Polyamine biosynthesis impacts cellular folate requirements necessary to maintain S-adenosylmethionine and nucleotide pools.

Folate (vitamin B9) is utilized for synthesis of both S-adenosylmethionine (AdoMet) and deoxythymidine monophosphate (dTMP), which are required for methylation reactions and DNA synthesis, respectively. Folate depletion leads to an imbalance in both AdoMet and nucleotide pools, causing epigenetic and genetic damage capable of initiating tumorigenesis. Polyamine biosynthesis also utilizes AdoMet, but polyamine pools are not reduced under a regimen of folate depletion. We hypothesized that high polyamine biosynthesis, due to the high demand on AdoMet pools, might be a factor in determining sensitivity to folate depletion. We found a significant correlation (P<0.001) between polyamine biosynthesis and the amount of folate required to sustain cell line proliferation. We manipulated polyamine biosynthesis by genetic and pharmacological intervention and mechanistically demonstrated that we could thereby alter AdoMet pools and increase or decrease demand on folate availability needed to sustain cellular proliferation. Furthermore, growing a panel of cell lines with 100 nM folate led to imbalanced nucleotide and AdoMet pools only in cells with endogenously high polyamine biosynthesis. These data demonstrate that polyamine biosynthesis is a critical factor in determining sensitivity to folate depletion and may be particularly important in the prostate, where biosynthesis of polyamines is characteristically high due to its secretory function.

Spermidine requirement for cell proliferation in eukaryotic cells: structural specificity and quantitation.

Six structural homologs of spermidine and five of its precursor, putrescine, were studied for their ability to prevent cytostasis of cultured L1210 leukemia cells induced by alpha-difluoromethylornithine (DFMO), a specific inhibitor of putrescine biosynthesis. High-performance liquid chromatography and competition studies with spermidine indicated that the homologs, which vary in the length of the carbon chain separating the amines, penetrated the cells. The structural specificity of the spermidine carrier was defined. Three of the six spermidine homologs supported cell growth during a 48-hour incubation in the presence of DFMO, indicating that a two-carbon extension of spermidine structure was tolerated for biological function. Two of the five putrescine homologs supported growth after being converted by the cells to their respective spermidine homologs. The central nitrogen of spermidine appears to be essential for function since diamines of chain length comparable to that of spermidine did not prevent DFMO cytostasis. No more than 15 percent of the spermidine normally present in L1210 cells was required for cell proliferation in the presence of DFMO.

Generation of a mouse model for arginase II deficiency by targeted disruption of the arginase II gene.

Mammals express two isoforms of arginase, designated types I and II. Arginase I is a component of the urea cycle, and inherited defects in arginase I have deleterious consequences in humans. In contrast, the physiologic role of arginase II has not been defined, and no deficiencies in arginase II have been identified in humans. Mice with a disruption in the arginase II gene were created to investigate the role of this enzyme. Homozygous arginase II-deficient mice were viable and apparently indistinguishable from wild-type mice, except for an elevated plasma arginine level which indicates that arginase II plays an important role in arginine homeostasis.

Antitumor activity of N,N'-bis(ethyl)spermine homologues against human MALME-3 melanoma xenografts.

The spermine analogues, N1,N12-bis(ethyl)spermine (BESPM), N1,N11-bis(ethyl)norspermine (BENSPM), and N1,N14-bis(ethyl)-homospermine (BEHSPM) behave similarly in down-regulating the key polyamine biosynthetic enzymes, ornithine and S-adenosylmethionine decarboxylase, but differ distinctly in their abilities to induce the polyamine catabolic enzyme, spermidine/spermine-N1-acetyltransferase; BENSPM is 6-fold more effective than BESPM in increasing spermidine/spermine-N1-acetyltransferase activity and BEHSPM is 10-fold less effective. Since MALME-3 human melanoma cells are extremely responsive to spermidine/spermine-N1-acetyltransferase induction (i.e., increases greater than 200-fold) and since this induction correlates with growth inhibition among melanoma cell lines, the ability of these homologues to inhibit the growth of MALME-3 xenografts was examined. Analogues were administered i.p. three times per day (i.e., every 8 h) for 6 days at the following doses per injection: BEHSPM, 1.5, 3, or 6 mg/kg; BESPM, 10, 20, or 40 mg/kg; BENSPM, 20, 40, or 80 mg/kg. At the highest tolerated doses, all of the analogues fully suppressed growth of established (100-200 mm3) MALME-3 tumor during treatment and sustained tumor growth inhibition following treatment as follows: BEHSPM, 14 days; BESPM, 27 days, and BENSPM, 37 days. The tumor delay (to reach 1000 mm3 relative to control) at the highest tolerated doses was as follows: BEHSPM, 20 days; BESPM, 34 days, and BENSPM, 63 days. The rank order of analogue host toxicity as indicated by weight loss was opposite that for antitumor activity, BEHSPM was most toxic, BESPM, intermediate, and BENSPM, least toxic. Thus, the most effective of the three homologues, BENSPM, was best tolerated, and produced an initial tumor regression, full suppression of tumor regrowth during treatment, and sustained inhibition of tumor regrowth for 37 days after treatment stopped. Owing to its potent antitumor activity, mild host toxicity, and novel apparent mechanism of action, BENSPM is being considered for further development toward clinical trial.

Isolation and uptake characteristics of human cell variants resistant to the antiproliferative effects of methylglyoxal bis(guanylhydrazone).

Four variants (VA2/MGBG) of the human cell line, VA2, have been isolated which are 10- to 20-fold more resistant than the parent line to the antiproliferative effects of the anticancer agent, methylglyoxal bis(guanylhydrazone) (MGBG). Drug resistance was not cytoplasmically transmitted by cytoplast cell fusion for any of the four sublines, suggesting that the genes responsible for resistance may be of nuclear rather than mitochondrial origin. Uptake properties were characterized in the VA2 cells and two of the four variant lines. Uptake of [14C]MGBG during long-term (0.5 to 28 hr) incubations was 3 to 4 times greater in the VA2 cells than in the VA2/MGBG sublines. However, during short-term (2 to 60 min) incubations, the uptake of [14C]MGBG or [3H]spermidine (which competes for MGBG uptake) was similar for all cell lines. This was further supported by kinetic data which indicated that, for [14C]MGBG uptake at 4 min, the apparent Km for all cell lines was 5.8 to 13.4 microM, and the Vmax, 44 to 53 pmol/mg/min. For [3H]spermidine uptake, the apparent Km values were approximately 1 microM and Vmax, 54 to 69 pmol/mg/min. Efflux studies performed on cells incubated for 30 min in 10 microM [14C]MGBG revealed that the two VA2/MGBG sublines released drug much more rapidly than did VA2 cells over an 8-hr period. Thus, while the variants may transport MGBG at a rate similar to VA2 cells, the drug is not accumulated to the same extent during long-term incubations, probably because of altered intracellular binding sites for MGBG. The identification of these sites may provide insight into the basis for antiproliferative action of MGBG.

Ultrastructural changes in the mitochondria of intestinal epithelium of rodents treated with methylglyoxal-bis(guanylhydrazone).

Ultrastructural studies of rats or mice treated for 24 hr with a toxic dose (100 mg/kg) of methylglyoxal-bis(guanylhydrazone) revealed the presence of damaged mitochondria in the crypt cells of the intestinal epithelium. Mitochondria were severely swollen and electron lucent, and appeared to be similar to those observed previously in a variety of cell types treated in vitro and in vivo with methylglyoxal-bis(guanylhydrazone). Since thymidine incorporation into the intestine was not found to be decreased until after 24 hr, it is concluded that the mitochondrial damage of methylglyoxal-bis(guanylhydrazone) could be responsible for the antiproliferative toxicities of the drug.

A selective effect of methylglyoxal-bis(guanylhydrazone) on the synthesis of mitochondrial DNA of cultured L1210 leukemia cells.

Methylglyoxal-bis(guanylhydrazone) (MGBG) is a polycationic drug which is useful in the chemotherapy of lymphoid and myeloid proliferative disorders. The drug has recently been shown to produce selective ultrastructural damage to the mitochondria of proliferating cell populations. It is important to understand the molecular basis for this action, since it may be related to the known ability of MGBG to block polyamine biosynthesis. Accordingly, the effect of MGBG treatment on the incorporation of [3H]thymidine into both mitochondrial and nuclear DNA has been examined. Exponentially growing L1210 leukemia cells were prelabeled with [14C]thymidine, treated with MGBG for 1.5 to 16 hr, and then pulse labeled with [3H]-thymidine. Incorporation of [3H]thymidine into mitochondrial DNA was selectively inhibited at 5 hr with concentrations of 1 to 10 microM MGBG. Nuclear DNA, however, was not similarly affected until 8 to 11 hr of drug treatment. Dye-CsCl gradients of mitochondrial DNA indicated that the inhibition of synthesis occurred in replicative forms of circular DNA. Uptake studies excluded the possibility of drug interference with cellular uptake of thymidine. Ultrastructural studies revealed a very close correlation between the dose-response curve for mitochondrial damage and that for MGBG inhibition of mitochondrial DNA synthesis. This correlation suggests a direct cause-and-effect relationship between inhibition of mitochondrial DNA synthesis and ultrastructural damage, but the possibility of both phenomena being related to another action by the drug, such as inhibition of polyamine biosynthesis, or a drug effect on mitochondrial function, must also be considered.

Aliphatic chain length specificity of the polyamine transport system in ascites L1210 leukemia cells.

A series of diamine homologues of putrescine and triamine homologues of spermidine was used to determine the structural specificity of the polyamine transport system in ascites L1210 leukemia cells by measuring their ability to compete with [3H]-putrescine, [3H]spermidine, or [3H]spermine for uptake. Transport specificity among the diamines (as indicated by K1 constants) was greatest for those having chain lengths similar to that of spermidine and least for those similar to putrescine. Among the triamines, transport specificity was greatest for those having an overall chain length similar to those of spermidine and spermine. The homologue competition profiles were relatively the same for [3H]putrescine, [3H]spermidine, or [3H]spermine, suggesting that all three polyamines utilize the same transport system. This was further substantiated by uptake kinetic plots which showed that the three polyamines were competitive inhibitors of one another. In terms of receptor specificity, the ranking order among the polyamines was as follows: spermine (apparent Km, 1.6 microM) greater than spermidine (apparent Km, 2.2 microM) greater than putrescine (apparent Km, 8.5 microM). This information should prove useful in designing anticancer agents which are intended to utilize this transport system.

Comparison of specificity of inhibition of polyamine synthesis in bovine lymphocytes by ethylglyoxal bis(guanylhydrazone) and methylglyoxal bis(guanylhydrazone).

Ethylglyoxal bis(guanylhydrazone) (EGBG) was compared as an inhibitor of polyamine biosynthesis with methylglyoxal bis(guanylhydrazone) (MGBG) in bovine small lymphocytes stimulated by concanavalin A. EGBG brought about a decrease in spermidine and spermine levels equal to that found with MGBG, but at a 5-fold lower intracellular drug concentration. Despite identical polyamine levels, the degree of inhibition of DNA and protein synthesis by EGBG was smaller than that observed with MGBG, in either the presence or absence of the ornithine decarboxylase inhibitor, alpha-difluoromethylornithine. It was found that in vitro protein synthesis and in vivo mitochondrial function were inhibited by concentrations of MGBG necessary to inhibit polyamine synthesis in cells (1 to 3 mM), but not by efficacious levels of EGBG (0.2 to 0.6 mM). These results suggest that EGBG is more suitable as a specific inhibitor of polyamine biosynthesis and that use of this drug, rather than MGBG, in combination with alpha-difluoromethylornithine may be useful for studying the physiological functions of polyamines in animal cells.

Potentiation of the antimitochondrial and antiproliferative effects of bis(guanylhydrazones) by phenethylbiguanide.

The ability of methylglyoxal-bis(guanylhydrazone) (MGBG) and 4,4'-diacetyldiphenylurea-bis(guanylhydrazone) to interact with the hypoglycemic agent, phenethylbiguanide (DBI), in affecting the bioenergetic functions of isolated rat liver mitochondria was studied. DBI was found to increase markedly the inhibitory effect of either 4,4'-diacetyldiphenylurea-bis(guanylhydrazone) or MGBG on respiration of isolated rat liver mitochondria. Conversely, these bis(guanylhydrazones) enhanced the inhibitory potency of DBI and increased the apparent affinity of mitochondria for the drug. As with MGBG and 4,4'-diacetyldiphenylurea-bis(guanylhydrazone), the potassium cationophore, valinomycin, increased the sensitivity of mitochondrial respiration to DBI. It is suggested that the enhancement of bis(guanylhydrazone) inhibition of mitochondrial respiration by DBI involves inhibition of proton fluxes across the inner mitochondrial membrane and the subsequent alkalinization of the mitochondrial matrix. This drug interaction was extended to the level of antiproliferative activity in which DBI was found to potentiate the growth-inhibitory effects of MGBG on murine L1210 leukemia in vivo.

Biochemical localization of aryl hydrocarbon hydroxylase in the intestinal epithelium of the rat.

The distribution of the carcinogen-metabolizing enzyme system, aryl hydrocarbon hydroxylase (AHH), was biochemically determined in the intestinal epithelium of the rat. A method of epithelial cell isolation in which fractions of cells are sequentially collected as a villus tip-to-crypt gradient was used. AHH activity was highest in the midvillus region, 40% lower at the villus tip, and practically nonexistent in the crypt region where active cell proliferation takes place. This distribution differed from those of sucrase and alkaline phosphatase (used here as markers for cellular differentiation), which were characteristically lowest in activity at the crypts and increased continuously to the villus tips. Conceivably, the midvillus peak of AHH activity may serve to protect or enhance the susceptibility of cells undergoing cell division in the nearby crypt regions, depending on whether the predominant function of AHH in the intestinal epithelium involves detoxication or activation of polyaromatic carcinogens.

Autoradiographic distribution of hematoporphyrin derivative in normal and tumor tissue of the mouse.

The distribution of isotopically labeled hematoporphyrin derivative (HPD) has been studied in mice bearing the spontaneous mammary tumor (fast growing). In stomach, liver, spleen, and pancreas, 3 hr after i.p. injection of [3H]HPD, grains were uniformly distributed over the tissue sections. After 24 hr, the grain density overlying parenchymous areas of these tissues was lower than that over the stromal or reticuloendothelial areas. In the spontaneous mammary tumor (fast growing), higher grain densities were seen over pseudocapsule, stromal septa, and necrotic areas at 3, 6, 12, 24, and 48 hr after injection. At 168 hr postinjection, only isolated stomal cells, presumably macrophages, showed high grain densities. From the temporal changes observed in the distributions of HPD in normal tissues and the relative stability of the distribution seen in the spontaneous mammary tumor (fast growing), we speculate that tissue factors such as vascular permeability, lack of an adequate lymphatic drainage, and nonspecific binding of serum proteins to stromal elements may be responsible for or contribute to the preferential uptake and/or retention of HPD observed in both human and animal tumors.

Morphological evidence for an antimitochondrial action by methylglyoxal-bis(guanylhydrazone).

Exposure of cultured leukemia L1210 cells to 0.1 micron methylglyoxal-bis(guanylhydrazone) resulted in a concentration-dependent inhibition of cellular proliferation, beginning after about 1 to 2 generation times (12 to 24 hr). Ultrastructural examination of the cells treated at and above 1.0 micron concentrations of drug for 24 hr revealed unifrom damage to mitochondria. The basic lesion involved extensive swelling of the mitochondrion, a deterioration and eventual loss of cristae, and a decrease in the matrix density. This damage preceded growth inhibition by about 12 hr and did not immediately affect cell viability as detected by trypan blue dye exclusion. Whether these findings are related to the known actions of the drug on polyamine metabolism is not clear at present.

Effect of pretreatment with alpha-difluoromethylornithine on the selectivity of methylglyoxal bis(guanylhydrazone) for tumor tissue in L1210 leukemic mice.

A number of studies have demonstrated that pretreatment of tumor-bearing animals with the inhibitor of polyamine biosynthesis, alpha-difluoromethylornithine (DFMO), potentiates the antitumor activity of methylglyoxal bis(guanylhydrazone) (MGBG). The present study examines whether this phenomenon is related to a DFMO-mediated increase in the selectivity of MGBG for tumor tissue. Specifically, the effect of DFMO pretreatment on the tissue distribution and content of MGBG was investigated in mice bearing ascites L1210 leukemia. At 3 and 18 h following a single i.v. injection of [14C]MGBG (50 mg/kg), L1210 cells and seven tissues from nonpretreated (control) and DFMO-pretreated (3% by drinking water for 3 days) animals were compared for their [14C]MGBG content. In control mice, the greatest amount of drug was found in L1210 cells, small intestine, and kidney (in decreasing order of magnitude) at both 3 and 18 h. This distribution was not altered following DFMO pretreatment, but the relative MGBG content of other tissues was shifted. On an average, DFMO pretreatment increased the accumulation of MGBG by 30% in normal tissues and 32% in tumor tissues at 3 h and 56% and 69%, respectively, at 18 h. Thus, pretreatment of leukemic mice with DFMO fails to improve the selectivity of MGBG for L1210 cells. It is possible that other tumor systems might demonstrate sufficient DFMO-mediated increases in MGBG uptake to enhance drug selectivity but not without significantly increasing MGBG uptake (and hence toxicity) in normal tissues.

Enhancement of 1,3-bis(2-chloroethyl)-1-nitrosourea-induced cytotoxicity and DNA damage by alpha-difluoromethylornithine in L1210 leukemia cells.

Polyamine depletion by pretreatment with alpha-difluoromethylornithine (DFMO), a specific and irreversible inhibitor of ornithine decarboxylase, potentiates the cytotoxicity of 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) in L1210 leukemia cells grown in a modified soft agar system. The dose enhancement ratio was 1.97 at a control colony formation level of 5%. The basis for this enhancement was investigated at the level of DNA damage using a modified fluorometric assay to quantitate the production of alkaline-labile strand breaks per relative DNA molecular mass. Pretreatment of cultured L1210 cells for 48 hr with 5 mM DFMO depleted intracellular putrescine and spermidine (but not spermine) pools and resulted in a 2.3-fold increase in BCNU-induced (10 micrograms/ml, 2 hr) DNA strand breaks per relative DNA molecular mass. The inclusion of 10 microM spermidine during the DFMO pretreatment fully prevented growth inhibition and enhancement of BCNU-induced DNA damage while maintaining cellular spermidine pools at control levels. The inclusion of 2 microM putrescine or spermidine also prevented growth inhibition and enhancement of DNA damage while maintaining spermidine pools at only 25 to 35% of control. Thus, the portion of spermidine essential for cell growth appears to be associated with DNA. BCNU itself was found to reduce cellular polyamine levels by causing their leakage from cells. In addition, BCNU was found to react directly with spermidine in a cell-free system, resulting in a major reaction product detectable by high-performance liquid chromatography. While decreased interaction of BCNU with polyamines could account, in part, for enhancement effects of DFMO, it is more probable that alterations in DNA structure secondary to polyamine depletion are responsible for these effects.

Correlations between polyamine analogue-induced increases in spermidine/spermine N1-acetyltransferase activity, polyamine pool depletion, and growth inhibition in human melanoma cell lines.

The polyamine analogue, N1,N12-bis(ethyl(-spermine (BESPM), is known to suppress ornithine and S-adenosylmethionine decarboxylase levels, deplete intracellular polyamine pools, and inhibit cell growth. Among human melanoma cell lines, MALME-3 cells were found to be typically sensitive to the antiproliferative activity of the BESPM, whereas LOX cells were atypically insensitive to the analogue. A comparison of polyamine-related parameters revealed that the most differentially altered activity between the 2 BESPM-treated cell lines was that of spermidine/spermine N1-acetyltransferase (SSAT), which increased from 50 pmol/min/mg to greater than 10,000 pmol/min/mg in MALME-3 cells and from 16 pmol/min/mg to only 120 pmol/min/mg in LOX cells over 48 h. The basis for the large difference seems to be related to increased enzyme synthesis in both cell lines coupled with differences in prolongation of SSAT half-life (greater than 12 h in MALME-3 cells versus 1.6 h in LOX cells) after BESPM treatment. In MALME-3 cells, SSAT accumulation was found to be differentially modulated by the BESPM homologues, N1,N11-bis-(ethyl)norspermine and N1,N14-bis-(ethyl)homospermine, which were 5-fold more and 9-fold less effective, respectively, than BESPM in increasing SSAT but similar in analogue uptake and effects on polyamine biosynthesis and cell growth inhibition. Treatment of MALME-3 cells with BESPM resulted in an accumulation of N-acetylspermidine in cells and the enhanced excretion of putrescine, spermidine, and N-acetylspermidine into the medium. The relationship between SSAT induction and growth sensitivity was deduced to be a possible function of increased excretion of acetylated polyamines leading to enhanced polyamine pool depletion. The data suggest that, in cell types in which it occurs, unusually high increases in SSAT activity may serve as a determinant of growth sensitivity to bis-ethyl spermine analogues or, alternatively, as a target for appropriately designed chemotherapeutic strategies.

Relative abilities of bis(ethyl) derivatives of putrescine, spermidine, and spermine to regulate polyamine biosynthesis and inhibit L1210 leukemia cell growth.

It has been shown previously (Porter et al., Cancer Res., 45: 2050-2057, 1985) that the N1,N8-bis(ethyl) derivative of spermidine has significant antiproliferative activity which appears to derive from its regulatory effects on the polyamine biosynthetic pathway, particularly on ornithine decarboxylase activity. In the present study, N1,N4-bis(ethyl)putrescine (BEP) and N1,N12-bis(ethyl)spermine (BESm) were compared with N1,N8-bis(ethyl)spermidine (BES) in their ability to inhibit cell growth and regulate polyamine biosynthesis. With cultured L1210 murine leukemia cells, the IC50 values at 48 h were approximately 2 mM for BEP, 30 microM for BES, and 1 microM for BESm making the latter the most effective polyamine inhibitor or analogue thus far identified. At concentrations which approximated IC50 values and yielded similar intracellular concentrations at 48 h (1500-2000 pmol/10(6) cells), the effects of the analogues on polyamine biosynthesis generally correlated with their antiproliferative activity. BEP, at 1 mM, exerted relatively minor effects on polyamine biosynthesis. By contrast, 100 microM BES totally eliminated ornithine decarboxylase activity, depleted putrescine and spermidine pools, and decreased spermine pools by 40%. AdoMet decarboxylase activity was lowered slightly. The most impressive effects were obtained with 10 microM BESm which decreased ornithine and AdoMet decarboxylase activities by 99 and 84%, respectively; depleted putrescine and spermidine pools; and decreased spermine pools by 73%. None of the analogues, at 1 or 3 mM, had significant direct inhibitory effects on the decarboxylase activities from untreated cells with the exception of BESm which inhibited ornithine but not AdoMet decarboxylase activity. Thus, the effects of the analogues on these enzymes in treated cells are presumed to be mainly mediated by regulatory mechanisms. In this regard, BESm was superior to BES since both ornithine and AdoMet decarboxylase activities were suppressed. Given its unique activities, BESm would seem to have potential as both an antiproliferative agent and also as an experimental probe for studying regulation of the polyamine pathway, particularly AdoMet decarboxylase.

Antitumor activity of N1,N11-bis(ethyl)norspermine against human melanoma xenografts and possible biochemical correlates of drug action.

In in vitro systems, the spermine analogue, N1,N11-bis(ethyl)norspermine (BENSPM), suppresses the polyamine biosynthetic enzymes, ornithine and S-adenosylmethionine decarboxylase (ornithine decarboxylase and S-adenosylmethionine decarboxylase, respectively), greatly induces the polyamine catabolic enzyme, spermidine/spermine N1-acetyltransferase (SSAT), depletes polyamine pools, and inhibits cell growth. Against MALME-3 M human melanoma xenografts, BENSPM and related homologues demonstrate potent antitumor activity that has been found to correlate positively with their ability to induce SSAT activity in vitro. Herein, we further evaluate the antitumor activity of BENSPM and at the same time characterize the biochemical effects of BENSPM treatment on polyamine metabolism of selected normal and tumor tissues. At 40 mg/kg 3 times/day for 6 days i.p., BENSPM suppressed growth of MALME-3 M human melanoma xenografts during treatment and for 65 days afterwards. Similar antitumor activity was obtained with 120 mg/kg once daily for 6 days and 40 mg/kg once daily for 6 days, indicating that against this tumor model, the dosing schedule can be relaxed up to sixfold without compromising antitumor activity. When MALME-3 M tumor-bearing mice were retreated with BENSPM 2 weeks after the first treatment at 40 mg/kg 3 times/day for 6 days, initial tumor volumes of 85 mm3 were reduced to < 10 mm3. Analysis of melanoma, liver, and kidney tissues from mice treated with 40 mg/kg 3 times/day for 6 days revealed relatively similar accumulations of BENSPM in all tissues at levels greater than the original total content of polyamine pools. By 2 weeks following treatment, BENSPM pools in normal tissues were almost gone, whereas in tumor tissues significant amounts (40%) were still retained. The biosynthetic enzymes, ornithine decarboxylase and S-adenosylmethionine decarboxylase, gave no indication of enzyme suppression (or increase) by the analogue as typically occurs in vitro. By contrast, SSAT was induced from an average of < 50 pmol/min/mg in control tissues to 320 pmol/min/mg in liver, 1255 pmol/min/mg in kidney, and 13,710 pmol/min/mg in MALME-3M tumor. Two weeks later, SSAT activity was still 12 times higher in tumor than in kidney. Polyamine pools (putrescine, spermidine, and spermine) were reduced after treatment in all tissues and approached near-total depletion in the tumor. Good antitumor activity and even more potent induction of SSAT (i.e., 26,680 pmol/min/mg) was also observed in PANUT-3 human melanoma xenografts. Overall, the findings reveal meaningful antitumor activity by BENSPM against 2 human melanoma xenografts and provide in vivo evidence consistent with SSAT-induced polyamine depletion playing a determining role in at least the initial phase of the antitumor response.

Biochemical and ultrastructural characterization of human cell variants resistant to the antiproliferative effects of methylglyoxal bis(guanylhydrazone).

Stable variants of the human cell line, VA2-B, have been developed which are 10- to 20-fold less sensitive to the antiproliferative effects of methylglyoxal bis(guanylhydrazone) (MGBG) than the parent cell lines and which are not drug transport deficient. The lines were characterized biochemically giving particular attention to parameters related to the two known sites of MGBG action, mitochondria and polyamine metabolism. Dose-response studies with MGBG (0 to 30 microM for 40 to 48 hr) revealed that, of the parameters related to polyamine metabolism (i.e., polyamine pools, S-adenosylmethionine, and ornithine decarboxylase activities), only spermine pool size reduction seemed to correlate with inhibition of cell growth by MGBG. By contrast, decreases in pyruvate oxidation (used here as a measure of mitochondrial function) closely paralleled growth inhibition in all cell lines. Similarly, MGBG-induced changes in mitochondrial ultrastructure were less conspicuous in the variants than in the parent cell line and also corresponded with growth inhibition. Respiration of isolated mitochondria from one of the variant lines was about 2-fold more resistant to the inhibitory effects of MGBG than mitochondria from the VA2 cells. Finally, treatment with alpha-difluoromethylornithine, a potent inhibitor of polyamine biosynthesis having no known effect on mitochondrial function, resulted in comparable inhibition of growth in variant and parent cell lines. Overall, the data suggest that a phenotypic alteration in mitochondrial function, rather than in polyamine metabolism, may represent the basis for MGBG resistance in these variant cell lines.

Polyamines and biosynthetic enzymes in the rat intestinal mucosa and the influence of methylglyoxal-bis(guanylhydrazone).

The use of methylglyoxal-bis(guanylhydrazone) (MGBG) in the clinical treatment of myeloid and lymphoid disorders has been limited by severe host toxicity to renewing tissues, particularly the intestinal mucosa. Since the drug is a potent inhibitor of spermidine biosynthesis, the distributions of ornithine and S-adenosylmethionine decarboxylases and polyamine pools have been characterized in the rat intestinal mucosa in an attempt to discern the basis for MGBG toxicity. A method of epithelial cell isolation in which fractions of cells are sequentially collected in a villus tip-to-crypt gradient was used. Ornithine decarboxylase activity was highest in the villus tip region and unexpectedly lowest in the crypts, while S-adenosylmethionine decarboxylase activity showed the opposite pattern. Intracellular polyamine pools were uniform along the gradient corresponding to the villus length and increased appreciably in the crypt region. The relative concentrations of the individual polyamines were highest in the crypts, with spermidine and spermine being nearly equivalent in all regions. Twenty-four hr after a single i.p. injection of MGBG (50 mg/kg), S-adenosylmethionine decarboxylase activity increased markedly, especially in the crypt region (approximately 50-fold), while ornithine decarboxylase activity also increased but to a lesser extent. Putrescine pools were most affected by MGBG and were elevated 5- to 6-fold, especially in the crypt region. The results are consistent with an alteration of polyamine biosynthesis by MGBG being involved in the antiproliferative toxicity of the drug.