J. Med. Chem. 1991,34,992-998
`992
`genated as for 22, affording a mixture of components: yield 76.5
`Plates were incubated for 72 h at 37 OC in a humidified atmo-
`sphere containing 5% COP After 72 h, 20 pL of 5 mg/mL MTl’
`mg, from which 5 was isolated via chromatotron with u& of ethyl
`acetate/hexane (41); total yield of 5 from 20b, 43.5 mg, 42%; mp
`(3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltstolium bromide)
`was added, and cells were incubated for 90 min to allow reduction
`219-220 “C; ‘H NMR (CDC13) 6 6.75 (1 H, s, H7), 6.5 (1 H, s, H4),
`6.31 (2 H, S, H2’, 6’), 6.00 (2 H, AB 9, OCHZO), 5.48 (1 H, 9, M H ) ,
`of the formasan by the surviving cells. Following washing and
`4.74 (1 H, d, H3), 4.37 (2 H, AB q, CH,OCO), 3.87 (2 H, br d,
`solubilization by DMSO, absorbance of each well was measured
`spectrophotometrically at 570 nm. The ICw is determined as the
`CHZOH), 3.82 (6 H, S, OCH3 X 2), 3.57 (1 H, d, H2), 1.81 (1 H,
`br t, OH); HRMS (FAB/HRP), calcd for CZ1HmO8 400.1156, found
`concentration of compound tested required to reduce the ab-
`sorbance to 50% of non-drug-treated control values.
`400.1162. Anal. (CzlHzoOs) C, H; C: calcd, 62.98; found, 61.04.
`Biological Assay. Cells were grown in RPMI 1640 supple-
`mented with 10% fetal calf serum. Test compounds were dis-
`solved in dimethyl sulfoxide (DMSO) and diluted first with &le’s
`Balanced Salt Solution, followed by culture medium, to twice the
`highest concentration of compound to be tested. From this
`concentrated stock, 2-fold serial dilutions were prepared in S w e l l
`microtiter trays. Each concentration was tested in triplicate and
`compared to triplicate drug-free controls. A 100-pL aliquot of
`cells (2.5 x 103 cells) was added to the wells of the microtiter plate
`containing 100 pL of growth medium with or without test drugs.
`
`Acknowledgment. We are grateful to Dr. R. Stephens
`in the Analytical division for helpful NOE analysis on
`several of our compounds and also thank Dr. M. Bures of
`CAMD division for molecular modeling analyses.
`
`Supplementary Material Available: Microanalysis and masa
`spectra for several compounds mentioned in the text (31 pages).
`Ordering information is given on any current masthead page.
`
`Relationships between the Structure of Taxol Analogues and Their Antimitotic
`Activity8
`
`Francoise GuBritte-Voegelein,**t Daniel GuBnard,? Francois Lavelle,* Marie-ThBrBse Le Goff,t Lydie Mangatal,+ and
`Pierre Potiert
`Znstitut de Chimie des Substances Naturelles, Centre National de la Recherche Scientifique, avenue de la Terrasse,
`91 198 Gif-sur- Yvette Ceden, France, and Centre de Recherches d e Vitry, RhGne-Poulenc SantB, 94403 Vitry-sur-Seine,
`France. Received April 2, 1990
`
`A variety of synthetic analogues of taxol, a naturally occurring antitumor diterpene, were examined for their potency
`to inhibit microtubule disassembly. For some of the compounds, the in vitro cytotoxic properties showed a good
`correlation with the tubulin assay. This structure-activity relationship study shows that inhibition of microtubule
`disassembly is quite sensitive to the configuration at (2-2’ and C-3’. A correlation between the conformation of the
`side chain at C-13 and the activity is suggested. Of all the compounds examined, one of the most potent in inhibiting
`microtubule disassembly and in inhibiting murine P388 leukemic cells, N-debenzoyl-N-tert-(butoxycarbony1)- 10-
`deacetyltaxol, named taxotere, was selected for evaluation as a potential anticancer agent.
`
`Several antitumor drugs prevent the formation of the
`mitotic spindle during cell division by interfering with the
`tubulin-microtubules system. Among the different classes
`of natural “mitotic spindle poisons”, the anticancer di-
`terpene taxol’ promotes the assembly of microtubules and
`inhibits the disassembly process of microtubules to tubu-
`lin2s3 in contrast to the vinblastine and colchicine type
`compounds which prevent microtubule assembly. Among
`natural substances, relationships between structure and
`microtubule assembly in vitro have been reported mostly
`for vinblastine! colchicine$ maytansine: podophyllotoxin,’
`and steganacine.8 A good correlation between the inhib-
`ition of tubulin assembly and cytotoxicity has been shown
`for some of these compounds. In the vinblastine series,
`a new hemisynthetic “Vinca alkaloid”, 5’-noranhydrovin-
`blastine or Navelbines was selected by using this in vitro
`assay as a possible useful chemotherapeutic agent and this
`compound is now used in clinics.1°
`In the taxol series, investigation of the structureactivity
`relationships has been limited because of the poor avail-
`ability of taxol (1) from natural sources (only 50-150
`mg/kg of dried trunk bark can be isolated from several
`
`8 Dedicated to Professor G. B. Marini-Bettolo on the occasion
`of his 75th anniversary.
`‘Centre National de la Recherche Scientifique.
`Centre de Recherches de Vitry.
`
`species of yew (genus Taxus, family Taxaceae)’J1).
`However, some closely related taxol congeners, mostly
`
`(1) Wani, M. C.; Taylor, H. L.; Wall, M. E.; Coggon, P.; McPhail,
`A. T. J. Am. Chem. SOC. 1971,93, 2325-2327.
`(2) Suffness, M.; Cordell, G. A. The Alkaloids, Chemistry and
`Pharmacology; Brossi, A., Ed.; Academic Press. Inc.: New
`York, 1985; Vol XXV, Chapter 1.
`(3) (a) Schiff, P. B.; Fant, J.; Horwitz, S. B. Nature 1979, 277,
`665-667. (b) Manfredi, J. J.; Fant, J.; Horwitz, S. B. Eur. J.
`Cell Biol. 1986, 42, 126.
`(4) Zavala, F.; Guhard, D.; Potier, P. Experientia 1978,34,1497.
`(5) Rosner, M.; Capraro, H. G.; Jacobson, A. E.; Atwell, L.; Brossi,
`A,; Ioro, M. A,; Williams, T. H.; Sik, R. M.; Chignell, C. F. J.
`Med. Chem. 1981,24, 257, and references cited therein.
`(6) Higashide, E.; Asai, M.; Ootsu, K.; Tanida, S.; Kozai, Y.; Ha-
`segawa, T.; Kishi, T.; Sugino, y.; Yoneda, M. Nature, Lond.
`1977, 270, 721.
`(7) Gender, W. J.; Murthy, C. D.; Trammell, M. H. J. Med. Chem.
`1977, 20, 635.
`(8) Zavala. F.: GuCnard. D.: Robin. J-P.: Brown. E. J. Med. Chen.
`.
`,
`
`.
`,
`
`-
`1980, 23, 546.
`(9) Maneenev. P.: Andriamialisoa. 2.: Lallemand. Y.: Lannlois. Y.:
`.
`.
`. .
`Lanaois,”N.; Potier, P. Tetrahekron 1979, 35, 2175.
`(10) For an overview of the clinical studies, see: Seminars in On-
`cology 1989, 16, No. 2, Suppl4.
`(11) (a) Miller, R. W.; Powell, R. G.; Smith, C. R., Jr.; Arnold, E.;
`Clardy, J. J. Org. Chem. 1981, 46, 1469-1474. (b) SCnilh, V.;
`Blechert, S.; Colin, M.; GuCnard, D.; Picot, F.; Potier, P.;
`Varenne, P. J. Nat. Prod. 1984,47, 131-137. (c) Magri, N. F.;
`Kingston, D. G. I. J . Org. Chem. 1986,51, 797-802.
`
`
`
`0022-2623/91/1834-0992$02.50/0 0 1991 American Chemical Society
`
`CabRef0001252
`
`AVENTIS EXHIBIT 2026
`Mylan v. Aventis
`IPR2016-00627
`
`

`
`Structure and Antimitotic Activity of Taxol Analogues
`
`Journal of Medicinal Chemistry, 1991, Vol. 34, No. 3 993
`
`Table I. Structures of Taxol Analogues and Their Inhibition of Pig Brain Microtubule Diassembly”
`
`kw6b
`RZ
`
`R1
`
`IDM/IDm
`R3
`ref
`compd
`23
`CH=CHC6H6
`H
`H
`20b, 21
`4
`100
`CH=CHCH3
`COCH3
`H
`20b
`5
`CH(OH)CH(OH)CeH,
`H
`H
`20b
`7‘
`3
`60
`CH(OH)CH(OH)CH3
`COCH3
`H
`20b
`8‘
`loo0
`CH=CHC6H6
`COZCHzCC13
`COZCHzCC13
`20,21
`1 la
`CH=CHCeHe
`COCH3
`COZCH2CCls
`20,21
`l l b
`-
`The isolation of tubulin from pig brain and inhibition studies were carried out as previously described.13 *ID, is the concentration of
`drugs leading to a 50% inhibition of the rate of microtubule disassembly. The ratio IDSo/ID, (taxol) gives the activity with regard to taxol
`itself (taxol: IDSO (pM) = 0.4). CThreo compounds 2’R,3’R + 2’S,3’S.
`
`acylated at C-2‘ and/or C-7,12-16 have been prepared as
`have water-soluble taxol derivatives, thereby showing that
`esters at C-2’ can serve as prodrugs of taxol.16
`Although previous studies have mostly highlighted the
`importance of the side chain in the 13-position for the in
`vitro disassembly13-assembly14 process and for the cyto-
`toxic activity,’J4 we were interested in studying in more
`details the effects of structural and/or configurational
`modifications at carbons 2’ and 3’ of the side chain and
`carbons 7 and 10 of the taxane skeleton, on the biological
`activity.
`To overcome the serious problems posed by the poor
`availability of taxol (1) and its derivatives, we have used
`simpler natural taxane-type compounds as “chemical
`precursors” to synthesize more complex taxol-like products.
`In this way, 10-deacetylbaccatin I11 (2a), easily extracted
`from the yew 1eave~’’J~~ and baccatin I11 (3a), isolated from
`the heartwood’* or prepared from 2a,19 can be used as
`
`(a) Kingston, D. G. I.; Magri, N. F.; Jitrangsri, C. New Trends
`Nat. Prod. Chem. 1986, 26, 219. (b) Mellado, W.; Magri, N.
`F.; Kingston, D. G. I.; Garcia-Arenas, R.; Orr, G. A.; Horwitz,
`S. B. Riochem. Riophys. Res. Commun. 1984, 224, 2; 329.
`(a) Lataste, H.; SBnilh, V.; Wright, M.; GuBnard, D.; Potier, P.
`Proc. Natl. Acad. Sci. U.S.A. 1984, 81, 4090. (b) SBnilh, V.
`PhD Thesis, March 2, 1984, UniversitC de Paris-Sud, Orsay.
`Parness, J.; Kingston, D. G. I.; Powell, R. G.; Harracksingh, C.;
`Horwitz, S. B. Biochena. Biophys. Res. Commun. 1982, 105,
`1082.
`Magri, N. F.; Kingston, D. G. I. J . Nat. Prod. 1988, 298-306.
`Deutsch, H. M.; Glinski, J . A.; Hernandez, M.; Haugwitz, R.
`D.; Narayanan, V. L.; Suffness, M.; Zalkow, L. H. J. Med.
`Chem. 1989,32, 788-792.
`Chauvisre, G.; GuCnard, D.; Picot, F.; SBnilh, V.; Potier, P. C.R.
`Acad. Sci. Paris 1981, 293, 501-503.
`(a) Della Casa de Marcano, D. P.; Halsall, T. G. J. Chem. SOC.
`Chem. Comm. 1975, 365. (b) Miller, R. W.; Powell, R. G.;
`Smith, C. R., Jr.; Arnold, E.; Clardy, J. J. Ore. Chem. 1981,46,
`1469-1474. (c) SCnilh, V.; Blechert, S.; Colin, M.; GuBnard, D.;
`Picot, F.; Potier, P.; Varenne, P. J. Not. Prod. 1984, 47,
`131-137.
`
`starting materials for the preparation of taxol and deriv-
`atives. Recently, we reported two partial syntheses of taxol
`(1) and structural analogues from 2a and 3a. One ap-
`proach relies on the double bond functionalization of
`cinnamoyl derivatives of taxane,20p21 the other makes use
`of the direct esterification of a taxol-like side chain on
`10-deacetylbaccatin III.’9921c
`Using these two synthetic approaches, we have been able
`to prepare a number of new taxol-like substances and
`consequently to further study the structure-activity rela-
`tionships in this series. The potential antimitotic activities
`of these new compounds have been investigated by using
`the in vitro tubulin assay. For some of the compounds,
`the in vitro cytotoxic properties were also determined.
`Chemistry
`Structures of compounds, literature references con-
`cerning their preparation together with their activity on
`tubulin are given in Tables I and 11.
`Esterification of cinnamic, crotonic, and 3-phenyl-
`propionic acids with compounds 2b or 3b, followed by
`
`(19) Denis, J-N.; Greene, A.; GuBnard, D.; Gugritte-Voegelein, F.;
`Mangatal, L.; Potier, P. J . Am. Chem. SOC. 1988, 120,
`5917-5919.
`(20) (a) SBnilh, V.; GuBritte, F.; GuBnard, D.; Colin, M.; Potier, P.
`(b)
`C.R. Acad. Sci. Paris, 1984, 299, SBrie 11, 1039-1049.
`GuBritte-Voegelein, F.; SBnilh, V.; David, B.; GuBnard, D.;
`Potier, P. Tetrahedron 1986, 42, 4451-4460.
`(21) (a) Colin, M.; GuBnard, D.; GuBritte-Voegelein, F.; Potier, P.
`(Rh6ne-Poulenc SantC) Eur. Pat. Appl. EP 253,738 (Cl.
`C07D305/14), January 20, 1988, Fr Appl86/10,400, 17 July
`1986; Chem. Abstr. 1988,209,22762~. (b) Mangatal, L.; Ad-
`eline, M-T.; Gudnard, D.; GuBritte-Voegelein, F.; Potier, P.
`Tetrahedron 1989, 42, 4451. (c) Mangatal, L., PhD Thesis,
`April 19, 1989, UniversitB de Paris XI Orsay. (d) Colin, M.;
`GuCnard, D.; GuBritte-Voegelein, F.; Potier, P. (Rhhe-Poulenc
`SantB) Eur. Pat. Appl. EP 253,739 (CI.C07D305/14), Jan 20,
`1988, Fr Appl86/10,401, Jul17,1986; Chem. Abstr. 1988,109,
`22763~.
`
`CabRef0001253
`
`

`
`994 Journal of Medicinal Chemistry, 1991, Vol. 34, No. 3
`
`GuBritte- Voegelein et al.
`
`Table 11. Inhibition of Pig Brain Microtubule Disassembly by Taxol Analogues at 4 O c a
`
`compd
`
`ref
`
`R1
`
`20a
`20a
`20a
`20a
`21
`21
`21
`21
`
`H
`6
`9a (2’R)
`H
`9b (2’s)
`H
`10a (3’s or 3’R)
`H
`10b (3’R or 3’s)
`H
`12a (2’R,3’S)
`H
`12b (2’S,3’R)
`H
`12c (2’R,3’S)
`H
`12d (2’S,3’R)
`H
`13a (2’R,3’S)
`H
`13b (2’S,3’R)
`H
`13c (2’R,3’S)
`H
`13d (2’S,3’R)
`H
`H
`1 3ea
`H
`13F
`14a (2’R,3’S)
`H
`14b (2’S,3’R)
`H
`14c (2’R,3’S)
`H
`14d (2’S,3’R)
`H
`H
`1 4ea
`H
`14F
`15 (2’R,3’S)
`H
`16a (2’R,3’S)
`H
`16b (2’S,3’R)
`H
`16c (2’R,3’S)
`H
`16d (2’S,3’R)
`H
`l a (2’R,3’S)
`H
`17b (2’S,3’R)
`H
`17c (2’R,3’S)
`H
`17d (2’S,3’R)
`H
`18 (2’R,3’S)
`H
`H
`19aa
`H
`19ba
`20 (2’R,3’S)
`H
`H
`21 (2’R,3’S)
`22 (2’R,3’S)
`CO(CH2)&02H
`23 (2’RJ’S)
`CO(CH2)&02H
`COCH2NH2
`24 (2’R,3’S)
`25 (2’R.3’S)
`COCH(NH9)CHGHc.
`-.
`, . ” ”
`a Erythro compounds 2’R,3’R or 2’S,3’S.
`
`21
`21
`21
`21
`
`21c
`21
`21
`21
`21
`21
`21
`21
`21
`
`R2
`
`H
`H
`H
`H
`H
`H
`H
`H
`H
`H
`H
`H
`H
`H
`H
`H
`H
`H
`H
`H
`H
`H
`COCHS
`COCH,
`COCH3
`COCH3
`COCH3
`COCH3
`COCH3
`COCH3
`COCH3
`H
`H
`H
`H
`COCH,
`CO(CH2)3C02H
`COCHzNHz
`H
`
`R3
`
`H
`OH
`OH
`H
`H
`OH
`OH
`NHTs
`NHTs
`OH
`OH
`NHCOztBu
`NHCOztBu
`OH
`OH
`OH
`OH
`NHCOPh
`NHCOPh
`OH
`OH
`OH
`OH
`OH
`NHCOztBu
`NHCOZtBu
`OH
`OH
`NHCOPh
`NHCOPh
`OH
`OH
`OH
`OH
`OH
`OH
`OH
`OH
`OH
`
`R4
`
`H
`H
`H
`NHC02tBu
`NHCOztBu
`NHTs
`NHTs
`OH
`OH
`NHC0,tBu
`NHC02tBu
`OH
`OH
`NHC02tBu
`NHC0,tBu
`NHCOPh
`NHCOPh
`OH
`OH
`NHCOPh
`NHCOPh
`NH2
`NHC0,tBu
`NHCOztBu
`OH
`OH
`NHCOPh
`NHCOPh
`OH
`OH
`NH,
`NH,
`NH,
`NHkO(CH2)3C02H
`NHCO(CGH&S03H
`NHCOPh
`NHC02tBu
`NHC0,tBu
`NHCOdBu
`
`ID,/ID,(taxol)
`17
`4.5
`3.5
`2.3
`4.1
`5
`-
`15
`-
`0.5
`30
`10
`160
`1.8
`4.3
`1.3
`4
`10
`170
`1.3
`1.3
`-
`0.5
`30
`10
`108
`1
`4.5
`10
`110
`44
`30
`30
`1
`5.5
`1
`2
`1.2
`1
`
`deprotection, led, respectively, to esters 4, 5, and 6.
`2’,3’-Dihydroxy derivatives 7 and 8 were easily obtained
`from esters 4 and 5. Coupling of racemic O-(l-ethoxy-
`ethyl)-3-phenyllactic acid with 2b led, after deprotection,
`to a mixture of esters 9a and 9b in a 82/18 ratio showing
`that an asymmetric induction in favor of the 2’R isomer
`took place during esterification.21c Under the same con-
`ditions, coupling of L-O-( l-ethoxyethyl)-3-phenyllactic acid
`yielded compounds 9a and 9b in a 60140 ratio, showing
`that epimerization at C-2’ also occurred during esterifi-
`cation. Racemic N-tert-(butoxycarbonyl)-3-amino-
`propionic acid yielded the (2-3’ functionalized derivatives
`10a and 10b after removal of the protecting groups.
`Application of the Sharpless vicinal oxyamination re-
`action2* to cinnamate derivatives such as lla using chlo-
`ramine T or tert-butyl-N-chloro-N-argentocarbamate
`followed by deprotection of the C-7 and C-10 hydroxyl
`groups afforded, respectively, a mixture of threo hydroxy
`p-toluenesulfonamide isomers 12a-d and threo hydroxy-
`carbamates 13a-d. Compound 13a was correlated with
`
`(22) Sharpless, K. B.; Patrick, D. W.; Truesdale, L. K.; Biller, S. A.
`J. Am. Chem. Soc. 1975,97, 2305-2307.
`
`10-deacetyltaxol(14a) while 2‘-epi,3‘-epi, 10-deacetyltaxol
`14b was obtained from 13b. In the same way regioisomers
`14c and 14d were prepared from the oxyaminated com-
`pounds 13c and 13d. Amino alcohol 15 was prepared from
`hydroxycarbamate 13a after deprotection of the amino
`group. The same reactions were applied to the cinnamate
`derivative of baccatin I11 1 l b to provide oxycarbamates
`16a-d, N-benzoyl-3-phenylisoserine isomers 1 (taxol),
`17b-d, and amino alcohol 18. Taxol (11, 10-deacetyltaxol
`(14a), and hydroxycarbamates were also obtained by direct
`esterification.
`Erythro isomers 13e,f and 14e,f were prepared from
`cinnamate ester 1 la. Epoxidation of 1 la yielded a mixture
`of two diastereoisomers which were treated with sodium
`azide. After reduction and deprotection with zinc dust in
`acetic acid, two amino alcohols 19a and 19b were obtained.
`Treatment of compounds 19a and 19b with benzoyl
`chloride or di-tert-butyl dicarbonate yielded, respectively,
`the erythro isomers of 10-deacetyltaxol 14e,f and the er-
`ythro hydroxycarbamates 13e,f. Compounds 20 and 21
`were obtained by condensing the appropriate anhydride
`with the amino alcohol 15. Esterification of taxol (1) and
`compound 13a with glutaric anhydride led, respectively,
`
`CabRef0001254
`
`

`
`Structure and Antimitotic Activity of Taxol Analogues
`to products 22 and 23. Finally, acylation of 13a with
`suitably protected amino acids such as glycine and phe-
`nylalanine gave compounds 24 and 25. All new compounds
`were principally characterized by NMR and MS.
`Results and Discussion
`Different in vitro assays have been used to determine
`the activity of taxol congeners on tubulin (promotion of
`microtubule assembly in the absence of GTP,14 inhibition
`of binding of tritiated taxol to microtubule^,'^ and inhib-
`ition of microtubule disassembly at 4 OCU$lc). Concerning
`the drug-tubulin interaction, the two procedures involving
`inhibition of microtubule assembly or inhibition of mi-
`crotubule disassembly gave the same results. However,
`because of the rapidity of the latter method (5 min per
`sample), all the compounds described in Tables I and I1
`were assayed for their ability to inhibit the disassembly
`process of microtubules at 4 “C.
`Previous structure-activity studies have shown that (a)
`the side chain at C-13 is necessary for a good drug-receptor
`interaction,lJ3J4 and (b) modification of substituents at
`C-10 and/or C-7 such as replacement of a hydroxyl group
`by an acyl or xylosyl group has little effect on the activi-
`ty.12b,13,14
`With respect to structural modifications on the side
`chain, structural modifications have mainly been made at
`C-2’ and have shown that acylation of this carbon results
`in a loss of activity in the tubulin assay12bJ3 but not in the
`cytotoxicity assay. These observations recently led to the
`preparation of water soluble derivatives of taxol.16
`The ID,, values obtained in this study (Tables I and 11)
`are in good agreement with those previously described and
`also provide much more information concerning the spe-
`cific influence of configuration and structural modifications
`of the side chain on the activity of taxol-type molecules:
`(a) Replacement of the C-10 acetoxy group with a hydroxyl
`group did not lead to loss of potency in different series of
`taxane derivatives (Compare series 13a-d with 16a-b, and
`14a-d and 1,17b-d). (b) Replacement of a 3’-phenyl group
`with a methyl group resulted in a major loss of activity.
`Thus the 2’,3’-dihydroxycrotonyl ester 8 is 20-fold less
`potent than its cinnamoyl analogue 7. (c) Branching of
`large groups such as [ (trichloroethyl)oxy]carbonyl at C-7
`and/or (2-10 (lla,b) resulted in a loss of activity whereas
`compounds with polar substituents such as xylo~yl,’~ glu-
`taryl (22,23), or aminoacyl (24,25) are as potent as taxol
`in the tubulin assay. (d) No loss of potency was observed
`for compounds having different kinds of hydrophobic
`substituents on the amido group at C-3’. Thus, replace-
`ment of the phenyl group in taxol with tiglyl (cephalo-
`manine),I4 tosyl (12a), or hexanoyl groups13b gave com-
`pounds as active as taxol. Moreover, the tert-butyloxy-
`carbonyl compounds (13a,16a) were shown to be the most
`potent inhibitors of microtubule disassembly so far pre-
`pared by us. In contrast, compounds having a free amino
`group at C-3’ are less potent than their N-amido analogues
`(compare 18 and taxol (1)).
`More interesting are the results obtained with com-
`pounds having different configurations at (3-2’ and/or C-3’.
`Both threo ((2’R,3’S) 13a,14a,16a and 1; (2’S,3’R)
`13b,14b,16b, and 17b) and erythro ((2’R,3’R; 2’S,3’S) 13e,
`13f, 14e, and 14f) diastereoisomers were assayed, showing
`that inhibition of microtubule disassembly is quite sen-
`sitive to these two configurations. Thus in all cases, the
`2’R,3‘S diastereoisomer (natural configuration), was found
`to be the most potent. If this is true, then one question
`remains unanswered with regard to our observations: do
`the differences in activity seen among the various ana-
`logues synthesized arise from the presence or absence of
`
`Journal of Medicinal Chemistry, 1991, Vol. 34, No. 3 995
`t
`
`IQH
`
`Figure 1. Three-dimensional view of compound 13a.
`particular functionalities per se or are these the result of
`differences in the side chain conformations imposed by
`these modifications, resulting in a favorable or unfavorable
`fit to‘the activity site?
`The conformation of taxol-like compounds is imposed
`by the three-ring system having the highly strained
`eight-membered ring B cis-fused to ring A and trans-fused
`to ring C and, in addition, a bridgehead double bond, a
`hindering geminal dimethyl group, and a planar oxetane
`ring. This particular structure gives rise to a “caged-type
`conformation”. A recent X ray analysis of the 2’R,3’S
`compound 13a,24 giving interatomic bond lengths, showed
`that the side chain adopts a particular conformation due
`to intramolecular hydrogen bonds and repulsive interaction
`between the substituents at carbons 2’ and 3’ and those
`of the taxane skeleton (Figure 1). Moreover, recent
`ROESY experiment^^^ with threo isomers showed that
`compounds having a 2’R,3’S configuration exhibit a num-
`ber of NOE’s, indicating interactions between C-3’H and
`the C-4 acetyl group and between C-2’H and C-18H3. In
`contrast, the 2’S,3’R diastereoisomers are characterized by
`NOE involving C-2’H/C-4 acetyl group and C-3’H/C-18H3.
`Though these observations can also be predicted by mo-
`lecular modeling, they need to be confirmed by other NMR
`experiments under different conditions of temperature and
`solvent. Configurations at C-2’ and/or C-3’ of compounds
`10a,b and erythro isomers 13e,f and 14e,f are still unknown
`
`(23) Beloeil, J-C.; Dubois, J.; Gillet, B.; Guenard, D.; Gueritte-
`Voegelein, F. Unpublished results.
`(24) Gugritte-Voegelein, F.; Gugnard, D.; Mangatal, L.; Potier, P.;
`Guilhem, J.; Cesario, M.; Pascard, C. Acta Crist. 1990, C46,
`781-784.
`(25) Lavelle, F.; Fizames, C.; Gueritte-Voegelein, F.; Gubnard, D.;
`Potier, P. Proc. Am. Assoc. Cancer Res. 1989,30, 2254.
`
`CabRef0001255
`
`

`
`996 Journal of Medicinal Chemistry, 1991, Vol. 34, No. 3
`
`Table 111. Cvtotoxicitv of Taxol Analogues
`P388O
`IDW/IDm(13a)'
`IC,/ICW(13a)*
`IC,
`compd
`0.27
`taxol (1)
`2
`2
`taxotere (13a)
`0.13
`1
`1
`> 10
`>77
`60
`13b
`20
`4
`31
`13c
`1
`0.17
`1.3
`16a
`7
`54
`17b
`9
`20
`6
`46
`17c
`>4
`>31
`21
`11
`1.4
`1.5
`0.19
`24
`0.12
`2
`1
`25
`"Murine P388 leukemic cells were obtained from the tumor
`bank of the National Cancer Institute. P388 cells were grown at 37
`"C as suspensions in RPMI 1640 medium containing 10 pM 2-
`mercaptoethanol, 2 mM L-glutamine, 200 U/mL penicillin, 200
`pg/mL streptomycin, and supplemented with 10% (v/v) fetal calf
`serum. Exponentially growing P388 cells suspended in complete
`medium were seeded in tubes at IO6 cells/mL.
`the compounds
`were added at different concentrations on day 0 (3 tubes/concen-
`tration). Cells were allowed to grow for 3 or 4 days at 37 "C under
`5 % COP Final cell numbers were measured by using a Coulter
`counter. The results were expressed as the concentration (pg/mL)
`which inhibits 50% of the cell proliferation (ICw). The ICM were
`estimated by regression analysis concentration-response data.
`*Cytotoxicity of drugs in comparison to compound 13a. cIDM for
`disassembly of microtubule: see Table 11. The ratio IDw/IDM-
`(13a) gives the activity with regard to 13a itself(l3a: IDM (pM) =
`0.2).
`but some information can also be provided by ROESY
` experiment^.^^ These studies could show the direct in-
`fluence of the side-chain conformation on the inhibition
`of microtubule disassembly. It is already interesting to
`note that the gain in activity, in going from compound 6,
`having no substituent at (2-2' and C-3', to compound 13a,
`bearing hydroxyl and hydroxycarbamate groups at C-2'
`and C-3', may be the product of the separate contributions
`brought by the substituents at carbons 2' (compound 9)
`and 3' (compound 10).
`Other structural features such as the benzoate group at
`C-2 or the oxetane ring are probably essential for exhibiting
`a good activity. Indeed, products lacking these groups,
`though identified by others in yew extracts (i.e. taxanes
`with C4-C20 double bond),26 were not isolated by using
`the tubulin bioassay guided fra~tionation."~J~~
`Previous results have shown that taxanes lacking the
`side chain at C-13 have the same affinity as taxol for
`Physarum t ~ b u 1 i n . l ~ ~ This may be the result of a point
`mutation on the peptide chain in the neighborhood of the
`binding site, allowing a specific noncovalent bond between
`the side chain and mammalian tubulin; moreover, the side
`chain by itself is inactive and has no effect on the binding
`of taxol or its analogues. These two observations allow us
`to imagine the binding process to occur as follows: (a) the
`taxane skeleton is recognized and binds to its site on tu-
`bulin, and (b) the drug-tubulin bond is stabilized by the
`specific interaction of the side chain, a situation which is
`probably not restricted to the field of taxane derivatives.
`Concerning the in vitro cytotoxic activity, a generally
`good correlation with the tubulin assay may be noted:
`compounds such as 1, 13a, 16a, 24, and 25 have the ability
`to inhibit both cell growth and disassembly of microtubules
`(Table 111). However, the C-2' esters, less active in di-
`sassembling microtubules, are promptly hydrolysed in
`intra- or extracellular media, resulting in good cytotoxic
`activity.I2b,l3,14,16
`
`(26) Ud- Khan, N.; Parveen, N. J. Sci. Ind. Res. 1987,46,512, and
`referenced cited therein.
`
`~~
`
`~
`
`GuBritte-Voegelein et al.
`Some attempts have been made to solubilize taxol de-
`rivatives in water, but the conditions used to obtain sol-
`ubilization are incompatible with the stability of the final
`products (Le. acidic or basic medium). The diglycine de-
`rivative 24, soluble in water as its hydrochloride salt, is a
`potent inhibitor of microtubule disassembly and in vitro
`cellular proliferation but has no effect in vivo on murine
`le~kemia.~' The other amino acid or acid derivatives are
`good inhibitors of disassembly but they are only partially
`soluble in an ethanol-water mixture.
`Conclusion
`About 40 synthetic taxane-type compounds have been
`tested as potential inhibitors of disassembly of mammalian
`tubulin. The tubulin assay has proven to be a convenient
`method for studying structure-activity relationships,
`particularly in regard to the conformations of the "active"
`molecules. Correlation with activity against P388 murine
`leukemia shows that the tubulin assay is also a very effi-
`cient tool for a preliminary evaluation of new active
`products in the taxane family.
`Of all the compounds examined in Tables I and 11, one
`the most potent, N-debenzoyl-N-(tert-butoxy-
`of
`carbonyl)-lO-deacetyltaxol(13a) named taxotere, was se-
`lected for evaluation as a potential anticancer agent and
`so far has shown excellent antitumor activity against
`several models of grafted murine tumors.26 Moreover
`taxotere (13a) showed a better solubility in excipient
`system (polysorbate 80/ethanol, 1:l) than the two others
`most active compounds taxol (1) and N-debenzoyl-N-
`(tert-butoxycarbony1)taxol (16a).
`Experimental Section
`The purity of the samples was checked by chromatographic
`methods (HPLC and TLC) and careful analysis of NMR spectra.
`Melting points were taken on a Kofler hot bench and are un-
`corrected. Optical rotations (c, cg/mL) were determined on a
`Perkin-Elmer 141 MC polarimeter using a 10 cm path length cell.
`Infrared spectra (cm-', CHC1, or Nujol) were recorded on a
`'H and 13C NMR were
`Perkin-Elmer 257 spectrophotometer.
`recorded on a Bruker AM 200 or AM 400 spectrometer. Chemical
`shifts are in ppm relative to TMS (0.00). Multiplicities are in-
`dicated in parentheses with coupling constants expressed in hertz.
`Mass spectra were recorded on an AEI MS9 (CI) or on a Kratos
`MS8O (FAB). The 10-deacetylbaccatin I11 used in this study was
`isolated from the leaves of the yew tree Taxus baccata L.
`General procedures for esterification of 7,10-bis[ [ (2,2,2-tri-
`chloroethyl)oxy]carbonyl]-l0-deacetylbaccatin I11 (2b) with
`different acids and for the deprotection of C-7 and/or C-10 troc
`group are described in refs 20b and 21.
`3-(Phenylpropionyl)-lO-deacetylbaccatin 111 (6). Esteri-
`fication of 2b (500 mg, 0.56 mmol) with 3-phenylpropionic acid
`(330 mg, 2.2 mmol) gave 93% of the corresponding ester which
`was deprotected to yield compound 6 (70%): MS-CI m / z 677
`(MH'); 'H NMR (CDCl,) 6 1.11 (s, C-17H3), 1.21 (s, C-16H3), 1.74
`(s, C-19H3), 1.82 (s, C-18H3), 2.20 (s, OCOCH,), 2.74 (m, C-2'H2),
`3.05 (m, C-3'Hz), 3.91 (d, J = 7, C-3H2), 4.18 and 4.31 (s d, J =
`9, C-20H2), 4.23 (m, C-7H), 4.95 (d, J = 9, C-5H), 5.20 (s, C-IOH),
`5.67 (d, J = 7, C-2H), 6.17 (t, J = 8, C-l3H), 7.25 and 7.33 (C,H,),
`7.50, 7.62, and 8.06 (OCOCGH6).
`Compounds 9a and 9b. O-(l-Ethoxyethyl)-3-phenyllactic acid
`was prepared by classical procedures from 3-phenyllactic acid
`(first, protection of the carboxyl group as a benzyl ester (MS-E1
`m / z 256 (M'+)), second, protection of the hydroxyl group as
`ethoxyethyl ether (MS-E1 m / z 328 (M'+)), third, cleavage of the
`benzyl ester by hydrogenolysis).
`Esterification of 2b (669 mg, 0.75 mmol) with DL-O-(l-eth-
`oxyethyl)-3-phenyllactic acid (714 mg, 3 mmol) gave 98% of a
`mixture of two esters which were deprotected to give 9a (56%)
`and 9b (12%). Diastereoisomers were separated by silica gel
`column chromatography using hexane/ethyl acetate (2:8) as
`
`(27) Lavelle, F. Personal communication.
`
`CabRef0001256
`
`

`
`Structure and Antimitotic Activity of Taxol Analogues
`
`Journal of Medicinal Chemistry, 1991, Vol. 34, No. 3 997
`
`eluant. Under the same conditions, esterification of 2b with
`L-O-( 1-ethoxyethyl)-3-phenyllactic acid gave after deprotection
`a 60/40 mixture of 9a and 9b.
`(C = 1.00, MeOH); MS-FAB' m/z 693 (MH');
`9a: [a]=D = -57'
`'H NMR (CDC13) 6 1.11 (8, C-17H3), 1.22 (8, C-16H3), 1.72 and
`1.73 (29, C-19H3 and C-18H3), 1.82 (m, C-6H), 2.12 (m, C-14Hz),
`2.21 (s, OCOCH3), 2.55 (m, C-6H), 3.12 (m, C-3'Hz), 3.85 (d, J
`= 7, C-3H), 4.18 (m, C-7H), 4.15 and 4.27 (2 d, J = 8, C-20Hz),
`4.50 (m, C-2'H), 4.92 (d, J = 9, C-5H), 5.17 (s, C-lOH), 5.62 (d,
`J = 7, C-2H), 6.11 (t, J = 9, C-l3H), 7.18,7.25,7.30 (CeH5), 7.48,
`7.60, and 8.05 (OCOC6H5).
`(c = 1.00, MeOH); MS-FAB' m/z 693 (MH+);
`9 b [a]l"D -37'
`'H NMR (CDCl3 + CD30D) 6 1.11 (8, C-17&), 1.22 (s, C-16H3),
`1.75 (8, C-19H3), 1.84 (s, C-18H3), 1.92 (m, C-6H), 2.15 (dd, J =
`15 and 9, C-l4H), 2.26 (8, OCOCH3), 2.28 (m, C-14H), 2.53 (m,
`C-6H), 3.03 and 3.24 (2 dd, C-3'Hz), 3.84 (d, J = 7, C-3H), 4.15
`(m, C-7H), 4.18 a i d 4.28 (2 d, J = 9, C-20Hz), 4.47 (dd, J = 8 and
`4.5, C-2'H), 4.96 (d, J = 9, C-SH), 5.20 (s, C-lOH), 5.66 (d, J =
`7, C-2H), 6.17 (t, J = 9, C-13H), 7.25 (C&5), 7.45, 7.60, and 8.00
`(OCOCsH,).
`Compounds 10a and lob. DL-3-[ (tert-Butoxycarbony1)-
`amino]-3-phenylpropionic acid was prepared in 86% from DL-3-
`amino-3-phenylpropionic acid.21c Esterification of this acid (999
`mg, 3.77 mmol) with 2b (838 mg, 0.94 mmol) gave a mixture of
`two diastereoisomeric esters in 98% yield. Removal of the pro-
`tecting groups at C-7 and C-10 provided diastereoisomers 10a
`(60%) and 10b (40%) after purification by HPLC (RP-18; mobile
`phase, H,O/MeOH 36:64; flow rate; 2 mL min-').
`loa: mp 160 'c (MeOH); [a]"D = -18'
`(c = 1.10, CHCl,);
`MS-FAB+ m/z 792 (MH+), 774, 692, 527, 509, 266; 'H NMR
`(CDC13) 6 1.08 (8, C-ITHa), 1.15 (9, C-l6H3), 1.42 (9, tBu), 1.72
`(C-18H3) and C-19H3), 1.82 (m, C-6H), 2.13 (m, C-14H2), 2.28 (e,
`OCOCH3), 2.56 (m, C-6H), 2.93-3.10 (m, C-2'Hz), 3.85 (d, J = 7,
`C-3H), 4.18 (m, C-7H), 4.15 and 4.25 (2 d, J = 9, C-20Hz), 4.94
`(d, J = 9, C-5H), 5.15 (s, C-lOH), 5.18 (br d, C-3'H), 5.60 (br d,
`NH), 5.70 (d, J = 7, C-2H), 6.05 (t, J = 8, C13H), 7.30 (C&), 7.50,
`7.60, and 8.05 (OCOC&,); 13C NMR (CDCl3) 69.90 (C19), 14.40
`(C18), 20.40 (C17), 22.60 (CH3-acetate), 26.50 (ClS), 28.40
`(CH,-tBu), 36.00 (C14), 36.80 (C6), 41.60 (C-2'), 43.00 (C15), 46.70
`(C3), 54.7 ((239, 57.80 (C8), 70.30 (C7), 71.90 (C13), 74.60 (ClO),
`75.10 (C2), 76.50 (C20 and C-l), 78.80 (C-tBu), 81.20 (C4), 84.30
`(C5), 126.10, 127.70, 128.70, 128.80, 130.10, 133.70 (0-Bz, m-Bz,
`p-Bz, o-Ph, m-Ph, p-Ph), 129.50 (Cl-Bz), 135.90,138.70 and 138.50
`(C11, C12, and C1-Ph), 155.20 ( C 4 of carbamate), 166.90 (C=O
`of Bz), 169.80 (C=O of Ac), 170.60 (Cl'), 211.13 (C9).
`lob: mp 156 'c (MeOH); [(YlzZD = -12'
`(c = 0.90, CHCI,);
`MS-FAB+ m/z 792 (MH'), 774, 692, 527, 509, 266; 'H NMR
`(CDC13) 6 1.10 (s, C-17H3), 1.18 (s, C-16H3), 1.43 (e,