Evaluation of the Effect of Suramin on Neural Cell Growth and N-CAM Expression1

Home , Suramin

(CANCER RESEARCH 30, 5164-5170, August 15. 1990] Evaluation of the Effect of Suramin on Neural Cell Growth and N-CAM Expression1

Xiao-Jun Guo, Jacques Fantini, Regine Roubin, Jacques Marvaldi, and GeneviÃ¨veRougon2

URA 202 CNRS, institut de Chimie Biologique, 3, Place V. Hugo, 13331 Marseille Cedex 3 Â¡X-J.G., J. F., J. M., G. K.J. and CIML, Case 906, 70, Route L. Lachamp, 13288 Marseille Cedex 9 [R. R.Â¡,France

ABSTRACT in vitro as well as on brain cell primary cultures to locate potential cellular targets of this toxicity. We report that, while Suramin, a polysulfonated naphthylurea, is currently under investiga suramin inhibits astrocytoma cell growth, it exhibited cytotoxic tion for treatment of advanced malignancy and has been shown to exhibit antiproliferative effects on some cells. We investigated its action on two effects on neuroblastoma cells and neurons. To shed some light cell lines of neural origin, one with neuronal (N2A) and the other with on the possible mode of action of suramin we also monitored glial (C6) phenotype, as well as on brain primary cultures. We showed the expression of cell adhesion proteins at both the mRNA and that suramin completely inhibited astrocytoma proliferation for an opti protein levels in cells maintained under drug treatment. We mal dose of 1000 Mg/ml but had the opposite effect on neuroblastoma observed that the effects were different depending on the phe- cells. For these cells, doses as low as 12.5 ng/ml first increased cell notypes of the cells studied. The pleiotropic action of suramin proliferation and then led to massive cell death. This cytotoxic effect, is discussed in relation to the possible role of growth factors in which could be compatible with an internalization of the drug by the modulating mRNA steady state levels, protein expression, and cells, was also observed for postmitotic neurons in brain primary cultures. metabolism. In both cell lines, suramin was responsible for an accumulation of the neural cell adhesion molecule at the cell surface. One of the causes was the inhibition by suramin on the liberation processes of the phosphati- MATERIALS AND METHODS dylinositol anchored M, 120,000 isoform. At the mRNA level, suramin (12.5 to 50 Mg/ml)induced an increase of all neural cell adhesion molecule Suramin (Specia, Paris, France), the hexasodium salt of 8,8'-|car- transcripts in N2A but not in C6 cells. Suramin did not have an overall bonylbis[imino-3,1 -phenylenecarbamylimino(4-methyl-2,1 -phenylene) effect on transcription rates or RNA stability as the levels of transcripts carbonylimino]jbis-l,3,5-naphthalenetrisulfonic acid, was prepared as coding for PrP*, another cell surface molecule, and actin were not affected. a stock solution of 100 mg/ml in H2O and stored at â€”¿20Â°Cunder Our data demonstrated pleiotropic action of suramin. The neurotoxic sterile conditions. [meiA>7-3H]Thymidine was from Amersham Inter effect exerted on neurons needs to be considered as possible outcomes national (Amersham, United Kingdom). Culture media and supple for the use of suramin in humans. ments were from Eurobio (Paris, France).

Cell Cultures INTRODUCTION Suramin has been used for several decades in the treatment The C6 rat glioma and N2A mouse neuroblastoma cell lines were routinely grown in DMEM supplemented with 10% fetal calf serum. of prophylaxis of Rhodesian and Cambian trypanosomiasis and In some experiments, cells were cultured in a chemically defined me for clearing the adult filariae in onchocerciasis (1, 2). It has dium consisting of DMEM/Ham's F-12 (1:l .v/v), 15 mM 4-(2-hydrox- been shown that suramin inhibits a wide variety of biological yethyl)-l-piperazineethanesulfonic acid at pH 7.4, supplemented with systems, presumably on the basis of forming stable complexes EGF (5 ng/ml), transferrin (10 Mg/ml), and selenium (10 ng/ml). with proteins (3). Recently there has been a resurgence of Similar results were obtained when insulin was used instead of EGF. interest in suramin primarily because of its efficacy in prevent After a 24-h adherence period suramin was added to the medium to ing T-lymphocyte infection by human immunodeficiency virus reach the required concentrations. in vitro, probably by inhibiting the virus reverse transcriptase Brain primary cultures were prepared from embryonic day 15 mice (4). Moreover, antiproliferation effects had been documented as described (12) and kept for 5 days in DMEM-10% fetal calf serum on poly-L-lysine coated coverslips with or without suramin before being on lymphoid cells (5) as well as on human colic cancer cells treated for immunofluorescence. HT29 (6) and on glioma cells (7). In the two latter cases, the drug was able to very quickly trigger differentiation. Growth Studies Suramin was also shown to tightly bind to a wide variety of tumor growth factors in vitro, including platelet derived growth Around 1 x IO5cells were plated in 60-mm tissue culture dishes; 1 factor (8), EGF,3 and transforming growth factor ÃŸ(9). All day after plating they were treated with suramin at the appropriate three factors are known to modulate the growth of tumor cells concentration. They were trypsinized and counted with a hemocytom- and to influence gene expression at the transcriptional level. eter at different time intervals in culture. Trypan blue exclusion was Based on these observations a clinical trial was conducted and used as an indicator of viability. it was concluded that suramin was potentially useful as an Immunofluorescence Studies anticancer drug (10) although an adverse neurotoxic effect was noticed (11). These observations prompted us to explore the The preparation and specificity of site-directed rabbit antibody rec effect of suramin on astrocytoma and neuroblastoma cell lines ognizing the NH2 terminal N-CAM domain, has been documented elsewhere (13). Antibody anti-GFAP was a kind gift of Dr. R. Pruss Received 2/26/90; revised 5/14/90. (NIH, Bethesda, MD). When double immunofluorescence staining was The costs of publication of this article were defrayed in part by the payment conducted with anti-GFAP and anti-N-CAM antibodies, a rat anti- of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. mouse N-CAM monoclonal antibody (H28) was used (14). N-CAM ' This work was supported by Grants ARC 6895 and AFM to G. R., FNLCC immunofluorescence was conducted on live cells. For double labeling, to J. M., CNRS, and University of Provence. N-CAM antibody was applied first; then the cells were fixed and 1To whom requests for reprints should be addressed. 3The abbreviations used are: EGF, epidermal growth factor; DMEM, Dulbec- permeabilized with acid/alcohol (5/95, v/v) prior to addition of the co's modified Eagle's medium; N-CAM. neural cell adhesion molecule; GFAP, anti-GFAP antibody (12). Bound antibodies were revealed with anti- glial fibrillary acidic protein; cDNA, complementary DNA. isotype second antibodies coupled to rhodamine or fluorescein. 5164

DNA Synthesis N2A cells were plated in Costar 96-well trays at 2 x IO3cells/well. One day after plating, the cells were treated with suramin at 25 and 50 Mg/ml. DNA synthesis was determined at different intervals after in corporation of 0.25 pCi [mef/i)>/-3H]thymidine (41 Ci/mmol) to each well for 4 h. The experiments were performed in quadruplicate. Expression of N-CAM, PrP% and Actin mRNAs Isolation and characterization of cDNA clones coding for N-CAM had been reported by Goridis et al. (15) and Barthels et al. (16). Insert et tÃ DW 3 was used in this study. The PrPÂ°cDNA clone (17) was a kind (O 5 gift of Dr. F. Meyer (Zurich, Switzerland). The actin cDNA clone (18) 3 was kindly supplied by Dr. M. Buckingham (Institut Pasteur, Paris, Z France). These purified inserts were labeled to high specific activity (3 x 10s U cpm/^g DNA) using the technique of Feinberg and Vogelstein (19). u Total RNA was extracted, and samples were electrophoresed on 0.8% agarose gels and transferred to nylon membranes (Nitroscreen; NEN). Hybridization was conducted according to the method of Gennarini et al. (20). In some experiments blots were dehybridized and reacted with labeled actin or PrP probes. The various bands of the autoradiograms were quantified by scanning using a GS 300 densitometer equipped with a GS 370 data system. Detection of N-CAM Proteins by Immunoblotting time (days) time (days) Fig. 1. Effect of suramin on N2A (A, B) and C6 (C, D) cell growth. Cells were This was conducted according to the method of Rougon et al. (12, plated in multidish plates (10" cells/well) and cultured in DMEM supplemented 13). Briefly, cells were pelleted and lysates were prepared in Tris-HCl with fetal calf serum (A, C) or in defined medium (B, D), in the presence of buffer, pH 7.4, containing 1% Nonidet P-40 and a cocktail of protease various doses of suramin added 24 h after plating. â€¢¿,untreatedcells. O, 25 Â¿tg/ inhibitors. After centrifugation at 100,000 x g for 90 min, protein ml; Â»,50 Mg/ml for N2A (A, B). A, 100 Mg/ml: A, 1000 ^g/ml for C6 (C, D). At concentrations of supernatants were measured and they were charged different time intervals, cells from three wells corresponding to each dose were on a 7% sodium dodecyl sulfate containing polyacrylamide gel and collected and counted. Trypan blue was used as an indicator of viability. blotted on Nitroscreen. N-CAM was revealed using rabbit antibody diluted 1:1000. Bound antibody was detected with I25l-protein A (0.5 Table 1 Effect of suramin on f'HJthymidine incorporation into DNA ofN2A x IO6cpm/ml) and subsequent autoradiography. cells grown in defined or serum containing medium Values (cpm/well) are average of 4 determinations. (^g/ml)Time Doseofsuramin RESULTS (h)0502512.5Serum Effect of Suramin on Cell Growth, DNA Synthesis, and Cell supplemented medium12244872Defined Morphology. Results of the effect on cell growth are shown in Â±2749 Â±2975 Fig. 1 for two culture conditions (with or without serum and Â±4925 Â±5079 for different doses of suramin). For doses up to 1000 ÃŸg/mlof Â±9924 Â±1105Â±1886Â±2253 Â±1331 suramin, C6 astrocytoma cells survived under both conditions for at least 8 days (Fig. 1, C and D). Suramin showed a strong medium122448722294 Â±1213 effect on proliferation as growth was inhibited by 75 and 100% +1446Â±1684Â±704Â°1592538941438911431633034827638817262235Â±Â±â€¢+-Â±519180421851Â±4433129Â±2913226 for doses of 100 and 1000 ng/m\, respectively, after 5 days in Â±402795Â±20326904037563519551615334234571139Â±Â±Â±Â±Â±Â±215151385703230360Â±445Â±5182424Â±1031 culture in defined medium (Fig. ID). Thus a dose of 1000 /ig/ Â±825341747158100224238 ml completely inhibited proliferation without major toxic ef ' Average Â±SD. fects. For lower doses (100 Mg/ml), efficacy in inhibiting growth was better in defined medium. By contrast, doses of 100 Mg/ml alteration in the C6 astrocytoma cell morphology, provided were very toxic for N2A neuroblastoma cells as most of the that the dose was above 10 ng/m\. This effect was detected cells died after 2 days of culture (not shown). With lower doses within 24 h of suramin addition and was more striking when such as 50, 25, and 12.5 /Â¿g/ml,cells survived for at least 2 days cells were grown in defined medium. Under treatment cells without any sign of cell loss although most of the cells died exhibited a multipolar shape associated with the appearance of after 3 days (Fig. 1, A and B). For doses lower than 12.5 fig/ml many processes (Fig. 2, B and C). This effect persisted through cytotoxicity was less evident and cells survived longer (not out the presence of the drug. shown). Interestingly, in neuroblastoma cells, suramin tran N2A neuroblastoma cells, when cultured in defined medium, siently activated cell proliferation (Fig. 1, A and B). This effect showed some signs of morphological differentiation as they was even more striking in serum free medium in which the started to extend processes after 24 h in culture (Fig. 2D). nontreated cells divided very slowly (Fig. IB). The effect of Suramin at doses of 12.5 or 50 Mg/ml inhibited neurite growth. proliferation was maximum between 24 and 48 h in the presence Cells remained round and after 48 h, clumps of cells adhering of the drug, before a cytotoxic effect could be detected. This to each other could be detected (Fig. 2F). was confirmed when DNA synthesis was evaluated by [methyl- Toxic Effect on Mouse Neural Primary Culture. In order to 3H]thymidine incorporation. As shown in Table 1, about a 50% verify whether the cytotoxic effect observed on neuroblastoma increase in thymidine incorporation was observed with drug cells also occurred on differentiated, postmitotic neurons, we concentrations ranging from 12.5 to 50 Mg/ml. prepared mouse brain primary cultures (Fig. 3, A-C) and ex In both types of culture media, suramin induced a marked amined the effect of suramin on cell survival and morphology 5165

Fig. 2. Effect of suramin on cell morphol ogy and N-CAM cell surface expression. C6 and N2A cells were grown for 2 days on poly- lysine coated coverslips in defined medium containing various concentrations of suramin, 0, 10. and lOO^g/ml for C6 cells M, B, C)and 0, 12.5. and 50 ng/ml for N2A cells (D, E, F). Live cells were incubated for 45 min with ani i N-CAM polyclonal antibody. Bound antibody was revealed with goat anti-rabbit |l-|.ib t.>l labeled with fluorescein. Experiments were conducted under strictly the same conditions with the same time of exposure for each pho tograph taken using Kodak Asa 800 films. Note the increase of N-CAM Â¡mmunoreactiv- ity and morphological changes in cells under suramin treatment. Bar. 20 i/m.

(Fig. 3, D-F). In such cultures the major types of cells observed and undergoing degeneration (Fig. 3E). In contrast, no differ were: (a) neurons which were strongly labeled with anti-N- ence in GFAP expression could be detected between controls CAM antibodies and which exhibited a recognizable morphol (Fig. 3C) and treated cultures (Fig. 3F). From this experiment ogy with long neurite extensions (Fig. 3Ã„);and (b) flat glial we concluded that suramin was neurotoxic at doses higher than cells, the majority of which were astrocytes as they expressed 12.5 Mg/ml and that this effect was independent of proliferation the specific intracellular marker GFAP (Fig. 3C). as neurons in primary cultures are postmitotic. Cultures were treated for 3 days with either 25 or 50 ng/m\ Effect of Suramin on N-CAM mRNA Expression. The obser of suramin and subsequently examined in double immunofluo- vation of increased cell-cell contacts in drug treated cells might rescence labeling with anti-N-CAM (Fig. 3, B and E) and anti- reflect a change in cell surface expression of intracellular adhe GFAP (Fig. 3, C and F) antibodies in order to reveal the sion molecules. Thus, we searched for the effect of suramin on phenotype of stained cells. It was clear that suramin completely mRNA expression of an adhesive molecule, N-CAM, expressed depleted the culture of cells with the neuronal phenotype as on the surface of both N2A and C6 cells. This molecule exists labeling with N-CAM antibody was almost negative. Occasion in several molecular forms that are selectively expressed in ally, round N-CAM positive cells could be seen; they most different cell types and during different stages of tissue devel likely corresponded to neurons having retracted their neurites opment (20, 21). The diversity is generated by alternative splic- 5166

Fig. 3. Effect of suramin on brain primary cultures. Embryonic mouse brain primary cul tures were kept on poh l\ sine coated coverslips for 3 days in serum containing medium, with out (A, B, C) or with (D, E, F) 50 Mg/ml of suramin. Double immunofluorescence labeling was performed with anti-N-CAM H28 rat monoclonal antibody (B, E) and anti-GFAP rabbit polyclonal antibody (C, F). H28 was applied on live cells and revealed with Â¡nui isotype antibody conjugated with rhodamine. After rinsing, cells were fixed and permeabilized and incubated with anti-GFAP antibody revealed with anti-isotype antibody conjugated with fluorescein. A and />, corresponding phase contrast micrographs for B, C and E, F, re spectively, with focus on neurons for A and on flat cells for I). Bar, 20 Â¡un.

ing and differential polyadenylation site selection within mRNA and 6.7 kilobases long; it is known that the 7.2-kilobase species products of a single complex gene. In neural tissues the major is specifically expressed by cells with a neuronal phenotype mRNA species found are 2.9, 5.2, 6.7, and 7.2 kilobases long (19). The 2.9-kilobase band coding for the M, 120,000 isoform (20). We prepared total RNA from C6 and N2A cells grown of N-CAM was very faintly expressed and hardly detectable. either in defined or serum containing medium and maintained With suramin, whatever the culture conditions, the transcript for various periods in the presence of different concentrations levels were increased by about 2 times over the value of the of suramin. In C6 cell RNA the N-CAM probes hybridized to control without suramin. This effect seemed to be dose depend major transcripts migrating at 6.7, 5.2, and 2.9 kilobases, ent and maximum for 48 h of drug treatment. A typical North respectively, when compared to rRNA markers. None of these ern blot is shown in Fig. 4; when normalized relative to the 28S transcripts seemed to be greatly altered in size or intensity rRNA content, values of 0.8, 1.2, and 1.8 were obtained for 0, following suramin treatment as shown for a dose of 100 ng/m\ 25, and 50 Mg/ml suramin in defined medium, respectively, and in Fig. 4. In defined medium, all N-CAM transcripts appeared 0.9 and 1.4 for 0 and 25 Mg/m' suramin in serum supplemented to be expressed in a lesser extent in comparison with serum medium. containing medium, but no effect of suramin could be detected It was interesting to test whether suramin had a general effect (not shown). in increasing steady state mRNA levels in N2A cells or whether In N2A cell RNA, the major transcripts detected were 7.2 it was selectively modulating expression of some of them. For 5167

N-CAM 6.7kb 5.2kb N-CAM

2.9 kb rRNA 0 100 18S

O 25 tÂ«28S rRNA

Act in 18S

O 50 O 100 Fig. 4. Effect ot MUamÂ¡nonthe expression of mRNA transcripts. Northern blot analysis of N-CAM. PrP0. and actin mRNA in suramin treated N2A and C6 cells. Total RNA was isolated from N2A cells grown in serum supplemented (S) or defined medium (D) for 48 h. with or without (0) various doses of suramin expressed as Â¿ig/ml.The experiment showing C6 transcripts was performed with cells grown in serum supplemented medium, without (0) or with 100 pg/ml suramin. Total RNA (10 jig) was denaturated with formaldehyde, electrophoresed on 0.8% agarose-formaldehyde gels, and blotted to a Nylon membrane as described. The same blot was subsequently stained with mÃ©thylÃ¨neblue;28S and 18S rRNAs were used as molecular weight markers and to check the amount of RNA in each lane, kb, kilobase. this purpose, we examined its effect on expression of mRNAs coding for molecules other than N-CAM. We examined mRNA N2A Co expression of actin and another cell surface protein, PrPc, mainly expressed by neurons (17). Actin mRNA (1.9 kilobases) expression was not affected by suramin treatment. This was â€”¿180 also the case for the 2.1-kilobase transcript coding for PrPc (Fig. 4). In some experiments, a small decrease was seen though it was not constantly observed. Thus, in N2A cells, the effect of suramin on mRNA levels is selective and varied according to the mRNA examined. Effects of Suramin on N-CAM Protein Expression. Expres sion of N-CAM isoforms by C6 and N2A cells was examined 6.25 125 25 100 by immunoblotting. Cells were cultured for 2 days in the pres Fig. 5. Immunoblot analysis of N-CAM contents in cell extracts. N2A and C6 cells were grown in defined medium for 48 h, in the absence (0) or the presence ence or absence of suramin at different doses (Fig. 5). N2A of various doses of suramin. Cells were collected and membrane extracts were cells mainly expressed the M, 180,000 and 140,000 isoforms, prepared as described in "Materials and Methods." Extracts (300 Mg) were submitted to electrophoresis on a 7% acrylamide gel and blotted onto nitroscreen. whereas C6 expressed the M, 140,000 and 120,000 isoforms. N-CAM was revealed using anti-N-CAM polyclonal antibody and I25l-labeled Suramin had no effect on the cellular content of the M, 140,000 protein A. isoform but appeared to increase the quantity of the M, 120,000 isoform in C6 cell homogenates. In N2A, by contrast, we clearly increased under drug treatment. This M, 120,000 iso- detected an increase in overall N-CAM proteins in agreement form is known to be a glycane-phosphatidylinositol anchored with mRNA increased levels observed in these cells under protein which could be selectively released under soluble form suramin treatment. We also noticed that the M, 120,000 iso- in the culture medium (22). form could be detected at the protein level and that it was In order to test the hypothesis that suramin interferes with 5168

Downloaded from cancerres.aacrjournals.org on September 27, 2021. © 1990 American Association for Cancer Research. SURAMIN EFFECT ON NEURAL CELL LINES the Afr 120,000 N-CAM isoform release, we collected every already been proposed for HT29 cells (6, 23). 24-h culture medium from C6 cultured with or without suramin For N2A cells, suramin transiently increased growth rate, and examined by immunoblot, their content in Mr 120,000 blocked neurite extention induced by culture in defined me isoform (Fig. 6). The quantity of M, 120,000 N-CAM detectable dium, and eventually led to massive cell death. This last effect in the supernatant was much lower under suramin treatment. was not associated with the mitotic state of the cells, inasmuch The difference between assay and control increased together as we also demonstrated that the drug was cytotoxic for post- with time of presence of the drug, the maximal difference being mitotic neurons in brain primary cultures. Our observations observed after a 5-day treatment. As expected from these data, differ from the data of Hensey et al. (24); these authors have a 5-day suramin treatment elicited an increase in the cell reported an inhibition of the rate of proliferation and a differ associated M, 120,000 isoform on top of that depicted for 2 entiation of N2A cells treated by suramin. One possible expla days in Fig. 5. It is noteworthy that prolonging times in culture nation for the discrepancies are differences in experimental even in the absence of suramin somewhat alters isoform expres conditions, their tests being conducted only for 2 days with a sion possibly by increased cell contacts. single dose of the drug, that might not have allowed them to Effect of Suramin on Cell Surface Expression of N-CAM. notice the cytotoxic effects. It is likely that neurons express Immunofluorescence labeling showed an increased expression "binding sites" for suramin which is known to bind tightly to of N-CAM immunoreactivity on both C6 and N2A cells; this various molecules that are positively charged (3). Alternatively, is shown in Fig. 2 for cells maintained in defined medium. This suramin may be selectively accumulated inside these cells which increase of N-CAM immunoreactivity at the surface was ex express a wide variety of channels. The action of suramin on pected for N2A cells, since both the mRNA and protein levels autocrine factors can, however, not be excluded. Although were higher after suramin treatment. For C6 cells we demon further experiments are needed to ascertain the action of sura strated that suramin interferes with the processes of liberation min on neurons, it is clear that this effect should be taken into of Mt 120,000 N-CAM in the medium, leading to its accumu account in clinical trials designed to test suramin as a therapeu lation at the cell surface. We cannot exclude either that suramin tic agent. To date, it is not known whether suramin is able to alters the turn over mechanisms, in particular endocytosis, of penetrate the blood-brain barrier. This ability would constitute the cell surface molecules. a very damaging action to the central nervous system. The availability of a radioactive derivative of the drug would help in clarifying these points. DISCUSSION The effects of suramin on steady state mRNA levels are interesting by a fundamental point of view. We demonstrated A report following a clinical trial of suramin as an anticancer that in N2A cells the action was different according to the drug mentioned its adverse neurotoxic effects (10). With the mRNA examined. We observed no changes or only a slight goal to locate its action on the nervous system, we examined decrease for PrPc as well as actin mRNAs and a noticeable and the action of suramin on two cell lines of neural origin, one of reproductible increase for N-CAM transcripts. One possibility neuronal (N2A) and the other of astroglial (C6) phenotype. would be that the transcription of these genes is controlled by When comparing the effects of the drug on growth and different pathways driven by different factors, which could or morphology of the two types of cells, it appeared that they were could not be affected by suramin. Differential splicing processes strikingly different. In both cases suramin interfered with pro occurring in N-CAM gene transcription are probably not af liferation in serum-supplemented and defined medium culture, fected since the ratio between transcripts coding for M, 180,000 but with opposite effects. The growth inhibitory effect of sura and 140,000 isoforms do not change markedly under suramin min on C6 cells suggests that the drug binds to exogenous and treatment. autocrine factors (i.e., EGF) controlling proliferation as has The differential effect of suramin on N-CAM mRNA levels in C6 and N2A cells suggests that their transcription could be -SN- under the control of different factors in neuron and glial cells. It would be tempting to speculate that N-CAM expression and + - i + - i + - cell proliferation are under the control of autocrine regulating Pf! mechanisms as already described for other cell types (25). Then, because of its binding properties, suramin could trap the in volved regulating factors, breaking the growth inhibitory auto 140^ crine loop, eventually leading to proliferation and increased N- CAM mRNA expression. Elucidation of the nature of mole 120^ cules bound by suramin followed by a test of their action on N2A cell behavior should be instrumental in testing this hy pothesis. The increased cell surface expression of N-CAM molecules 1 2 3 in both types of cells and not correlated with an increase in Fig. 6. Immunoblot analysis of M, 120,000 N-CAM isoform released in C6 expression of N-CAM mRNA transcripts in C6 astrocytoma is culture supernatants. Culture media (SW) were collected every 24 h from C6 cells likely to be due to another mode of action of suramin. Indeed, maintained without or with 100 /ig/nil of suramin, starting from 2 days after addition of the drug. Three successive collections were done, 1, 2, and 3. Super this drug had been reported to exert a lysosomotropic effect natants were dialyzed, lyophilized, and immunoblotted as in Fig. 5. Each lane (1), which could induce perturbation of the endocytotic path corresponded to a volume of medium normalized for an equivalent number of ways. This could result in the blockade of N-CAM molecule cells in assay and control. Immunoreactivity for N-CAM in the M, 120,000 area was less intense in suramin treated (+) than in control (â€”)media. The difference internalization leading to its accumulation of the molecules at between assay and control increased in parallel with the time of drug treatment. the cell surface without interfering with the regulation of their Cells were collected at the end of the experiment and their immunoreactivity for anti-N-CAM antibody was examined. The content in M, 120,000 isoform was biosynthesis. It is also well known that the M, 120,000 glycane- higher in suramin treated than in control cells. phosphatidylinositol anchored isoform is spontaneously re- 5169

Suramin inhibits cell growth, glycolytic activity and triggers differentiation leased into the culture medium (22). The physiological mecha of human colic adenocarcinoma cell clone HT29-D4. J. Biol. Chem., 264: nisms of its release are not yet understood; however, they are 10282-10286, 1988. likely to involve glycane-phosphatidylinositol enzymes either 7. Fantini, J., Guo, X. J., Marvaldi, J., and Rougon, G. Suramin inhibits proliferation of rat glioma cells. Int. J. Cancer, 45: 554-561, 1990. of C or D specificity (26). Such enzymatic mechanisms could 8. Hosang, M. Suramin binds to platelet-derived growth factor and inhibits its be impaired by suramin, here again leading to an accumulation biological activity. J. Cell. Biochem., 29: 265-273, 1985. 9. Coffey, R. J., Jr., Leof, E. B., Shipley, G. D., and Moses, H. L. Suramin of the M, 120,000 isoform at the cell surface. This hypothesis inhibition of growth factor receptor binding and mitogenicity in AKR-2D is in agreement with our results showing a lower amount of this cells. J. Cell. Physiol., 132: 143-148, 1987. protein in the medium of cells treated with suramin, and with 10. Stein, C. A., LaRocca, R. V., Thomas, R., McAtee, N., and Myers, C. Suramin an anticancer drug with a unique mechanism of action. J. Clin. the increase of the M, 120,000 isoform observed on immuno- Oncol. 7:499-508, 1989. blots of suramin treated C6 and N2A cells. Moreover, this 11. Cooper, M., La Rocca, R., Stein, C., and Meyers, C. Pharmacokinetic correlates with observations made on carcinoembryonic anti monitoring is necessary for the safe use of suramin as an anticancer drug. Proc. Am. Assoc. Cancer Res., 30: 242, 1989. gen, a phosphatidylinositol anchored molecule that is expressed 12. Rougon, G., Hirsch, M. R., Hirn, M., Guenet, J. L., and Goridis, C. by HT29 cells. When these cells are treated with suramin, Monoclonal antibody to neural cell surface protein; identification of a glycoprotein family of restricted cellular localization. Neuroscience, IO: 511- release of carcinoembryonic antigen was decreased and it ac 520, 1983. cumulated at the cell surface (27). On the other hand, we could 13. Rougon, G., and Marshak D. Structural and immunological characterization not exclude the possibility that the action of suramin in mod of the amino terminal domain of mammalian neural cell adhesive molecules. J. Biol. Chem., 261: 3396-3401, 1986. ulating the release of phosphatidylinositol anchored molecules 14. Hirn, M., Pierres, M. Deagostini-Bazin, H., Hirsch, M. R., and Goridis, C. would occur through the binding of external soluble factors. Monoclonal antibody against cell surface glycoprotein of neurons. Brain Res. 2/4:433-439, 1981. Indeed, in some instances release of phosphatidylinositol an 15. Goridis, C., Hirn, M., Santoni, M. J., Gennarini, G. F., Deagostini-Bazin. chored proteins had been shown to be increased by factors such H., Jordan, B., Keif, M., and Steinmetz, M. Isolation of mouse N-CAM as insulin (26). related cDNA. Detection and cloning using monoclonal antibodies. EMBO J., 4:631-635, 1985. In conclusion, our data demonstrate pleiotropic effects of 16. Barthels, D., Santoni, M. J., Wille, W., Ruppert, C, Chaix, J. C, Hirsch, suramin on neural cells. Its differential effect on cells with M. R., Fonticella-Camps, J. C., and Goridis, C. Isolation and nucleotide sequence of mouse NCAM cDNA that codes for a M, 79,000 polypeptide neuronal or astroglial phenotype may have clinical relevance without a membrane spanning region. EMBO J., 6:907-914, 1987. for physicians using this agent in the treatment of trypanoso- 17. Basler, K., Oesch, B., Scott, M., Westaway, D., Walchli, M., Groth, D., miasis, onchocerciasis, and human immunodeficiency versus McKinley, M. P., Prusiner, S., and Weismann, C. Scrapie and cellular PrP" isoforms are encoded by the same chromosomal gene. Cell, 46: 417-428, associated syndromes and its intended use in cancer therapy. 1986. The massive toxic effect observed on neurons in primary cul 18. Alonso, S., Minty, A., Bourlet, V., and Buckingam, M. Comparison of three actin coding sequences in the mouse; evolutionary relationship between the tures cautions against the use of high doses of suramin in actin genes of warm-blooded vertebrates. J. Mol. Evol. 23: 11-22, 1986. humans. Such selectivity for neurons may be related to cell 19. Feinberg, A., and Vogelstein, B. A technique for radiolabeling DNA restric entry and should be the subject of future studies. We also tion endonuclease fragments to high specific activity. Anal. Biochem. 137: 266-267, 1984. confirm the potency of suramin to inhibit the growth of glioma 20. Gennarini, G. F., Hirsch, M. R., He, H. T., Hirn, M., Finne, J., and Goridis, cells in a similar manner to that reported for adenocarcinoma C. Differential expression of mouse neural cell adhesion molecule (NCAM) mRNA species during brain development and in neural cell lines. J. Neurosci. cells (6) by probably interfering with growth stimulatory path 6: 1983-1990, 1986. ways. Learning the mode of action of suramin should also help 21. Rougon, G., Deagostini-Bazin, H., Hirn, M., and Goridis, C. Tissue and in understanding mechanisms regulating growth and gene developmental stage-specific forms of a neural cell surface antigen linked to differences in glycosylation of a common polypeptide. EMBO J. /: 1239- expression in a variety of cells. 1244,1982. 22. He, H. T., Finne, J., and Goridis, C. Biosynthesis, membrane association and release of NCAM 120, a phosphatidylinositol linked form of neural cell adhesion molecule. J Cell Biol., 105: 2489-2500, 1987. REFERENCES 23. Culouscou, J. M., Garrouste, F., Remacle-Bonnet, M., Bettetini, D., Mar valdi, J., and Pommier, G. Autocrine secretion of a colorectum-derived 1. Hawking, F. Suramin with special reference to onchocerciasis. Adv. Phar- growth factor by HT29 human colon carcinoma cell line. Int. J. Cancer, 42: macol. Chemother., 15: 289-322, 1987. 895-901, 1988. 2. Roll, I. M. Miscellaneous drug used in the treatment of protozoal infections. 24. Hensey, C., Boscoboinik, D., and Azzi, A. Suramin, an anti-cancer drug, In: S. Goodman and A. Gilmon (eds.). The Pharmacological Basis of Ther inhibits protein kinase C and induces differentiation in neuroblastoma cell apeutics, Ed. 5, pp. 1081-1082. New York: MacMillan Publishing Co., 1975. clone NB2A. FEBS Lett. /: 156-158, 1989. 3. Muller, W. E., and Wollen, J. Spectroscopic studies on the complex forma 25. Toribio, M., Gutierrez-Ramos, J., Pezzi, L., Marcos, M., and Martinez, C. tion of suramin with bovine and human serum albumin. Biochim. Biophys. Interleukin-2-dependent autocrine proliferation in T-cell development. Na Acta, 427: 465-480, 1976. ture (Lond.), 342: 82-84, 1989. 4. Jentsch, K., Hunsmann, G., Hartmann, H., and Nickel, P. Inhibition of 26. Ferguson, M. A., and Williams, A. F. Cell surface anchoring of proteins via human deficiency virus type 1 reverse transcriptase by suramin-related com glycosyl phosphatidyl inositol structure. Annu. Rev. Biochem. 57: 285-320, pounds. J. Gen. Virol. 68: 2183-2192, 1987. 1988. 5. Spigelman, Z., Dowers, A., Kennedy, S., DiSorbo. D., O. Brien, M., Barr, 27. Fantini, J., Rognoni, J. B., Theveniau, M., Pommier, G., and Marvaldi, J. R., and McCaffrey, R. Antiproliferative effects of suramin on lymphoid cells. Impaired Carcinoembryonic antigen release during the process of suramin- Cancer Res., 47: 4694-4698. 1987. induced differentiation of the human colic adenocarcinoma cell clone HT29- 6. Fantini, J., Rognoni, J. B., Roccabianca, M., Pommier, G., and Marvaldi, J. D4. J. Cell. Physiol., in press, 1990.

5170

Xiao-Jun Guo, Jacques Fantini, Régine Roubin, et al.

Cancer Res 1990;50:5164-5170.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/50/16/5164

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/50/16/5164. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.