Halichondrin B and homohalichondrin B, marine natural products binding in the vinca domain of tubulin. Discovery of tubulin-based mechanism of action by analysis of differential cytotoxicity data

Data generated in the new National Cancer Institute drug evaluation program, which is based on inhibition of cell growth in 60 human tumor cell lines, were used to compare new compounds with agents of known mechanism of action in terms of their differential cytotoxicity. Two marine natural products, halichondrin B and homohalichondrin B, appeared repeatedly when the data base was probed with known antimitotic agents. We confirmed that both compounds were highly cytotoxic (IC50 values for L1210 murine leukemia cells of 0.3 and 1 nM, respectively), with accumulation of cells arrested in mitosis at toxic concentrations, that both inhibited the polymerization of purified tubulin, and that both inhibited microtubule assembly dependent on microtubule-associated proteins. Limited amounts of homohalichondrin B, the less active agent, were available, so only halichondrin B was studied in detail. Halichondrin B did not interfere with colchicine binding to tubulin, but it was a noncompetitive inhibitor of the binding of vinblastine to tubulin (apparent Ki, 5.0 microM). Halichondrin B was therefore compared with other agents which interfere with the binding of vinca alkaloids to tubulin (vinblastine, maytansine, dolastatin 10, phomopsin A, rhizoxin) in terms of its effects on tubulin polymerization, inhibition of GTP hydrolysis, inhibition of nucleotide exchange, and stabilization of tubulin, as well as the quantitative assessment of its effects on vinca alkaloid binding and inhibition of cell growth. Since halichondrin B was originally isolated from the same organism as the phosphatase inhibitor okadaic acid, and since it is about 50-fold more effective than okadaic acid as an inhibitor of L1210 cell growth, perturbations of cellular microtubules observed following treatment with okadaic acid should be interpreted cautiously.     

Temperature dependent binding of mebendazole to tubulin in benzimidazole-susceptible and -resistant strains of Trichostrongylus colubriformis and Caenorhabditis elegans

The binding of [³H]mebendazole ([³H]MBZ) to tubulin in benzimidazole-susceptible (BZ-S) and benzimidazole-resistant (BZ-R) strains of Trichostrongylus colubriformis and Caenorhabditis elegans was examined in order to investigate the biochemical changes to tubulin that result in BZ resistance in parasitic and free-living nematodes. In both species the extent of [³H]MBZ binding to tubulin was significantly reduced in the BZ-R strain compared with the BZ-S strain. The decrease in [³H]MBZ binding in the BZ-R strain of each species was the result of a significant reduction in the amount of charcoal stable [³H]MBZ-tubulin complexes and was not related to a change in the association constant of the [³H]MBZ-tubulin interaction. [³H]MBZ binding to tubulin was temperature dependent, reaching maximum levels at 37°C in BZ-S T. colubriformis and 10°C in BZ-R T. colubriformis. Both the BZ-S and BZ-R strains of C. elegans displayed maximum [³H]MBZ binding at 4°C. Resistance ratios derived from the amount of [³H]MBZ binding in the BZ-S and BZ-R strains and in vitro development assays demonstrated that the temperature dependence and extent of drug binding was indicative of BZ resistance status and was species specific in the BZ-S isolates. These results indicate that biochemical differences exist in the binding of benzimidazole carbamates to tubulin in nematode species, and suggest that the susceptibility of the parasitic nematodes to the benzimidazole anthelmintics is the result of a unique high affinity and/or high capacity interaction ofbenzimidazole carbamates with tubulin.     

Electrogenic properties of the cloned Na+/glucose cotransporter: I. Voltage-clamp studies

Summary Cystic fibrosis (CF) is characterized by abnormal epithelial Cl^– conductance (G_Cl). In vitro studies that have shown that cAMP regulation is an intrinsic property of the CF-affected G_Cl(CF-G_Cl) have been carried out previously on cultured secretory cells and on nonepithelial cells. Even though G_Cl in absorption is defective in CF, a clear demonstration of cAMP regulation of CF-G_Cl in a purely absorptive tissue is lacking. We studied the cAMP regulation of CF-G_Cl in the microperfused intact human reabsorptive sweat duct. About 40% of the ducts responded to cAMP (responsive) while the remainder of the ducts did not. In responsive ducts, cAMP-elevating agents: -adrenergic agonist isoproterenol (IPR), CPT-cAMP, forskolin, theophylline or IBMX increased G _tby about 2.3-fold (n = no. of ducts = 8). Removal of media Cl^–, but not amiloride pretreatment (in the lumen), abolished the cAMP response, indicating exclusive activation of G_Cl. cAMP activated both apical and basolateral G_Cl. cAMP hyperpolarized gluconate: Cl^– (lumen: bath) transepithelial bionic potentials (V _t=–20.3±5.2 mV, mean ±se, n=9) and transepithelial 3 1 luminal NaCl dilution diffusion potentials (V _t=–8.8±2.9 mV, n=5). cAMP activated basolateral G_Cl as indicated by increased bi-ionic (gluconate: Cl^–, bath: lumen) diffusion potentials (by about 12 mV). The voltage divider ratio in symmetric NaCl solutions increased by 60%. Compared to responsive ducts, nonresponsive ducts were characterized by smaller spontaneous transepithelial potentials in symmetrical Ringer's solution (V _t=–6.9±0.8 mV, n=24, nonresponsive vs. –19.4±1.8 mV, n=22, responsive ducts) but larger bi-ionic potentials (–94±6 mV, n=35, nonresponsive vs. –65±5 mV, n=17, responsive ducts) and dilution diffusion potentials (–40±5 mV, n=11, nonresponsive vs. –29±3 mV, n=7, responsive ducts). These results are consistent with an inherently (prestimulus) maximal activation of G_Cl in nonresponsive ducts and submaximal activation of G_Cl in responsive ducts. We conclude that cAMP activates CF-G _Cl which is expressed and abnormal in both apical and basal membranes of this absorptive epithelium in CF.Abbreviations CF cystic fibrosis - G _t transepithelial conductance - V _b electrical potential across the basolateral membrane - V _a electrical potential across the apical membrane - V _t transepithelial potential - V _b transepithelial currentinduced voltage deflections across the basolateral membrane - V _a transepithelial current-induced voltage deflections across the apical membrane - V _t transepithelial current-induced voltage deflection across the epithelium - VDR voltage divider ratio - G_Cl transepithelial Cl^– conductance - CF-G_Cl cystic fibrosis-affected Cl^– conductance - EMF electromotive force - IPR isoproterenol - IBMX 3-isobutyl-1-methylxanthine - CPT-cAMP chlorophenylthio-adenosine 3-5 cyclic monophosphate - PGE₂ prostaglandin E₂     

Identification of the metal associated with the insulin degrading enzyme.

Insulin degrading enzyme (IDE) is a thiol-dependent metalloendoprotease that is responsible for initiation of cellular insulin degradation. However, its exact mode of action and the factors controlling it are poorly understood. Since IDE is a metal requiring enzyme, we have examined which metal(s) is(are) endogenously associated with it. Using neutron activation analysis, we studied the metal content of a partially purified enzyme from three different tissues: rat skeletal muscle, rat liver, and human placenta. Our results indicate that zinc and manganese are associated with the enzyme with approximately 10 times more zinc as manganese being present. These results suggest that one or both of these two metals are endogenously associated with this enzyme and are a means of controlling the enzyme's activity.     

Calmodulin structure and function: Implication of arginine in the compaction related to ligand binding

Calmodulin (CAM) is a modulatory protein that regulates cellular activity by binding to a large number of proteins. Key elements in the Ca²⁺-dependent mechanism of interaction between CAM and the proteins it activates are the selectivity for Ca²⁺ ions and the requirement for Ca²⁺-dependent conformational changes. We report on results from a series of molecular dynamics simulations that identified discrete steps in the mechanism of structural rearrangement of CAM. The findings implicate the side chains of arginine residues in the bending of the central alpha helix. Structural and energetic considerations point to a dynamic hydrogen bonding pattern around the arginine residues as a ratcheting-type mechanism, causing the kinking of the central helix in consecutive steps stabilized by each new pattern of hydrogen bonds. Initial model building studies to locate potential binding sites of ligands such as trifluoperazine (TFP) indicate that the compaction of CAM results in several structural changes, that explain the selective binding of molecules such as TFP in the N-terminal domain. The present studies identify specific residues involved in the process of compaction and point to specific CAM residues involved in the binding of the ligand. These insights lead directly to propositions for experimental engineering of the molecular structure of CAM in order to probe the hypotheses and their consequences for the function of this important protein.     

Human red blood cell insulin-degrading enzyme and rat skeletal muscle insulin protease share antigenic sites and generate identical products from insulin.

The mechanisms of cellular insulin degradation remain uncertain. Considerable evidence now exists that the primary cellular insulin-degrading activity is a metallothiol proteinase. Two similar degrading activities have been purified and characterized. Insulin protease has been purified from rat skeletal muscle and insulin-degrading enzyme from human red blood cells. Whereas the two degrading activities share a number of similar properties, significant differences have also been reported; and it is not at all established that they are the same enzyme. To examine this, we have compared antigenic and catalytic properties of the two enzymatic activities. Monoclonal antibodies against the red blood cell enzyme adsorb the skeletal muscle enzyme; and on Western blots, the antibodies react with an identical 110-kDa protein. Immunoaffinity-purified enzymes from both red blood cells and skeletal muscle degrade [125I]iodo(B26)insulin to the same products as seen with purified insulin protease and with intact liver and kidney. Chelator-treated muscle and red blood cell enzymes can be reactivated with either Mn2+ or Ca2+. Thus, insulin-degrading enzyme and insulin protease have similar properties. These results support the hypothesis that these activities reside in the same enzyme.     

Chemical properties of the N-termini of human haemoglobin.

The chemical properties, namely pK and reactivity, of the N-termini of oxyhaemoglobin and deoxyhaemoglobin toward acetic anhydride and 1-fluoro-2,4-dinitrobenzene (Dnp-F) were determined by the competitive-labelling approach [Kaplan, Stevenson & Hartley, (1971) Biochem. J. 124, 289-229; Duggleby & Kaplan (1975) Biochemistry 14, 5168-5175]. At physiological pH and temperature, the valine-1 alpha and valine-1-beta amino groups had unusually low pK values, but showed only minimal changes in their pK values on deoxygenation. Between pH 7.5 and pH 8.0 a deviation was observed in the pH-reactivity profiles and the apparent pK values became markedly pH-dependent. It was found that Dnp-F, but not acetic anhydride, had an abnormally high reactivity toward the N-termini. It is concluded that the valine-1 alpha and valine-1 beta N-termini make little or no contribution to the alkaline Bohr effect at physiological pH values. The high reactivity toward Dnp-F is attributed to an interaction or binding near the N-terminal region, and the discontinuity in the pH-reactivity profile at moderate alkaline pH values to a conformational change which alters the environment of these groups.     

MITOTIC ACTIVITY AT THE SHOOT APEX OF AZOLLA FILICULOIDES

The mosquito fern, Azolla filiculoides Lam., was grown in a growth chamber on a nitrogen-free culture solution at 24 C under the following photoperiod: 16 hr light/8 hr darkness. Shoot tips were fixed every 2 hr for 24 hr to determine the mitotic index for the apical cell, immediate derivatives, and remaining cells to the level of the first leaf or lateral shoot primordium. Mitotic indices were 6.9%, 6.5% and 6.3%, respectively. The colchicine method was employed to determine the cell-cycle durations and duration of mitosis for the same populations of cells. The cell-cycle duration and duration of mitosis of the apical cell were 28.2 hr and 2.8 hr, respectively; for the immediate derivatives, 26.7 hr and 2.5 hr; for the remaining cells, 23.6 hr and 2.1 hr. Conclusions: the apical cell is as mitotically active as its immediate derivatives, and there is no evidence of a quiescent apical cell.     

Differing mechanisms for leukotriene D4-induced bronchoconstriction in guinea pigs following intravenous and aerosol administration

Leukotriene D₄ (LTD₄) when administered intravenously or by aerosol to guinea pigs produced changes in pulmonary mechanics including a decrease in dynamic compliance and an increase in pulmonary resistance. The effects of intravenous LTD₄ (0.5 μg kg⁻¹) were short lived and abolished by pretreatment of the animal with either cyclooxygenase inhibitors, a thromboxane synthetase inhibitor (OKY 1555) or an SRS-A antagonist (FPL 55712). These findings suggest that bronchoconstriction produced by the intravenous infusion of LTD₄ at 0.5 μg kg⁻¹ is due to the release of thromboxane A₂. However, in animals treated with indomethacin, LTD₄ at higher doses (>0.8 μg kg⁻¹) still elicited a bronchoconstriction which could be blocked by FPL 557112. Nebulization of 0.1 – 1.0 μg of LTD₄ into the lung produced prolonged changes in pulmonary mechanics which were inhibited by FPL 55712 and were potentiated indomethacin. LTD₄, therefore, when administered by aerosol produced effects on the lung which were not mediated by cyclooxygenase products. Responses to nebulized rather than intravenous LTD₄ in the guinea pig may more closely resemble those seen in human tissues.     

Poliovirus RNA-Dependent RNA Polymerase Synthesizes Full-Length Copies of Poliovirion RNA, Cellular mRNA, and Several Plant Virus RNAs In Vitro

The poliovirus RNA-dependent RNA polymerase was active on synthetic homopolymeric RNA templates as well as on every natural RNA tested. The polymerase copied polyadenylate. oligouridylate [oligo(U)], polycytidylate . oligoinosinate, and polyinosinate. oligocytidylate templates to about the same extent. The observed activity on polyuridylate. oligoadenylate was about fourfold less. Full-length copies of both poliovirion RNA and a wide variety of other polyadenylated RNAs were synthesized by the polymerase in the presence of oligo(U). Polymerase elongation rates on poliovirion RNA and a heterologous RNA (squash mosaic virus RNA) were about the same. Changes in the Mg(2+) concentration affected the elongation rates on both RNAs to the same extent. With two non-polyadenylated RNAs (tobacco mosaic virus RNA and brome mosaic virus RNA3), the results were different. The purified polymerase synthesized a subgenomic-sized product RNA on brome mosaic virus RNA3 in the presence of oligo(U). This product RNA appeared to initiate on oligo(U) hybridized to an internal oligoadenylate sequence in brome mosaic virus RNA3. No oligo(U)-primed product was synthesized on tobacco mosaic virus RNA. When partially purified polymerase was used in place of the completely purified enzyme, some oligo(U)-independent activity was observed on the brome mosaic virus and tobacco mosaic virus RNAs. The size of the product RNA from these reactions suggested that at least some of the product RNA was full-sized and covalently linked to the template RNA. Thus, the polymerase was found to copy many different types of RNA and to make full-length copies of the RNAs tested.