[CANCER RESEARCH 47, 3656-3661, July 15, 1987] Effect of Adenosine Analogues on Protein Car boxy Imethyltransferase, S-Adenosylhomocysteine Hydrolase, and Ribonucleotide Reducíase Activity in Murine Neuroblastoma Cells1

Robert F. O'Dea,2 Bernard L. Mirkin, Harry P. Hogenkamp, and Donna M. Barten

Division of Clinical Pharmacology, Departments of Pediatrics, Pharmacology, and Biochemistry, University of Minnesota Health Sciences Center, Minneapolis, Minnesota 55455

ABSTRACT DZA, 2'-deoxyadenosine, and l-/3-D-arabinofuranosyladenine and the irreversible inhibitors AD, neplanocin A, and 3-deaza- The noncompetitive 5-adenosylhomocysteine (AdoHcy) hydrolase an neplanocin (3-7). AdoHcy hydrolase catalyzes the NAD+-de- tagonist adenosine dialdehyde (AD) has been shown to suppress the pendent oxidation of the 3'-hydroxyl group of AdoHcy (hydro- growth of cultured C-1300 murine neuroblastoma (MNB) cells. The lytic direction) or Ado (synthetic direction), resulting in the enzymatic sites at which AD and other nucleoside analogues exert their generation of 3'-keto-AdoHcy, 3'-ketoadenosine, and 3'-keto- cytotoxic effects have been postulated to include protein carboxyl- 4',5'-dehydroadenosine (2). Adenosine dialdehyde, generated methyltransferase (PCM), AdoHcy hydrolase, and ribonucleotide reduc íase. by periodate oxidation of adenosine, contains structural com AD (10~5 M) increased PCM activity 350% in suspensions prepared ponents (e.g., 3'-keto group) which are similar to the proposed from disrupted cells after 72 h of drug exposure; in contrast, 3-deaza- functional moieties of these enzyme-bound intermediates and adenosine ( IO-1 M) increased PCM activity 57%, whereas AdoHcy and has been reported to be a potent, irreversible inhibitor of sinefungin had no effect. When intact MNB cells were incubated with AdoHcy hydrolase activity from liver (5) and murine L929 AD for varying time periods up to 72 h and then pulse labeled with the leukemia cells (8). 5-adenosylmethionine precursor i,-|3H|-methionine, AD (10"' to 5 x 10"* Phosphorylated AD (S'-adenosine dialdehyde diphosphate) M) produced a concentration-dependent inhibition of protein car boxy I- inhibits bacterial ribonucleotide reducíase in vitro, the rate- methylation which persisted for up to 6 h. Following extended periods of AD treatment (48 to 72 h), AD (10~* to 10~5 M) produced a 250% limiting synthetic step in the production of deoxynucleotides increment in protein carboxylmethylation, similar in magnitude to that (9). However, the nonphosphorylated form of AD is a poor observed in disrupted cell preparations. This increase in carboxylmeth inhibitor of bacterial ribonucleotide reducíasein vitro (9), and ylation was observed at time-points when AdoHcy hydrolase activity it is uncertain whether AD is even phosphorylated in intact remained suppressed. The inhibitory effect of AD on AdoHcy hydrolase eukaryotic cells (6). These data suggest that AD may exert its activity was maximal within 4 h and still apparent after 72 h of incubation. cytotoxic effects on eukaryotic cells at other enzymatic sites. In contrast, AD treatment had no effect on the ribonucleotide reducíase Among these, PCM (EC 2.1.2.24) has been implicated as a activity of MNB cells. potential site at which AD and other Ado analogues exert their These data suggest that the cytotoxic effect of AD on MNB cells action (10). This enzyme catalyzes the AdoMet-dependent post- results directly from its inhibition of AdoHcy hydrolase activity and translational methylation of the carboxyl side chains of gluta- indirectly through its suppression of methyltransferase enzyme systems. The potential linkage between the observed long-term elevations in PCM myl and/or aspartyl residues of cellular proteins, resulting in activity and AD-induced cytotoxicity remains to be defined. the formation of labile protein methylesters (11). In prokar- yotes, the PCM system regulates chemotaxis by methylating and demethylating specific membrane proteins which modify INTRODUCTION flagellar rotation (12,13). A defined role for PCM in eukaryotes has not been provided; however, various reports have linked it The role of AdoMet3-dependent methyltransferase enzymes to the regulation of leukocyte (14,15) and, possibly, slime mold as potential targets for chemotherapeutic agents has been inten chemotaxis (16), cellular secretion (17,18), racemization repair sively examined in recent years (1). These investigations have of aging proteins (19), phototransduction (20), and cellular been focused on the development of competitive methyltrans growth and development (21-24). ferase inhibitors and agents that indirectly block intracellular The cytotoxic potency of AD and other nucleoside dialde- methylation reactions by inhibiting either AdoMet biosynthesis hydes has been determined in cultured C-1300 MNB cells and or AdoHcy metabolism. The latter approach logically derives shown to be dependent upon the presence of an adenine base from the defined biochemical function of AdoHcy hydrolase and an intact ribosyl dialdehyde moiety, structural elements (EC 3.3.1.1), the enzyme which catalyzes the reversible hydrol present in AD (30). In the following report, the effects of AD ysis of AdoHcy to adenosine and homocysteine (2). and other nucleoside analogues on PCM, AdoHcy hydrolase, A number of Ado analogues have been reported to inhibit and ribonucleotide reducíaseactivity were studied in disrupted AdoHcy hydrolase. They include the competitive antagonists cell preparations obtained from MNB cells after incubation with these agents. Received 10/24/86; revised 3/27/87; accepted 4/16/87. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in MATERIALS AND METHODS accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1This investigation was supported by Grants NS-17194, National Institute of Adenosine Nucleoside and Dialdehyde Analogues Neurological and Communicable Diseases and Stroke, GM33776, National In stitute of General Medical Science, and the University of Minnesota Institutional All of the compounds investigated in this study were obtained from Award from the American Cancer Society. 1To whom requests for reprints should be addressed, at Division of Clinical readily available commercial sources (Sigma, St. Louis, MO; Calbi- Pharmacology, University of Minnesota Health Sciences Center, Mayo Box 468, ochem, San Diego, CA; or Lilly Research Labs, Indianapolis, IN), as 420 Delaware Street S.E., Minneapolis, MN 55455. gifts (3-deazaadenosine, Burroughs-Wellcome Research Lab, Research 3The abbreviations used are: AdoMet, S-adenosylmethiomne; AdoHcy, S- adenosylhomocysteine; MNB, murine neuroblastoma; DZA, 3-deazaadenosine; Triangle Park, NC), or synthesized in our laboratories, utilizing previ AD, adenosine dialdehyde; Ado, adenosine; PCM, protein carboxylmethyltrans- ously published procedures (5, 25). The analogues were categorized as ferase; ANOVA, analysis of variance. follows: 3656

Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 1987 American Association for Cancer Research. EFFECT OF ADENOSINE ANALOGUES IN MNB CELLS at 50 x g for 10 min at 4*C, and the pellets were resuspended in Nucleoside Analogues phosphate-buffered saline (pH 7.4). The washed cells were pelleted Sinefungin; DZA; AdoHcy; 5'-chloro, 5'-deoxyadenosine; 5'-deox- again by low-speed centrifugation, and the pellets were lysed in a yadenosine; 5'-chloro,5'-deoxy-ara-adenosine; 3',5'-dichloro, 2',3',5'- mixture consisting of 0.05% TRITON X-100 and 2 HIMphenylmeth- trídeoxyadenosine; methylfuryladenine; adenosine S'-carboxylic acid; ylsulfonyl fluoride in phosphate-buffered saline at a linai concentration and Ado. of IO7cells/ml. The cellular lysates were immediately frozen and stored at -70"C prior to further use. Proteins were measured by the method Dialdehyde Analogues of Lowry et al. (27). Enzyme assays for PCM and ribonucleotide AD; reduced adenosine dialdehyde; 5'-deoxyadenosine dialdehyde; reducíasewere performed with these disrupted cell preparations. inosine dialdehyde; guanosine dialdehyde; undine dialdehyde and cyti- Enzyme Assays dine dialdehyde. Dialdehydes were prepared by periodic acid oxidation of the parent nucleoside (5). PCM Activity. Aliquots of each thawed disrupted cell preparation were incubated for 10 min at 37'C in a reaction mixture containing 80 Other Chemicals HIM sodium acetate buffer, pH 6.0, and 5 x 10"' M 5-adenosyl-L- S-Adenosyl-L-[mefA>>/-3H]inethionine (specific activity, 14 Ci/mmol) (mcíA.v/-'HlnicthioniiK1(10 ^Ci/nmol) in the presence and absence of a was purchased from ICN Radiochemicals, Irvine, CA. L^methyPH] saturating concentration of the exogenous substrate, gelatin (2 mg/ml). methionine (specific activity, 80 Ci/mmol), and [2,8,5'-3H]adenosine The reactions were terminated by the addition of 1.0 ml of 10% (specific activity, 51 Ci/mmol) were purchased from New England trichloroacetic acid, followed by the addition of 100 ^1 of 1% bovine Nuclear Corp., Boston, MA. serum albumin as a coprecipitant. After centrifugation at 20,000 x g for 10 min, the supermit atils were removed, and the precipitates con MNB Tumor Cell Culture taining methylated proteins were dispersed in 0.3 ml of borate buffer (pH 11) containing 2.7% methanol as a carrier. The released methanol MNB tumors were grown in male A/J mice after s.c. implantation was extracted into 3 ml of a 3:2 toluenensoamyl alcohol mixture. After of 10" viable tumor cells. Tumors were excised from mice under sterile centrifugation, the organic (upper) phase was divided into two I ml conditions (14 to 21 days after implantation) and placed in sterile fractions. One portion was added to 10 ml of Omnifluor, and the total culture medium (RPMI-1640) supplemented with 10% fetal calf serum, radioactivity was measured by liquid scintillation counting. The other 1-ml fraction was evaporated to dryness at 37°Cunder a stream of air, 10,000 units/ml penicillin, and 100 /¿g/mlstreptomycin. The tumor was minced with a sterile scalpel in a IViri dish containing medium to and the nonvolatile radioactivity was determined. The difference be prepare a suspension of MNB cells. The cells were grown in monolayers tween the two counts represents the amount of tritiated methylester in 75 cm2 Falcon tissue flasks containing 20 ml of medium per flask in formed during the incubation. Virtually identical results were obtained an atmosphere of 5% CO2/95% air at 37"C. Subcultures were initiated when the 3H-labeled methanol released from the precipitated protein by the addition of 1 x 10'' MNB cells to each flask, and the medium was collected via a microdistillation method similar to one previously was changed every 3 days or as required, depending upon the experi described (28). mental conditions. Unless stated otherwise all statistical comparisons AdoHcy Hydrolase Activity. The assay of AdoHcy hydrolase activity were performed using a one-way ANOVA design, utilizing the Bonfer- in disrupted MNB cellular preparations was performed in the hydrolytic roni test of critical difference between group means (26). direction utilizing a previously described method (3). This assay re quired synthesis of the radiolabeled substrate [2,8-3H]AdoHcy, as de scribed previously (29). [2,8-3H]Adenosine was diluted to a specific Experimental Design activity of 32.2 nC\/^mo\ by the addition of unlabeled Ado, and the All drugs were added to tissue culture flasks at the initiation of each mixture was evaporated to dryness. The residue was dissolved in 1 ml subculture (zero time), and counts of viable cells were performed at 24, of hexamethyl phosphoramide followed by the addition of 0.1 ml of 48, and 72 h. In those experiments designed to measure AdoHcy thionylchloride. The mixture was incubated overnight at room temper ature, and the 5'-chloroadenosine product was generated by the addi hydrolase activity following AD treatment, cells were grown for 24 h tion of 5 ml of distilled water. The entire mixture was applied to a prior to drug addition, so as to obtain adequate amounts of protein for Dowex 50 x 2 (II') column, and the 5'-[2,8-3H]chIoroadenosine was enzyme analysis. Drugs were dissolved in a vehicle consisting of 3.3% eluted with 50 ml of 1.5 N NH4OH. The 5'-[2,8-3H]chloroadenosine dimethyl sulfoxide in 50% ethanol. Final concentrations of the diluent ranged from 0.33 to 0.0033% dimethyl sulfoxide and 0.5 to 0.005% was dissolved in 1 ml of 2 N NaOH containing 50 mg of L-homocysteine ethanol, which did not affect MNB cell growth. Each experiment had thiolactone. This mixture was applied to a Dowex 50 x 2 (M U )column its own control to which only the vehicle was added. Experimental and washed with distilled water to remove unreacted homocysteine. flasks contained drug concentrations ranging from 10" to IO' M. After The [2,8-3H]AdoHcy synthesized by this procedure was 67% pure. A various periods of incubation, each flask was agitated, and the cells preparation of >99% purity was achieved by paper chromatography after separation of the [2,8-3H]AdoHcy from the other radioactive were removed for counting and quantitation of the appropriate enzyme activity. Viable cell counts were determined on a regular basis at the contaminant in a solvent system consisting of normal butyl alco- hol:H20:methanol:NH4OH (55:25:20:2). The final purified [2,8-3H)- same time each day after dilution with 0.1 % trypan blue in 0.9% sodium chloride and then counted using a Speirs-Levy eosinophil counting AdoHcy product had a calculated specific activity of 15 /tCi//imol after chamber (C. A. Hausser, Philadelphia, PA). Average cell viability determination of the isotopie content and absorbance at 254 nm of the chromatographically separated product. ranged between 95 and 98% in control flasks as assessed by trypan blue MNB cells (IO7) were sonicated at 4'C in cold reaction mixture dye exclusion. This value decreased in treated flasks, and only viable cells were used for statistical comparisons and calculation of enzyme buffer, followed by centrifugation at 8000 x g for 2 min to obtain a activities. Culture flasks containing less than 85% viable cells were not plasma membrane-free supernatant. The incubation mixture (final vol included in any of the analyses. ume, 500 fi\) contained 20 DIMsodium phosphate buffer (pH 7.4), 1.0 HIMEDTA, 2 mM dithiothreitol, 40 /IM [2,8-3H]AdoHcy, and 1 unit of calf intestinal adenosine deaminase (Sigma). The hydrolytic reaction Determination of Enzyme Activity was initiated by the addition of 100 /

Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 1987 American Association for Cancer Research. EFFECT OF ADENOSINE ANALOGUES IN MNB CELLS of [2,8-3H]inosine following incubation and Chromatographie separa tion was consistently 90 to 95%. Ribonucleotide Redactase Activity. Ribonucleotide reductase activity 400 was determined using the method of Tsai and Hogenkamp (9). Each incubation mixture contained 100 DIM dimethylglutarate buffer (pH 7.5), 40 mM dithiothreitol, 0.06 imi adenosylcobalamin. 0.1 HIM[5- £ 300 111< I>r. and variable amounts of cellular protein in a final volume of 50 (¡I-Aftera 10-min incubation at 37°C,each reaction was terminated i

by freezing in a dry ice-isopropanol bath and stored at —70'Cprior to O assay. The radioactive product. | 'I I|d( I)!'. was separated from other £ 200 reaction constituents by thin-layer chromatography on polyethylenei- e o mine cellulose using 0.5 M LiCI (pH 7.0) saturated with boric acid as o the solvent system. Following chromatography, the radioactive spots comigrating with authentic dCDP were cut from the plate and eluted 100 with 1 ml of LiCI-borate solvent, and the eluate was counted in a liquid scintillation counter.

PCM Activity in Intact MNB Cells -7 -6 -3 Adherent MNB cells were incubated with L-[3H]methionine (10 i<(i ml; 15 fi( i nmol) in methionine-free RPMI medium (final methionine concentration, 1.9 x 10"* M) for 60 min at 37'C. After pulsing, the Adenosine Dialdehyde(log M) radioactive medium was discarded, and the cells were washed with 10 Fig. 1. Effect of adenosine dialdehyde on PCM activity of disrupted MNB ml of RPMI containing 0.1 HIMunlabeled L-methionine, into which cells. Disrupted cell suspensions were prepared from cultured MNB cells after incubation for 72 h in the presence of varying concentrations of AD. PCM activity the cells were dislodged by gentle agitation. After centrifugation at 50 was determined utilizing gelatin as the exogenous substrate (see "Materials and x g for 10 min, the cell pellets were washed with isotonic phosphate- Methods"). Enzyme activity is expressed as the percentage of control where control PCM activity = 62 ±7.1 pmol [3H]methanol/30 min/106 cells or 385 ± buffered saline (pH 7.4) and recentrifuged, and the pellets were dis 44 pmol [3H]methanol/30 min/mg protein (mean ±SE: n = 20). Points, mean of rupted by homogenization following the addition of 1 ml of 10% data from 4 to 8 individual experiments; bars, SE. **, significance greater than trichloroacetic acid containing 10 HIMunlabeled L-methionine. Precip control, IO"'. 10"'. or 10—MAD groups (P < 0.01). itated methylesters were quantitated as described above after hydrolysis in 1 M borate buffer (pH 11.0), containing 2.7% methanol. Table 1 Effect of adenosine analogues on PCM activity in disrupted MNB cell preparations RESULTS activity AnalogueAdenosine (% ofcontrol)*115± Effect of Nucleoside Analogues on Protein Carboxylmethyltrans- dialdehydeDeazaadenosine5-AdenosylhomocysteineSinefunginConcentration(M)io-*10-' 16»(»Y ferase, .S-Adenosylhomocysteine Hydrolase and Ribonucleotide 294 ±75 (6)' io-io-« 186 ±59(6)98 Reductase Activities of Disrupted and Intact Murine Neuroblas toma Cells io-' ±3 (4) io-io-* 105 ±10(5) Protein Carboxylmethyltransferase Activity. The effects of 157 ±24(5Y98 nucleoside analogues on enzyme activity were determined after io-' ±6 (3) incubation of MNB cells for 72 h, utilizing concentrations of 10-io-* 104 ±14(4) analogue previously shown to have significant cytotoxic effects 138 ±14(4)i on C-1300 MNB cell growth over the same time period. The IS •18(4) AdoHcy hydrolase antagonist AD (10~8 to IO"6 M) exerted no io-'IO-PCM 122 ±36(6) effect on PCM activity. However, higher concentrations of AD 92 ±10(6) (10~5 M), which inhibited MNB growth 73% after 72 h, pro * Data expressed as percentage of control where PCM activity equals pmol |'H)methanol/30 mm 10' cells, after 72-h incubation of cultured MNB cells in duced a 350% increase in the PCM activity of suspensions the presence of each analogue. " Mean ±SE. prepared from disrupted MNB cells (Fig. 1). An increment in ' Numbers in parentheses, number of experiments at each concentration. PCM activity was observed whether cell suspensions were in 'Significantly greater than 10—MAD (/>< 0.01) by ANOVA. cubated in the presence (Fig. 1) or absence (data not shown) of ' Significantly greater than 10—or 10"' M DZA (P < 0.05) by ANOVA. the exogenous substrate gelatin. This stimulatory response was present when enzyme activity was normalized to either cell determined in the presence or absence of gelatin. number or protein content. The competitive AdoHcy hydrolase The increase in PCM activity of cell suspensions prepared antagonist. DZA. increased PCM activity 57%. However, this from disrupted MNB cells that had been incubated with AD elevation was observed only at a concentration of 10 ' M and led us to examine whether protein carboxylmethyiation, deter required the presence of gelatin (Table 1). The competitive mined by pulse-labeling techniques in intact viable cells, was PCM antagonists sinefungin and AdoHcy had no effect on also modified by this analogue and if it exerted any effect at PCM activity in this preparation (Table 1). The following cytotoxic concentrations (10~8 to IO"5 M). Cultured MNB cells nucleoside analogues, previously evaluated for their cytotoxic were pulse labeled with the AdoMet precursor L-[methyl-3H]- potency (30) were also investigated: 5'-chloro, 5'-deoxyadeno- methionine following varying periods of incubation with AD sine; 5'-deoxyadenosine; 5'-chloro, S'-deoxy-ara-adenosine; (10~8 to 5 x 10~* M). A concentration-dependent inhibition of 3',5'-dichloro, 2',3',5'-trideoxyadenosine; methylfuryladenine; protein carboxylmethyiation, which persisted from 2 to 6 h, adenosine 5'-carboxylic acid; uridine dialdehyde; inosine di- was observed after incubation with 10~" to 10"' M AD (Fig. 2). aldehyde; cytidine dialdehyde: guanosine dialdehyde; reduced The suppressive effect of AD on protein carboxylmethyiation adenosine dialdehyde; and Ado. These compounds exerted no was no longer apparent after longer periods of drug exposure effect on PCM activity when determined in disrupted MNB cell (24 to 72 h). In fact, a significant increment in protein carbox preparations after 72-h exposure to each analogue, whether ylmethyiation was observed following extended incubation of 3658

Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 1987 American Association for Cancer Research. EFFECT OF ADENOSINE ANALOGUES IN MNB CELLS

400 100

300

200

100

20 40 60 80 Incubation Time (hr) Incubation Time (hr) Fig. 3. Effect of AD on the AdoHcy hydrolase activity of disrupted MNB Fig. 2. Time course of protein carboxylmethylation in intact MNB cells after incubation with AD. Cultured MNB cells were pulse labeled with L-[JH]methio- cells. MNB cells were incubated with AD for varying time periods, and disrupted nine for 60 min after varying periods of incubation with AD (10~* to in MI. cell suspensions were prepared as described. AdoHcy hydrolase activity was determined using [3H]AdoHcy as substrate (see "Materials and Methods"). En and [3H]-methylester formation was determined (see "Materials and Methods"). zyme activity is expressed as percentage of control where control AdoHcy hydro Points, mean for data obtained in 4 to 14 individual experiments and expressed lase activity = 339 ±25 pmol [3H]inosine/mg protein/min. Points, mean for as percentage of control carboxylmethylation activity; bars, SE. *, significance less than 10"* M AD (P < 0.05); **, significance greater than 10"' M (P < 0.05); duplicate determinations in 3 individual experiments; bars, SE. AHH, AdoHcy hydrolase. *, significance less than control activity il' < 0.01); **, significance ***, significance greater than IO"7 M or 5 x IO"7 M (P < 0.01). less than control or IO"7 M AD (P< 0.01). MNB cells with AD concentrations from IO"6 to IO"5 M (Fig. 400 2). In the [3H]methionine pulse-labeled cell model, 10"', 5 x 10~6, and 10~5 M AD concentrations were equally effective at stimulating carboxylmethylation, a response seen only with H» M AD in disrupted cells (Fig. 1). 300 Adenosylhomocysteine Hydrolase Activity. Incubation of MNB cells with AD (10"' M) decreased AdoHcy hydrolase o activity of disrupted cell suspensions within 30 min (Fig. 3). This response was dose dependent; AD (10~7 M) reduced en S o zyme activity 62%, whereas AD (10~* M) diminished activity 200 98% 2 h after addition of drug. Prolonged periods of incubation (24 to 72 h) with AD (10"* or 10~7M) attenuated the inhibitory effect of this analogue on AdoHcy hydrolase activity; e.g., enzyme activity at 24 h was 30% of control, at 48 h 45%, and o 100- at 72 h 70%, after exposure to AD (10~7 M). Ribonucleotide ReducíaseActivity. Incubation of MNB cells with AD (10~7 M and 10~6 M) for periods up to 72 h did not cause any significant changes in the ribonucleotide reducíase activity of disrupted cell suspensions (Fig. 4).

DISCUSSION Incubation Time (hr) Fig. 4. Effect of AD on the ribonucleotide reducíaseactivity of disrupted MNB Adenosine analogues have been reported to inhibit AdoHcy cells. MNB cells were incubated with AD (IO"7 or 10"* M) for varying time periods, and ribonucleotide reducíaseactivity was determined in disrupted cell hydrolase by competitive and noncompetitive mechanisms. preparations (see "Materials and Methods"). Points, mean of duplicate determi Compounds known to act as competitive antagonists include nations in 3 individual experiments and expressed as percentage of observed DZA, 2'-deoxyadenosine, and l-/3-D-arabinofuranosyladenine, control activity at each separate time point; bars, SE. whereas AD, neplanocin A, and 3-deazaneplanocin inhibit AdoHcy hydrolase in a time-dependent, noncompetitive man Iular mechanism(s) of action have not been clearly defined. ner (3-7). Recently performed studies on the structure-action relation Investigative interest in AD is due, in part, to the presence ship of nucleoside analogues have demonstrated that the growth of certain structural components (e.g., 3'-keto group) which are of MNB cells in tissue culture is effectively suppressed by Ado similar to the proposed functional moieties of the 3'-keto analogues.4 These effects on cell growth exhibit a high degree intermediates thought to be generated during the AdoHcy hy- of structural specificity. Only those nucleosides possessing an drolytic reaction (2, 5). Although AD appears to inactivate intact adenine base and di-keto functions on the 2'- and 3'- AdoHcy hydrolase in murine L929 leukemia cells (8) and positions of the ribose were effective cytotoxic agents. Blockade cultured murine neuroblastoma cells (see Fig. 3), its intracel- of kinase-mediated intracellular phosphorylation by substitu- 3659

Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 1987 American Association for Cancer Research. EFFECT OF ADENOSINE ANALOGUES IN MNB CELLS tion of a —CH3 group at the 5'-position or oxidation of the expression of the "differentiated" MNB phenotype and its 3'—4'—C—C—bondyielded AD derivatives whose cytotoxic apparent association with enhanced protein carboxylmethyla potencies were similar to that of the parent compound. These tion are quite consistent with other observations, demonstrating data suggest that intracellular 5'-phosphorylation of AD is not increases in PCM and its methylacceptor proteins after treat a prerequisite for the expression of its cytotoxic effect. ment of N1E-115 neuroblastoma cells with differentiating In order to examine whether inhibition of specific enzymes agents (22). The significance of this association between AD- correlated with the cytotoxic effects of the purine nucleoside induced differentiation and regulation of cellular protein car analogues, MNB cells were incubated with AD and other nu- boxylmethylation remains to be defined. cleosides at appropriate cytotoxic concentrations of each agent. This hypothesis was tested by measuring the changes in PCM, ACKNOWLEDGMENTS AdoHcy hydrolase, and ribonucleotide reducíase activity of disrupted MNB cells following incubation with the aforemen The authors gratefully acknowledge the excellent technical assistance tioned analogues. Pretreatment with AD produced a 350% of Judy Bratt. Special thanks are due to Miriam Stormo for her skillful elevation in PCM activity after 72 h of incubation (Fig. 1). The preparation of the manuscript. increment in enzyme activity noted above was associated with a consistent increase in protein content from 150 to 175 /ig/ REFERENCES IO6cells in control cultures to 240 to 260 /tg/106 cells in AD- 1. Bon-hardi. R. T. S-Adenosyl-L-methionine-dependent macromolecule meth- treated cultures. The increase in PCM activity was statistically yltransferases: potential targets for the design of chemotherapeutic agents. J. greater than control whether enzyme activity was expressed per Med. Chem., 23: 347-357, 1980. cell number (P < 0.01) or protein content (P < 0.05). These 2. Palmer, J. L., and Abeles, R. H. The mechanism of action of 5-adenosylhom- ocysteinase. J. Biol. Chem., 254:1217-1226, 1976. observations suggested that the increment in PCM activity 3. Chiang, P. K., Richards, H. H., and Cantoni, G. L. S-Adenosyl-L-homocys- produced by AD might, in part, be due to nonspecific increases teine hydrolase: analogues of S-adenosyl-L-homocysteine as potential inhib in protein synthesis. itors. Mol. Pharmacol., 13:939-947, 1977. 4. Hershfield, M. S. Apparent suicide inactivation of human lymphoblast .V The effects of AD on PCM activity were also examined in adenosylhomocysteine hydrolase by 2'-deoxyadenosine and adenine arabi- intact MNB cells utilizing AD concentrations previously shown noside. J. Biol. Chem., 254: 22-25, 1979. 5. Grant, A. J., and Lerner, L. M. Dialdehydes derived from adenine nucleosides to be cytotoxic 30. In these studies, protein carboxylmethylation as substrates and inhibitors of adenosine amino hydrolase. Biochemistry, was quantitatively estimated by pulse-labeling MNB cells with 18:2838-2842, 1979. L-[3H]methionine after incubation with AD. AD (10~8 to 10~6 6. Borchardt, R. T., Keller, B. T., and Patel-Thombre, U. Neplanocin A: a potent inhibitor of 5-adenosylhomocysteine hydrolase and of vaccinia virus M) significantly depressed protein carboxylmethylation (Fig. 2). multiplication in mouse L929 cells. J. Biol. Chem., 259:4353-4358,1984. The inhibitory effect persisted from 2 to 6 h following addition 7. Glazer, R. I., Hartman, K. D., Kunde. M. C., Richard, M. M., Chiang, P. K., Tseng, C. K. H., and Marquez, V. E. 3-Deazaneplanocin: a new and of AD and returned to control levels by 24 h (Fig. 2). This potent inhibitor of S-adenosylhomocysteine hydrolase and its effect on hu pattern of reduced carboxylmethylation paralleled the inhibi man promyelocytic leukemia cell line III,60. Biochem. Biophys. Res. Com tion of AdoHcy hydrolase and the accumulation of AdoHcy mun., 35:688-694, 1986. 8. Buriel. R. L., and Borchardt, R. T. Effects of adenosine dialdehyde on 5- that has been reported after treatment of murine L929 cells adenosylhomocysteine hydrolase and S-adenosylmethionine-dependent with AD (8). Recently, the time course of AdoHcy accumulation transmethylations in mouse L929 cells. Mol. Pharmacol., 25:418-424,1984. in MNB cells following incubation with 10~6 M AD has been 9. Tsai, R. K., and Hogenkamp, H. P. C. Affinity labeling of ribonucleotide reducíase by the 2',3'-dialdehyde derivatives of ribonucleotides. Arch. studied, and an 80-fold increase in intracellular AdoHcy con Biochem. Biophys., 226: 276-284, 1983. centration was observed." Supporting the concept that methyl- 10. Mirkin, B. L., O'Dea, R. F., and Hogenkamp, H. P. C. Effect of adenosine analogues on the growth and protein carboxylmethyltransferase activity of ation reactions are influenced by intracellular levels of AdoHcy, C-1300 murine neuroblastoma in tissue culture. In: A. E. Evans, C. F. these studies have demonstrated that AD produces a rapid D'Angio, and R. C. Seegar (eds.). Advances in Neuroblastoma Research, pp. time-dependent inactivation of AdoHcy hydrolase activity in 69-77. New York: Alan R. Liss, 1985. 11. O'Dea, R. F., Viveros, O. H., and Diliberto, E. J. Protein carboxylmethyla MNB cells. tion: role in the regulation of cell functions. Biochem. Pharmacol., 30:1163- An interesting response in the PCM activity of disrupted 1168, 1981. MNB cell suspensions (Fig. 1) and the protein carboxylmeth 12. Black, R. A., Hobson, A. C, and Adler, J. Regulation of the level of methylation of a protein involved in bacterial chemotaxis. In: E. Usdin, R. ylation of intact MNB cells (Fig. 2) was observed following 24- T. Borchardt, and C. R. Creveling (eds.), Biochemistry of S-Adenosylme- to 72-h exposure to AD. While short-term incubation inhibited thionine and Related Compounds, pp. 91-98. London: Macmillan Press, protein carboxylmethylation in intact cells (Fig. 2), a 200 to Ltd., 1982. 13. Wang, E. A., and Koshland, D. E., Jr. Methylation of receptors in chemotaxis 300% increment in PCM activity and protein carboxylmeth of bacteria. In: E. Usdin, R. T. Borchardt, and C. R. Creveling (eds.), ylation was noted after long-term treatment with AD at con Biochemistry of 5-adenosylmethionine and Related Compounds, pp. 99- centrations exceeding 10~6 M. This paradoxical response ap 106. London: Macmillan Press, Ltd., 1982. 14. O'Dea, R. F., Viveros, O. H., Axeirod, J., Aswanikumar, S., Schiffman, E., peared to be limited only to PCM, since AdoHcy hydrolase and Corcoran, B. A. Rapid stimulation of protein carboxylmethylation in leukocytes by a chemotactic peptide. Nature (I.onil.). 272:462-464, 1978. activity remained below control activity, and ribonucleotide 15. Venkatasubramanian, K., Mirata, F., Gagnon, C., Corcoran, B. A., O'Dea, reducíaseactivity was unchanged from controls after 72-h in R. F., Axeirod, J., and Schiffman, E. Protein methylesterase and leukocyte cubation with AD. chemotaxis. Mol. Immunni.. 17: 201-207, 1980. The present investigation has demonstrated that AD inhib 16. Van Waarde, A. Rapid, transit methylation of four proteins in aggregative amoebae of Dictyostelium discoideum as a response to stimulation with cyclic ited AdoHcy hydrolase and PCM activities during the first 6 h AMP. FEBS Lett., 149:266-270, 1982. following addition of drug. Prolonged incubation of MNB cells 17. Strittmatter, W. J., Gagnon, C., and Axeirod, J. Beta adrenergic stimulation of protein carboxylmethylation and amylase secretion in rat parotid gland. with AD, however, produced a 250 to 350% elevation in PCM J. Pharmacol. Exp. Ther., 207:419-424, 1978. activity, whether measured in intact cells or in disrupted cell 18. Eiden, L. E., Borchardt, R. T., and Rutledge, C. O. Protein carboxylmeth preparations. Morphological changes in MNB cells observed ylation in neurosecretory processes. In: E. Usdin, R. T. Borchardt, and C. R. after exposure to AD (10~6 M) were manifest by a 50 to 60% Creveling (eds.), Transmethylation, pp. 539-545. New York: Elsevier/North- Holland, 1979. increase in neurite expression after 72-h incubation.5 This 19. Clarke, S. Protein carboxylmethyltransferases: two distinct classes of en zymes. Annu. Rev. Biochem., 54: 479-506, 1985. 4 P. M. Gaffney, B. L. Mirkin, and R. F. O'Dea, unpublished observations. 20. Swanson, R. J., and Applebury, M. L. Methylation of proteins in photore- 5D. M. Barten, B. L. Mirkin, and R. F. O'Dea, unpublished observations. ceptor rod outer segments. J. Biol. Chem., 25Õ:10599-10605, 1983. 3660

Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 1987 American Association for Cancer Research. EFFECT OF ADENOSINE ANALOGUES IN MNB CELLS 21. Zuckerman, S. II.. O'Dea, R. F., Olson, J. M., and Douglas, S. D. Protein to Lens culinarii lectin. Biochem. Biophys. Res. Commun., 77: 1552-1558, carboxylmethylation during in vitro culture of human peripheral blood mon- 1977. ocytes and pulmonary alveolar macrophages. Mol. Immu mil., 19: 281-286, 26. Wallenstein, S., Zucker, C. L., and Fleiss, J. L. Some statistical methods 19g2 useful in circulation research. Circ. Res., 47:1-9, 1980. 22. Kloog, Y., Axelrod, J., and Spector, I. Protein carboxylmethylation increases 27- L0*^- O. H., Rosebrough, N J., Farr, A. L., and Randall, R. J. Protein in parallel with differentiation of neuroblastoma cells. J. Neurochem., 40: measurement with the Folm phenol reagent. J. Biol. Chem., 193: 265-275, 552—579IQR^ 1"3I. ,, ;. ••••"•'• . 28. Stock,J.B.,andKoshland,D.E.,Jr.Changingreactivityofreceptorcarboxyl 23. Duerre, J. A, and Fetters H. A. Protein carboxylmethylatjon-deinethylat.on durj bac,eria, ^ _, Bio, ^ 2¡6.,¿826_,o833, 1981. system in developing rat livers. Biochemistry, 24:6848-6854,1985. 29 Wang Y and Hogenkamp, H. P. C. Reinvestigation of the synthesis of T- 24. Fetters, H. A., Kelleher, J., and Duerre, J. A. Protein carboxyl methylation- deoxyadenosylhomocysteine. J. Org. Chem., 43:998-999, 1978. demethylation in rat thymocytes. Can. J. Biochem. Cell Biol., 63: 1112- 30. Mirkin, B. L., O'Dea, R. F., and Hogenkamp, H. P. Cytotoxic action of 1119, 1985. adenosine nucleoside and dialdehyde analogues on murine neuroblastoma in 25. Edstrom, R. D., Prody, C. A., and Hogenkamp, H. P. C. Halogenated tissue culture: structure-activity relationships. Cancer Res., 47: 3650-3655, nucleoside derivatives as inhibitors of the mitogenic response of lymphocytes 1987.

3661 Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 1987 American Association for Cancer Research. Effect of Adenosine Analogues on Protein Carboxylmethyltransferase, S-Adenosylhomocysteine Hydrolase, and Ribonucleotide Reductase Activity in Murine Neuroblastoma Cells

Robert F. O'Dea, Bernard L. Mirkin, Harry P. Hogenkamp, et al.

Cancer Res 1987;47:3656-3661.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/47/14/3656

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/47/14/3656. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 1987 American Association for Cancer Research.