A common pharmacophore for epothilone and taxanes: Molecular basis for drug resistance conferred by tubulin mutations in human cancer cells Paraskevi Giannakakou*†, Rick Gussio‡, Eva Nogales§, Kenneth H. Downing§, Daniel Zaharevitz‡, Birgit Bollbuck¶, George Poyʈ, Dan Sackett**, K. C. Nicolaou¶, and Tito Fojo* *Medicine Branch, National Cancer Institute, and ʈGenetics and Biochemistry Branch, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD 20892; ‡Target Structure-Based Drug Discovery Group, Information Technology Branch, and **Laboratory of Drug Discovery Research and Development, Developmental Therapeutics Program, National Cancer Institute, National Institutes of Health, Frederick, MD 21702; §Life Science Division, Lawrence Berkeley National Laboratory, and Molecular and Cell Biology Department, University of California, Berkeley, CA 94720; and ¶Department of Chemistry and The Skaggs Institute for Chemical Biology, The Scripps Research Institute, La Jolla, CA 92037 Contributed by Kyriacos C. Nicolaou, December 15, 1999 The epothilones are naturally occurring antimitotic drugs that The PTX-binding site on ␤-tubulin has been identified both share with the taxanes a similar mechanism of action without by electron crystallography, showing PTX bound on the of apparent structural similarity. Although photoaffinity labeling ␣␤-tubulin dimer (10), and by photoaffinity labeling, which has and electron crystallographic studies have identified the taxane- identified amino acids ␤1–31 and ␤217–233 as important areas binding site on ␤-tubulin, similar data are not available for for PTX binding (11, 12). Similar studies, however, for epothi- epothilones. To identify tubulin residues important for epothi- lones are not available. In this study, we present a report of lone binding, we have isolated two epothilone-resistant human epothilone-resistant (EpoR) human cancer cell lines with ovarian carcinoma sublines derived in a single-step selection acquired ␤-tubulin mutations. The residues involved, ␤274 and with epothilone A or B. These epothilone-resistant sublines ␤282, both map near the taxane-binding site in the atomic exhibit impaired epothilone- and taxane-driven tubulin poly- model of tubulin (10). These mutations affect the ability of merization caused by acquired ␤-tubulin mutations (␤274Thr3 Ile epothilones to induce tubulin polymerization as well as inhibit and ␤282Arg3 Gln) located in the atomic model of ␣␤-tubulin near cell growth. Using molecular modeling and guided by the the taxane-binding site. Using molecular modeling, we investi- mutation data and the activity profile of several MT-stabilizing gated the conformational behavior of epothilone, which led to agents against the cell lines with mutant tubulins, we were able the identification of a common pharmacophore shared by tax- to identify a common pharmacophore shared by taxanes and anes and epothilones. Although two binding modes for the epothilones and in turn model Epo binding onto tubulin. In epothilones were predicted, one mode was identified as the addition, we have identified an MT-active agent of the sarco- preferred epothilone conformation as indicated by the activity dictyin family which, in contrast to taxanes and epothilones, is of a potent pyridine-epothilone analogue. In addition, the preferentially active against the tubulin mutants, and we have structure–activity relationships of multiple taxanes and epothi- modeled the way it binds tubulin. lones in the tubulin mutant cells can be fully explained by the The unification of taxane, epothilone, and sarcodictyin chem- model presented here, verifying its predictive value. Finally, istries in a single pharmacophore, and the fact that their these pharmacophore and activity data from mutant cells were structure–activity relationship profile against both the parental used to model the tubulin binding of sarcodictyins, a distinct and mutant tubulins can be successfully explained by our mod- class of microtubule stabilizers, which in contrast to taxanes and eling and docking studies, provide a framework to study drug– the epothilones interact preferentially with the mutant tubulins. tubulin interactions that should assist in the rational design of The unification of taxane, epothilone, and sarcodictyin chemis- agents targeting tubulin. tries in a single pharmacophore provides a framework to study drug–tubulin interactions that should assist in the rational Materials and Methods design of agents targeting tubulin. Materials. PTX and docetaxel were obtained from the Drug Synthesis and Chemistry Branch of the National Cancer Insti- he clinical success of paclitaxel (PTX) and docetaxel has tute, Bethesda, MD. Epo A and B (13), Epo B pyridine †† Tstimulated a search for compounds with a similar mode of analogue, and sarcodictyin A analogue (14) were prepared by action and has resulted in the identification of three nontaxane chemical synthesis. chemical classes of natural products: the soil bacteria-derived epothilones A and B (Epo A and Epo B) (1), the marine sponge-derived discodermolide (2, 3), and the coral-derived Abbreviations: PTX, paclitaxel; MT, microtubule; Epo A, epothilone A; Epo B, epothilone B; eleutherobins͞sarcodictyins (4, 5). All three classes stabilize EpoR, epothilone-resistant. microtubules (MTs) and competitively inhibit the binding of †To whom reprint requests should be addressed at: Medicine Branch, NCI, NIH, Bldg. 10, PTX to tubulin polymers, indicating overlap of binding sites Room 12N226, 9000 Rockville Pike, Bethesda, MD 20892. E-mail: [email protected]. †† (1–3, 6, 7). The epothilones, however, display some superior The synthesis of this Epo B analogue will be reported elsewhere. qualities: they are water soluble, can be produced in large The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. quantities through bacterial fermentation, and retain activity §1734 solely to indicate this fact. against multidrug-resistant (MDR) cell lines and tumors Article published online before print: Proc. Natl. Acad. Sci. USA, 10.1073͞pnas.040546297. (5, 8, 9). Article and publication date are at www.pnas.org͞cgi͞doi͞10.1073͞pnas.040546297 2904–2909 ͉ PNAS ͉ March 14, 2000 ͉ vol. 97 ͉ no. 6 Downloaded by guest on September 25, 2021 Table 1. Cytotoxicity profile of 1A9͞EpoR cells to drugs acting on MTs IC50,* nM Drug 1A9 1A9͞A8 (␤274) 1A9͞B10 (␤282) PTX 2 Ϯ 0.23 19.5 Ϯ 1.14 (10) 13 Ϯ 0.58 (6.5) Docetaxel 1 Ϯ 0.15 7 Ϯ 0.45 (7) 5 Ϯ 0.25 (5) Baccatin 550 Ϯ 70 3950 Ϯ 1060 (7) 3000 Ϯ 870 (5.4) Epo A 3.2 Ϯ 0.42 127 Ϯ 14.7 (40) 182 Ϯ 6.25 (57) Epo B 0.22 Ϯ 0.03 5.4 Ϯ 0.64 (25) 5.2 Ϯ 0.35 (24) B-pyridine 0.1 Ϯ 0.015 3 Ϯ 0.25 (30) 6 Ϯ 0.2 (60) Sarc. A analogue† 3.75 Ϯ 0.02 1.18 Ϯ 0.12 (0.3)‡ 0.45 Ϯ 0.018 (0.12)‡ *IC50 values (ϮSEM) represent the mean of 3–13 independent 96-h growth inhibition assays. Values in parentheses are fold-resistance factors, defined as the IC50 value obtained for the ␤-tubulin mutant line divided by that ob- tained for the parental line. Baccatin and mouse monoclonal anti-␤-tubulin antibody †Sarcodictyin A analogue is methylated at the C-3 oxygen atom of the parent (Tub2.1) were from Sigma, and horseradish peroxidase- compound. This analogue is shown as an active member representative of the conjugated sheep anti-mouse IgG antibody was from Amersham. sarcodictyin family. ‡ The fold resistance factors in parentheses are less than 1 because the IC50 values for the ␤-tubulin mutant cells are lower than the IC for the 1A9 Cell Culture and Cytotoxicity Assay. R ͞ 50 The Epo cell lines, 1A9 A8 parental cells. and 1A9͞B10, were isolated in a single step after exposure of the human ovarian carcinoma cell line A2780 (1A9) to lethal concentrations (IC99) of either Epo A or Epo B (IC99 concen- trations: 6 nM for Epo A and 0.5 nM for Epo B). After an initial the crystal structure. High-temperature (1000 K) molecular expansion, the concentration of epothilones in the culture dynamics (MD) was used to generate Epo conformations. Con- medium was gradually increased to 30 nM for Epo A and 5 nM formers of Epo were analyzed by tracking distances between the for Epo B. Cells were maintained in drug, in a 5% CO2 epoxide O and other atoms to identify a common pharmacoph- humidified atmosphere at 37°C in RPMI medium 1640 ore with taxanes. The Epo͞docetaxel overlap model was formed (GIBCO͞BRL) containing 10% fetal bovine serum, penicillin by using template forcing. The deposited tubulin structure (10) (100 units͞ml), and streptomycin (100 ␮g͞ml) (GIBCO͞BRL). (Refcode 1TUB) was energy refined by using tethered minimi- Before an experiment, the 1A9͞A8 and 1A9͞B10 cells were zations. Epo was docked into the binding site formerly occupied cultured for at least 7–10 days in drug-free medium. Cytotoxicity by docetaxel and optimized. Docetaxel was completely ligated by assays using the protein-staining sulforhodamine B method were water to the tubulin structure and optimized. performed in 96-well plates as described previously (15), by PHARMACOLOGY seeding 500 cells per well and incubating with cytotoxic agents Results and Discussion for 4 days. Two EpoR clones were isolated by selecting 1A9 human ovarian carcinoma cells with Epo A or Epo B. The two clones, designated Tubulin Polymerization Assay. Quantitation of the degree of in vivo 1A9͞A8 and 1A9͞B10, are 25- to 57-fold resistant to Epo A and tubulin polymerization in response to MT-stabilizing agents was B and 5- to 10-fold cross-resistant to PTX, docetaxel, and their performed as previously described (15). Briefly, cells were precursor, baccatin (Table 1). Neither of the EpoR clones express plated in 24-well plates and the following day were exposed to MDR1 mRNA (data not shown), showing that epothilones do not increasing concentrations of PTX or epothilones for 5 h.