J. Biochem. 101, 1247-1252 (1987)

Inhibition of Microtubule Polymerization by Synthetic Estrogens :

Formation of a Ribbon Structure1

Yoshihiro SATO,* Tomoko MURAI,* Taiko ODA,* Hazime SAITO,** Masahiko KODAMA,** and Aiko HIRATA ***

*Biochemistry Division , Kyoritsu College of Pharmacy, Minato-ku, Tokyo 105; **Biophysics Division , National Cancer Center Research Institute, Chuo-ku, Tokyo 104; and ***Institute of Applied Microbiology, The University of Tokyo, Bunkyo-ku, Tokyo 113

Received for publication, December 26, 1986

Dienestrol, meso-hexestrol, and dl-hexestrol, synthetic nonsteroidal estrogens, were shown to be inhibitors of microtubule assembly in vitro using microtubule proteins isolated from porcine brains. The order of activity of the synthetic estrogens as inhibitors of microtubule assembly is: dienestrol > diethylstilbestrol > meso-hex estrol > dl-hexestrol > isodienestrol. The activity of dienestrol as an inhibitor was of the same order as that of (+ )-griseofulvin, as determined by turbidity measure ment. Electron microscopic observation revealed that twisted ribbon structures are formed from microtubule proteins in the presence of some synthetic estrogens (dienestrol, meso-hexestrol, and dl-hexestrol).

In a preceding paper (1), we showed that diethyl to disturbed assembly of microtubule proteins. stilbestrol (DES), a synthetic estrogen as well as Recently, Sharp and Parry (2) and Hartley-Asp a carcinogen, is an inhibitor of microtubule as et al. (3) also reported the effect of DES on mi sembly in vitro. The carcinogenicity of DES, crotubules from the same point of view. which exhibits no mutagenicity in Salmonella, In this report, we examined the inhibitory could be best explained in terms of aneuploidy due activity of a variety of DES analogues (dienestrol, isodienestrol, meso-hexestrol, and dl-hexestrol and 1 This study was supported in part by Grant-in-Aid for their methylether derivatives, Fig. 1) on micro Scientific Research (No. 60303026) from the Ministry of tubule assembly. It was found that dienestrol, Education, Science and Culture of Japan. meso-hexestrol, and dl-hexestrol (4) not only have Abbreviations: pp(CH2)pG, guanyl-5•Œ-methylene di an inhibitory effect on microtubule assembly, but phosphonate; DES, diethylstilbestrol; MES, 2-(morpho also accumulate twisted ribbon structures. lino)ethanesulfonic acid; EGTA, ethyleneglycol-bis(2 aminoethylether)-N,N,N•Œ,N•Œ-tetraacetic acid; EDTA, ethylenediamine tetraacetate; DMSO, dimethylsulfox MATERIALS AND METHODS

ide; DMF, N,N-dimethylformamide; NMR, nuclear magnetic resonance; MAPs, microtubule associated Preparation of Microtubule Proteins-Micro

proteins. tubule proteins were prepared from porcine brains

Vol. 101, No. 5, 1987 1247 248 Y. SATO, T. MURAI, T. ODA, H. SAITO, M. KODAMA, and A. HIRATA

later, carbon-coated collodion 150 mesh copper grids were placed on the drops of diluted and fixed sample solution, and rinsed with the same buffer. The samples were then negatively stained with 1 % uranyl acetate solution and air-dried. Speci mens were examined on a JEOL 200 CX electron microscope at 100 kV. Protein Concentration-Concentration of mi crotubule proteins was determined by the method of Lowry et al. (11) using bovine serum albumin as the standard. Materials-meso-Hexestrol was obtained from Wako Pure Chemical Industries, Ltd. (Osaka). DES and dienestrol were obtained from Tokyo Chemical Industry Co., Ltd. (Tokyo). dl-Hex Fig. 1. Structures of synthetic estrogens: a, diethyl estrol was prepared (12) by catalytic hydrogenation stilbestrol; b, meso-hexestrol; c, dl-hexestrol (3S, 4S of DES over 5 % palladium-charcoal in ethyl isomer) (4); d, dienestrol; and e, isodienestrol. acetate at ambient pressure, and purified through silica gel chromatography. Isodienestrol was pre by temperature-dependent cycles of assembly pared after Liao and Ashman (13). Mono and disassembly after Shelanski et al. (5) and Ishikawa dimethylethers of these compounds were prepared et al. (6) with some modifications. The micro according to Wilds and McCormack (14) and tubule proteins were stored at -50•Ž after one purified by silica gel chromatography. All DES cycle of assembly and they were thawed imme derivatives were proved pure by iH-NMR spec diately before use to receive a second cycle of troscopy and some of them were further confirmed assembly-disassembly. structurally by mass spectroscopy. ATP and GTP

Preparation of Tubulin and MAPS-Tubulin were obtained from Boehringer Mannheim Co., was prepared from two-cycle microtubule proteins Ltd. (B.R.D.), and the materials for electron

by phosphocellulose chromatography after Kuma microscopy were obtained from Nissin EM Co., Ltd. (Tokyo). All other reagents were obtained gai and Nishida (7). MAPs were prepared as "boiled factors" from one -cycle microtubule pro from Wako Pure Chemical Industries Ltd. (Osaka).

teins by the method described by Fellous et al.

(8). RESULTS AND DISCUSSION Assembly Assay-The effect of the test com

pounds on microtubule proteins or purified tubulin Effect of DES Derivatives on Microtubule Pro at 37•Ž was determined by turbidity measurement teins-We examined the effect of DES, meso hexestrol, dl-hexestrol, and dienestrol on micro (9) at 400 nm using a UVIDEC 430B double-beam spectrophotometer equipped with a thermostati tubule polymerization. Turbidity measurement

cally controlled cell holder. Viscometric analysis showed that the inhibitory activity of these com

was performed by using Ostwald-type viscometers. pounds at 50 ƒÊM decreases in the following order: Assembly of purified tubulin was performed after dienestrol > DES > meso-hexestrol > dl-hexestrol

addition of 3 volumes of the buffer consisting of (Fig. 2A). When DES derivatives were added to

10.6% DMSO, 0.1 M MES, 1 mM EGTA, 8 mM preformed microtubules, the order of disassembling MgCl2, 0.5 mM GTP (pH 6.8). Each test com activity is: DES > meso-hexestrol > dienestrol > dl

pound was dissolved in a I : 1 mixture of DMSO hexestrol at 100 ƒÊM (Fig. 2B). On the other and DMF (10) and this solution was added to the hand, isodienestrol showed no significant activity

protein solution at a volume ratio of 2 %. (data not shown). It appears that there is a close Electron Microscopy-Samples were fixed by correlation between the inhibitory and disassembling

addition of 9 volumes of the buffer for assembly activities, except that dienestrol with the highest

containing 1% glutaraldehyde. A few minutes inhibitory activity was less active than DES in

J. Biochem. EFFECT OF NONSTEROIDAL ESTROGENS ON MICROTUBULE ASSEMBLY 1249

Fig. 2. Turbidimetric analysis of assembly-inhibition and disassembly of micro tubules by DES derivatives at 37•Ž. (A) Inhibitory effect on microtubule assembly. Test compounds (50 ECM) were added to microtubule proteins (3

mg/ml) at O min. (B) Disruptive effect on preformed microtubules . Test compounds (100ƒÊM) were added to microtubule proteins (3 mg/ml) preincu bated for 30 min at 37•Ž. The turbidity change was monitored at 5 min intervals but is plotted at 10 min intervals in the figure. •›, control; • , dienestrol; •¡, DES; •¢, meso-hexestrol; •£, dl-hexestrol.

Fig. 3. Electron micrographs of microtubule proteins incubated at 37•Ž for 20 min in the presence of dienestrol. (A) Ribbon-microtubules and ribbon-sheet-microtubules were observed at concentration of 50ƒÊM and (B) only long ribbons were observed at a concentration of 100ƒÊM. Bar, 100 nm.

disassembly. It is worth noting that twisted rib The formation of these ribbon structures undoubt bon structures were formed when microtubule pro edly influenced the turbidity data to some extent teins were incubated in the presence of dienestrol (9). These structures were not observed with either (Fig. 3), meso-hexestrol (Fig. 4) or dl-hexestrol. DES or isodienestrol. Further, the inhibitory ac-

Vol. 101, No. 5, 1987 250 Y. SATO, T. MURAI, T. ODA, H. SAIT6, M. KODAMA, and A. HIP. A

Fig. 4. Electron micrographs of microtubule proteins incubated at 37•Ž in the presence of 100ƒÊM meso hexestrol. (A) Many short ribbons and ribbon-sheets were observed after 5 min of incubation and (B) ribbons, ribbon-sheets, and (ribbon-microtubules; not shown) were observed after 20 min. Bar, 100 nm.

tivity of dienestrol at 50 ƒÊM was turbidimetrically such sites. almost the same as that of (+)-griseofulvin, an In contrast to the case of estrogenic actions, antifungal antibiotic, at 100 ƒÊM (data not shown). the inhibitory activity on microtubule proteins is

We also checked the activity of mono and not always in parallel with the carcinogenic action dimethylethers of DES, dienestrol, and meso of DES derivatives, while the latter data obtained hexestrol in order to estimate the contribution of in vitro and in vivo were at variance (15, 18, 19). The process of carcinogenesis should be more phenolic hydroxyl groups in DES derivatives. The dimethylethers of these compounds had no activity, complex than the interaction of the drugs with and the monomethylethers were less inhibitory on the microtubule proteins. microtubule assembly than their parent phenolic Concentration-Dependence of the Effect of compounds. Figure 5 shows the results of vis meso-Hexestrol on Microtubule Proteins and Puri cometric analysis in the case of DES and its fied Tubulin-Since the effects of dienestrol, meso methylethers. It is clearly indicated that the pres hexestrol, and dl-hexestrol on microtubule proteins ence of hydroxyl groups is indispensable for the were qualitatively similar, more detailed examina activity of DES derivatives. tion was performed with meso-hexestrol. meso

Among a series of DES derivatives, the order Hexestrol inhibited the microtubule assembly in a of inhibitory activity on microtubule proteins is concentration-dependent manner below 50 ƒÊM. almost in parallel with that of estrogenic action At concentrations higher than 100ƒÊM, the turbi

(15). Although microtubule proteins (16, 17) and dity slightly increased with increasing drug con estrogen receptors should not have a common centration (Fig. 6A). sequence of amino acids, they might have similar Microtubule proteins incubated for 20 min in tertiary structures for drug interaction. The pres this experiment were examined by electron mi ence of phenolic hydroxyl groups in DES deriva croscopy. At a concentration lower than 50 ƒÊM, tives could contribute to molecular interactions at normal microtubules were observed. At 100 ƒÊM,

J. Biochem. EFFECT OF NONSTEROIDAL ESTROGENS ON MICROTUBULE ASSEMBLY 1251

ribbons and ribbon-microtubules were observed

along with some microtubules, whereas at 200 or

300ƒÊM only ribbons were observed. Next, the

process of formation of ribbon structures in the

presence of 100ƒÊM meso-hexestrol was checked by electron microscopy. At 0 min, right after the

addition of the drug, microtubule proteins were

observed as small particles. They developed into

numerous short ribbons and some ribbon-sheets

(Fig. 4A) after 5 min of incubation, elongated to

form ribbon-sheet-microtubules and ribbon-micro

tubules (Fig. 4B) after 20 min, and finally accumu

lated microtubules and a small amount of aggre

gates at 60 min (not shown). On the other hand, in the presence of 200 and 300 ƒÊM meso-hexestrol,

no microtubules but only ribbons were formed in

the process (20 min) of incubation, with some

particles, and after further incubation (60 min) they gathered to form aggregates. These results

are not contradictory to the almost linear increase

of turbidity at concentrations higher than 100ƒÊM. Thus, it appears that ribbon structures are tran

sient assembly intermediates which are formed

from tubulin dimers and develope into micro Fig. 5. Viscometric analysis of the effect of DES methylethers on microtubule proteins at 37•Ž. (A) tubules by cylindrical folding. It is postulated Inhibitory effect on microtubule assembly. Test com that synthetic estrogens affect such initial steps of pounds were added at 0 min. (B) Disruptive effect on microtubule formation. preformed microtubules. Test compounds were added after 30 min of preincubation. The protein concen The effect of meso-hexestrol was also examined tration was 5.5 mg/ml. •œ control; •¢, DES dimethyl on purified tubulin. When tubulin solution was ether; •£, DES monomethylether; • , DES. Final concentration of drugs was 100 ƒÊM. incubated in the presence of meso-hexestrol, mi-

Fig. 6. Turbidimetric analysis of the effect of meso-hexestrol on the assembly of micro tubules and purified tubulin. meso-Hexestrol was added to microtubule proteins (3 mg/ml) (A) or purified tubulin (1.8 mg/ml) (B) at 0 min. The final concentration of meso-hexestrol was: •›, 0ƒÊM; •¢, 20ƒÊM; •£ 50ƒÊM; ••¡,100ƒÊM; • , 200ƒÊM; •œ 300ƒÊM.

Vol. 101, No. 5, 1987 1252 Y. SATO, T. MURAI, T. ODA, H. SAITO, M. KODAMA, and A. HIRATA

c rotubule assembly was inhibited in a concentra Katzenellenbogen, J.A. (1981) Mol. Pharmacol. 19, tion-dependent manner up to 50 ƒÊM (Fig. 6B). 388-398

At concentrations higher than 100ƒÊM, the turbi 5. Shelanski, M.L., Gaskin, F., & Cantor, C.R. (1973) Proc. Natl. Acad. Sci. U.S. 70, 765-768 Jity was suppressed to below the control level, but 6. Ishikawa, M., Murofushi, H., & Sakai, H. (1983) there was a transient increase at 20 min and some J. Biochem. 94, 1209-1217 precipitation was observed, which was confirmed 7. Kumagai, H. & Nishida, E. (1979) J. Biochem. 85, to be aggregates by electron microscopy. Inter 1267-1274 e stingly, ribbon structures were not formed from 8. Fellous, A., Francon, J., Lennon, A.-M., & Nunez, purified tubulin without MAPs but they were J. (1977) Eur. J. Biochem. 78, 167-174 formed in the presence of MAPS (8). Although 9. Gaskin, F., Cantor, C.R., & Shelanski, M.L. (1974) ribbon structures as transient intermediates were J. Mol. Biol. 89, 737-758 also demonstrated during normal assembly of 10. Sato, Y., Saito, Y., Shiratori, Y., Shoda, S., & microtubules (20), they were usually detected in Hosoi, J. (1981) Nippon Kagaku Kaishi (in Japanese), abnormal processes of polymerization. Sandoval 746-754 11. Lowry, O.H., Rosebrough, N.J., Farr, A.L., & and Weber (21, 22) reported that ribbon structures Randall, R.J. (1951) J. Biol. Chem. 193, 265-275 were formed from microtubule proteins in the 12. Schwenk, E., Papa, D., Whitman, B., & Ginsberg, presence of pp(CH2)pG instead of GTP. They H.F. (1944) J. Org. Chem. 9, 175-177 initially employed colchicine or podophyllotoxin 13. Liao, S. & Ashman, H.G.W. (1962) Biochim. Blo but later dispensed with these drugs and used phys. Acta 59, 705-707 purified MAP2. On the other hand, Matsumura 14. Wilds, A.L. & McCormack, W.B. (1948) J. Am. and Hayashi (23) have detected ribbon structures Chem. Soc. 70, 4127-4132 in the tubulin which was polymerized at acidic pH 15. Li, J.J., Li, S.A., Klicka, J.K., Parson, J.A., &

(5.8-6.3) in the presence of 1 mM CaCl2. Lam, L.K.T. (1983) Cancer Res. 43, 5200-5204 The present study has demonstrated that rib 16, Ponsting]e, H., Krauhs, E., Little, M., & Kempf, T. bons, ribbon-sheets, and/or ribbon-sheet-micro (1981) Proc. Natl. Acad. Sci. U.S. 78, 2757-2761 tubules are formed in the presence of synthetic 17. Krauhs, E., Little. M., Kempf, T., Hofer-Warbinek, R., Ade, W., & Ponstingl, H. (1981) Proc. Natl. estrogens and GTP, providing important informa Acad. Sci. U.S. 78,4156-4160 tion relating to the mechanism of microtubule 18. McLachlan, J.A., Wong, A., Degen, G.H., & assembly. Barrett, J.C. (1982) Cancer Res. 42, 3040-3045 19. Lacomba, T. & Gabaldon, M. (1971) Cancer Res.

REFERENCES 31,1251-1256 20. Kirschner, M.W., Honig, L.S., & Williams, R.C. 1. Sato, Y., Murai, T., Tsumuraya, M., Saito, H., & (1975) J. Mol. Biol. 99, 263-276 Kodama, M. (1984) Gann 75,1046-1048 21. Sandoval, I.V. & Weber, K. (1979) J. Mol. Biol. 2. Sharp, D.C. & Parry, J.M. (1985) Carcinogenesis 6, 134,159-172 865-871 22. Sandoval, IN. & Weber, K. (1980) J. Biol. Chem. 3. Hartley-Asp, B., Deinum, J., & Wallin, M. (1985) 255,8952-8954 Mutat. Res. 143, 231-235 23. Matsumura, F. & Hayashi, M. (1976) Biochim. Bio 4. Kilbourn, M.R., Arduengo, A.J., Park, J.T., & phys. Acta 453, 162-175

J. Biochem.