Control of irritable bowel syndrome with polyamine analogs: a structure-activity study

The evaluation of a group of polyamine analogs as agents to ameliorate diarrhea-predominant irritable bowel syndrome is described. Each compound was assessed when administered subcutaneously in a psychological stress-induced model of irritable bowel syndrome in rodents for its ability to reduce stool output in a dose-dependent manner. The spermine pharmacophore is shown to be an excellent platform from which to construct compounds to treat irritable bowel syndrome. The activity of the compounds is very dependent on both the nature of the terminal alkyl groups and the geometry of the methylene spacers separating the nitrogens. In addition to the subcutaneous studies, several compounds, N1,N11-diethylnorspermine, N1,N12-diethylspermine, N1,N12-diisopropylspermine, N1,N14-diethylhomospermine, N,N'-bis[5-(ethylamino)pentyl]-1,4-butanediamine, N,N'-bis[2-(4-piperidinyl)ethyl]-1,4-diaminobutane, and N,N'-bis[3-(ethylamino)propyl]-trans-1,4-cyclohexanediamine, were subsequently evaluated for oral efficacy. The remarkable activity of N,N'-bis[3-(ethylamino)propyl]-trans-1,4-cyclohexanediamine underscores the need to explore this framework further as a pharmacophore for the construction of other analogues to relieve the symptoms of diarrhea-predominant IBS.

Synthesis and evaluation of hydroxylated polyamine analogues as antiproliferatives

A new means of accessing N(1)-cyclopropylmethyl-N(11)-ethylnorspermine (CPMENSPM) and the first synthesis of (2R,10S)-N(1)-cyclopropylmethyl-2,10-dihydroxy-N(11)-ethylnorspermine [(2R,10S)-(HO)(2)CPMENSPM] are described. Both of these polyamine analogues are shown to be more active against L1210 murine leukemia cell growth than either N(1),N(11)-diethylnorspermine (DENSPM) or (2R,10R)-N(1),N(11)-diethyl-2,10-dihydroxynorspermine [(2R,10R)-(HO)(2)DENSPM] after 96 h of treatment; the activity was comparable to that of (2S,10S)-N(1),N(11)-diethyl-2,10-dihydroxynorspermine [(2S,10S)-(HO)(2)DENSPM] at 96 h. Both cyclopropyl compounds reduced putrescine and spermidine pools, but less effectively than did DENSPM and its derivatives. Only CPMENSPM, and not (2R,10S)-(HO)(2)CPMENSPM, lowered spermine pools. As with DENSPM and (2R,10R)-(HO)(2)DENSPM, both cyclopropyl analogues diminished ornithine decarboxylase and S-adenosylmethionine decarboxylase activity. Unlike the hydroxylated DENSPM compounds, both cyclopropyl norspermines substantially upregulated spermidine/spermine N(1)-acetyltransferase. The most interesting effect of hydroxylating CPMENSPM is the profound reduction in toxicity compared with that of the parent drug. The same phenomenon had been observed for the DENSPM/(2R,10R)-(HO)(2)DENSPM pair. Thus, hydroxylation of norspermine analogues appears to be a way to maintain the compounds' antiproliferative activity while reducing their toxicity.

Significance of asymmetric sites in choosing siderophores as deferration agents

The syntheses of the microbial iron chelators L-fluviabactin, its unnatural enantiomer, D-fluviabactin, L-homofluviabactin, and L-agrobactin, are described. The key steps involve the selective bis-acylation of the terminal nitrogens of norspermidine, spermidine, or homospermidine with 2,3-bis(benzyloxy)benzoic acid in the presence of 1,1-carbonyldiimidazole, followed by coupling of the N-hydroxysuccinimide ester of CBZ-protected L- or D-threonine with the central nitrogen. The effectiveness of each of these ligands in supporting the growth of Paracoccus denitrificans in a low-iron environment and the ability of these compounds to promote iron uptake are evaluated. The stereochemical configuration of the oxazoline ring is shown to be the major structural factor controlling both microbial growth stimulation and iron uptake. L-Fluviabactin, L-homofluviabactin, and L-agrobactin all promoted growth and iron uptake; D-fluviabactin was only marginally active. As with the microorganism's native siderophore, L-parabactin, all three ligands in the L-configuration investigated exhibited biphasic, i.e., both high-affinity and low-affinity, kinetics. The high-affinity system (iron concentration < 1 microM) yielded K(m) values between 0.11 and 0.23 microM and V(max) values from 157 to 129 pg-atoms Fe min(-1) (mg of protein)(-1), whereas the low-affinity scheme (iron concentration > 1 microM) gave K(m) values from 0.53 to 3.5 microM and V(max) values between 96 and 413 pg-atoms Fe min(-1) (mg of protein)(-1). Both L- and D-fluviabactin are very effective at clearing iron from the bile duct-cannulated rodent; when given subcutaneously at a dose of 150 micromol/kg, both ligands had iron clearing efficiencies of >13%, which is much greater than that of desferrioxamine in this model. Thus, by altering the stereochemistry of certain microbial siderophores, it is possible to generate deferration agents that are still effective at clearing iron from animals, yet do not promote microbial growth.

Polyamine analogue antidiarrheals: a structure-activity study

The syntheses of a group of spermine polyamine analogues and their evaluation as antidiarrheals are described. Each compound was assessed in a rodent castor oil-induced diarrhea model for its ability to reduce stool output and weight loss in a dose-dependent manner. The spermine pharmacophore is shown to be an excellent platform from which to construct antidiarrheals. The activity of the compounds is very dependent on both the nature of the terminal alkyl groups and the geometry of the methylene spacers separating the nitrogens. The toxicity profile is also quite dependent on these same structural features. On the basis of subcutaneous dose-response data and toxicity profiles, two compounds, N(1),N(12)-diisopropylspermine and N(1),N(12)-diethylspermine, were taken forward into more complete evaluation. These measurements included formal acute and chronic toxicity trials, drug and metabolic tissue distribution studies, and assessment of the impact of these analogues on tissue polyamine pools. Finally, the remarkable activity of N,N'-bis[3-(ethylamino)propyl]-trans-1,4-cyclohexanediamine underscores the need to further explore this framework as a pharmacophore for the construction of other antidiarrheal agents.

Metabolism and pharmacokinetics of N1,N11-diethylnorspermine in a Cebus apella primate model

The tissue distribution, metabolic profile, and pharmacokinetic parameters of i.v.-administered N1,N11-diethylnorspermine (DENSPM) are evaluated in Cebus apella primates, and the results are compared with data gathered from canine and human studies. Although the metabolic processing of DENSPM (i.e., deethylation and deaminopropylation) in dogs and primates is very similar, there are some significant differences in tissue distribution of the parent drug. In dogs, the organ concentration of DENSPM follows the order kidney >> liver approximately = lung > spleen. In the primate, the order is liver >> kidney approximately = spleen > lung. The difference in pharmacokinetic parameters between the species is profound with (area under the time-concentration curve)primate << (area under the time-concentration curve)dog; (terminal elimination half-life)primate << (terminal elimination half-life)dog; and (mean residence time)primate << (mean residence time)dog. The most notable difference between dogs and primates is seen in the fraction of parent drug excreted unchanged in the urine, 50% in the dog and < 1% in the primate. However, the pharmacokinetic parameters and urinary drug clearance in C. apella primates are remarkably similar to those in humans. Thus, C. apella is established as an excellent model for assessing the metabolism, tissue distribution, and pharmacokinetic properties of polyamine analogues.

Synthesis and evaluation of hydroxylated polyamine analogues as antiproliferatives

The synthesis of four hydroxylated polyamine analogues, (2R, 10R)-N(1),N(11)-diethyl-2,10-dihydroxynorspermine, (2S,10S)-N(1), N(11)-diethyl-2,10-dihydroxynorspermine, (3S,12S)-N(1), N(14)-diethyl-3,12-dihydroxyhomospermine, and (3R,12R)-N(1), N(14)-diethyl-3,12-dihydroxyhomospermine, is described along with their impact on the growth and polyamine metabolism of L1210 murine leukemia cells. Four different synthetic approaches are set forth, two each for the hydroxylated norspermines and for the hydroxylated homospermines. The key step in the assembly of the norspermines was the coupling of either N-[(2R)-2,3-epoxypropyl]-N-ethyl p-toluenesulfonamide or N-[(2S)-2,3-epoxypropyl]-N-ethyl trifluoromethanesulfonamide to N,N'-dibenzyl-1,3-diaminopropane. The key step with homospermines employed alkylation of putrescine with (3S)-N-(benzyloxycarbonyl)-N-ethyl-3,4-epoxybutylamine or of N, N'-bis(mesitylenesulfonyl)-1,4-butanediamine with (2R)-2-benzyloxy-4-[N-(mesitylenesulfonyl)ethylamino]-O-tosyl-1-++ +butan ol. All of the hydroxylated analogues were active against L1210 cells with 96-h IC(50) values of </=2 microM, and they also effectively reduced putrescine and spermidine, although the effect on spermine pools ranged from moderate to insignificant. Interestingly, the impact of the hydroxylated analogues on ornithine decarboxylase (ODC) was significantly less than that of unhydroxylated parent drug (e.g., N(1),N(11)-diethylnorspermine [DENSPM]) at 1 microM; however, S-adenosylmethionine decarboxylase (AdoMetDC) depletion was nearly identical to what was observed in cells treated with parent drug. The most notable difference between the parent and hydroxylated analogues was seen with spermidine/spermine N(1)-acetyltransferase (SSAT) upregulation in the DENSPM series. The hydroxylated analogues, especially (R, R)-(HO)(2)DENSPM, were much less effective at upregulation than the parent DENSPM. Finally, a comparison of the toxicity of (R, R)-(HO)(2)DENSPM with that of DENSPM at subchronic doses revealed that the neurological effects seen with DENSPM were now absent.

Pharmacokinetics of orally administered desferrithiocin analogs in cebus apella primates

The pharmacokinetic behavior of three iron chelators based on the desferrithiocin (DFT) pharmacophore, (S)-4, 5-dihydro-2-(2-hydroxyphenyl)-4-thiazolecarboxylic acid (desmethyldesferrithiocin, DMDFT, 2); (S)-4,5-dihydro-2-(2, 4-dihydroxyphenyl)-4-thiazolecarboxylic acid [4-(S)-hydroxydesazaDMDFT, 3); and (R)-2-(2-hydroxyphenyl)-4-oxazolinecarboxylic acid, the oxazoline analog of desazaDMDFT, 4, is described. Although 2 and 3 are comparably effective in inducing iron excretion upon oral administration, they exhibit markedly different plasma pharmacokinetics. Ligand 2 achieves a substantially higher plasma concentration than does 3, yet the renal clearance of these compounds is similar. The oxazoline analog 4 shows poor iron clearance when administered orally, although it remains in the plasma for extended periods. Chelator 4 demonstrates a marked capacity to bind to human serum albumin compared with the thiazoline derivatives. The possible implications for designing ligands for the treatment of transfusional iron overload are discussed.

Evaluation of the desferrithiocin pharmacophore as a vector for hydroxamates

A series of (S)-desmethyldesferrithiocin (DMDFT, 1) hydroxamates and a bis-salicyl polyether hydroxamate are evaluated for their iron-clearing properties in rodents; some of these are further assessed in primates. These hydroxamates include (S)-desmethyldesferrithiocin, N-methylhydroxamate (2); (S)-desmethyldesferrithiocin, N-[5-(acetylhydroxyamino)pentyl]hydroxamate (3); desmethyldesferrithiocin, N-benzylhydroxamate (4); (S,S)-N(1), N(8)-bis[4,5-dihydro-2-(3-hydroxy-2-pyridinyl)-4-thiazoyl]-N(1), N(8)-dihydroxy-3,6-dioxa-1,8-octanediamine (5); and N(1), N(8)-bis(2-hydroxybenzoyl)-N(1),N(8)-dihydroxy-3,6-dioxa-1, 8-octanediamine (6). The ligands are evaluated when given both orally (po) and subcutaneously (sc) in the bile-duct-cannulated rodent model. In iron-overloaded primates, ligands 1-4 are assessed when administered po and sc. The efficiencies of the hydroxamates are shown to vary considerably; giving the compounds sc consistently resulted in greater chelating efficiency in vivo. After oral administration in the primate, compound 3, a pentacoordinate unsymmetrical dihydroxamate, produces iron excretion sufficient to warrant further preclinical evaluation both as a potential orally active iron-chelating agent and as a parenteral iron chelator. The increased iron clearance of several of these ligands when administered sc versus po also underscores the idea that parenteral administration is a reasonable alternative to a less efficient, orally active device which would require large and frequent doses.

Effects of C-4 stereochemistry and C-4' hydroxylation on the iron clearing efficiency and toxicity of desferrithiocin analogues

Additional structure-activity studies of desferrithiocin analogues are carried out. The effects of stereochemistry at C-4 on the ligands' iron clearing efficiency are reviewed and assessed using the enantiomers 4,5-dihydro-2-(2, 4-dihydroxyphenyl)thiazole-4(R)-carboxylic acid and 4,5-dihydro-2-(2, 4-dihydroxyphenyl)thiazole-4(S)-carboxylic acid. The utility of 4'-hydroxylation as a method of reducing the toxicity of desazadesferrithiocin analogues is also examined further with the synthesis and in vivo comparison of 4, 5-dihydro-2-(2-hydroxyphenyl)-4-methylthiazole-4(S)-carboxylic acid, which is the natural product 4-methylaeruginoic acid, and 4, 5-dihydro-2-(2,4-dihydroxyphenyl)-4-methylthiazole-4(S)-carboxylic acid. The stereochemistry at C-4 is shown to have a substantial effect on the iron clearing efficiency of desferrithiocin analogues, as does C-4'-hydroxylation on the toxicity profile. All of the compounds are evaluated in a bile-duct-cannulated rodent model to determine iron clearance efficiency and are carried forward to the iron-overloaded primate for iron clearing measurements. On the basis of the results of the present work, although 4,5-dihydro-2-(2, 4-dihydroxyphenyl)thiazole-4(S)-carboxylic acid is still the most promising candidate for clinical evaluation, 4,5-dihydro-2-(2, 4-dihydroxyphenyl)-4-methylthiazole-4(S)-carboxylic acid (4'-hydroxydesazadesferrithiocin) also merits further preclinical assessment.

Polyamine analogue induction of the p53-p21WAF1/CIP1-Rb pathway and G1 arrest in human melanoma cells

Although polyamines are well recognized for their critical involvement in cell growth, the cell cycle specificity of this requirement has not yet been characterized with respect to the newly delineated regulatory pathways. We recently reported that polyamine analogues having close structural and functional similarities to the natural polyamines produce a distinct G1 and G2-M cell cycle arrest in MALME-3M human melanoma cells. To determine a molecular basis for this observation, we examined the effects of N1,N11-diethylnorspermine on cell cycle regulatory proteins associated with G1 arrest. The analogue is known to deplete polyamine pools by suppressing biosynthetic enzymes and potently inducing the polyamine catabolic enzyme spermidine/spermine N1-acetyltransferase. Treatment of MALME-3M cells with 10 microM N1,N11-diethylnorspermine caused an increase in hypophosphorylated Rb, which correlated temporally with the onset of G1 arrest at 16-24 h. Rb hypophosphorylation was preceded by an increase in wild-type p53 (approximately 100-fold at maximum) and a concomitant increase in the cyclin-dependent kinase inhibitor, p21WAF1/CIP1 (p21; approximately 5-fold at maximum). Another cyclin-dependent kinase inhibitor, p27KIP1, and cyclin D increased slightly, whereas proliferating cell nuclear antigen and p130 remained unchanged. Induction of p21 protein was accompanied by an increase in p21 mRNA, whereas induction of p53 protein was not, suggesting transcriptional activation of the former and posttranscriptional regulation of the latter. SK-MEL-28 human melanoma cells, which contain a mutated p53, failed to induce p53 or p21 and did not arrest in G1. Rather, these cells rapidly underwent programmed cell death within 48 h. Overall, these findings provide the first indication of the cell cycle regulatory pathways by which polyamine antagonists such as analogues might inhibit growth in cells containing wild-type p53 and further suggest a mechanistic basis for differential cellular responses to these agents.

Desazadesmethyldesferrithiocin analogues as orally effective iron chelators

Further structure-activity studies of desferrithiocin analogues are carried out. (S)-Desazadesmethyldesferrithiocin, 2-(2-hydroxyphenyl)-Delta2-thiazoline-4(S)-carboxylic acid, serves as the principal framework in the current paper. Desazadesmethyldesferrithiocin can be structurally altered with facility, and data are already available on its iron-clearing properties and toxicity parameters. Four different kinds of structural modifications of this framework are undertaken: introduction of hydroxy, carboxy, or methoxy groups on the aromatic ring; alteration of the thiazoline ring; increasing the distance between the ligand donor atoms; and benz-fusion of the aromatic rings. The structural modifications described are shown to have a tremendous impact on both the iron clearance and toxicity profiles of the desazadesmethyldesferrithiocin molecule. All of the compounds are assessed in a bile-duct-cannulated rodent model to determine iron clearance efficiency. Ligands which demonstrate an efficiency of greater than 2% are carried forward to the iron-overloaded primate for iron-clearing measurements. Ligands with efficiencies greater than 3% in the primate are then evaluated in a formal toxicity study in rodents. On the basis of the results of the present work, 2-(2, 4-dihydroxyphenyl)-Delta2-thiazoline-4(S)-carboxylic acid is a promising candidate for clinical evaluation.

HBED: the continuing development of a potential alternative to deferoxamine for iron-chelating therapy

To further examine the potential clinical usefulness of the hexadentate phenolic aminocarboxylate iron chelator N, N'-bis(2-hydroxybenzyl)ethylenediamine-N,N'-diacetic acid (HBED) for the chronic treatment of transfusional iron overload, we performed a subchronic toxicity study of the HBED monosodium salt in rodents and have evaluated the iron excretion in primates induced by HBED. The HBED-induced iron excretion was determined for the monohydrochloride dihydrate that was first dissolved in a 0.1-mmol/L sodium phosphate buffer at pH 7.6 and administered to the primates either orally (PO) at a dose of 324 micromol/kg (149.3 mg/kg, n = 5), subcutaneously (sc) at a dose of 81 micromol/kg (37.3 mg/kg, n = 5), sc at 324 micromol/kg (n = 5), and sc at 162 micromol/kg (74.7 mg/kg) for 2 consecutive days for a total dose of 324 micromol/kg (n = 3). In addition, the monosodium salt of HBED in saline was administered to the monkeys sc at a single dose of 150 micromol/kg (64.9 mg/kg, n = 5) or at a dose of 75 micromol/kg every other day for three doses, for a total dose of 225 micromol/kg (n = 4). For comparative purposes, we have also administered deferoxamine (DFO) PO and sc in aqueous solution at a dose of 300 micromol/kg (200 mg/kg). In the iron-loaded Cebus apella monkey, whereas the PO administration of DFO or HBED even at a dose of 300 to 324 micromol/kg was ineffective, the sc injection of HBED in buffer or its monosodium salt, 75 to 324 micromol/kg, produced a net iron excretion that was nearly three times that observed after similar doses of sc DFO. In patients with transfusional iron overload, sc injections of HBED may provide a much needed alternative to the use of prolonged parenteral infusions of DFO. Note: After the publication of our previous paper (Blood, 91:1446, 1998) and the completion of the studies described here, it was discovered that the HBED obtained from Strem Chemical Co (Newburyport, MA) that was labeled and sold as a dihydrochloride dihydrate was in fact the monohydrochloride dihydrate. Therefore, the actual administered doses were 81, 162, or 324 micromol/kg; not 75, 150, or 300 micromol/kg as was previously reported. The new data have been recalculated accordingly, and the data from our earlier study, corrected where applicable, are shown in parentheses.

Polyamine analogue antiarrhythmics

A group of polyamine analogues was assessed for their ability to prevent isoproterenol-induced ventricular fibrillation and death in a desoxycorticosterone acetate (DOCA)/saline rodent model. The compounds tested included polyamine antimetabolites and putrescine mimics. A structure-activity analysis revealed that tetraamines that are dicationic at physiological pH with their terminal nitrogens incorporated into pyridine rings are the most active analogues. It is clear from this study that there was no correlation between the compounds' ability to diminish polyamine metabolism and their effects on the electrical properties of the heart. In fact, the most potent polyamine antimetabolites were among the least effective antiarrhythmics. The most active of the compounds investigated, N1, N3-bis(4-pyridyl)-1,3-diaminopropane, PYR(3,3,3), was shown to both prevent isoproterenol-induced arrhythmias in DOCA/saline-treated rodents and reverse the progression of arrhythmic events that would otherwise culminate in ventricular fibrillation and death. Electrocardiographic tracings demonstrated that PYR(3,3,3) and propranolol both protect from and reverse the progression of arrhythmic events to ventricular fibrillation. In addition, cardiac pathologies from rats treated with both drugs are similar, but are substantially different from the control (isoproterenol)-treated animals. (c) 1998 The Italian Pharmacological Society.

Effect of polyamine analogues on hypusine content in JURKAT T-cells

The availability of synthetic hypusine and deoxyhypusine has made it possible to develop analytical methods which allow for the measurement of these compounds in various tissues. The methods involve dansylation of extracts from the pellet remaining after perchloric acid precipitation of cell or tissue homogenates, followed by high-performance liquid chromatography. To demonstrate the utility of this approach, the impact of four polyamine analogues, N1,N11-diethylnorspermine (DENSPM), N1,N14-diethylhomospermine (DEHSPM), 1,6,12-triazadodecane [(4,5) triamine], and 1,7, 13-triazatridecane [(5,5) triamine], on hypusine levels in a human T-cell line (JURKAT) is evaluated. All four analogues are active in controlling cell growth and compete well with spermidine for the polyamine transport apparatus. After 144 h of exposure to JURKAT cells, DENSPM reduces putrescine to below detectable limits and spermidine to 10% of the level in control cells. The other three analogues diminish both putrescine and spermidine to below detectable limits. The effectiveness with which the compounds lower spermine levels is DENSPM > DEHSPM > (4,5) triamine > (5,5) triamine. The analogues decrease the activities of ornithine decarboxylase and S-adenosylmethionine decarboxylase in a similar fashion. Of the four polyamines, DENSPM and DEHSPM are potent at lowering intracellular hypusine levels after 144 h: 59 +/- 9% and 73 +/- 12% of control levels, respectively. The other two analogues have marginal effects.

Synthesis of reagents for the construction of hypusine and deoxyhypusine peptides and their application as peptidic antigens

Two new synthetic methods which allow access to (2S)-deoxyhypusine, natural (2S,9R)-hypusine, (2S,9S)-hypusine, and deoxyhypusine- and hypusine-containing peptides are described. The methods involve both the construction of a deoxyhypusine reagent in which the alpha-nitrogen protecting group is orthogonal to the N-7 and N-12 protecting groups and an alternate synthesis of our previous hypusine reagent, a synthesis which provides for better stereochemical control at C-9. Synthetic hypusine and deoxyhypusine can be generated from these reagents. The hypusine-containing hexapeptide (Cys-Thr-Gly-Hpu-His-Gly) is conjugated to ovalbumin (OVA), keyhole limpet hemocyanin (KLH), and a bis-maleimide; KLH conjugates are also made with the deoxyhypusine- and lysine-containing hexapeptides. Monoclonal antibodies are generated to the hypusine-containing hexapeptide-OVA conjugate in mice. These are isolated and screened against the hypusine-containing hexapeptide-KLH and hypusine-containing hexapeptide-bis-maleimide conjugates, as well as against the deoxyhypusine-containing and lysine-containing hexapeptide-KLH conjugates. These antibodies may be useful in localizing intracellular hypusine-containing peptides as well as peptides containing hypusine analogues.

Lysosomal sequestration of polyamine analogues in Chinese hamster ovary cells resistant to the S-adenosylmethionine decarboxylase inhibitor, CGP-48664

CGP-48664, an inhibitor of the polyamine biosynthetic enzyme S-adenosylmethionine decarboxylase (AdoMetDC), is presently undergoing Phase 1 clinical trials as an experimental anticancer agent. We have shown previously (D. L. Kramer et al., J. Biol. Chem., 270: 2124-2132, 1995) that Chinese hamster ovary (CHO) cells that are made resistant to the growth inhibitory effects of the drug overexpress AdoMetDC because of a stable gene amplification. Unexpectedly, these same cells (CHO/644) were found to be insensitive to the growth inhibitory effects of N1,N11-diethylnorspermine (DENSPM)-a polyamine analogue also undergoing Phase 1 clinical trials-despite accumulating approximately 5 times more analogue than parental cells. We now report that treatment of CHO/664 cells with DENSPM results in the formation of numerous large cytoplasmic vacuoles, which on the basis of electron microscopy and cytochemical staining seem to be lysosomal in origin. A series of newly established CHO cell lines made differentially resistant to 1, 3, 10, 30, and 100 microM CGP-48664 by chronic exposure were used to demonstrate that vacuole formation correlated with the accumulation of extremely high levels of DENSPM without increasing growth inhibition. These same cells were used to show that AdoMetDC gene overexpression as indicated by mRNA levels was unrelated to vacuole formation; cells resistant to 100 microM CGP-48664 displayed a 170-fold increase in AdoMetDC mRNA levels and formed vacuoles in response to DENSPM, whereas those resistant to 10 microM CGP-48664 displayed a 120-fold increase in AdoMetDC mRNA levels and failed to form vacuoles. Despite accumulating to high intracellular levels, DENSPM was much less effective than spermine at down-regulating ornithine decarboxylase and polyamine transport activities in highly resistant cells. Similarly, DENSPM was less able to induce spermidine/spermine N1-acetyltransferase activity in cells that formed vacuoles than in those that did not. Overall, natural polyamines failed to induce vacuoles and various analogues of DENSPM were used to probe the structural specificity of the effect. The data are consistent with the probability that DENSPM is sequestered to high concentrations in lysosomal vacuoles of CGP-48664-resistant cells and is, therefore, not available to interact with polyamine regulatory sites or to cytotoxically affect cell growth. In addition to implicating the lysosome as a potential new site of CGP-48664 drug action that could be involved in antitumor activity and/or host toxicities, the findings also suggest a potential mechanism of cell resistance to analogues such as DENSPM.

The origin of the differences in (R)- and (S)-desmethyldesferrithiocin. Iron-clearing properties

The iron clearance properties, toxicity, and pharmacokinetics of (R)- and (S)-desmethyldesferrithiocin (DMDFT) are described. The studies were performed in rodent and primate models. While both enantiomers were found to be effective iron chelators with minimal toxicity in the rodents, only (S)-DMDFT was able to induce the clearance of any iron in the primates. In addition, two out of nine of the monkeys given (R)-DMDFT died within 24 h of drug administration. The reason for the differences in iron clearance properties and the apparent toxicity of the (R)-enantiomer in the primates is likely related to the disparities in the pharmacokinetics of the two analogues. The pharmacokinetic data suggest enantioselectivity in renal clearance of the desferrithiocins and their iron complexes with (S)-DMDFT clearance 3.5 times greater than that of (R)-DMDFT, and FeIII [(S)-DMDFT]2 clearance 6.8 times greater than that of FeIII [R-DMDFT]2. In all primates studied FeIII [(R)-DMDFT]2 in the plasma exceeded 25 mg/L (50 microM) for several hours and remained above 10 mg/L (20 microM) at 8 h while levels of FeIII [(S)-DMDFT]2 never exceeded 50 microM and were at or below the limits of detection 8 h post-injection.

HBED: A potential alternative to deferoxamine for iron-chelating therapy

To examine the potential clinical usefulness of the hexadentate phenolic aminocarboxylate iron chelator N, N-bis(2-hydroxybenzyl)ethylenediamine-N,N-diacetic acid (HBED) for the chronic treatment of transfusional iron overload, we compared the iron excretion induced by subcutaneous (SC) injection of HBED and deferoxamine (DFO), the reference chelator, in rodents and primates. In the non-iron-overloaded, bile-duct-cannulated rat, a single SC injection of HBED, 150 micromol/kg, resulted in a net iron excretion that was more than threefold greater than that after the same dose of DFO. In the iron-loaded Cebus apella monkey, a single SC injection of HBED, 150 micromol/kg, produced a net iron excretion that was more than twice that observed after the same dose of SC DFO. In patients with transfusional iron overload, SC injections of HBED may provide a much needed alternative to the use of prolonged parenteral infusions of DFO.

Human prostatic carcinoma cell lines display altered regulation of polyamine transport in response to polyamine analogs and inhibitors

BACKGROUND

The possibility was investigated that complex homeostatic mechanisms which maintain polyamine pools in prostate-derived tumors may differ from those which are typically seen in other tissues and tumors.

METHODS

Growth sensitivity and various regulatory responses were investigated in three human prostate carcinoma cell lines (LNCaP, DU145, and PC-3) treated with the inhibitor of S-adenosylmethionine decarboxylase CGP-48664 or the polyamine analog N1,N11-diethylnorspermine (DENSPM), both of which are currently undergoing phase I clinical trial.

RESULTS

Prostate tumor cell lines were all similarly growth-inhibited by the inhibitor CGP-48664 (IC50 values, 1-5 microM at 72 hr), but varied considerably in their sensitivity to DENSPM. The rank-order for cell-line growth inhibition by the analog was DU145 > PC-3 > LNCaP, with IC50 values of 1, 30, and 1,000 microM, respectively. Both compounds depleted intracellular polyamine pools to levels which seemed sufficient to account for inhibition of cell growth. While polyamine enzyme regulatory responses to both CGP-48664 and DENSPM were typical of those seen in other cell types, regulation of polyamine transport differed distinctly. Based on Vmax determinations, LNCaP cells failed to upregulate transport in response to CGP-48664, while PC-3 and LNCaP cells failed to downregulate transport in response to DENSPM.

CONCLUSIONS

Relative to other cell lines, polyamine transport in prostate carcinoma cell lines was found to be uniquely insensitive to regulation by polyamines or analogs. Although this did not seem to correlate with growth sensitivity to polyamine analogs in vitro, it should be therapeutically exploitable in in vivo systems.

Effects of novel spermine analogues on cell cycle progression and apoptosis in MALME-3M human melanoma cells

On the basis of encouraging preclinical findings, polyamine analogues have emerged as a novel class of experimental antitumor agents. The spermine derivative N1,N11-diethylnorspermine (DE-333, also known as DENSPM) is currently undergoing Phase I clinical trials against solid tumors. A series of systematically modified DE-333 analogues differing in intra-amine carbon distances and in N-alkyl terminal substituents (i.e., methyl, ethyl, and propyl) were evaluated in MALME-3M human melanoma cells, a cell line known to be cytotoxically affected by DE-333 and especially responsive to analogue induction of the polyamine catabolic enzyme spermidine/spermine N1-acetyltransferase. Analogues accumulated to comparable intracellular concentrations and similarly affected cell growth with IC50 values in the 0.5-1.0 microM range. During prolonged incubations, diethyl and dipropyl analogues were cytotoxic, whereas two dimethyl analogues were cytostatic. Cell cycle analysis following treatment with the cytotoxic analogues revealed a prominent G1 block apparent as an accumulation of cells in G0/G1 and depletion of S-phase cells as well as a less restrictive G2 block. By contrast, cytostatic analogues incompletely arrested cells in G1, leaving a significant number of S-phase cells. Morphological and immunocytochemical analysis of detached cells revealed a far greater proportion of apoptotic cells with cytotoxic analogues than with cytostatic analogues. Although spermidine/spermine N1-acetyltransferase activity was differentially induced by the analogues, there was no obvious correlation with cell cycle effects. Overall, these data indicate a previously unrecognized combined effect of polyamine analogues on cell cycle progression and apoptosis. On the basis of structure-function relationships, these activities may be manipulated to optimize therapeutic efficacy.

The influence of molecular conformation upon the self-assembly of cyclohexane diamide diacids

BACKGROUND

Information regarding the self-association of small peptide motifs can be used in the design of peptide microstructures. Previous work in our laboratories illustrated the self-association of certain diamide diacids into microcapsules. In this report a series of cyclohexane diamide diacids are investigated. The cyclohexylene (R-C6H10-R) system (with its axial and equatorial requirements) provided an opportunity to study the influence of molecular conformation upon the self-aggregation process.

RESULTS

Condensation of the respective cis- and trans-1,2-, 1,3-, and 1,4- cyclohexane dicarboxylic acid platforms with two equivalents of a L-Phe ester followed by deprotection gave the desired diamide diacids. Basic solutions of cis-1,2-, trans-1,3-, and cis-1,4-diamide diacids generated solid microspheres when acidified to pH 2.4. Molecular modeling revealed that 1,3-diaxial interactions favor a helical turn within these diamides.

CONCLUSIONS

Access to 'complementary' molecular geometries is needed to self-associate into microscopic architectures.

Structural basis for differential induction of spermidine/spermine N1-acetyltransferase activity by novel spermine analogs

The spermine analog N1,N11-diethylnorspermine (DE-333, also known as DENSPM or BENSPM) is regarded as the most potent known inducer of the polyamine catabolic enzyme, spermidine/spermine N1-acetyltransferase (SSAT), increasing activity by more than 200- to 1000-fold in certain cell types. The relative ability of a series of eight systematically modified DE-333 analogs to affect SSAT expression was examined in Malme-3M human melanoma cells, one of several cell lines known to be especially responsive to induction of this enzyme. In particular, we examined the relative contribution of induction of enzyme mRNA and prolongation of enzyme half-life to analog-mediated increases in enzyme activity. Induction of enzyme mRNA was most influenced by intra-amine carbon distances; relative effectiveness was found to be proportional to the number of three-carbon units. Stabilization of enzyme was most determined by the terminal N-alkyl substituent size; among methyl, ethyl and propyl groups, methyl was least effective. Thus, DE-333, which most potently induces SSAT mRNA and effectively stabilizes SSAT enzyme activity, produces the greatest increase in enzyme activity. Although other contributing mechanisms may be involved, the relative abilities of the various analogs to induce enzyme activity is at least partially attributable to their combined effects on enzyme mRNA and protein half-life. These data reveal the highly sensitive structure-activity relationships that underlie and control spermine analog induction of SSAT activity. Pending further definition of the relationship between SSAT induction and antitumor growth and toxicity in vivo, these relationships may be used to optimize therapeutic efficacy.

A comparison of structure-activity relationships between spermidine and spermine analogue antineoplastics

A systematic investigation of the impact of spermidine analogues both in vitro and in vivo is described. The study characterizes the effects of these analogues on L1210 cell growth, polyamine pools, ornithine decarboxylase, S-adenosyl-L-methionine decarboxylase, spermidine/spermine N1-acetyltransferase, the maintenance of cellular charge, i.e., cationic equivalence associated with the polyamines and their analogues, and compares their ability to compete with spermidine for transport. The findings clearly demonstrate that the activity of the linear polyamine analogues is highly dependent on the length of the triamines and the size of the N(alpha),N(omega)-substituents. It appears that there is an optimum chain length for various activities and that the larger the N(alpha),N(omega)-alkyls, the less active the compound. Metabolic transformation including N-dealkylation of these compounds is also evaluated. While there is no monotonic relationship between chain length and the ability of the analogue to be metabolized, the dipropyl triamines are clearly more actively catabolized than the corresponding methyl and ethyl systems. A comparison of the triamines with the corresponding tetraamines is made throughout the text regarding both in vitro activity against L1210 cells and in vivo toxicity measurements, suggesting that several triamine analogues may offer therapeutic advantages over the corresponding tetraamines.

Polyamine analogue regulation of NMDA MK-801 binding: a structure-activity study

A series of analogues and homologues of spermine were synthesized, and their impact on MK-801 binding to the N-methyl-D-aspartate (NMDA) receptor was evaluated. These tetraamines encompass both linear and cyclic compounds. The linear molecules include norspermine, N1, N11-diethylnorspermine, N1,N12-bis(2,2,2-trifluoroethyl)spermine, homospermine, and N1,N14-diethylhomospermine. The cyclic tetraamines consist of the piperidine analogues N1,N3-bis(4-piperidinyl)-1,3-diaminopropane, N1,N4-bis(4-piperidinyl)-1,4-diaminobutane, N1,N4-bis(4-piperidinylmethyl)-1,4-diaminobutane, and N1,N4-bis[2-(4-piperidinyl)ethyl]-1,4-diaminobutane and the pyridine analogues N1,N3-bis(4-pyridyl)-1,3-diaminopropane, N1,N4-bis(4-pyridyl)-1,4-diaminobutane, N1,N4-bis(4-pyridylmethyl)-1,4-diaminobutane, and N1,N4-bis[2-(4-pyridyl)-ethyl]-1,4-diaminobutane. This structure-activity set makes it possible to establish the importance of charge, intercharge distance, and terminal nitrogen substitution on polyamine-regulated MK-801 binding in the NMDA channel. Four families of tetraamines are included in this set: norspermines, spermines, homospermines, and tetraazaoctadecanes. Calculations employing a SYBYL modeling program revealed that the distance between terminal nitrogens ranges between 12.62 and 19.61 A. The tetraamines are constructed such that within families cyclics and acyclics have similar lengths but different nitrogen pKa's and thus different protonation, or charge, states at physiological pH. The pKa values for all nitrogens of each molecule and its protonation state at physiological pH are described. The modifications at the terminal nitrogens include introduction of ethyl and beta,beta,beta-trifluoroethyl groups and incorporation into piperidinyl or pyridyl systems. The studies clearly indicate that polyamine length, charge, and terminal nitrogen substitution have a significant effect on how the tetraamine regulates MK-801 binding to the NMDA receptor. Thus a structure-activity basis set on which future design of MK-801 agonists and antagonists can be based is now available.

Metabolically programmed polyamine analogue antidiarrheals

The design, synthesis, and testing of a novel class of antidiarrheal drugs based on a tetraamine pharmacophore are reported. While N1,N14-diethylhomospermine (DEHSPM) (5 mg/kg) completely prevents diarrhea in rodents, tissue distribution studies demonstrated that the principal metabolite of DEHSPM, homospermine (HSPM), accumulates and persists in tissues for a protracted period of time. This accumulation accounts for a large part of the chronic toxicity of DEHSPM. Thus a major objective was to develop a metabolically labile analogue of DEHSPM which retained the desirable biological properties of the parent drug. Hydroxyl groups, sites vulnerable to further metabolic transformation, were introduced into the external aminobutyl segments providing N1,N14-diethyl-(3R),(12R)-dihydroxyhomospermine [(HO)2-DEHSPM]. The design concept was assisted by molecular modeling, which predicted that (HO)2DEHSPM would have a Ki for polyamine transport essentially identical with that of DEHSPM. The experimentally measured Ki and also the observed values of other biological properties of (HO)2DEHSPM were in fact identical with those of DEHSPM, including IC50 against L1210 cells, impact on the NMDA receptor, and impact on L1210 native polyamine pools. Most significantly, however, there was no accumulation of the dideethylated metabolite in tissues from mice treated chronically with (HO)2DEHSPM, and (HO)2DEHSPM was 3-fold less toxic than DEHSPM. Finally, (HO)2DEHSPM completely prevented diarrhea in the castor oil-treated rat model at a dose of 5 mg/kg, just as did DEHSPM.

Synthesis and biological evaluation of naphthyldesferrithiocin iron chelators

The synthesis and iron-clearing properties of the naphthyldesferrithiocins 2-(2'-hydroxynaphth-1'-yl)-delta2-thiazoline-(4R)-carboxylic acid, 2-(2'-hydroxynaphth-1'-yl)-delta2-thiazoline-(4S)-carboxylic acid, 2-(3'-hydroxynaphth-2'-yl)-delta2-thiazoline-(4R)-carboxylic acid, and 2-(3'-hydroxynaphth-2'-yl)-delta2-thiazoline-(4S)-carboxylic acid are described. While the bile duct-cannulated rat model clearly demonstrates that the 3'-hydroxynaphthyl-2'-yl compounds are orally active iron-clearing agents and the corresponding 2'-hydroxynaphthyl-1'-yl compounds are not, in the primate model none of the benz-fused desazadesferrithiocin analogues are active. Oral versus subcutaneous administration of these ligands strongly suggests that metabolism is a key issue in their iron-clearing properties and that these benz-fused desferrithiocins are not good candidates for orally active iron-clearing drugs.

Metabolism and pharmacokinetics of N1,N14-diethylhomospermine

The pharmacokinetics and metabolism of N1,N14-diethylhomospermine (DEHSPM) is described. Analysis of 15 min constant rate intravenous infusion data in dogs gave mean values of: plasma t1/2 = 1.04 hr; Vd = 0.514 liter/kg; CL = 0.343 liter/hr/kg; and AUC0-infinity = 43.2 mg/hr/liter. The renal t1/2 = 0.99 hr, with 36% of the drug recovered in the urine between 0-4 hr unchanged. In other experiments, the drug was administered to dogs by subcutaneous injection. Noncompartmental analysis of plasma concentration-time data showed a mean residence time (MRT) of 4.67 hr (subcutaneous) vs. 1.93 hr (intravenous). Mice and dogs received DEHSPM chronically to evaluate tissue distribution of DEHSPM and its metabolites. All tissues examined contained DEHSPM and its N-deethylated metabolites, N1-ethylhomospermine (MEHSPM) and homospermine (HSPM). On day 1 posttreatment, 35% of the total dose administered to mice was present in the liver (25%) and kidney (10%). The DEHSPM present declines rapidly (liver t1/2 = 1.6 days). The majority of the original dose was present as HSPM, which persisted in tissues for weeks (liver t1/2 = 15.4 days). These data suggest that DEHSPM and MEHSPM are metabolized by N-deethylation, but that HSPM is not susceptible to further degradation by polyamine catabolic enzymes that involves stepwise removal of aminopropyl equivalents by spermine/spermidine N1-acetyltransferase/polyamine oxidase. Thus, chronic DEHSPM dosing regimens in both dogs and mice may result in the accumulation of HSPM, which is retained by tissues for an extended period of time resulting in disruption of normal polyamine homeostasis in these tissues. These findings correlate with clinical and histopathological signs of toxicity in dogs and in mice.

The impact of polyamine analogues on the blood pressure of normotensive and hypertensive rats

The impact of the antineoplastic polyamine analogues N1N14-diethylhomospermine (DEHSPM) and N1N11-diethylnorspermine (DENSPM) on the blood pressure and heart rate of normotensive and hypertensive rats are described. DEHSPM was administered to both normotensive and spontaneously hypertensive rats (SHR), while the DENSPM analogue was given only to the normotensive animals. The intravenous administration of DEHSPM at doses of 5 or 10 mg/kg resulted in a profound and long-lasting drop in the test animals' blood pressure, with no appreciable change in their heart rate. This was true for both the normotensive and the hypertensive animals. When administered at equivalent molar dosages, DENSPM was one fifth as effective as DEHSPM at reducing blood pressure. The impact of NG-nitro-L-arginine-methyl ester (L-NAME) and L-arginine on the analogues' activity is consistent with the involvement of nitric oxide.

Metabolism and pharmacokinetics of N1,N11-diethylnorspermine

The pharmacokinetics and metabolism of N1,N11-diethylnorspermine (DENSPM) is described. When administered to dogs as an intravenous bolus, DENSPM was shown to have a plasma half-life of 72.8 +/- 11.8 min, with an early distribution phase half-life of approximately 4 min and an apparent volume of distribution of 0.216 +/- 0.032 liter/kg. The renal clearance half-life was 59.7 +/- 7.6 min, with 48.8 +/- 12.5% of the drug recovered in the urine between 0-4 hr unchanged. In three other experiments, the drug was administered to dogs by constant rate intravenous infusion over periods ranging from 10 min to 2 hr. Analysis of plasma concentration-time data and urinary excretion data yielded pharmacokinetic parameters in general agreement with the intravenous bolus experiments. DENSPM metabolites were identified in both beagle dog and mouse tissues. Tissues were sampled from a single beagle 24 hr posttreatment, and rodent samples were examined at 12, 24, 48, and 96 hr posttreatment. Both the concentration of DENSPM and the metabolic profile were shown to vary in the lung, liver, spleen, and kidney. Although all the tissues examined contained DENSPM and its metabolites, the liver and kidney had the highest level of metabolites that included N1-ethylnorspermine, N1-ethylnorspermidine, N1-ethyl-1,3-diaminopropane, and norspermidine. These data suggest that DENSPM is metabolized by N-deethylation and step-wise removal of aminopropyl equivalents by spermine/spermidine N1-acetyltransferase/polyamine oxidase, a metabolic pathway unique to the polyamines.

Preclinical antitumor efficacy of the polyamine analogue N1, N11-diethylnorspermine administered by multiple injection or continuous infusion

Certain N-alkylated analogues of the natural polyamine spermine have been found to disrupt polyamine pool homeostasis and inhibit tumor cell growth. The most effective of these analogues, N1, N11-diethylnorspermine (DENSPM), apparently depletes intracellular polyamine pools primarily by inducing the polyamine acetylating enzyme spermidine/spermine N1-acetyltransferase, which contributes to polyamine depletion via increased polyamine excretion and catabolism. In this report, the experimental therapeutic efficacy of DENSPM was further examined with the use of other human solid tumor xenografts, including A121 ovarian carcinoma, A549 lung adenocarcinoma, HT29 colon carcinoma, and SH-1 melanoma, and compared with previously obtained findings with MALME-3M and PANUT-3 human melanomas. In vitro studies indicated that the growth sensitivity of most tumor cell lines to DENSPM was similar, with characteristically flat dose-response curves and IC50s ranging between 0.1 and 1 micrometer the only exception was the HT29 colon carcinoma cell line, which had an IC50 of >100 micrometer. For in vivo studies, DENSPM was administered by i.p. injection to female nude athymic mice at 40 and/or 80 mg/kg 3 times a day (every 8 h) for 6 days or by continuous s.c. infusion with the use of Alzet pumps at 120, 240, or 360 mg/kg/day for 4 days. Treatment began after s.c. tumor xenografts had reached 100-200 mm3. The SH-1 melanoma, A549 lung adenocarcinoma, and A121 ovarian carcinoma xenografts responded well to the i.p. administration of analogue with obvious tumor regressions, long-term tumor growth suppressions, and a significant proportion (up to 40%) of apparent cures (i.e., lack of tumor regrowth). However, in similarity to in vitro findings, HT29 colon carcinoma xenografts responded poorly to DENSPM treatment. Massive induction of N1-acetyltransferase activity and extensive depletion of polyamine pools were consistent findings in most tumor types after in vivo or in vitro treatment with DENSPM. The rapidly growing human LOX melanoma xenograft, however, demonstrated poor induction of N1-acetyltransferase activity and the poorest response to DENSPM treatment. In nude athymic mice with MALME-3M melanoma xenografts, constant infusion delivery of DENSPM resulted in prolonged inhibition of tumor growth and long-term tumor regressions comparable to those produced by multiple i.p. injections. On the basis of the unique structure of DENSPM, novel target and mode of intervention, mild host toxicity, and activity against different human solid tumor xenografts, DENSPM is currently being developed as an antitumor agent in humans.

The role of charge in polyamine analogue recognition

A series of analogues and homologues of N1,N12-diethylspermine (DESPM) was synthesized, and their biological properties were evaluated. These tetraamines include a simple linear analogue of DESPM, N1,N12-bis(2,2,2-trifluoroethyl)spermine (FDESPM), the cyclic analogues of DESPM, N,N'-bis(4-piperidinylmethyl)-1,4-diaminobutane [PIP(4,4,4)] and N,N'-bis[2-(4-piperidinyl)ethyl]-1,4-diaminobutane [PIP(5,4,5)], and their aromatic counterparts, N,N'-bis-(4-pyridylmethyl)-1,4-diaminobutane [PYR(4,4,4)] and N,N'-bis[2-(4-pyridyl)ethyl]-1,4-diaminobutane [PYR(5,4,5)]. The analogues FDESPM, PIP(4,4,4), and PYR(4,4,4) have distances between their nitrogen atoms almost identical to those of DESPM. The longer analogues PIP(5,4,5) and PYR(5,4,5) are very similar in the spacing of their amino groups. However, the pKa of the nitrogens in the groups differ; thus, the extent of protonation and the charge characteristics among the members of the groups differ. A comparison of the biological properties of these compounds clearly demonstrates that the tetraamines must be charged to be "recognized" by the cell. Analogues with low nitrogen pKa's such that the nitrogens are poorly protonated at physiological pH do not compete well with spermidine for uptake and, as expected, have high 96 h IC50 values and have little effect on S-adenosylmethionine decarboxylase, ornithine decarboxylase, and spermidine/spermine N1-acetyltransferase activities and on intracellular polyamine pools.

Impact of polyamine analogues on the NMDA receptor

Several N,N'-terminal dialkylated homologs of the tetraamine spermine exhibit a pronounced biphasic activity at the N-methyl-D-aspartate (NMDA) receptor-channel complex in rat cerebral cortex membranes in the presence of 100 microM L-glutamate and 100 microM glycine. At low micromolar polyamine concentrations, these analogs enhance binding of [3H]-(+)-5-methyl-10,11-dihydro-5H-dibenzo[a,d]cyclohepten-5,10-im ine ([3H]MK-801) similar to spermine (SPM). At higher concentrations (e.g., > or = 10 microM), the analogs are antagonists of [3H]MK-801 binding. The most potent analog, N1,N14-bis(1-adamantyl)homospermine, is almost totally devoid of agonist activity and is a potent antagonist at concentrations > or = 5 microM. Three structural features of the tetraamines studied appear to correlate with potency of inhibition: (1) N-terminally alkylated polyamines > terminal primary amines (e.g., SPM); (2) length of the polyamine backbone, e.g., DMHSPM > DMNSPM; and (3) size of the terminal alkyl groups, i.e., adamantyl > tert-butyl > ethyl > methyl. These findings emphasize the potential of the tetraamine backbone as a pharmacophore to modulate NMDA receptor-channel function.

Design and testing of novel cytotoxic polyamine analogues

We designed three polyamine analogues, 1,14-diamino-N5-methyl-5,10- diazatetradecane (5me-4-4-4), 1,14-diamino-N5,N5-dimethyl-5,10-diazatetradecane (Q-Amm-4-4-4), and 1,14-bis-(ethylamino)-N5,N5-dimethyl-5,10-diazatetradecane (BE-Q-Amm-4-4-4), on the basis of computer modeling and physical-chemical studies of polyamine-DNA interactions. These analogues differ from natural polyamines and from one another in the charge distribution on their aliphatic backbone. We found that 10 microM 5me-4-4-4 did not inhibit growth and was not cytotoxic to the human brain tumor cell lines SF-767 and SF-126. The same concentrations of Q-Amm-4-4-4 and BE-Q-Amm-4-4-4 inhibited cell growth and killed more than 90% of each cell type on day 7 of the experiment. BE-Q-Amm-4-4-4 was slightly more toxic than Q-Amm-4-4-4 in both cell lines. All three agents either decreased or completely depleted intracellular putrescine and spermidine. Q-Amm-4-4-4 and BE-Q-Amm-4-4-4 each also lowered spermine. The fact that 5me-4-4-4 was nontoxic but that Q-Amm-4-4-4 was cytotoxic and inhibited growth suggests that the charge distribution along the surface of the aliphatic backbone of polyamines is important in determining growth inhibition and cytotoxicity.

Collateral sensitivity of human melanoma multidrug-resistant variants to the polyamine analogue, N1,N11-diethylnorspermine

Certain N-alkylated analogues of the natural polyamine spermine, such as N1,N11-diethylnorspermine (DENSPM), rapidly deplete intracellular polyamine pools by down-regulating the biosynthetic enzymes, ornithine decarboxylase and S-adenosylmethionine decarboxylase, and by potently up-regulating the polyamine catabolizing enzyme, spermidine/spermine N1-acetyltransferase. On the basis of previously reported antitumor activity in human tumor xenograft model systems, DENSPM is currently undergoing Phase I clinical trials against human melanoma and other solid tumors. The antiproliferative activity of this analogue against the multidrug resistance (MDR) phenotype was examined in three MDR sublines of human melanoma RPMI-7932 cells, which were shown to be 2-to 10-fold resistant to classical MDR agents. These MDR lines had been separately derived using different selecting agents (Lemontt et al., Cancer Res., 48: 6344-6353, 1988). Subline functional resistance due to P-glycoprotein was confirmed by decreased retention of rhodamine 123 relative to parent cells as detected by flow cytometry. Although the three sublines were 2- to 10-fold less sensitive than the parent line to classical MDR-type agents, they were found in dose-response studies to be significantly more sensitive to DENSPM than the parent line. In addition, they showed a distinct cytotoxic response after a 48-h treatment with 10 microM DENSPM, which was not apparent in the parent line. Growth sensitivity of the sublines to the ornithine decarboxylase inhibitor, alpha-difluoromethylornithine, or the S-adenosylmethionine decarboxylase inhibitor, CGP-48664, was found to be similar to parent cells. The ratio of the key biosynthetic enzyme activities for ornithine decarboxylase and S-adenosylmethionine decarboxylase was found to be 3.5- to 5-fold higher in all three sublines, due mainly to increases in the former enzyme. This imbalance produced unusually high putrescine pools. Although DENSPM down-regulation of decarboxylase activities and potent up-regulation of spermidine/spermine N1-acetyltransferase activity occurred similarly in both parent and variant lines, polyamine depletion was greater in the variant lines. Collateral sensitivity of the MDR sublines to DENSPM is partially attributable to the finding that analogue (and spermidine) uptake in the sublines was about 2-fold higher (after 2 h) than in the parent cells. The presence of disturbances in polyamine homeostasis and increased sensitivity to DENSPM in three independently selected cell line variants suggests that they may be generally associated with the MDR phenotype in human melanoma and possibly other tumor cells. The collateral sensitivity of human melanoma MDR variants to DENSPM represents a possible therapeutic indication which should be considered during the ongoing clinical evaluation of this drug.

Antiproliferative properties of polyamine analogues: a structure-activity study

A basis set of polyamine analogues was designed and synthesized. These compounds were used to initiate a systematic investigation of the role of chain length, terminal nitrogen alkyl group size, and symmetry of the methylene backbone in the antineoplastic properties of polyamine analogues. New synthetic methods predicated on our earlier polyamine fragment synthesis are described for accessing the tetraamines of interest. An unsymmetrically substituted diamine reagent, N-(tert-butoxycarbonyl)-N,N'-bis(mesitylenesulfonyl)-1,4-diaminobu tane, was developed for entry into unsymmetrical tetraamines. All of the tetraamines synthesized were first evaluated in a murine leukemia L1210 cell IC50 assay at 48 and 96 h. In an attempt to correlate this behavior with some aspect of polyamine metabolism, each compound was tested for its ability to compete with spermidine for the polyamine uptake apparatus, its impact on the polyamine biosynthetic enzymes ornithine decarboxylase (ODC) and S-adenosylmethionine decarboxylase (AdoMetDC), and its effect on the polyamine-catabolizing enzyme spermidine/spermine N1-acetyltransferase (SSAT) and on polyamine pools. While there was no obvious correlation between the 48 and 96 h IC50's and the impact of the analogues on polyamine metabolism, there were other structure-activity relationships. Correlations were observed to exist between chain length and IC50's and between terminal alkyl substituents and impact on Ki, ODC, and AdoMetDC. Also, preliminary studies suggest a relationship may exist between the 48 and 96 h IC50 activities and the analogue's chronic toxicity in vivo. Finally, when the overall length of the polyamine backbone was held constant, the symmetry of the methylene chains of the polyamine fragments was shown to be unimportant to the compound's activity.

An investigation of desferrithiocin metabolism

The hydrolyses of (S)-desferrithiocin (DFT, 1), (R)-desmethyl-DFT (2), and (R)-desazadesmethyl-DFT (3) were studied at pH 2.5 and 7.2 in order to access the stability of the thiazolines at the pH of the stomach and the serum. At 37 degrees C and pH 2.5, DFT (1) (t1/2 = 18.6 h), desmethyl-DFT (2) (t1/2 = 8.74 h), and desazadesmethyl-DFT (3) (t1/2 = 31.7 h) were shown to open principally to the thiol amides with trace amounts of the corresponding thioesters, < or = 2%. The thiazolines were resistant to hydrolysis at pH 7.2. Iron(III) stabilized significantly the thiazolines in the complexes 16a/b of 3 in regard to hydrolysis at pH 2.5 (t1/2 > 20 days). The iron(III) complexes 16a/b were shown to be stable at pH 7.2. While the thiol amides 13 and 14 of 1 and 2 were isolated from the hydrolysis of the parent desferrithiocins, the thioester 4 and the thiol amide 5 of 3 were synthesized and their stability in aqueous solution, iron-clearance properties, and toxicity were evaluated. Thioester 4 was shown to rearrange to thiol amide 5 at pH 2.5 and 37 degrees C with a half-life of 4.18 h and instantaneously at pH 7.2. Thiol amide 5 is in equilibrium with 4 (5/4 = 49:1) at pH 2.5 and was shown to be stable at pH 7.2. Thioester 4 and thiol amide 5 demonstrated neither iron-clearance activity in iron-overloaded rats nor toxic side effects in mice. Hydrolysis products of the drug, which might be generated in the stomach, seem unlikely to be the source of the drug's toxicity or iron-clearing properties.

Radiopotentiation of human brain tumor cells by the spermine analog N1,N14-bis(ethyl)homospermine

PURPOSE

To determine whether the cytotoxicity produced by radiation can be increased by the spermine analog N1,N14-bis(ethyl)homospermine (BE-4-4-4).

METHODS AND MATERIALS

Two human tumor cell lines, SF-126 and U-251 MG, were either treated with 0.1 or 0.4 microM BE-4-4-4 for 3 or 4 days, or with 0.2 microM BE-4-4-4 for 4 days. At the end of BE-4-4-4 treatment, cells were irradiated and assayed immediately. Polyamine levels, cell survival, and cell number were determined.

RESULTS

In SF-126 cells, treatment with 0.2 microM BE-4-4-4 for 4 days killed about 50% of the cells and also increased the cytotoxicity of radiation. The dose enhancement ratio was approximately 1.3:1.5, which is similar to that reported for alpha-difluoromethylornithine. Polyamine levels were partially depleted, and growth was inhibited to about 60% of control levels. Pretreatment of cells with either 0.1 or 0.4 microM BE-4-4-4 for 3 or 4 days produced less of an increase in radiation-induced cytotoxicity, even though these exposures killed 30-40% or 60-90% of the cells, respectively. Similar treatment with 0.1-0.4 microM BE-4-4-4 in U-251 MG cells had minimal effects on cytotoxicity and growth inhibition, while treatment with 1.0 microM and 2.0 microM BE-4-4-4 for 4 days produced more than a 50% depletion in polyamine levels and partial inhibition in growth, but failed to demonstrate radiopotentiation.

CONCLUSION

The cytotoxic polyamine analog BE-4-4-4 can increase the cytotoxicity caused by radiation in at least one cell line. The amount of potentiation depends on the concentration of the analog, with the most occurring at the intermediate concentration. Because we did not observe potentiation in both cell lines, and because of the dose dependence seen in SF-126 cells, the clinical efficacy produced by combined BE-4-4-4 and radiation protocols may be limited.

The desferrithiocin pharmacophore

The (S)-desferrithiocin (DFT) skeleton is shown to be a useful pharmacophore on which to design orally effective iron chelators. While the study clearly indicates that formal reduction of the desazadesmethyldesferrithiocin thiazoline to a thiazolidine (6), expansion of the desmethyldesferrithiocin thiazoline to a thiazine (7), or substitution of the thiazoline sulfur of of desazedes-methyldesferrithiocin by an oxygen (8 and 9) lead to a substantial loss of activity, conversion of (S)-desmethyldesferrithiocin (1) to an N-methylhydroxamate (4) or to the hexacoordinate dihydroxamate ligand (5) results in active compounds. This investigation thus demonstrates which structural components of the siderophore are required for iron clearance after oral administration and suggests the use of the desferrithiocin platform as a vector for other chelators.

Role of ribonucleotide reductase in inhibition of mammalian cell growth by potent iron chelators

Ribonucleotide reductase consists of two nonidentical subunits, proteins R1 and R2, the latter of which contains an iron-tyrosyl free radical center essential for activity. We have studied the in vivo effects on the R2 protein of the potent iron chelators parabactin and desferrioxamine using R2-overproducing mouse cells with a tyrosyl free radical signal easily quantifiable by EPR spectroscopy. Both chelators inhibited cell growth, and the inhibition was reversible by iron. Furthermore, both chelators, which penetrate cells and chelate the intracellular iron pool, caused a disappearance of the R2 tyrosyl free radical. In parallel, there was an accumulation of apo-R2 protein in the inhibited cells. In vitro studies using pure, 59Fe-labeled recombinant mouse R2 protein unexpectedly showed that its iron center is labile at physiological temperatures and that iron is spontaneously lost from the protein even in the absence of chelators in a temperature-dependent process. Our conclusion is that parabactin or desferrioxamine inhibits ribonucleotide reduction and cell growth not by directly attacking the iron-radical center of the R2 protein, but instead by chelating the intracellular iron pool. This prevents the regeneration of the iron-radical center both in newly synthesized apo-R2 protein and in apo-R2 protein continuously formed from active R2 protein by the loss of iron.

Effects of diethyl spermine analogues in human bladder cancer cell lines in culture

The biochemical and antiproliferative effects of two recently developed N-alkylated analogues of naturally-occurring polyamines, N1,N14-diethylhomospermine (DEHSPM) and N1,N11-diethylnorspermine (DENSPM), were investigated in two human transitional cell carcinoma (TCC) lines, T24 and J82. Parallel studies with the ornithine decarboxylase enzyme inhibitor alpha-difluoromethylornithine (DFMO) were included for comparison. DENSPM displayed greater antiproliferative activity than DEHSPM in both TCC cell lines. Both analogues were strikingly more potent than DFMO. DEHSPM and DENSPM suppressed the activity of the major biosynthetic enzymes, ornithine decarboxylase and S-adenosylmethionine decarboxylase. However, differences in the resulting polyamine depletion suggest that the substantial antiproliferative activity of these analogues may result from mechanisms other than polyamine depletion. The greater polyamine depletion seen with DENSPM is thought to result from its striking induction of spermidine/spermine N1-acetyltransferase. DENSPM is an attractive agent for further preclinical and clinical development, possibly as a chemopreventive agent, in TCC of the bladder.

Polyamine and polyamine analog regulation of spermidine/spermine N1-acetyltransferase in MALME-3M human melanoma cells

In MALME-3M human melanoma cells the polyamine analog N1,N12-bis(ethyl)spermine (BESPM) suppresses the key polyamine biosynthetic enzymes, ornithine and S-adenosylmethionine decarboxylase, and increases the polyamine catabolizing enzyme, spermidine/spermine N1-acetyl-transferase (SSAT) by more than 200-fold. In the present study increases in SSAT activity in MALME-3M cells treated with 10 microM BESPM were found to be accompanied by a substantial (up to 45-fold) accumulation of SSAT mRNA. By Northern blot analysis three RNA transcripts were found to hybridize with the coding region of human SSAT cDNA: a minor high molecular weight (approximately 3.5 kilobases) species designated form A and two lower molecular weight species designated forms B and C (approximately 1.5 and approximately 1.3 kilobases, respectively). Form A increased uniformly during BESPM treatment and was most obvious in nuclear RNA preparations. On the basis of size similarity to the transcribing region of the gene and hybridization with the coding region of SSAT cDNA and its prevalence in nuclear mRNA preparations, form A is thought to represent precursor SSAT RNA. Form C is present in control cells and increases steadily during treatment, whereas form B increases transiently during early treatment (1-3 h). By RNase H digestion assay, form B was found to have a 200-base pair longer poly(A) tract and as such may represent a precursor to form C. Accumulation of SSAT mRNA was found to be a result of increased gene transcription and stabilization of SSAT mRNA. Nuclear run-on studies indicated a 2-4-fold increase in the transcription rate of the SSAT gene. As indicated by actinomycin D studies, the SSAT mRNA half-life increased with BESPM treatment from 17 to 64 h. The natural polyamine, spermine, also increased SSAT mRNA (5.5-fold at 24 h) and behaved similarly to BESPM in inducing the appearance of the same three transcript forms. The polyamine was much less effective than the analog at increasing enzyme activity. Lowering intracellular polyamine pools with inhibitors of biosynthesis decreased basal SSAT mRNA levels by at least 70% indicating, that the gene can be down-regulated as well as up-regulated by polyamines. These findings indicate that SSAT represents a unique example of gene expression being positively influenced at the RNA level by polyamines and their analogs.

Growth inhibition of hormone-responsive and -resistant human breast cancer cells in culture by N1, N12-bis(ethyl)spermine

Previous studies have documented differential sensitivity of human lung cancer and melanoma cell lines to the cytotoxic effects of N1, N12-bis(ethyl)spermine (BESpm). We show here that BESpm can significantly inhibit the growth of six human breast cancer cell lines with 50% inhibitory concentration in the microM range. The degree of inhibition does not correlate with estrogen receptor status. Detailed studies with estrogen receptor-positive MCF-7 and estrogen receptor- negative Hs578t cells show a similar dose-response curve with concentrations of 1-10 microM resulting in maximal growth inhibition. Growth inhibition in both lines is associated with an 8-12-fold induction of the polyamine catabolic enzyme, spermidine/spermine N1-acetyltransferase, and a progressive decrease in intracellular polyamine levels over 6 days even though steady-state levels of BESpm are achieved within 24 h. Similar studies on WTMCF7 and AdrRMCF7 cells show that the acquisition of resistance to hormonal or doxorubicin therapy is not associated with resistance to the growth-inhibitory effects of BESpm. These results suggest that BESpm exerts similar growth-inhibitory effects against both hormone-responsive and -unresponsive human breast cancer cells, a finding which has significance for the potential use of polyamine analogues in treating human breast cancer.

Regulation of polyamine transport by polyamines and polyamine analogs

Regulation of polyamine transport in murine L1210 leukemia cells was characterized in order to better understand its relationship to specific intracellular polyamines and their analogs and to quantitate the sensitivity by which it is controlled. Up-regulation of polyamine uptake was evaluated following a 48-hr treatment with a combination of biosynthetic enzyme inhibitors to deplete intracellular polyamine pools. The latter declined gradually over 48 hr and was accompanied by a steady increase in spermidine (SPD) and spermine (SPM) transport as indicated by rises in Vmax to levels approximately 4.5 times higher than control values. Restoration of individual polyamine pools during a 6-hr period following inhibitor treatment revealed that SPD and SPM uptake could not be selectively affected by specific pool changes. The effectiveness of individual polyamines in reversing inhibitor-induced stimulation of uptake was as follows: putrescine < SPD < SPM = the SPM analog, N1, N12-bis(ethyl)spermine (BESPM). In contrast to stimulation of transport, down-regulation by exogenous polyamines or analogs occurred rapidly and in response to subtle increases in intracellular pools. Following a 1-hr exposure to 10 microM BESPM, Vmax values for SPD and SPM fell by 70%, whereas the analog pool increased to only 400-500 pmol/10(6) cells--about 15-20% of the total polyamine pool (approximately 2.8 nmol/10(6) cells). SPM produced nearly identical regulatory effects on transport kinetics. Both BESPM and SPM were even more effective at down-regulating transport that had been previously stimulated four to fivefold by polyamine depletion achieved with enzyme inhibitors. A dose response with BESPM at 48 hr revealed a biphasic effect on uptake whereby concentrations of analog < 3 microM produced an increase in SPD and SPM Vmax values, whereas concentrations 3 microM and higher produced a marked suppression of these values. Cells treated with 3 microM BESPM for 2 hr and placed in analog-free medium recovered transport capability in only 3 hr. Thus, whereas stimulation of polyamine transport is a relatively insensitive and slowly responsive process that tends to parallel polyamine depletion, down-regulation of polyamine transport by exogenous polyamines and analogs and its reversal are rapidly responsive events that correlate with relatively small (i.e., 15-20%) changes in intracellular polyamine pools.

A comparative study of the iron-clearing properties of desferrithiocin analogues with desferrioxamine B in a Cebus monkey model

A comparative study of the iron-clearing properties of subcutaneously administered desferrioxamine B (DFO) with those of orally administered desferrithiocin sodium salt (1), desmethyl desferrithiocin (2), desazadesmethyl desferrithiocin sodium salt (3), desazadesmethyl desferrithiocin pivaloyloxymethyl ester (4), and desazadesmethyl-5,5-dimethyl desferrithiocin (5) in an iron-loaded Cebus monkey model and a non-iron overloaded bile duct-cannulated rat model is presented. All six drugs, which performed well in rodent studies, demonstrated increased efficiency in the Cebus monkey model. When administered to rodents at a daily dosage of 384 mumol/kg over a period of 10 days, drug 1 demonstrated severe renal toxicity. whereas drugs 3, 4, and 5 exhibited severe gastrointestinal (GI) toxicity. Under the same experimental protocol, drug 2 did not show significant toxic side effects. In addition, to further evaluate the iron-clearing properties of analogue 2, a dose-response study was performed in the primates that showed that iron excretion increased in a dose-dependent fashion.