the determination of small of isomeric chondroitin sulfates. J. Biol. Chem. 243, 1536-1542. SANTOIANNI, P., and AYALA, M. (1965). Fluorometric ultramicroanalysis of deoxyribonucleic acid in human skin. J. Invest. Dermatol. 45,99-103. SCHWARTZ, N. B., and DORFMAN, A. (1975). Stimulation of chondroitin sulfate proteoglycan production by chondrocytes in monolayer. Conn. Tissue Res. 3, 115-122. SEARLS, R. L. (1965). An autoradiographic study of the uptake of S-35 sulfate during the differentiation of limb bud cartilage. Develop. Biol. 11, 155-168. SHAPIRO, S. S., and POON, J. P. (1976). Effect of retinoic acid on chondrocyte glycosaminoglycan biosynthesis. Arch. Biochem. Biophys. 174, 74-81. SHULMAN, H. J., and MEYER, K. (1968). Cellular differentiation and the aging process in cartilaginous tissues. J. Exp. Med. 128, 1353-1362. SMITH, G. N., JR., TOOLE, B. P., and GROSS, J. (1975). Hyaluronidase activity and glycosaminoglycan synthesis in the amputated newt limb: Comparison of denervated, nonregenerating limbs with regenerates. Develop. Biol. 43, 221-232. SOLURSH, M., and MEIER, S. (1973a). A conditioned medium (CM) factor produced by chondrocytes that promotes their own differentiation. Develop. Biol. 30,279-289. SOLURSH, M., and MEIER, S. (1973b). The selective inhibition of mucopolysaccharide synthesis by vitamin A treatment of cultured chick embryo chondrocytes. Calcif Tissue Res. 13, 131-142. SOLURSH, M., and MORRISS, G. M. (1977). Glycosaminoglycan synthesis in rat embryos during the formation of the primary mesenchyme and neural folds. Develop. Biol. 57, 75-86. SOLURSH, M., VAEREWYCK, S., and REITER, R. R. (1974). Depression by hyaluronic acid of glycosaminoglycan synthesis by cultured chick embryo chondrocytes. Develop. Biol. 41,233-244. TOOLE, B. P. (1972). Hyaluronate turnover during chondrogenesis in the developing chick limb and axial skeleton. Develop. Biol. 29,321-329. TOOLE, B. P. (1973a). Hyaluronate and hyaluronidase in morphogenesis and differentiation. Amer. Zool. 13, 1061-1065. TOOLE, B. P. (1973b). Hyaluronate inhibition of chondrogenesis: Antagonism of thyroxine, growth hormone, and calcitonin. Science 180,302-303. TOOLE, B. P., and GROSS, J. (1971). The extracellular matrix of the regenerating newt limb: Synthesis and removal of hyalmonate prior to differentiation. De-

velop. Biol. 26,57-77. TOOLE, B. P., and TRELSTAD, R. L. (1971). Hyaluronate production and removal during corneal development in the chick. Develop. Biol. 26,28-35. TOOLE, B. P., JACKSON, G., and GROSS, J. (1972). Hyaluronate in morphogenesis: Inhibition of chon-

NATHANSON

AND

HAY

drogenesis in vitro. Proc. Nat. Acad. Sci. USA 69, 1384-1386. URIST, M. R. (1971). The substratum for bone morphogenesis. In “Changing Syntheses in Development” (M. N. Runner, ed.), 29th Symposium of the Society for Developmental Biology, pp. 125-163. Academic Press, New York. URIST, M. R., HAY, P. H., DUBUC, F., and BURING, K. (1969). Osteogenic competence. Clin. Orthopaed.

In Vitro Chondrogenesis

351

64, 194-220. URIST, M. R.. TERASHIMA, Y., NAKACAWA, M., and STAMOS, C. (1978). Cartilage tissue differentiation from mesenchymal cells derived from mature muscle in tissue culture. In Vitro 14, 697-706. YAMAGATA, T., SAITO, H., HABUCHI, O., and SUZUKI, S. (1968). Purification and properties of bacterial chondroitinaaes and chondrosulfataaes. J. Biol. Chem. 243, 1523-1535.