Tricyclic pyrone analogs: a new synthetic class of bifunctional anticancer drugs that inhibit nucleoside transport, microtubule assembly, the viability of leukemic cells in vitro and the growth of solid tumors in vivo.

Tricyclic pyrones (TPs) may represent a novel synthetic class of microtubule (MT) de-stabilizing anticancer drugs previously shown by us to inhibit macromolecule synthesis, tubulin polymerization, and the proliferation of leukemic and mammary tumor cells in vitro. A linear skeleton with a N-containing aromatic ring attached at C3 of the top A-ring, a central pyran B-ring and a six-membered bottom C-ring with no alkylation at C7 are required for the antitumor activities of the lead compounds, a 3-pyridyl benzopyran (code name H10) and its somewhat weaker 2-pyridyl regioisomer (code name H19). Increasing concentrations of H10 do not alter the binding of [3H]vinblastine and [3H]GTP to tubulin but mimic the ability of unlabeled colchicine (CLC) to reduce the amount of [3H]CLC bound to tubulin, suggesting that TPs may interact with the CLC binding site to inhibit tubulin polymerization. Exogenous Mg2+ cations absolutely required for the binding of GTP to tubulin and MT assembly cannot overcome the antitubulin action of H10. H10 reduces the viability of L1210 cells in vitro (IC50: 0.5 microM) but its antitumor activity may be related to its ability to inhibit tubulin polymerization and rapidly increase the mitotic index rather than to induce DNA cleavage and apoptosis. The anticancer potential of TPs in vivo is demonstrated by the fact that i.p. injections of the water-soluble H10-HCl decrease the growth of solid tumors in mice inoculated s.c. with Lewis lung carcinoma. A critical finding is that the antimitotic H10 is a bifunctional anticancer drug, which also blocks the cellular transport of nucleosides (IC50: 6 microM) to inhibit DNA synthesis. Since few CLC site-binding antimitotic agents are active in solid tumor models in vivo, the ability of these new MT destabilizing TPs to totally block nucleoside transport might be valuable in polychemotherapy to arrest tumor cells at several phases of their cycle, potentiate the action of antimetabolites and sensitize multidrug-resistant tumor cells.

Microtubule-disrupting effects of gallium chloride in vitro.

Gallium chloride (GaCl3), an antitumor agent with antagonistic action on iron, magnesium and calcium, was tested for its ability to alter the polymerization of purified tubulin (2.2 mg/ml) in a cell-free system in vitro. GaCl3 (250 microM) does not mimic the effect of 10 microM paclitaxel and, therefore, is not a microtubule (MT)-stabilizing agent that can promote tubulin polymerization in the absence of glycerol and block MT disassembly. In contrast, GaCl3 mimics the effect of 1 microM vincristine (VCR) and inhibits glycerol-induced tubulin polymerization in a concentration-dependent manner (IC50: 125 microM), indicating that GaCl3 is a MT de-stabilizing agent that prevents MT assembly. However, 150 microM GaCl3 must be used to match or surpass the inhibitions of tubulin polymerization caused by 0.25 microM of known MT de-stabilizing agents, such as colchicine (CLC), nocodazole, podophyllotoxin, tubulozole-C and VCR. The inhibitory effect of 250 microM GaCl3 persists in the presence of up to 9 mM MgCl2, suggesting that the exogenous Mg2+ cations absolutely required for the binding of GTP to tubulin and MT assembly cannot overcome the antitubulin action of Ga3+ ions of a higher valence. The binding of [3H]vinblastine (VBL) to tubulin (0.5 mg/ml) is inhibited by unlabeled VBL but enhanced by concentrations of GaCl3 > 200 microM. However, increasing concentrations of GaCl3 mimic the ability of cold CLC to reduce the amount of [3H]CLC bound to tubulin, suggesting that GaCl3 may interact with the CLC binding site to inhibit tubulin polymerization. The binding of [3H]GTP to tubulin is decreased by unlabeled GTP but markedly enhanced by GaCl3, especially when concentrations of this metal salt of 32 microM or higher are added to the reaction mixture before rather than after the radiolabeled nucleotide. These data suggest that changes in protein conformation following GaCl3 binding might increase the interactions of tubulin with nucleotides and Vinca alkaloids. After a 24 h delay, the viability of GaCl3-treated L1210 leukemic cells is reduced in a concentration-dependent manner at days 2 (IC50: 175 microM), 3 (IC50: 35 microM) and 4 (IC50: 16 microM). Since GaCl3 (100-625 microM) increases the percentage of mitotic cells at 2-4 days, it might arrest tumor cell progression in M phase, but its antimitotic activity is much weaker than that of 0.25 microM VCR. Because the concentrations of GaCl3 that inhibit tubulin polymerization also increase the mitotic index and decrease the viability of L1210 cells in vitro, the antitubulin and antimitotic effects of GaCl3 might contribute, at least in part, to its antitumor activity.

Antitumor activity of novel octalactin A analogs in murine leukemia cells in vitro.

Octalactin A and B (code names K1 and K2) are eight-membered-ring lactones from a marine bacterium. K1 is reportedly cytotoxic. Since access to this natural product is severely limited, the entire synthesis of K1 has been achieved in K. Buszek's laboratory, and several of its structural and stereochemical analogs (code names K3-K9) have been tested for their ability to prevent murine L1210 leukemic cells from synthesizing macromolecules and growing in vitro. At 50 microM, K1 is inactive and the eight-membered lactone K4, an oxocene, is the only compound found to inhibit tumor cell growth by about 90% in the L1210 system. The long-term inhibition of L1210 cell growth by K4 is concentration dependent (IC50 around 10 microM) and not reversible following drug removal. The delayed and weaker cytotoxic effects of K4 suggest that the inhibition of tumor cell proliferation observed 1-4 days after K4 treatment is not solely caused by drug cytotoxicity. When compared to a spectrum of representative anticancer drugs, higher concentrations of K4 must be used to maximally inhibit tumor cell growth. In contrast to its antiproliferative activity, 50 microM K4 fails to alter the rates of DNA, RNA and protein synthesis in L1210 cells. This discrepancy between the ability of K4 to inhibit macromolecule synthesis and leukemic cell growth suggests that other molecular targets are involved in the antitumor action of this drug. At 50 microM, K4 inhibits the polymerization of purified tubulin by about 45%, and therefore may be a novel microtubule de-stabilizing drug weaker than vincristine. Even though other mechanisms may be involved in its antitumor action, the ability of K4 to partially disrupt microtubule dynamics indirectly suggests that this synthetic oxocene may be a cell cycle-specific anticancer drug that blocks mammalian cells in M-phase.

Tricyclic pyrone analogs: a new class of microtubule-disrupting anticancer drugs effective against murine leukemia cells in vitro.

Novel 1H,7H-5a,6,8,9-tetrahydro-1-oxopyrano [4,3-b][1]benzopyrans were synthesized in Hua's laboratory (code names H5, H10, H14 and H15) and tested for their ability to prevent L1210 leukemic cells from synthesizing macromolecules and growing in vitro. The aryl groups of these tricyclic pyrone (TP) analogs are either 3, 4-dimethoxyphenyl in H5 and H15 or 3-pyridyl in H10 and H14. Since 50 M H5 and H10 both inhibit DNA synthesis and tumor cell growth by 79-100%, concentrations 25 M were used in this study to assess the structure-activity relationships for this class of compounds. At 10-25 M, H5 and H14 are more potent inhibitors of DNA, RNA and protein synthesis than H10. In contrast, at 5-25 M, H10 is much more effective than H5 and H14 at inhibiting the growth of L1210 cells over a 4-day period. Interestingly, H15 inhibits DNA synthesis as much as H10 but fails to alter tumor cell growth. This discrepancy between the ability of TPs to inhibit macromolecule synthesis and leukemic cell growth suggests that other molecular targets may be involved in the antitumor action of these drugs. Their short-term inhibition of nucleic acid synthesis is reversible following drug removal but their long-term inhibition of tumor cell growth is not. Moreover, 25 M H5 and H10 are not cytotoxic at 2 days but equally decrease cell viability at 4 days, suggesting that the potent and irreversible inhibition of cell proliferation observed 1-4 days after H10 treatment is not solely caused by drug cytotoxicity. The effectiveness of H10 as inhibitor of L1210 cell growth is comparable to that of a spectrum of representative anticancer drugs. A critical finding is that 5 M H10 blocks the polymerization of purified tubulin by 90% and, therefore, may be a novel microtubule de-stabilizing drug. Indeed, H10 inhibits tubulin polymerization and L1210 cell growth as much as 5 M of vincristine (VCR). In contrast, 5 M H5 alters neither tubulin polymerization nor tumor cell growth. The ability of H10 to disrupt microtubule dynamics indirectly suggests that TPs may be novel cell cycle-specific anticancer drugs useful for arresting mammalian cells in mitosis.

Antitumor activity of tricyclic pyrone analogs, a new synthetic class of microtubule de-stabilizing agents, in the murine EMT-6 mammary tumor cell line in vitro.

Novel tricyclic pyrone (TP) analogs synthesized in Hua's laboratory (code names H10, H14 and H16) were tested against a spectrum of known antimitotic drugs for their ability to disrupt microtubule (MT) dynamics, alter the mitotic index, and prevent murine EMT-6 mammary sarcoma cells from synthesizing DNA and proliferating in vitro. At 2-10 microM, H10 inhibits DNA synthesis, tubulin polymerization and tumor cell growth to a greater degree than H14, whereas H16 has no effect. A linear skeleton with a pyridyl ring at C-3 of the A-ring, a pyran B-ring and no alkylation at C-7 of the C-ring is required for the antitumor activity of these TPs. Since H10 mimics the effect of vincristine (VCR), but not that of paclitaxel, on tubulin polymerization, TPs may represent a novel synthetic class of MT de-stabilizing anticancer drugs. H10 is less potent than VCR against tubulin polymerization (IC50: 1.5 microM versus 0.15 microM) and tumor cell proliferation (IC50: 1.5 microM versus 5 nM) but inhibits DNA synthesis (IC50: 10 microM) more effectively than all other MT-disrupting agents tested, except tubulozole-C. Although TPs disrupt DNA synthesis and might affect several phases of the cell cycle, the ability of H10 to increase the percentage of mitotic cells indicates that these novel compounds may be cell cycle-specific anticancer drugs useful for arresting mammalian cells in M-phase.

Antitumor activity of novel tricyclic pyrone analogs in murine leukemia cells in vitro.

New tricyclic pyrone derivatives were synthesized and tested for their ability to prevent L1210 leukemic cells from synthesizing DNA and growing in vitro. At 50 microM, a pyripyropene analog has no effect, whereas four pentahydro-3-aryl-1-oxopyrano[4,3-b][1]benzopyrans all inhibit DNA synthesis by 79-91% and tumor cell growth by 93-100%. These inhibitory effects are concentration dependent with IC50 around 8.5 microM for DNA synthesis at 2 hours and 1.1 microM for tumor cell growth at 4 days. The aryl groups of these antitumor agents are either 3,4-dimethoxyphenyl or 3-pyridyl. Introduction of a methyl group at C5a and a formyloxy or hydroxy group at C6 does not alter the antitumor effects of the 3,4-dimethoxyphenyl benzopyrans but reduces those of the 3-pyridyl benzopyrans, which, at 50 microM, inhibit DNA synthesis by only 32-49% and fail to alter tumor cell growth. The 4-hydroxy-6-(3-pyridyl)-2-pyrone has no effect and the tricyclic pyrones lacking aryl groups have very little inhibitory effects on DNA synthesis, suggesting that a greater conjugation is required for the antitumor activity. These molecules have never been reported and might be valuable to develop a new class of anticancer drugs.