A novel P450-catalyzed transformation of the 2,2,6,6-tetramethyl piperidine moiety to a 2,2-dimethyl pyrrolidine in human liver microsomes: characterization by high resolution quadrupole-time-of-flight mass spectrometry and 1H-NMR

We describe herein a novel metabolic fate of the 2,2,6,6-tetramethyl-piperidine (2,2,6,6-TMPi) moiety to a ring-contracted 2,2-dimethyl pyrrolidine (2,2-DMPy) in human liver microsomal incubations. The existence of this pathway was demonstrated for three compounds (I-III) of varied structures suggesting that this may be a general biotransformation reaction for the 2,2,6,6-TMPi moiety. The 2,2-DMPy metabolites formed in incubations of the three compounds with human liver microsomes were characterized by online high performance liquid chromatography coupled to a high resolution hybrid quadrupole-time-of-flight mass spectrometer. Suggested elemental composition obtained from accurate mass measurements of the molecular ions and fragment ions of the metabolites clearly indicated the loss of a mass equivalent to C(3)H(6) from the parent 2,2,6,6-TMPi functionality. Additional accurate tandem mass spectrometry data indicated that one of the original two gem-dimethyl groups was intact in the metabolite structure. Proof of a ring-contracted 2,2-DMPy structure was obtained using (1)H-NMR experiments on a metabolite purified from liver microsomal incubations, which showed only two geminal methyl groups, instead of four in the parent compound. Two-dimensional correlation spectroscopy and decoupling experiments established aliphatic protons arranged in a pyrrolidine ring pattern. The fact that the formation of 2,2-DMPy metabolites in human liver microsomes was NADPH-dependent suggested that this novel metabolic reaction was catalyzed by the cytochrome P450 (P450) enzyme(s). Immunoinhibition studies in human liver microsomal incubations using anti-P450 monoclonal antibodies and experiments with insect cell microsomes containing individually expressed recombinant human P450 isozymes indicated that multiple P450 isozymes were capable of catalyzing this novel metabolic transformation.

Addressing the metabolic activation potential of new leads in drug discovery: a case study using ion trap mass spectrometry and tritium labeling techniques

Metabolic activation of drug candidates to electrophilic reactive metabolites that can covalently modify cellular macromolecules may result in acute and/or idiosyncratic immune system-mediated toxicities in humans. This presents a significant potential liability for the future development of these compounds as safe therapeutic agents. We present here an example of an approach where sites of metabolic activation within a new drug candidate series were rapidly identified using online liquid chromatography/multi-stage mass spectrometry on an ion trap mass spectrometer. This was accomplished by trapping the reactive intermediates formed upon incubation of compounds with rat and human liver microsomes as their corresponding glutathione conjugates and mass spectral characterization of these thiol adducts. Based on the structures of the GSH adducts identified, potential sites and mechanisms of bioactivation within the chemical structure were proposed. These metabolism studies were interfaced with iterative structural modifications of the chemical series in order to block these bioactivation sites within the molecule. This strategy led to a significant reduction in the propensity of the compounds to undergo metabolic activation as evidenced by reductions in the irreversible binding of radioactivity to liver microsomal material upon incubation of tritium-labeled compounds with this in vitro system. With the efficiency and throughput achievable with such an approach, it appears feasible to identify and address the metabolic activation potential of new drug leads during routine metabolite identification studies in an early drug discovery setting.

A nonpeptide oxytocin receptor antagonist radioligand highly selective for human receptors

A novel, potent nonpeptide oxytocin receptor antagonist (1-(1-(2-(2,2,2-trifluoroethoxy)-4-(1-methylsulfonyl-4-piperidinyloxy) phenylacetyl)-4-piperidinyl)-3,4-dihydro-2(1H)-quinolinone) has been identified that can be labeled to high specific activity with [35S]. In binding studies, this compound exhibits sub-nanomolar affinity and a high degree of selectivity (900-1800-fold) for human oxytocin receptors compared to human vasopressin receptors. This compound appears suitable for studying the pharmacology of oxytocin receptors in human and nonhuman primate tissues, for which there is currently a paucity of highly selective tools. It may also be useful as a nonlabeled competitor or as a radioligand in autoradiographic studies of oxytocin receptor localization in these tissues.

Ivermectin and nodulisporic acid receptors in Drosophila melanogaster contain both gamma-aminobutyric acid-gated Rdl and glutamate-gated GluCl alpha chloride channel subunits

35S-labeled derivatives of the insecticides nodulisporic acid and ivermectin were synthesized and demonstrated to bind with high affinity to a population of receptors in Drosophila head membranes that were previously shown to be associated with a glutamate-gated chloride channel. Nodulisporic acid binding was modeled as binding to a single population of receptors. Ivermectin binding was composed of at least two kinetically distinct receptor populations, only one of which was associated with nodulisporic acid binding. The binding of these two ligands was modulated by glutamate, ivermectin, and antagonists of invertebrate gamma-aminobutyric acid (GABA)ergic receptors. Because solubilized nodulisporic acid and ivermectin receptors comigrated as 230-kDa complexes by gel filtration, antisera specific for both the Drosophila glutamate-gated chloride channel subunit GluCl alpha (DmGluCl alpha) and the GABA-gated chloride channel subunit Rdl (DmRdl) proteins were generated and used to examine the possible coassembly of these two subunits within a single receptor complex. DmGluCl alpha antibodies immunoprecipitated all of the ivermectin and nodulisporic acid receptors solubilized by detergent from Drosophila head membranes. DmRdl antibodies also immunoprecipitated all solubilized nodulisporic receptors, but only approximately 70% of the ivermectin receptors. These data suggest that both DmGluCl alpha and DmRdl are components of nodulisporic acid and ivermectin receptors, and that there also exists a distinct class of ivermectin receptors that contains the DmGluCl alpha subunit but not the DmRdl subunit. This co-association of DmGluCl alpha and DmRdl represents the first biochemical and immunological evidence of coassembly of subunits from two different subclasses of ligand-gated ion channel subunits.

An interleukin-1 genotype is associated with fatal outcome of meningococcal disease

To determine whether known variants of the interleukin-1 (IL-1) and tumor necrosis factor (TNF) gene families are associated with severe manifestations of meningococcal disease, 276 white patients 4-70 years of age (median, 17 years) were genotyped. All patients had microbiologically proven Neisseria meningitidis infection; 39 died and 237 survived. A significant association (P<.001) was found between fatal outcome and genotype at IL1B (nucleotide position -511). Homozygous individuals, both for the common (1/1) and the rare (2/2) alleles, had increased odds ratios (ORs) for death, compared with heterozygous individuals (1/2): ORs (95% confidence intervals [CIs]) were 3.39 (1.39-8.29) and 7.35 (2.51-21.45), respectively. The mortality rates according to genotype at IL1B (-511) were 18.0% (1/1), 6.1% (1/2), and 32.3% (2/2), compared with 14.2% overall. The composite genotype, consisting of heterozygosity of IL1B (-511) together with homozygosity of the common allele of the IL-1 receptor antagonist gene (IL1RN) at +2018, was significantly associated with survival (P=.018; OR, 7.78 [95% CI, 1. 05-59.05]). There was no association between TNF genotype and fatal outcome. These data suggest that IL-1 genotype influences the severity of meningococcal disease.

The influence of interleukin gene polymorphism on expression of interleukin-1beta and tumor necrosis factor-alpha in periodontal tissue and gingival crevicular fluid

BACKGROUND

A specific composite genotype of the polymorphic interleukin-1 (IL-1) gene cluster has recently been associated with severe periodontitis. One polymorphism of the composite periodontitis-associated genotype (PAG) has been functionally linked with expression of high levels of IL-1. The purpose of this study was to test whether gingival crevicular fluid (GCF) levels of IL-1beta and tumor necrosis factor-alpha (TNFalpha), and gingival tissue levels of IL-1alpha, IL-1beta, and TNFalpha correlate with PAG, and to examine the effect of conservative periodontal therapy on these levels.

METHODS

Twenty-two adults with moderate to advanced periodontal disease were enrolled. Polymerase chain reaction amplification and restriction enzymes were used to identify specific polymorphisms from peripheral blood samples. GCF samples were collected at baseline and 3 weeks following conservative treatment and analyzed by ELISA for IL-1beta and TNFalpha. An interproximal gingival biopsy was collected at baseline and follow-up and analyzed for IL-1alpha, IL-1beta, and TNFalpha by ELISA.

RESULTS

The genotyping identified 7 as PAG(+) and 15 as PAG(-). The 2 groups were comparable in terms of existing periodontitis and age. In shallow sites (<4 mm), total IL-1beta in GCF was 2.5 times higher for PAG(+) patients prior to treatment (P=0.03), and 2.2 times higher after treatment (P=0.04), while differences were less apparent in deeper sites. Following treatment, a reduction in IL-1beta concentration in GCF was seen for PAG(-) but not for PAG(+) patients. While not statistically significant, a trend was observed in mean tissue levels of IL-1beta which were 3.6 times higher in PAG(+) versus PAG(-) patients (P=0.09).

CONCLUSIONS

These data suggest that PAG(+) patients may demonstrate phenotypic differences as indicated by elevated levels of IL-1beta in GCF.

Synthesis and biological evaluation of 2-acyl analogues of paclitaxel (Taxol)

The anticancer drug paclitaxel (Taxol) has been converted to a large number of 2-debenzoyl-2-aroyl derivatives by three different methods. The bioactivities of the resulting analogues were determined in both tubulin polymerization and cytotoxicity assays, and several analogues with enhanced activity as compared with paclitaxel were discovered. Correlation of cytotoxicity in three cell lines with tubulin polymerization activity showed reasonable agreement. Among the cell lines examined, the closest correlation with antitubulin activity was observed with a human ovarian carcinoma cell line.

Distances between the paclitaxel, colchicine, and exchangeable GTP binding sites on tubulin

Distances between the paclitaxel, colchicine, and exchangeable GTP binding sites on tubulin polymers have been probed using fluorescence spectroscopy. Techniques for measuring fluorescence resonance energy transfer (FRET) between fluorescent or chromophoric ligands for each binding site were employed. 2-Debenzoyl-2-(m-aminobenzoyl)paclitaxel (2-AB-PT) was the fluorophore ligand for the paclitaxel binding site; thiocolchicine, allocolchicine, and MDL 27048 were probes for the colchicine site, and 2'(or 3')-O-(trinitrophenyl)guanosine 5'-triphosphate (TNP-GTP) was the fluorophore ligand for the exchangeable GTP site. The distance between the colchicine and paclitaxel binding sites was determined with two different acceptor ligands in the colchicine site. An average distance distribution of 17 A was found in both cases. Energy transfer between 2-AB-PT bound to the paclitaxel site and TNP-GTP (acceptor) bound to the exchangeable GTP site was observed in the polymer. The average distance distribution between the fluorophores was 16.0 A, but the half-width of the distribution was large (17.9 A), which indicates that energy transfer between more than one donor-acceptor pair occurred in the system. One interpretation of this result is that 2-AB-PT serves as an energy transfer donor for two GTP sites, one contained on the same subunit and one on an adjacent protofilament. No FRET was observed between ligands bound to the colchicine and exchangeable GTP sites, indicating that the result of colchicine binding on the GTP region of beta-tubulin is a long range, allosteric effect. The results from these experiments are interpreted in terms of known structural features of microtubules.

Identification of the structural region of taxol that may be responsible for cytokine gene induction and cytotoxicity in human ovarian cancer cells

PURPOSE

Interleukin-8 (IL-8) is a pleiotropic chemokine with both chemoattractant and angiogenic properties. In addition to its cytotoxic effects on ovarian cancer cells, taxol can transcriptionally activate genes such as IL-8 that may play a role in tumorigenesis. Utilizing IL-8 as a prototypic marker of tumor-derived modulators of growth, we undertook a systematic study of taxol and 11 structurally modified taxol analogs to identify the region of the taxane skeleton responsible for IL-8 gene induction.

METHODS

The human ovarian cancer cell line OVCA-420 was exposed to taxol or taxol analogs. IL-8 gene induction was assessed by Northern blot analysis after 6 h and cytotoxicity after 72 h.

RESULTS

Changes in the southern hemisphere (C-1 to C-4) of the taxane skeleton had greater effects on IL-8 induction than changes in the northern hemisphere (C-7 to C-11). Some of the taxol analogs modified at positions C-1 and/or C-2 with increased hydrophobicity induced IL-8 expression more than threefold over that induced by taxol or taxotere and more than 20-fold over control cells. Cells that failed to induce IL-8 gene expression in response to taxol were only marginally responsive to the analogs unless first primed with IL-1beta. Modifications to the northern hemisphere did not alter taxol's effect on IL-8 expression in human cells, but did influence TNFalpha expression in murine macrophage cells, suggesting species and/or gene specificity. We found a direct correlation between IL-8 induction and cytotoxicity, in that analogs that dramatically upregulated IL-8 expression proved to be the most cytotoxic, inhibiting cell survival by > 90%.

CONCLUSION

Taken together our results demonstrate that changes in the southern hemisphere of the taxane skeleton influence both the gene induction and cytotoxic potential of taxol in human ovarian cancer cells.

Interaction of a fluorescent derivative of paclitaxel (Taxol) with microtubules and tubulin-colchicine

A fluorescent derivative of paclitaxel, 2-debenzoyl-2-(m-aminobenzoyl)paclitaxel (2-AB-PT), has been prepared. 2-AB-PT induces microtubule assembly in vitro, but is about 3-fold less potent than paclitaxel itself. The absorption and emission characteristics of 2-AB-PT were analyzed as a function of solvent. It was found that both spectra were perturbed by specific solvent effects when the solvent contained a hydrogen bond donor. The absorption and fluorescence spectra of 2-AB-PT bound to microtubules could not be mimicked by a single solvent, but the absorption and emission maxima of the tubulin-bound species could be duplicated by a solvent mixture of DMSO and water. These results indicate that the fluorophore binding site on the microtubule is in an environment of intermediate polarity that is accessible to a hydrogen bond donor in the vicinity of the m-amino group. In addition, tubulin fluorescence is quenched in the 2-AB-PT/microtubule complex, and energy transfer from tubulin to 2-AB-PT is apparent. These results indicate that substituents on the C-2 position of paclitaxel associate with tubulin when bound to the microtubule. 2-AB-PT binding to microtubules was quantitatively analyzed by fluorescence titrations. Two classes of binding sites for 2-AB-PT on microtubules were found. The high affinity site has an apparent association constant (K1app) of 2.0 (+/- 0.9) x 10(7) M-1 and an apparent binding stoichiometry (n1app) of 0.8 (+/- 0.1) sites/tubulin dimer in the microtubule. The apparent association constant for the lower affinity site is about 100-fold less than that of the higher affinity site (K2app = 2.1 (+/- 0.7) x 10(5) M-1), and the stoichiometry of the lower affinity site or class of sites (n2app) was found to be 1.3 +/- 0.1. Paclitaxel blocked 2-AB-PT binding to the high affinity site. No binding of 2-AB-PT to unassembled tubulin was observed, but the emission spectrum of 2-AB-PT in the presence of the tubulin-colchicine complex resembled the emission spectrum of the ligand bound to microtubules. It was previously shown that paclitaxel can induce GTPase activity in the tubulin-colchicine complex, indicating that paclitaxel can bind to unassembled tubulin in its complex with colchicine [Carlier, M.-F., & Pantaloni, D. (1983) Biochemistry 22, 4814-4822]. Rigorous characterization of the aggregation state of the protein under these conditions demonstrates that 2-AB-PT is also capable of binding to the tubulin-colchicine complex.

A convenient tubulin-based quantitative assay for paclitaxel (Taxol) derivatives more effective in inducing assembly than the parent compound

A room temperature biochemical assay, based on centrifugal removal of tubulin polymer, was developed to permit ready detection of paclitaxel analogs more active than the parent compound and to permit reliable quantification of differences in activity relative to paclitaxel in terms of drug concentration. The assay was validated by comparing paclitaxel to two compounds (docetaxel and 2-debenzoyl-2-meta-azidobenzoylpaclitaxel) known to be more active under multiple reaction conditions. The assay was designed to yield a relatively high EC50 (23 microM) for paclitaxel. This was possible because paclitaxel only weakly induced tubulin assembly at room temperature in 0.4 M glutamate without exogenous GTP. Under these same reaction conditions 50% assembly occurred with 4.7 microM 2-debenzoyl-2-meta-azidobenzoylpaclitaxel and 11 microM docetaxel. These biochemical EC50 values were in agreement with the relative cytotoxicity of the three compounds for human Burkitt lymphoma CA46 cells (IC50 values for paclitaxel, docetaxel, and 2-debenzoyl-2-meta-azidobenzoylpaclitaxel were 40, 10, and 3 nM, respectively).

Characterization of the taxol binding site on the microtubule. 2-(m-Azidobenzoyl)taxol photolabels a peptide (amino acids 217-231) of beta-tubulin

Photoaffinity labeling methods are being used to define the molecular contacts between taxol and its target protein, tubulin. Our laboratory has demonstrated previously that [3H]3'-(p-azidobenzamido)taxol photolabels the N-terminal 31 amino acids of beta-tubulin (Rao, S., Krauss, N.E., Heerding, J.M., Swindell, C.S., Ringel, I., Orr, G.A., and Horwitz, S.B. (1994) J. Biol. Chem. 269, 3132-3134). The interaction of a second photoaffinity analogue of taxol, [3H]2-(m-azidobenzoyl)taxol, with tubulin has been investigated. This analogue specifically photolabels beta-tubulin and the photolabeling is completed by both taxol and unlabeled 2-(m-azidobenzoyl)-taxol indicating a common binding domain. To identify the site(s) of photoincorporation, [3H]2-(m-azidobenzoyl)taxol-photolabeled beta-tubulin was subjected to sequential cyanogen bromide and tryptic digestions. Radiolabeled peptides were purified by reverse phase high performance liquid chromatography, and amino acid sequencing studies identified a peptide containing amino acid residues 217-231 of beta-tubulin as the major photolabeled domain.

Differential effects of paclitaxel (Taxol) analogs modified at positions C-2, C-7, and C-3' on tubulin polymerization and polymer stabilization: identification of a hyperactive paclitaxel derivative

Our finding that an analog of paclitaxel (Taxol) modified at position C-2 (2-debenzoyl-2-(m-azidobenzoyl)paclitaxel) was substantially more active than paclitaxel in promoting tubulin assembly [Chaudhary et al. (1994) J. Am. Chem. Soc. 116, 4097-4098] led us to perform an analysis of the modulating effects of microtubule-associated proteins, GTP, and temperature on assembly and polymer stability. The analog always showed superior activity to paclitaxel in inducing polymerization where it fails to occur without drug, probably indicating a greater ability than paclitaxel to "hypernucleate" assembly. In contrast, much smaller differences in effects on polymer stability were observed. The analysis was extended to a large series of derivatives modified at positions C-2, C-7, C-10, and C-3', including docetaxel, a clinically important analog of paclitaxel. While analog stabilization of polymer was frequently observed, neither qualitative nor quantitative analysis of this property reliable predicted whether a compound would have enhanced hypernucleation activity relative to that of paclitaxel. Stabilization was often observed at substoichiometric analog concentrations, while even superstoichiometric concentrations of most compounds failed to induce extensive tubulin polymerization at low temperatures or in the absence of microtubule-associated proteins or GTP. Docetaxel was intermediate in activity between paclitaxel and 2-debenzoyl-2-(m-azidobenzoyl)paclitaxel in promoting assembly reactions. We conclude that the hypernucleation of tubulin assembly and polymer stabilization observed with paclitaxel represent two distinct properties of the drug. Our findings suggest that paclitaxel, docetaxel, and 2-debenzoyl-2-(m-azidobenzoyl)paclitaxel are able to interact with progressively smaller assemblages of tubulin at low temperatures or in the absence of microtubule-associated proteins or GTP.

Modified taxols, 9. Synthesis and biological evaluation of 7-substituted photoaffinity analogues of taxol

The 7-substituted taxol analogues 7, 19, 27, and 32 have been prepared as potential photoaffinity-labeled derivatives for studies of the nature of the binding site of taxol on polymerized tubulin. The analogue 32 has been prepared in both deuterium- and tritium-labeled versions. Tubulin-assembly studies were carried out with these compounds, and it was found that they showed some but not all of the properties of taxol. We conclude that these specific taxol analogues labeled at the 7 position are not ideal derivatives for photoaffinity labeling studies.