Bone, 12, 7-1.5 (1991) Printed in the USA. All rights reserved.

Copyright

8756-3282191 $3.00 + .OO 0 1991 Pergamon Press plc

a,-HS-glycoprotein: Expression in Chondrocytes and Augmentation of Alkaline Phosphatase and Phospholipase A, Activity F. YANG’,

Z. SCHWARTZ2*3,

L. D. SWAIN*,

C.-C. LEE’, B. H. BOWMAN’

and B. D. BOYAN*

‘Deparhnent ofCellular and Structural Biology ‘Deparhnent of Orthopaedics, The University of Texas Health Science Center at San Antonio 3Department of Periodontics Hebrew Universio Hadassah Faculty of Dental Medicine, Jerusalem, Israel Address for correspondence and reprints: Funmei Yang, Ph.D., Department Science Center, 7703 Floyd Curl Drive, San Antonio, TX 78284, USA.

of Cellular and Structural

Abstract

nous proteins in the organic bone matrix. It has been suggested that a,-HS-glycoprotein plays important roles in two biological functions, bone mineralization (Triffitt et al. 1976) and the immune response (Bradley et al. 1977; van Oss et al. 1974). Synthesis of cY,-HS-glycoprotein changes dynamically in various physiological conditions. It is one of the few negative acute phase reactants; the plasma level of this protein significantly decreases in certain malignancies (Bradley et al. 1977) and inflammatory diseases (Lebreton et al. 1979). Malnutrition (Schelp et al. 1980) and Paget’s disease of bone also cause decreased concentrations of a,-HS-glycoprotein in serum (Ashton & Smith 1980), while in some patients with osteogenesis imperfecta, abnormally high concentrations of a,-HS-glycoprotein have been observed (Dickson et al. 1983). Human o.,-HS-glycoprotein has a molecular weight of 5 1,000 and circulates in the plasma as two polypeptide chains, A and B , held together covalently through disulfide bonding (Gejyo et al. 1983). The A chain consists of 282 amino acids (Yoshioka et al. 1986), while the B chain has only 27 amino acids (Gejyo et al. 1983). Recently our laboratory cloned, characterized, and chromosomally mapped the human AHSG gene (Lee at al. 1987). Characterization of the AHSG cDNA revealed a single mRNA sequence that encodes both A and B polypeptide chains. Of additional interest was the discovery that an intervening sequence between the A and B chains encodes a 40 amino acid polypeptide not seen in the mature a,-HS-glycoprotein protein. The function and fate of this short peptide are presently unknown. It was suggested, because of its high affinity for barium (Schmid & Burgi 1961) and calcium ions (Triffitt et al. 1976), that a,-HS-glycoprotein influences the mineral phase of bone. Evidence has been reported that a,-HS-glycoprotein is involved in recruitment of monocytes or osteoclast precursors to sites of bone resorption (Malone et al. 1982) and may, therefore, participate in bone remodeling. The concentration of a,-HSglycoprotein in neonatal bone is three times higher than in bone from children and seven times higher than in adult bone (Quelch et al. 1984). Accordingly, high concentrations of a,-HS-glycoprotein are found only in immature or newly formed bone (Dickson & Bagga 1985), implying a correlation between the level of a,-HS-glycoprotein and bone development and maturation.

The a,-HS-glycoprotein is a plasma protein synthesized in liver and enriched in bone. The concentration of a,-HSglycoprotein dynamically changes in various physiological conditions and is highest in bone during growth, suggesting that it is involved in regulation of endochondral ossification. Northern blot analysis demonstrated that mRNA transcripts from growth zone and resting zone costochondral chondrocyte cultures hybridized with cl,-HS-glycoprotein cDNA. However, a difference of mRNA transcript size was observed, with chondrocyte mRNA transcripts being 2.2 kb, while mRNA isolated from liver was 1.6 kb. Presence of a,-HS-glycoprotein in cartilage cells was found by immunohistochemical staining of human fetal epiphyses using antihuman a,-HS-glycoprotein antibody. To understand the role of cY,-HS-glycoprotein in cartilage growth, the effects of exogenous cl,-HS-glycoprotein were correlated with alkaline phosphatase (ALPase) and phospholipase A, (PA,) activity in the chondrocyte cultures. Alkaline phosphatase specific activity was stimulated by a,-HS-glycoprotein at concentrations between 0.25 and 1.25 pg/mL in the growth zone and resting zone cultures 2.7 and 2.0-fold, respectively. Matrix vesicle PA, activity was increased only in the growth zone chondrocyte cultures. These results suggested that a,-HSgiycoprotein may contribute to the regulation of the expression of the chondrocyte phenotype. Steady state mRNA levels of ALPase were analyzed in chondrocytes after additions of a,-HS-glycoprotein. The ALPase mRNA levels remained stationary during the stimulation of enzymatic activity, indicating that the effect of a,-HS-glycoprotein upon alkaline phosphatase activity is not at the transcriptional level. Key Words: a,-HS-glycoprotein

mRNA-Chondrocyte--in vitro-Alkalinephosphatase-Immunohistochemistry-CartilagePhospholipase A,

Introduction Human cw,-HS-glycoprotein (AHSG) is a plasma protein synthesized in liver (Schultze et al. 1962). It is selectively concentrated up to 300-fold in bone and is one of the major noncollage-

Biology,

The University

of Texas Health

8

The experiments reported in this paper were designed to investigate the functional role of cx,-HS-glycoprotein in the process of endochondral ossification involved in bone growth and development. In order to examine this, we utilized the chondrocyte culture system using cells derived from the growth zone and resting zone of cartilage in which specific enzyme activities can be correlated with the stage of cell differentiation (Boyan et al. 1988a; Boyan et al. 1988b; Schwartz et al. 1988a; Schwartz & Boyan 1988; Schwartz et al. 1989). Two of these enzymes, alkaline phosphatase and phospholipase A,, are enriched in matrix vesicles produced by these chondrocytes in culture and are differentially regulated depending on the state of chondrocyte maturation (Schwartz & Boyan 1988; Schwartz et al. 1989). In the present study, we examined the effect of a,-HS-glycoprotein on chondrocyte differentiation as determined by the mRNA level of alkaline phosphatase and by its specific activity in the matrix vesicles. In addition, we examined the effect of cw,-HS-glycoprotein on the specific activity of phospholipase A,. The capacity of chondrocytes to produce a,-HS-glycoprotein was studied using a human (r,HS cDNA probe; localization of cY,-HS-glycoprotein in human fetal cartilage was studied by immunoperoxidase staining. The results suggest a role for a,-HS-glycoprotein in the regulation of chondrocyte phenotype.

F. Yang et al.: a,-HS-glycoprotein

amplified without loss of phenotype, number of animals necessary.

and chondrocytes

thereby

decreasing

the

Preparation of cell fractions Matrix vesicles and plasma membranes were prepared as described previously (Boyan et al. 1988a). At harvest, cultures were trypsinized (1% in HBSS), the reaction was terminated with DMEM containing 10% FBS. Cells were collected by centrifugation at 500 X g for five minutes, resuspended in 0.9% NaCl, washed twice, and counted. The trypsin digest supematant was centrifuged for ten minutes at 13,000 X g to produce a mitochondria/membrane pellet, and the resultant supematant centrifuged for one hour at 100,000 X g to obtain a pellet of matrix vesicles. The chondrocytes were homogenized in a Ten Broeck homogenizer (Fisher Scientific, Pittsburgh, PA), and the plasma membranes were isolated by differential centrifugation, followed by sucrose density centrifugation. Plasma membranes and matrix vesicles were resuspended in 500 FL phosphate buffered saline (PBS). All samples used in subsequent assays represent the combination of five cultures (i.e., five T-75 flasks). The protein content of each fraction was determined (Helenius and Simons 1971). Preparation of cell layer

Materials and Methods Chondrocyre culrures The chondrocyte culture system has been described in detail by Boyan et al. (1988a). Rib cages were removed from thirty 125-g Sprague-Dawley rats per experiment and placed in Dulbecco’s modified Eagle’s medium (DMEM). The resting zone and adjacent growth zone of the costochondral junction were separated, sliced, and incubated overnight in DMEM with 5% CO, in air at 37°C. DMEM was replaced by two 20-minute washes in Hank’s balanced salt solution (HBSS), followed by sequential incubations in 1% trypsin (Gibco, Grand Island, NY) for one hour and 0.02% collagenase (Worthington type II, Freehold, NJ) for three hours. All enzymes were prepared in HBSS. After enzymatic digestion, cells were separated from debris by filtration (40-mesh nylon), collected by centrifugation at 500 X g for five minutes, resuspended in DMEM, and plated at a density of 10,000 cells/cm* for resting zone cells or 25,000 cells/cm’ for growth zone cells. Cultures were incubated in DMEM containing 10% fetal bovine serum (FBS) and 50 p,g/mL vitamin C in an atmosphere of 5% CO, in air at 37°C and 100% humidity. Medium was changed after 24 hours and then again at 72-hour intervals. At confluence (7 to 10 days), cells were subcultured and allowed to return to confluence using the plating densities and techniques described above. Fourth passage cells were used for the following reasons. These cells and their isolated plasma membranes and matrix vesicles show similar phospholipase A, and alkaline phosphatase activities to that of the primary cultures (Boyan et al. 1988b). The fourth passage cells produce type II collagen and cartilage specific proteoglycan (Schwartz et al. 1989). The matrix produced by the cells in culture is similar in appearance at the transmission electron microscopic level to cartilage matrix in vivo (Swain et al. unpublished data). In addition, the cells retain their differential responsiveness to vitamin D metabolites (Boyan et al. 1988a; Schwartz et al. 1988a; Schwartz et al. 1989). Moreover, when the fourth passage cells were implanted intramuscularly in vivo, they formed cartilage nodules. As a result, the number of cells available for the study could be

Cell layers were prepared following the method of Hale et al. (1986). Cells were cultured in 24-well culture dishes (Coming). At harvest, the medium was decanted and the cell layer was removed using a cell scraper and suspended in PBS. After centrifugation, the cell layer pellet was washed with PBS and resuspended by vortexing in 500 PL of deionized water and 25 pL of 1% Triton X-100. Protein content was determined using the micro BCA technique (Smith et al. 1985). Enzvme assays Alkaline phosphatase (ortho pyrophosphohydrolase alkaline: EC 3.1.3.1) was measured in the cell layer and in isolated matrix vesicles and plasma membranes as a function of release of para-nitrophenol from para-nitrophenylphosphate at pH 10.2 (Bretaudiere and Spillman 1984). Data are expressed as kmoles Pi/mg protein/minute to facilitate comparison with previously published literature (Boyan et al. 1988a; Schwartz et al. 1988a), as 1 mole of inorganic phosphate is released for each mole of para-nitrophenol. Phospholipase A2 (EC 3.1.1.4) was measured in isolated plasma membranes and matrix vesicles as described previously (Schwartz and Boyan 1988). Activity was measured as the hydrolysis of 14C-arachidonate from phosphatidylethanolamineL-2-R-palmitoyl-2, arachidonyl-R’4C (New England Nuclear, Boston, MA). Each sample was analyzed in triplicate. Data presented are from a typical experiment. For the isolated matrix vesicles and plasma membranes, each data point is the mean ? SEM for six samples, each sample representing the combined membrane fractions of five T-75 flasks. For the cell layers, each data point is the mean ? SEM for six individual cultures. Statistical significance between data points and controls was determined by Student’s t-test, using P < 0.05 confidence limits. Incubation with a,-HS-glycoprotein The a,-HS-glycoprotein (Calbiochem) was solubilized in DMEM immediately before use and added to fourth passage confluent

F. Yang et al.: a,-HS-glycoprotein

and chondrocytes

9

growth zone and resting zone chondrocyte cultures that had been maintained in serum-free DMEM for 24 hours. Cells were incubated with the DMEM containing a,-HS-glycoprotein for 24 hours. Cells cultured in DMEM plus 10% FBS were used as a positive control; negative controls were cells in serum-free medium with no additives.

123466789lOll

Isolation and Northern blot analysis of RNA from chondrocytes RNA was extracted by lysing cells in guanidinium thiocyanate followed by phenol-chloroform extraction (Chomczynski and Sacchi 1987). Total cellular RNA was isolated from fourth passage, confluent cultures of growth zone and resting zone chondrocytes, as well as neonatal rat muscle mesenchymal cells, which had been cultured as described previously (Schwartz et al. 1989). Poly(A+) RNA was obtained by fractionating total RNA using oligo(dT) cellulose chromatography (Aviv and Leder 1972). Northern blot analysis was performed as described by Foumey et al. (1988), with a modified procedure for formaldehyde agarose gel electrophoretic separation of RNA (Lehrach et al. 1977; Davies et al. 1986). Preparations of RNA loaded on 1% agarose gels were electrophoresed at 4 volts/cm gel length for 6 hours and transferred to Nitroplus 2000 filters (Micron Separation Inc.) in 10X SSC at room temperature. The filter was baked at 80°C under vacuum and prehybridized in 50% formamide, 5X SSPE, 5X Denhardt, and 250 I.rg/mL denatured E. coli DNA at 37°C for 3 to 5 hours. The filter was then hybridized at 37°C overnight in the same solution plus 10% dextran sulfate and the 32P-labeled probe. The cDNA probes were labeled with 32P as described by Feinberg and Vogelstein (1983). After hybridization, filters were washed and autoradiographed (Foumey et al. 1988). The RNA blots were analyzed by densitometric quantitation (Quick Scan R & D, Helena Laboratories). Relative intensities of the hybridization signals were calculated with the aid of a GS 370 program (Hoefer Scientific Instruments). The relative amount of mRNA loaded in each lane of the agarose gel was determined by the amount of B-actin mRNA detected in each lane. Filters were stripped in 50% formamide, 1OmM Tris, pH 7.5, 1mM ethylene diamine tetraacetic acid (EDTA), 0.1% sodium dodecyl sulfate (SDS) for 2 to 4 hours at 65°C before hybridization with a second probe. Three cDNA probes were used: human a,-HS cDNA isolated from a liver library (Lee et al. 1987); human liver/bone/ kidney-type alkaline phosphatase cDNA (Weiss et al. 1986); and chicken B-actin cDNA (Cleveland et al. 1980).

Fig. 1. Detection of AHSG-like mRNA in rat chondrocytes. Poly(A+)

RNAs (IO pg) isolated from human liver (lane 1), rat liver (lanes 2 and 3), neonatal rat muscle mesencymal cells (lane 6). resting zone chondrocytes (lane 8). and growth zone chondrocytes (lane 10) were used for Northern blot analysis using human cDNA as a probe. Lanes 4, 7, 9 and 11 are 10 pg of poly (A-) RNA from rat liver, mesenchymal cells, resting zone chondrocytes and growth zone chondrocytes respectively. Molecular weight markers in kb are indicated. The mRNA detected in liver (lanes 1, 2 and 3) was 1.6 kb, while the mRNA detected in chondrocytes and mesenchymal cells (lanes 6, 8 and 10) was 2.2 kb in size.

human o,-HS-glycoprotein (Calbiochem) for one hour at 37°C. After extensive washing, the sections were incubated for 90 minutes in peroxidase-labeled Fab from sheep anti-mouse IgG, diluted 1:50 in PBS + 1% bovine serum albumin and 0.05% saponin (Onnanns and Schaeffer 1985), followed by incubation in 0.1% diaminobenzidine and 0.0 1% H,O, in Tris-HCI-sucrose buffer for 5 to 10 minutes to visualize the peroxidase-antibody conjugate.

Results Detection of AHSG-like mRNA in chondrocvtes

Immunoperoxidase

staining

In order to determine whether chondrocytes could synthesize a,-HS-glycoprotein in situ, immunoperoxidase labeling was used. Because antisera to rat (r,-HS were unavailable, antisera to human a,-HS were used. These antisera (Atlantic Antibodies), as well as an antiserum provided by Dr. David Malone (St. Louis Unrversity, St. Louis, MO), failed to cross-react with rat cw,-HS-glycoprotein. Accordingly, human fetal tissues (gift of Dr. Asher Omoy, Hebrew University, Jerusalem, Israel) were analyzed. Epiphyses derived from 11-week-old human fetuses were fixed in one of the following: 4% buffered paraformaldehyde, Comoy’s fixative, or Bouin’s fixative (Baker 1958). Fixed tissues were either embedded in paraffin or frozen in liquid nitrogen, then serial-sectioned. Sections were incubated with 3% hydrogen peroxide in methanol to block endogenous peroxidase activity, washed, incubated with 10% chicken albumin, and then treated with the appropriate diluted goat anti-

Transcripts of AHSG-like mRNA were detected in rat chondrocyte cultures derived from resting zone and growth zone cartilage by Northern blot analysis and hybridization to a radiolabeled human AHSG cDNA probe. The hybridization pattern is shown in Fig. 1. A discrete fragment hybridized to AHSG was detected in both resting zone and growth zone chondrocytes. The hybridization signal in resting zone chondrocytes (lane 8) is stronger than that in growth zone chondrocytes (lane 10). Transcripts of AHSG-like mRNA were also present in mesenchymal cells of one-day-old rats (lane 6). The molecular weight of the rat mesenchymal (lane 6) and chondrocyte mRNA (lanes 8 and 10) was larger (2.2 kb) than the AHSG mRNA (1.6 kb) observed in livers from human (lane 1) or rat (lanes 2 and 3). Chondrocytes cultured in the presence or absence of 10% FBS and in the presence of 1 pg/mL AHSG for 24 hours before harvest gave similar results (data not shown). No AHSG mRNA transcripts could be detected in RNAs derived from two other bone-derived

F. Yang et al.: a,-HS-glycoprotein

OGC -RC

30

!T

3 ;

and chondrocytes

*

t

i

rL

2.4+

P

a

18

? L a

12

z

I

0.6

3

00 0

01

001

1

10

~zHS bLg/m~)

100

10 %

FBS

Fig. 3. Effect of a,-HS-glycoprotein (0.01 to 100 pg/mL) on the specific activity of alkaline phosphatase in the cell layers of confluent, fourth passage resting zone (RC) and growth zone (GC) chondrocytes. Data represent the mean 2 SEM of six samples. *Denotes p < 0.05 for sample versus control. Data shown are from a single experiment. Each experiment was repeated three times.

binding of the antibody to its antigen. The intensely dark staining material within the chondrocyte lacunae represents poorly fixed cell bodies, an artifact of the delay in fixation following collection of the fetal tissue, and is not reaction product. The greatest amount of reaction product was found in association with hypertrophic chondrocytes, of which Fig. 2 is an example. Effect of exogenous a,-HS-glycoprotein on chondrocyte alkaline vhomhatase activitv ‘

B

+-

Fig. 2. Micrograph of human fetal hypertrophic

epiphysial cartilage fixed in Bouin’s fixative, paraffin embedded, sectioned at 2 pm and immunoperoxidase stained. Arrows indicate reaction product following incubation with goat anti-human cr,-HS-glycoprotein (a). The negative control shown in 2b was incubated with normal goat serum. Initial magnification: 350 X .

cell lines studied, normal human osteoblasts. (total RNA provided by Dr. Marian Young, National Institute of Dental Research, Bethesda, MD), and human osteosarcoma U-2 OS cells (data not shown). The hybridization signal detected in rat liver is rather weak compared to that of human liver. While the rat AHSG gene has not been cloned, we have noted that the degree of nucleotide identity between human and mouse AHSG cDNA sequences is 68%. The rat AHSG mRNA could share even lower nucleotide homology with human AHSG mRNA in comparison with that of mouse, as judged by repeated Northern analysis in our laboratory. The hybidization signal is expected to be weak after the Northern filter has been washed under high stringency. Localization of ct,-HS-alvcovrotein in human fetal cartilage tissue a,-HS-glycoprotein was detected in human fetal cartilage tissue by immunoperoxidase labeling using anti-human a,-HS-glycoprotein antibodies (Fig. 2a). Most reaction product was restricted within the chondrocyte lacunae, although there was some evidence of antibody-reactive protein in the intercellular matrix. Absence of reaction product in the control sections (Fig. 2b) demonstrated that the reaction product was due to specific

‘

,

Addition of a,-HS-glycoprotein to cultures of growth zone and resting zone chondrocytes resulted in a dose dependent stimulation of alkaline phosphatase specific activity in the cell layer of each cell type after 24 hours. Figure 3 compares control chondrocyte cultures incubated in serum-free media for 24 hours to cultures incubated with 0.01 to 100 p,g/mL 02HS-glycoprotein. Stimulation of alkaline phosphatase was apparent at levels as low as 1 Fg/mL in both growth zone and resting zone cell cultures, although the magnitude of response differed. There was a 2.7-fold increase in enzyme activity in the growth zone chondrocyte cultures and a 2.0-fold increase in resting zone cell cultures. There was no statistically significant difference in the stimulation of alkaline phosphatase activity by 1 pg/mL a,-HS-glycoprotein and that observed in cultures incubated with 10% FBS. Growth zone chondrocyte cultures stimulated with 10% FBS increased in alkaline phosphatase activity 3.4-fold. There was a 25fold elevation of enzyme activity in resting zone cell cultures under comparable conditions. In humans, the level of a,-HS-glycoprotein in the serum of healthy individuals ranges from 570 to 760 pg/mL (Dickson et al. 1983). The serum concentration of fetuin, the bovine homologue of human au,-HS-glycoprotein, is around 50 mg/mL in the fetuses and decreases to 50 p,g/mL shortly after birth (Dziegielewska et al. 1980). Although a,-HS-glycoprotein and fetuin, both single copy genes, are derived from the same gene in two different species (Dziegielewska et al. 1990; Yang et al. unpublished result), their modes of gene expression during development are very different (Yang et al. unpublished results). It is not known whether they have similar biological functions. The concentration of o,-HS-glycoprotein in the microenvironment of chondrocytes in vivo is not known. Therefore, it is not possible to correlate the optimal concentration of the effect of au,-HS-glycoprotein to physiological conditions. Data presented in Fig. 4 demonstrate that the a,-HS-glyco-

F. Yang et al.: a,-HS-glycoprotein 7.0

r

and chondrocytes

.-.GC W n RC

6.0

*

i 4.0

*

;

/!/‘\T

50

11

l

30 I 2.0-1_

0 0.03

0 06 9HS

Fig.

0 13

0.25

0.50

1.00

0 01

0.1

10.00

(dml)

4. Effect of a,-HS-glycoprotein (0.03 to 10 kg/mL) on the specific

activity of alkaline phosphatase in confluent, fourth passage resting zone (RC) and growth zone (GC) chondrocytes. Data represent the mean ? SEM of six samples. *Denotes p < 0.05 for sample versus control. Data shown are from a single experiment. Each experiment was repeated three times.

protein-dependent elevation in alkaline phosphatase specificactivity in both growth zone and resting zone chondrocyte cultures occurred over a dose range of 0.25, 0.5, and 1 kg/mL+. In growth zone cell cultures, 0.5 Fg/mL a,-HS-glycoprotein caused the greatest increase in enzyme activity. However, in the resting zone cell cultures, each of the three concentrations of CQHS-glycoprotein demonstrated comparable effects. The matrix vesicles isolated from control cultures were enriched more than 2-fold in alkaline phosphatase specific activity when compared to the plasma membranes. The effect of exogenous a,-HS-glycoprotein on alkaline phosphatase activity was comparable in the plasma membrane and matrix vesicle fractions. Although 0.1 pg/mL a,-HS-glycoprotein caused a slight but statistically significant elevation (1.3-fold) of alkaline phosphatase specific activity in the matrix vesicles isolated from growth zone chondrocyte cultures, there was no significant effect on the plasma membrane enzyme activity (Fig. 5). At concentrations of 1 pg/mL a,-HS-glycoprotein, a 1.6-fold

Fig. 5. Effect of cY,-HS-glycoprotein (0.01 to 100 pg/mL) on the specific activity of alkaline phosphatase in matrix vesicles (MV) and plasma membranes (PM) isolated from confluent, fourth passage rat costochondral growth zone chondrocytes. Each sample represents the combined plasma membranes or matrix vesicles from three T-75 flasks. *Denotes p < 0.05 for sample versus control. Data are from a single experiment. Each experiment was repeated three times.

increase was seen in matrix vesicle enzyme activity and a 2.3-fold elevation in the plasma membrane fraction. Addition of 10% FBS to the culture media resulted in a 1.5-fold increase in enzyme activity in the matrix vesicles and a 2.6-fold increase in the plasma membranes isolated from growth zone chondrocyte cultures. Similar results were observed in cultures of resting zone cells (Table I). The effect of a,-HS-glycoprotein on alkaline phosphatase activity was independent of cell number. Whether enzyme activity was calculated as kmol Pi/mg protein/minute or as a

function of cell number, 0.63 and 1.25 pg/mL of a,-HSglycoprotein stimulated enzyme activity in matrix vesicles and plasma membranes from both resting zone (Table I) and growth zone (Table II) chondrocytes. Administration of a,-HS-glycoprotein had no effect on cell number at the concentrations examined (data not shown). Specific activity of phospholipase A, was enriched in the matrix vesicles isolated from growth zone chondrocyte cultures

Table I. The effect of cc,-HS-glycoprotein on alkaline phosphatase specific activity in matrix vesicles (MV) and plasma membranes (PM) isolated from resting zone chondrocyte cultures.

WSGI

MV 0.00

0.32 0.63 1.25 2.50 5.00 10.00

~Mol Pi/l@ Cells/Minute

~Mol Pi/mg Protein/Minute

63.6 f 4.6 79.7 f 9.4 107.3 f 15.8’ 123.0 f 18.7’ 50.0 f 8.0 64.1 f 10.5 55.7 ? 6.1

*p < 0.05: samples vs. control. Data are the means f SEM of six samples.

PM 11.4 + 3.1 20.6 i 6.7 23.7 2 2.3’ 24.6 t 4.6’ 9.2 f 2.2 18.1 i 2.6 18.5 i 2.5

MV 0.7 f 0.9 f 1.2 f 1.3 i 0.5 i 0.6 i 0.6 f

0.0 0.2 0.2’ 0.2’ 0.1 0.1 0.1

PM 0.03 t 0.02 0.03 f 0.00 0.05 ?:0.00’ 0.05 i 0.00’ 0.04 i 0.00 0.03 t 0.00 0.04 f 0.00

F. Yang et al.: a,-HS-glycoprotein

12

and chondrocytes

Table II. The effect of a,-HS-glycoprotein on alkaline phosphatase specific activity in matrix vesicles (MV) and plasma membranes (PM) isolated from growth zone chondrocyte cultures.

[.4HSGl

aMo1 Pi/mg Protein/Minute

PM

MV

fig/ml

~Mol Pi/l@ Cells/Minute

MV

PM

0.00

103.0 i 6.2

45.1 * 10.4

0.8 i 0.1

0.07 * 0.02

0.32

106.9 + 5.4

38.5 i 17.9

0.7 i 0.1

0.05 f 0.03

0.63

148.8 i 9.3’

92.2 i 16.2’

1.0 + 0.1’

0.13 + 0.07’

1.25

135.3 i 8.7’

83.1 i 11.4’

1.1 f 0.1’

0.13 f 0.02’

2.50

x1.1 Lt 9.9

24.2 i 10.9

0.8 t 0.3

0.04 ? 0.00

5.00

126.3 r 15.6

21.0 t 16.1

0.7 i 0.1

0.02 i 0.01

10.00

112.2 ? 19.3

51.4 * 14.5

0.7 f 0.1

0.08 ?L0.03

*p < 0.05: Sample vs. Control. Data used are the means + SEM of six samples.

compared to the plasma membranes (Table III). When data were calculated as a function of cell number, enzyme activity was enriched in matrix vesicles isolated from cultures incubated with 0.63 pg/mL a,-HS-glycoprotein. Specific activity in these matrix vesicles was enriched 8.3-fold. The effect of a,-HSglycoprotein was greatest in the matrix vesicle fraction. Addition of cu,-HS-glycoprotein (0.63 to 1.25 pg/nL) to growth zone chondrocyte cultures resulted in a 3.1-fold stimulation in matrix vesicle phospholipase A, specific activity. When the data were calculated as a function of cell number, stimulation was still significant, but the apparent magnitude of the increase was smaller (1.7X). No statistically significant effect on the enzyme activity in the plasma membrane was observed; however, at 2.5 p.g/mL cw,-HS-glycoprotein, there was an apparent increase in plasma membrane specific activity. There was no effect of a,-HS-glycoprotein on phopholipase A, activity in matrix vesi-

cles or plasma membranes cyte cultures (Table IV).

isolated from resting zone chondro-

The effect of a,-HS-glycoprotein administration on alkaline phosphatase mRNA levels in chondrocytes There appeared to be no significant increase in the amount of alkaline phosphatase mRNA when growth zone or resting zone chondrocytes cultured in the presence of 1 p,g/mL o,-HSglycoprotein were compared to either type of cells in serum-free medium (Fig. 6). Slightly more alkaline phosphatase mRNA was detected in resting zone chondrocytes grown in the absence of a,-HS-glycoprotein (lane 5) when compared to cells grown in the presence of o,-HS-glycoprotein (lane 6). No detectable alkaline phosphatase mRNA was observed in chondrocytes

Table III. Tbe effect of a,-HS-glycoprotein membranes ImSGl

on phospholipase A, activity in matrix vesicles (MV) and plasma (PM) isolated from growth zone chondrocyte cultures. % Hydrolysis/mg

MV

Protein/Minute

PM

% Hydrolysis/l06

MV

Cells/Minutt

PM

0.00

20.9 i 6.4

5.5 f 1.5

0.14 i 0.03

0.32

18.7 +. 4.4

9.1 ? 4.3

0.09 i 0.03

0.13 * 0.06

0.63

64.9 t 9.4’

7.8 t 3.8

0.24 f O.Ol*

0.09 f 0.03

1.25

59.6 t 11.9’

8.5 ? 4.6

0.22 * 0.02’

0.19 i 0.04

2.50

29.9 ? 2.5

11.1 i 4.5

0.15 * 0.02

0.14 f 0.06

5.00

36.6 t 7.3

10.5 ? 2.9

0.13 i 0.04

0.02 f 0.01

10.00

38.8 + 9.8

4.4 i 1.4

0.11 i 0.01

0.11 f 0.04

*p