Evaluation of arylmethylaminopropanediols by a novel in vitro pharmacodynamic assay: correlation with antitumor activity in vivo.

The pharmacodynamics of a new series of antitumor DNA intercalators, known as arylmethylaminopropanediols (AMAPs), has been evaluated in vitro against adherent (MCF-7 human breast cancer) and nonadherent (P388 murine leukemia) cell lines. Previous work had shown that the in vitro antitumor activity of the model AMAP crisnatol was a function of exposure (Cn x T), rather than concentration alone. A unique exposure parameter, the minimum C x T, was proposed as an end point for antitumor activity in cell culture. Comparison of crisnatol to several established agents by the minimum C x T versus the standard concentration producing 10% survival indicated that these end points were not equivalent. The current work examined the validity of the pharmacodynamic approach using AMAP isomers from three different ring systems that were known to exhibit a spectrum of activity against the P388 tumor in vivo. The results indicated that antiproliferative, but not cytotoxic, activity of AMAPs in the pharmacodynamic assay correlated with their differential activity in vivo, expressed as percentage of increase in life span. In contrast, the concentration producing 10% survival either at 1 h or after continuous exposure did not show a similar correlation. The pharmacodynamic assay also revealed that certain AMAPs, while equipotent by concentration alone, required significantly less time and therefore less overall exposure for efficacy. Finally, the activity of AMAP isomers in P388 cells differed from that in MCF-7 cells, which may indicate AMAP selectivity for certain tumor types. Since AMAP action was a function of exposure, drug effects on cellular targets could likewise depend on exposure rather than concentration. These findings emphasize the importance of relating drug mechanisms to the pharmacodynamics of anticancer agents.

Phase I and clinical pharmacology trial of 502U83 using a monthly single dose schedule.

502U83 is an arylmethylaminopropanediol derivative exhibiting significant antineoplastic activity in a number of murine and human tumor models. In this Phase I trial, a 1-h or 4-h infusion of the agent was administered i.v. in 250 ml of 5% dextrose in water every 28 days. Fifty-three courses at doses of 25 to 2000 mg/m2 were administered to 36 patients with refractory solid tumors. Prolongation of the PR, QRS, and QT intervals on electrocardiograms was dose limiting at 2000 mg/m2. This prolongation appeared dose related and was reversible upon discontinuation of the infusion. No hematological toxicity was observed. Other toxicities included only sporadic and mild to moderate nausea and vomiting. No tumor responses were noted. 502U83 plasma concentrations were determined by high-pressure liquid chromatography. Complete pharmacokinetic profiles were obtained for 21 of the 36 patients. After infusion, plasma concentrations declined in a biexponential or in a triexponential manner with a harmonic mean terminal t 1/2 of 8.83 h. Using a three-compartment model, the mean apparent volume of distribution at steady state and total-body clearance were 195 liters/m2 and 42.5 liters/h/m2, respectively, indicative of extensive tissue distribution. No correlation could be found between the pharmacokinetic parameters and prolongation of the cardiac conduction intervals. Because of the cardiac effects with the drug, the schedule of administration of 502U83 used in this study cannot be recommended.

Protein-protein interactions: lessons learned.

Protein-Protein (P-P) interactions play a pivotal role in determining cellular structure and in all cellular processes. The nature of P-P interactions is complex, and despite the large amount of research that has occurred in the field, is still poorly understood. Abnormal P-P interactions are particularly important because of their association with a variety of diseases, including cancer. This review examines P-P interactions with particular emphasis on the discovery of new anti-tumor drugs, including underlying physical forces that determine affinity and specificity and discusses classical and newer strategies used to discover inhibitors of P-P interactions, providing a number of recent cancer-related case studies and commentary.

(1-Pyrenylmethyl)amino alcohols, a new class of antitumor DNA intercalators. Discovery and initial amine side chain structure-activity studies.

In the series of 1-pyrenylmethylamines studied in this work the relationships among structure, interaction with DNA, and murine antitumor activity were examined. Binding studies show that all of these 1-pyrenylmethylamine derivatives bind to some extent to DNA by intercalation. The presence of additional basic amine groups in the side chain enhances DNA binding due to electrostatic interactions. Those compounds containing only a single basic benzylic amine bind similarly to DNA. Only the presence of bulky side chains appears to decrease the DNA interactions in the compounds examined. Although antitumor activity is seen for (1-pyrenylmethyl)amino alcohols, useful antitumor activity in the series is limited to those congeners bearing the 2-amino-1,3-propanediol-type side chain. These derivatives bind moderately to DNA. DNA binding is a necessary but not sufficient criterion for antitumor activity in the series. In addition, the strength of DNA binding does not correlate with the antitumor activity in the group of active compounds. Three related 2-[(arylmethyl)amino]-1,3-propanediol derivatives (AMAPs) [crisnatol (770U82), 773U82, and 502U83] are currently in clinical trials as potential antitumor agents.

Structural and conformational requirements for high-affinity binding to the SH2 domain of Grb2(1).

Following earlier work on cystine-bridged peptides, cyclic phosphopeptides containing nonreducible mimics of cystine were synthesized that show high affinity and specificity toward the Src homology (SH2) domain of the growth factor receptor-binding protein (Grb2). Replacement of the cystine in the cyclic heptapeptide cyclo(CYVNVPC) by D-alpha-acetylthialysine or D-alpha-lysine gave cyclo(YVNVP(D-alpha-acetyl-thiaK)) (22) and cyclo(YVNVP(D-alpha-acetyl-K)) (30), which showed improved binding 10-fold relative to that of the control peptide KPFYVNVEF (1). NMR spectroscopy and molecular modeling experiments indicate that a beta-turn conformation centered around YVNV is essential for high-affinity binding. X-ray structure analyses show that the linear peptide 1 and the cyclic compound 21 adopt a similar binding mode with a beta-turn conformation. Our data confirm the unique structural requirements of the ligand binding site of the SH2 domain of Grb2. Moreover, the potency of our cyclic lactams can be explained by the stabilization of the beta-turn conformation by three intramolecular hydrogen bonds (one mediated by an H2O molecule). These stable and easily accessible cyclic peptides can serve as templates for the evaluation of phosphotyrosine surrogates and further chemical elaboration.

2-[(arylmethyl) amino]-2-methyl-1,3-propanediol DNA intercalators. An examination of the effects of aromatic ring variation on antitumor activity and DNA binding.

The effects of variation of aromatic ring size, shape, and side-chain position on antitumor activity and DNA binding in a series of carbocyclic 2-[(arylmethyl)amino]-2-methyl-1,3-propanediols (AMAPs) were examined. In general, the interaction of AMAPs with DNA increases as the intercalating ring system grows in area, with three distinct binding levels evident. Isomers from a specific ring system appear to bind DNA similarly. DNA binding is not the sole criterion for antitumor activity for the AMAPs studied; the magnitude of the delta Tm does not correlate with the antitumor activity observed. Significant in vivo P388 activity was seen for AMAP congeners from several tetracyclic ring systems. However, isomers from each of the specific ring systems produced a wide range of in vivo P388 activity. Thus, AMAP antitumor activity is not a function of the ring system per se, but rather appears to be related to the shape of the specific molecule. Three AMAP congeners (crisnatol (770U82, 773U82, and 502U83) are currently in clinical trials.

Synthesis and antitumor activity of ester-modified analogues of bengamide B.

Bengamide B, a novel sponge-derived marine natural product with broad spectrum antitumor activity, was not suitable for further preclinical development because of its difficult synthesis and very poor water solubility. Bengamide B produced a 31% T/C at its solubility-limited maximum intravenous dose of 33 micromol/kg in MDA-MB-435 breast carcinoma implanted subcutaneously as a xenograft in nude mice. Compound 8a, a bengamide B analogue with three structural changes (t-Bu alkene substituent, unsubstituted lactam nitrogen, and inverted lactam 5'-myristoyloxy group), was as potent as bengamide B in vitro and more efficacious than bengamide B in vivo. A series of ester-modified analogues based on 8a were synthesized and tested in vitro and in vivo (MDA-MB-435). The cyclohexyl- and phenethyl-substituted esters, 8c and 8g, respectively, had in vitro and in vivo activities similar to that of 8a and enhanced water solubility (ca. 1 mg/mL). Consequently, 8c and 8g were tested in the MDA-MB-435 xenograft model at 100 micromol/kg and produced 29% and 57% tumor regression, respectively.

Polyploidy induction as a consequence of topoisomerase inhibition. A flow cytometric assessment.

Following recovery from a 4-hr exposure to clinically achievable concentrations of the topoisomerase II inhibitors Adriamycin, teniposide, or amsacrine or the putative topoisomerase II inhibitor crisnatol, murine erythroleukemic cells remained viable for up to 48 hr, but did not proliferate. Cell cycle analysis after a 24-hr recovery revealed blocks in G2 (4N DNA) or greater than G2 (up to 8N DNA) polyploid stages. The relative percentages of cells in either stage was a function of drug concentration and cell cycle stage at time of exposure: typically, cells exposed during S phase became blocked in G2, whereas those exposed during G2/M progressed into greater than G2 polyploid stages. G2-blocked cells exhibited a 2- to 3-fold increase in nuclear protein content and cellular/nuclear volume (i.e. unbalanced growth) and approximately 5% more DNA stainability (as a consequence of nuclear conformational changes rather than redundant DNA synthesis). In all cases, at the drug concentrations studied, mitotic figures were absent and G2 and greater than G2 blocks were irreversible, indicating that the mechanism of polyploidy induction differs from that of microtubule inhibitors. These findings suggest that although topoisomerase inhibitors interfere with DNA synthesis in the S phase, their induction of greater than G2 polyploid blocks may involve direct or indirect inhibition of chromosome condensation.

Analysis of the interactions of SDZ PSC 833 ([3'-keto-Bmt1]-Val2]-Cyclosporine), a multidrug resistance modulator, with P-glycoprotein.

Multidrug resistance (MDR) is considered to be an important impediment to the effective treatment of cancer. P-glycoprotein, the drug efflux pump that mediates this resistance, can be inhibited by a wide variety of pharmacological agents, resulting in the circumvention of the MDR phenotype. SDZ PSC 833 ([3'-keto-Bmt1]-Val2]-cyclosporine), a nonimmunosuppressive cyclosporine D derivative, was identified to be a potent MDR modulator (Gaveriaux et al. J. Cell Pharmacol. 2:225-234; 1991). In this study, the interactions of P-glycoprotein with two cyclosporine derivatives, SDZ PSC 833 and cyclosporine A (CsA, Sandimmune), were analyzed. SDZ PSC 833 enhanced the sensitivity of the MDR cells to anticancer drugs by increasing the accumulation and inhibiting the efflux of cytotoxic agents from resistant cells more efficiently than CsA. The two cyclosporine analogs competed with the labeling of P-glycoprotein by a photoactive cyclosporine derivative. In addition, membrane vesicles derived from resistant cells bound SDZ PSC 833. However, CsA was transported by P-glycoprotein, whereas SDZ PSC 833 was not actively transported. This resulted in a prolonged inhibitory effect by SDZ PSC 833. The studies suggest that the binding of SDZ PSC 833 to P-glycoprotein in the absence of its transport from MDR cells mediated its high potency as an MDR reversing agent. In addition, the comparison of the two cyclosporine analogs indicated that limited chemical modifications of MDR reversing agents can affect their potential to inhibit P-glycoprotein function.

The reversal of multidrug resistance in multicellular tumor spheroids by SDZ PSC 833.

Multidrug resistance (MDR) is a major impediment to the effective treatment of cancer. We have used multicellular tumor spheroids (MTS) as a model to investigate whether MDR can be reversed in a three dimensional structure. MTS are tightly associated aggregates of tumor cells that exhibit many of the properties of solid tumors. A human MDR breast carcinoma cell line was selected by exposure to taxol under monolayer conditions. The sensitive (parental) and drug-resistant phenotypes were retained when the cells were grown as MTS. Thus, the three dimensional conditions in this novel model system did not affect the MDR phenotype. SDZ PSC 833 is an efficient MDR reversing agent as determined under monolayer conditions and is currently being evaluated in clinical trials. Resistance to taxol and doxorubicin of the MDR cells grown as MTS was almost completely reversed by SDZ PSC 833. Our results suggest that SDZ PSC 833 has the potential to reverse the MDR phenotype in solid tumors.

Effects of isomeric 2-(arylmethylamino)-1,3-propanediols (AMAPs) and clinically established agents on macromolecular synthesis in P388 and MCF-7 cells.

The in vitro effects of the 2-(arylmethylamino)-1,3-propanediols (AMAPs) on macromolecular synthesis have been examined using the murine leukemia, P388, and the human mammary adenocarcinoma, MCF-7, under conditions of short-term drug exposure. AMAPs that were observed to inhibit macromolecular synthesis produced nearly equipotent inhibition of DNA and RNA synthesis. Equivalent inhibition of protein synthesis generally required significantly greater concentrations of AMAP. There is a general correlation between inhibition of polynucleotide synthesis and in vivo antitumor activity. The effects of four clinical candidate AMAPs (crisnatol, 773U82, 502U83, and 7U85) on macromolecular synthesis were further compared with those of actinomycin D, doxorubicin, mitoxantrone, etoposide, amsacrine, and cisplatin in MCF-7 cells. The pattern of AMAP action was most similar to that observed for doxorubicin and mitoxantrone. Finally, the effects of these four AMAPs on the size, specific activity, and rate of incorporation of [3H]-dTTP into DNA of MCF-7 cells synchronized by pretreatment with hydroxyurea was determined. It was found that DNA synthesis was inhibited by AMAPs independent of inhibition of the uptake, phosphorylation, or retention of the metabolic precursors. These results support the theory that antitumor AMAPs interfere with the normal functioning of enzymes, such as topoisomerase II or DNA and RNA polymerases, which interact with DNA.

Correlation of cytotoxicity and protein-associated DNA strand breaks for 2-(arylmethylamino)-1,3-propanediols.

A mechanism of action study was performed with 14 novel DNA binding agents characterized structurally as 2-(arylmethylamino)-1,3-propanediols (AMAPs). Correlations between 8226 myeloma cell colony formation and DNA damage were performed using soft agar colony-forming assays and alkaline elution filter techniques respectively. The frequency of double-stranded breaks (DSBs), single-stranded breaks (SSBs) and DNA-protein cross-links were compared with cell growth inhibitory potency. Highly potent AMAPs in the colony formation assays included 91U86, an N-methyl-5-benzo(c)carbazole derivative, 773U82, a 3-substituted fluoranthene derivative, and crisnatol (770U82), the 6-substituted chrysene derivative. There was a high frequency of SSBs and DSBs with many analogues, but only SSBs occurred in a concentration-dependent fashion. Using regression analysis, the degree of single-strand damage correlated with cytotoxic potency for the AMAPs, with an R-value of 0.57 (P = 0.04). By gel electrophoresis assays, three clinically tested AMAPs, crisnatol BW 770U82, BW 502U83 and BW 773U82, were shown to inhibit the decatenation of pBR 322 DNA by purified topoisomerase-II (TOPO-II) enzymes. These results suggest that while some active AMAPs, such as crisnatol (BW 770U82), BW 502U83 and BW 773U82, inhibit TOPO-II enzymes, leading to protein-associated SSBs, other mechanisms, which do not involve DNA strand damage, must also contribute to the cytotoxic effects of this class of antitumor compounds. Intercalation has been well documented for these drugs and this may explain some of the growth inhibitory activity of the AMAPs.

Limitations of the fluorescent probe viability assay.

In vitro assessment of the efficacy/capacity of toxicants (e.g., cancer chemotherapeutic agents, environmental pollutants, etc.) to damage/kill cells and/or inhibit growth (cell duplication) requires accurate measurement of target cell viability as a function of exposure. Rapid measurement of viability, such as can be achieved employing fluorescent probes of metabolic function in combination with instrumental analysis, is highly desirable. However, we observe that exposure to chemicals (of unrelated type) complicates the interpretation of viability data and, in the case of perturbed cells, questions the validity of viability growth assays based on intrinsic enzyme activity. Viability commonly is determined flow cytometrically (FCM) by the carboxyfluorescein diacetate (CFDA)/propidium iodide (PI) assay. Nonfluorescent CFDA is taken up by diffusion and converted via cytoplasmic esterase-catalyzed hydrolysis to carboxyfluorescein (CF), a negatively charged fluorescent molecule that is retained (incompletely) by the cell. As such, if CF fluorescence intensity is a relative measure of enzyme activity, it also can be considered an index of cellular vigor (metabolic rate). It is generally accepted that the viable cell excludes both basic dyes, such as PI, and acidic dyes, such as trypan blue, and uptake is indicative of irreversible cellular injury presaging cell death. We observe that, following incubation for 4 h with 0.5-1.0 microM tributyltin (TBT), a potent environmental toxicant, murine erythroleukemic cells (MELC) exhibit enhanced (supranormal) CF fluorescence compared to control cells. Apparent cell volume (ACV) is unaltered, and because such cells exclude PI, they are considered viable in terms of the CFDA/PI assay. However, rate of growth (increase in cell number over 48 h) is depressed, suggesting that supranormal CF fluorescence, even in the absence of PI uptake, is indicative of cellular perturbation. In effect, although CF fluorescence is the product of an enzyme-catalyzed reaction and, therefore, an indicator of vital function (enzyme activity), it apparently is not a reliable index of cellular vigor. At higher TBT concentrations (greater than 1.0, but less than 50.0 microM), the cells exhibit both increased CF fluorescence and PI fluorescence and are growth inhibited. MELC exposed to the cancer chemotherapeutic agents adriamycin, m-AMSA, or crisnatol (Burroughs Wellcome 770U82) also exhibit increased cellular CF fluorescence. However, rate of growth is decreased and ACV increased. The latter, measured either as a function of electrical resistance (Coulter volume) or by the FCM parameter axial light loss could account for the increase in mean CF fluorescence.(ABSTRACT TRUNCATED AT 400 WORDS)

An efficient multiple-exposure analysis of the toxicity of crisnatol, a DNA intercalator in phase II clinical trials.

To investigate the toxicity and mechanism of action of crisnatol (CRS), a new DNA intercalator currently in phase II clinical trials, we analyzed cellular and nuclear flow cytometric (FCM) parameters of murine erythroleukemic cells (MELC) exposed to a range of CRS concentrations over three exposure conditions: short-term (4 h), long-term (24 h), and short-term with recovery (4 h+/19 h-). At 0.5-1.0 microM CRS, 4 h exposure results in a reversible G2-phase block, while 24 h exposure results in greater than G2 polyploidy. At 5-10 microM CRS concentrations, cells exhibit persistent retardation of S-phase progression or irreversible G2 and/or greater than G2 blocks, depending on duration of exposure. Cells terminally blocked in G2 exhibit increased nuclear/cellular volumes and increased nuclear fluorescein isothiocyanate (protein) staining, suggestive of unbalanced growth. At 25-50 microM CRS concentrations, MELC exhibit severe membrane perturbation (loss of viability) regardless of exposure. In contrast, following similar exposures to an inactive isomer of CRS, MELC exhibit minimal cell cycle effects, suggesting that cell cycle kinetics may be a useful criterion for assessing potential efficacy. Similar analyses with different classes of chemotherapeutic agents reveal that the range of induced cellular/nuclear perturbations varies with the class of compound used. Taken together, these results suggest that drug toxicity can vary with both concentration and duration of exposure and, as such, a selective multiple-exposure FCM analysis may better represent the spectrum of drug action for drug development and pharmacodynamic studies.

Solid-phase total synthesis of polymyxin B1.

The total solid-phase synthesis of polymyxin B1 (PMB1) has been achieved in 20% yield using the orthogonal protecting group N-1-(4,4-dimethyl-2,6-dioxocyclohexylidene)ethyl-(Dde). This report demonstrates that a complex peptide macrocycle can be synthesized in high yields using solid-phase synthesis. According to MS and HPLC, the synthetic peptide was identical to the naturally occurring antibiotic.

Identification of E2F-1/Cyclin A antagonists.

A simple method for the synthesis of a rationally designed (S,S)-[Pro-Leu]-spirolactam scaffold is described. This was expanded to a small biased library of compounds mimicking the 'ZRXL' motif in order to identify E2F-1/Cyclin A antagonists. The synthesized compounds were evaluated in an E2F-1/Cyclin A binding assay and moderately active analogues were identified. In addition, the critical roles of Phe, Leu, Lys, and Arg residues of the identified motif were determined.

Selective killing of transformed cells by cyclin/cyclin-dependent kinase 2 antagonists.

Recent studies identified a short peptide motif that serves as a docking site for cyclin/cyclin-dependent kinase (cdk) 2 complexes. Peptides containing this motif block the phosphorylation of substrates by cyclin A/cdk2 or cyclin E/cdk2. Here we report that cell membrane-permeable forms of such peptides preferentially induced transformed cells to undergo apoptosis relative to nontransformed cells. Deregulation of E2F family transcription factors is a common event during transformation and was sufficient to sensitize cells to the cyclin/cdk2 inhibitory peptides. These results suggest that deregulation of E2F and inhibition of cdk2 are synthetically lethal and provide a rationale for the development of cdk2 antagonists as antineoplastic agents.

Phase I evaluation of 773U82.HCl, a member of a new class of DNA intercalators.

The arylmethylaminopropanediols (AMAPs) are a new class of DNA intercalators. 773U82.HCl is the second of these compounds to enter clinical trial. Significant antitumor activity for 773U82.HCl was documented in a variety of murine and human tumor models. This phase I study examined a 1-, 2- and 6-hour infusion given every 28 days. Thirty-six patients received 58 courses of drug at doses ranging from 15 mg/m2 to 980 mg/m2. The dose-limiting toxicity of 773U82.HCl was hemolysis noted at 980 mg/m2. Change in color of the plasma and decreases in haptoglobin were correlated with drug concentrations of the infusate greater than or equal to 3 mg/ml. Clinically significant changes in hemoglobin levels requiring blood transfusions did not occur. Neurologic toxicity occurred at 720 mg/m2 with the most severe neurologic toxicity occurring in a patient with the highest peak plasma concentration (4.1 micrograms/ml). With an increase in duration of the infusion and amount of fluid administered, the neurologic toxicity resolved. Other toxicities included mild nausea and vomiting and a dose-related phlebitis. Pharmacokinetic studies were completed in 22 patients. The mean terminal t1/2 beta was 4.4 h with a mean apparent volume of distribution at steady state (Vdss) of 314 l/m2. The mean total body clearance was 72 l/h/m2. Peak plasma levels ranged from 0.04 to 4.14 micrograms/ml. Further studies with 773U82.HCl on this schedule at the doses studied are not recommended. Hematologic monitoring for evidence of intravascular hemolysis should be included in future studies with 773U82.HCl.

Bengamides revisited: new structures and antitumor studies.

The structural chemistry and biological activity of the bengamide class of compounds have been further characterized. Extracts prepared from recollected Jaspis cf. coriacea from five sites in Fiji were pooled. Six new bengamides, M (7b), N (8a), O (8b), P (9a), Q (9b), and R (10), were identified, accompanied by the known bengamides A (1a), B (1b), E (3a), F (3b), Y (5), Z (6), L (7a), G (11a), H (11b), and I (12). The structures of the new compounds were determined from spectroscopic data, and some were additionally confirmed by semisynthesis. Cytotoxicity screening data were obtained from the NCI-DTP 60 cell screen for bengamides A, B, and P. Bengamides A and B were more potent than bengamide P, with average IC(50) values of 0.046, 0.011, and 2.70 FM, respectively. The in vitro antitumor activity against MDA-MB-435 human mammary carcinoma was also determined for natural bengamides A, B, E, F, P, M, O, and Z and for synthetic samples of B and O. The best activity was observed for the natural bengamides A (IC(50) = 1 nM) and O (IC(50) = 0.3 nM).

Total syntheses of bengamides B and E.

Total syntheses of the cytotoxic marine natural products bengamides B and E are described. Both bengamides are prepared via amide coupling of a protected polyhydroxylated lactone intermediate 9 with a suitably substituted aminocaprolactam intermediate. Lactone 9 is prepared in five steps from commercially available alpha-D-glucoheptonic gamma-lactone. The key reactions are a selective deprotection of a 1,2-acetonide in the presence of a 1,3-acetonide and an (E)-selective olefination of an unstable aldehyde using a gem-dichromium reagent. The bengamide B lactam intermediate 10 is prepared in seven steps from commercially available (5R)-5-hydroxy-L-lysine (12). The desired S-configuration at the gamma-OH lactam position is established using the Mitsunobu reaction.