monensin and tunicamycin-induced inhibition of ht29 cell spreading ...

J. Cell Sci. SO, 269-280 (1986) Printed in Great Britain © The Company of Biologists Limited 1986

269

MONENSIN AND TUNICAMYCIN-INDUCED INHIBITION OF HT29 CELL SPREADING AND GROWTH A. EL-BATTARI, J. M. MULLER, J. FANTINI, F. BELLOT, A. TIRARD, F. D U C R E T A N D J . MARVALDI Laboratoire de Chimie Biologique (UA. n° 202), Universite d'Pax-Marseille I, place V. Hugo, F-13331 Marseille cedex 3, France

SUMMARY HT29 cells originating from a human colon adenocarcinoma, spread very rapidly after seeding on their own extracellular matrix (ECM). Preincubation of cells with the inhibitor of protein glycosylation tunicamycin (TM) or with the ionophore monensin substantially suppressed cell spreading in serum-free medium without affecting cell adhesion to ECM. Addition of the drugs after cell attachment and spreading inhibited cell growth. TM-treated cells remained viable after 6 days of exposure to 2/igml~' of TM and resumed their normal growth rate and shape after removing the drug from the medium. On the contrary, monensin inhibition of cell growth was not reversible: after 3 days, cells detached from the ECM and were unable to exclude Trypan Blue. At the ultrastructural level, a swollen Golgi apparatus with numerous vacuoles was observed after treatment for 2h in either TM or monensin-preincubated cells. These results suggest that TM and monensin interfere with the insertion and, or, function of one or more cell surface glycoproteins, possibly interacting with cytoskeleton and involved in cell spreading and growth.

INTRODUCTION

Cell-substrate adhesion is generally considered to be a multistep process involving recognition of extracellular matrix components by cell surface receptors, followed by cytoskeletal rearrangements that lead to cell spreading (Grinnel, 1978; Hynes, 1981). Cell spreading can be performed on surfaces coated with various attachment factors such as fibronectin, laminin, collagen and vibronectin, presumably through different cell surface receptors (for a review, see Roos, 1984). One or more of these molecules together with many others that remain to be identified are known to constitute the extracellular matrices (ECMs) secreted by cells to promote their own spreading. Interference of drugs with metabolic pathways is an interesting approach towards understanding the role of certain molecules in cellular properties such as adhesion, spreading and growth. Here we report that spreading of HT29 cells, a malignant epithelial cell line originating from human colon, was obtained very rapidly in serumfree medium when cells were seeded on ECM secreted by HT29 cells themselves. The antibiotic tunicamycin (TM) widely used as an inhibitor of protein iV-glycosylation (Elbein, 1981; Olden, Bernard, White & Parent, 1982) was found to prevent Key words: tunicamycin, monensin, HT29 cells, extracellular matrix

270

El-Battari and others

cell spreading of HT29 on their own ECM. The monovalent ionophore monensin, known to inhibit the intracellular transport of secretory or membrane proteins (Rotundo & Fambrough, 1980) and to disrupt the Golgi complex (Tartakoff, 1983; Tartakoff & Vassali, 1977) was also shown to prevent HT29 cell spreading on ECM. Neither TM nor monensin was an inhibitor of cell adhesion. We suggest that these effects on cell spreading resulted from impaired synthesis and, or, secretion of cell surface iV-glycosyl-containing proteins, or from insertion of carbohydrate-poor glycoproteins within the cell membrane. These glycoproteins (receptors) could interact directly or indirectly with the cytoskeleton. MATERIALS AND METHODS

Chemicals Tunicamycin was purchased from Boehringer (Mannheim, W. Germany). Monensin was from Calbiochem-Behring Corp. (La Jolla, CA, U.S.A.). D-[2-3H]mannose (19-5 CimmoP 1 ) and aqueous counting scintillant (ACS) were from Amersham France S.A. L-[3sS]methionine (1098 Ci mmol" 1 ) and 51chromium (25mCi) were obtained from New England Nuclear (Boston, MA, U.S.A.). Dulbecco's Modified Eagle's Medium (DMEM) and foetal calf serum (FCS) were purchased from Eurobio. Phosphate-buffered saline (PBS) was from Oxoid.

Cell culture The human colonic adenocarcinoma cell line HT29 (Fogh & Trempe, 1975), a gift from Dr A. Zweibaum (Paris), was routinely cultured in DMEM containing 25 mM-glucose and 10% foetal calf serum, in a humidified atmosphere of 95% air/5% CO2 with a medium change every 2 days.

Extracellular matrix preparation Extracellular matrix was prepared as previously described elsewhere (Bellote/ al. 1985). Briefly, a monolayer of HT29 cells grown in 24-well culture plates was washed twice with PBS and incubated for lOmin with 0-5% Triton X-100 in PBS. The cell layer was removed by aspiration and the culture flasks were carefully rinsed four times with PBS. Extracellular matrices (ECM) were kept at —20°C for at least 1 month without any loss of efficiency.

Exposure of cells to inhibitors Stock solutions of monensin and TM were prepared in absolute ethanol to a concentration of 6 i 5 m g m l " ' and lOmgml""1, respectively. The final concentration of ethanol, in all experiments, never exceeded 0*1 %. In all cases, cell suspensions were divided into samples that contained either monensin (0-07-3-5/ig ml" 1 ) or TM (0-05-2^gml~') or the corresponding concentrations of ethanol (control). Unless indicated otherwise, cell monolayers were exposed to inhibitors in serumsupplemented medium (10% FCS) 2h at 37°C in a humidified atmosphere of 95% air/5% CO2 before the spreading assay, isotope incorporation or cell viability determination. Cell viability was based on Trypan Blue exclusion (Philips, 1973) and on the chromium release assay (Brunner, Mauel, Cerottini & Chapuis, 1968).

Cell spreading assay All cultures used were grown to near confluence, washed in 10% FCS-supplemented medium and incubated in the presence or absence of inhibitors as specified above. Cells were then harvested with 0-05% trypsin in PBS containing 0-53 mM-EDTA, washed and resuspended at a final concentration of 5X10 4 cells per ml in serum-free medium, in the presence or absence of inhibitors. Cells were seeded in 24-well culture plates containing ECM and allowed to attach and spread during 4h. Fifty to eighty cells were counted in six different microscopic fields. The non-spread cells were clearly distinguishable from the spread ones since they are perfectly spherical.

Tunicamycin and monensin inhibition ofHT29 cells

271

Cell growth For the measurement of the population doubling time, cells were suspended in 10% FCSsupplemented medium and seeded in 24-well culture plates containing ECM (SxlO 4 cells per well). Cells were allowed to attach and spread overnight without inhibitors. TM (2//gml~') and monensin (1-5 figml"1) were then added and culture continued for 6 additional days. Media were changed every day. The effects of drugs were determined by observing cell morphology and by recording the cell number at indicated days.

Electron microscopy Control and treated cells were harvested from culture plates with 0'53mM-EDTA in PBS, peletted and rinsed three times with PBS. Pellets were fixed with 2-5% glutaraldehyde in 0-2Msodium cacodylate buffer (pH 7-4) with 7-5% saccharose, post-fixed in 1% osmium tetroxide in 0-2M-sodium cacodylate buffer for 1 h, dehydrated in ethanol and embedded in Epon.

Radioactive labelling Control and treated cells were washed three times with PBS and incubated 3 h in serum-free medium containing isotopes and 2/igml" 1 TM or 1-5 (ig ml" 1 monensin. For labelling experiments with [ 3s S]methionine, cell monolayere were incubated in Eagle's minimal essential medium lacking methionine but containing 0-2% bovine serum albumin (BSA), and supplemented with 25 flQA [35S]methionine and inhibitors as indicated above. For labelling with pHJinannose, cells were incubated in DMEM containing O56mM-glucose (low glucose medium), 0-2% BSA and 10/jCi [ 3 H]mannose. After incubation with isotopes cells were washed six times with cold PBS, dissolved in 0-1 M-NaOH and the incorporation of radiolabelled methionine or mannose into proteins was determined as 10% trichloroacetic acid-precipitable material. The precipitates were rinsed three times and dissolved in lM-NaOH. The radioactivity was determined by liquid scintillation spectrometry using aqueous counting scintillant (ACS). Protein contents were determined by the method of Lowry, Rosebrough, Fair & Randall (1951).

RESULTS

Spreading ofHT29 cells on ECM HT29 cells in serum-free medium were totally unable to spread and grow when seeded on the plastic culture dishes. We tested for the possibility that HT29 cells secreted their own extracellular matrix (ECM), which could permit cells to spread in an optimal way in serum-free medium. Monolayere of HT29 cells were removed by Triton X-100 and the culture dishes were washed thoroughly. We failed to see anything left on the plastic under optical-microscopic observation. When HT29 cells were seeded in serum-free medium on these ECMs, they spread very rapidly (75%, 4 h after seeding) (Table 1). Control of cell viability HT29 cells were exposed to TM or monensin for 2h at 37 °C in serumsupplemented medium. The exclusion of Trypan Blue or the release of radiolabelled chromium was measured. The results obtained by the two techniques were in good agreement. No cytotoxicity wa9 observed for the two drugs at the concentrations used in this study, i.e. below 2/igmP 1 TM and 3-S/ig ml monensin (Fig. 1).


272

Effect of TM and monensin on [3SS]methionine and [3H]mannose incorporation Addition of TM (2/igmP 1 ) or monensin (1-5 jigml" 1 ) to the cell culture in serumfree medium induced an inhibition of [ H]mannose incorporation (Table 2) of, respectively, 51-5% and 59-6% versus control. This inhibition cannot be due to an inhibition of sugar transport because the uptake of radioactive precursors was not affected by the drugs (data not shown). The incorporation of [3SS]methionine into proteins was slightly decreased by the two drugs. On the contrary 55% of inhibition of methionine incorporation was obtained in the presence of O-ljigmF 1 of cycloheximide, a well-known inhibitor of protein synthesis (Table 2). These results Table 1. Spreading velocity ofHT29 cells on their extracellular matrix Time (h)

% HT29 spread cells

1 2 3 4 20

15 40 60 75 90

The average percentage of spread cells was calculated after counting six different microscopic fields.

13

o

3

10

20

30

50 Drug (fig ml"1)

10

20

30

Fig. 1. Cell mortality of HT29 cells versus drug concentrations. After 2h of preincubation with TM ( • — • ) or monensin (O—O) cell mortality was checked by the Trypan Blue exclusion method (A) and by chromium release (B). Arrows indicate the maxima] doses of TM and monensin used in this study.


273

indicate that at the concentrations tested in this study neither TM nor monensin had an important inhibitory effect on protein synthesis. Effect of TM and monensin on cell spreading Well-dissociated HT29 cells were seeded on ECM after incubation with or without various amounts of drugs. The results are shown in Fig. 2A,B. The two drugs prevented the spreading of cells in a dose-dependent manner. The morphology of control and drug-treated cells is shown in the micrographs of Fig. 3. In order to determine whether a reduction in protein synthesis was responsible for cell spreading Table 2. Incorporation of methionine and mannose in trichloroacetic acid-insoluble fractions in the presence of drugs L-[35S] methionine (cts min" 1 jjg protein"'xlO~ 3 )

Cells Control Tunicamycin (2/4gml l) : Monensin (1-5/igmF •) Cycloheximide (0-1 mjml ')

% Inhibition

D-[2-3H] mannose (cts min" /ig protein" 1 X10" 3 ) 97

188 163

147 84-S

13-3 21-8 55-0

47

39-2 78-6

0-5 Tunicamycin

% Inhibition

1 Monensin

51-5 59-6 19-0

1-5

Fig. 2. Cell spreading of HT29 cells on their own extracellular matrix as a function of TM ( • — • ) or monensin (A—A) concentration. Cells were incubated for 2 h with drugs at indicated concentrations in 10% FCS-supplemented DMEM, then harvested, rinsed and further seeded on ECM in serum-free medium containing drugs at the same concentrations. After an additional incubation of 4h spread cells were counted. Each experimental value represents the percentage of spread cells compared with the control. Values are the mean of six different cell counts in different microscope fields. Bars indicate S.E.M.

3-5

274


O o

o°o


275

inhibition we tested the effect of cycloheximide on HT29 cells. At a concentration of 0-1 jUgml"1 cycloheximide, which inhibited 55% of methionine incorporation, there was no change in cell adhesion, spreading or morphology (data not shown). Effect of prolonged exposure to inhibitors on cell morphology and growth Addition of TM (2jug ml~') 24 h after seeding produced a marked inhibition in the growth rate of HT29 cells (Fig. 4A) without cell death, in either sparse or dense culture after 6 days of exposure to the drug. When TM was removed from the medium, HT29 cells resumed their normal growth rate. Addition of monensin (1'5/Jgmr 1 ) also prevented cell growth but, unlike the TM effect, the inhibition

.2

50 -

Fig. 4. Effect of chronic treatment with 2 fig ml ' TM (A) or monensin 1-5 fig ml ' (B) on HT29 cell growth. Cells were grown in DMEM supplemented with 10% FCS and treated with drugs (arrows pointing down). After 4 days of incubation, drug-containing medium was replaced by fresh medium (arrows pointing up). Cells were counted in the presence of Trypan Blue. (O—O) Growth in the absence of the drugs; ( • — # ) growth in the presence of drugs. Fig. 3. Micrographs of cells after drug treatment. Untreated HT29 cells (A), TM- (B) and monensin-treated cells (c). After 2 h of preincubation with drugs, cells were seeded in serum-free medium and allowed to attach and spread on ECM for 4 h.

276


was not reversible (Fig. 4B) and after 3 days of exposure cells detached from the ECM; 70% of the detached cells were unable to exclude Trypan Blue and to spread again after seeding in monensin-free medium on fresh ECM. On the contrary, TMtreated cells remained attached to the substratum, and between 4 and 6 days they underwent striking morphological changes as shown in Fig. 5. Ultrastructure ofHT29 cells after treatment vrith TM and monensin In order to control the effects of TM and monensin at the ultrastructural level, cells were observed by electron microscopy after 2h of exposure to the drugs. TMand monensin-treated cells conserved the normal morphology of all the visible structures except the Golgi apparatus, which appeared distended and surrounded by swollen vacuoles (Fig. 6). DISCUSSION

HT29 cells attach and spread very rapidly (3—4h) when seeded on their own extracellular matrix (ECM) in serum-free medium. Exposure of cells, in the absence

Fig. 5. TM-treated HT29 cell after 4—6 days of exposure to the drug (A). Note the striking change of morphology compared to control cells (B). Fig. 6. Electron micrographs of HT29 cells. Golgi apparatus (G), nucleus (w), vacuoles (v) and intermediate filaments (if). A. Control cell with stacked G and sparse vacuoles; X5000. B,C. TM- and monensin-treated cells showing swollen G surrounded by numerous vacuoles; x 17000.

Tunicamydn and monensin inhibition of HT29 cells

278


of FCS, to TM or monensin resulted in a dose-dependent inhibition of cell spreading (see Fig. 2). Within the concentration range used in these experiments, neither TM nor monensin had an effect on cell adhesion to their own ECM. Inhibition of cell spreading induced by the two drugs was observed at concentrations that slightly decreased protein glycosylation, as shown by the reduction of [3H]mannose incorporation. According to the results of methionine incorporation experiments there was no significant effect on protein synthesis. It is worth noting that cycloheximide, a well-known protein synthesis inhibitor, at the concentration that prevented 55% methionine incorporation, had no effect on cell spreading. We suggest that inhibition of cell spreading induced by TM was only due to the impaired protein glycosylation within HT29 cells. The incomplete inhibition of [3H]mannose incorporation by TM may reflect: (1) a metabolic conversion of mannose to fucose, which retains 3H label and is very poorly incorporated in glycoproteins of HT29 cells (unpublished observation); (2) the addition of sugars to the ./V-acetylglucosamine—lipid carrier, which is unaffected by TM (Tkacz & Lampen, 1975; Lehle & Tanner, 1976). The data from this inhibition study indicate that ECM receptors, which are involved in cell spreading, contain substantial amounts of N-linked carbohydrates. The inhibitory effect of monensin on cell spreading could be explained by the inhibition of the addition of terminal sugars (galactose, fucose, sialic acid) to mannose-containing oligosaccharide chains (Alonso-Caplen & Campans, 1983). Interpretation of the monensin inhibition mechanism is complicated by the multiple effects of the drug, especially on intracellular transport (Tartakoff, 1983). Neither TM nor monensin seems to have an effect on cell adhesion to ECM. As described by others, some findings suggest that the molecules involved in cell adhesion to the substratum are O-glycoproteins or ganglioside-like structures. In fact, studies on fibronectin-cell interaction have shown an inhibitory effect of di- and tri-sialogangliosides on the attachment of cells to fibronectin (Kleinman, Martin & Fishman, 1979). The inhibitory activity was found to reside in the carbohydrate moiety of the glycolipid. The cytotoxicity of TM and monensin was checked carefully by two different techniques: the Blue Trypan exclusion method and the 5'chromium release technique. Within the concentration range of TM or monensin used in this study, there was no cytotoxicity toward HT29 cells and, hence, inhibition of cell spreading cannot be explained by the toxicity of the drugs. The effect of TM or monensin on HT29 cell growth was also investigated in order to demonstrate cell viability after long-term exposure to the drugs. Our results demonstrate that TM (2/igml" 1 ) inhibited cell growth. This inhibition was fully and rapidly reversible after removal of TM. The inhibition of cell growth by TM might be the result of the loss of growth factor receptors, as demonstrated for epidermal growth factor (EGF) receptor (Bhargava & Makman, 1980) and insulin receptor (Ronnet & Lane, 1981; Jacobs, Kull & Cuatrecasas, 1983). On the other hand, monensin (1-5 /igml~ ), which also inhibits cell growth, was cytotoxic after exposure


279

of the cells to the drug for a long period, as shown by the non-reversibility of its effect. Finally, we have observed the morphology of HT29 cells after TM or monensin treatment, by electron microscopy. The Golgi apparatus was distended in TMtreated cells as already observed (Kuo & Lampen, 1974). Swollen vacuoles derived from the Golgi apparatus were clearly visible after 2h of treatment with monensin. This characteristic of monensin has been described elsewhere (Geisow & Bourgoyne, 1982). In conclusion, the results reported here demonstrate the implication of Nglycoproteins in the spreading process. The structure of the cell surface receptor(s) that interact with ECM is now under investigation and we have already selected a monoclonal antibody that prevents the spreading of HT29 cells on their own ECM. This work was supported in part by the CNRS (UA 202) and by a grant from la F6d6ration Nationale des Centres de Lutte Contre le Cancer (F.N.C.L.C.C). We thank Miss J. Secchi for technical assistance and Mrs J. Charpentier for carefully preparing the text. REFERENCES ALONSO-CAPLEN, F. V. & CAMPANS, R. W. (1983). Modulation of glycosylation and transport of viral membrane glycoproteins by a sodium ionophore. J. Cell Biol. 97, 659-668. BELLOT, F., PICHON, J., LUIS, J., E L BATTARI, A., SECCHI, J., CAU, P. & MARVALDI, J. (1985).

Extracellular material secreted by human colonic adenocarcinoma cell lines (HT29 and HRT18) promotes their own very rapid spreading in serum free medium and induces neurite outgrowth of PC 12 cells. Int. J. Cancer (in press). BHARGAVA, G. & MAKMAN, M. H. (1980). Effect of tunicamycin on the turnover of epidermal EGF receptors in cultured calf aorta muscle cells. Biochim. biophys. Acta 629, 107—112. BRUNNER, K. T., MAUEL, J., CEROTTINI, J.-C. & CHAPUIS, B. (1968). Quantitative assay of the

lytic action of immune lymphoid cells on 51Cr labeled allogenic target cell in vitro. Inhibition by . isoantibody and by drugs. Immunology 14, 181. ELBEIN, A. D. (1981). The tunicamycins - useful tools for studies on glycoproteins. Trends Biochem. Sci. 6,219-221. FOGH, J. & TREMPE, G. (1975). Human tumor cellsm vitro. lnHuman Tumor Cells in vitro (ed. J. Fogh), pp. 115-159. New York: Plenum Publ. Corp. GEISOW, M. J. & BURGOYNE, R. D. (1982). Effect of monensin on chromaffin cells and the mechanism of organelle swelling. Cell Biol. Int. Rep. 6, 933-939 GRINNEL, F. (1978). Cellular adhesiveness and extracellular substrata. Int. Rev. Cytol. 53, 65-144. HYNES, R- D. (1981). Relationships between fibronectin and the cytoskeletal. In Cell Surface Reviews (ed. G. Poste & G. L. Nicolson), vol. 7, pp. 97-136. Amsterdam: Elsevier/North Holland. JACOBS, S., KULL, F. JR & CUATRECASAS, P. (1983). Monensin blocks the maturation of receptors for insulin and somatomedin C: Identification of receptors precursors. Proc. natn. Acad. Sci. U.SA. 80, 1228-1231. KLEINMAN, H. K., MARTIN, G. R. SCFISHMAN, P. H. (1979). Ganglioside inhibition of fibronectin

mediated cell adhesion to collagen. Proc. natn. Acad. Sci. U.S A. 76, 3367—3371. Kuo, S. C. & LAMPEN, J. O. (1974). Tunicamycin an inhibitor of yeast glycoprotein synthesis. Biochem. biophys. Res. Commun. 58, 287-295. LEHLE, L. & TANNER, W. (1976). The specific site of tunicamycin inhibition in the formation of dolichol-bound ./V-acetyl glucosamine derivatives. FEBS Lett. 71, 167-170. LOWRY, O. H., ROSEBROUGH, N. J., FARR, A. L. & RANDALL, R. J. (1951). Protein measurement

with the Folin phenol reagent. J. biol. Chem. 193, 265-275. OLDEN, K., BERNARD, B. A., WHITE, S. L. & PARENT, J. B. (1982). Function of the carbohydrates

moieties of glycoproteins..?. Cell Biochem. 18, 313-335.

280


PHILIPS, H. J. (1973). Dye exclusion test for cell viability. In Tissue Culture Methods and Applications (ed. P. F. Kruse, Jr & M. K. Patterson, Jr), pp. 406-408. New York: Academic Press. RONNET, G. V. & LANE, M. D. (1981). Post-transitional glycosylation induced activation of aglycoinsulin receptor accumulated during tunicamycin treatment. J. biol. Chem. 256, 4704-4707. Roos, E. (1984). Cellular adhesion, invasion and metastasis. Biochim. biophys. Acta 738, 263-284. ROTUNDO, R. L. & FAMBROUGH, D. M. (1980). Secretion of acetylcholinesterase: relation to acetylcholine receptor metabolism Cell 22, 595-602. TAKTAKOFF, A. M. (1983). Perturbation of vesicular traffic with the carboxylic ionophore monensin. Cell 32, 1026-1028. TARTAKOFF, A. M. & VASSALI, P. (1977). Plasma cell immunoglobulin secretion: arrest is accompanied by alteration of the Golgi complex. J. exp. Med. 145, 1332-1345. TKACZ, J. S. & LAMPEN, J. O. (1975). Tunicamycin inhibition of polyisoprenyl N-acetylglucosaminyl pyrophosphate formation in calf liver microsomes. Biochem. biophys. Res. Comrnun. 65, 248-257.

(Received22 July 1985 -Accepted 2 September 1985)