Search

A series of 6-demethylmitomycins and 6-demethyl-6-halomitomycins having various mitomycin skeletons were synthesized, taking into account the electronic effect toward the quinone moiety and the partition coefficients. Treatment of enones 15 and 16 with selenenamide or N-halosuccinimide-Et2NH afforded the 6-demethyl intermediates 17, 18, and 21-24 via the tandem Michael addition/retro-Mannich reaction sequence. Subsequent conversions into the mitomycin skeletons resulted in the formation of the desired derivatives 7a-c, 8a-c, 11a-c, and 12a,b. These mitomycin derivatives including 3a-c and 4a-c were evaluated for their anticellular activity against HeLa S3 cells and antitumor activity against Sarcoma 180 in mice. The anticellular activity of 1 and 3a-c depends on the substituent at the C-6 position and the order of increasing activity is H < CH3 < Br < Cl. A similar tendency was observed in their antitumor potency (ED50). The activities of 9 and 11a-c also follow a pattern similar to that of 1 and 3a-c. Compounds 4b,c, 8b,c, and 12b having both a halogen at the C-6 position and a methoxy group at the C-7 position did not show the activities because of the instability of the compounds. Interestingly, a correlation between the anticellular activity (IC50) and the partition coefficients (log kappa') determined by HPLC was observed within the compounds studied except the unstable compounds, while their antitumor activity (ED50 or T/C) did not correlate with the quinone reduction potential (E1/2). These results would indicate the importance of the C-6 substituents and the mitomycin skeletons for exhibiting both anticellular and antitumor activities.

A series of 4-hydroxy-3-quinolinecarboxylic acid derivatives (6) and 4-hydroxy-2-oxo-1,2-dihydro-3-quinolinecarboxylic acid derivatives (7) were designed and synthesized as 5-HT3 receptor antagonists. Molecular modeling studies suggested that the 3-carbonyl moiety in 6 was almost coplanar to the plane of an aromatic ring, but in 7 there was a 30 degrees deviation. 4-Hydroxy substitution in quinoline derivatives enhanced affinity for the 5-HT3 receptors, and endo-N-(8-methyl-8-azabicyclo[3.2.1]oct-3-yl)-4-hydroxy-3- quinolinecarboxamide (6f) exhibited the most potent activity in the Bezold-Jarisch (B-J) reflex test (ED50 = 0.1 micrograms/kg, iv) among quinoline derivatives 6. Although 4-hydroxy-2-oxo-1,2-dihydro-3-quinolinecarboxamide derivatives (7a) exhibited higher affinity (e.g., 7d: Ki = 0.48 nM) for the 5-HT3 receptors than ondansetron (Ki = 7.6 nM) or granisetron (Ki = 2.1 nM), these amides showed less potent activity in the B-J reflex test than the reference compounds. Interestingly, the ester derivatives 7c, 7f, and 7h eliminated affinity for the 5-HT3 receptors. These unusual structure-activity relationships and the deviation of the 3-carbonyl moiety from the plane of an aromatic ring suggest that the active conformation of 7a might be different from the proposed one for the preceding 5-HT3 antagonists. Thus, 6f was chosen for further studies. No receptor binding for a variety of ligands was significantly antagonized by 6f. Comparing the ratios of the ED50 value in the B-J reflex test (rat, iv) with the LD50 value in acute lethal toxicity (mouse, iv), 6f was proved to have a 600-fold wider margin of safety than ondansetron. Compound 6f dose-dependently attenuated both the incidence and frequency of emetic episodes induced by cisplatin in the dog (ED50 = 14 micrograms/kg, iv) more potently than ondansetron (ED50 = 210 micrograms/kg, iv). Compound 6f (KF-20170) is now under further investigation as a drug for treating gastrointestinal disorder.

After inspection of the X-ray crystallographic analyses of mitomycin A (1a) and albomitomycin A (2), an efficient chemical conversion of 1a to isomitomycin A (3) via 2 was accomplished in 68% overall yield. First, 1a was converted to (8aS)-8a-bromoalbomitomycin A (4b) in 89% yield, which was reduced with tributyltin hydride and AIBN to give 2 in 85% yield. Second, the C4a-N4 bond of 2 was selectively cleaved under hydrogenolysis conditions, and subsequent air oxidation afforded 3 in 90% yield from 2. Treatment of 2 or 3 with NH 3 in methanol afforded mitomycin C (1b) in good yield

A series of C-6-substituted methyl mitomycins was synthesized and evaluated for anticellular and antitumor activities. These novel compounds were prepared by Michael addition of various alcohols or thiols to 6-demethyl-7,7-(ethylenedioxy)-6,7-dihydro-6-methylidenemitosanes followed by treatment with NH3 or MeOH/K2CO3. Most compounds were potent against HeLa S3, and some of them showed superior activity to that of mitomycin C (MMC) against P388 leukemia and sarcoma 180 in mice. In addition, some compounds exhibited remarkable activity against MMC-resistant P388 in mice. FAB-MS spectra of these mitomycin derivatives showed the elimination of the C-6-methyl substituents from the mitomycin skeletons to form quinonemethides. Interestingly, treatment of 6-demethyl-6-[[(2-pyrimidinyl)thio]methyl ]mitomycin C (12v) with diethylamine afforded 6-demethyl-6-[(diethylamino)methyl]mitomycin C (31) in good yield. These results suggested that the C-6-substituted methyl mitomycins would have different biological character from that of MMC.

A series of mitomycin derivatives 1-3 having unique condensed-ring structures was synthesized and evaluated for their anticellular and antitumor activity. These compounds were synthesized by the Michael addition of 1, 3-dicarbonyl compounds to 6-demethyl-7, 7-(ethylenedioxy)-6, 7-dihydro-6-methylenemitosanes (4-6, and 14) and the subsequent cyclization. For the preparation of 1, the allyloxycarbonyl (Aloe) group was employable for the protection of the aziridine (1a-N-H), since the deprotection proceeded without decomposition of the substrates under the mild conditions with Pd(0) and HCO2H-NEt3. Among these structurally unique derivatives, compounds 1a, 1b, 1d and 1e were quite potent against HeLa S3 human tumor cells and sarcoma 180 solid tumor in mice.

A series of 6-demethylmitomycins and 6-demethyl-6-halomitomycins having various mitomycin skeletons were synthesized, taking into account the electronic effect toward the quinone moiety and the partition coefficients. Treatment of enones 15 and 16 with selenenamide or N-halosuccinimide-Et2NH afforded the 6-demethyl intermediates 17, 18, and 21-24 via the tandem Michael addition/retro-Mannich reaction sequence. Subsequent conversions into the mitomycin skeletons resulted in the formation of the desired derivatives 7a-c, 8a-c, 11a-c, and 12a,b. These mitomycin derivatives including 3a-c and 4a-c were evaluated for their anticellular activity against HeLa S3 cells and antitumor activity against Sarcoma 180 in mice. The anticellular activity of 1 and 3a-c depends on the substituent at the C-6 position and the order of increasing activity is H < CH3 < Br < Cl. A similar tendency was observed in their antitumor potency (ED50). The activities of 9 and 11a-c also follow a pattern similar to that of 1 and 3a-c. Compounds 4b,c, 8b,c, and 12b having both a halogen at the C-6 position and a methoxy group at the C-7 position did not show the activities because of the instability of the compounds. Interestingly, a correlation between the anticellular activity (IC50) and the partition coefficients (log kappa') determined by HPLC was observed within the compounds studied except the unstable compounds, while their antitumor activity (ED50 or T/C) did not correlate with the quinone reduction potential (E1/2). These results would indicate the importance of the C-6 substituents and the mitomycin skeletons for exhibiting both anticellular and antitumor activities.

A series of esters and amides of 1-alkyl-2-oxo-1,2-dihydroquinoline-4- carboxylic acid or 2-alkoxy-quinoline-4-carboxylic acid containing a basic azabicycloalkyl moiety has been synthesized and evaluated for affinity for the [3H]quipazine-labeled 5-HT3 receptors. Most of the esters exhibited 10-fold more potent activity than that of ondansetron (1; Ki = 7.6 nM). Lipophilic substituents at the 1- or 2-position of the quinoline ring enhanced affinity for the receptors. Compounds 21 and 37 showed the highest affinity (Ki = 0.32 and 0.31 nM, respectively) among them. On the other hand, most of the amides showed 100-fold lower affinity than that of the esters. Molecular modeling studies indicated that the carbonyl moiety in 19 (ester) or 31 (amide) was not coplanar to the plane of an aromatic ring (over 20 degrees deviation). Although some of the selected compounds exhibited potent activity in the Bezold-Jarisch (B-J) reflex test, good correlation was not observed between the affinity for the 5-HT3 receptors and the activity in the B-J reflex test (in vivo). From these data, it was suggested that our quinoline derivatives might interact with the 5-HT3 receptors in a different way from that of the reported 5-HT3 receptor antagonists presumably due to the presence of the heterogeneity of the 5-HT3 receptors between brain and heart.

The first example of C-3 alkoxylation in mitomycins has been achieved. 3 alpha-iso-Propoxy-10-O-decarbamoylmitomycin D (4) and 3 alpha-iso-propoxymitomycin D (5) were derived from mitomycin D (3) under decarbamoylation conditions with iso-propoxide. Under similar conditions 3 alpha-iso-propoxy-10-O-decarbamoylporfiromycin (8) and 3 alpha-methoxy-10-O-decarbamoylmitomycin B (11) were also derived from porfiromycin (6) and mitomycin B (9), respectively. The mechanism of generation of these novel analogs was based on the premise that the key intermediate of hydroquinone iminium salt (14) was led through the iminium salt (13), followed by alkoxide addition and oxidation.

Mitomycin G (2) was derived from porfiromycin (10b) in 3 steps via the methanesulfonate (14b) in an overall yield of 39%. On the basis of the established method for the introduction of an exomethylene group in mitomycins with a 9a-methoxy group, the preparation of biologically more important 1a-demethylmitomycin G (5) from mitomycin C (1) was accomplished by the use of a protective acetyl group on the aziridine in an overall yield of 57%. 1a-Demethylmitomycin K (6) was obtained from 5 in a yield of 42%. In a preliminary evaluation of their antitumor activity, compound 5 showed superior activity against sarcoma 180 (sc-ip) to its 1a-methyl congener, i.e., mitomycin G (2).

The histamine H1 receptor has an aspartate (Asp) residue in transmembrane helix 3 (TM3), which is well-conserved among biogenic amine receptors. The Asp residue is one of the most crucial amino acids for ligand binding. The tested histamine H1 receptor antagonists with tri- and tetracyclic structures were not selective for histamine H1 receptors and showed affinity for several other biogenic amine receptors. In contrast, KW-4679 ((Z)-11-[3-(dimethylamino)propylidene]-6,11-dihydrodibenz[b, e]oxepin-2-acetic acid hydrochloride), a tricyclic compound, was a selective histamine H1 receptor antagonist. [3H]KW-4679 had high affinity (Kd value of 2.5 +/- 0.12 nM) for wild-type human histamine H1 receptors. In the [3H]KW-4679 binding assay, replacement of Asp107 by alanine by site-directed mutagenesis greatly reduced the affinities (280-2100-fold) of tri- and tetracyclic compounds, whereas this mutation led to a comparatively small reduction (14-fold) in KW-4679 affinity. These results demonstrate that the tested tri- and tetracyclic histamine H1 receptor antagonists which have a tight interaction with the Asp residue are not selective for the histamine H1 receptor. Furthermore, the high selectivity of KW-4679 might be explained by a unique binding pocket, which consists of the Asp residue and other acceptor sites, in the histamine H1 receptor.

The search for 3-D requirements for the adenosine A1 receptor affinity is useful to aid in the design of more potent and/or novel ligands as pharmacological tools and therapeutics for the receptor. To emboss 3-D requirements for adenosine A1 receptor affinity among adenosine receptor antagonists, adenosine and xanthine analogs, conformations for the N6-substituted adenosine analogues and related adenosine A1 receptor antagonists were thoroughly searched by semi-empirical quantum mechanics calculations. Newly established global minima for these compounds (C1'-N6-C6-N1 torsion: 10 degrees) are consistent with retrieved structures from the Cambridge Structural Database and previously published NMR data on the solution conformation of N6-substituted adenosine analogues. However, these newly studied global minima for adenosine analogues are found to be different from those previously reported (C1'-N6-C6-N1 torsion: +/-75 degrees).

The three-dimensional (3-D) requirements for A1 adenosine receptor affinity have been studied based on hydrogen-bonding functionality correlation between a group of twelve A1 adenosine receptor ligands representing ten structurally different classes of compounds. Electrostatic potential similarity indices and shape similarity indices strongly support the proposed receptor-bound orientations of the ligands. We conclude, in areas common to both agonist and antagonist binding at the A1 receptor, that the ligands are recognized by a similar physicochemical 3-D environment. The finding of similar 3-D requirements for agonists and antagonists suggests a fairly static receptor structure in the region common to agonist and antagonist binding. The ribose moiety is remote from antagonist binding site. Such a 3-D environment rationalizes the binding of a number of potent novel antagonists including KW-3902, not previously reported in modeling studies.

A series of C-6-substituted methyl mitomycins was synthesized and evaluated for anticellular and antitumor activities. These novel compounds were prepared by Michael addition of various alcohols or thiols to 6-demethyl-7,7-(ethylenedioxy)-6,7-dihydro-6-methylidenemitosanes followed by treatment with NH3 or MeOH/K2CO3. Most compounds were potent against HeLa S3, and some of them showed superior activity to that of mitomycin C (MMC) against P388 leukemia and sarcoma 180 in mice. In addition, some compounds exhibited remarkable activity against MMC-resistant P388 in mice. FAB-MS spectra of these mitomycin derivatives showed the elimination of the C-6-methyl substituents from the mitomycin skeletons to form quinonemethides. Interestingly, treatment of 6-demethyl-6-[[(2-pyrimidinyl)thio]methyl ]mitomycin C (12v) with diethylamine afforded 6-demethyl-6-[(diethylamino)methyl]mitomycin C (31) in good yield. These results suggested that the C-6-substituted methyl mitomycins would have different biological character from that of MMC.

The reaction between mitomycin A (1) and cysteamine afforded 7-N,7'-N'-bis(2-thioethyl)dimitomycin C (7), 7-N-[2-[(2-aminoethyl)dithio]ethyl]mitomycin C (8), and 7-methoxy mitosenes (10, 11). The structures of 7 and 8 were elucidated on the basis of spectroscopy and reactions between 1 and 8, and 1 and cystamine. The observation of the time course for the reaction revealed the mechanism of the formation of 7 and 8. The rapid oxidation of cysteamine by the quinone of 1 gave cystamine, which was trapped by 1 to give 8, and 8 was additionally reacted with 1 to give 7. Since 7 showed significant antitumor activities, related 7-N,7'-N'-bis(omega-thioalkyl)dimitomycins were synthesized. They also showed remarkable antitumor activities against HeLa-S3 in vitro, sarcoma 180 (sc-ip), leukemia P388 (ip-ip) in vivo. In these evaluations, compound 7 demonstrated unique potency.

Barriers to exchange between dl and meso forms via rotation about the C9–Csubst bond of 9-(arylmethyl)-1,4-dimethoxytriptycenes are obtained by the use of the methoxyl signals in the 1H NMR spectra. It is found that, although the population ratios are affected by the substituent in the 9 position, the enthalpy of activation for the rotation is almost invariant at ca. 10 kcal/mol.

Benzyl methylene signals, which split into an AB quartet and a singlet and show temperature dependence, are obtained at various temperatures and are simulated with the use of two independent rates of rotation. Certain limitations of accuracy are found in this method. The uncertainty becomes less if additional data are provided from other sources, but never becomes negligible. Contribution of the slower process is found difficult to detect especially when the population involved in the process are the less.