Metabolic Stability of α-Methylated Polyamine Derivatives and Their Use as Substitutes for the Natural Polyamines*

Metabolically stable polyamine derivatives may serve as useful surrogates for the natural polyamines in studies aimed to elucidate the functions of individual polyamines. Here we studied the metabolic stability of α-methylspermidine, α-methylspermine, and bis-α-methylspermine, which all have been reported to fulfill many of the putative physiological functions of the natural polyamines. In vivo studies were performed with the transgenic rats overexpressing spermidine/spermine N1-acetyltransferase. α-Methylspermidine effectively accumulated in the liver and did not appear to undergo any further metabolism. On the other hand, α-methylspermine was readily converted to α-methylspermidine and spermidine; similarly, bis-α-methylspermine was converted to α-methylspermidine to some extent, both conversions being inhibited by the polyamine oxidase inhibitor N1, N2-bis(2,3-butadienyl)-1,4-butanediamine. Furthermore, we used recombinant polyamine oxidase, spermidine/spermine N1-acetyltransferase, and the recently discovered spermine oxidase in the kinetic studies. In vitro studies confirmed that methylation did not protect spermine analogs from degradation, whereas the spermidine analog was stable. Both α-methylspermidine and bis-α-methylspermine overcame the proliferative block of early liver regeneration in transgenic rats and reversed the cytostasis induced by an inhibition of ornithine decarboxylase in cultured fetal fibroblasts.

Although the requirement of the natural polyamines spermidine, spermine, and their precursor putrescine for the growth of mammalian cells is extremely well documented, their specific functions in proliferative processes are largely unknown (1). Some of the published data appear to assign a central role to spermidine, whereas putrescine is supposed to serve as its precursor and spermine as a storage pool convertible back to spermidine. For the elucidation of the physiological roles of individual polyamines, metabolically stable derivatives of polyamines fulfilling their specific cellular functions would be ex-tremely valuable. Methyl derivatives of spermidine and spermine have been used as substitutes for the natural polyamines both in vitro and in vivo. ␣-Methylspermidine (MeSpd), 1 ␣-methylspermine (MeSpm), and bis-␣-methylspermine (1,12dimethylspermine, Me 2 Spm) are equally effective as the natural polyamines in inducing the conversion of right-handed B-DNA to left-handed Z-DNA (2). In addition to spermidine and spermine, cytostasis that resulted from the inhibition of the S-adenosylmethionine decarboxylase can be reversed by MeSpd but not by Me 2 Spm (3). Spermidine, spermine (because of its conversion to spermidine), and MeSpd serve as substrates for the synthesis of deoxyhypusine (an integral part of eukaryotic initiation factor 5A), whereas Me 2 Spm does not (3). Interestingly, all of the mentioned methylated derivatives of spermidine and spermine have been reported to reverse the cytostasis induced by difluoromethylornithine, a specific inhibitor of mammalian ornithine decarboxylase (4). MeSpd appears to undergo slow conversion to MeSpm. Me 2 Spm has not been reported to be metabolized. MeSpd is not a substrate for spermidine/spermine N 1 -acetyltransferase (SSAT) (4). In line with the above studies, we found that MeSpd prevents zinc-induced pancreatitis and restores liver regeneration in transgenic rats overexpressing SSAT under the control of the metallothionein promoter (5). Under these conditions where polyamine catabolism was intensely activated, the natural polyamines could not be used as they were rapidly acetylated and degraded with no net tissue accumulation (5).
Here we studied the metabolic stability of the three methylated polyamine derivatives, namely MeSpd, MeSpm, and Me 2 Spm. Transgenic rats with activated polyamine catabolism because of overexpression of SSAT were used for the experiments in vivo, immortalized fibroblasts derived from the same animals and liver extracts from normal rats were used for experiments in vitro. The metabolisms of these polyamine analogs were also studied by using purified recombinant human polyamine oxidase, human spermine oxidase, and mouse SSAT. We found no evidence indicating that MeSpd would be further metabolized in vivo and the compound was also a poor substrate for the studied enzymes in vitro. Surprisingly, both MeSpm and Me 2 Spm were catabolized at the methylated ends both in vivo and in vitro although Me 2 Spm was far more stable. Furthermore, MeSpd and Me 2 Spm were competitive inhibitors of SSAT.
Both MeSpd and Me 2 Spm restored early liver regeneration in transgenic rats with activated polyamine catabolism, in which a profound spermidine and spermine depletion developed after partial hepatectomy. Although Me 2 Spm was metabolized to MeSpd to some extent, the restoration of the proliferative activity was attributable to Me 2 Spm and not to the low contents of MeSpd.
Transgenic Rats and the Studies of ␣-Methylated Spermine Analogs-The production of transgenic rats has been described in detail earlier (8). Partial hepatectomy of the transgenic rats was carried out essentially as described in Ref. 9. Treatments before partial hepatectomy and the determination of DNA synthesis were carried out as described in the legend to Ref. 5. Transgenic 10-week-old male rats were injected twice with MDL72527 (50 mg/kg intraperitoneal) at 16-h intervals to inactivate PAO according to Bolkenius et al. (10) and further with MeSpd, MeSpm, or Me 2 Spm twice (25 mg/kg intraperitoneal) 2 and 8 h after the second MDL72527 treatment. Animals not treated with the PAO inhibitor were injected with ␣-methylated spermine analogs or MeSpd at the same time points as MDL72527-pretreated animals. The animals were sacrificed 24 h after the second MDL72527 injection; liver pieces were frozen in liquid nitrogen and homogenized in the standard buffer (25 mM Tris-HCl, pH 7.4, 0.1 mM EDTA, 1 mM dithiothreitol). An aliquot of the homogenates was used for the polyamine assays. The homogenates were centrifuged (at 13,000 ϫ g, for 30 min, at 4°C) and the supernatant fractions were used for the enzyme activity assays. The Institutional Animal Care and Use Committee of the University of Kuopio and the Provincial Government approved the animal experiments.
Immortalized Fibroblasts Overexpressing SSAT-Rat fibroblasts from MT-SSAT transgenic rats were derived from 13-day-old fetuses. The fibroblast processing was basically the same as described by Mackintosh et al. (11), and transfection was performed with the same plasmid and the commercial kit Lipofectamine TM Plus. Plasmid expressing SV40 large and small T antigens was a kind gift from Dr. M. J. Tevethia of the Department of Microbiology (Pennsylvania State University College of Medicine, Hershey, PA). The transgenic cell populations were identified with the aid of the quantitative PCR. The immortalized fibroblasts were plated in 6-well culture plates in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with heat-inactivated 10% fetal bovine serum, gentamycin (50 g/ml; Invitrogen), and geneticin G418 (500 g/ml; Sigma). The cells were left to adhere for 24 h before the growth medium was replaced with fresh medium and drugs. After the incubation, cells were washed with phosphate-buffered saline, detached with trypsin, counted, and the polyamines were determined after sulfosalisylic acid precipitation from the supernatant fractions with the aid of high performance liquid chromatography.
The Back-conversion Studies of the ␣-Methylated Polyamine Analogs with the Liver Extracts in Vitro-Wild-type Wistar male rats were sacrificed and livers were processed as above. All supernatants were combined and eluted at 4°C through Amicon Ultra-15 centrifugal filter devices (Millipore) with a nominal molecular weight limit of 30,000 with high-salt (500 mM NaCl) buffer to remove the natural polyamines. The resulting eluates were desalted in the same columns with non-salt buffer three times to remove excess salt. The final eluates were pooled and the protein concentrations were determined. Experiments were carried out in triplicates where 40 l of liver extract was used per total reaction volume of 180 l. Reaction buffer containing 0.1 M glycine-NaOH, pH 9.5, and 5 mM dithiothreitol was used to study the metabolism of the polyamines and their ␣-methylated analogs in 1 mM concentrations. The reactions were initiated with the studied polyamine or analog addition and the reaction tubes were incubated at 37°C water bath for 60 min. Five millimolar freshly distilled benzaldehyde was used to decrease the K m values and to increase the reaction velocity of PAO according to Hölttä (12). Where indicated, the reaction mixtures were preincubated for 10 min with 250 M MDL72527 to inactivate PAO before polyamine or analog addition ( Table I). The reactions were stopped with the addition of 20 l of 100 M diaminohexane in 50% (w/v) sulfosalisylic acid.
Production of Recombinant PAO and SMO, Expression Systems-The cDNA coding for the hSMO open reading frame and containing His 6 at the N terminus was synthesized by PCR using the following primers: 5Ј-GAAGGAGATATACATATGCACCATCATCACC-ATCACATTGAGGGTCGCATGCAAAGTTGTGAATCCAGTGG-3Ј and 5Ј-GAGAAGGTCGTCCCCTGGACTGAGCTCTTCGAACATACG-3Ј. The cDNA coding for the hPAO open reading frame and containing His 6 at the N terminus was synthesized by PCR using the following primers: 5Ј-GAAGGAGATATACATATGCACCATCATCACCATCACATTGAGG-GTCGCATGGAGTCGACCGGCAGCGTC-3Ј and 5Ј-CGGGTCCGGCT-CCGAGATCGAGCTCTTCGAACATACG-3Ј. The PCR fragments were cloned into NdeI/XhoI of the pET 30a vector (Novagen, Inc., Madison, WI). Bacteria containing plasmids were cultured and collected according to the Qiagen Qiaexpressionist TM manual and the proteins were purified under native conditions using nickel-nitrilotriacetic acid His Bind resin (Novagen) according to the manufacturer's instructions.
Both PAO and SMO proteins were further purified with affinity chromatography. For the affinity matrix NHS-activated Sepharose TM 4 Fast Flow was used according to the manufacturer's instructions (Amersham Biosciences). N 8 -Acetylspermidine and spermidine were used as ligands for PAO, whereas for SMO, spermidine and Me 2 Spm were used. The SMO elutes were concentrated and washed with the standard buffer (25 mM Tris-HCl, pH 7.4, 0.1 mM EDTA, 1 mM dithiothreitol) several times using Amicon Ultra-15 centrifugal filter devices (Millipore). PAO, however, could not be concentrated nor washed in a similar way as it bound to the filter membrane very tightly. Desalting of the PAO elutes after affinity chromatography were performed with NAP-10 TM gel filtration columns according to the manufacturer's instructions (Amersham Biosciences).
Production of Recombinant SSAT-To obtain the SSAT cDNA (exons one to six) the pool of the first strand cDNA was PCR amplified using primers 5Ј-TACGTCGACACGAATGAGGAACCACC-3Ј and 5Ј-CTAGC-GGCCGCAGGTTGTCATTGTCTAC-3Ј. The resultant PCR product was gel-purified, restriction enzyme digested, and cloned. For protein expression the coding sequence of SSAT was amplified by PCR using the cloned SSAT cDNA as a template. The primers used are: 5Ј-TTAGCC-ATATGCATCATCATCATCATCATGATGACGACGACAAGATGGCTA-AATTTAAGATCCG-3Ј and 5Ј-CTACTCGAGCTCACTCCTCTGCTGC-C-3Ј. The upstream primer contained His 6 and an enterokinase cleavage site. The NdeI/XhoI-digested PCR products were cloned into the pET 30a vector (Novagen, Inc., Madison, WI) and sequenced. For the protein production Escherichia coli strain BL21(DE3) was used as the host and the recombinant protein was purified under native conditions using nickel-nitrilotriacetic acid His Bind resin (Novagen) according to the manufacturer's instructions.
All purified proteins were analyzed by SDS-PAGE, and the protein concentrations were measured using the Bio-Rad Protein Assay (Bio-Rad). The kinetic studies of the oxidizing enzymes were performed in duplicates with 4 to 6 different 10 to 1000 M substrate concentrations. The SMO and PAO reactions were carried out in a total volume of 180 l in the same buffer as the liver extract reactions and were allowed to proceed for 10 to 30 min at 37°C before addition of 20 l of 100 M diaminohexane in 50% (w/v) sulfosalisylic acid. The kinetic study of SSAT was performed with 100, 400, 700, and 1000 M spermidine or spermine as a substrate and 100, 400, 700, and 1000 M MeSpd or Me 2 Spm as a competitive inhibitor.
Determination of Ornithine Decarboxylase, SSAT, PAO Activities, and Polyamines-The activity of ornithine decarboxylase was assayed as described previously (13), and SSAT activity was assayed as described in Ref. 14. The PAO activity was assayed essentially as described by Kumazawa et al. (15) using radioactive N 1 ,N 11 -diacetylnorspermine, instead of N 1 -acetylspermine, as the substrate. High performance liquid chromatography was used to determine the concentrations of the polyamines and their methylated analogs essentially as described by Hyvönen et al. (16).
Statistical Analyses-The data were expressed as mean Ϯ S.D. Oneway analysis of variance with Dunnett's post-hoc test for multiple comparisons was used for statistical analyses with the aid of a software package, GraphPad Prism 3.0 (GraphPad Software, Inc., San Diego, CA).

Purity of ␣-Methylated Polyamine Analogs: NMR and High
Performance Liquid Chromatography Studies-1 H and 13 C spectra were recorded on an Avance DRX spectrometer operating at 500.13 MHz. NMR showed Ͼ99.5% purity for all studied polyamine analogs. MeSpd, MeSpm, and Me 2 Spm were further tested with high performance liquid chromatography for typical orthophtalaldehyde reactive impurities. In 10,000 pmol of each sample, the amount of orthophtalaldehyde reactive impurities were less than 50 pmol (results not shown).
Stability of the Analogs in the Liver of SSAT Transgenic Rats- Fig. 1A displays the typical polyamine pattern in the liver of SSAT overexpressing rats. Putrescine pool was greatly increased, whereas the spermine level was decreased when compared with wild-type rat liver where putrescine remains almost undetectable (8). An exposure of the rats to the polyamine oxidase inhibitor MDL72527 expectedly greatly reduced putrescine content and expanded the pool of N 1 -acetylspermidine (Fig. 1A). All the analogs accumulated in the liver of the transgenic rats, MeSpd showing the highest tissue concentrations (Fig. 1, A-D). The latter analog likewise appeared to be metabolically stable and, in contrast to an earlier report (4), we found no evidence that MeSpd was converted to MeSpm (Fig.  1B). As indicated in Fig. 1C, MeSpm was effectively converted to MeSpd. The latter conversion was markedly inhibited by the MDL72527 compound (Fig. 1C). Fig. 1D shows that also Me 2 Spm was converted to MeSpd, but to a much lesser extent than MeSpm. The conversion of Me 2 Spm to MeSpd was totally prevented by MDL72527 (Fig. 1D).
Effect of the Analogs on SSAT and PAO Activities in the Liver of SSAT Transgenic Rats-The methylated polyamine analogs did not appear to be very effective inducers of SSAT activity. In fact, only Me 2 Spm significantly increased the hepatic SSAT activity, which was further enhanced by combining the latter analog with MDL72527 ( Fig. 2A). However, the modest induction of the SSAT activity in response to the methylated analogs may be tissue-specific as the analogs much more effectively induced SSAT activity in the pancreas of the transgenic rats (results not shown). The analogs had little effect on PAO activity, whereas the used doses of MDL72527 alone or in combination with the analogs virtually completely inhibited PAO activity (Fig. 2B). All the analogs stimulated ornithine decarboxylase activity to some extent (results not shown).
Studies with Liver Extracts Obtained from Wild-type Rats-We used crude liver extracts to study the stability of analogs in vitro under conditions corresponding to the situation in vivo. High-salt extraction was performed (see "Materials and Methods") to remove or decrease the content of small molecular weight compounds. The extraction procedure certainly removed all acetyl-CoA (and probably also inactivated the labile SSAT protein) thus preventing acetylation of the polyamine substrates during the experiments. We used benzaldehyde supplementation as it is known to greatly enhance PAO activity when nonacetylated spermidine or spermine is used as a substrate (but not when N 1 -acetylspermidine is used) (12). Benzaldehyde in all likelihood forms a Schiff base with the polyamines mimicking the structure and charge distribution of the acetylated polyamines, which are much more preferred substrates for PAO than the unmodified polyamines (12).
As shown in Table I, substantial amounts of spermidine and spermine (but no putrescine) remained in the extracts even after the high-salt procedure. During the 60-min incubation, some putrescine was formed and the production was greatly enhanced upon an inclusion of 5 mM benzaldehyde in the incubation mixture. Interestingly, the formation of putrescine was associated with the disappearance of the endogenous spermine, but not spermidine. However, the view that spermine was directly converted to putrescine was ruled out by tracer studies indicating that labeled spermine was first converted to spermidine followed by the appearance of the label in putrescine (results not shown). Putrescine formation without benzaldehyde addition was probably because of mono-and/or diaminooxidases as we did not use any inhibitors. Inclusion of 1 mM N 1 -acetylspermidine yielded large amounts of putrescine, the formation of which was not enhanced by benzaldehyde. In fact, the observed slight inhibition in the presence of benzaldehyde could be attributable to the formation N 1 -acetyl/N 8 -benzyl-Schiff base-spermidine and N 8 -acetylspermidine is an inhibitor for PAO (17). MDL72527 prevented any formation of putrescine from N 1 -acetylspermidine (Table I). In comparison with N 1 -acetylspermidine, 1 mM spermidine produced relatively little putrescine even in the presence of benzaldehyde, probably because Schiff base can be formed in both primary amino groups of spermidine (N 1 /N 8 ). Inclusion of 1 mM MeSpd appeared to yield some putrescine in the presence of benzaldehyde, but this putrescine was in all likelihood derived from endogenous spermine (Table I). Exogenous spermine was converted to spermidine even in the absence of benzaldehyde but the conversion was greatly enhanced by the latter compound. Inclusion of 1 mM MeSpm yielded substantial amounts of spermidine and MeSpd already without benzaldehyde, which further enhanced the conversions by a factor of about 3 ( Table I). The only product derived from 1 mM Me 2 Spm was MeSpd, the formation of which was stimulated nearly 5-fold by benzaldehyde. It is interesting to note that Me 2 Spm was even more effectively converted to MeSpd in the presence of benzaldehyde than was MeSpm. The metabolism of spermine and its methylated derivatives was fully or partially blocked by MDL72527 (Table I).
Polyamines and Their Analogs as Substrates for Recombinant SSAT, PAO, and SMO-We also produced purified recombinant mouse SSAT, human PAO, and human SMO to study the substrate specificities of these enzymes. SSAT and SMO purifications yielded very pure enzymes according to SDS-PAGE, whereas purification of PAO resulted in about 80% pure enzyme. However, the contaminating bacterial protein did not have any PAO-like activity (data not shown).
K m and V max values of spermidine and spermine as substrates for SSAT were 52 M and 2.4 nmol/min/g; 33 M and 0.43 nmol/min/g, respectively. The K i value of MeSpd with spermidine as substrate was 144 M, K i values of Me 2 Spm with spermidine and spermine as substrates were 30 M and 34 M, respectively. MeSpm was not tested as it could serve as a substrate for SSAT at the unmethylated end. The kinetic values for different analogs are listed in Table II in respect to both recombinant oxidases. PAO strongly preferred N 1 -acetylspermidine; furthermore, PAO was able to oxidize methylated spermine derivatives, but not MeSpd, to spermidine and MeSpd. In the presence of benzaldehyde, the K m values for PAO decreased and the enzyme readily used spermine and its analogs but not very effectively spermidine or MeSpd, probably because of the inhibitory N 8 Schiff base. Spermidine, N 1 -acetylspermidine, and MeSpd did not serve as substrates for SMO, whereas spermine and its mono-and dimethyl derivatives did. Expectedly, neither oxidase produced spermidine from Me 2 Spm. Interestingly, the effect of methylation on the kinetic values seems to be opposite for the oxidases; the activity of PAO without benzaldehyde increased with the number of the methyl groups, whereas that of SMO decreased. As a conclusion it seems that ␣-methylation prevents polyamine derivatives from acetylation but not their oxidation.
Restoration of Early Liver Regeneration in Transgenic Rats Overexpressing SSAT-We have earlier found that partial hep- atectomy of the transgenic rats overexpressing SSAT results in a profound spermidine and spermine depletion because of SSAT induction at 24 h postoperatively and failure to initiate liver regeneration (5,18). Liver regeneration could be restored by a prior administration of MeSpd (5). Table III shows thymidine incorporation in livers of transgenic rats before and 24 h after partial hepatectomy. As shown, thymidine incorporation remained at the preoperative level at 24 h in untreated rats, whereas small doses (5 mg/kg) of MeSpd only insignificantly increased DNA synthesis. Higher doses (25 mg/kg) of both MeSpd and Me 2 Spm resulted in about a 10-fold stimulation of DNA synthesis. Table III also lists the hepatic pools of poly-TABLE I Metabolism of polyamines and their analogs in rat liver extracts The liver extracts were prepared as described under "Materials and Methods" and incubated for 60 min at ϩ37°C with the indicated additions. Reactions with MDL72527 were preincubated for 10 min before other drug addition. The concentration of added polyamine was 1 mM, benzaldehyde 5 mM, and MDL 72527 250 M. The abbreviations used are: Put, putrescine; N 1 -AcSpd, N 1 -acetylspermidine; Spd, spermidine; Spm, spermine; BA, benzaldehyde.

TABLE II Kinetic values of recombinant polyamine and spermine oxidases
The enzyme activities were measured as described under "Materials and Methods." Methylspermine can be catabolized from either end of the spermine backbone. The abbreviations used are: SMO, spermineoxidase; BA, benzaldehyde; N 1 -AcSpd, N 1 -acetylspermidine; Spd, spermidine; Spm, spermine. amines and their analogs before and after partial hepatectomy. As found earlier, partial hepatectomy brought about profound depletions of the natural polyamines spermidine and spermine at 24 h after the operation in the transgenic rats, whereas the analogs readily accumulated in the liver. N 1 -Acetylspermidine levels remained unaffected in all animal groups at about 100 pmol/mg of tissue (results not shown). Even though Me 2 Spm was converted to MeSpd, the hepatic concentration of the latter compound remained clearly below (about half) the level achieved with the smaller dose of MeSpd that failed to stimulate thymidine incorporation (Table III). It thus appears that the observed stimulation of DNA synthesis after Me 2 Spm was attributable to the spermine derivative and not to MeSpd. In fact, Me 2 Spm appeared to be even more effective than MeSpd on the molar basis.

Reversal of Difluoromethylornithine-induced Cytostasis by Bis-␣-methylspermine in
Immortalized Fibroblasts-We next tested whether difluoromethylornithine (DFMO)-induced cytostasis in fibroblasts derived from transgenic fetuses overexpressing SSAT could be reversed with Me 2 Spm. As shown in Fig. 3, 5 mM DFMO distinctly inhibited cell growth. An addition of 20 M Me 2 Spm fully reversed growth inhibition. Polyamine analyses showed that under these conditions, Me 2 Spm was not converted to MeSpd to any appreciable extent. As also shown in Fig. 3, a combination of Me 2 Spm with MDL72527 (inhibiting PAO and SMO) equally effectively reversed DFMO-induced growth inhibition (Me 2 Spm, MDL72527, and their combination did not have any effect on cell growth in the absence of DFMO; results not shown). These results can be understood in terms that Me 2 Spm fulfills the requirements for the natural polyamines without being converted to MeSpd.

DISCUSSION
The natural polyamines spermidine and spermine are ultimately converted to putrescine via the concerted action of SSAT and PAO. As PAO strongly prefers acetylated polyamines to the unmodified spermidine and spermine, SSAT is generally considered as the rate-controlling enzyme of the back-conversion pathway. However, when working with SSATdeficient mouse embryonic stem cells we found that SSAT is absolutely necessary for the conversion of spermidine to putrescine, but not for the degradation of spermine to spermidine (19). In fact, the targeted cells appeared to convert spermine to spermidine much more efficiently than did their wild-type counterparts (19). The conversion of spermine to spermidine in the absence of SSAT activity is obviously attributable to a recently discovered oxidase, which, when first cloned, was thought to be PAO (20), but was later identified as a novel flavin-containing spermine oxidase (21). Unlike PAO, SMO strongly prefers spermine to its acetylated derivatives. Spermidine is not degraded at all, but monoethylspermine is as a good substrate for SMO as is spermine (21,22). SMO, although inhibited to some extent, is much more resistant to the PAO inhibitor MDL72527 (21).
The present results have revealed that both MeSpm and Me 2 Spm serve as substrates for both PAO and SMO in vivo and in vitro. MeSpm yielded both spermidine and MeSpd, whereas Me 2 Spm was converted only to MeSpd. Like in the case of MeSpd (4), it is highly unlikely that doubly methylated spermine would serve as a substrate for SSAT. This view is supported by the findings that in the presence of benzaldehyde in vitro (Table I) Me 2 Spm was converted to MeSpd much more effectively than MeSpm, whereas in vivo (Fig. 1,C and D), the latter compound yielded many times more MeSpd than its bis-methylated counterpart.
The results obtained with crude liver extracts and transgenic animals were largely confirmed with the use of purified recombinant PAO, SMO, and SSAT. Whereas PAO strongly preferred N 1 -acetylspermidine as a substrate, SMO most effectively degraded spermine, but did not use spermidine or any of its tested analogs as substrates. The methyl derivatives of spermine were readily metabolized by SMO and PAO (in the presence of benzaldehyde) (Table II). PAO did not use MeSpd as a substrate and TABLE III Accumulation of polyamines, their analogs, and thymidine incorporation in regenerating liver Transgenic rats were partially hepatectomized and sacrificed at 24 h. The polyamine analogs were given 20 and 4 h before operation (5 mg or 25 mg/kg). Ten Ci of tritiated thymidine was given 30 min before sacrifice. There were 3 to 4 animals in each group. The abbreviations used are: Put, putrescine; Spd, spermidine; Spm, spermine.  it was only poorly metabolized when supplemented with benzaldehyde (Table II). The recombinant SSAT studies confirmed that MeSpd and Me 2 Spm were not acetylated. However, they were competitive inhibitors of SSAT.
Our earlier studies with transgenic rats overexpressing SSAT have indicated that partial hepatectomy of these animals resulted in an induction of SSAT and a profound depletion of the hepatic spermidine pool that was associated with failure to initiate liver regeneration (5,18). Liver regeneration could be fully restored with prior administration of MeSpd (5). Based on these studies and earlier work indicating that partial hepatectomy rapidly elevates the hepatic spermidine, but not spermine, pool, we assigned a critical role to spermidine in liver regeneration (5,18). However, the present results indicated that liver regeneration, as judged by thymidine incorporation, was equally well restored by Me 2 Spm, even at lower concentrations than required for MeSpd (Table III), indicating that spermidine and spermine may be fully exchangeable in this system. Straightforward interpretation of the experimental results with the transgenic animals are somewhat complicated by the fact that Me 2 Spm was converted to MeSpd to some extent. However, according to the present results this conversion appeared to be slow yielding insufficient levels of MeSpd for the correction of the polyamine depletion and initiation of liver regeneration (Table III). To minimize Me 2 Spm conversion to MeSpd, we used cell cultures with low Me 2 Spm concentration. In cultured cells, DFMO-induced growth inhibition was fully reversed by Me 2 Spm in the absence of any conversion of the latter compound to MeSpd. The combination of Me 2 Spm with MDL72527 equally effectively reversed the antiproliferative effect of DFMO thus totally excluding the contribution of MeSpd. These results confirmed the observed restoration of liver regeneration by Me 2 Spm as directly attributable to a polyamine (spermine) function, as Me 2 Spm cannot be converted to deoxyhypusine (3). Our present study shows the necessity of combining in vivo and in vitro methods to indisputably confirm the conclusions drawn in modern biochemical experiments and clearly shows the value of the analogs in the polyamine metabolism studies.