Printed in Sweden Copyright Q 1975 by Academic Press, Inc. All rights of reproduction in any form reserved

Experimental Cell Research92 (1975) 87-94

AN EVALUATION FACTOR

OF LYSOSOMAL

INFLUENCING

ENZYME

THE BEHAVIOUR

LEAKAGE

AS A

OF TRANSFORMED

CELLS JOANNE

D. ELLIGSEN,

J. E. THOMPSON

and H. E. FREY

Departments of Biology and Physics University of Waterloo, Waterloo, Ont., Canada N2L3GI

SUMMARY Commercially available fetal calf serum has been found to contain substantial amounts of protease, acid phosphatase, and glycosidase activity. All such activity can be inactivated by heating serum to 70°C for 15 min. Moreover, this heat inactivation has no influence on the ability of serum to support growth of pyBHK cells, and only marginally affects the growth properties of BHK cells. Glycosidases and acid phosphatase were also detectable in medium containing heat-inactivated serum and conditioned by growth of either BHK or pyBHK cells. This latter activity can be attributed solely to enzymes derived from the growing cells and, for most of the enzymes monitored, was present to a greater extent in medium conditioned by transformed cells as compared to normal cells. In all cases, however, enzyme activity derived solely from cells was only 1 to 10 % of that indigenous to commercial tissue culture medium by virtue of its fetal calf serum component. The relevance of this information to current speculation about a role for lysosomal enzymes as mediators of the uncontrolled growth characterizing transformed cells is discussed.

The prospect that lysosomal enzymes released into the growth medium by transformed cells may alter the architecture of the plasmalemma, and thus contribute to the loss of growth control characterizing these cells, has recently been advocated [3, 5, 12, 221. This possibility is supported by the observations that mild treatment of normal cells with protease renders them as agglutinable by plant lectins as their transformed counterparts, and also allows their escape from the inhibition of growth featured at confluence [7, 81. Moreover, it has been noted that the plasma membranes of tumour cells are leakier than those of normal cells [I] and that lysosoma1 enzymes do in fact leak from tumour cells in suspension [13, 181. However, no definitive demonstration of their leakage from cultured virally transformed cells nor

of their effect on cell growth has yet been reported. The present study was designed to more fully evaluate this question by comparing the extent of lysosomal enzyme leakage from normal and virally transformed BHK cells and examining its significance in the light of information about levels of lysosomal enzymes present as contaminants in commercially available fetal calf serum. As well, the influence that such enzyme activity in culture medium has on the growth of both normal and transformed cells has been documented. MATERIALS

AND METHODS

Cell culturing BHK and polyoma transformed (pyBHK) cells were routinely cultured at 37°C in 6 or 32 oz. glass prescripExptl Cell Res 92 (1975)

88

Elligsen, Thompson and Frey

tion bottles containing Basal Medium Eagle’s (BME) supplemented with L-glutamine, penicillin-streptomycin and 10% fetal calf serum. Stocks were subcultured at split ratios of 1: 10 to 1 :20 every 3 or 4 days to prevent selection of spontaneous transformants. Cells were removed from the glass surfaces bv treatment for 5 to 10 min at 37°C with a solution of 0.225 % trypsin and 0.09% disodium ethylene diamine tetraacetic acid (EDTA) in Ca-%+ and Mg2+free phosphate-buffered saline (PBS).

Leakage of lysosomal enzymes For experiments in which the extent of lysosomal enzyme leakage from the cells into growth medium was being monitored, normal or transformed cells were seeded in 4 glass roller bottles, each containing 100 ml of medium, at 5 x 109 cell&me. After 24 h, a period of sufficient duration to allow the cells to attach and initiate growth, the medium was removed and the cells washed twice with sterile phosphatebuffered saline (PBS) previously warmed to 37°C. One hundred ml of fresh medium containing no phenol red and supplemented with 10% heat-inactivated fetal calf serum were then added to each bottle. The phenol red was omitted to preclude its interference in calorimetric assays for lysosomal enzymes, and the serum was heated to 70°C for 15 min to inactivate its indigenous enzvme activitv. This treated medium was l;ft on the cells for 48 h and then collected. filtered through a 0.22 urn Millipore filter and lyophilized to concentrate the enzyme activity which had leaked from the growing cells. The freeze-dried material thus derived from 400 ml of conditioned medium was then dissolved in approx. 12 ml of 1 mM NaHCO,, pH 7.5, and the final volume recorded. The total number of cells, including those which had sloughed off into the medium, was also determined so that leaked enzyme activity could be expressed on a per cell basis. As a control, 400 ml of medium prepared in exactly the same manner were also incubated at 37°C for 48 h, but not exposed to cells, and then concentrated by lyophilization as described above. Both control and conditioned medium were stored at -10°C until required for assays. Assays were done within 3 days after media collection.

3.2.1.20). Each reaction mixture contained 6 pmole of the p-nitrophenyl substrate, 165 pmole of sodium citrate, and 0.5 ml of the medium being assayed for enzyme activity in a total volume of 3.3 ml adjusted to pH 4.3. The reactions were stopped after 1 h at 37°C by the addition of 2 ml of 0.4 M glycine, pH 10.5, and the amount of p-nitrophenol released was determined by measuring absorbance at 420 nm and referring to a standard curve. Under these conditions the enzyme activities varied linearly as a function of time for at least the duration of the assay. Controls in the form of medium which had been heat-inactivated, but not conditioned, were run simultaneously and subtracted. Each assay was run in duplicate. Protease activity was also monitored using Azocoll (Calbiochem) as a general protease substrate according to a procedure published by Calbiochem [9]. Azocoll is an insoluble, powdered cowhide to which a dye is attached. The cowhide contains the assortment of peptide linkages characteristic of proteins and when a proteolytic enzyme breaks any of these linkages, the bound dye is released into the suspending medium and can be measured spectrophotometrically. The assays were run in duplicate. Each assay tube contained 25 mg of Azocoll, 500 pmole of phosphate and 0.1 to 0.5 ml of the medium being assayed for enzyme activity in a total volume of 5.1 ml adjusted to pH 7.0. Under these conditions the enzyme activity varied linearly as a function of time. The reaction mixtures were incubated at 37°C for 15 min and then cooled to 0°C by immersion in crushed ice. The suspensions were cleared by centrifugation in an International clinical centrifuge and the optical density of the supernatants determined at 520 nm. These readings were equated to units of proteolytic activity by reference to a standard curve prepared by using pronase from Sigma Chemical Co. In addition to determining levels of lysosomal enzymes released into conditioned medium, the extent to which the fetal calf serum supplement of growth medium might contain lysosomal enzymes as contaminants was explored. Serum from two companies-Grand Island Biological and Microbiological Associates-was examined. Enzyme assays were performed as described above using 0.1 ml to 0.5 ml of undiluted serum. Protein was assayed as described previously by Lowry et al. [14] using bovine serum albumin as a standard.

Enzyme and protein measurements To quantitate amounts of hydrolytic enzyme activity leaked from cells during growth, conditioned and control media obtained as described above were assayed for lysosomal enzymes according to a procedure described previously in which artificial substrates are employed [2]. The following substrates obtained from Sigma Chemical Co. were used: p-nitrophenyl phosphate for acid phospha%&yz 3.1.3.2); p-nitrophenyl-N-acetyl-/I-W&osamr N-acetyl+-glucosaminidase (EC 3.2.1.29); p-nitrophenyl-N-acetyl-P-D-galactosaminide for N-acetyl$D-galactosaminidase (EC 3.2.1.53); p-nitrophenyl-pD-galactoside for j??-D-galactosidase (EC 3.2.1.23); p-nitrophenyl-a-D-glucoside for a-n-ghtcosidase (EC

Exptl

Cell Res 92 (1975)

Growth curves The effects of heat inactivation on the growth supporting properties of serum were determined by comparing growth curves for normal and transformed BHK cells in medium containina heat-treated and untreated serum. Each curve was &n in triplicate. The cells were mated at a density of 1 or 2 x 109/cm’ on plastic tissue-culture plates (sb mm diameter) to which were added 5 ml of BME supplemented with 10% heatinactivated or untreated serum. The medium was changed every 3 days. Cell counts were determined at intervals ranging from 6 to 24 h depending on the stage of growth by using a Celloscope electronic cell counter.

Lysosomal enzyme leakage

89

Table 1. Acid phosphatase, glycosidase and protease activities of commercially available fetal calf serum Enzyme

A

B

C

Acid phosphatase c+o-Glucosidase /I-D-Galactosidase N-Acetyl-&oglucosaminidase N-Acetyl+ogalactosaminidase Brotease

28.4 51.8 29.5

39.5

52.8 47.0 18.8

7390 a79 20.0

X:! 9 410

6800

1 085 1 205 20.0 7.8

Glycosidases and acid phosphatase are expressed as nmole p-nitrophenol released/h/ml serum and protease as mg equivalents of pronase/h/ml serum. A, B and C represent different batches of fetal calf serum, A and B from Grand Island Biological Co. and C-from Microbiological Associates.

RESULTS Enzymatic analyses of commercially available fetal calf serum from two Companies revealed a spectrum of strongly active glycosidases, acid phosphatase and non-specific protease activity (table 1). These enzymes are normally housed in lysosomes and are presumably contaminants in the serum. Two of the enzymes monitored showed surprisingly high activity. For example, the specific activity of [email protected] for serum from both Companies, when expressed on a specific activity basis rather than as activity per ml of serum, ranged from 260-380 nmol p-nitrophenol released per mg protein per hour, yet reported figures for corresponding activities of the same enzyme in homogenates of BHK and pyBHK cells are 233 and 280, respectively [4]. Values for the specific activity of N-acetyl$-D-galactosaminidase are also comparable to those reported previously for cell homogenates [4]. a-D-Glucosidase, which was not detectable in BHK cell homogenates, was at least measurable in fetal calf serum. In addition serum from both Companies featured non-

Figs I, 2. Abscissa: cell no./cm2; ordinate: time (hours). Growth curves for (fig. 1) BHK cells, and (fig. 2) pyBHK cells, in medium supplemented with untreated and heat-inactivated fetal calf serum (Microbiological Associates). BME supplemented with o, 10% untreated serum; x , 10 % heat-inactivated Exptl Cell Res 92 (I975)

90

Elligsen, Thompson and Frey

Table 2. Acid phosphatase

and glycosidase

activities of cellular origin in conditioned growth

medium

Enzyme

Expt

Medium from BHK cells nmole p-nitrophenol/cell/h

Acid phosphatase

A B C A B C A B C A B C A B

7.8 ND 34.0 12.7 ND 2.1 191 111 393 26.9 39.5 54.4 25.4 33.2

a-n-Glucosidase N-Acetyl-p-nglucosaminidase N-Acetyl-&ngalactoseaminidase p-n-Galactosidase

Medium from pyBHK cells nmole p-nitrophenol/cell/h 31.8 21.6 50.9 7.8 ND ND 435 375 524 50.2 49.0 65.7 52.3 71.3

Medium for both BHK and pyBHK cells contained 10 % heat-inactivated serum from Microbiological and was conditioned for 48 h; ND, not detectable.

specific proteolytic activity which, on the basis of four separate analyses, was consistently higher in the serum from Grand Island Biological Co. than in that from Microbiological Associates. Measurement of this activity using Azocoll as a substrate was possible despite the fact that the serum, added to the assay mixture as a source of enzyme, would also serve as substrate and tend to competitively inhibit Azocoll hydrolysis. Thus the proteolytic activities are almost certainly underestimations of the true protease level in serum. To elucidate any contribution by the hydrolytic enzyme activity of fetal calf serum to its growth supporting properties, conditions which would inactivate the enzymes were sought. It was found that all six enzymes being monitored were rendered completely inactive by heating serum to 70°C for 15 min. Comparisons of the growth sustaining properties of media supplemented with untreated and heat-inactivated serum Exptl

Cell Res 92 (197.5)

Associates

are illustrated for both BHK and pyBHK cells in figs 1 and 2 respectively. Heatinactivated and untreated serum supported growth of the transformed cells equally well, but for normal cells the heat-treated serum was marginally less effective at sustaining growth in that it gave rise to a slightly lower density saturation. The growth plateau seemingly exhibited by the transformed cells at confluence reflects a high incidence of cells falling off into the medium. In experiments designed to evaluate the extent to which lysosomal enzymes leak into the growth medium from cells, fetal calf serum was heat-inactivated prior to its use as a supplement in order to increase the probability of detecting leakage. In addition, batches of conditioned media were concentrated about 30-fold by lyophilization before being assayed. With these provisions it was possible to measure hydrolytic enzyme activity in medium conditioned by growth of either normal or transformed cells, al-

Lysosomaf enzyme leakage

91

Table 3. Relative levels of acidphosphatase and glycosidases indigenous to unconditioned medium and of cellular origin in conditioned medium Activity/ml of medium

Enzyme Acid phosphatase a-D-Ghtcosidase

Control medium 47.9 36.0 34.9

i.1: 4:o

N-Acetyl-B-D-glucosaminidase

613 849 672

[email protected] p-D-Galactosidase

79.6 103.8 103.6

17.0 26.5 28.6

Heat-treated, conditioned medium BHK cells ND 0.4 1.8

pyBHK cells 0.4 1.5 2.6 0.4

It:

0.1

Fig

8.7 3.1 21.2

2.1 18.1 26.7

0.3 1.2 2.9

0.8 2.3 3.3

1.2 0.2 1.8

2.4 1.2 3.6

Glycosidase and acid phosphatase activities are expressed as nmol p-nitrophenol/h/ml of medium; control medium, that which has been neither heat-treated nor conditioned by cell growth; heat-treated conditioned medium, that which has been heated to inactivate indigenous enzyme activity and then conditioned by 48 h of either BHK or pyBHK cell growth; ND, not detectable.

though not all of the enzymes monitored were consistently detectable among experiments (table 2). Since any enzymes innate to the medium had been previously inactivated, this hydrolytic activity could only have been derived from the growing cells, presumably by leakage, although the prospect of some cell lysis is not precluded. With the exception of a-D-glucosidase the activity per cell was consistently higher within experiments in the conditioned medium from transformed cells than in that from normal cells, sometimes by more than 3-fold (table 2). Moreover there were marked differences in the levels of individual enzymes in the medium. Activities for N-acetyl-j3-D-glucosaminidase averaged 232 nmole of p-nitropheno1 released per cell/h for normal cells and 445 nmole/cell/h for transformed cells, yet corresponding values for acid phosphatase,

cc-D-glucosidase, /?-D-galactosidase and Nacetyl-#?-D-galactosaminidase were consistently less than 75 nmole of p-nitrophenol released per cell/h. Protease activity was not detectable in conditioned medium from either transformed or normal cells using Azocoll as a substrate. A significant derivation from this information is that levels of the enzymes in heatinactivated, conditioned medium are markedly less than those contributed normally by fetal calf serum. This phenomenon is clearly evident from table 3 which compares the relative concentrations of glycosidase and acid phosphatase activities for control medium and heat-inactivated, conditioned medium. Control medium had been neither heat-treated nor conditioned and thus could only have manifested indigenous enzyme activity. Any enzymatic activity in the heatExptl Cell Res 92 (1975)

92

Elligsen, Thompson and Frey

inactivated, conditioned medium can be attributed to leakage from growing cells. All enzymes measured were 10 to 100 times more concentrated in the control medium than in the heat-treated, conditioned medium irrespective of whether the latter had been conditioned by normal or transformed cells.

fluent 3T3 cells with pronase does not stimulate cell division [17]. Moreover the plasma membranes of certain transformed and tumour cells have been found to be leaky to enzymes. Holmberg [13] working with ascites tumour cells in suspension and Sylvtn [18] with tumour fluids of animals observed soluble and lysosomal enzymes in the cell suspension media and interpreted their data DISCUSSION as reflecting leakage rather than cell lysis. It has been proposed that enzymes of lysoso- However, it has also been reported that leakma1 origin may serve as agents rendering age of macromolecules from transformed architectural modifications on the surfaces chicken cells is no greater than that from of transformed cells which in turn give rise corresponding normal cells and amounts to uncontrolled growth [3, 12, 13, 18, 221. to less than 3% during a 12 h period [19]. This proposal is indirectly supported by the Moreover it must be recognized that lysosoobservation that elevated glycosidase, acid ma1 glycosidases would be substantially less phosphatase and protease levels accompany active at cell surfaces than in their native cell transformation mediated by either DNA milieu in view of their low pH optima. In or RNA tumour viruses [2, 31. Glycosidases, any case there is at least indirect evidence enzymes which are normally housed in to support the theory that enzymes of lysolysosomes, are capable of cleaving bonds somal origin may be partially responsible between monosaccharides, or between a for the distinctive growth behaviour of monosaccharide and some other moiety transformed cells by reason of their capabilin glycoproteins, glycolipids, glycosaminoity for modifying the cell surface. The spectrum of lysosomal enzymes, other glycans and oligo- and polysaccharides. Consequently, were such enzymes to gain than protease, selected for examination in access to the cell surface, they would certainly the current study comprised those prebe capable of rendering alterations, perhaps viously shown to have substantially higher to the point of accounting for the different activity in pyBHK cell homogenates than growth modes of normal and transformed in those for BHK cells [4]. Measurable cells if accessibility to the cell surface were activities were present in medium conditioned unique or even preferential as a result of by growth of either normal or transformed transformation. Indeed, breakdown of glyco- cells, although with the exception of Q-Dproteins and glycolipids on the surface glucosidase, the glycosidases monitored showmembrane of density-inhibited normal cells ed consistently higher released activity on a by neuraminidase stimulates cell division per cell basis within experiments for trans[21]. It has also been observed that protease formed cells. Bosmann was unable to detect action on the cell surface endues normal cells cc-D-glucosidase in BHK cell homogenates with a sufficiently altered surface architecture and the enzyme showed only minimal activity to enable escape from contact inhibition of in homogenates of pyBHK cells [4]. This growth [8], although this observation is very low activity of the enzyme in the cells controversial in the light of a more recent themselves doubtless contributed to our report to the effect that treatment of con- inability to detect significant differences in Exptl Cell Res 92 (1975)

Lysosomal

its released activity between medium conditioned by normal cells and that conditioned by transformed cells. While it is possible that the plasma membrane of pyBHK cells is leakier than its counterpart for normal cells, it seems likely that the higher activities released from transformed cells also reflect higher innate concentrations of the enzymes in the cells themselves. Protease activity was not detectable in the heat-inactivated medium conditioned by either normal or transformed cell growth presumably because the assay procedure was insufficiently sensitive to detect low levels of activity. However increased concentrations of proteases have been found in a number of virally transformed cells compared with their normal counterparts [2, 3, 61 and release of a specific protease during M phase of the mitotic cycle has been reported for murine leukaemic cells [5]. Moreover virally transformed cells exhibit fibrinolytic activity that is undetectable in corresponding cultures of normal cells [19, 201. Glycosidase and even protease activities derived from growing cells are relatively low in comparison with levels of the same enzymes indigenous to the growth medium by reason of its fetal calf serum component. Glycosidase activity in the medium itself was from 10 to 100 times greater than that released by the cells. Measurable levels of protease were present in control medium, yet levels of protease released into the medium by either normal or transformed cells were too low to be detected by the assay procedure used. Indeed for two of the enzymes monitored, N-acetyl-/?-D-glucosaminidase and Nacetyl-/?-D-galactosaminidase, the specific activities in commercially available serum were of comparable magnitude to those previously reported for cell homogenates of BHK and pyBHK cells [4]. In addition it would appear that the high glycosidase and

enzyme leakage

93

protease activities indigenous to tissue culture medium do not overtly influence cell growth. BHK cell growth in medium containing heatinactivated fetal calf serum was only marginally inferior to that in medium supplemented with untreated serum. We suspect this probably reflects a general deterioration of the fetal calf serum brought on by the heat treatment rather than inactivation of its enzyme activity. Indeed, no differences were apparent at all when the growth behaviour of pyBHK cells was similarly examined, although this is doubtless attributable at least in part to the fact that the growth requirement of transformed cells for serum is generally less stringent than for normal cells [lo, 11, 151. Taken at face value, the observation that levels of degradative enzymes indigenous to the growth medium are markedly higher than those derived from the cells by leakage is inconsistent with the proposal [3, 161 that the transformed state is maintained through sublethal autolysis. However, it is possible that the serum glycosidases act largely on serum glycoproteins rather than on cell surfaces, a prospect which is supported by the observation that both normal and transformed cells grow equally well in either heat-inactivated or untreated serum. Granted this contingency, the substantially higher levels of external glycosidase activity derived from transformed as compared with normal cells assume greater significance, for these cellular glycosidases could well be acting on surface membrane glycoconjugates during their release and thus rendering a more extensive degradation of the transformed cell surface than of the normal cell surface, Viewed in this context the higher levels of glycosidases released from the transformed cells constitute supporting evidence for the involvement of sublethal autolysis in maintenance of the neoplastic state. Exptl Cell Res 92 (1975)

94

Elligsen, Thompson and Frey

The authors gratefully acknowledge a grant-in-aid of this research from the Medical Research Council of Canada. They are also indebted to Mrs Margaret Gruber for expert technical assistance, to Dr Jack Kruuv for critically reviewing the manuscript and to Dr C. Basilica, New York University Medical School, for donating the cell lines used in the study.

13. 14. 15. 16. 17.

REFERENCES 1. Bissell, M J, Rubin, H & Hatie, C, Exptl cell res 68 (1971) 404. 2. Bosmann, H B, Exptl cell res 54 (1969) 217. 3. - Biochim biophys acta 264 (1972) 339. 4. Bosmann, H B & Pike, G Z, Life sci US 9 part II (1970) 1433. 5. Bosmann, H B, Nature 249 (1974) 144. 6. Bosmann, H B, Lockwood, T & Morgan, H, Exptl cell res 83 (1974) 25. 7. Burger, M M, Proc natl acad sci US 62 (1969) 994. 8. - Fed proc 32 (1973) 91. 9. Calbiochem (1972) Document No. 3805. 10. Clarke, G D, Stoker, M G P, Ludlow, A & Thornton, M, Nature 227 (1970) 798. 11. Dulbecco, R, Nature 227 (1970) 802. 12. Dingle, J T, Lysosomes in biology and pathology

Exptl Cell Res 92 (1975)

18.

19. 20. 21. 22.

(ed J T Dingle & H B Fell) vol. 2, p. 241. NorthHolland, Amsterdam (1969). Holmberg, B, Cancer res 21 (1961) 1386. Lowry, 0 H, Rosebrough, N J, Farr, A L & Randall, R J, J biol them 193 (1951) 265. Paul, D, Lipton, A & Klinger, I, Proc natl acad sci US 68 (1971) 645. Poste, G, Exptl cell res 67 (1971) 11. Glynn, R D, Thrash, C R & Cunningham, D D, Proc natl acad sci US 70 (1973) 2676. Sylven, B, Endogenous factors influencing hosttumour balance (ed R W Wissler, T L Dao & S Wood) p. 267. Univ Chicago Press, Chicago, Ill. (1967). Unkeless, J C, Tobia, A, Ossowski, L, Quigley, J P, Rifkin, D B & Reich, E, J exptl med I37 (1973) 85. Ossowski, L, Unkeless, J C, Tobia, A, Quigley, J P, Rifkin, D B & Reich, E, J exptl med 137 (1973) 112. Vaheri, A, Ruoslahti, E & Nordling, S, Nature new biol 238 (1972) 211. Wallach, D F H, The plasma membrane: dynamic perspectives genetics and pathology, p. 116. English Universities Press Ltd., London (1972).

Received October 31, 1974

An evaluation of lysosomal enzyme leakage as a factor influencing the behaviour of transformed cells.

Printed in Sweden Copyright Q 1975 by Academic Press, Inc. All rights of reproduction in any form reserved Experimental Cell Research92 (1975) 87-94...
641KB Sizes 0 Downloads 0 Views