Growth-inhibitory glycopeptides obtained from the cell surface of cultured chick embryo fibroblasts

Experimental Cell Research 154 (1984) 147-154

Growth-Inhibitory Surface

Glycopeptides Obtained of Cultured Chick Embryo YOSHIHITO

Biology

from the Cell Fibroblasts

YAOI

Division, National Cancer Center Research Institute, Tsukiji 5-l-1, Chuo-ku, Tokyo 104, Japan

Cell surface glycopeptides were obtained from cultured chick embryo fibroblasts (CEF) by digestion with Pronase E, and a fraction exerting growth-inhibitory activity on CEF was isolated by high performance gel permeation chromatography. The active fraction, tentatively termed cell surface glycopeptide-2 (CSGP-2), was soluble in 5 % trichloroacetic acid (TCA) or 75 % ethanol. It inhibited the growth of CEF reversibly at 10-20 pg sugar/ml, but did not inhibit BALB/c mouse 3T3, SV40-transformed 3T3, and human diploid cells at similar concentration. The growth-inhibitory activity of CSGP-2 was reduced or lost after digestion with neuraminidase or oxidation with sodium metaperiodate. Cellulose acetate electrophoresis revealed that CSGP-2 was a mixture of sialoglycopeptides. A similar growth inhibitor was also isolated from chicken embryonic tissues.

Density-dependent growth inhibition of cultured animal cells has attracted particular interest because loss or decrease of this property usually accompanies malignant transformation of cells. Several mechanisms have been proposed to explain how increased cell density exerts the regulatory effect. One theory proposes that the growth restriction may be caused by some growth inhibitors that accumulate on cell surface membranes in dense cultures. For instance, Whittenberger & Glaser [I] reported that the growth of Swiss mouse 3T3 cells was inhibited by a membrane-enriched fraction obtained from 3T3 cells, and they also showed that the inhibitory activity was solubilized by octylglucoside [2]. Natraj & Datta [3] reported that a growth inhibitor was released from the cell surface of BALB/c mouse 3T3 cells by treatment with 0.2 M urea. Generally, however, it is more difficult to investigate growth inhibitors of this kind than growth-promoting factors and their chemical properties and action mechanisms have not yet been fully understood. In our previous study [4], we found that a glycopeptide fraction isolated from the cell surface membrane of cultured CEF by mild digestion with proteolytic enzymes exerted some growth-inhibitory activity on CEF. The inhibitor was soluble in 5 % trichloroacetic acid (TCA) or 75 % ethanol and thus was separable from undigested proteins or proteoglycans. A partially purified preparation inhibited the growth of CEF at a concentration of 100 pg sugar/ml [4]. Since this phenomenon seemed to be unique and interesting with respect to the regulatory mechanisms of cell proliferation, we further purified and characterized the inhibitory factor. The present paper describes results indicating that the inhibitor is a Copyright @ 1984 by Academic Press, Inc. All rights of reproduction in any form reserved 0014-4827/&l 103.00

148 Y. Yuoi mixture of sialoglycopeptides and that a carbohydrate moiety of the glycopeptides might be essential for the inhibitory activity. MATERIALS

AND METHODS

Cell Culture The primary CEF culture was prepared from 9- or IO-day-old embryos. Eagle’s minimal essential medium (MEM) (Nissui Seiyaku Co.) supplemented with 7.5 % fetal bovine serum (Gibco) was used throughout the experiments. Radioactively labelled growth inhibitor was extracted from cells grown to full confluence in the presence of 0.025 @i/ml of [6-3H] D-ghCOSaIUine (26.8 Ci/mmol, Amersham) in Nunclon square culture flasks (530 cm*). Growth-inhibitory activity was determined on aliquots of test samples that were sterilized in Falcon multiwell culture plates by brief UV irradiation. The secondary CEF culture was seeded at 2-3 x lo3 cells/cm *. Cell numbers were determined using a Coulter counter after cultivation for 6 days, when the cells in control wells reached saturation density. The growth rate of BALB/c mouse 3T3 cells (strain A-31) and their SV40 transformants (supplied by Dr A. Hiragun, Tokyo Metropolitan Institute of Medical Sciences) and human diploid cells, strain TIG-1 (supplied,by Dr M. Ohashi, Tokyo Metropolitan Institute of Gerontology, [5], was similarly determined.

High-performance

Liquid

Chromatography

Approx. 200 pl of sample was loaded on a column of Toyo Soda TSK G-3000 SW (0.75~60 cm, equipped with a 5-cm precolumn) and eluted with IO-fold diluted Ca*+- and Mg*+-free phosphatebuffered saline (PBS) at a flow rate of 0.5 ml/mm. Radioactivity in each fraction was determined with a liquid scintillation counter using ACS II (Amersham) as a scintillator. Sugar content was determined by the phenol-sulfuric acid method using glucose as a standard [6].

Autoradiography Secondary CEF were seeded in Lab-Tek tissue culture chamber slides (2 cm*) at a density of 2.5X lo3 cells/cm*. The cells were labelled at intervals with 0.5 @i/ml of [methyl, l’,2’-3H]thymidine (113 Ci/mmol, New England Nuclear) for 24 h. The labelled cells were washed with Ca*‘- and Mg*+free PBS, fixed with ethanol-acetic acid (3 : 1, v/v), washed with 70% ethanol, and dried. Autoradiography was performed using Sakura NR-M2 nuclear emulsion, and developed after a 5-day exposure. The preparations were stained with Giemsa dye, and the labelling index was obtained by counting 300 cell nuclei in each chamber.

Cellulose Acetate Electrophoresis An aqueous solution of crude or purified glycopeptides labelled metabolically with [3H]glucosamine was spotted on a sheet of cellulose acetate (Oxoid, 3.5-cm wide) and electrophoresed in pyridine-acetic acid-water (3 : 1 : 387, v:v:v, pH 5.4) at 1 mA/cm until the marker dye (xylenecyanol FF) migrated 3.75 cm toward the anode (25 min). After drying, the cellulose acetate film was cut into 0.25 cm pieces, and radioactivity in each piece was determined in a Beckman liquid scintillation counter.

Periodate

Oxidation

and Mild Acid Hydrolysis

About 100 Pg of salt-free glycopeptides was dissolved in 0.5 ml of 0.05 M sodium metaperiodate in 0.1 M acetate buffer, pH 4.5, and allowed to react at 4°C for 48 h. Chemical reagents were removed by passing through a small column of Sephadex G-25. Mild acid hydrolysis was performed in 0.01 N HCl at 80°C for 1 h, and HCl was removed by lyophilization.

Enzymes and Chemicals Pronase E and Arthrobacter ureafaciens neuraminidase (free from protease, glycosidase, and Nacetylneuraminic acid aldolase) were purchased from Kaken Chemicals and Nakarai Chemicals, respectively. N-Acetylneuraminic acid was obtained from Sigma. Exp Cell Res 154 (1984)

Growth-inhibitory

f

250

‘p 2

200

$

150

; z

100

2 J m

glycopeptides

149

Fig. 1. Purification of growth-inhibitory factor from cultured CEF by high performance gel permeation chromatography. A crude glycopeptide fraction obtained from 20 culture dishes (530 cm2) was loaded on a column of Toyo Soda TSK G-3000 SW (0.75~60 cm, equipped with a 5-cm precolumn) and eluted with IO-fold diluted Ca*+- and Mg2+-free PBS at 0.5 ml/min. Fractions of 0.5 ml were collected and each tube was tested for 0, radioactivity; 0, growth-inhibitory activity, and sugar content (lower figure).

50

15

10 Effluent

2E

20 volume

(ml)

RESULTS Isolation of Growth-Inhibitory

Glycopeptides

A confluent monolayer culture of CEF, which had been metabolically labelled with [3H]glucosamine throughout the cultivation period, was washed twice with Ca2+- and Mg 2+-free PBS and incubated with 0.05% Pronase E and 0.02% ethylenediaminetetraacetic acid (EDTA) in Ca’+- and Mg2+-free PBS at 37°C for 20 min. Dispersed cells were removed by centrifugation, and the supernatant solution was digested with 0.5 mg/ml of freshly added Pronase E at 37°C for another 3 h. The enzyme and residual proteins were removed by precipitation with 5 % TCA. The supernatant solution was neutralized and proteoglycans were removed by precipitation with 75 % ethanol in the presence of 5% sodium acetate. The soluble fraction was concentrated, dialysed against distilled water,

Fig. 2. Purification of growth-inhibi-

tory factor from chicken embryonic tissues by high performance gel permeation chromatography. Crude glycopeptides from 20 embryos were fractionated as described in fig. 1.0, Sugar content; 0, growthinhibitory activity in each tube were determined. 5

10

15 Effluent

voiume

20 (InI

25 )

Exp Cell Res 154 (1984)

150 Y. Yuoi

0 5 Dose

(pg

10

15

sugar/ml)

a0

0123456769 Cultivation

(days)

Fig. 3. Growth-inhibitory activity of CSGP-2 on several types of cultured cells. Cells were cultivated with or without various concentrations of CSGP-2 obtained from cultured CEF. Cell numbers were counted after cultivation for 6 days, and those in control plates were referred to as 100%. 0, CEF; A, BALB/c mouse 3T3, strain A-31; 0, SV40-transformed 3T3 and A, human diploid cells, strain TIG-1 were tested. Fig. 4. Effect of CSGP-2 on the growth rate and [3H]thymidine uptake of cultured CEF. Secondary cultures were seeded in Lab-Tek chamber slides (2 cm’) 0, with; or 0, without 20 pg/ml of CSGP-2. (A) Cell numbers were counted by using a Coulter counter at 24-h intervals. Medium containing CSGP-2 in the test cultures was replaced by the fresh one without CSGP-2 on the fifth day (arrow). Each point represents the mean value of two cultures. (B) [3H]Thymidine (0.5 @i/ml) was added at 24-h intervals and the cells were labelled for 24 h. Numbers of labelled nuclei were determined by autoradiography. Each result, representing the mean value of three experiments, was plotted at the end of the labelling period.

and fractionated by high-performance gel permeation chromatography on a column of TSK G-3000 SW. Each fraction was examined for radioactivity, sugar content and growth-inhibitory activity, as described in Materials and Methods. As shown in fig. 1, growth-inhibitory activity was detected at three or four positions in the elution profile, although the second peak, tentatively termed cell surface glycopeptide-2 (CSGP-2), usually exerted the most profound inhibitory activity. This fraction was pooled and used in the subsequent experiments. The yield of CSGP-2 was 50 to 100 ug from lo9 cells as determined by sugar content. CSGP-2 was also isolated from chicken embryonic tissues by the same procedures. Decapitated and eviscerated embryos (16day-old) were minced and washed repeatedly with Ca2+- and Mg2+- free PBS and then incubated with 0.05% Pronase E and 0.02% EDTA in Ca2+- and Mg2+-free PBS at 37°C for 20 min. After cells and debris were removed by centrifugation, the growth-inhibitory factor was purified by successive precipitation with 5% TCA and 75% ethanol, Exp Cell Res 154 (1984)

Growth-inhibitory

glycopeptides

15 1

followed by high-performance gel permeation chromatography as described above. The elution profile (fig. 2) agreed well with that obtained with cultured CEF: the most profound growth-inhibitory activity was eluted at the coincident position with CSGP-2 (indicated by an arrow in fig. 2). The yield of the active material was about 50 mg sugar from 20 embryos. Inhibitory

Properties

of CSGP-2

Fig. 3 shows the effect of CSGP-2 on the growth rate of CEF, BALB/c mouse 3T3, SV40-transformed 3T3, and human diploid cells. CEF growth was inhibited markedly by 10-20 &ml of CSGP-2 in a dose-dependent manner. Although the data are not shown here, corresponding inhibitory fraction obtained from embryonic tissues showed similar specific activity on CEF growth. In contrast, the growth of 3T3, SV40-3T3 and human diploid cells was not affected significantly by up to 20 ug/ml of CSGP-2.. Fig. 4 shows the effect of CSGP-2 that was obtained from cultured cells on the growth curve and [3H]thymidine uptake of CEF. Neither growth rate nor thymidine incorporation was affected by CSGP-2 during the period of the first 2 days. The initial growth rate in the control cultures was relatively low because the cells were seeded at rather low cell density [7]. Inhibition of DNA synthesis in the test cultures became apparent on the third day, and the decline of the growth rate followed. More than 90% of the cells in the inhibited cultures were viable on the fifth day, as judged by the vital staining with trypan blue. When the medium containing CSGP-2 was replaced by the fresh one without CSGP-2, active DNA synthesis commenced in more than 70% of the cells within 24 h. The cells continued growing thereafter and reached a saturation density (about lo5 cells/cm2) 4 days later. Chemical

or Enzymatic

Modification

of CSGP-2

Purity of the glycopeptides was examined by cellulose acetate electrophoresis and the results are shown in fig. 5. The crude glycopeptide fraction was shown to contain a number of neutral and acidic components (fig. 5a). CSGP-2 gave a relatively simple, although still heterogeneous, band having an acidic electrophoretic mobility (fig. 5 b). When CSGP-2 was digested with neuraminidase (0.5 unit/ml, 37°C 22 h) and electrophoresed, the original acidic band disappeared and was replaced by two neutral and one acidic band (fig. SC). Electrophoretic mobility of the acidic band coincided with that of standard N-acetylneuraminic acid (indicated by an arrow). The two neutral bands seemed to be asialoglycopeptides. These results suggested that CSGP-2 was a heterogeneous mixture of sialoglycopeptides. This was also supported by the results of mild acid hydrolysis (0.01 N HCl, 80°C 1 h), which allows specific cleavage of sialyl bonds in sugar chains (fig. 54. Electrophoresis showed that although the original CSGP-2 band did not disappear completely, probably because of incomplete hydrolysis of sialyl Exp Cell Res 154 (1984)

152 Y. Yuoi NANA

Origin

,-10

12 Mobility

34

56 (cm)

I 10

0

Dose

1 6

20

(pa sugar/ml)

Fig. 5. Cellulose acetate electrophoresis of glycopeptides. Electrophoresis was carried out in pyridine-acetic acid buffer (pH 5.4) at 1 mA for 25 min, and glycopeptides were detected by radioactivity of [3H]glucosamine. (n) Crude glycopeptides; (6) CSGP-2; (c) CSGP-2 digested with 0.5 unit/ml of A. ureufaciens neuraminidase at 37°C for 22 h; (d) CSGP-2 hydrolysed with 0.01 N HCI at 80°C for 1 h. The arroul (NANA) indicates the mobility of standard N-acetylneuraminic acid. Fig. 6. Effect of neuraminidase digestion and periodate oxidation on the growth-inhibitory activity of CSGP-2. About 50 ug of CSGP-2 was incubated A, with or 0, without 0.5 unit/ml of A. ureafuchs neuraminidase at 37°C for 24 h. Growth rate of cultured CEF was determined as described in fig. 3. Control cultures (referred to as 100%) received the same concentration of neuraminidase. Oxidation with 0.05 M sodium metaperiodate (A) was performed at pH 4.5, 4°C for 48 h.

bonds, the electrophoretic profile of the newly formed components resembled that obtained with neuraminidase. We tested the function of the carbohydrate moiety of CSGP-2 in the growthinhibitory phenomenon by examining the effects of neuraminidase digestion and periodate oxidation on the inhibitory activity. Enzymatic digestion was performed with 0.5 units of neuraminidase at 37°C for 24 h. In the control experiments, either the enzyme or CSGP-2 alone was incubated under the same conditions. As shown in fig. 6, marked reduction of the growth-inhibitory activity of CSGP-2 was found after digestion with neuraminidase. The enzyme alone did not affect the growth rate of CEF, and incubation of CSGP-2 without the enzyme did not cause any reduction in its growth-inhibitory activity. We also confirmed that N-acetylneuraminic acid had no effect on the growth rate of CEF at a concentration below 100 &ml (data not shown). Residual inhibitory activity of neuraminidase-digested CSGP-2 is therefore not due to free sialic acid released from CSGP-2. The growth-inhibitory activity of CSGP-2 was destroyed completely by oxidation with sodium metaperiodate (4°C 48 h). Exp Cell Res 154 (1984)

Growth-inhibitory

glycopeptides

153

DISCUSSION Proteolytic enzymes cleave a number of glycoproteins exposed on cell surface membranes, resulting in the release of glycopeptides of various sizes. In experiments reported previously, we obtained those glycopeptides from cultured CEF and suggested that some of them might exert growth-inhibitory activity on CEF [4]. In the present study, we used high performance gel permeation chromatography to further purify glycopeptides showing the inhibitory activity. The main fraction, termed CSGP-2, inhibited the growth of cultured CEF at 10-20 pg/ml, indicating that the specific activity of this preparation was approximately ten times higher than that of the one previously reported [4]. Although the elution position of CSGP-2 indicated that its molecular weight was about 25 K, this might not be a true estimate because glycopeptides tend to aggregate at low ionic concentrations. Preliminary analysis by Sephadex G-50 chromatography indicated that the average molecular weight of CSGP-2 was about 4 K. The present results demonstrated that the growth inhibition by CSGP-2 was reversible. The cell cycle was probably blocked at Gl in the presence of CSGP-2, since the activation of DNA synthesis preceded the increase in cell number upon removal of the inhibitor. It was also suggested that CSGP-2 affected the saturation density of cultured CEF rather than the initial rate of growth or thymidine uptake. Cellular viability was not significantly affected during 5-day cultivation in the presence of 20 CLg/mlof CSGP-2. Growth of BALB/c mouse 3T3, SV40-3T3 and human diploid cells was not inhibited by 20 pg/ml of CSGP-2. So far, cultured CEF was the only cell type sensitive to growth inhibition by CSGP-2. These results seem to exclude the possibility that the inhibition might be due to nonspecific cytotoxity of some contaminants present in low concentration. Cellulose acetate electrophoresis revealed that CSGP-2 was a mixture of sialoglycopeptides. The growth-inhibitory activity was markedly reduced by digestion with neuraminidase and completely lost by prolonged oxidation with sodium metaperiodate. In contrast, this activity of CSGP-2 was retained after extensive digestion with Pronase E used in the isolation procedures. These data strongly suggest that the carbohydrate moiety of CSGP-2 is essential to the growth-inhibitory activity. The growth inhibitor reported here is different from the other inhibitors obtained so far from the surface membranes of cultured cells [l-3] in protease sensitivity and cell type specificity and may be an entirely new class of growthregulating substance. A similar inhibitor was also isolated from chicken embryonic tissues, suggesting that it might play regulatory roles in vivo as well as in vitro. In considering the biological function of CSGP-2 in growth control of cultured animal cells, it would be relevant to note that several proteolytic enzymes [8-lo] and neuraminidase [ 1l] are known to induce DNA synthesis and cell division in density-inhibited CEF. It is tempting to infer that the mitogenic effect of these hydrolytic enzymes might be explained in part by removal of the growth-inhibiExp Cell Res 154 (1984)

154 Y. Yuoi tory glycopeptides from cell surface membranes or their inactivation by removal of sialic acid residues. Several pieces of evidence suggest the possible existence of glycoproteins on the cell surface that are directly involved in growth regulation. For example, Kinders et al. [12, 131reported that mild proteolysis of mouse or bovine cerebral cortex resulted in the release of at least two kinds of large glycopeptides from the cell membrane, which could inhibit growth and protein synthesis in cultured cells. Properties of their glycopeptides, however, seem to be considerably different from CSGP-2, since they are precipitated with 2 vol of ethanol and inactivated by high concentration of proteases. CSGP-2 seems to be the first case in which sugar moiety has been suggested to be essential to growth regulation. Since CSGP-2 is not yet pure, more rigorous purification is necessary to substantiate those assumptions. It would be also important to identify the cell surface glycoprotein(s) containing the growth-inhibitory glycopeptides. Those studies are in progress to gain confidence on the physiological importance of the growth-inhibitory glycopeptides. REFERENCES 1. 2. 3. 4. 5.

Whittenberger, B & Glaser, L, Proc natl acad sci US 74 (1978) 2251. Whittenberger, B, Raben, D, Lieberman, M A & Glaser, L, Proc natl acad sci US 75 (1978) 5457. Natraj, C V & Datta, P, Proc natl acad xi US 75 (1978) 6115. Yaoi, Y & Motohashi, K, Gann monogr cancer res 25 (1980) 29. Ohashi, M, Aizawa, S, Oota, H, Ohsawa, T, Kaji, K, Kondo, H, Kobayashi, T, Noumura, T, Matsuo, M, Mitsui, Y, Murota, S, Yamamoto, K, Ito, H, Shimada, H & Utakoji, T, Exp geront 15 (1980) 121. 6. Dubois, M, Gilles, K A, Hamilton, J K, Rebers, P A & Smith, F, Anal them 28 (1956) 350. 7. Yaoi, Y & Kanaseki, T, Nature 237 (1972) 283. 8. Sefton, B M & Rubin, H, Nature 227 (1970) 843. 9. Blumberg, P M & Robbins, P W, Cell 6 (1975) 137. 10. Zetter, B R, Chen, L B & Buchanan, J M, Cell 7 (1976) 407. 11. Vaheri, A, Ruoslahti, E & Nordling, S, Nature 238 (1978) 211. 12. Kinders, R J, Milenkovic, A G, Nordin, P & Johnson, T C, Biochem j 190 (1980) 605. 13. Kinders, R J & Johnson, T C, Biochem j 206 (1982) 527. Received December 16, 1983 Revised version received March 26. 1984

Exp Cell Res 154 (1984)

Printed

in Sweden