Antiproliferative Effects of Inhibitors of

Val. 269, No. 45, Issue of November 11, pp. 27827-27632.1994 Printed in U.S.A.

THEJOURNAL OF BIOLWICAL CHEMISTRY

Antiproliferative Effects of Inhibitors of Deoxyhypusine Synthase INHIBITION OF GROWTH OF CHINESE HAMSTER OVARY CELLS BY GUANYL DIAMINES* (Received for publication, May 16, 1994, and in revised form, September 9, 1994)

Myung Hee Park$, EdithC. Wolff, Young Bok Lee,and J. E. Folk From the Enzyme Chemistry Section, Laboratory of Cellular Deuelopment and Oncology, NIDR, National Institutes of Health, Bethesda, Maryland 20892

form theintermediate, deoxyhypusine. Deoxyhypusine hyCertain guanyl diamines are effective inhibitors of dehydroxylation of this interoxyhypusine synthase (Jakus, J., Wolff, E. C., Park, M. droxylase (7) in turn mediates the H., andFolk,J. E. (1993)J. BioZ. Chem. 268,13151-13159), mediate to complete hypusine synthesis and eIF-5A maturathe first enzyme involved in the biosynthesis of the untion. Normally, hypusine biosynthesisoccurs efficiently, and no usualaminoacidhypusine (NC-(4-amino-2-hydroxybuaccumulation of either unmodified eIF-5A precursor (8, 9) or ty1)lysine). Evidence that hypusine is implicated in cell the deoxyhypusine-containingintermediate (3) is detectable. growth prompted this study of the cellular effects of Hypusine has been shown to be essential for the activity of these inhibitors.In Chinese hamster ovary (CHO) cells, eIF-5A in stimulating methionyl puromycin synthesis, an in inhibition of hypusine biosynthesis followed by progres- vitro model assay for translation initiation (10, 11).However, sive arrest in cellular proliferation was observed with the precise physiological function of eIF-5A in eukaryoticpro1,6both N-mono- andN,N'-bisguanylderivativesof tein synthesis, or in other cellular metabolism, is as yet undiaminohexane,1,7-diaminoheptane,andl,$?-diaminoknown (2, 12). The marked increasein hypusine synthesis oboctane. Cells treated with these compounds showed no served following mitogen treatment of human peripheralblood significant change in polyamine distribution, suggestis not mediated lymphocytes (13), together with a general correlation noted ing that the observed growth inhibition through an interference with polyamine metabolism. N- betweenhypusinesynthesisand cell proliferation inother guanyl-1,7-diaminoheptane,the most potent inhibitor ofmammalian cells (2, 14-17), suggested an important role for deoxyhypusine synthase both in vitro and in cells, ex- hypusine in cellular proliferation. Recently, a requirement for hypusine in yeast was implied by a complete loss of growth hibitedthehighestantiproliferativeactivitytoward potential following inactivation of both eIF-SA genes in this CHO cells. No early cytotoxic effects were observed with organism (18). Expression of at least one of the two eIF-5A thisinhibitor,and its antiproliferativeactivityappeared to be reversible. Transport studies showed that genes to produce an eIF-5A precursor protein (12, 18, 19)and W-guanyl-1,7-diaminoheptaneis actively taken up by its subsequentmodification to form hypusine (18)are required the polyamine transport system. Mutant CHO cells de- for yeast growth. Additional support for the involvement of fective in polyamine transport were found to be resis- hypusine in cell growth derives from studies in which cellular tant to growth inhibition by this compound. The find- spermidine, the polyamine substrate for hypusine formation, ingssuggestthattheantiproliferativeeffect of N- was depletedby the useof inhibitors of polyamine biosynthesis guanyl-l,7-diaminoheptaneis exertedintracellularly (20, 21x2 through inhibition of hypusine synthesis. Targeted inhibitionof the enzymes responsible for hypusine biosynthesis offers a promising means for controlling production of this aminoacid and thusfor further defining its specific Hypusine (N"(4-amino-2-hydroxybutyl)lysine)is a unique cellular role. In addition, selective inhibitors should provide a amino acid that is formed posttranslationally in only one pro- mechanism for regulation of cell proliferation, if indeed hytein, eukaryotic translation initiation factor5A (eIF-5A,' older pusine is essential for this process (22). We recently reported nomenclature eIF-4D) (Ref. 1and, for a review, see Ref. 2). Its (23) several effective inhibitors of deoxyhypusine synthase. biosynthesis, which occurs by modification of a single lysyl Among these are mono- and bisguanyl diamines structurally residue, involves two enzymes (3). Deoxyhypusine synthase related t o spermidine, which, inpreliminarystudies,were (4-6) catalyzes the transferof the 4-aminobutylmoiety of the found to be effective deterrents to hypusineformation in cells. polyamine spermidine to the €-amino group of the lysine to The most potent of these compounds is N1-guanyl-1,7-diaminoheptane (GC,), the K, value of which, 10 I", is approximately * A preliminary account of a portion of this work has been presented 400-fold lower than the K , of spermidine. Here we report the at the ASBMB meeting in WashingtonD. C. on May 22,1994 (Park, M. H., Wolff, E. C., Lee, Y. B., and Folk, J. E. (1994) FASEB J. 8, 1310 effects of mono- and bisguanyl derivatives of several diamines (Abstr. 305). The costs of publication of this article were defrayed in part on hypusine formation, protein synthesis, polyamine metabolism, and growth inCHO cells with special focus on GC,. by the payment of page charges. This article must therefore be hereby marked "aduertisernent"in accordance with 18 U.S.C.Section 1734 solely to indicate this fact. EXPERIMENTALPROCEDURES $ To whom correspondence should be addressed: NIH, Bldg. 30, Rm. Materials 211, 30 Convent Dr., MSC 4330, Bethesda, MD 20892-4330. "el.: 301496-5056; Fax: 301-402-0823. [1,8-3HlSpermidine.HCl(15 Ci/mmol)and ~-[4,5-~Hllysine (94 Ci/ ' T h e abbreviations used are: eIF-SA, eukaryotic translation initia- mmol) were purchased from DuPont NEN. N1,P-bisguanyl-1,6-diamition factor 5 A CHO, Chinese hamster ovary; 2, benzyloxycarbonyl; nohexane(GC,G),GC,, N1,N"bisguanyl-1,7-diaminoheptane(GC,G), MGBG, methylglyoxal-bis(guany1-hydrazone); GC,, N1-guanyl-1,6-dia- N1-guanyl-l,8-diaminooctane (GC,), and N'$P-bisguanyl-l,B-diaminoGC,, N1-guanyl-1,8- octane (GC,G) were synthesized as described previously (23). minohexane; GC,, N1-guanyl-l,7-diaminoheptane; diaminooctane; GC,G, N',N6-bisguanyl-l,6-diaminohexane;GC,G, N', W-bisguanyl-l,7-diaminoheptane; and GC,G, N1,h@-bisguany1-l,8-diaminooctane. A. Shirahata, personal communication. ~~

27827

27828

Antiproliferative Effects of Guanyl Diamines

The MGBG-resistant mutant of CHO cells (CHO-MG) that is defec- with 6 N HCl for 18 h at 108 "C. The amount of radiolabeled hypusine or tive in polyamine transport (24) and the corresponding wild type CHO deoxyhypusine in the hydrolysate was determined &er ion exchange cells (CHO-P5) werekindly provided by Dr. Wayne F. Flintoff, Univer- chromatographic separation (28, 30). For the samples labeled with sity of Toronto, Toronto, Canada. L3H1lysine, the trichloroacetic acid precipitates were washed once with ~-Guanyl-l,6-diaminohexane (GC,) was prepared through the fol- 5% trichloroacetic acid and then dissolved in 1 mlof 0.1 N NaOH. A lowing series of reactions: N~-bisbenzyloxycarbonyl-I,6-diaminohe~-portion of this solution was used forthe determination of radioactivity. ane, m.p. 125 "C, wasobtained in 90% yieldby the method described for Determination of Polyamine Pools-Untreated cells, or cells treated preparation of the bis-Z derivative of cadaverine (25). Attempts to con- with a guanyl diamine, were washed, harvested, and extracted with 5% vert this compound to mono-Z-diaminohexane by the published proce- trichloroacetic acid. The basic amine compounds in the trichloroacetic dure using HBr in glacial acetic acid (25) were unsuccessful. Mono-N- acid supernatant solution were resolvedby ion exchangechromatogra2-1,6-diaminohexane0.5 HzS04, m.p.160 "C dec.,was prepared by phy according to a published procedure (28) and quantitated fluoroshort term hydrogenation (5 min at 30 lbs/in2) of 10 g of the bis-Z metrically after postcolumn derivitazation with o-phthalaldehyde. compound in 100mlof a 1:l (v/v) mixture of methanol and dichloMeasurement of the Dansport of Spermidine and GC, romethane containing 1.4 equivalents of concentrated H,SO, using 0.5 r3H1spermidine or VHIGC, was added to dishes of cells with densities g of 10%palladium on charcoal catalyst. The compound wasobtained in near confluencybut still in exponential growth, either in the presence of 17%yield after two recrystallizations from watedmethanol. Its conver- or in the absence of unlabeled GC, or of unlabeled spermidine, respecsion to N1-Z,A@-guanyl-l,6-diaminohexane.0.5 H,SO, was accomplished tively. The cells werethen incubated at 37 or 4 "C, for periods up to 2 h. in 34% yield by guanylation in aqueous methanol using O-methyli- After the incubation period, the cells were washed twice with ice-cold sourea hydrogen sulfate and triethylamine. Removal of the Z group by phosphate-buffered saline containing 1 mM each of nonlabeled polycatalytic hydrogenation in methanol containing 0.5 equivalent of H,SO, amines putrescine, spermidine, and spermine. The washed cells were gave N1-guanyl-l,6-diaminohexane~HzS04~H,0, m.p.295 "C dec., in dissolved in 1 ml of 0.1 N NaOH, and aportion of this solution was used 48% yield from water/methanol/acetone, CIMS 159 (M + 1). for the determination of radioactivity. Determination of GC,-This compound was measured either fluoroC,Hl,N4~H,S0,~H20 metrically or by measurement of the radioactivity of the labeled compound after its separation by ion exchange chromatography (28). EluCalculated: C 30.64 H 8.08 N 20.42 tion of GC, was significantly later than spermine in a buffer mixture Found: C 30.54 H 8.14 N 20.29 containing nine parts of sodium citrate buffer (3.0 N Na'), pH 5.55, and N'-Guanyl-1,7-diamin0[7-~H]heptane ([3HlGC,), (-18 Ci/mmol) was one part of 2-propanol. prepared by vapor-phase HCl hydrolysis of its W-acetyl derivative formed by custom catalytic tritiation (Amersham Corp.) of 7-guanidiRESULTS noheptanonitrile in a mixture of glacial acetic acid and acetic anhydride Effects of Guanyl Diamines on the Synthesisof Hypusine a n d at 1 atm with platinum oxide catalyst. The 7-guanidinoheptanonitrile, m.p. 140-140.5 "C, was produced from 7-aminoheptanonitrile (26) by Protein in CHO Cells-The effects of each memberof a series of guanylation with 0-methylisourea hydrogen sulfate and triethylamine guanyl diamineson cellular synthesis of hypusine and protein in water. The product, isolated as the free base by extraction from basic are shown graphically in Fig. 1, A and B, respectively. For solution with dichloromethane, was dissolved in water, acidified with comparative purposes, the Kivalues of these guanyl diamines, acetic acid, dried in vacuo, and crystallized as the acetate salt from as measures of their efficacies as in vitro inhibitors of deoxyethanollether in 46% yield, FABMS 169 (M + 1). hypusine synthase, are given in Fig. 1C. Hypusine synthesis in

cells (Fig. L4)was followed by measuring the incorporation of radioactivityfrom[3H]spermidine into [3H]hypusine in the eIF-5A p r ~ t e i nThe . ~ specific radioactivities of the cellular spermidine pools after the 18-h incubation were not significantly C3HIGC, was purified by silica gel chromatography as outlined (27), different in cells treated with the guanyl diamines at the conexcept that only water was used to wash radioactive impurities from the column beforeelution of the labeled product. Upon ion exchange chro- centrations tested, compared with controls. Because deoxyhymatography (28), together with unlabeled carrier N1-guanyl-1,7-diami- pusine is efficiently converted to hypusine in cells by the action noheptane in a buffer mixture containing nine parts of sodium citrate of deoxyhypusine hydroxylase(3), and since no accumulation of buffer (3.0N Na'), pH 5.55, and one part of 2-propanol, 96% of the eluted labeled deoxyhypusine was observed in the cells in these exradioactivity was found to coelutewith the carrier compound. periments, thereduction in r3H1hypusineformed (Fig. lA)was taken as a measure of the degree of inhibition of deoxyhypusine Methods Cell Culture-CHO cells were cultured in the a-modification of Ea- synthase. Incorporation of [3H]lysine into total protein, measserves as a measure of (Fig. B), gle's medium supplemented with 10%fetal bovine serum, penicillin (50 ured in parallel cultures IU/ml), streptomycin (50 mg/ml), and 1 m~ aminoguanidine.Aminogua- general protein synthetic activity and, inaddition, provides a n nidine was included because in preliminary experiments it was found index of eIF-5Aprecursor protein synthesis, asdiscussed below. that the deoxyhypusine synthase inhibitor GC, was rapidly degraded in It is evident from Fig. L4 that the inhibition of hypusine certain cell culture media, especially those containing calf serum, as formation in CHO cells by the guanyl diamines is concentrameasured fluorometricallyafter separation by ionexchange chromatogmonoguaraphy. This loss of GC,,which occurredin media without cells as rapidly tion-dependent within the1-10 p~ range tested. The as in those with cells, wasprevented by inclusion in the media of 1mM nyl derivatives proved to bemore inhibitory than theirbisguaaminoguanidine, a well known inhibitor of amine oxidases (291, sug- nyl counterparts, and the compound in each group with the gesting that GC, is degraded by an amine oxidase. This activity was seven-member carbon chain was the most effective inhibitor. much lowerin media containing fetal calf serum but was not negligible. Inhibition of hypusine synthesis in cells was most pronounced Therefore in all of the experiments described in this paper, aminogua- with GC,, the strongest inhibitor of deoxyhypusine synthase in nidine was added to the culture medium. vitro, as reflected by its Ki value (Fig. 1C). A 1 p~ level of GC, Exponentially growing cells were incubated with the testcompounds as described in the figure legends.Cell viability was tested by the caused more than 97% inhibition of hypusine production in to this degree of inhibition, CHO cells in 18 h.4 In order achieve trypan blue exclusion method. Determination of Hypusine and Protein Synthesis-A radiolabeled the effect of GC, on deoxyhypusine synthase must have been precursor, [3Hlspermidineor [3H]lysine, was added to the cells 10-20 exerted almost immediately after its addition to the cells. Only C,Hl,N,.C2H,OZ Calculated: C 52.61 H 8.83 N 24.54 Found: C 52.43 H8.87 N 24.46

min after the addition of the compound being tested. After incubation for the indicated times, the cells were washed twice with phosphatebuffered saline and harvested, and the cellular macromolecules were precipitated with 5% trichloroacetic acid. In the case of the samples labeled with [3Hlspermidine,the trichloroacetic acid precipitates were washed twice with 5% trichloroacetic acid containing 1 mM each of nonlabeled putrescine, spermidine, and spermine and then hydrolyzed

It should be pointed out that the data shown in Fig. lA illustrate the effects of the guanyl diamines on synthesis of hypusine but do not accurately reflect the hypusine content of the cells. Lower concentrations of GC, were also effectivewith 90% inhibition at 0.3 p ~ 71% , at 0.1 p ~ and , 21% at 0.03 p ~ .

Antiproliferative Effects

27829

ofDiamines Guanyl TABLEI

Polyamine contents of CHO cells treated with guanyl diamines The guanyl diamines were added at a 10 p~ level to cultures of CHO cells growing at a density of -7 x lo6 cells/60-mm dish, and incubation was continued. After 24 h, the poIyamine levels were determined as

described under “Experimental Procedures.“ Inhibitor

Putrescine

Spermidine

None

0.6 0.8

8.6

2.2

7.5 8.0 8.9

Spermine

nmol / m g protein G c 6

4.3 4.4

8.8 8.3

GC, GC, GC,G GC,G GC,G

1.3 1.4 1.6 1.5

7.0

4.5

5.6 6.2 6.1 4.6

for inhibition of deoxyhypusine synthase invitro (Fig. 1C).For example, GC, and GC,, which have K, values significantly lower than the K, of spermidine (4JM I ), are strong inhibitorsof cellular hypusine formation, whereas GC, and GC,G, with K, values much greaterthan the K, of spermidine, are much less effective. In the caseof the latter inhibitors, themodest reduction in hypusine formation maybe largely due to thereduced amount of eIF-5A precursor protein synthesized (compare panels A and B of Fig. 1).With the stronger inhibitors,GC,, GC,, and GC,G, however, the effect seen offers a clear indication of efficient inhibition of deoxyhypusine synthase incells. Polyamine Leuels in Cells Deated with Guanyl Diamines-In view of the structural similarities of the guanyl diamines t o polyamines, it was considered that these compounds might affect enzymes (in addition to deoxyhypusine synthase) that are involved in polyamine biosynthesis or metabolism. However, no pronounced changes in the polyamine patterns of CHO cells were seen after treatment with the guanyl diamines, as KI (PW 24 0.01 0.24 35 5.7 1.7 shown in Table I. Some elevations in the levels of putrescine Gc6 GC7 GC6G GQ Gc$ Gc8G were observed in treated cells, but relatively small differences were found in the amounts of spermidine and spermine in Inhibitor treated and untreatedcells. FIG.1. The inhibition of hypusine synthesis (A) and of total Inhibition of Proliferation of CHO Cells by Guanyl Diaprotein synthesis( B )in CHO cells by guanyl diamines; comparison with the efficiency of these guanyl diamines as in vitro mines-Each of the compounds tested inhibited the growth of inhibitors of deoxyhypusine synthase(C). CHO cells were cultured CHO cells in a concentration- and time-dependent manner as described under “Experimental Procedures.” When the cells reached !Fig. 2). Consistent with theireffects on cellular hypusinesyna density of 1x lo6 cells/60-mm dish,the guanyl diaminewas added to twodishes at the indicatedconcentration.After 10-20 min, [1,8- thesis, the monoguanyl compounds were more effective in in3Hlspennidine(15 pCi/dish) was added to one of the dishes (A), and hibiting growth than thecorresponding bisguanyl compounds. t3HIlysine (15 pCi/dish) was added to the other dish ( B ) .After 18 h of GC,, the most effective inhibitor of deoxyhypusine synthesis incubation, the amounts of radioactivity in protein-bound hypusine andboth in vitro and in cells, exhibited the greatest antiproliferatotal protein were measured as described under “ExperimentalProcedures.” The K, values for the inhibition of deoxyhypusine synthase tive activity. In the24-h period after the addition of the inhibitor, the cell activity in vitro (C) were reported previously (231, except for the K, numbers continued to increase in all cases, albeit a t different value for GC,. rates. Upon continued incubation (48-72 h), further inhibition negligible changes in protein synthesis andin DNA synthesis of growth was observed, depending on the compound and the (asmeasured by r3H1thymidineincorporation) were seen withinconcentration. Whereas untreated cells approached maximum 3 h after addition of 1p~ GC, (data notshown). During 18 h, 1 density (3.5 x lo6 celld60 mm dish) by 72 h in this period PM GC, exerted only a small inhibitionof total protein synthe- (48-72 h), total arrestof growth occurred at concentrations as sis (-lo%, Fig. 1B). Clearly in this time period, a large differ- low as 1and 3 PM for GC, and GC,, respectively. Levels 210 VM ential existsbetween the inhibitionof hypusine formation and were requiredfor total growth suppression by other compounds this circum- (GC,, GC,G, GC,G, and GC,G). In the concentration range inhibition of general protein synthesis. Under bestance, a significant accumulation of the unmodified eIF-5A tween l and 10 p ~ there , appeared tobe no early toxic effects precursor proteinwould be expected. Indeed, whereas this pre- of these compounds, and a large portion of the treated cells cursor was not detectable in untreated cells, CHOcells treated remained viable for 3-4 days in culture. with 1 p~ GC, for 20 h were found to contain -10 pmol of In Fig. 3 is shown a more detailed examination of the time eIF-5A precursor/mg of protein, an amount comparable with course of growth inhibition by the most effective of the comthat found in CHO cells depleted of spermidine (9) by a 42-h pounds, GC,, at levels of 0.2-3 p ~ The . IC,, for growth inhibitreatment with a-difluoromethylornithine,an inhibitor of poly- tion a t 48 h after GC, addition was estimatedas approximately amine biosynthesis (31, 32). 0.3 p ~ . Aat concentration of 1p ~cytostasis , wasobserved after As noted in Fig. L4, the individual guanyl diamines exhibited the initial doubling period. When the culture medium was remarkedly different degreesof inhibition of hypusine formation placed with a GC,-free medium at 96 h, cell growth was rein thesecells. This is as expected onthe basisof their K, values sumed andcontinued until thecells reached confluence, show-

IC

27830

Antiproliferative Effects of Guanyl Diamines

None

10

FIG.2. The inhibition of growth of CHO cells by guanyl diamines. CHO cellswere cultured as described under “Experimental Procedures.” The guanyl diamine wasadded at 1, 3, and 10 PM levels to the media at 24 h (vertical arrow) when the cells had reached a density of -3 x IO5 cellsf60-mm dish. After an additional 24 and 48 h of incubation, the cells werecollected after trypsinization. Cell numbers (expressed on a logarithmic scale) were determined by the use of a Coulter counter.

GC7G None

48

2424

48 72 72

48

24

I

GCg G

I

None1

72

Time (hours)

in Fig. 3 ) than in a GC,-free medium containing 1 mM aminoguanidine (data not shown). Cellular Dunsport and Intracellular Stability of GCFWith the use of radiolabeled GC, and spermidine, we were able to show that GC, is transported efficiently and actively into CHO cells by way of the polyamine transport system (Fig. 4). The None affinity of this transport system for GC, appears to be lower t O.* IrM than for spermidine, judging from the fact that inhibition of GC, uptake by spermidine (Fig. 4A)is greater than the reverse (Fig. 4B). Consistent with this mechanism for GC, transport U are the results shown in Fig. 4, C and D, in which CHO cells X In defective in polyamine transport were tested for GC, and sperb r midine uptake. The cells used, a mutant strain of CHO cells (24), are resistant to MGBG, a cytotoxic inhibitor of S-adenosylmethionine decarboxylase known to be transported by the polyamine transporter. This mutant, CHO-MG, was reported to removal be deficient in the transportof polyamines, but not in general 1 transport of unrelated materials (33). As expected, CHO-MG 2448 72 96 120 144 168 proved to be defective in the uptakeof GC, (Fig. 40)and, infact, was able togrow normally inthe presence of a high level (10PM) Time (hours) of GC, (data not shown). It seemslikely, based on this finding, FIG.3. T h e concentration dependence of inhibition of growth of CHO cells by GC,. Cells werecultured as described under “Exper- that transport of GC, into cells is required for its effects. When CHO cells were incubated withr3H]GC, (15 pCi) for 20 imental Procedures.”The indicated concentration of GC, was added to the cells at 24 h, at a density of -2 x lo5 cells/60-mm dish. Cell numbers h and the cells were collected, washed, and extracted with 5% (expressed on a logarithmic scale) were determined as in Fig. 2 . Two trichloroacetic acid,virtually all the radioactivity in the extract sets of cells were treated with GC, at 1 p~ at 24 h. In one set, the at the position of GC,. Thus, GC, was found to chromatograph medium was changed atomedium freeof both GC, and aminoguanidine at 96 h as indicated by the dashed arrow and dashed line. In the other appears not to undergo any significant metabolic conversion largely intact in these cells. The release of set, the medium was not changed. Incubation was continued for 72 but rather to remain additional hours, and cell numbers were determined during this time. r3H]GC, from CHO cells preloaded with this compound (0.2-1 p ~ 2-20 , h) wasalso followed. After a 20-h chase in a medium ing that the growth inhibition is reversible (Fig. 3). Still, the free of GC, and aminoguanidine, approximately 50% of the growth during the first24-h period of recovery from 1 p~ GC, radioactivity was found in the medium, indicating that the was slower than thatof the untreatedcells, probably because of efflux of GC,, like that of polyamines, is slow. The amount of the low rate of efflux of GC, from cells, as discussed in the next radioactivity exported was even less if the chase medium consection. The reversal of the growth inhibition was more effec- tained 1 m~ aminoguanidine.Thesefindings are consistent tive in media lacking both GC, and aminoguanidine (as shown with the delayed recovery of growth rates described above.

f

d

27831 . efficient uptake byway of the trations as low as 1 p ~ Its polyamine transporter, together with its extremely high affinA. 13H]Spd Uptake B. ijH]GC, Uptake ity for deoxyhypusine synthase, probably accounts for the ef30 . the radioacti~ty associfectiveness of this i ~ i b i t o rBecause ated with cells at 20 h after exposure to [‘HIGC, is entirely in the form of it seems clear that theintracellular effects of 20 and not by its potential this compound are caused by m ~ t a b o l i ~Furthermore, s. the relatively minor effects of 1p~ fO GC, on DNA and protein synthesis, up to 3 h after its addition, suggest that this guanyl diamine does not interfere directly with macromolecular syntheses but rather that there is a gradIt ual cellular change that leads to suppression of proliferation. 0.5 1.0 2.0 0.5 7.0 2.0 is reasonable t o assume that at the time GC, is added to cells Time (hours) there exists sufficient pre-formed mature eIF-5A to sustain cell growth for a protracted period of time since the half-life of 20 eIF-SA in mammalian cells appears tobe relatively long (8, 17, Uatake c. 13HlSpd 34). Hence the lag time for suppression of cell growthof 18-24 15 h seen in Figs. 2 and 3 may represent the time needed for reduction of eIF-SA below the threshold level through normal intracellular degradation and dilution by growth. Since the 10 precise function of eIF-5A is unknown, it is, of course, impossible to delineate the sequence of cellular events that leads to growth arrest by GC,. Even though the effects of GC, on pro5 cesses other than hypusine synthesis cannot be excluded, the correlation seen here between inhibition of hypusine synthesis and cell growth is consistent with deoxyhypusine synthase as the primary cellular target of GC,. In contrast to GC, and GC,, which are the most potent inhibitors of deoxyhypusine synthase, growth suppression by GC,, GC,G, and GC,G may be due to causes other than inhibiIncubation Conditions FIG.4. Cellular transport of GC, in CHO cells and CHOmMG tion of hypusine synthesis. These latter compounds all discells. CHO cells were cultured as described under ‘Experimental Pro- played strong inhibition of growth at 10 p~ without a commencedures,” ‘Ib cells, growing at a density of -2 x lo6 cells/60-mm dish surate reduction in synthesisof hypusine. Intervention in other were added the following: 15 pCi of 13HJspe~idine (0.25 p ~together ) with ( ) or without 10) 5 p unlabeled GC, (A); 22 pCi of 13HJGC,(0.3 cellular processes in addition to hypusine synthesis probably p ~ together ) with (W or without (x) 5 p~ unlabeled spermidine(B), and contributes to the arrestof growth by these compounds. incubation was continued. Radioactivity in the cells was measured as An important aspect of the growth inhibition by the guanyl outlined under “Experimental Procedures”; theamounts indicated are diamines reported here is its lack of dependence on depletion of the total radioactivity fin dpm) incorporated per 2 x lo6 cells. Spd, spermidine. In C and D the experiments describedin A and B, respec- spermidine or the other polyamines. It i s well estab~ished that tively, were repeated, except in each case two sets of cells, wild type certain polyamine antimetabolites (31, 321, for example a-di(CHO-P5) and CHO-MG were used; unlabeled spermidine and unla- fluoromethylornithine, MGBG, and bisethylnorspermine (35), ; were at 4 or 37 “C for 20 halt cell proliferation through depletion of the polyamine sperbeled GC, were added at 50 p ~ incubations min. midine. Although recent studies suggest that thegrowth arrest following spermidine depletion is a consequence of hypusine DISCUSSION deprivation (20, 21); it is likely that depletion of spermidine The present study, undertaken with the notion that hypusine also leads t o disturbance of a number o f other cellular promay play a vital role in cell proliferation, was designed to cesses. If this disruption in cell proliferation by polyamineandetermine if in vitro inhibitors of deoxyhypusine synthase timetabolites is primarily due to the lack of hypusine or eIF-5A, could be used in cells to prohibit both hypusine formation and selective inhibition of deoxyhypusine synthase should accomgrowth and, if so, t o establish t o what extent the two events plish the same end without alteration in other cell functions could be related. The findings confirm the effectiveness of cer- that accompany depletion of spermidine. In this regard, the tain guanyl diamine inhibitors, particularly GC,, in cells and inhibitors that specifically target deoxyhypusine synthase, but provide substantial supportfor a critical role of hypusine (and not other polyamine biosynthetic or metabolic enzymes, may its precursor, the polyamine spermidine) in eukaryotic cell have advantages over the polyamine antimetabolites. proliferation. The information given in this report is limited t o studies Experiments in which cells were depleted of either eIF-5A carried out with CHO cells. However,growth inhibition by guaprecursor protein (12, 18, 19) or spermidine (20, 21); while nyl diamines was observed with other mammalian cells, includsupplying information consistent with a strategic role of hy- ing a number of human cancer cell lines. Sensitivity to the pusine, did not eliminate the possibility that the precursor guanyl diamines varied with the cell line. Although the basis protein, or the polyamine, per se, functions in a vital manner. for the differences in sensitivity is not yet fully understood, it The finding reported here that the deoxyhypusine synthase could berelated todifferent rates of cellular uptake andefflwr, inhibitor GC, curtails mammalian cell growth without deplet- to different intracellular target(s) andlor to differences in the ing cellular levels of spermidine or of the eIF-SA precursor intracellular metabolism of the compounds. These factors may protein adds support to our contention that hypusine plays a form the basis of a selective inhibition and open the possibility key role in cellular proliferation in eukaryotes (22). for practical application of these compounds as chemotherapeuGC,is a remarkable inhibitor of cellular hypusine f o ~ a t i o n , tic agents againstthose tumor cells with, for example, a highly essentially abolishing production of deoxyhypusine a t concen- efficient polyamine transport system. 40

GC,,

1

1

GC,

27832

Diamines Guanyl Antiproliferative of Effects REFERENCES

1. Cooper, H. L., Park, M. H., Folk, J. E., Safer, B., and Braverman, R. (1983) Proc. Natl. Acad. Sci. U.S. A. SO, 1854-1857 2. Park, M. H., Wolff, E.C., and Folk, J. E. (1993) Biofactors 4,95-104 3. Park, M. H., Cooper, H. L., and Folk, J. E. (1982) J. Biol. Chem. 267, 72177222 4. Murphey, R. J., and Gerner, E. W. (1987) J. Biol. Chem. 262,15033-15036 5. Park, M. H., and Wolff, E. C. (1988) J. B i d . Chem. 263,15264-15269 6. Wolff, E. C., Park, M. H., and Folk, J. E. (1990) J. Biol. Chem. 286,47934799 7. Abbruzzese,A,, Park, M. H., and Folk, J. E. (1986) J. Biol. Chem. 261,30853089 8. Duncan, R. F., and Hershey, J. W. B. (1986) J. Biol. Chem. 261, 12903-12906 9. Park, M. H. (1987) J. B i d . Chem. 262,12730-12734 10. Smit-McBride,Z.,Schnier, J., Kaufman, R. J., and Hershey, J. W. B. (1989) J. Biol. Chem. 264, 18527-18530 11. Park, M. H. (1989) J. Biol. Chem. 264, 18531-18535 12. Kang, H. A,, and Hershey, J. W. B. (1994) J. Biol. Chem. 269,3934-3940 13. Park, M. H., Cooper, H. L., andFolk, J. E.(1981) Proc. Natl. Acad. S c i U. S. A. 78,2869-2873 14. Cooper, H. L.,Park, M. H., and Folk, J. E. (1982) Cell 29,791-797 15. Chen, R-Y (1983) Biochim. Biophys. Acta 756,395402 16. lbrrelio, B. M.,Paz, M. A., and Gallop, P. M.(1984)Exp. Cell Res. 154,454-463 17. Gerner, E. W,, Mamont, P. S.,Bernhardt, A., and Siat, M. (1986) Biochem. J. 239,379-386 18. Schnier, J., Schwelberger, H., Smit-McBride,Z., Kang, H. A., and Hershey, J. W. B. (1991) Mol. Cell. Biol. 11,31053114 19. Wiihl, T., Klier, H., Ammer,H., Lottspeich, F., and Magdolen, V. (1993) Mol. & Gen. Genet. 241,305-311 20. Byers, T.L., Ganem, B., and Pegg, A. E. (1992) Biochem. J. 287, 717-724

21. Eyers, T.L., Wiest, L., Wechter, R. S., and Pegg, A. E. (1993) Biochem. J. 290, 115-121 22. Park, M. H., Wolff, E. C., and Folk, J. E. (1993) Zknds Biochern. Sci. 18, 475-479 23. Jakus, J., Wolff, E. C., Park, M. H., and Folk, J. E. (1993) J. Biol. Chern. 288, 13151-13159 24. Mandel, J.-L., and Flintoff, W.F. (1978) J. Cell. Physiol. 97, 335-344 25. Clarke, D. D., Mycek, M. J.,Neidle,A., and Waelsch, H. (1959)Arch. Biochem. Biophys. 79,338-354 26. McKay,A. F.,Garmaise, D. L., Gaudry, R., Baker, H. A,, Paris, G. Y., Kay, R. W., Just, G. E., and Schwartz, R. (1959) J. Am. Chem. SOC.81, 4328-4335 27. Grettie, D. P., Bartos, D., Bartos, F., Smith, R. G., Campbell, R.A. (1978) in

Advances inPolyamineResearch, (Campbell,R.A., Moms, D. R., Bartos, D., Daves, G. D., and Bartos, E , eds) Vol. 2, pp. 13-21, Raven Press Ltd., New York 28. Folk, J. E., Park, M. H., Chung, S. I., Schrode, J., Lester, E. P., and Cooper, H. L. (1980) J. Biol. Chem. 256,369S3700 29. Seiler, N. (1987) in Inhibition of Polyamine Metabolism, Biological Sign$%cance and Basis for New Therapies (McCann, P. P., Pegg, A. E., and Sjoerdsma, A,, eds) pp. 49-78, Academic Press, Orlando, FL 30. Park, M. H., Cooper, H. L., and Folk, J. E. (1983) MethodsEnzymol. 94, 458-462 31. Tabor, C . W., and Tabor, H. (1984)Annu. Rev. Biochem. 53,749-790 32. Janne, J., Alhonen, L., and Leinonen (1991) Ann. Med. 23,241-259 33. Byers, T.L., Kameji, R., Rannels, D. E., and Pegg,A. E. (1987)Am. J. Physiol. 252, C663-C669 34. lbrrelio, B. M., Paz, M. A,, and Gallop, P.M. (1987) Biochem. Biophys. Res. Commun. 145, 1335-1341 35. Porter, C . W., Bernacki, R. J., Miller, J.,and Bergeron, R. J. (1993) Cancer Res. 63,581-586