Polyamines play major roles in ionic and osmotic regulation, but their exact involvement in specific ion transport processes is poorly defined. Treatment of L1210 mouse leukaemia cells with either 5 mM alpha-difluoromethylornithine (DFMO), a suicide substrate of ornithine decarboxylase, or 25 microM N1,N12-bis(ethyl)spermine (BE-3-4-3), a dysfunctional polyamine analogue, caused a stable decreased in intracellular pH (pHi) by 0.1-0.4 unit from steady-state control values between 7.4 and 7.6, as measured either by partition of a weak acid or with a fluorescent pH-sensitive probe. This effect was not related to cell growth status or differences in metabolic acid generation, and was observed in either the presence or absence of HCO3-. Exogenous spermidine (10-25 microM) or putrescine (25-50 microM) fully reversed DFMO- or BE-3-4-3-induced acidification within 2 and 8 h respectively. Recovery of pHi in L1210 cells after a nigericin- or NH4(+)-mediated acid load in HCO3(-)-free buffers was mediated by Na+/H+ antiporter activity, in addition to a minor Na(+)-independent and amiloride-insensitive pathway. Decreased steady-state pHi was maintained in polyamine-depleted L1210 cells after recovery from acid stress. Moreover, the pHi-dependence of the rate of Na(+)-dependent H+ extrusion after an acid stress was altered by DFMO and BE-3-4-3, resulting in a set-point which was lower by 0.25-0.30 pH unit in polyamine-depleted cells. On the other hand, neither the rate nor the magnitude of Na+/H(+)-exchanger-mediated alkalinization induced by hypertonic shock was decreased by polyamine depletion. Thus polyamine depletion induces a persistent defect in pHi homeostasis which is due, at least in part, to a stable decrease in the pHi set-point of the Na+/H+ exchanger.

Deregulation of polyamine transport in L1210 cells overexpressing ornithine decarboxylase leads to a lethal accumulation of spermidine. We now provide evidence that over-accumulation of natural and synthetic polyamines, but not putrescine, rapidly induces apoptosis, as shown by hypercondensation of peripheral chromatin and internucleosomal cleavage, followed by nuclear fragmentation. Polyamine oxidation is not responsible for the apoptosis observed. Thus, abnormally high polyamine pools could be an important physiological trigger of apoptosis.

Spermidine/spermine N1-acetyltransferase (SSAT) is the rate-limiting enzyme for the degradation and excretion of polyamines in mammalian cells, and its activity is known to be increased enormously on exposure to polyamines and polyamine analogues. The mechanism by which such an analogue, BESM [N1N12-bis(ethyl)spermine], increases the content of SSAT was investigated by transfecting COS-7 cells with plasmids containing SSAT cDNA in the pEUK expression vector. Despite a large increase in mRNA production, there was only a very small increase in SSAT activity in the transfected cells. When BESM was added at 36 h after transfection, there was a large and very rapid increase in SSAT protein amounting to 380-fold in 12 h without any increase in the mRNA. SSAT protein turned over very rapidly, with a half-life of about 20 min. In the presence of BESM, this turnover was greatly reduced, and the half-life increased to more than 13 h. However, this increase was not sufficient to account for all of the increase in SSAT protein, suggesting that there is also regulation of the translation of the mRNA by BESM. Further evidence for such translation regulation was obtained by studying the polysomal distribution of the SSAT mRNA. In the absence of BESM, most of the mRNA was present in fractions which sedimented more slowly than the monoribosome peak. In BESM-treated cells, a significant proportion of the SSAT mRNA was moved into the small-polysome region of the gradient. The expression of SSAT and the effects of BESM on the polysomal distribution of SSAT mRNA were not affected by the 5′- or 3′-untranslated regions of the mRNA, since constructs which lacked all of these regions gave similar results to constructs containing the entire mRNA sequence. These results show that the increased transcription of the SSAT gene that occurs in the presence of polyamine analogues such as BESM is not sufficient for SSAT expression and that post-transcriptional regulation is critical for the control of SSAT content.

The abilities of the natural polyamines, spermidine and spermine, and of the synthetic analogues, 1-methylspermidine and 1,12-dimethylspermine, to reverse the effects of the S-adenosyl-L-methionine decarboxylase inhibitor 5′-([(Z)-4-aminobut-2-enyl]methylamino)-5′-deoxyadenosine (AbeAdo) on L1210-cell growth were studied. L1210 cells were exposed to AbeAdo for 12 days to induce cytostasis and then exposed to spermidine, spermine, 1-methylspermidine or 1,12-dimethylspermine in the continued presence of AbeAdo. AbeAdo-induced cytostasis was overcome by the natural polyamines, spermidine and spermine. The cytostasis was also reversed by 1-methylspermidine. 1,12-Dimethylspermine had no effect on the AbeAdo-induced cytostasis of chronically treated cells, although it was active in permitting growth of cells treated with the ornithine decarboxylase inhibitor, alpha-difluoromethylornithine. The initial 12-day exposure to AbeAdo elevated intracellular putrescine levels, depleted intracellular spermidine and spermine, and resulted in the accumulation of unmodified eukaryotic translation initiation factor 5A (eIF-5A). Exposure of these cells to exogenous spermidine, which is the natural substrate for deoxyhypusine synthase, resulted in a decrease in the unmodified eIF-5A content. 1-Methylspermidine, which was found to be a substrate of deoxyhypusine synthase in vitro, also decreased the levels of unmodified eIF-5A in the AbeAdo-treated cells. Although spermine is not a substrate of deoxyhypusine synthase, spermine was converted into spermidine in the L1210 cells, and spermine addition to AbeAdo-treated cells resulted in the appearance of both intracellular spermine and spermidine and in the decrease in unmodified eIF-5A. Exogenous 1,12-dimethylspermine, which was not metabolized to spermine or to 1-methylspermidine and was not a substrate of deoxyhypusine synthase in vitro, did not decrease levels of unmodified eIF-5A. The finding that AbeAdo-induced cytostasis was only reversed by polyamines and polyamine analogues that result in the formation of hypusine or an analogue in eIF-5A is consistent with the hypothesis [Byers, Wiest, Wechter and Pegg (1993) Biochem. J. 290, 115-121] that AbeAdo-induced cytostasis is due to the depletion of the hypusine-containing form of eIF-5A, which is secondary to the depletion of spermidine by inhibition of S-adenosyl-L-methionine decarboxylase.

The regulation of polyamine transport and the roles of polyamine transport and synthesis in cell growth were investigated using cultured Chinese hamster ovary (CHO) cells and CHOMG cells which are mutants lacking polyamine-transport activity. Metabolically stable methylated polyamine analogues were used to measure polyamine accumulation, and the irreversible S-adenosyl-L-methionine decarboxylase inhibitor, 5′-([(Z)-4-amino-2-butenyl]methylamino)-5′-deoxyadenosine (AbeAdo), was used to inhibit synthesis. Exposure to AbeAdo lead to a dose-dependent decrease in growth for both cell lines, although CHOMG cells were more sensitive. Intracellular putrescine levels were greatly increased in AbeAdo-treated CHO cells and to a lesser extent in CHOMG cells, whereas intracellular spermidine and spermine levels were substantially reduced in both. Treatment with AbeAdo increased putrescine content in the culture medium to a much greater extent in CHOMG cultures indicating that a portion of the excess putrescine synthesized in response to AbeAdo treatment is excreted, but that CHO cells salvage this putrescine whereas it is lost to CHOMG cells which cannot take up polyamines. AbeAdo treatment increased polyamine transport into CHO cells despite high intracellular putrescine, suggesting that spermidine and/or spermine, and not putrescine, are the major factors regulating transport activity. The accumulation of either 1-methylspermidine or 1,12-dimethylspermine was significantly increased by AbeAdo treatment. Accumulation was increased even further when protein synthesis was blocked by cycloheximide, indicating that a short-lived protein is involved in the regulation of polyamine uptake. In the presence of cycloheximide and AbeAdo or alpha-difluoromethylornithine, methylated polyamine derivatives accumulated to very high levels leading to cell death. These results show that the polyamine-transport system plays an important role in retaining intracellular polyamines and that down-regulation of the transport system in response to increased intracellular polyamine content is necessary to prevent accumulation of toxic levels of polyamines.

S-Adenosylmethionine decarboxylase (AdoMetDC), a rate-limiting enzyme in polyamine biosynthesis, is regulated by polyamines at the levels of both transcription and translation. Two unusual features of AdoMetDC mRNA are a long (320 nt) 5′-untranslated region (5′UTR), which is thought to contain extensive secondary structure, and a short (15 nt) open reading frame (ORF) within the 5′UTR. We have studied the effects of altering these elements on both the expression of AdoMetDC and its regulation by n-butyl-1,3-diaminopropane (BDAP), a spermine synthase inhibitor. Human AdoMetDC cDNAs containing alterations in the 5′UTR, as well as chimaeric constructs in which the AdoMetDC 5′UTR was inserted ahead of the luciferase-coding region, were transfected into COS-7 cells. Construct pSAM320, which contains all of the 5′UTR, the AdoMetDC protein-coding region and the 3′UTR, was expressed poorly (2-fold over the endogenous activity). Deletion of virtually the entire 5′UTR, leaving nt -12 to -1, increased expression 59-fold, suggesting that 5′UTR acts as a negative regulator. The same effect was seen when the 27 nt at the extreme 5′ end were removed (pSAM293, 47-fold increase), or when the internal ORF which is present in this region was destroyed by changing the ATG to CGA (pSAM320-ATG, 38-fold increase). The expression and regulation of pSAM44 (made by deleting nt -288 to -12), which has very little predicted secondary strucutre, was very similar to that of pSAM320 indicating that the terminal 27 nt including the internal ORF rather than extensive secondary structure may be responsible for the low basal levels of AdoMetDC expression. These results, confirmed using luciferase constructs, suggest that the negative effect on expression is predominantly due to the internal ORF. Depletion of spermine by BDAP increased the expression from pSAM320 more than 5-fold without affecting AdoMetDC mRNA levels. Expression from pSAM293 was unchanged by spermine depletion, whereas that from pSAM320-ATG was increased 2.5-fold. These results indicate the presence of a spermine response element in the first 27 nt of the 5′UTR that may include but is not entirely due to the internal ORF.

N1N12-Bis(ethyl)spermine (BESM) and related compounds are powerful inhibitors of cell growth that may have potential as anti-neoplastic agents [Bergeron, Neims, McManis, Hawthorne, Vinson, Bortell and Ingeno (1988) J. Med. Chem. 31, 1183-1190]. The mechanism by which these compounds bring about their effects was investigated by using variant cell lines in which processes thought to be altered by these agents are perturbed. Comparisons between the response of these cells and of their parental equivalents to BESM, N1N11-bis(ethyl)norspermine, N1N14-bis(ethyl)homospermine and N1N8-bis(ethyl)spermidine were then made. It was found that D-R cells, an L1210-derived line that over-expresses ornithine decarboxylase, were not resistant to these compounds. This indicates that the decrease in ornithine decarboxylase is not critical for the action of the compounds on cell growth. Furthermore, although polyamine levels were decreased in the D-R cells, the content was not totally depleted, indicating that such depletion is also not essential for the anti-proliferative effect. Two cell lines lacking mitochondrial DNA (human 143B206 cells and chicken DU3 cells) did not differ in sensitivity to BESM from their parental 143BTK- and DU24 cells. Furthermore, the inhibition of respiration in L1210 cells in response to BESM developed more slowly than the inhibition of growth. Thus it appears that the inhibitions of mitochondrial DNA synthesis and of mitochondrial respiration are also not primary factors in the anti-proliferative effects of these polyamine analogues. The inhibition of growth did, however, correlate with the intracellular accumulation of the analogues. It appears that the bis(ethyl)polyamine derivatives act by binding to intracellular target molecules and preventing macromolecular synthesis. The decline in normal polyamines may facilitate such binding, but is not essential for growth arrest.

We have previously reported that prolonged chronic exposure to the S-adenosyl-L-methionine decarboxylase (AdoMetDC) inhibitor, 5′-([(Z)-4-amino-2-butenyl]methylamino)-5′-deoxy-adenosine (MDL 73811, AbeAdo), leads to cytostasis of L1210 cells [Byers, Ganem and Pegg (1992) Biochem. J. 287, 717-724]. Further studies to investigate the mechanism by which these effects are brought about were carried out by comparing an L1210-derived cell line (R20) that is resistant to AbeAdo with the parent cells. The R20 cells were derived by two rounds of AbeAdo-induced cytostasis followed by rescue with exogenous polyamines. Cytostasis was induced in L1210 cells treated for 12 days with 10 microM AbeAdo; however, exposure to up to 40 microM AbeAdo did not induce cytostasis in R20 cells. Putrescine levels were elevated and spermine levels were depleted in both treated L1210 and treated R20 cells. Spermidine was depleted in treated L1210 cells but was only partly reduced in treated R20 cells. AdoMetDC activity was below the limit of detection in treated L1210 cells but, although greatly reduced, could be measured in the treated R20 cells. The resistance of the R20 cells to the effects of AbeAdo on cell growth and spermidine depletion correlated with reduced AbeAdo accumulation by R20 cells. In the absence of spermidine synthesis, unhypusinated eukaryotic translation initiation factor 5A (eIF-5A) accumulated in AbeAdo-treated L1210 cells. There was no detectable accumulation of unhypusinated eIF-5A in R20 cells. Unhypusinated eIF-5A accumulated during AbeAdo treatment was depleted in L1210 cells rescued by exogenous spermidine. These findings are consistent with the hypothesis that AbeAdo-induced cytostasis is due to the loss of hypusinated eIF-5A. However, spermine was able to rescue AbeAdo-treated L1210 cells without significantly reducing the unhypusinated eIF-5A accumulated during AbeAdo treatment, suggesting that only a small amount of the unmodified protein must be hypusinated to restore cell growth.

The effects of addition of exogenous spermidine and spermine and of two inhibitors of polyamine biosynthesis, alpha-difluoromethylornithine (DFMO), which decreases spermidine concentrations, and n-butyl-1,3-diaminopropane, which depletes spermine, on the expression of S-adenosylmethionine decarboxylase (AdoMetDC) activity were studied in mammalian cell lines (HT29, CHO and COS-7). AdoMetDC levels were inversely related to the polyamine content, and spermine was the more potent repressor of AdoMetDC activity, but only spermidine affected the amount of AdoMetDC mRNA. Transfection of COS-7 cells or CHO cells with plasmid constructs containing a chloramphenicol acetyltransferase (CAT) reporter gene driven by portions of the AdoMetDC promoter region indicated that CAT expression was altered by spermidine, but not by spermine, suggesting that there is a spermidine-responsive element in this promoter. Transient transfection of COS-7 cells with pSAMh1, a plasmid containing the AdoMetDC cDNA in a vector with the SV40 promoter and origin of replication, led to a large increase in AdoMetDC expression. Although treatment of COS-7 cells with n-butyl-1,3-diaminopropane greatly increased endogenous AdoMetDC activity, the spermine depletion brought about by this inhibitor did not stimulate AdoMetDC expression from pSAMh1. The pSAMh1 cDNA is missing 72 nucleotides from the 5′ end of the AdoMetDC mRNA, and it is possible that translational regulation by spermine involves this region. The expression of AdoMetDC from pSAMh1 in COS-7 cells was greatly inhibited by DFMO treatment, although endogenous AdoMetDC activity was increased. The expression of other plasmids containing the SV40 origin of replication was also inhibited by DFMO in COS-7 cells, but not in CHO cells. DFMO treatment did not interfere with the expression of plasmids driven by the RSV promoter. These results suggest that low spermidine levels interfere with the replication of plasmids containing the SV40 origin of replication.

The effects of inhibition of the capacity to form spermidine and spermine on cell growth were investigated using murine leukaemia L1210 cells and 5′-([(Z)-4-amino-2-butenyl]methylamino)-5′-deoxyadenosine (MDL 73811, AbeAdo), an enzyme-activated irreversible inhibitor of S-adenosyl-L-methionine decarboxylase. Putrescine levels were increased 80-fold, and spermidine and spermine levels were greatly reduced after a 3-day exposure to a maximally inhibitory dose of 10 microM-AbeAdo. Addition of AbeAdo to the culture medium inhibited the growth of L1210 cells measured 3 days later in a dose-dependent manner, but, even at a dose of 10 microM, which was maximally effective, exposure to AbeAdo was not immediately cytostatic. However, the growth rate of L1210 cells chronically exposed to 10 microM-AbeAdo declined steadily until day 12, when the cells stopped growing. L1210 cells exposed to AbeAdo for 12 days could not be rescued from cytostasis by removal of the drug from the culture, but could be rescued by exposure to exogenous spermidine or spermine, indicating that the growth-inhibitory effects of AbeAdo were a result of spermidine and/or spermine depletion. It is suggested that elevated intracellular putrescine in AbeAdo-treated cells sustained limited growth in the absence of physiological levels of spermidine and spermine until certain critical and specific physiological role(s) fulfilled by spermidine (and/or spermine) became deficient resulting in cytostasis. N-(3-Aminopropyl)-1,4-diamino-cis-but-2-ene, a spermidine analogue that is a substrate for deoxyhypusine synthase, was able to mimic the effects of spermidine in reversing AbeAdo-induced cytostasis. Spermidine analogues such as 5,5-dimethylspermidine, which are not substrates for deoxyhypusine synthase, were not active in this way. These results provide evidence that the formation of hypusine in the protein-synthesis initiation factor eIF-5A may be a critical role of spermidine essential for cell growth.

The importance of certain amino acid residues in mammalian ornithine decarboxylase activity and degradation was studied by site-specific mutagenesis. Changes were made to the mouse ornithine decarboxylase cDNA in a plasmid containing a T7 RNA polymerase promoter. The plasmid was then used for the synthesis of RNA, which was translated in a reticulocyte lysate system. The activity of the ornithine decarboxylase formed and the stability of the protein to degradation in a reticulocyte lysate system were determined. Changes of lysine-169 or of histidine-197 to alanine completely abolished enzyme activity, indicating that these residues are essential for enzyme activity. The removal of the C-terminal 36 residues, the mutation of lysine-349 to alanine, of lysine-298 to alanine or the double change of serine-303 and glutamic acid-308 to alanine residues still resulted in an active enzyme. The last-mentioned finding indicates that the phosphorylation of serine-303 does not play an essential role in the catalytic activity of ornithine decarboxylase. The control ornithine decarboxylase protein was degraded rapidly in a reticulocyte lysate provided that ATP was added. The truncated protein missing the 36 residues from the C-terminus was much more stable in this system, and the protein containing the double change of serine-303 and glutamic acid-308 to alanine residues was slightly more stable than control ornithine decarboxylase protein. These results indicate that the altered residues may play a role in interaction with factors responsible for the rapid turnover of ornithine decarboxylase.

Synthetic unsaturated analogues of the natural polyamine were examined as possible substrates for spermine synthase and as replacements for spermidine in supporting the growth of SV-3T3 cells. It was found that N-(3-aminopropyl)-1,4-diamino-cis-but-2-ene [the cis isomer of the alkene analogue of spermidine] was a good substrate for spermine synthase, but that the trans isomer [N-(3-aminopropyl)-1,4-diamino-trans-but-2-ene] and the alkene analogue [N-(3-aminopropyl)-1,4-diaminobut-2-yne] were not substrates. These results provide the first demonstration of stereospecificity in the spermine synthase reaction. All three of the unsaturated spermidine analogues described above and the cis-alkene analogue of spermine [N1N4-bis-(3-aminopropyl)-1,4-diamino-cis-but-2-ene] were able to support the growth of SV-3T3 cells that were prevented from the endogenous synthesis of spermidine by treatment with alpha-difluoromethylornithine. Since N-(3-aminopropyl)-1,4-diamino-trans-but-2-ene] and N-(3-aminopropyl)-1,4-diaminobut-2-yne were not converted into a spermine derivative, it is apparent that this conversion is not needed for the stimulation of growth. However, since N1N4-bis-(3-aminopropyl)-1,4-diamino-cis-but-2-ene was also able to support growth and was not degraded to the spermidine derivative, it appears that either polyamine can be effective in this respect. All of the unsaturated analogues tested accumulated in the SV-3T3 cells to a much greater extent than spermidine itself. This indicates that these compounds are substrates for the polyamine transport system, but that they are less effective than the natural polyamines in the feedback regulation of this system.

The cytotoxic response of the human large cell lung carcinoma line NCI H157 to exposure to the polyamine analogue N1 N12-bis(ethyl)spermine (BESpm) is preceded by an extremely high induction of spermidine/spermine N1-acetyltransferase (SSAT). The human enzyme has now been purified greater than 300-fold to apparent homogeneity and found to cross-react with antisera raised against rat liver SSAT. Although other acetylases are capable of acetylating polyamines using acetyl-CoA as the acetyl donor, the greater than 600-fold induction within 24 h was found to be specifically SSAT, since essentially all activity was precipitable by the specific antisera. The human enzyme appears to be similar to the rat enzyme in subunit size under reducing conditions (approximately 20 kDa), substrate specificity and kinetic parameters. Preliminary results using actinomycin D and cycloheximide suggested that the unusually high induction by N1 N12-bis(ethyl)spermine in the human lung cancer line result from new mRNA and protein synthesis. This hypothesis is further substantiated here by ‘in vitro’ translation experiments comparing poly(A) mRNA from control and treated cells. The large cell lung carcinoma line NCI H157 represents a useful system to produce large amounts of the SSAT protein and to study the molecular events responsible for the induction and control of this important polyamine-metabolic enzyme. By using this rich source of SSAT protein, a partial amino acid sequence was determined by N-terminal sequencing of endoproteinase Lys-C digestion fragments. Further, this system should be useful in determining whether there is an association between the unusually high induction of the acetylase and the observed cytotoxicity in the NCI H157 cells.

In the present study, the spermine (SPM) analogue N1N12-bis(ethyl)spermine (BESPM) is compared with SPM in its ability to regulate ornithine decarboxylase (ODC) and S-adenosyl-L-methionine decarboxylase (AdoMetDC) activities in intact L1210 cells and in the mechanism(s) by which this is accomplished. Unlike the comparable spermidine (SPD) analogue N1N8-bis(ethyl)spermidine, which regulates only ODC, BESPM suppresses both ODC and AdoMetDC activities. With 1 microM-SPM or -BESPM, near-maximal suppression of enzyme activity (i.e. less than 70%) was achieved after 2 h for ODC and 12 h for AdoMetDC. After such treatment, ODC activity fully recovered within 2-4 h, and that of AdoMetDC within 12 h, when cells were reseeded into drug-free media. It was deduced that an intracellular accumulation of BESPM or SPM equivalent to only approximately 200-450 pmol/10(6) cells was sufficient to fully invoke ODC regulatory mechanisms. Decreases in both enzyme activities after BESPM or SPM treatment were closely paralleled by concomitant decreases in the amount of enzyme protein. Since cellular ODC or AdoMetDC mRNA was not similarly decreased by either BESPM or SPM treatment, it was concluded that translational and/or post-translational mechanisms were probably responsible for enzyme regulation. In support of the former of these possibilities, it was demonstrated that both BESPM and SPM preferentially inhibited the translation in vitro of ODC and AdoMetDC relative to albumin in a reticulocyte-lysate system. On the basis of the consistent similarities between BESPM and SPM in all parameters studied, it is concluded that the analogue most likely acts by mechanisms identical with those by which SPM acts in suppressing polyamine biosynthesis.

Treatment of Chinese-hamster ovary (CHO) cells with N1N11-bis(ethyl)norspermine (BENSM) led to a very large increase in the activity of spermidine/spermine N1-acetyltransferase (SAT), which rose by about 600-fold within 48 h. Smaller, but still very large increases, were also produced in decreasing order of potency by 3,7,11,15,19-penta-azaheneicosane, N1N12-bis(ethyl)spermine and by N1N14-bis(ethyl)homospermine. The rise in acetyltransferase activity was due to an increase in enzyme protein, as indicated by immunoblotting using antibodies directed against rat liver SAT. There was an increase in the content of mRNA for SAT, indicating that BENSM regulates the level of enzyme protein partly by means of a change in transcription or stability of the mRNA. There was also a decreased rate of degradation of the protein in CHO cells trated with the drug. This may be due to the binding of BENSM, which is a competitive inhibitor of the enzyme with a Ki of 120 microM. Exposure to BENSM led to an increased conversion of spermidine into N1-acetylspermidine and putrescine, a rapid fall in the content of intracellular polyamines and the excretion from the cell of putrescine, N1-acetylspermidine and spermidine. When polyamine oxidase activity in the treated cells was blocked, increases in N1-acetylspermidine and N1-acetylspermine were much greater, and the formation of putrescine was prevented. These results indicate that the induction of SAT facilities the degradation of spermine and spermidine to putrescine and the subsequent excretion of putrescine from the cell. When the degradation of the N1-acetyl derivatives by polyamine oxidase is blocked, the cells excrete N1-acetylspermidine instead of putrescine. CHO cells also contained and excreted N8-acetylspermidine, but its synthesis was not increased in cells treated with BENSM, confirming data obtained in vitro that SAT does not produce this derivative.

A molecular-genetic approach towards isolating mammalian polyamine-transport genes and their encoded proteins was devised involving the production of Chinese-hamster ovary (CHO) cells expressing a human polyamine-transport protein. CHO cells and a polyamine-transport-deficient CHO mutant cell line (CHOMG) were equally sensitive to the antiproliferative effects of alpha-difluoromethylornithine (DFMO), which blocked endogenous polyamine synthesis. Exposure to exogenous polyamines increased intracellular polyamine levels and reversed this DFMO-induced cytostasis in the CHO cells, but not in the CHOMG cells. CHOMG cells were therefore transfected with human DNA (isolated from HT-29 colon carcinoma cells) and cells expressing the human polyamine-transport system were identified by the ability of these cells to grow in a medium containing DFMO and polyamines. A number of different positive clones were identified and shown to have the capacity for polyamine uptake and an increased sensitivity to the toxic effects of the polyamine analogue methylglyoxal bis(guanylhydrazone). Differences in these properties between the clones are consistent with a multiplicity of polyamine-transport systems. Some clones also showed a change in growth characteristics, which may indicate a relationship between genes involved in the polyamine-transport system and in cell proliferation.

1-Amino-oxy-3-aminopropane (AOAP) was reported to inhibit several mammalian polyamine-biosynthetic enzymes in vitro, including ornithine decarboxylase (ODC) and S-adenosylmethionine decarboxylase (AdoMetDC) [Khomutov, Hyvönen, Karvonen, Kauppinen, Paalanen, Paulin, Eloranta, Pajula, Andersson & Pösö (1985) Biochem. Biophys. Res. Commun. 130, 596-602]. In order to clarify its mechanism of action in intact cells, the inhibitory properties of AOAP on the growth and polyamine metabolism of L1210 cells were compared with those seen in a variant subline (D-R cells) which overproduces ODC. As little as 20 microM-AOAP completely blocked proliferation of L1210 cells, and this effect was reversed by the concomitant addition of exogenous putrescine or spermidine. Growth of D-R cells was not affected by AOAP at concentrations up to 0.5 mM. There was no difference in the uptake of AOAP between the L1210 and the D-R cells. Exposure of L1210 or D-R cells to AOAP greatly decreased ODC activity in undialysed cell extracts, but did not decrease AdoMetDC. Activities of both enzymes were increased severalfold by AOAP treatment when activity was measured in dialysed extracts. Treatment with AOAP depleted intracellular putrescine and spermidine contents of L1210 cells, while inducing a massive accumulation of decarboxylated AdoMet. The 8-fold higher putrescine pool present in untreated D-R cells was depleted in a dose-dependent manner by AOAP, but a significant decrease in spermidine and accumulation of decarboxylated AdoMet required 10 times higher drug concentrations, and the changes were much less dramatic than in L1210 cells. These results indicate that in L1210 cells AOAP behaves primarily as a reversible inhibitor of ODC.

The rate-limiting enzymes in polyamine biosynthesis, ornithine decarboxylase (ODC) and S-adenosylmethionine decarboxylase (AdoMetDC), are negatively regulated by the polyamines spermidine and spermine. In the present work the spermidine synthase inhibitor S-adenosyl-1,8-diamino-3-thio-octane (AdoDATO) and the spermine synthase inhibitor S-methyl-5′-methylthioadenosine (MMTA) were used to evaluate the regulatory role of the individual polyamines. Treatment of Ehrlich ascites-tumour cells with AdoDATO caused a marked decrease in spermidine content together with an accumulation of putrescine and spermine. Treatment with MMTA, on the other hand, gave rise to a marked decrease in spermine, with a simultaneous accumulation of spermidine. A dramatic increase in the activity of AdoMetDC, but not of ODC, was observed in MMTA-treated cells. This increase appears to be unrelated to the decrease in spermine content, because a similar rise in AdoMetDC activity was obtained when AdoDATO was given in addition to MMTA, in which case the spermine content remained largely unchanged. Instead, we show that the increase in AdoMetDC activity is mainly due to stabilization of the enzyme, probably by binding of MMTA. Treatment with AdoDATO had no effects on the activities of ODC and AdoMetDC, even though it caused a precipitous decrease in spermidine content. The expected decrease in spermidine-mediated suppression of ODC and AdoMetDC was most probably counteracted by the simultaneous increase in spermine. The combination of AdoDATO and MMTA caused a transient rise in ODC activity. Concomitant with this rise, the putrescine and spermidine contents increased, whereas that of spermine remained virtually unchanged. The increase in ODC activity was due to increased synthesis of the enzyme. There were no major effects on the amount of AdoMetDC mRNA by treatment with the inhibitors, alone or in combination. However, the synthesis of AdoMetDC was slightly stimulated in cells treated with MMTA or AdoDATO plus MMTA. The present study demonstrates that regulation of neither ODC nor AdoMetDC is a direct function of the polyamine structure. Instead, it appears that the biosynthesis of the polyamines is feedback-regulated by the various polyamines at many different levels.

A number of synthetic polyamine derivatives that included five achiral gem-dimethylspermidines and two analogous tetramethylated spermines were tested for their abilities to serve as substrates for enzymes metabolizing polyamines and for their capacities to substitute for the natural polyamines in cell growth. It was found that none of the compounds were effective substrates for spermine synthase, and only one, namely 8,8-dimethylspermidine, was a substrate for spermidine/spermine N1-acetyltransferase. However, all of the spermidine derivatives and 1,1,12,12-tetramethylspermine were able to support the growth of SV-3T3 cells in which endogenous polyamine synthesis was prevented by the addition of alpha-difluoromethylornithine. These results suggest that either spermidine or spermine can support cell growth without the need for metabolic interconversion. In contrast with the result with 1,1,12,12-tetramethylspermine, 3,3,10,10-tetramethylspermine did not restore growth of polyamine-depleted SV-3T3 cells. Comparison of the properties of these derivatives may prove valuable in understanding the physiological role of polyamines.

Treatment of L1210 cells with either of two inhibitors of S-adenosylmethionine decarboxylase (AdoMetDC), namely 5′-deoxy-5′-[N-methyl-N-[2-(amino-oxy)ethyl])aminoadenosine or 5′-deoxy-5′-[N-methyl-N-(3-hydrazinopropyl)]aminoadenosine, produced a large increase in the amount of ornithine decarboxylase (ODC) protein. The increased enzyme content was due to a decreased rate of degradation of the protein and to an increased rate of synthesis, but there was no change in its mRNA content. The inhibitors led to a substantial decline in the amounts of intracellular spermidine and spermine, but to a big increase in the amount of putrescine. These results indicate that the content of ODC is negatively regulated by spermidine and spermine at the levels of protein translation and turnover, but that putrescine is much less effective in bringing about this repression. Addition of either spermidine or spermine to the cells treated with the AdoMetDC inhibitors led to a decrease in ODC activity, indicating that either polyamine can bring about this effect, but spermidine produced effects at concentrations similar to those found in the control cells and appears to be the physiologically important regulator. The content of AdoMetDC protein (measured by radioimmunoassay) was also increased by these inhibitors, and a small increase in its mRNA content was observed, but this was insufficient to account for the increase in protein. A substantial stabilization of AdoMetDC occurred in these cells, contributing to the increased enzyme content, but an increase in the rate of translation cannot be ruled out.

Treatment of SV-3T3 cells with the spermine synthase inhibitor S-methyl-5′-methylthioadenosine [AdoS+(CH3)2] led to a large increase in the activity of S-adenosylmethionine decarboxylase (AdoMetDC) without affecting ornithine decarboxylase. The elevation of AdoMetDC activity was due to an increased amount of enzyme protein, as demonstrated by radioimmunoassay and by immunoblotting. The increase in AdoMetDC protein was caused by at least three factors: an increase in the amount of translatable mRNA, an increased translation efficiency of the mRNA and an increase in the half-life of the protein. The depletion of spermine brought about by AdoS+(CH3)2 appeared to be responsible for the increased synthesis of AdoMetDC and for part of the decrease in its rate of degradation. An additional stabilization of the enzyme protein was probably due to the binding of AdoS+(CH3)2, which is also a weak inhibitor of AdoMetDC. These results demonstrate the importance of cellular spermine concentrations in regulating the activity of AdoMetDC, which is a key enzyme controlling polyamine synthesis.

The synthesis of S-adenosylmethionine (AdoMet) decarboxylase was studied by translating the rat prostate mRNA for this enzyme in a reticulocyte lysate. The protein was formed as a precursor of Mr 37,000, which was converted into the enzyme subunit of Mr 32,000 in the lysates. The presence of putrescine had no effect on the synthesis of the precursor of AdoMet decarboxylase, but accelerated its conversion into the enzyme subunit. Spermidine, spermine, decarboxylated AdoMet, AdoMet and methylglyoxal bis(guanylhydrazone) were not able to substitute for putrescine in this effect. These results indicate that, in addition to its direct activation of mammalian AdoMet decarboxylase, putrescine could increase the amount of the enzyme by increasing its production.

Polyamine biosynthesis in intact cells can be exquisitely controlled with exogenous polyamines through the regulation of rate-limiting biosynthetic enzymes, particularly ornithine decarboxylase (ODC). In an attempt to exploit this phenomenon as an antiproliferative strategy, certain polyamine analogues have been identified [Porter, Cavanaugh, Stolowich, Ganis, Kelly & Bergeron (1985) Cancer Res. 45, 2050-2057] which lower ODC activity in intact cells, have no direct inhibitory effects on ODC, are incapable of substituting for spermidine (SPD) in supporting cell growth, and are growth-inhibitory at micromolar concentrations. In the present study, the most effective of these analogues, N1N8-bis(ethyl)SPD (BES), is compared with SPD in its ability to regulate ODC activity in intact L1210 cells and in the mechanism(s) by which this is accomplished. With respect to time and dose-dependence of ODC suppression, both polyamines closely paralleled one another in their response curves, although BES was slightly less effective than SPD. Conditions of minimal treatment leading to near-maximal ODC suppression (70-80%) were determined and found to be 3 microM for 2 h with either SPD or BES. After such treatment, ODC activity was fully recovered within 2-4 h when cells were re-seeded in drug-free media. By assessing BES or [3H]SPD concentrations in treated and recovered cells, it was possible to deduce that an intracellular accumulation of BES or SPD equivalent to less than 6.5% of the combined cellular polyamine pool was sufficient to invoke ODC regulatory mechanisms. Decreases in ODC activity after BES or SPD treatment were closely paralleled by concomitant decreases in ODC protein. Since cellular ODC mRNA was not similarly decreased by either BES or SPD, it was concluded that translational and/or post-translational mechanisms, such as increased degradation of ODC protein or decreased translation of ODC mRNA, were probably responsible for regulation of enzyme activity. Experimental evidence indicated that neither of these mechanisms seemed to be mediated by cyclic AMP or ODC-antizyme induction. On the basis of the consistent similarities between BES and SPD in all parameters studied, it is concluded that the analogue most probably acts by the same mechanisms as SPD in regulating polyamine biosynthesis.

The mechanism of inactivation of rodent ornithine decarboxylase by alpha-difluoromethylornithine (DFMO) was studied using the inhibitor labelled with 14C in both the 1 and the 5 positions. [1-14C]DFMO was a substrate and was decarboxylated by the enzyme yielding 14CO2. A radioactive metabolite derived from [5-14C]DFMO was bound to the enzyme, and the extent of binding paralleled the irreversible inactivation of ornithine decarboxylase. The partition ratio of decarboxylation to binding was approx. 3.3. These results provide support for the postulated mechanism of action of DFMO [Metcalf, Bey, Danzin, Jung, Casera & Vevert (1978) J. Am. Chem. Soc. 100, 2551-2553], in which enzymic decarboxylation of the inhibitor leads to the generation of a conjugated imine, which then alkylates a nucleophilic residue on the enzyme.

Exposure of rat L6 cells in culture to exogenous polyamines led to a very large increase in the activity of spermidine/spermine N1-acetyltransferase. Spermine was more potent than spermidine in bringing about this increase, but in both cases the elevated acetyltransferase activity increased the cellular conversion of spermidine into putrescine. The N1-acetyltransferase turned over very rapidly in the L6 cells, with a half-life of 9 min after spermidine and 18 min after spermine. A wide variety of synthetic polyamine analogues also brought about a substantial induction of spermidine/spermine N1-acetyltransferase activity. These included sym-norspermidine, sym-norspermine, sym-homospermidine, N4-substituted spermidine derivatives, 1,3,6-triaminohexane, 1,4,7-triaminoheptane and deoxyspergualin, which were comparable with spermidine in their potency, and N1N8-bis(ethyl)spermidine, N1N9-bis(ethyl)homospermidine, methylglyoxal bis(guanylhydrazone), ethylglyoxal bis(guanylhydrazone) and 1,1′-[(methylethanediylidene)dinitrilo]bis(3-amino-guanidine), which were even more active than spermidine. It is suggested that these polyamine analogues may bring about a decrease in cellular polyamines not only by inhibiting biosynthesis but by stimulating the degradation of spermidine into putrescine.

S-Adenosylhomocysteine/5′-methylthioadenosine nucleosidase (EC 3.2.2.9) was purified to homogeneity from Escherichia coli to a final specific activity of 373 mumol of 5′-methylthioadenosine cleaved/min per mg of protein. Affinity chromatography on S-formycinylhomocysteine-Sepharose is the key step of the purification procedure. The enzyme, responsible for the cleavage of the glycosidic bond of both S-adenosylhomocysteine and 5′-methylthioadenosine, was partially characterized. The apparent Km for 5′-methylthioadenosine is 0.4 microM, and that for S-adenosylhomocysteine is 4.3 microM. The maximal rate of cleavage of S-adenosylhomocysteine is approx. 40% of that of 5′-methylthioadenosine. Some 25 analogues of the two naturally occurring thioethers were studied as potential substrates or inhibitors of the enzyme. Except for the analogues modified in the 5′-position of the ribose moiety or the 2-position of the purine ring, none of the compounds tested was effective as a substrate. Moreover, 5′-methylthioformycin, 5′-chloroformycin, S-formycinylhomocysteine, 5′-methylthiotubercidin and S-tubercidinylhomocysteine were powerful inhibitors of the enzyme activity. The results obtained allow the hypothesis of a mechanism of enzymic catalysis requiring as a key step the protonation of N-7 of the purine ring.

Treatment of rats with spermidine, spermine or sym-norspermidine led to a substantial induction of spermidine/spermine N1-acetyltransferase activity in liver, kidney and lung. The increase in this enzyme, which was determined independently of other acetylases by using a specific antiserum, accounted for all of the increased acetylase activity in extracts from rats treated with these polyamines. Spermine was the most active inducer, and the greatest effect was seen in liver. Liver spermidine/spermine N1-acetyltransferase activity was increased about 300-fold within 6 h of treatment with 0.3 mmol/kg doses of spermine; activity in kidney increased 30-fold and activity in the lung 15-fold under these conditions. The increased spermidine/spermine N1-acetyltransferase activity led to a large increase in the liver putrescine content and a decline in spermidine. These changes are due to the oxidation by polyamine oxidase of the N1-acetylspermidine formed by the acetyltransferase. Our results indicated that spermidine was the preferred substrate in vivo of the acetylase/oxidase pathway for the conversion of the higher polyamines into putrescine. The induction of the spermidine/spermine N1-acetyltransferase by polyamines may provide a mechanism by which excess polyamines can be removed.

Comparisons were made of ornithine decarboxylase isolated from Morris hepatoma 7777, thioacetamide-treated rat liver and androgen-stimulated mouse kidney. The enzymes from each source were purified in parallel and their size, isoelectric point, interaction with a monoclonal antibody or a monospecific rabbit antiserum to ornithine decarboxylase, and rates of inactivation in vitro , were studied. Mouse kidney, which is a particularly rich source of ornithine decarboxylase after androgen induction, contained two distinct forms of the enzyme which differed slightly in isoelectric point, but not in Mr. Both forms had a rapid rate of turnover, and virtually all immunoreactive ornithine decarboxylase protein was lost within 4h after protein synthesis was inhibited. Only one form of ornithine decarboxylase was found in thioacetamide-treated rat liver and Morris hepatoma 7777. No differences between the rat liver and hepatoma ornithine decarboxylase protein were found, but the rat ornithine decarboxylase could be separated from the mouse kidney ornithine decarboxylase by two-dimensional gel electrophoresis. The rat protein was slightly smaller and had a slightly more acid isoelectric point. Studies of the inactivation of ornithine decarboxylase in vitro in a microsomal system [Zuretti & Gravela (1983) Biochim. Biophys. Acta 742, 269-277] showed that the enzymes from rat liver and hepatoma 7777 and mouse kidney were inactivated at the same rate. This inactivation was not due to degradation of the enzyme protein, but was probably related to the formation of inactive forms owing to the absence of thiol-reducing agents. Treatment with 1,3-diaminopropane, which is known to cause an increase in the rate of degradation of ornithine decarboxylase in vivo [Seely & Pegg (1983) Biochem. J. 216, 701-717] did not stimulate inactivation by microsomal extracts, indicating that this system does not correspond to the rate-limiting step of enzyme breakdown in vivo .

The effects of alpha-difluoromethylornithine, a specific inhibitor of ornithine decarboxylase, on cell growth rate, polyamine content and the content of decarboxylated S-adenosylmethionine in SV-3T3 transformed mouse fibroblasts were studied. DL-alpha-Difluoromethylornithine at 1 mM or higher concentrations decreased the growth rate by over 90% after 2 or more days of exposure, but the cells remained viable, although quiescent for at least 9 days. Addition of 10 microM-spermidine or -spermine or 50 microM-putrescine at any time throughout this period completely reversed the inhibition of growth. Treatment with alpha-difluoromethylornithine decreased putrescine and spermidine contents by more than 98% and that of spermine by 60%, but cells exposed to exogenous polyamines did not require complete replenishment of the polyamine pools to resume growth. In fact, a virtually normal growth rate was obtained in cells lacking putrescine, having 2% of normal spermidine content and 156% of normal spermine. These results suggest that the well-known increase in putrescine and spermidine in cells stimulated for growth is not essential for this to occur and that mammalian cells can utilize spermine as their only polyamine. A substantial reversal of the growth-inhibitory effect of alpha-difluoromethylornithine was produced by a number of polyamines not normally found in mammalian cells, including the spermidine analogues aminopropylcadaverine and sym-homospermidine, which were partially converted into their respective spermine analogues by addition of an aminopropyl group within the cell. The spermine analogue sym-norspermine was also effective, but the maximal growth rate produced by these unphysiological polyamines was only 60-70% of that produced by the normal polyamines. These results indicate that spermidine and spermine have the optimal length for activation of the cellular processes critically dependent on polyamines and should help in identifying these processes. Exposure to alpha-difluoromethylornithine leads to an enormous rise in the concentration of decarboxylated S-adenosylmethionine, which reached a peak at 530-fold after 3 days of exposure and steadily declined to 140-fold after 11 days. This increase was abolished by addition of exogenous polyamines, which rapidly decreased the activity of S-adenosylmethionine decarboxylase. The increase in decarboxylated S-adenosylmethionine is unlikely to be solely responsible for the decrease to the same extent by spermine, sym-norspermidine and sym-homospermidine, which produce 97%, 16% and 60% of the control growth rate, respectively.(ABSTRACT TRUNCATED AT 400 WORDS)

A monoclonal antibody of the immunoglobulin M class was produced against mouse kidney ornithine decarboxylase. Screening for the antibody was carried out using alpha-difluoromethyl[5-3H]ornithine-labelled ornithine decarboxylase. The antibody reacted with this antigen and with native ornithine decarboxylase. The antibody attached to Sepharose could be used to form an immunoaffinity column that retained mammalian ornithine decarboxylase. The active enzyme could then be eluted in a highly purified form by 1.0M-sodium thiocyanate. The monoclonal antibody could also be used to precipitate labelled ornithine decarboxylase from homogenates of kidneys from androgen-treated mice given [35S]methionine. Only one band, corresponding to Mr of about 55000, was observed. The extensive labelling of this band is consistent with the rapid turnover of ornithine decarboxylase protein, since this enzyme represents only about 1 part in 10000 of the cytosolic protein.

A radioimmunoassay for ornithine decarboxylase was used to study the regulation of this enzyme in rat liver. The antiserum used reacts with ornithine decarboxylase from mouse, human or rat cells. Rat liver ornithine decarboxylase enzyme activity and enzyme protein (as determined by radioimmunoassay) were measured in thioacetamide-treated rats at various times after administration of 1,3-diaminopropane. Enzyme activity declined rapidly after 1,3-diaminopropane treatment as did the amount of enzyme protein, although the disappearance of enzyme activity slightly preceded the loss of immunoreactive protein. The loss of enzyme protein after cycloheximide treatment also occurred rapidly, but was significantly slower than that seen with 1,3-diaminopropane. When 1,3-diaminopropane and cycloheximide were injected simultaneously, the rate of disappearance of enzyme activity and enzyme protein was the same as that seen with cycloheximide alone. These results show that the rapid loss in enzyme activity after 1,3-diaminopropane treatment is primarily due to a loss in enzyme protein and that protein synthesis is needed in order for 1,3-diaminopropane to exert its full effect. A macromolecular inhibitor of ornithine decarboxylase that has been termed antizyme is induced in response to 1,3-diaminopropane, but our results indicate that the loss of enzyme activity is not due to the accumulation of inactive ornithine decarboxylase-antizyme complexes. It is possible that the antizyme enhances the degradation of the enzyme protein. Control experiments demonstrated that the antiserum used would have detected any inactive antizyme-ornithine decarboxylase complexes present in liver since addition of antizyme to ornithine decarboxylase in vitro did not affect the amount of ornithine decarboxylase detected in our radioimmunoassay. Anti-(ornithine decarboxylase) antibodies may be useful in the purification of antizyme since the antizyme-ornithine decarboxylase complex can be immunoprecipitated, and antizyme released from the precipitate with 0.3 M-NaCl.

The substrate specificity and kinetic mechanism of spermidine N1-acetyltransferase from rat liver was investigated using a highly purified (18 000-fold) preparation from the livers of rats in which the enzyme was induced by treatment with carbon tetrachloride (1.5 ml/kg body wt. 6h before death). The enzyme catalysed the acetylation of spermidine, spermine, sym-norspermidine, sym-norspermine, N-(3-aminopropyl)-cadaverine, N1-acetylspermine, 3,3′-diamino-N-methyldipropylamine and 1,3-diaminopropane, but was inactive with putrescine, cadaverine, sym-homospermidine and N1-acetylspermidine. These results suggest that the enzyme is highly specific for the acetylation of a primary amino group that is separated by a three-carbon aliphatic chain from another nitrogen atom (i.e. the substrates are of the type H2N[CH2]3NHR). The maximal rates of acetylation of 1,3-diaminopropane and 3,3′-diamino-N-methyldipropylamine were much lower than the maximal rates with spermidine or sym-norspermidine as substrates, suggesting a preference for a secondary amino group bearing the aminopropyl group that is acetylated. The best substrates for acetylation were sym-norspermidine and sym-norspermine, which had Km values of about 10 micrograms and Vmax. values of about 2 mumol of product/min per mg of enzyme compared with Km of 130 microM and Vmax. of 1.3 mumol/min per mg for spermidine. N1-Acetylspermidine (the product of the reaction) and N8-acetylspermidine were weak inhibitors and were competitive with spermidine, having Ki values of about 6.6 mM and 0.4 mM respectively. N1-Acetylspermidine was a non-competitive inhibitor with respect to acetyl-CoA. CoA was also inhibitory to the reaction, showing non-competitive kinetics when either [acetyl-CoA] or [spermidine] was varied. These results suggest that the reaction occurs via an ordered Bi Bi mechanism in which spermidine binds first and N1-acetyl-spermidine is the final product to be released.

Rat liver spermidine/spermine N1-acetyltransferase was found to be strongly inhibited by the dyes Cibacron F3GA, Coomassie Brilliant Blue and Congo Red. Inhibition was competitive with respect to acetyl-CoA and Ki values of 0.7 microM and 52 microM were determined for Cibacron F3GA and Coomassie Brilliant Blue respectively. The enzyme was strongly retained by columns of Affi-Gel Blue, which contains Cibacron F3GA linked to agarose. It was not eluted from this adsorbent in the presence of 10 mM-spermidine/0.5 M-NaCl/50 mM-Tris/HCl, pH 7.5, but was released by 1 mM-CoA in 10 mM-spermidine/50 mM-Tris/HCl, pH 7.5. These results are consistent with the presence in the enzyme of a dinucleotide fold that binds acetyl CoA and has a high affinity for Cibacron F3GA. The spermidine/spermine N1-acetyltransferase was irreversibly inactivated by exposure to butane-2,3-dione in sodium borate, pH 7.8, or by exposure to phenylglyoxal or camphorquinone-10-sulphonic acid. All of these reagents are known to interact with arginine residues in proteins under the conditions in which they inactivated the acetyltransferase. Inactivation was prevented by the presence of acetyl-CoA or CoA, but to a lesser extent by 3′-dephospho-CoA and not at all by NAD or adenosine. This protection suggests that an arginine residue at the active site is involved in the binding of the acetyl-CoA substrate. Treatment of the assay mixture but not the spermidine N1-acetyltransferase with alkaline phosphatase prevented the reaction taking place. This suggests that the apparent loss of enzyme activity in response to alkaline phosphatase reported by Matsui, Otani, Kamei & Morisawa [(1982) FEBS Lett. 150, 211-213] is due to dephosphorylation of the acetyl-CoA substrate and that the 3′-phosphate group is essential for activity.

S-Adenosyl-L-methionine decarboxylases were purified from rat ventral prostate, yeast (Saccharomyces cerevisiae), slime mould (Physarum polycephalum) and bacteria (Escherichia coli) and tested for inhibition by a variety of nucleosides related to S-adenosylmethionine and by methyl- and ethyl-glyoxal bis(guanylhydrazone). Although the enzymes from these different sources are markedly different with respect to activation by cations, the inhibition by nucleosides was quite similar. Very little inhibition was seen when analogues of S-adenosylmethionine with a different base were tested or when the ribose ring was opened or the positive charge on the sulphur atom was not present. Some derivatives in which the amino acid portion of the molecule was altered were more potent inhibitors, but again there was little difference between the enzymes from different sources. 5′-(Dimethylsulphonio)-5′-deoxyadenosine and S-adenosyl-3-methylthiopropylamine were the most inhibitory substances and had similar Ki values, suggesting that the aminopropyl group does not contribute significantly to the binding. All of the S-adenosylmethionine decarboxylases were strongly competitively inhibited by methylglyoxal bis(guanylhydrazone) and even more powerfully by its ethyl analogue, although the putrescine-activated enzymes from prostate and yeast were more sensitive than the bacterial and slime-mould enzymes. All of the S-adenosylmethionine decarboxylases tested bound to a column of methylglyoxal bis(guanylhydrazone) linked to Sepharose and were not eluted by 0.5 M-NaCl, but could be released by 1 mM concentrations of the drug, providing a rapid and efficient method for their purification.

The effects of a number of nucleosides related to 5′-methylthioadenosine on the activities of S-adenosylhomocysteine hydrolase, 5′-methylthioadenosine phosphorylase, spermidine synthase and spermine synthase were investigated. Both 5′-methylthioadenosine and 5′-isobutylthioadenosine gave rise to an enzyme-activated irreversible inhibition of S-adenosylhomocysteine hydrolase, but 5′-methylthiotubercidin (5′-methylthio-7-deaza-adenosine), 5′-deoxy-5′-chloroformycin, 5′-ethylthio-2-fluoro-adenosine and 1,N6-etheno-5′-methylthioadenosine were totally ineffective in producing this inactivation. Of the nucleosides tested, only 5′-methylthioadenosine, 5′-methylthiotubercidin and 5′-isobutylthioadenosine were inhibitory towards the aminopropyltransferases responsible for the synthesis of spermine and spermidine. 5′-Methylthiotubercidin, 5′-deoxy-5′-chloroformycin and 5′-isobutylthioadenosine were inhibitors of the degradation of 5′-methylthioadenosine by 5′-methylthioadenosine phosphorylase, but only 5′-isobutylthioadenosine was also a substrate for this enzyme. These results suggest that the effects of 5′-isobutylthioadenosine of the cell may result from the combination of inhibitory actions on polyamine synthesis, 5′-methylthioadenosine degradation and S-adenosylhomocysteine degradation. The resulting increased concentrations of S-adenosylhomocysteine could bring about inhibition of methyltransferase reactions. A new convenient method for the assay of S-adenosylhomocysteine hydrolase in the direction of synthesis is described.

1. Administration of spermidine or sym-norspermidine decreased the activity of AdoMet (S-adenosylmethionine) decarboxylase in extracts prepared from rat liver, Kidney, psoas, diaphragm, soleus and small intestine, but not heart. The decline in psoas, diaphragm and soleus was much greater than that in liver and kidney. The difference in sensitivity to spermidine could not be explained by changes in the uptake and accumulation of the polyamine, because much higher contents were found in liver and kidney that in diaphragm and psoas. 2. Spermidine administration also led to a substantial increase in putrescine in all tissues examined. However, the rise in putrescine was not responsible for the decline in AdoMet decarboxylase activity, since norspermidine, which cannot form putrescine, also produced the decline. Also, administration of putrescine or 1,3-diaminopropane did not decrease AdoMet decarboxylase. 3. The decline in skeletal-muscle AdoMet decarboxylase activity in response to spermidine may be due to an increased rate of degradation of the enzyme protein. The t1/2 (half-time) for the decline in activity after inhibition of protein synthesis by cycloheximide was almost halved in the psoas of spermidine-treated rats. Spermidine treatment did not change the t1/2 in liver. 4. These results raise the possibility that there are at least two different forms of AdoMet decarboxylase and that the enzyme from psoas or diaphragm differs from that in liver. Additional support for this hypothesis was obtained by comparing the activation by putrescine of AdoMet decarboxylase from these tissues. The liver enzyme was stimulated 10-fold, but the muscle enzyme was stimulated 30-fold.

1. The specificity of rat prostatic spermidine synthase and spermine synthase with respect to the amine acceptor of the propylamine group was studied. 2. Spermidine synthase could use cadaverine (1,5-diaminopentane) instead of putrescine, but the Km for cadaverine was much greater and the rate with 1mM-cadaverine was only 10% of that with putrescine. 1,3-Diaminopropane was even less active (2% of the rate with putrescine) and no other compound tested (including longer alpha, omega-diamines, spermidine and its homologues and monoacetyl derivatives) was active. 3. Spermine synthase was equally specific. The only compounds tested that showed any activity were 1,8-diamino-octane, sym-homospermidine, sym-norspermidine and N-(3-aminopropyl)-cadaverine, which at 1mM gave rates 2, 17, 3 and 4% of the rate with spermidine respectively. 4. The formation of polyamine derivatives of cadaverine and to a very small extent of 1,3-diaminopropane was confirmed by exposing transformed mouse fibroblasts to these diamines when synthesis of putrescine was prevented by alpha-difluoromethylornithine. Under these conditions the cells accumulated significant amounts of N-(3-aminopropyl)cadaverine and NN'-bis(3-aminopropyl)cadaverine when exposed to cadaverine and small amounts of sym-norspermidine and sym-norspermine when exposed to 1,3-diaminopropane.

1. The activity of an enzyme catalysing the loss of O6-methylguanine from methylated DNA was increasing during liver regeneration after partial hepatectomy. Activity was increased 3-fold by 24h and was maximal (6-fold increase) over the period 48-72h after operation. 2. This activity could also be induced by chronic treatment with dimethylnitrosamine, but the maximal response amounted to a 2-3-fold change (with the greater effect in male rats) after 4-6 weeks of exposure to daily doses of 2 mg of dimethylnitrosamine/kg. 3. Neither partial hepatectomy nor treatment with dimethylnitrosamine increased the activities of two other enzymes repairing alkylated DNA, DNA (7-methylguanine-)glycosylase and DNA (3-methyladenine-)glycosylase. 4. These results therefore indicate that there is a selective induction of the O6-methylguanine removal system during hepatocyte proliferation. Since this product is known to lead to mutations and its persistence in DNA throughout cell replication has been implicated in tumour initiation, this induction may play a role in resistance to carcinogenesis by alkylating agents.

1. The induction of ornithine decarboxylase activity in mouse 3T3 fibroblasts or an SV-40 transformed 3T3 cell line by serum was prevented by addition of the naturally occurring polyamines putrescine (butane-1,4-diamine) and spermidine. Much higher concentrations of these amines were required to fully suppress ornithine decarboxylase activity in the transformed SV-3T3 cells than in the 3T3 fibroblasts. 2. Synthetic alpha omega-diamines with 3–12 carbon atoms also prevented the increase in ornithine decarboxylase activity induced by serum in these cells. The longer chain diamines were somewhat more potent than propane-1,3-diamine in this effect, but the synthetic diamines were less active than putrescine in the 3T3 cells. There was little difference between the responses of 3T3 and SV-3T3 cells to the synthetic diamines propane-1,3-diamine and heptane-1,7-diamine. 3. These results are discussed in relation to the control of polyamine synthesis in mammalian cells.

Rat liver ornithine decarboxylase activity was decreased by administration of putrescine (1,4-diaminobutane) or other diamines, including 1,3-diaminopropane, 1,5-diaminopentane and 1,6-diaminohexane. This effect was seen in control rats and in rats in which hepatic ornithine decarboxylase activity had been increased by administration of growth hormone (somatotropin) or thioacetamide. Loss of activity was not dependent on the conversion of putrescine into polyamines and was short-lived. Within 6h after intraperitoneal administration of 0.8 mmol/kg body wt., ornithine decarboxylase activity had returned to normal values. This return correlated with the rapid loss of the diamines from the liver, and the decrease in activity could be slightly prolonged by treatment with aminoguanidine, a diamine oxidase inhibitor. A decrease in ornithine decarboxylase activity by these diamines was accompanied by the accumulation in the liver of a nondiffusible inhibitor that decreased the activity of a purified ornithine decarboxylase preparation. The possibility that administration of non-physiological diamines that are not converted into polyamines might be useful for the inhibition of polyamine synthesis is discussed.

A new method for the assay of aminopropyltransferase activity is described. The method measures the formation of [methyl-14C]methylthioadenosine from decarboxylated S-adenosyl[methyl-14C]methionine in the presence of an amine acceptor. When used with extracts from rat ventral prostate, kidney, liver or brain, and with putrescine or spermidine as amines, the method gave results in excellent agreement with those obtained by the much more time-consuming conventional method. It was found that 1,3-diamino-propane and 1,8-diamino-octane were not acceptors for the prostatic enzyme fraction, but 1,5-diaminopentane (cadaverine) was active and 1,9-diaminononane and 1,12-diaminododecane also lead to the production of [methyl-14C]methylthioadenosine.

1. Polyamine concentrations were decreased in rats fed on a diet deficient in vitamin B-6. 2. Ornithine decarboxylase activity was decreased by vitamin B-6 deficiency when assayed in tissue extracts without addition of pyridoxal phosphate, but was greater than in control extracts when pyridoxal phosphate was present in saturating amounts. 3. In contrast, the activity of S-adenosylmethionine decarboxylase was not enhanced by pyridoxal phosphate addition even when dialysed extracts were prepared from tissues of young rats suckled by mothers fed on the vitamin B-6-deficient diet. 4. S-Adenosylmethionine decarboxylase activities were increased by administration of methylglyoxal bis(guanylhydrazone) (1,1′-[(methylethanediylidine)dinitrilo]diguanidine) to similar extents in both control and vitamin B-6-deficient animals. 5. The spectrum of highly purified liver S-adenosylmethionine decarboxylase did not indicate the presence of pyridoxal phosphate. After inactivation of the enzyme by reaction with NaB3H4, radioactivity was incorporated into the enzyme, but was not present as a reduced derivative of pyridoxal phosphate. 6. It is concluded that the decreased concentrations of polyamines in rats fed on a diet containing vitamin B-6 may be due to decreased activity or ornithine decarboxylase or may be caused by an unknown mechanism responding to growth retardation produced by the vitamin deficiency. In either case, measurements of S-adenosylmethionine decarboxylase and ornithine decarboxylase activity under optimum conditions in vitro do not correlate with the polyamine concentrations in vivo.

1. A comparison was made of the binding of 5α-dihydrotestosterone (17β-hydroxy-5α-androstan-3-one) and cyclic AMP in the rat prostate gland. Distinct binding mechanisms exist for these compounds, and cyclic AMP cannot serve as a competitor for the 5α-dihydrotestosterone-binding sites and vice versa. In contrast with the results obtained with 5α-dihydrotestosterone, very small amounts of cyclic AMP are retained in nuclear chromatin and the overall binding of this cyclic nucleotide is not markedly affected by castration. 2. Androgenic stimulation does not lead to major increases in the adenylate cyclase activities associated with any subcellular fraction of the prostate gland. Accordingly, changes in the concentration of cyclic AMP in the prostate gland after hormonal treatment are likely to be small, but these were not measured directly. 3. When administered to whole animals in vivo , small amounts of non-degraded cyclic AMP are found in the prostate gland but sufficient to promote an activation of certain carbohydrate-metabolizing enzymes in the cell supernatant fraction. The stimulatory effects of cyclic AMP were not evident with cytoplasmic enzymes engaged in polyamine synthesis or nuclear RNA polymerases. These latter enzymes were stimulated solely by the administration of testosterone. 4. By making use of antiandrogens, a distinction can be drawn between the biochemical responses attributable to the binding of 5α-dihydrotestosterone but not of cyclic AMP. Evidence is presented to suggest that the stimulation of RNA polymerase, ornithine decarboxylase and S -adenosyl- l -methionine decarboxylase is a consequence of the selective binding of 5α-dihydrotestosterone. Only the stimulation of glucose 6-phosphate dehydrogenase can be attributed to cyclic AMP or other metabolites of testosterone. 5. Overall, this study indicates that the formation of cyclic AMP is not a major feature of the androgenic response and affects only a restricted number of biochemical processes. Certainly, cyclic AMP cannot be considered as interchangeable with testosterone and its metabolites in the control of the function of the prostate gland. This difference is additionally emphasized by the failure of cyclic AMP to restore the morphology of the prostate gland in castrated animals; morphological restoration only follows the administration of androgens.

The effect of methylglyoxal bis(guanylhydrazone), a substance known to inhibit putrescine-dependent S -adenosyl- l -methionine decarboxylase, on polyamine metabolism in liver and kidney was investigated. Almost complete inhibition of the incorporation of putrescine into spermidine was obtained up to 8h after administration of 80mg of methylglyoxal bis(guanylhydrazone)/kg body wt. by intraperitoneal injection. However, by 20h after administration of the inhibitor spermidine synthesis was resumed. Considerable accumulation of putrescine occurred during this period (up to 3 times control concentrations in both tissues), but there was only a slight fall in the spermidine content. These results suggest that the putrescine-activated S -adenosyl- l -methionine decarboxylase plays an essential role in spermidine biosynthesis in rat liver and kidney, and the possibility of using methylglyoxal bis(guanylhydrazone) to study the role of polyamine synthesis in growth is discussed.

1. The ethylated nucleosides present in tRNA isolated from the livers of rats treated with 0.5g of l -ethionine/kg body wt. were investigated. Evidence that this tRNA contained N 2 -ethylguanine, N 2 N 2 -diethylguanine, N 2 -ethyl- N 2 -methylguanine, 7-ethylguanine, two ethylated pyrimidines and ethylated ribose groups was obtained. 2. Ethylation of bacterial tRNA was catalysed by extracts containing tRNA methylases prepared from rat liver by using S -adenosyl- l -ethionine as an ethyl donor, but the rate of ethylation was 20 times less than the rate of methylation with S -adenosyl- l -methionine as a methyl donor. 3. The principal product of such ethylation in vitro was N 2 -ethylguanine and traces of the other ethylated guanines and pyrimidines found in tRNA isolated from rats treated with ethionine in vivo were also found. 1-Ethyladenine was not formed, although 1-methyl-adenine is a major product of methylation of bacterial tRNA by these extracts, and 1-ethyladenine was not present in the rat liver tRNA isolated from ethionine-treated animals. 4. After injection of actinomycin D (15mg/kg body wt.) or l -methionine (1.0g/kg body wt.) before the ethionine, ethylation of tRNA was diminished by about 80% but not completely abolished. Administration of 1-aminocyclopentanecarboxylic acid (2.5g/kg body wt.) to inhibit the formation of S -adenosyl- l -ethionine inhibited ethylation of tRNA by 44%. 5. These results suggest that not all of the ethylation of tRNA that occurs in the livers of rats treated with ethionine is mediated by the action of tRNA methylases acting with S -adenosyl- l -ethionine as a substrate, but that this pathway does occur and accounts for a major part of the observed ethylation. 6. The results are discussed with reference to ethionine-induced hepatocarcinogenesis.

1. Administration of a large dose (500mg/kg body wt.) of 3 H-labelled l -ethionine to rats resulted in the incorporation of a small amount of radioactivity into the liver DNA. Considerable evidence that this radioactivity was not due to contamination of the isolated DNA with labelled protein, RNA, S -adenosyl- l -ethionine or l -ethionine was obtained. 2. After acidic hydrolysis of the DNA isolated from the livers of rats treated with labelled l -ethionine, virtually all of the radioactivity present in the DNA was found in a fraction with similar chromatographic properties to 7-ethylguanine. 3. Treatment of rats with comparable doses of l -methionine did not lead to the formation of 7-methylguanine in the liver DNA. 4. These results are discussed in relation to the induction of liver tumours by ethionine.

1. Castration of adult rats resulted in marked decreases in the amounts of putrescine, spermidine and spermine in the ventral prostate gland. Spermidine concentrations decline rapidly over the first 11 days after androgen withdrawal, reaching a value of only 12% of normal controls. Spermine concentrations diminish more slowly, reaching 24% of normal within 11 days. The spermidine/spermine molar ratio falls from 0.9 to 0.46 under these conditions. Putrescine concentrations decrease by 70% at 7 days after castration and then remain constant for some days. 2. After daily injections of testosterone propionate to rats castrated 7 days previously, prostatic spermidine and putrescine concentrations increase significantly within 24h; normal or even greater values are observed within 8 and 4 days respectively. In contrast, the spermine concentration does not increase until 5 days after commencement of androgen treatment. 3. The activities of two enzymes involved in polyamine biosynthesis (ornithine decarboxylase and a putrescine-activated S -adenosyl- l -methionine decarboxylase system) were greatly decreased soon after castration: after 7 days the respective values were 15% of normal for ornithine decarboxylase and 7% of normal for putrescine-dependent decarboxylation of S -adenosyl- l -methionine. Injection of testosterone propionate into animals castrated 7 days previously induced a rapid increase in both enzymic activities: ornithine decarboxylase was doubled in 6h, and increased three- to four-fold within 48h, whereas the putrescine-dependent decarboxylation of S -adenosyl- l -methionine doubled in 3h and increased tenfold within 48h of commencement of daily androgen treatments. 4. The activity of these enzyme systems was very low in the ventral prostates of hypophysectomized rats and was increased by administration of testosterone in a manner similar to that found in castrated rats. 5. Alterations in the activity of two ventral-prostate enzymes involved in ornithine production (arginase) and utilization (ornithine–2-oxoglutarate transaminase) that result from changes in the androgenic status of rats are described. 6. The findings presented suggest that the activities of ornithine decarboxylase and the putrescine-dependent S -adenosyl- l -methionine decarboxylase system, rather than ornithine concentrations, are rate-limiting for the formation of putrescine and polyamines in rat ventral prostate. 7. The relation of polyamines to androgen-induced prostatic growth is discussed with particular reference to the biosynthesis of proteins and nucleic acids.