We present a new approach to carbonic anhydrase II (CA II) inhibitor design that enables close interrogation of the regions of the CA active site where there is the greatest variability in amino acid residues among the different CA isozymes. By appending dual tail groups onto the par excellence CA inhibitor acetazolamide, compounds that may interact with the distinct hydrophobic and hydrophilic halves of the CA II active site were prepared. The dual-tail combinations selected included (i) two hydrophobic moieties, (ii) two hydrophilic moieties, and (iii) one hydrophobic and one hydrophilic moiety. The CA enzyme inhibition profile as well as the protein X-ray crystal structure of compound 3, comprising one hydrophobic and one hydrophilic tail moiety, in complex with CA II is described. This novel dual-tail approach has provided an enhanced opportunity to more fully exploit interactions with the CA active site by enabling these molecules to interact with the distinct halves of the active site. In addition to the dual-tail compounds, a corresponding set of single-tail derivatives was synthesized, enabling a comparative analysis of the single-tail versus dual-tail compound CA inhibition profile.

Diversely functionalized, fused aryl-alkyl ring systems hold a prominent position as well-established molecular frameworks for a variety of anti-cancer agents. The benzosuberene (6,7 fused, also referred to as dihydro-5H-benzo[7]annulene and benzocycloheptene) ring system has emerged as a valuable molecular core component for the development of inhibitors of tubulin assembly, which function as antiproliferative anti-cancer agents and, in certain cases, as vascular disrupting agents (VDAs). Both a phenolic-based analogue (known as KGP18, compound 39) and its corresponding amine-based congener (referred to as KGP156, compound 45), which demonstrate strong inhibition of tubulin assembly (low micromolar range) and potent cytotoxicity (picomolar range for KGP18 and nanomolar range for KGP156) are noteworthy examples of such benzosuberene-based compounds. In order to extend the structure-activity relationship (SAR) knowledge base related to benzosuberene anti-cancer agents, a series of eleven analogues (including KGP18) were prepared in which the methoxylation pattern on the pendant aryl ring as well as functional group incorporation on the fused aryl ring were varied. The synthetic approach to these compounds featured a sequential Wittig olefination, reduction, Eaton's reagent-mediated cyclization strategy to achieve the core benzosuberone intermediate, and represented a higher-yielding synthesis of KGP18 (which we prepared previously through a ring-expansion strategy). Incorporation of a fluorine or chlorine atom at the 1-position of the fused aryl ring or replacement of one of the methoxy groups with hydrogen (on the pendant aryl ring of KGP18) led to benzosuberene analogues that were both strongly inhibitory against tubulin assembly (IC50 approximately 1.0 M) and strongly cytotoxic against selected human cancer cell lines (for example, GI50 = 5.47 nM against NCI-H460 cells with fluorobenzosuberene analogue 37). A water-soluble phosphate prodrug salt of KGP18 (referred to as KGP265, compound 44) and a water-soluble serinamide salt (compound 48) of KGP156 were also synthesized and evaluated in this study.

No abstract

The discovery of 3-methoxy-9-(3′,4′,5′-trimethoxyphenyl)-6,7-dihydro-5H-benzo[7]annulen-4-ol (a benzosuberene-based analogue referred to as KGP18) was originally inspired by the natural products colchicine and combretastatin A-4 (CA4). The relative structural simplicity and ease of synthesis of KGP18, coupled with its potent biological activity as an inhibitor of tubulin polymerization and its cytotoxicity (in vitro) against human cancer cell lines, has resulted in studies focused on new analogue design and synthesis. Our goal was to probe the relationship of structure to function in this class of anticancer agents. A series of twenty-two new benzosuberene-based analogues of KGP18 was designed and synthesized. These compounds vary in their methoxylation pattern and separately incorporate trifluoromethyl groups around the pendant aryl ring for the evaluation of the effect of functional group modifications on the fused six-membered aromatic ring. In addition, the 8,9-saturated congener of KGP18 has been synthesized to assess the necessity of unsaturation at the carbon atom bearing the pendant aryl ring. Six of the molecules from this benzosuberene-series of compounds were active (IC50 < 5 μM) as inhibitors of tubulin polymerization while four analogues were comparable (IC50 approximately 1 μM) in their tubulin inhibitory activity to CA4 and KGP18. The potency of a bis-trifluoromethyl analogue 74 and the unsaturated KGP18 derivative 73 as inhibitors of tubulin assembly along with their moderate cytotoxicity suggested the potential utility of these compounds as vascular disrupting agents (VDAs) to selectively target microvessels feeding tumors. Accordingly, water-soluble and DMSO-soluble phosphate prodrug salts of each were synthesized for preliminary in vivo studies to assess their potential efficacy as VDAs.

The natural products combretastatin A-4 (CA4) and combretastatin A-1 (CA1) are potent cancer vascular disrupting agents (VDAs) and inhibitors of tubulin assembly (IC50 = 1–2 μM). The phosphorylated prodrugs CA4P and CA1P are undergoing human clinical trials against cancer. CA1 is unique due to its incorporation of a vicinal phenol, which has afforded the opportunity to prepare both diphosphate and regioisomeric monophosphate derivatives. Here, we describe the first synthetic routes suitable for the regiospecific preparation of the CA1-monophosphates, CA1MPA (8a/b) and CA1MPB (4a/b). The essential regiochemistry necessary to distinguish between the two vicinal phenolic groups was accomplished with a tosyl protecting group strategy. Each of the four monophosphate analogues (including Z and E isomers) demonstrated in vitro cytotoxicity against selected human cancer cell lines comparable to their corresponding diphosphate congeners. Furthermore, Z-CA1MPA (8a) and Z-CA1MPB (4a) were inactive as inhibitors of tubulin assembly (IC50 > 40 μM), as anticipated in this pure protein assay.