A selective inhibitor of N8-acetylspermidine deacetylation in mice and hela cells without effects on histone deacetylation

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 273, No. 1, August 15, pp. 12%136,1989

A Selective Inhibitor of NE-Acetylspermidine Deacetylation in Mice and HeLa Cells without Effects on Histone Deacetylation PAMELA MARCHANT,* SASI DREDAR,* VICTOR MANNEH,* OTHMAN ALSHABANAH,* HARRY MATTHEWS,? DAVID FRIES.* AND JIM BLANKENSHIP**’ *Departments

of Pharmacology and Medicinal Chemistry, School of Pharmaq, University Stockton, California 95211, and iDepartment of Biological Chemistry, University of California, Davis, CalQbrnia 95616

of the Pa&tic,

Received December 19,1988, and in revised form April 21,1989

The inhibitory effects of 7-[N-(3-aminopropyl)amino]heptan-2-one (APAH) on N8acetylspermidine deacetylation were studied. In in vitro studies, APAH produced inhibition (apparent Ki of 0.18 PM) of Ns-acetylspermidine deacetylation by the 100,000~ supernatant fraction of rat liver. This apparent Ki was 60-fold less than the apparent Km (11 PM) for deacetylation of the substrate, N8-acetylspermidine, suggesting that APAH could be a potent, effective inhibitor in vivo. APAH was administered to mice by intraperitoneal injection at a dose of 200 mg/kg, and polyamine and acetylpolyamine levels in liver and spleen were measured. In tissues of control mice, N8-acetylspermidine was not detectable but increased to detectable levels 30-360 min after APAH treatment. These data are consistent with inhibition of the deacetylase by APAH. Increases in putrescine and N’-acetylspermidine levels occurred in liver after APAH treatment with increases in N1-acetylspermidine levels observed in spleen. In HeLa cells, a significant increase in N8-acetylspermidine was observed following 24 h exposure to 10 ELM APAH while no change occurred in the acetylation level of HeLa cell histones. In contrast, 24 h levels exposure to 10 mM sodium butyrate produced no change in Ns-acetylspermidine and an increase in the acetylation level of histones H4 and H2B. These results suggest that APAH has a relatively selective inhibitory effect on N8-acetylspermidine but not histone deacetylation. This is the first report of significant levels of N8-acetylspermidine in animal tissues and of the effects of in viva inhibition of N8-acetylspermidine deacetylase. 0 1989 Academic Press, Inc.

Spermidine undergoes acetylation by two different routes to yield Ns-acetylspermidine or N1-acetylspermidine. The formation of N’-acetylspermidine is catalyzed by a cytoplasmic enzyme spermidine/spermine N’-acetyltransferase (1) while NE-acetylspermidine formation occurs in the cell nucleus catalyzed by a chromosomal enzyme N*-acetylspermidine acetyltransferase (2,3). The two acetylated forms of spermidine are metabolized by different routes in the body with N’-acetylspermidine converted to putres’ To whom correspondence

should be addressed.

0003-9861/89 $3.00 Copyright All rights

0 1989 by Academic Press. Inc. of reproduction in any form reserved.

128

tine by polyamine oxidase (4, 5) and N8acetylspermidine deacetylated by N8-acetylspermidine deacetylase to yield spermidine (6, 7). The metabolism of N’-acetylspermidine has received more attention over recent years because of its role in the interconversion of the polyamines (4, 5). N*-Acetylspermidine metabolism is less well understood, and although it has been proposed that this acetylation could alter polyamine binding to DNA (8, 9), the normal cellular function of this process is still unknown. To increase our understanding of N*acetylspermidine metabolism, we have

INHIBITION

OF NS-ACETYLSPERMIDINE

studied the inhibition of N*-acetylspermidine deacetylase and its consequences on polyamine levels in vivo. We chose this deacetylase as the target for our inhibitor studies in an attempt to increase tissue levels of N*-acetylspermidine. This compound occurs in significant levels in urine (10, ll), but the concentration in tissues is much lower making detection difficult if not impossible with most detection systems (7). The development of an inhibitor that can increase the tissue levels of N*acetylspermidine would provide a useful, new tool for studying the function of nuclear acetylation of polyamines. In a previous study (12), we have synthesized and screened a number of polyamine-related analogs for inhibitory activity on in vitro deacetylation of N*-acetylspermidine by rat liver cytosol. The present study focuses on one of these compounds, 7[N-(3-aminopropyl)amino]heptan-2-one (APAH)? designed as a substrate analog resistant to N-deacetylation due to replacement of the N*-nitrogen by a carbon. The in vitro deacetylase assay was used to compare this compound with a variety of active-site-directed agents known to inhibit other deacetylases or proteases. APAH proved to be a potent inhibitor in our in vitro tests, and we went on to determine its effects on acetylated polyamine levels in mice and HeLa cells. We found that APAH does produce an increase in N*-acetylspermidine levels in both mice and HeLa cells permitting detection of N*acetylspermidine in tissues. Histones are N-acetylated in the cell nucleus by chromosomal enzymes similar and possibly identical to the nuclear N*-acetyltransferase for spermidine described above (3,13). Also as described for N*-acetylspermidine, acetylated histones undergo enzyme-catalyzed deacetylation (14). A number of studies have focused on the ability of sodium butyrate to inhibit histone deacetylation in vitro and in viva increasing the levels of acetylated histones in chromatin (15,16). In fact, sodium butyrate has become a commonly used tool to ’ Abbreviations used: APAH, ‘I-EN-(3-aminopropyl)amino]heptan-2-one; PAGE, polyacrylamide gel electrophoresis.

DEACETYLASE

129

induce morphological and biochemical changes in cultured cells (17). Unlike the nuclear acetyltransferases for histones and polyamines, there is very little information on the specificity or uniqueness of the deacetylase enzymes for these two classes of compounds. To determine if differences occur in inhibitor sensitivity between polyamine and histone deacetylase activities, we have examined the levels of acetylated polyamines and histones in HeLa cells grown in the presence of APAH or sodium butyrate. METHODS Animals and chemicals. Male, Swiss-Webster mice, 25-30 g in weight, and male, Wistar rats, 150-250 g in weight, were obtained from Simonsen Laboratories, Inc. (Gilroy, CA). N’- and N8-acetylspermidine and N8-[acet@H]acetylspermidine (25 mCi/mmol) were prepared following procedures previously used in this laboratory (6). The synthesis of APAH.2HCl was also performed in our laboratories (12). All media, horse serum and sterile salts or amino acid solutions for tissue culture were obtained from GIBCO Laboratories (Grand Island, NY). Cell culture. HeLa cells (53 strain) were obtained from Dr. Peter Yau (Department of Biological Chemistry, University of California, Davis) and from the American Type Culture Collection (Rockville, MD). Cells were grown at 37°C under 5% CO* in 75 cm2 plastic culture flasks in minimum essential medium with Earle’s salts with L-glutamine (GIBCO) containing horse serum (100 ml/l) and 0.01 mM nonessential amino acids. At 24 h before cells were harvested, fresh media were added containing APAH (10 PM) or sodium butyrate (10 mM) or unaltered (control). Cells were harvested (2.4 X lo7 cells/plate) by scraping cells from plates into ice-cold perchloric acid for polyamine analysis or into ice-cold phosphate-buffered saline (containing 10 InM sodium butyrate) for histone isolation. Cell counts were determined from parallel treated flasks by suspending cells using incubation in 0.25% trypsin and counting in a hemacytometer and viability was determined by trypan blue exclusion. In vitro enzyme assays. Assays for Ns-acetylspermidine deacetylase activity followed procedures used previously in this laboratory (6). Each assay mixture consisted of 0.125 mmol sucrose, 2.5 pmol MgCla, and 29 pmol NaH2P0, (pH 7.4) (Buffer A). Cytoplasmic protein (1.8-2.2 mg) various amounts of inhibitors, and 2.5-12.5 nmol of N8-[ucetyl-3H]acetylspermidine in a total volume of 0.5 ml. The enzyme preparation was the 100,090gsupernatant fraction from rat liver in Buffer A (6). Assays were incubated for 10 min at 37°C and halted on ice with addition of 500 pl of 1.0 N

130

MARCHANT

HCl-0.05 M acetic acid. The free 3H-acetic acid was extracted with ethyl acetate and radioactivity measured by liquid scintillation spectrometry (6). All assays were performed under conditions that ensured a linear rate of reaction throughout the lo-min incubation period. Each inhibitor was tested in preliminary experiments over a wide range of concentrations to get an estimate of the apparent K,, and the reported apparent Ki values were determined over a concentration range bracketing this value (&lo fold). The apparent Ki values were determined from the data obtained with each inhibitor using Dixon plots (18). Protein determinations were done by the method of Lowry et al (19). Treatment of mice with inhibitor APAH. A preliminary experiment was conducted to examine acute toxicity of APAH in mice. The compound was dissolved in 0.9% NaCl and the pH was adjusted to 7.5. Mice were injected intraperitoneally with various doses (50-500 mg/kg) of APAH in volumes not exceeding 0.3 ml. Mice were monitored for time of death and other signs of toxicity for up to 7 days after injection. For studies on the effects of APAH on polyamine tissue levels, mice were administered 200 mg/kg of the compound as described above. At various times after injection;mice were sacrificed by decapitation and livers and spleens were removed and placed immediately in ice-cold 0.25 N perchloric acid. In order to obtain sufficient tissue for analyses, spleens and livers from three mice were weighed and pooled prior to homogenization. Tissues were homogenized in 4 vol ice-cold 0.25 N perchloric acid containing 2.5 mM 1,7diaminoheptane (internal standard) using a Polytron homogenizer (Brinkmann Instruments). Homogenates were centrifuged at 10,OOOgand the supernatants were filtered through 0.22-rm cellulose nitrate filters. The filtrates were stored at 4°C for later analyses. Determination of mome tissue and HeL.a cell polyamine levels by high-pe&rmnnce liquid chromutography. The filtered tissue extracts were analyzed for polyamine and acetylpolyamine content by the HPLC technique of Seiler and Knodgen (20). The procedures, chemicals, and equipment used for this assay in our laboratories have been described previously (7). For HeLa cells, the standard gradient did not permit quantification of acetylspermidine levels, but we were able to measure these levels using gradient III from a more recent paper by Seiler and Knodgen (21). Quantitation was based on comparison with standard polyamines and acetylpolyamines with individual analyses compared using the internal standard. This assay system will permit detection of tissue levels of polyamines or acetylpolyamines as low as 0.5 nmol/g wet wt. Tissue polyamine levels at various treatment times were tested for significance compared to control levels by the use of Dunnett’s test (22).

ET AL. Isolution of nuclei from HeLa c&s. HeLa cell nuclei were isolated by a procedure described by Birnie (23) modified so that all solutions contained 10 mM sodium butyrate to prevent histone deacetylation during isolation. All steps were performed at 4°C and cells or nuclei were collected by centrifugation at 5OOOgfor 10 min. HeLa cells were harvested and washed three times in 20 vol of phosphate-buffered saline. Cells were resuspended in 20 vol of lysis buffer (10 mEdTrisHCI, pH 6.85,l mM CaClz, 1.5 mM MgClz, 0.2 mM phenylmethanesulfonyl fluoride, 0.25 M sucrose, 10 mM sodium butyrate and 0.5% Triton X-100) and allowed to swell for 5 min. Lysis was promoted by homogenization (15 strokes) in a tight Dounce homogenizer, and nuclei were collected by centrifugation. Nuclei were washed six times by resuspension in 20 vol of isolation buffer (lysis buffer without Triton X-100) and collected as a pellet on centrifugation. Analysis of acetylnted histonesfrm HeLa cells. Histones were extracted from isolated HeLa cell nuclei with 5 ml of 0.25 N HCl and homogenization for 15 s with a Polytron homogenizer. After having stood overnight at 4”C, the HCl extract was dialyzed against deionized water and lyophilized. Levels of histone acetylation were analyzed by separation of histones using polyacrylamide gel electrophoresis (PAGE) and scanning on a densitometer. The PAGE system used acid urea/Triton X-100 gels (15% acrylamide) following the procedures of Bonner et al (24). Electrophoresis was carried out on 0.5-mm-thick gels at 200 V/15 mA (maxima) for 5 h. Gels were stained in 1% naphthalene black, 20% methanol, 7% acetic acid and destained in 20% methanol, 7% acetic acid. The destained gels were photographed on Polaroid film, and the negatives were scanned with a Cary 210 spectrophotometer equipped with a gel scanner and digital interface port. RESULTS

In Vitro Inhibitor

Studies

From our studies on the subcellular localization of IV*-acetylspermidine deacetylase, we have found that about 90% of the enzyme activity is in the 100,OOOg supernatant fraction (6). We have used this crude enzyme preparation to screen for potential inhibitors with the assumption that compounds with inhibitory activity in this in vitro system and with proper physicochemical properties should also be an effective inhibitor in vivo (6,25). One consequence of the use of this crude enzyme preparation instead of a highly purified enzyme is that we can not obtain absolute Ki values, and we have, therefore, expressed our data as apparent Ki values. For our

INHIBITION

OF N*-ACETYLSPERMIDINE

purposes in screening and designing potential in vivo inhibitors, these apparent Ki values are acceptable and in fact can be more useful than results from a purified enzyme preparation that might lack other enzymes or cofactors that could affect inhibitor activity in vivo. The assay was set up under conditions to measure rates of initial velocity and the reaction was linear with time (up to 30 min) and protein concentration as shown in our earlier study (6). Apparent Ki values were obtained from Dixon (18) plots of the data as was done in earlier studies (6,25). From a series of acetylpolyamine analogs synthesized and tested in our laboratories, APAH was selected for further study as an inhibitor of Ns-acetylspermidine deacetylation. In this compound, the nitrogen in the NE position is replaced by a carbon rendering the compound resistant to N-deacetylation. APAH produced competitive inhibition in this in vitro deacetylase assay with an apparent Ki of 0.13 pM. This apparent Ki (0.18 PM) is significantly less than the apparent Km (11 PM) for the substrate determined in our study and indicates that it might be an effective selective inhibitor in vivo. One other compound, 7-amino-2-heptanone, has been reported to have inhibitory activity on this deacetylase, but the Ki (2.2 PM) is lo-fold greater than that of APAH (26). Sodium butyrate is a widely used inhibitor of histone deacetylation producing 50% inhibition of calf thymus histone deacetylase at a concentration of 60 PM (27). In the present study, sodium butyrate produced an apparent Ki of 38.5 mM in our in vitro N8-acetylspermidine deacetylase assay indicating a relatively low level of inhibition of this enzyme activity. In Vivo Mouse Studies on Deacetylase Inhibition Our initial toxicity screening of APAH was done over a range of four doses (50, 100, 200, and 500 mg/kg) with three mice receiving each dose. We observed no deaths at the three lower doses, but all three of the animals receiving 500 mg/kg died within 5 min. We followed with injection

DEACETYLASE

131

of six mice at 200 mg/kg and six mice at 300 mg/kg. All of the animals receiving 200 mg/kg survived while three of the six receiving 300 mg/kg died within 10 min. There were no overt signs of toxicity at 50100 mg/kg, but at 200 mg/kg, mice appeared excited with some mild tremors occurring in the first 10 min and then mice were sedated for 30-45 min. Given this preliminary toxicity screening, we chose 200 mg/kg of APAH given intraperitoneally for our remaining mouse studies. The levels of polyamines and acetylpolyamines in liver and spleen at various times after treatment with APAH are shown in Table I. N*-Acetylspermidine levels increased significantly in both liver and spleen as would be expected with inhibition of the deacetylase. The levels reached were higher in spleen than in liver. Our HPLC assay system appears to be as sensitive for tissue levels of N8-acetylspermidine as any other procedures published with detection limits of less than 0.5 nanomol/g of tissue, but in tissues from control mice, we could not detect this compound. This was expected since there have been no reports of detectable levels of this compound in tissues. The increases were seen in both tissues within 30 min after treatment and remained significantly elevated at 360 min in spleen even though inhibitor concentrations were no longer detectable after 90 min (see below). Although Mamont et aZ. (26) have reported Ns-acetylspermidine in cultured HTC cells, this is the first report of significant tissue levels of N*-acetylspermidine or of changes in tissue levels of this compound. An effect of the inhibitor was also observed on putrescine levels in liver and to a lesser extent in spleen. A 20-fold increase was observed in liver in the first 30 min after treatment with the peak effect occurring at 60 min. These levels returned to near the normal range at 240 min. Although the control levels of putrescine in spleen are higher than those in liver, there was no dramatic increase in levels after APAH with significant increases observed at only two widely separate times, 60 and 360 min. Increases in hepatic putrescine levels have been reported for animals receiving polyamines, polyamine analogs,

132

MARCHANT

ET AL.

TABLE I CONCENTRATIONS OF POLYAMINES AND ACETYLPOLYAMINES IN LIVERSAND SPLEENSOFMICE AT VARIOUS

TIMESAFTERADMINISTRATIONOF200 mg/kg 0~ 7-[N-(3-AMlNOPROPYL)AMINO@IEPTAN-2-ONE

Tissue Liver

Spleen

Treatment time (min) 0 30 60 90 120 240 360 0 30 60 90 120 240 360

Polyamine concentration (nmol/g wet wt)” Putrescine 18.0 + 0.3 312 f15** 353 + 16** 278 f41** 224 +36** 42.1 + 12.2 28.5 + 4.5 61 + 8 75 f 5 99 f 6** 87 f 6 82 + 8 81 + 8 113 f 5**

N’-Acetylspermidine 2.5 t 0.2 2.7 k 0.2 3.8 + 0.4** 3.9 f 0.2** 4.7 + 0.4** 2.0 + 0.3 2.8 zk0.4 2.8 -t 0.2 6.1+ 0.3** 8.7 t l.l** 8.0 + l.O** 7.2 f 0.3** 5.3 f 0.5* 4.9 + 0.7 *

N8-Acetylspermidine 10.5 b 2.3 f 0.3 * 3.1 + 0.7** 3.2 f 0.5** 3.9 f 0.3 ** 2.6 + 0.2* 0.8 f O.l* <0.5” 3.4 + 0.4** 6.0 f 0.9** 7.0 f 0.9 ** 7.9 + 0.4 ** 7.4 + 0.3** 3.2 + 1.3*

Spermidine 11z3* 52 1167k 77 1186k 75 1268+ 75 113Ok 97 1437 + 70 1076 + 43 11912 59 1312 f 87 1418+ 88** 1345 + 131 1395 + 123* 1213zk 42 1238k 84

a Values are the means + standard error from six to nine tissue samples. Each tissue sample consisted of the pooled tissues from three mice. bfl-Acetylspermidine was detected in only two of nine liver samples from control mice (detection limit i 0.5 nmol/g). “N*-Acetylspermidine was not detectable in spleen samples from control mice (detection limit i 0.5 nmol/g). * Significantly different from 0 time at P < 0.05 level. ** Significantly different from 0 time at P < 0.01 level.

and other compounds. For instance, spermidine and spermine caused an increase in putrescine levels in liver at 3-6 h (28) and in cultured L6 cells at 6 h (29) after administration. Methylglyoxal bis(guanylhydrazone), an inhibitor of S-adenosylmethionine decarboxylase, and carbon tetrachloride have also been shown to increase tissue levels of putrescine when administered to rats (1,28). In each of these studies, effects were observed at 3 h or longer after treatment and significant changes were observed in spermidine concentrations as well. The increases in putrescine levels observed in the present study after 30 min of treatment with APAH occur much more rapidly than those in any of the previous reports. N1-Acetylspermidine levels also increased in livers and spleens of mice treated with APAH. The change occurring in spleen was much greater and occurred earlier than that seen in liver. In the stud-

ies of Pegg and co-workers (1,28), N1-acetylspermidine levels were also found to increase under conditions leading to increased putrescine levels. They proposed that an increase in spermidineispermine N1-acetyltransferase activity was responsible for the increased levels of both putrescine and N1-acetylspermidine. Spermidine levels increased significantly in spleen at two of the five treatment times but did not change in livers of mice treated with APAH. The changes observed were small relative to those in previous reports in which polyamines or related compounds were administered (28,29). The HPLC procedure that we used permitted us to measure tissue levels of APAH in the same chromatographic runs which yielded concentrations of putrescine and spermidine. Thus we were able to determine the tissue concentrations of APAH at the same treatment times and

INHIBITION

OF N8-ACETYLSPERMIDINE

DEACETYLASE

133

TABLE II C~NCENTRATIONSOFPOLYAMINESANDACET~LPOLYAMINESIN HeLa CELLSAFTER~NCUBATIONFOR~~~WITH ~~~M~-[~V-(~-AMINOPROP~L)AMIN~~A~TAN-~-~NE(APAH) 0~10 mM SODIUMBU~RATE Polyamine concentration (nmol/106 cells)” Inhibitor (24-h treatment)

Putrescine

None APAH Butyrate

0.64 f 0.11 0.81+- 0.04 1.10 + 0.09*

N’-Acetylspermidine 0.16 + 0.02 0.18 + 0.03 0.59 f 0.003*

N8-Acetylspermidine <0.01* 0.04 + 0.015* <0.01*

Spermidine 2.05 + 0.34 2.14 + 0.22 2.00 + 0.33

a Each value is the mean f standard error of separate analyses from five culture flasks. * N*-Acetylspermidine was not detected in control or sodium butyrate-treated HeLa cells (detection limit < 0.01 nmoV1OGcells). * Significantly different from controls at P < 0.05.

even in the same tissue samples from which the data were obtained for Table I. For liver, tissue concentrations of APAH were 329 + 45 nmol/g wet wt at 30 min (mean f standard error), 50 + 11 at 60 min, and 29 + 7 at 90 min with no APAH detectable at later times. For spleen, tissue concentrations of APAH were 102 + 24 nmol/ g wet wt at 30 min and 37 f 4 at 60 min with no APAH detectable at later times. These results indicate that much higher levels occurred in liver than in spleen as would be expected after intraperitoneal injection. The concentrations are highest at 30 min in both tissues and fall rapidly over the next 30-60 min. These results indicate a short half-life for the compound with no detectable levels after 90 min; however, the treatment with APAH results in effects for 6 h or longer as indicated by the NEacetylspermidine levels in both liver and spleen. Inhibition of Polyamine and Histone Deacetylation in HeLa Cells The effects of APAH and sodium butyrate on polyamine and acetylpolyamine concentrations in HeLa cells are shown in Table II. The concentration of APAH to be used in the experiments with HeLa cells was determined in preliminary experiments. We found that concentrations of APAH of 10 PM and greater produced maximum increases in N*-acetylspermidine levels in HeLa cells after 24-h exposure times. The concentration for sodium buty-

rate (10 mM) and the exposure time (24 h) correspond to those used in a number of previous studies of inhibition of histone deacetylation in HeLa cells (15-17). Incubation of HeLa cells for 24 h in the presence of 10 PM APAH resulted in N*-acetylspermidine levels of 0.04 nmol/106 cells, at least four times the detection limit of ~0.01 nmol/106 cells. In both control cells and those treated with sodium butyrate, no N8-acetylspermidine was detected. These results are consistent with the proposed mechanism for APAH as an inhibitor of N8-acetylspermidine deacetylase. While we cannot rule out the possibility that a smaller increase in N8-acetylspermidine concentrations also occurred with sodium butyrate treatment, we were unable to detect such a change. APAH produced no significant change in any of the other three polyamines shown in Table II, nor were changes seen in spermine levels (data not shown). On the other hand, sodium butyrate produced significant increases in N1acetylspermidine and putrescine levels, but no changes were observed in spermidine or spermine (data not shown). The increases in putrescine and Nl-acetylspermidine levels produced by sodium butyrate are similar to effects produced by polyamines and polyamine analogs in L6 cells (28) and in rat liver (29). These effects have been attributed, at least in part, to increases in spermidine/spermine N’-acetyltransferase activity and the resultant breakdown of spermidine to putrescine

134

MARCHANT

60

66

Migration -,

70

76

mm

FIG. 1. Densitometric tracings of the stained H4 and H2B histone bands from HeLa cells following 24h exposures to 10 mM sodium butyrate or 10 pM APAH compared with controls. Nuclei were isolated from control and treated HeLa S3 cells, and histones were extracted and analyzed by electrophoresis. Levels of histone acetylation are shown for butyratetreated cells (top), APAH-treated cells (middle), and control cells (bottom). The bands represent histones H4 and H2B (60-75 mm from top of gel); this part of the gel pattern most clearly displays changes due to histone deacetylation. The degree of acetylation of histone H4 is indicated by the acetate/H4 ratios 4-O for tetraacetylated-nonacetylated.

with N’-acetylspermidine serving as an intermediate (5,28,29). The APAH results give no indication that Nl-acetylspermidine synthesis or further metabolism is affected by exposure to APAH in the concentrations and times used in these experiments. The extent of histone acetylation in HeLa cells exposed for 24 h to 10 mM sodium butyrate or 10 PM APAH in comparison with that of control HeLa cells is shown in Fig. 1. The concentrations of inhibitors, exposure times, and other conditions for these experiments are identical to those used to examine the inhibitors’ effects on polyamines (Table II). The top gel scan in Fig. 1 obtained using histones from sodium butyrate-treated cells illustrates the effects of inhibition of histone deacetylation. There are five peaks corresponding to H-4 with the largest peak (numbered 4) representing the tetraacetylated form and peaks of decreasing heights

ET AL.

(numbered 3 to 0) representing the tri-, di-, mono-, and nonacetylated forms. To the left of the H4 pattern, the gel scan produced by histone H2B follows the same order with the fastest migrating peak representing the nonacetylated form and the peaks to the left representing successively greater levels of acetylation. When compared with the scan from the control cells in which the largest peaks are the nonacetylated forms, it is evident that the level of acetylation is much greater in cells following sodium butyrate treatment. These results are similar to those obtained by many other workers indicating that sodium butyrate treatment results in hyperacetylation of histones (15-17) and are consistent with the inhibition of histone deacetylase(s). The gel scan of histones from APAH-treated HeLa cells indicates that this compound has no significant effect on the extent of histone acetylation with a pattern very similar to that seen from control cells. Thus these results indicate that APAH at the concentration and exposure time which produce significant increases in tissue levels of N8-acetylspermidine does not inhibit histone deacetylase(s) and does not alter the extent of acetylation of histones H4 and H2B. Thus APAH appears to have significant specificity for deacetylation of Ns-acetylspermidine rather than histones H4 or H2B. DISCUSSION

We have used both in vitro and in vivo studies to characterize the effects of an inhibitor, APAH, on polyamine and histone deacetylation. APAH was designed and synthesized as a structural analog of N*acetylspermidine that would be resistant to the deacetylase activity because of the replacement of the N*-nitrogen by a carbon (12). The in vitro studies indicated that this compound is a potent, competitive inhibitor as we had predicted and that the Ki was significantly less than the Km for the substrate and thus might serve as an effective inhibitor in viva. The studies using APAH in mice were initiated to test for these inhibitory effects in viva, and our results showed an increase in N*-acetylspermidine levels in liver and spleen as would

INHIBITION

OF N*-ACETYLSPERMIDINE

be expected with deacetylase inhibition. We observed a similar increase in N8-acetylspermidine levels in HeLa cells following exposure to APAH. In both mice and HeLa cells, spermidine levels remain unaffected by APAH indicating that NE-acetylspermidine levels are not increasing as a consequence of changes in spermidine concentrations. Thus APAH appears to be an effective inhibitor of N*-acetylspermidine deacetylation under both in vitro and in vivo conditions. The finding that APAH inhibited N8acetylspermidine deacetylase raises another question concerning the specificity of this compound for acetylpolyamine versus histone deacetylation. It has been found that acetylation of histones and N8-acetylation of spermidine are catalyzed by the same or closely related enzymes (3, 13). While evidence indicates some differences in polyamine versus histone deacetylases since they occur in the cytoplasm and the nucleus (14), respectively, the similarity of the N-deacetylation reactions with these two classes of compounds points to the possibility that closely related enzymes may be involved in these reactions. The usefulness of APAH for the study of polyamine metabolism requires that it be selective for N8-acetylspermidine and not histone deacetylation. Our in vitro deacetylase assays comparing apparent Ki values for APAH and butyrate offered results suggesting some specificity of APAH for the polyamine deacetylase reactions. On the other hand, butyrate appears to be a much less effective inhibitor of N8-acetylspermidine deacetylase than it is of histone deacetylation (27). The HeLa cell studies were designed to examine more directly the effects of APAH on the levels of polyamine and histone acetylation and thus indirectly evaluate this compound’s specificity for these two deacetylase processes. The results of these studies showed little or no effect of APAH on histone deacetylation at concentrations which produced significant increases in N*-acetylspermidine. The results also confirmed our in vitro finding that butyrate is not an effective inhibitor of N8-acetylspermidine deacetylation. Thus the enzymes involved in histone and N8-acetylspermi-

DEACETYLASE

135

dine deacetylation appear to be distinctly different with respect to substrate and inhibitor specificity, and APAH exhibits a potent selectivity for the N8-acetylspermidine deacetylase. In the mouse studies, the effects of APAH on polyamine and acetylpolyamine levels in tissues indicate that this compound does not produce simply an increase in N*-acetylspermidine. For instance, significant increases were observed in N’acetylspermidine levels in both liver and spleen and dramatic increases occurred in putrescine levels in liver. These changes were not expected; however, it is not uncommon that an inhibitor produces effects on tissue levels of more than one compound or on several metabolic pathways. Such effects can be due to a lack of inhibitor specificity resulting from direct effects on other enzymes or processes, or in other cases, effects can be a secondary consequence of the inhibition of the target enzyme and the resultant changes in normal levels of compounds. We can speculate that the increase in N1-acetylspermidine and putrescine levels could be a result of induction of the N’-acetyltransferase as has been reported by Pegg and co-workers (1, 28,29) but why should the increase in putrescine levels be so much greater in liver than in spleen? Bolkenius and Seiler (4) have shown that N8-acetylspermidine is a noncompetitive inhibitor of polyamine oxidase with a Kc of 11 pM while the Km for N’-acetylspermidine is 14 pM. This should normally be of no consequence since N8acetylspermidine levels are far below either the Ki or Km values. However, between the 60- and 240-min treatment times, levels of N*-acetylspermidine in spleen approach the Ki value. Thus the increase in putrescine levels in spleen might be blocked because of inhibition of polyamine oxidase by NE-acetylspermidine. This leaves open the question as to whether the concentration in spleen is sufficient to produce inhibition of polyamine oxidase while that in liver is not. In the reports by Pegg and co-workers (1,28, 29), the time course for the polyamine-induced increases in N’-acetylspermidine and putrescine is much slower than that observed with APAH in the present study.

136

MARCHANT

This raises another question concerning the mechanism for the increased levels of these compounds and points to the need for additional experiments to answer these questions. While there have been several reports on the levels of N8-acetylspermidine in urine (10, ll), there have been no previous reports of detectable levels of this compound in. tivo in mammalian tissues in the presence or absence of inhibitors. The present study confirms that NE-acetylspermidine can be detected in mouse tissues and in HeLa cells under certain conditions and points to the usefulness of APAH as a tool in determining the physiological role for N8-acetylation and deacetylation of spermidine. ACKNOWLEDGMENTS We thank Peter Yau for his help and advice on the tissue culture techniques. This study was supported in part by National Science Foundation Grants DCB8313188,DCB8711590, and DCB8705378. REFERENCES 1. MATSUI, I,, WIEGAND, L., AND PEGG, A. E. (1981)

J. Biol Chem. 256,2454-2459. 2. BLANKENSHIP, J., AND WALLE, T. (1977)Arch Biohem. Biophys. 179,235-242. 3. LIBBY, P. R. (1980) Arch. Biockm. Biophys. 203,

384-389.

ET AL. 9. MORGAN, J. E., BLANKENSHIP, J. W., AND MATTHEWS, H. R. (1987) Biochemists 26, 3643-

3649. 10. TSUJI, M., NAKAJIMA, T., AND SANO, I. (1975) Clin

Chim Acta 59,161-16’7. 11. ABDEL-MONEM,

M. M., AND OHNO, K. (1978) J.

Pharm. Soi. 67,1671-1673. 12. DREDAR, S., BLANKENSHIP, J., MARCHANT, P., MANNEH, V., AND FRIES, D. (1989) J. Med. Chem. 32,984-989. 13. LIBBY, P. R. (1978) J. Biol. Chem. 253,233-237. 14. MATTHEWS, H. R., AND WATERBORG, J. H. (1985) in Enzymology of Post-translational Modification of Proteins (Freedman, R., and Hawkins, H., Eds.), Vol. 2, pp. 125-184, Academic Press, San Diego. 15. VIDALI, G., BOFFA, L. C., BRADBURY, E. M., AND ALLFREY, V. G. (1978) Proc. Natl. Acad Sci USA 75.2239-2243. 16. WHITLOCK, J. P., GALEAZZI, D., AND SCHULMAN, H. (1983) J. Biol Chem. 258,1299-1304. 17. KRUH, J. (1982) Mol. Cell B&hem. 42,65-82. 18. DIXON, M. (1953) Biochem. J. 55,170-175. 19. LOWRY, 0. H., ROSEBROUGH, N. J., FARR, A. L., AND RANDALL, R. J. (1951) J. Bid Chem. 193, 265-275. 20. SEILER, N., AND KNODGEN, B. (1980) J. Chrometog-. 221,227-235. 21. SEILER, N., AND KNODGEN, B. (1985) J. Chroma-

togr. 339,45-57. 22. DUNNETT, C. W. (1964) Biometrics 20,482-494.

23. BIRNIE, G. D. (1978) in Methods in Ceil Biology (Stein, G., and Stein, J., Eds.). Vol. 16, pp. 1326, Academic Press, New York. 24. BONNER, W. M., WEST, M. P. H., AND STEDMAN, J. D. (1980) Eur. J. B&hem. 109,17-23. 25. SANTACROCE, M. J., AND BLANKENSHIP, J. (1982)

Proc. West.Pha rwzocol Sot. 25,113-118.

4. BOLKENIUS, F. N., AND SEILER, N. (1981) Int. J. Biochem. 13,287-292. 5. PEGG, A. E. (1986) B&hem J. 234.249-262. 6. BLANKENSHIP, J. (1978) Arch B&hem. Biqphys. 189,20-27. 7. BLANKENSHIP, J., AND MARCHANT, P. (1984) Proc Sot Exp. BioL Med. 177,180-187. 8. BLANKENSHIP, J., AND WALLE, T. (1977) Advances

26. MAMONT, P. S., BOLKENIUS, F., SEILER, N., BEY, P., AND KOLB, M. (1984) International Confer-

in Polyamine Research (Campbell, R. A., Ed.), Vol. 2, pp. 97-110, Raven Press, New York.

29. ERWIN, B. G., AND PEGG, A. E. (1986) Biochem. J. 238,581-587.

ence on Polyamines, Budapest, Hungary. [Abstract No. 61 27. COUSENS,L. S., GALLWITZ, D., AND ALBERT& B. M.

(1979) J. Biol. Chem. 254,1716-1723. 28. PEGG, A. E., ERWIN, B. G., AND PERSSON, L. (1985)

Biochim. Biophys. A& 842,111-118.