Total synthesis of tubulysins U and V

Article abstract of DOI:10.1002/anie.200604557

Meeting the challenge: A reliable and modular reaction sequence has been developed for the synthesis of the challenging tubulysin framework. This route allows preparation of hundreds of milligrams of the stereochemically pure tetrapeptides (see picture), which are produced in small amounts by two different species of myxobacteria. Thus, full biological evaluation of the tubulysins and their analogues is now a real possibility. (Chemical Equation Presented).

Full text of DOI:10.1002/anie.200604557

Communications
DOI: 10.1002/anie.200604557
Total Synthesis
Total Synthesis of Tubulysins U and V**
Monica Sani, Giacomo Fossati, Florent Huguenot, and Matteo Zanda*
Tubulysins (Scheme 1) are a family of tetrapeptides produced
in rather small quantities (< 4 mgLÀ1 culture broth) by two
different species of myxobacteria: Archangium gephyra and
Angiococcus disciformis.[1] The structure, stereochemistry,
and biosynthetic pathway of the tubulysins were recently
determined by Höfle, Reichenbach, and co-workers, who also
reported the potent cytotoxicity (in the nanomolar concen-
tration range) of these compounds.[2] The cytotoxicity of the
tubulysins stems from their ability to bind tubulin and
disintegrate microtubules of dividing cells, thus inducing
apoptosis.[3]
From a structural point of view, tubulysins are derived
from an N-methylpipecolic acid residue (Mep) at the
N terminus, isoleucine (the only proteinogenic amino acid)
at the second position, an unusual thiazole-containing amino
acid, tubuvaline (Tuv), which features two stereogenic centers
at the third position, and two possible g-amino acids at the
C terminus: either tubutyrosine (Tut, tubulysins A, B, C, G,
and I) or tubuphenylalanine (Tup, tubulysins D, E, F, and H).
Additionally, the N-terminal residue of Tuv is functionalized
with a highly unusual N,O-acetal substituent with different
ester groups (Scheme 1).
Scheme 1. Structure of tubulysins.
Very recently, Dömling, Wessjohann, and co-workers
described the first total synthesis of tubulysins U (1) and V
(2) (Scheme 1) without the N,O-acetal ester fragment.[9]
According to these authors, the series of tubulysins lacking
this synthetically challenging function is less active, but the
cytotoxicity of some of these tubulysins still reaches the
activity range of, for example, taxol or the epothilones. This
synthesis allowed the synthesis of just a few milligrams of the
target tubulysins. Furthermore, it is affected by several
drawbacks, such as 1) the use of a hard to cleave N-tosyl
protecting group in the synthesis of Tup; 2) several epimeri-
zations occurring at various levels in the synthesis of the Tuv
unit and peptide fragments thereof; and 3) several rather
inefficient and complex synthetic steps, such as the assembly
of the tetrapeptide framework and the final removal of the
protecting groups. Thus, we felt that a more direct, flexible,
and modular synthesis was needed that would allow the
preparation of different tubulysin stereoisomers and structur-
ally modified tubulysin-like structures.
Although at first sight the synthesis of tubulysins might
not appear to be extremely challenging, to date there are only
two published reports of their total syntheses, which suggests
that several important challenges must be present.[4] Indeed, a
number of challenging synthetic issues have been described in
a recent review by Dömling and Richter,[5] who reported that
the major hurdles are: 1) the installation of the acid and base-
labile N,O-acetal ester, 2) the synthesis of the configuration-
ally and chemically sensitive thiazole fragment, and 3) the
assembly of the peptide framework, as a consequence of the
steric congestion of the central Tuv region. However, the
research groups of Höfle and Dömling have claimed two
different total syntheses of tubulysins that are so far described
only in patents.[6] The synthesis of the Tuv, Tup, and dipeptide
fragments of tubulysins were also reported by the groups of
Wipf[7] and Friestad.[8]
We describe herein an effective synthesis of tubulysins U
(1) and V (2) that can be used to prepare hundreds of
milligrams of pure compounds. We also show that the
tubulysins U and V previously obtained by Dömling et al.[9]
are actually stereoisomers of the natural molecules.
The g-amino acid tubuphenylalanine (Tup) was synthe-
sized in a multigram amount as a mixture of epimers at the a-
carboxy stereocenter (Scheme 2). l-Phenylalaninol (3) was
transformed into the phthaloylimide 4, and then treated with
triflic anhydride to afford the corresponding triflate 5.
Nucleophilic substitution by the sterically hindered sodium
salt of a-methyl diethylmalonate produced a clean SN2
reaction to give the quaternary g-amino malonate 6. A one-
[*] Dr. M. Sani, G. Fossati, Dr. F. Huguenot, Dr. M. Zanda
C.N.R.—Istituto di Chimica del Riconoscimento Molecolare and
Dipartimento C.M.I.C.
Politecnico di Milano
Via Mancinelli 7, 20131 Milano (Italy)
Fax: (+39)02-2399-3080
E-mail: matteo.zanda@polimi.it
[**] We thank the European Commission (Integrated Project “STROMA”
LSHC-CT-2003-503233), Politecnico di Milano, and C.N.R. for
financial support. We also thank Dr. L. Malpezzi for the X-ray
analyses, and Prof. A. Mele and Dr. G. Fronza for their help with the
NMR spectroscopic analysis.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
3526
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 3526 –3529

Angewandte
Chemie
pure H-Tup-OMe·HCl (8b), which was suitable for the
subsequent coupling reactions.
The synthesis of stereochemically pure tubuvaline (Tuv) is
shown in Scheme 4. l-Cysteine (11) was condensed with
pyruvaldehyde to afford the 2-acetyl-4-ethoxycarbonylthi-
azolidine 12 which was oxidized in a multigram amount to
Scheme 2. Synthesis of epimeric Tup. a) Phthalic anhydride, Et3N,
toluene, Dean–Stark trap, 1458C, 40 h; b) (CF3SO2)2O, Py, CH2Cl2,
À788C to RT, 3 h; c) diethyl 2-methylmalonate, NaH, 08C, RT, 7 h;
d) 6n HCl, AcOH, 1458C, 2 days. Py=pyridine, Tf=trifluoromethane-
sulfonyl.
pot simultaneous cleavage of all the protecting groups and
decarboxylation provided a 1:1 epimeric mixture of H-Tup-
OH hydrochloride (7) in just four straightforward steps and
an overall yield of 68% from 3.
We performed extensive experimentation to obtain 7 in a
stereochemically pure form. Eventually, we found that the N-
Boc-protected (À)-menthyl esters 10a,b (Scheme 3) were
separable by flash chromatography. These compounds were
obtained by conversion of 7 into Boc-Tup-OMe (8), saponi-
fication of the methyl ester to give Boc-Tup-OH 9, and
conversion of the carboxylic acid epimers (9) into the (À)-
menthyl esters 10a and 10b. X-ray analysis of 10a allowed
unambiguous assignment of the stereochemistry of these
compounds.[10] Cleavage of all the protecting groups by acidic
hydrolysis gave the stereochemically pure H-Tup-OH·HCl
(7b), which was resubjected to esterification to give stereo-
Scheme 4. Synthesis of stereopure Tuv. a) Pyruvaldehyde, NaHCO3,
EtOH/H2O (1:1), 18 h, RT; b) activated MnO2, acetonitrile, 508C, 2 h;
c) isobutyraldehyde, TiCl4, Et3N, dry THF, À788C to RT; d) BocNH2,
Sn(OTf)2, acetonitrile, 3 h, RT; e) (S)-CBS, BH3·Me2S, dry THF, 08C to
RT, 2 h; f) LiOH, THF/H2O 4:1, 5 h, RT.
thiazole 13 by using MnO2.[11] The thiazole 13 gave rise to a
rather efficient crotonic condensation with isobutyraldehyde
under Lewis acid catalysis, particularly with TiCl4. The
intermediate aldol could only be detected by TLC when the
reaction was carried out at low temperature, since dehydra-
tion to the a,b-unsaturated thiazolyl ketone 14 took place
spontaneously upon heating the reaction mixture at room
temperature. The amino group was installed by an aza-
Michael addition[12] of Boc-NH2 to 14 catalyzed by tin(II)
triflate to give the b-amino ketone 15.[13]
To obtain stereochemically pure Tuv derivatives we used
the Corey, Bakshi, and Shibata oxazaborolidine (CBS cata-
lyst) mediated reduction.[14] (S)-CBS, in the presence of
BH3·Me2S, reacted nearly exclusively with the Si face of 15 to
give the diastereomeric alcohol (R,R)-16a in 90% ee from
(R)-15, and to the (S,R)-16b diastereomer in 96% ee from
(S)-15. The epimeric alcohols 16a and 16b could be easily
separated by flash chromatography. The relative stereochem-
istry could be assigned unambiguously by X-ray analysis of
racemic 16b.[10] The stereopure target Boc-Tuv-OH (17a) was
obtained after mild saponification of the methyl ester 16a.
The epimeric alcohol 17b was obtained by the same protocol.
To prove the feasibility of the method for the preparation of
the whole set of stereoisomers of Tuv, we also performed the
reaction of (R)-CBS with 15, which afforded, as expected, ent-
Scheme 3. Preparation of stereopure Tup-OMe. a) 1. 2,2-dimethoxypro-
pane, conc. HCl, dry MeOH, 608C, 1 day; 2. Boc2O, Et3N, acetonitrile,
6 h; b) LiOH, H2O/THF (1:1), 1 day; c) (À)-menthol, DCC, DMAP,
CH2Cl2, 6 h; then flash chromatography to separate 10b from 10a;
d) 6n HCl, 1308C, 1.5 h; e) 2,2-dimethoxypropane, conc. HCl, dry
MeOH, 608C, 1 day. Boc=tert-butoxycarbonyl, DCC=N,N’-dicyclo-
hexylcarbodiimide, DMAP=4-dimethylaminopyridine.
Angew. Chem. Int. Ed. 2007, 46, 3526 –3529ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3527

Communications
16a and ent-16b. The absolute stereochemistry of alcohols 16
proclivity of the Ile amino acid center to undergo epimeriza-
tion. The same problem was observed by Dömling et al. who
could not suppress the epimerization.[9] A number of different
conditions were tried, including a HATU/HOAt-promoted
coupling, but the formation of the tetrapeptide methyl ester
20 was always accompanied by at least 40% epimerization of
the Ile stereocenter. Eventually, we found that the use of ethyl
chloroformate and N-methylmorpholine (NMM) in ethyl
acetate afforded the target 20 in very good yields, with very
little epimerization (< 5%).[16] Simple flash chromatographic
purification allowed 20 to be obtained in a purity of greater
than 98%. Careful hydrolysis of the methyl ester with LiOH
gave the target tubulysin V (2) in good yield and again
without any detectable epimerization. Finally, acetylation of
the Tuv hydroxy group afforded the other target tubulysin U
(1) in a chemically and stereochemically pure form. So far, the
method has allowed the preparation of up to 100 mg of pure
tubulysin U (1) from a single reaction sequence. Not surpris-
(whose relative configuration was determined by X-ray
diffraction, see above) was assessed by chemical correlation
with the enantiopure O-Ac-Cbz-Tuv-OEt described by Wipf
et al. (see the Supporting Information).[7]
Completion of the synthesis of tubulysins U and V is
shown in Scheme 5. Initial attempts to assemble the peptide
by standard stepwise methodology, namely by coupling the
1
ingly, both the H and 13C NMR spectra did not match those
previously described in the literature,[9] which must therefore
belong to a different stereoisomer, and not to the natural
molecule.
In summary, we have developed an effective synthesis of
tubulysins U and V in chemically/stereochemically pure form
and with the correct stereochemistry. This route should allow
the synthesis of all the possible stereoisomers in amounts
suitable for a thorough investigation of the biological proper-
ties of these peptides. The efficient access to tubulysins U and
V should also allow for further structure–activity relation-
ships and biological studies, to investigate possible applica-
tions in anticancer therapy.
Scheme 5. Synthesis of tubulysins U and V. a) HOAt, HATU, TMP,
DMF, 3 h; b) CH2Cl2, 20% TFA, 1 h; c) isoBuO2CCl, NMM, AcOEt,
8 h; d) LiOH, THF/H2O (4:1), 3 d, then TFA; e) Ac2O, Py, 14 h.
HOAt=7-aza-1-hydroxy-1H-benzotriazole, HATU=N-[dimethylamino)-
1H-1,2,3-triazole[4,5-b]pyridin-1-ylmethylene]-N-methylmethanaminium
hexafluorophosphate, TMP=2,4,6-trimethylpyridine, TFA=trifluoroace-
tic acid.
Experimental Section
Synthesis of 6: Diethyl methylmalonate (6.00 mL, 34.89 mmol) was
added dropwise to a suspension of NaH (1.37 g, 60% dispersion in
mineral oil, 34.15 mmol) in dry THF (30 mL), cooled at 08C and
under nitrogen. The reaction was warmed to RT and stirred for
15 min. The resulting solution was added over a period of 30 min to a
solution of 5 (6.41 g, 15.52 mmol) in dry THF (70 mL), cooled at 08C
and in a nitrogen atmosphere. The resulting mixture was stirred for
6 h at the same temperature and then warmed at 308C for 1 h. The
reaction was quenched by adding a saturated aqueous solution of
NH4Cl (20 mL) and water (20 mL). The layers were separated and the
aqueous phase was extracted with AcOEt (4 ” 30 mL). The collected
organic layers were dried over anhydrous Na2SO4, filtered, and the
solvent was removed in vacuo. The residue was purified by flash
chromatography (n-Hex/AcOEt 3:1) to give 5.48 g of 6 as a yellowish
oil (80% yield).
Tuv-Tup dipeptide with Ile and then with Mep, failed because
the latter coupling did not occur. The coupling of the two
dipeptides Mep-Ile-OH and H-Tuv-Tup-OMe was therefore
explored. The former was obtained by standard methods (see
the Supporting Information). The coupling of Boc-Tuv-OH
(17a) with H-Tup-OMe (8b) using HATU/HOAt delivered
the dipeptide Boc-Tuv-Tup-OMe (18) without any detectable
loss of stereochemical purity and very satisfactory yields.
Interestingly, comparison of the 1H NMR spectrum of 18 (see
the Supporting Information) with that reported by Dömling
et al. under identical conditions showed subtle but clear
differences. Furthermore, polarimetric analysis showed nearly
opposite rotations, namely [a]2D3 = + 15.2 degcm3 gÀ1 dmÀ1 (c =
0.7 gcmÀ3, CHCl3) for 18, and [a]2D3 = À20.67 degcm3 gÀ1 dmÀ1
(c = 1.0 gcmÀ3, CHCl3) for Dömlingꢀs dipeptide. These find-
ings strongly suggest that the Boc-Tuv-Tup-OMe dipeptide
previously described by Dömling et al.[9] should actually be a
different stereoisomer, with the incorrect stereochemistry.[15]
Treatment of 18 with TFA delivered H-Tuv-Tup-OMe
(19) in a stereopure form and excellent yield. Unfortunately,
the key coupling between Mep-Ile-OH and H-Tuv-Tup-OMe
(19) was found to be extremely challenging, because of the
Synthesis of 15: Sn(OTf)2 (1.10 g, 2.64 mmol) and BocNH2
(1.55 g, 13.2 mmol) were added to
a solution of 14 (3.38 g,
13.2 mmol) in CH3CN (50 mL). The resulting mixture was stirred
for 3 h at RT. The solvent was removed in vacuo and the crude
product was purified by flash chromatography (n-Hex/AcOEt 8:2) to
give 2.87 g of 15 (60% yield) as a white solid.
Synthesis of 20: Trifluoroacetic acid (2 mL) was added to a
solution of 18 (350 mg, 0.64 mmol) in CH2Cl2 (8 mL). The solution
was stirred for 1 h and all the volatile compound removed in vacuo to
give 352 mg of 19. N-methylmorpholine (NMM) (90 mL, 0.82 mmol)
was added to a suspension of Mep-Ile-OH (210 mg, 0.82 mmol) in dry
AcOEt (5 mL). The mixture was cooled at À108C and isoBuO2CCl
(106 mL, 0.82 mmol) was added. After 5 min, a solution of 19 (185 mg,
3528
www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 3526 –3529

Angewandte
Chemie
0.41 mmol) in dry AcOEt (2 mL) was added. The resulting suspension
was allowed to warm to RTand the stirring was continued for 8 h. The
reaction was diluted with H2O (10 mL) and the phases were
separated. After drying the organic phase over anhydrous Na2SO4,
and subsequent filtration, the solvent was removed in vacuo. The
residue was purified by flash chromatography (CH2Cl2/MeOH 96:4)
to afford 170 mg of 20 (60% yield) as a white foam.
Synthesis of 2: A 1n aqueous solution of LiOH·H2O (184 mL,
0.18 mmol) was added to a solution of 20 (84 mg, 0.12 mmol) in THF
(5 mL). The reaction was stirred for 3 days and then acidified with
TFA to pH 1–2. The resulting mixture was washed with H2O (5 mL)
and extracted whit AcOEt (10 mL). The organic phase was dried over
anhydrous Na2SO4, filtered, and the solvent removed in vacuo to give
92 mg of pure 2 (> 98% yield) as a white foam.
McMahon, A. W. Patterson, J. A. Ellman, J.Am.Chem.Soc.
2006, 128, 16018 – 16019; see also: F. Sasse, D. Menche, Nat.
Chem.Biol. 2007, 3, 87—89.
[5] A. Dömling, W. Richter, Mol.Diversity 2005, 9, 141 – 147.
[6] For a review on the synthesis of tubulysins, see D. Neri, G.
Fossati, M. Zanda, ChemMedChem 2006, 1, 175 – 180.
[7] P. Wipf, T. Takada, M. Rishel, Org.Lett. 2004, 6, 4057 – 4060.
[8] G. K. Friestad, J.-C. MariØ, A. D. Deveau, Org.Lett. 2004, 6,
3249 – 3252.
[9] A. Dömling, B. Beck, U. Eichelberger, S. Sakamuri, S. Menon,
Q.-Z. Chen, Y. Lu, L. A. Wessjohann, Angew.Chem. 2006, 118,
7393 – 7397; Angew.Chem.Int.Ed. 2006, 45, 7235 – 7239; .
[10] Detailed information about the X-ray diffraction data will be
published in a forthcoming full paper.
Synthesis of 1: Ac2O (1 mL) was added to a solution of 2 (79 mg,
0.10 mmol) in pyridine (2 mL) and the solution was stirred overnight.
The mixture was quenched with H2O (10 mL) and extracted with
AcOEt (1 ” 10 mL). The organic phase was washed with brine (3 ”
10 mL) and the solvent removed in vacuo. The crude product was
purified by flash chromatography (CH2Cl2/MeOH 9:1) to give 70 mg
of 1 (97% yield) as a white powder.
[11] R. A. Hughes, S. P. Thompson, L. Alcaraz, C. J. Moody, J.Am.
Chem.Soc. 2005, 127, 15644 – 15651.
[12] For a recent review, see J. L. Vicario, D. Badꢁa, L. Carrillo, J.
Etxebarria, E. Reyes, N. Ruiz, Org.Prep.Proced.Int. 2005, 37,
513 – 538.
[13] For conjugate additions by carbamates, see a) L.-W. Xu, C.-G.
Xia, Eur.J.Org.Chem. 2005, 633 – 639; b) T. C. Wabnitz, J. B.
Spencer, Tetrahedron Lett. 2002, 43, 3891 – 3894; c) C. Palomo,
M. Oiarbide, R. Halder, M. Kelso, E. Gomez-Bengoa, J. M.
Garcꢁa, J.Am.Chem.Soc. 2004, 126, 9188 – 9189.
Received: November 7, 2006
Revised: February 22, 2007
Published online: April 2, 2007
[14] E. J. Corey, C. J. Helal, Angew.Chem. 1998, 110, 2092 – 2118;
Angew.Chem.Int.Ed. 1998, 37, 1986 – 2012.
[15] Careful inspection of the supporting information of the article by
Dömling et al. (Ref. [8]) reveals several obscure issues, such as
1) an exchange of stereoisomers in the hydrolysis of Boc-Tuv-
OMe to Boc-Tuv-OH and 2) strong discrepancies between the
analytical and spectroscopic data of the Tup fragment precur-
sors, with respect to the identical reference compounds de-
scribed in J. L. Vicario, D. Badꢁa, L. Carrillo, J.Org.Chem. 2001,
66, 5801 – 5807.
[16] C. Pesenti, A. Arnone, S. Bellosta, P. Bravo, M. Canavesi, E.
Corradi, M. Frigerio, S. V. Meille, M. Monetti, W. Panzeri, F.
Viani, R. Venturini, M Zanda, Tetrahedron 2001, 57, 6511 – 6522,
and references therein.
Keywords: antitumor agents · enantioselectivity ·
natural products · peptide coupling · total synthesis
.
[1] F. Sasse, H. Steinmetz, J. Heil, G. Höfle, H. Reichenbach, J.
Antibiot. 2000, 53, 879 – 885.
[2] H. Steinmetz, N. Glaser, E. Herdtweck, F. Sasse, H. Reich-
enbach, G. Höfle, Angew.Chem. 2004, 116, 4996 – 5000; Angew.
Chem.Int.Ed. 2004, 43, 4888 – 4892.
[3] M. W. Khalil, F. Sasse, H. Luensdorf, Y. A. Elnekady, H.
Reichenbach, ChemBioChem 2006, 7, 678 – 683.
[4] After submission of this paper, the first total synthesis of
tubulysin D appeared in the literature: H. M. Peltier, J. P.
Angew. Chem. Int. Ed. 2007, 46, 3526 –3529ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3529