Suramin Increases P53 Protein Levels but Does Not Activate the P53-Dependent G1 Checkpoint I

Vol. 2, 269-276, February 1996 Clinical Cancer Research 269

Advances in Brief

Suramin Increases p53 Protein Levels but Does Not Activate the p53-dependent G1 Checkpoint I

Steven P. Howard, 2 Sonia J. Park, effect on tumor cells (3-6). The mechanism by which suramin Luke Hughes-Davies, C. Norman Coleman, exerts its antineoplastic effects is poorly understood, in part because of the wide range of cellular processes that it inhibits. and Brendan D. Price For example, suramin antagonizes the mitogenic actions of Stress Protein Group, Joint Center for Radiation Therapy, Dana- epidermal growth factor (7), platelet-derived growth factor (8), Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts 02115 insulin-like growth factor I (9), and serum (7), leading to cell cycle arrest in vitro. In HeLa cells, suramin inhibits both DNA synthesis and DNA polymerase activity (10). Mitochondrial Abstract toxicity following exposure to suramin disrupts the cellular Suramin is an antineoplastic agent which has a cyto- energy balance (11). Suramin can competitively inhibit PKC, 3 static effect on both normal and tumor-derived cells. We an enzyme intimately involved in the regulation of cellular have investigated whether the induction of growth arrest by signaling events (12, 13). Suramin can also alter the metabolism suramin requires the p53 protein, a tumor suppressor gene of phosphoinositides (14) and may block increases in cellular product involved in the initiation of growth arrest following Ca 2÷ following growth factor stimulation of cells (15). Another DNA damage. Activation of the p53 protein by genotoxic report indicates that suramin may inhibit topoisomerase II, agents causes increased p53 protein levels and p53-depen- which is required for the replication of DNA (16). dent transcription of the p21 gene. The p21 protein then This wide range of biological targets for suramin indicates inhibits cyclin-dependent kinases, initiating G~ arrest. Ex- that the antineoplastic activity of the drug may be exerted at posure of NIH-3T3 cells to suramin caused a rapid (1-2 h) multiple cellular sites and may vary according to cell type. The increase in the level of p53-DNA-binding activity. Flow cy- exact mechanism by which suramin exerts its antiproliferative tometric analysis indicated that suramin arrested NIH-3T3 and antineoplastic effects is therefore unclear. We have recently cells in Go-G 1. However, suramin did not increase the p53- demonstrated that suramin sensitizes human prostatic epithe- dependent transcription of the p21 gene or inhibit cyclin- lium to radiation-induced DNA damage in a manner indepen- dependent kinase 2 kinase activity. If NIH-3T3 cells were dent of cellular proliferation (17). This suggests that suramin exposed to radiation or suramin plus radiation, p21 mRNA may be interacting with the DNA damage detection pathways in levels were increased and cyclin-dependent kinase 2 kinase the cell. In the present study, to further evaluate the interactions activity was inhibited, indicating that suramin does not between DNA damage and suramin, we sought to examine the block the cells' ability to increase p21 levels. To determine mechanism by which suramin induces cellular growth arrest. whether the Go-G1 arrest induced by suramin required p53, Many chemotherapeutic agents target the DNA of tumor cells, NIH-3T3 cells transfected with a dominant negative mutant causing lesions in the genomic DNA. Cells respond to DNA p53 gene to eliminate wild-type p53 function (NMP cells) damage by pausing at the G~ or G2 stages (18-20). These cell were exposed to suramin. NMP cells still exhibited Go-G 1 cycle checkpoints allow the cell to monitor the integrity of the arrest after suramin treatment. Suramin increases p53 pro- genomic DNA before committing to either DNA synthesis or cell tein levels, but fails to increase p21 mRNA levels or to division (18-21). It is now clear that the GI checkpoint is con- activate the G~ checkpoint. These data suggest that suramin trolled by the wtp53 protein (22-26). Irradiation of human cells induces growth arrest in NIH-3T3 cells by a mechanism that induces G1 arrest (19, 20) and increases the level of wtp53 protein is independent of cellular p53 status. in the cell (22, 24-26). Cells which have a mutation or deletion in the p53 gene do not undergo radiation-induced G~ arrest, but Introduction instead proceed on to G 2 before arresting (22, 23). The induction of Suramin is a polysulfonated napthylurea which has been G1 arrest by p53 may afford the cell time to repair DNA damage used to treat trypanosomiasis and is currently under investiga- before proceeding to DNA synthesis and cell division (21). The tion as an antineoplastic agent (1). Suramin is effective in the failure of cells with mtp53 to undergo G~ arrest may cause point treatment of hormone refractory metastatic prostate cancer (2), mutations and chromosomal deletions to be incorporated into the and additional studies have shown that suramin has a cytostatic genome during DNA replication and division (18, 22, 23). The loss of wtp53 function in human malignancies may therefore be a key step in the progression of human cancer. We have previously shown that a range of chemotherapeutic agents are able to activate the p53 response and initiate Received 9/15/95; revised 11/10/95; accepted 11/15/95. This work was supported with funds from the Joint Center for Radi- ation Therapy and by NIH Grant CA64585 (to B. D. P). 2 To whom requests for reprints should be addressed, at Division of Radiation Oncology, Department of Oncology, Johns Hopkins School of 3 The abbreviations used are: PKC, protein kinase C; wt, wild type; mt, Medicine, 600 North Wolfe Street, Baltimore, MD 21287-8922. Phone: mutant; EMSA, electrophoretic mobility shift assay; BrdUrd, bromode- (410) 955-8777; Fax: (410) 955-8780. oxyuridine; cdk2, cyclin-dependent kinase 2.

growth arrest (25). In the current study, we investigated whether (0.04% in 0.1 M HC1) followed by incubation with 2 M HC1 at suramin's ability to induce growth arrest was mediated by the 37°C for 30 min. Nuclei were washed in sodium borate (0.1 M), p53 protein. The data indicate that suramin activates the p53 resuspended in PBS containing goat serum (0.1%)/Tween 20 response and induces G 1 arrest in NIH-3T3 cells, but that the (0.5%), then incubated with anti-BrdUrd monoclonal antibody growth arrest mediated by suramin does not require functional (Becton Dickinson, Mountain View, CA) for 30 min at 37°C. p53. In addition, the activation of p53 by suramin does not lead Nuclei were washed with PBS-TB (PBS containing 0.1% BSA to the activation of the normal down stream effectors of the p53 and 0.5% Tween 20) and incubated with FITC-conjugated an- protein. timouse IgG for 30 min at 37°C. Nuclei were washed in PBS-TB and resuspended in PBS-TB containing 10 ixg/ml propidium Materials and Methods iodide and 10 I~g/ml RNase. Flow cytometric analyses of the Cells. NIH-3T3 mouse f'lbroblasts were cultured in nuclei were performed using a FACScan (Becton Dickinson); DMEM with 10% calf serum and antibiotics (HycLOne, Los nuclei were excited at 488 nm. Gated data were acquired and Angeles, CA). Cells were used when 80% confluent. Suramin analyzed using LYSIS II software (Becton Dickinson). (Mobay Chemical Corporation, New York, NY) was prepared cdk2 Kinase Assays. NIH-3T3 cells were solubilized in fresh each day in culture medium. Radiation was delivered using 400 I~1 lysis buffer [50 mM Tris (pH 7.4), 0.15 M NaC1, 0.5% a 137Cs irradiator at a dose rate of 100 cGy/min. NP40, 2 mM EDTA, 20 ~g/ml aprotinin and leupeptin, 1 mM Electrophoretic Gel Mobility Shift Assays. The EMSA phenylmethylsulfonylfluoride, 50 mM NaF, and 100 IXM sodium was carried out as previously described (26). EMSA reactions o-vanadate], and cell extracts were prepared as described else- contained 32p-p53Cs oligonucleotide (0.5 ng, GGACATGC where (30, 31). Cell extract (150 Ixg) and cdk2 antibody M-2 (2 CCGGGCATGTCC; Ref. 27), nuclear extract (20 I~g), and Ixg; Santa Cruz Biotech, Santa Cruz, CA) were incubated in sonicated salmon sperm DNA (0.2 ~Lg; Sigma Chemical, St. lysis buffer (500-txl final volume) for 2 h at 4°C. Immune Louis, MO) in 25 ~1 (final volume) binding buffer [20 mM complexes were collected on protein A-Sepharose beads and 4-(2-hydroxyethyl)-l-piperazine-ethanesulfonic acid (pH 7.6), washed in lysis buffer, followed by two washes in 0.5 ml kinase 20% glycerol, 10 mM NaC1, 1.5 mM MgC12, 0.2 mM EDTA, 1 buffer [50 mM Tris (pH 7.6, 10 mM MgC12, and 1 mM DTT]. The mM DTT, and 1 mM phenylmethylsulfonyl fluoride]. Antibody remaining liquid was removed with a hypodermic needle and 35 pAb421 (Oncogene Science, Manhasset, NY) was used at 100 ~1 kinase buffer containing ATP (0.7 IXM), [32p]ATP (10 IxCi), ng/assay. Binding assays were incubated for 20 min at 25°C and and histone H1 (2 txg) were added. After 20 min at 25°C, separated on 4% nondenaturing polyacrylamide gels as previ- reactions were terminated, and the phosphorylated histone H1 ously described (26). substrate was separated using SDS-PAGE. Histone H1 bands p53 Mutant Cell Lines. NIH-3T3 cells were transfected were cut from the polyacrylamide gel, and the amount of radio- by calcium phosphate precipitation with 5/xg pC53-SCX3 (con- active incorporation was assessed by liquid scintillation taining a valine to alanine substitution at codon 143; Ref. 28) or counting. 5 p~g pCMVneo vector lacking the mutant p53 cDNA. Cells Northern Blot Analysis. Total cellular RNA was pre- were trypsinized, replated at 1:5-1:20 dilution, and exposed to pared using the method of Chirgwin et al. (32). RNA (10 Ixg) G418 (400 ixg/ml) for 12 days. Surviving colonies were was separated on 1% agarose/2.2 M formaldehyde gels, trans- trypsinized, pooled, and tested for expression of mtp53 using ferred to nylon membranes, and probed as previously described Western blotting. Pooled cell populations (containing approxi- (33) using an oligonucleotide complementary to amino acids mately 100 individual colonies) were maintained in G418 (250 111-120 (TCAGACACCAGAGTGCAAGACAGCGACAA) ~g/ml). Plasmid pC53-SCX3 was kindly supplied by Dr. B. of the mouse p21 cDNA clone (34), which specifically detects Vogelstein (Johns Hopkins University, Baltimore, MD). the 2-kb p21 mRNA. Equal loading was confirmed by measur- Immunoprecipitation and Western Blot Analysis. ing the concentration of RNA (absorbance, 260 nm) and by Preparation of cell extracts and immunoprecipitation (with p53 staining the agarose gels with ethidium bromide and assessing antibody Ab421) were carried out as previously described (26). the intensity of the 18S and 28S rRNA bands. Membranes were For Western blotting, cell extracts (5 ~g; prepared as in Ref. 26) washed in 0.2 × SSC (33) at 56°C for 30 min. were separated by SDS-PAGE, transferred to Immobilon poly- vinylidene difluoride membranes (Millipore Corp., Bedford, Results MA), and blocked in 10% dried milk. p53 was detected with p53 Many chemotherapeutic agents increase cellular p53 pro- antibody pAb421. Antigens were detected using a coupled bi- tein levels, which in turn induces a p53-dependent G 1 arrest (22, otin-avidin immunoenzyme method (Vector Laboratories, Bur- 23). Accordingly, we investigated whether the antineoplastic lingame, CA). agent suramin may exert its antiproliferative effects through Cell Cycle Distributions. Flow cytometric analysis of modulation of cellular p53 levels. Murine NIH-3T3 fibroblasts total DNA content and the simultaneous detection of BrdUrd were used since they have a well-defined cell cycle and possess incorporation is described elsewhere (29). NIH-3T3 cells were wtp53 alleles (25, 26). Asynchronously growing NIH-3T3 cells pulse labeled with BrdUrd (10 ~M for 20 min), trypsinized, and were exposed to suramin, and the levels of p53-DNA-binding resuspended in PBS containing 10% serum. Cells were collected activity were assessed using an EMSA. Low levels of p53- by centrifugation, resuspended in PBS, and fixed with 70% DNA-binding activity were detected in untreated NIH-3T3 cells ethanol/0.5% Tween 20. Nuclei were isolated by centrifugation, (Fig. 1A, Lane 0). Cells exposed to <100 p~g/ml suramin and the DNA was denatured by digesting for 30 min with pepsin showed no significant increase in p53-DNA-binding activity.

A_. B SURAMIN(pg/ml) TIME (h)

Fig. 1 Suramin increases the level of p53-DNA- 50 lOO 200 400 800 0 1 3 4 6 8 10 12 binding activity in NIH-3T3 cells. A, NIH-3T3 cells were exposed to increasing concentrations of suramin (50-800 lxg/ml) for 2 h. Cells were harvested, and nuclear extracts were prepared for EMSA as described in "Materials and Methods." Control cells (0) were not treated with suramin. B, NIH-3T3 cells were incubated with suramin (200 ~g/ml) for 1 to 12 h, followed by preparation of nuclear extracts and analysis using an EMSA. Cells at time 0 (0) were not exposed to suramin.

However, suramin concentrations above 200 ~g/ml increased A_ _B the levels of p53-DNA-binding activity, with maximum activa- C S XR S÷XR tion occurring at 800 Ixg/ml. Above 800 p~g/ml, suramin begins C S XR S÷XR to exhibit toxicity toward NIH-3T3 cells. Subsequent experi- ments were carried out at 200 Ixg/ml suramin. In Fig. 1B, NIH-3T3 cells were exposed to suramin for the indicated times. Increased p53-DNA-binding activity can be detected after a 1-h ..... ~i ': ",i ~ ~ !i¸':: ~ ~!" : exposure to suramin. Maximum activation occurred by 4-6 h, and this was sustained for at least 10-12 h during continuous exposure to suramin. Fig. 1 demonstrates that suramin induces a rapid sustained increase in the levels of p53-DNA-binding activity in NIH-3T3 cells. In NIH-3T3 cells, increased p53-DNA- binding activity can also be detected when the cells are exposed to DNA-damaging agents, including X-ray irradiation (26). In Fig. 2, we investigated whether the levels of p53 detected after exposure to suramin were similar to those seen following X-ray Fig. 2 Suramin is a more effective inducer of p53 than ionizing radi- irradiation, p53 levels were monitored by measuring the rate of ation. A, NIH-3T3 cells were either untreated (C) or exposed to suramin [35S]methionine incorporation into immunoprecipitable p53. In (S; 200 txg/ml), X-ray radiation (XR; 5 Gy), or both (S + XR) for 90 min Fig. 2A, low levels of methionine-labeled p53 were detected in in the presence of [35S]methionine. Cell extracts were prepared and control cells (Lane C). Both suramin-treated (Fig. 2A, Lane S) subjected to immunoprecipitation with p53 antibody Ab421. B, NIH- 3T3 cells were either untreated (C) or exposed to suramin (S; 200 and irradiated cells (Fig. 2A, Lane XR) showed increased levels &g/ml), X-ray radiation (XR; 5Gy), or both (S + XR) for 90 min. of methionine-labeled p53 protein, although suramin treatment Nuclear extracts were prepared, and the levels of p53-DNA-binding yielded a much larger accumulation of 35S-labeled p53. The activity were measured using an EMSA. effects of suramin and radiation appeared to be additive, since irradiation of suramin-treated cells increased the labeling of the p53 protein above that seen with suramin alone (Fig. 2A, Lane S + XR). In Fig. 2B, the levels of p53-DNA-binding activity inhibition of cdk2 kinase and increased transcription of the p21 detected after suramin treatment (Lane S) are greater than those gene. The effect of suramin on cdk2 kinase activity was mea- detected following irradiation (Lane XR). When suramin-treated sured by immunoprecipitating cdk2 and monitoring the ability cells were irradiated (Fig. 2B, Lane S + XR), the level of of cdk2 to phosphorylate histone HI. In Fig. 3, NIH-3T3 cells p53-DNA-binding activity was greater than that seen with were irradiated, and the rate of histone H1 phosphorylation was suramin alone (Fig. 2B, Lane S). Fig. 2 indicates that suramin is an followed over the next 6 h (O). cdk2 kinase activity was rapidly effective activator of the p53 response, giving rise to higher cellular decreased following irradiation, reaching a minimum at 1-2 h, levels of p53 than those seen after irradiation. In addition, the followed by a slow recovery to normal levels by 6 h. In contrast, effects of suramin and radiation appeared to be additive. cells exposed to suramin showed only a small decline in cdk2 Increased p53 levels transcriptionally activate the p21 kinase activity during a 6-h suramin exposure (O), despite the gene, which encodes a specific inhibitor of cdk kinases (30, 34, induction of p53 within 1 h of suramin addition (Fig. 1B). To 35). Since cdk2 kinase activity is required for entry into S-phase ensure that the effects of suramin were not due to a nonspecific (36), inhibition by the p21 protein leads to G1 arrest. Because effect such as inhibition of transcription or translation, cells suramin induces p53 and causes growth arrest (7-9), we inves- were incubated with suramin, and then irradiated (Fig. 3, A). tigated whether the activation of p53 by suramin leads to the Cells exposed to suramin plus radiation exhibited a prompt

3000 - A Time (h) I I 2500 1 2 4

j+~+ ,: :: 2000 Sur

.m 5Gy ,< 1500 Sur + 5Gy r~

.m 1000

B ,°°0 I i I ! I i I I 0 2 4 6 .T.L.~ a- Time (hrs) -- Z Z Z Z Fig. 3 cdk2 kinase activity in cells exposed to radiation and suramin. NIH-3T3 cell extracts were ilnmunoprecipitated with anti-cdk2 anti- ~; i~: :++~ 69kd body. The washed immune complexes were then allowed to phosphorylate histone HI, and the amount of radioactive phosphate incorporation was measured as described in "Materials and Methods." Results are expressed as cpm/50 ~g protein/20 min. @, suramin (200 Ixg/ml); O, irradiated (6 Gy); (A), suramin plus radiation. :~ 46kd

Fig. 4 Northern blot analysis of p21 mRNA levels in NIH-3T3 cells and Western blot analysis of transfected NIH-3T3 cells. A, NIH-3T3 decrease in cdk2 activity, reaching a minimum value 1 h after cells were either exposed to suramin (Sur; 200 ~g/ml), X-ray radiation (5 Gy), or both of these agents simultaneously (Sur + 5 Gy). RNA was treatment, which was virtually indistinguishable from that seen prepared at the indicated time (in h), separated on agarose gels, trans- with radiation alone (Fig. 3, O). This indicates that suramin does ferred to nylon membranes, and probed for p21 mRNA levels as not block the ability of radiation to inhibit cdk2 kinase. described in "Materials and Methods." B, NIH-3T3 cells were trans- In Fig. 4, we investigated whether suramin was able to fected with pCMVneo or pC53-SCX3 (mtp53), as described in "Mate- increase the levels of p21 mRNA. NIH-3T3 cells were exposed dais and Methods," and stable cell lines were prepared. Extracts from equal numbers of cells were prepared, separated by SDS-PAGE, and to suramin for the indicated times, and the amount of p21 transferred to Immobilon membranes. Western blot analysis was carried mRNA was measured using Northern blotting. In Fig. 4A, cells out using Ab421 to detect the mtp53 protein. NIH, normal NIH-3T3 exposed to suramin showed no increase in p21 mRNA levels cells; NNC, NIH-3T3 cells bearing the neomycin selection marker; during a 4-h incubation, which is in agreement with the data NMP, NIH-3T3 cells bearing a mtp53 gene. indicating that cdk2 kinase activity is unaffected by suramin (Fig. 3). If the length of incubation with suramin was extended for up to 12 h, no accumulation of p21 mRNA was observed over endogenous wtp53 and results in functional inactivation of (data not shown). In contrast, irradiated cells (Fig. 4A, 5 Gy) had the cellular p53 response and loss of the DNA damage-induced increased p21 mRNA levels at 1 h and 2 h after irradiation, and G~ arrest (22, 23, 24). In Fig. 4B, NIH-3T3 cells (NIH), NIH- these had returned to basal levels by 4 h. These changes in the 3T3 cells bearing only the neomycin selection plasmid (NNC) levels of p21 mRNA are consistent with the rapid inhibition and and NIH-3T3 cells transfected with a mtp53 expression vector recovery of cdk2 kinase activity following irradiation (Fig. 3). (NMP) were analyzed using Western blotting to detect p53 When cells were treated with suramin and then irradiated (Fig. protein. Only the NMP cells, transfected with the mtp53 expres- 4, Sur + 5 Gy) p21 mRNA levels still increased, although there sion vector, displayed significant levels of the M r 53,000 p53 was a slight delay in the peak of p21 mRNA accumulation protein. These NMP and NIH cells were then compared to compared to radiation alone. Because suramin-treated cells still determine their response to DNA damage. In Table 1, the effect respond to radiation by induction of the p21 mRNA and cdk2 of radiation on the cell cycle distribution of asynchronously inhibition, it is unlikely that suramin's effect is due to inhibition growing NIH-3T3 cells was determined. Irradiation increased of the transcriptional or translational apparatus. the percentage of NIH-3T3 cells in Gz-M , indicating a DNA To determine whether the induction of p53 by suramin is damage-induced Ge block. In addition, there was a decrease in required for the initiation of growth arrest by suramin, NIH-3T3 the percentage of S-phase cells, but no decrease in the number cells were stably transfected with a mtp53 gene as described in of cells in G 1, indicating that the cells were also arrested in Gj. "Materials and Methods." Transforming mtp53 is dominant NMP cells, expressing mtp53, had a larger percentage of cells in

Table I Loss of radiation-induced G~ arrest in NMP cells a NIH NMP G1-Go (%) S (%) G2-M (%) G1-Go (%) S (%) G2-M (%) Control 52 28 20 36 44 20 5 Gy 55 12 33 21 31 48 "Asynchronously growing NIH and NMP cells were either untreated (control) or irradiated (5 Gy). Two-hundred min after irradiation, cells were collected and analyzed using flow cytometry to determine the cell cycle distribution. The percentage of cells in each phase is indicated.

the S-phase than NIH cells, but fewer cells in G 1. After irradi- induction of p53 by suramin does not lead to the activation of ation of NMP cells, there was an accumulation of cells in Gz-M, the downstream effectors of p53. Suramin treatment of NIH-3T3 indicating that the G2-M block, which is independent of p53, cells did not increase the level of p21 mRNA or cause inhibition remained intact (Table 1, NMP). However, the percentage of of cdk2 kinase, although suramin caused higher levels of p53 NMP cells in G 1 and the S-phase was decreased after irradiation, accumulation than ionizing radiation (Fig. 2). When NIH-3T3 indicating that NMP cells were continuing to progress from G 1 cells were exposed to suramin, they became arrested in Go-G1, into the S-phase. The absence of a G~ block in NMP cells is with >75% being in this phase within 13 h of suramin addition. consistent with the absence of functional p53 protein in NMP If wtp53 function was abolished by overexpression of mtp53 cells. Therefore, the NMP cells have lost the G 1 checkpoint, as protein, the ability of suramin to induce Go-G 1 arrest was has been demonstrated previously for other cell types expressing unaffected. This is the first report, to our knowledge, of an mtp53 proteins (22, 23). antineoplastic agent that can increase p53 protein levels but does To determine whether the induction of p53 by suramin not activate the associated downstream effectors. could induce growth arrest, asynchronously growing NIH-3T3 When cells are exposed to DNA damage, including ioniz- cells were treated with suramin. The position of the cells in the ing radiation, p53 levels are increased, and the p53 protein cell cycle was then determined by pulse labeling with BrdUrd transcriptionally activates the p21 gene (30, 34, 35). The p21 followed by staining with propidium iodide as described in gene product binds to and inhibits cdk2 kinase, causing the cells "Materials and Methods." In untreated NIH-3T3 cells (Fig. to arrest at the G1-S boundary (35). Suramin-induced p53 ex- 5A), the percentage of cells in G1, S-phase, and G2-M remained hibited DNA-binding activity in the in vitro EMSA assay, which relatively constant over the 13-h time course of the experiment. was indistinguishable from radiation-induced p53 (Figs. 1 and In Fig. 5B, NIH-3T3 cells continuously exposed to suramin 2). However, suramin did not increase p21 mRNA levels or began to accumulate in G O or G1 after 3-4 h of exposure. We inhibit cdk2 kinase activity (Figs. 3 and 4). Suramin has a wide refer to this as Go-G t arrest since we have not determined the range of biochemical effects on cells, including alterations in the precise point in the cell cycle that suramin arrests cells. The cellular energy balance (11) and inhibition of topoisomerase percentage of cells in Go-G~ became increasingly more pro- activity (16), both of which have the potential to inhibit the nounced with increasing length of exposure to suramin until a transcription and translation of the p21 gene. However, irradi- majority (approximately 75%) was arrested in Go-G l by 13 h. ation of suramin-treated cells increased p21 mRNA levels and Suramin appeared to cause a selective Go-G 1 arrest, since cells inhibited cdk2 kinase activity, indicating that the mechanisms did not accumulate in either S-phase or G2-M. Similar results required for the p53-dependent activation of the p21 gene are were obtained when NNC cells (bearing only the neomycin intact in the presence of suramin. The failure of suramin to selection marker) were exposed to suramin (data not shown). To increase p21 mRNA levels is therefore not the result of a determine whether suramin-induced growth arrest was depen- nonspecific effect on the transcription of the p21 gene. dent on wtp53, NMP cells, which lack the p53-dependent G 1 The ability of suramin and radiation to increase p53-DNA- checkpoint (Table 1), were examined to determine whether binding activity were additive, suggesting that they may in- suramin would induce growth arrest. In untreated cultures of crease the p53 levels by distinct pathways. The activation of the NMP cells, the percentage of cells in G1, S-phase, and G2-M p53 response normally occurs through the generation of single- phase were unchanged throughout the experiment, although the or double-strand breaks in DNA by genotoxic compounds (25, NMP cells had a higher percentage of cells in G1 then the 37). However, the mechanism by which suramin increases the parental NIH-3T3 cell line (Fig. 5C). When NMP cells were p53 levels is unclear. Suramin can inhibit topoisomerase II, an exposed to suramin, there was an increase in the percentage of enzyme involved in the unwinding of DNA (16). Inhibitors of cells in G~ (Fig. 5D), with at least 85% in Go-G ~ after 13-h topoisomerase II can stabilize the cleaveable complex formed exposure to suramin, which was similar to that seen in suramin- between topoisomerase II and DNA, leading to the formation of treated NIH-3T3 cells (Fig. 5B). This indicates that suramin- single-strand breaks (25, 37), which are thought to be the signal induced growth arrest is independent of wtp53. for p53 activation (25). Suramin inhibition of topoisomerase II does not lead to the formation of a cleaveable complex (16), Discussion suggesting that suramin does not induce p53 through topoi- The results demonstrate that suramin induces a sustained somerase II-induced strand breaks. This does not exclude the increase in wtp53 protein levels in NIH-3T3 cells. However, the possibility that suramin may cause DNA damage by a separate

I00 in cell cycle distribution (7-9). In Go-arrested murine fibroblast cells, the p53 levels are minimal, but increase as cells progress 80 through G 1 and S-phase, before decreasing during Gz-M (38). However, increased p53 levels can be detected within 1 h of 60 suramin addition, which precedes the onset of Go-G1 arrest 40 (Figs. 1 and 5). In addition, serum starvation of NIH-3T3 cells, which arrests the cells in Go, does not increase the p53 levels 20 - Go/G1 (data not shown), indicating that the changes in the p53 levels I I I I I I I I I I I I seen after suramin treatment cannot be accounted for by changes 0 in cell cycle distribution. I00 Analysis of the cell cycle distribution of suramin-treated 80 NIH-3T3 cells indicates that the cells are arrested in G o or G 1. Studies in the LNCaP (39) and DU-145 (40) human prostate cell 6O lines and in primary cultures of normal human prostatic epithe- 0 .,w-I lium (17) indicate that suramin causes G1-G o arrest. However, suramin treatment of the MCF-7 human breast cancer cell line 20 GoIGa (41) and the PC-3 human prostate cell line (42) appears to result .iml in a selective G2-M arrest. In addition, the magnitude of the I I I I I I I I I I I I °f0 observed growth arrest could be modulated by manipulating 100 components of the culture medium (42). Hence, suramin can cause cell cycle arrest at either G1-G o or Gz-M, depending on 80 the cell type and growth conditions. rd Suramin can antagonize the action of growth factors which U are required for NIH-3T3 cells to progress through the cell cycle, including platelet-derived growth factor (1, 7-9). The m-i onset of G]-G o arrest by suramin may reflect the loss of growth U 20 Go/G1 factor action in these cultures, causing the cells to remain in G o. I I I I I I I I I I I I Suramin also arrested NMP cells (expressing mtp53) in G1-G o, 0 °t indicating that the growth arrest properties of suramin are inde- 100 pendent of p53 status. Suramin therefore exerts its cytostatic effects independently of its ability to induce the p53 protein, p21 80 is transcriptionally induced by the p53 gene product and medi- 60 ates the observed G 1 arrest associated with DNA-damaging agents (30, 34, 35). Other cyclin-dependent kinase inhibitors 40 - function independently of the p53 suppressor gene. The p27 Ge/G1 protein participates in the growth arrest following serum depri- 20 - vation, contact inhibition, or exposure to transforming growth 0 !. I I I I I I I I I I I factor [3 (43, 44). p15, pl6, p18, and p19 (45-48) induce cell cycle arrest by competing with D-type cyclins for binding to the -1 0 1 2 3 4 5 6 7 8 9 10 11 12 appropriate cyclin-dependent kinase. All of these p53-indepen- Time 0a) dent pathways of cell cycle regulation represent potential target Fig. 5 Flow cytometric analysis of asynchronous NIH-3T3 and NMP sites for suramin-induced growth arrest. Interestingly, suramin cells exposed to suramin. Cell cultures were pulse labeled with BrdUrd does not appear to inhibit either DNA synthesis or mitosis in at the times indicated, fixed, stained with propidium iodide and anti- BrdUrd antibody, and analyzed using flow cytometry. The area plots NIH-3T3 cells, since suramin-treated cells already in S-phase indicate the percentage of cells in each phase of the cell cycle at the progressed through mitosis before arresting in G1-G o (Fig. 5). indicated time. Suramin was present throughout the time course of the This is in contrast to studies showing that suramin can inhibit experiment. A, asynchronously growing NIH-3T3 cells. B, NIH-3T3 the activity of DNA polymerases (10). The most likely expla- cells treated with suramin (200 ixg/ml) at time 0. C, asynchronously nation for suramin-induced growth arrest is the blockade of growing NMP cells (mtp53). D, NMP cells treated with suramin (200 Ixg/ml) at time 0. growth factor action, preventing the cells from reinitiating cell cycle progression. The increase in p53-DNA-binding activity seen with suramin may therefore have an alternative function than the activation of p21 transcription and initiation of G1 (unknown) mechanism. However, since suramin-induced p53 arrest. The p53 gene product can initiate the apoptotic cell death does not activate the p21 gene or the DNA damage-dependent seen after radiation exposure (49). Suramin can potentiate radi- G t checkpoint, it is possible that the increased p53 occurs ation-induced apoptosis (3), suggesting a possible interaction through a mechanism that does not involve DNA damage. with other p53 target genes involved in the apoptotic pathway. Previous studies have indicated that the p53 levels are subject to Irradiation of suramin-treated cells increased p21 mRNA cell cycle regulation (38), suggesting that the effects of suramin levels and inhibited cdk2 activity, indicating that the p53-de- on the p53 levels may be explained by suramin-induced changes pendent G1 checkpoint is still intact after suramin treatment.

This observation may be clinically important since the integra- 4. Peehl, D. M., Wong, S. T., and Stamey, T. A. Cytostatic effects of tion of cytostatic agents such as suramin with fractionated suramin on prostate cancer cells cultured from primary tumors. J. Urol., 145: 624-630, 1991. radiation therapy is currently being evaluated (50). In both in 5. Vignon, F., Prebois, C., and Rochefort, H. Inhibition of breast cancer vitro (51) and animal solid tumor systems (52), the loss of a growth by suramin. J. Natl. Cancer Inst., 84: 38-42, 1991. functional p53 response has been shown to disrupt an important 6. Kraft, A., Reid, L. M., and Zvibel, I. Suramin inhibits growth and yet apoptosis pathway, resulting in refractoriness to standard antin- promotes insulin-like growth factor II expression in HepG2 cells. Can- eoplastic therapies. In prostate cell lines, prior treatment with cer Res., 53: 652-657, 1993. suramin decreased radiation-induced cell death, whereas expo- 7. Betsholtz, C., Johnson, A., Heldin, C-H., and Westermark, B. Effi- sure to suramin after radiation increased the cytotoxic effects of cient reversion of simian sarcoma virus-transformation and inhibition of growth factor-induced mitogenesis by suramin. Proc. Natl. Acad. Sci. radiation (3). These results are in direct contrast to our own USA, 83: 6440-6444, 1986. observations using cultured normal human prostatic epithelium 8. Hosang, M. Suramin binds to platelet-derived growth factor and (17). In this model system, pretreatment with suramin resulted inhibits biological activity. J. Cell. Biochem., 29: 265-273, 1985. in both G 1 arrest and increased radiosensitivity. The reasons for 9. Pollak, M., and Richard, M. Suramin blockade of insulin-like growth these discrepancies are unclear. However, these data suggest factor I-stimulated proliferation of human osteosarcoma cells. J. Natl. Cancer Inst., 82: 1349-1352, 1990. that a combination of the cytostatic effect of suramin with 10. Jindal, H. K., Anderson, C. W., Davis, R. G., and Vishwanatha, radiation could potentially yield a beneficial therapeutic effect J. K. Suramin affects DNA synthesis in HeLa Cells by inhibition of in cancer therapy. DNA polymerases. Cancer Res., 50: 7754-7757, 1990. Why suramin induces a transcriptionally inactive p53 pro- 11. Rago, R., Mitchen, J., Cheng, A-L., Oberley, T., and Wilding, G. tein is unclear. DNA damage causes G 1 and G 2 arrest in cells Disruption of the cellular energy balance by suramin in intact human prostatic carcinoma cells, a likely antiproliferative mechanism. Cancer (20, 22). Suramin does not appear to activate the G 1 or G 2 Res., 51: 6629-6635, 1991. checkpoints, suggesting that DNA damage is not the primary 12. Hensey, C., Boscoboinik, D., and Azzi, A. Suramin, an anti-cancer signal for suramin induction of p53. One potential mechanism drug, inhibits protein kinase C and induces differentiation in neuroblas- by which suramin may increase p53 could involve modulation toma cell clone NB2A. FEBS Lett., 258: 156-158, 1989. of PKC activity. We have previously shown that activation of 13. Mahoney, C., Azzi, A., and Huang, K. Effects of suramin, an anti-immunodeficiency virus reverse transcriptase agent, on protein PKC by phorbol esters blocks the increase in p53 protein seen kinase C. J. Biol. Chem., 265: 5424-5428, 1990. after DNA damage (26), and it has been demonstrated that 14. Kopp, R., and Pfeiffer, A. Suramin alters phosphoinositide synthe- suramin can inhibit PKC (12, 13, 53). Inhibition of PKC by sis and inhibits growth factor receptor binding in HT-29 cells. Cancer suramin may relieve the inhibitory effects of this kinase on p53 Res., 50: 6490-6496, 1990. levels, causing accumulation of the p53 protein. Suramin may 15. Seewald, M., Olsen, R., and Powis, G. Suramin blocks intracellular Ca 2÷ release and growth factor induced increases in cytoplasmic free therefore allow accumulation of p53 in the cell by an indirect Ca 2+ concentration. Cancer Lett., 49: 107-113, 1989. mechanism that does not involve DNA damage. Alternatively, 16. Bojanowski, K., LeLievre, S., Morkovits, J., Couprie, J., Jacque- the activation of p53 following DNA damage involves a number min-Sablon, A., and Larsen, A. Suramin is an inhibitor of DNA topoi- of steps, including stabilization of the p53 protein (25, 26), and somerase II in vitro and in Chinese hamster fibrosarcoma cells. Proc. may include alterations in the phosphorylation status of p53. Natl. Acad. Sci. USA, 89: 3025-3029, 1992. Many of the kinases responsible for p53 phosphorylation are 17. Howard, S. P., Groch, K. M., Lindstrom, M. J., Messing, E. M., and Gould, M. N. Proliferation-independent growth factor modulation of regulated by DNA damage, including cdk2 kinase (31, 54), radiation sensitivity of human prostate cells. Radiat. Res., 143: 229- mitogen-activated protein kinase (55), and the double-strand 233, 1995. DNA kinase (56). Suramin may fail to activate these kinase 18. Weinert, T. A., and Hartwell, L. H. The rad9 gene controls the cell pathways, resulting in accumulation of p53 in a form unable to cycle response to DNA damage in saccharomyces cerevisiae. Science activate transcription of the p21 gene. Suramin may provide (Washington DC), 241: 317-322, 1988. important functional information on the regulation of the p53 19. Lau, C. C., and Pardee, A. B. Mechanism by which caffeine potentiates lethality of nitrogen mustard. Proc. Natl. Acad. Sci. USA, protein levels and the activation of p53-dependent transcription. 79: 2942-2946, 1982. 20. Tolmach, L. J., Jones, R. W., and Busse, P. M. The action of Acknowledgments caffeine on X-ray irradiated HeLa cells: delayed inhibition of DNA We thank Dr. Roy Tishler for his helpful discussions and thought- synthesis. Radiat. Res., 71: 653-665, 1977. ful review of the manuscript. 21. Lane, D. P. p53, guardian of the genome. Nature (Lond.), 358: 80-83, 1992. 22. Kastan, M. B., Onyekwere, O., Sidransky, D., Vogelstein, B., and References Craig, R. W. Participation of p53 in the cellular response to DNA 1. Stein, c. A. Suramin: a novel antineoplastic agent with multiple damage. Cancer Res., 51: 6304-6311, 1991. potential mechanisms of action. Cancer Res., 53: 2239-2248, 1993. 23. Kuerbitz, S. J., Plunkett, B. S., Walsh, W. V., and Kastan, M. B. 2. Myers, C., Cooper, M., Stein, C., LaRocca, R., Walther, M., Weiss, Wild-type p53 is a cell cycle checkpoint determinant following irradi- G., Choyke, P., Dawson, N., Steinberg, S., Ehrich, M., Cassidy, J., ation. Proc. Natl. Acad. Sci. USA, 89: 7491-7495, 1992. Kohler, K., Trepel, J., and Linehem, W. Suramin: a novel growth factor 24. Kastan, M. B., Zhan, Q., E1-Deiry, W. S., Carrier, F., Jacks, T., antagonist with activity in hormone refractory metastatic prostate can- Walsh, W., Plunkett, B. S., Vogelstein, B., and Fornace, A. J. A cer. J. Clin. Oncol., 10: 881-889, 1992. mammalian cell cycle checkpoint utilizing p53 and gadd45 is defective 3. Sklar, G. N., Eddy, H. A., Jacobs, S. C., and Kyprianou, N. Com- in ataxia telangiectasia. Cell, 71: 587-597, 1992. bined effect of suramin plus irradiation in human prostate cancer cells: 25. Tishler, R. B., Calderwood, S. K., Coleman, C. N., and Price, B. D. the role of apoptosis. J. Urol., 150: 1526-1532, 1993. Increases in sequence specific DNA binding by p53 following treatment

with chemotherapeutic and DNA damaging agents. Cancer Res., 53: of the human prostate carcinoma cell line PC-3. Cancer (Phila.), 71: 2212-2216, 1993. 1151-1158, 1993. 26. Price, B. D., and Calderwood, S. K. Increased sequence-specific 43. Toyoshima, H., and Hunter, T. p27, a novel inhibitor of G1 cyclin- p53-DNA binding activity after DNA damage is attenuated by phorbol Cdk protein kinase activity, is related to p21. Cell, 78: 67-74, 1994. esters. Oncogene, 8: 3055-3062, 1993. 44. Polyak, K., Lee, M. H., Erdjument-Bromage, H., Koff, A., Roberts, 27. Funk, W. D., Pak, D. T., Karas, R. H., Wright, W. E., and Shay, J. M., Tempst, P., and Massague, J. Cloning of p27 Kipl, a cyclin- J. W. A Transcriptionally active DNA-binding site for human p53 dependent kinase inhibitor and a potential mediator of extracellular protein complexes. Mol. Cell. Biol., 12: 2866-2871, 1992. antimitogenic signals. Cell, 78: 67-74, 1994. 28. Baker, S. J., Markovitz, S., Fearon, E. R., Wilson, J. K. V., and 45. Serrano, M., Hannon, G. J., and Beach, D. A new regulatory motif Vogelstein, B. Suppression of human colorectal carcinoma cell growth in cell-cycle control causing specific inhibition of cyclin D/CDK4. by wild-type p53. Science (Washington DC), 249: 912-915, 1990. Nature (Lond.), 366: 704-707, 1993. 29. Gratzner, H., Monoclonal antibody to 5 bromo- and 5-iodode- oxyuridine. A new reagent for detection of DNA replication. Science 46. Hannon, G. J., and Beach, D. p15 INK4B is a potential effector of (Washington DC), 218: 474-475, 1982. TGF-beta-induced cell cycle arrest. Nature (Lond.), 371: 257-261, 1994. 30. Harper, J. W., Adami, G. R., Wei, N., Keyomarsi, K., and Elledge, S. J. The p21 cdk-interacting protein is a potent inhibitor of G~ cyclin- 47. Hirai, H., Roussel, M. F., Kato, J. Y., Ashmun, R. A. and Sherr, dependent kinase. Cell, 75: 805-816, 1993. C. J. Novel INK4 proteins, p19 and p18, are specific inhibitors of the 31. Price, B. D., Hughes-Davies, L., and Park, S. J. Cdk2 kinase cyclin D-dependent kinases CDK4 and CDK6. Mol. Cell. Biol., 15: phosphorylates serine 315 of human p53 in vitro. Oncogene, 11: 73-80, 672-681, 1995. 1995. 48. Chan, F. K., Zhang, J., Cheng, L., Shapiro, D. N., and Winoto, A. 32. Chirgwin, J. A., Pryzbyla, R., McDonald, R., and Rotter, W. Iso- Identification of human and mouse p19, a novel CDK4 and CDK6 lation of biologically active ribonucleic acid from sources rich in ribo- inhibitor with homology to p16 INK4. Mol. Cell. Biol., 15: 2682-2688, nuclease. Biochemistry, 18: 5294-5295, 1979. 1995. 33. Price, B. D., and Calderwood, S. K. Ca z+ is essential for multistep 49. Lowe, S. W., Schimtt, S. W., Smith, S. W., Osborne, B. A., and activation of the heat shock factor in permeabilized cells. Mol. Cell. Jacks, T. p53 function is required for radiation-induced apoptosis in Biol., 6: 3365-3368, 1991. mouse thymocytes. Nature (Lond.), 362: 847-849, 1993. 34. EI-Deiry, W. S., Tokino, T., Velcelescu, V. E., Levy, D. B., Par- 50. Kinsella, T. J., Gould, M. N., Mulcahy, R. T., Ritter, M. A., and sons, R., Trent, J. M., Lin, D., Mercer, W. E., Kinzler, K. W., and Fowler, J. F. Keynote Address: Integration of cytostatic agents and Vogelstein, B. WAF1, a potential mediator of p53 tumor suppression. radiation therapy: A different approach to "proliferating" human tu- Cell, 75: 817-825, 1993. mors. Int. J. Oncol. Biol. Phys., 20: 295-302, 1991. 35. Dulic, V., Kaufman, W. K., Wilson, S. J., Tlsty, T. D., Lees, E., 51. Lowe, S. W., Ruley, H. W., Jacks, T., and Housman, D. E. p53- Harper, J. W., Elledge, S. J., and Reed, S. I. p53-dependent inhibition of dependent apotosis modulates the cytotoxicity of anticancer agents. cyclin-dependent kinase activities in human fibroblasts during radiation Cell, 74: 957-967, 1993. induced G~ arrest. Cell, 76: 1013-1023, 1994. 52. Lowe, S. W., Bodis, S., McCarthy A., Remington, L., Ruley, H. W., 36. Sherr, C. J. Mammalian G 1 cyclins. Cell, 73: 1059-1063, 1993. Fisher, D. E., Housman, D. E., and Jacks, T. p53 status and the efficacy 37. Nelson, W. G., and Kastan, M. B. DNA strand breaks: the DNA of cancer therapy in vivo. Science (Washington DC), 266: 807-810, template alterations that trigger p53 dependent DNA damage response 1994. pathways. Mol. Cell. Biol., 14: 1815-1823, 1994. 38. Reich, N. C., and Levine, A. J. Growth regulation of a cellular 53. Khaled, Z., Rideout, D., O'Driscoll, K. R., Petralyk, D., Cacace, A., tumor antigen, p53, in non-transformed cells. Nature (Lond.), 308: Patel, R., Chiang, L-C., Rotenberg, S., and Stein, C. A. Effects of 199-201, 1984. suramin related and other clinically therapeutic polyanions on protein kinase C activity. Clin. Cancer Res., 1: 113-122, 1995. 39. Berns, E. M., Schuurmans, A. L., Bolt, J., Lamb, D. J., Foekens, J. A., and Mulder, E. Antiproliferative effects of suramin on androgen 54. Bischoff, J. R., Friedman, P. N., Marshak, D. R., Prives, C., and responsive tumor cells. Eur. J. Cancer, 26: 470-474, 1990. Lee, W-H. Genetic mechanisms of tumor suppression by the p53 gene. 40. Qiao L., Pizzolo, J. G., and Melamed, M. R. Effects of suramin on Science (Washington DC), 250: 1576-1580, 1990. expression of proliferation associated nuclear antigens in DU-145 pros- 55. Stevenson, M. A., Pollock, S. S., Coleman, C. N., and Calderwood, tate carcinoma cells. Biochem. Biophys. Res. Commun., 201:581-588, S. K. X-irradiation, phorbol esters and H202 stimulate mitogen activated 1994. protein kinase activity in NIH-3T3 cells through the formation of 41. Foekens, J. A., Sieuwerts, A. M., Stuurman-Smeets, E. M. J, Peters, reactive oxygen intermediates. Cancer Res., 54: 12-15, 1994. H. A., and Klijn, J. G. M. Effects of suramin on cell-cycle kinetics of 56. Lees-Miller, S. P., Chen, Y-R., and Anderson, C. W. Human cells MCF-7 human breast cancer cells in vitro. Br. J. Cancer, 67." 232-236, contain a DNA activated protein kinase that phosphorylates simian virus 1993. 40 T antigen, mouse p53 and the human Ku antigen. Mol. Cell. Biol., 42. Ewing, M. W., Liu, S. C., Gnarra, J. R. Walther, M. M., Meyers, 10: 6472-6481, 1990. C. E., and Linehan, W. M. Effect of suramin on the mitogenic response

Downloaded from clincancerres.aacrjournals.org on September 27, 2021. © 1996 American Association for Cancer Research. Suramin increases p53 protein levels but does not activate the p53-dependent G1 checkpoint.

S P Howard, S J Park, L Hughes-Davies, et al.

Clin Cancer Res 1996;2:269-276.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/2/2/269

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://clincancerres.aacrjournals.org/content/2/2/269. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.