Total synthesis of the antimicrotubule agent (+)-discodermolide using boron-mediated aldol reactions of chiral ketones

Article abstract of DOI:10.1002/(SICI)1521-3773(20000117)39:2<377::AID-ANIE377>3.0.CO;2-E

With a similar mechanism of action to taxol, the title compound 1 is a particularly promising candidate for development in cancer chemotherapy. This efficient synthesis, based on stereocontrolled aldol reactions, should help to overcome the scarce natural

Full text of DOI:10.1002/(SICI)1521-3773(20000117)39:2<377::AID-ANIE377>3.0.CO;2-E

COMMUNICATIONS
[6] B. Lippert, M. Leng in Topics in Biological Inorganic Chemistry, Vol. 1
(Metallopharmaceuticals 1) (Eds.: M. J. Clarke, P. J. Sadler), Springer,
Berlin, 1999, pp. 117 ± 142, and references therein.
[7] R. Manchanda, S. U. Dunham, S. J. Lippard, J. Am. Chem. Soc. 1996,
118, 5144 ± 5145.
[8] J. Schliepe, U. Berghoff, B. Lippert, D. Cech, Angew. Chem. 1996, 108,
705 ± 707; Angew. Chem. Int. Ed. Engl. 1996, 35, 646 ± 648.
[9] Removal of the 1-N-cyclohexylmethylthyminate ligand is more
efficient than reported for 1-N-methylthyminate: U. Berghoff, K.
Schmidt, M. Janik, G. Schröder, B. Lippert, Inorg. Chim. Acta 1998,
269, 135 ± 142.
[10] S. B. Rajur, J. Robles, K. Wiederholt, R. G. Kuimelis, L. W. McLaugh-
lin, J. Org. Chem. 1997, 62, 523 ± 529.
[11] Recently the synthesis of a thymine-tethered PtIV complex has been
reported: S. Ren, L. Cai, B. M. Segal, J. Chem. Soc. Dalton Trans. 1999,
1413 ± 1422. The authors present, however, no evidence sustaining the
compatibility of this complex with automated DNA synthesis.
[12] J. A. Zoltewicz, D. F. Clark, T. W. Sharpless, G. Grahe, J. Am. Chem.
Soc. 1970, 92, 1741 ± 1750.
[13] D. T. Browne, J. Eisinger, N. J. Leonard, J. Am. Chem. Soc. 1968, 90,
7302 ± 7323.
[14] C. J. L. Lock, P. Pilon, B. Lippert, Acta Crystallogr. Sect. B 1979, 35,
2533 ± 2537.
[15] O. Krizanovic, M. Sabat, R. Beyerle-Pfnür, B. Lippert, J. Am. Chem.
Soc. 1993, 115, 5538 ± 5548.
[16] M. J. Gait in An Introduction to Modern Methods of DNA Synthesis
(Ed.: M. J. Gait), IRL, Oxford, 1984, pp. 1 ± 22.
Scheme 1. Retrosynthetic analysis. Arˆ 2,6-dimethylphenyl.
resistant cells.[3] Hence, discodermolide is a particularly
promising candidate for development in cancer chemother-
apy, but its use is severely limited by its scarce supply (0.002%
w/w isolation yield)[1] from the rare sponge source.
To date, two total syntheses of the natural (‡)-discodermo-
lide and three syntheses of the antipodal ( )-discodermolide
have been reported,[4] together with various synthetic ap-
proaches.[5] Schreiber and co-workers[4a] have synthesized
both antipodes and established the absolute configuration, as
well as preparing a number of structural analogues.[4b] Despite
these impressive efforts, there is still a pressing demand for
developing a more practical and efficient synthetic route to
(‡)-discodermolide.
Total Synthesis of the Antimicrotubule Agent
(‡)-Discodermolide Using Boron-Mediated
Aldol Reactions of Chiral Ketones**
Ian Paterson,* Gordon J. Florence, Kai Gerlach, and
Jeremy P. Scott
Discodermolide (1) is a unique polyketide isolated in 1990
by Gunasekera et al. at the Harbor Branch Oceanographic
Institute, Florida, from the Caribbean sponge Discodermia
dissoluta.[1] Structurally (Scheme 1), it bears 13 stereogenic
centers, a tetrasubstituted d-lactone (C1-C5), one di- and one
tri-substituted Z-alkene, a carbamate moiety, and a terminal
Z-diene. Discodermolide displays potent activity as an anti-
mitotic agent, with a similar mechanism of action to taxol
(paclitaxel), namely by stabilizing microtubules and promot-
ing the polymerization of tubulin.[2] It inhibits the growth of
human breast cancer cells in vitro, as well as paclitaxel-re-
sistant ovarian and colon cancer cells, and other multidrug
Herein, we describe a highly convergent total synthesis,
which has the potential to provide useful quantities of (‡)-
discodermolide. Notably, our route is entirely different from
earlier syntheses and is based on a novel aldol-coupling
strategy, which also employs aldol reactions of chiral ketones
to construct the three key subunits 2, 3, and 4 (Scheme 1). Our
retrosynthetic analysis is based on a C6-C7 disconnection
which leads back to the (C1-C6) methyl ketone 2 and the enal
5. Further disassembly of 5 gives the (C9-C16) ester 3 and the
(C17-C24) aldehyde 4.
As shown in Scheme 2 the synthesis of the C1-C6 subunit 2
began with a boron-mediated anti-selective aldol reaction
between the readily available[6] ethyl ketone (S)-6 and
acetaldehyde. The intermediate aldolate was reduced in situ
with LiBH4 to give diol 7 (98%, >97%diastereoselectivity
(ds)).[6a] Protection as the bis-tert-butyldimethylsilyl (TBS)
ether 8, followed by methanolysis using catalytic CSA in
MeOH/CH2Cl2 at 08C, gave C5-alcohol 9 in 70% yield,
together with starting material and diol 7, which were recycled
accordingly. After debenzylation, diol 10 was converted into
subunit 2 by sequential double Swern oxidation, NaClO2
oxidation, and esterification. This three-step operation was
carried out without chromatography in 93% yield. The
[*] Dr. I. Paterson, G. J. Florence, Dr. K. Gerlach, Dr. J. P. Scott
University Chemical Laboratory
Lensfield Road, Cambridge CB2 1EW (UK)
Fax : (‡ 44)1223-336-362
E-mail: ip100@cus.cam.ac.uk
[**] We thank the EPSRC (Grants GR/M46686 and L41646, and student-
ships for G.J.F. and J.P.S.), the DFG (postdoctoral fellowship for K.G.),
Merck, and Novartis for support, and Dr. S. P. Gunasekera (Harbor
Branch Oceanographic Institute) for kindly providing a sample of
natural (‡)-discodermolide. We also thank Prof. A. B. Holmes
(Cambridge) for helpful discussions regarding the Claisen rearrange-
ment chemistry and Dr. P. A. Wallace for contributions to early
studies.
Angew. Chem. Int. Ed. 2000, 39, No. 2
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0570-0833/00/3902-0377 $ 17.50+.50/0
377

COMMUNICATIONS
and simultaneously introduce a carbonyl group attached to
C16, to enable a later aldol coupling.[7] The synthesis of the aryl
ester 3 started from the ethyl ketone (S)-11 (that is, the para-
methoxybenzyl (PMB) equivalent of (S)-6; available in
3 steps and 80% yield from methyl (S)-3-hydroxy-2-methyl-
propionate).[8] Using our standard conditions,[6b,c] the boron-
mediated anti-selective aldol reaction between (S)-11 and
methacrolein gave adduct 12 in high diastereoselectivity
(>97% ds) and 99% yield.
An Evans ± Tishchenko reduction with SmI2 and propio-
naldehyde formed exclusively the corresponding 1,3-anti-diol
(96%) protected as the mono-propionate 13.[9] Saponification
with K2CO3 in MeOH, followed by transacetalization with
phenylselenyl acetaldehyde diethyl acetal,[10] provided a
mixture of diastereomeric acetals 14 in 94% yield, which
were oxidized with NaIO4 and submitted directly to the
rearrangement conditions. Elimination of the selenoxide took
place in refluxing xylenes (0.01m) in the presence of DBU
giving an intermediate ketene acetal, which underwent
Claisen rearrangement smoothly to give the lactone 15 in
82% yield together with 10% of the recyclable seleno acetals
14. The exclusive formation of the Z-alkene can be attributed
to the preferred chairlike transition state shown in Scheme 3.
Next, the 8-membered lactone 15 was converted in a four-step
sequence (85% overall) by methanolysis and TBS protection
of the resultant hydroxy ester into 16, followed by saponifi-
cation and esterification with 2,6-dimethylphenol by using
Steglichꢁs protocol,[11] to give the required subunit 3 (61%
overall yield for 9 steps on a multigram scale).
Scheme 2. Synthesis of the C1-C6 subunit (2). a) 1) (cHex)2BCl, Et3N,
Et2O, 08C, 3 h; MeCHO, 78 ! 278C, 16 h; 2) LiBH4, 788C, 3 h;
3) H2O2(30%aq)/MeOH, NaOH(10%aq), 08C, 2 h; b) TBSOTf, 2,6-
lutidine, CH2Cl2, 788C, 2 h; c) CSA, MeOH/CH2Cl2, 08C, 8 h; d) 20%
Pd(OH)2/C, H2, EtOH, 208C, 20 h; e) 1) (COCl)2, DMSO, CH2Cl2, 788C,
1.5 h; 2) Et3N, 78 ! 208C, 20 min; f) NaClO2, NaH2PO4, 2-methyl-2-
butene, tBuOH, H2O, 208C, 2 h; g) CH2N2, Et2O, 208C, 5 min. Bn ˆ benzyl,
cHex ˆ cyclohexyl, TBSOTf ˆ tert-butyldimethylsilyl trifluoromethanesul-
fonate, CSA ˆ camphorsulfonic acid.
synthesis can be conveniently performed on a multigram scale
with the overall yield of 2 being 62% over the seven steps.
It was of special interest to have a solution in hand to
establish the Z-configuration of the C13-C14 trisubstituted
alkene in discodermolide with complete selectivity
(Scheme 3). Commonly used methods such as Wittig olefina-
tions give varying degrees of selectivity. On the basis of the
work of Holmes et al., a Claisen-type ring expansion was used
advantageously to achieve this alkene substitution pattern
The preparation of the remaining C17-C24 subunit 4 is shown
in Scheme 4. The boron-mediated aldol reaction between the
lactate-derived ketone (S)-17 and aldehyde 18 gave the
corresponding adduct 19 in excellent yield and with anti
selectivity (99%, >97%ds).[12] Protection as the PMB ether
Scheme 3. Synthesis of the C9-C16 subunit (3). a) 1) (cHex)2BCl, Et3N,
ˆ
Et2O, 3 h; CH2 C(Me)CHO,
78 ! 278C, 16 h; 2) H2O2(30%aq)/
MeOH, pH7 buffer, 08C, 2 h; b) SmI2, EtCHO, THF, 108C, 2.5 h;
c) K2CO3, MeOH, 208C, 3 h; d) PhSeCH2CH(OEt)2, toluene, cat. PPTS,
reflux, 4.5 h; e) NaIO4, NaHCO3, MeOH/H2O, 208C, 2 h; f) DBU,
Scheme 4. Synthesis of the C17-C24 subunit (4). a) 1) (cHex)2BCl, Et3N,
Et2O, 3 h; 18, 78 ! 278C, 16 h; 2) H2O2(30%aq)/MeOH, 08C, 1 h;
b) PMBTCA, cat. TfOH, THF, 208C, 9 h; c) LiAlH4, THF 78 ! 278C,
3 h; d) NaIO4, MeOH, 208C, 30 min; e) 1) CrCl2, THF, 208C, 16 h; 2) KH,
THF, 08C, 1.5 h; f) CSA (0.2 equiv), MeOH/CH2Cl2, 208C, 5 h; g) DMP,
CH2Cl2, 208C, 15 min. Bz ˆ benzoyl, PMBTCA ˆ para-methoxybenzyl
trichloroacetimidate, TfOH ˆ trifluoromethanesulfonic acid, DMP ˆ
Dess ± Martin periodinane.
ˆ
CH2 C(OMe)OTBS, xylenes, reflux, 6 h; g) NaOMe, MeOH, 08C, 1 h;
h) TBSOTf, 2,6-lutidine, CH2Cl2, 788C, 1.5 h; i) KOH(1m aq), MeOH,
reflux, 1 h; j) 2,6-dimethylphenol, DCC, 4-DMAP, CH2Cl2, 208C, 16 h.
PPTS ˆ pyridinium para-toluenesulfonate, DBU ˆ 1,8-diazabicyclo[5.4.0]-
undec-7-ene, DCC ˆ N,N'-dicyclohexylcarbodiimide, 4-DMAP ˆ 4-N,N-di-
methylaminopyridine.
378
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0570-0833/00/3902-0378 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2000, 39, No. 2

COMMUNICATIONS
under acidic conditions gave 20, which was reduced with
LiAlH4 to provide diols 21. NaIO4-mediated oxidative
cleavage gave the aldehyde 22 in preparation for the
introduction of the Z-diene moiety. This sequence was
achieved in a two-step procedure, following our previously
established protocol by using a Nozaki ± Hiyama reaction.[5f]
Addition of the aldehyde 22 and allylic bromide 23 to a
suspension of CrCl2 in THF produced an intermediate b-
hydroxy silane, which on treatment with KH underwent a
Peterson-type syn elimination to generate the required Z-
diene 24 in 74% yield (Z/E >98:2). Deprotection with CSA
in MeOH/CH2Cl2 (94%), followed by Dess ± Martin oxida-
tion, gave aldehyde 4. The overall yield of 4 is 54% for 8 steps
carried out on a multigram scale.
Now, having all three major subunits in hand, we turned to
coupling these together while at the same time installing the
three stereogenic centers at C7, C16, and C17 (Scheme 5). The
Heathcock-type ester 3 was enolized with LiTMP at 1008C
to give exclusively the E-enolate.[13] The subsequent anti-
selective aldol reaction with the C17-C24 aldehyde 4 gave the
Felkin ± Anh adduct 25 in 69% yield and high diastereose-
lectivity (>97%ds). A three-step sequence of reduction with
LiAlH4, selective sulfonation in the presence of the secondary
C19-alcohol, followed by an additional LiAlH4 reduction, gave
the deoxygenated product 26 with the C16-methyl group in
place (78% yield; all reactions gave 95 ± 100% yield based on
recovered starting materials). Protection of the remaining
alcohol with a TBS group and a double deprotection with
DDQ provided the diol 27 (86% yield). Selective oxidation of
the primary alcohol in the presence of the secondary C19-
alcohol was achieved using TEMPO in CH2Cl2 at 208C.[14] The
crude aldehyde obtained was converted directly into the Z-
enoate 28 by using the Still modification of the Horner-
Wadsworth-Emmons procedure (79% yield for two steps).[15]
At this stage, the carbamate moiety was installed following a
modification of the Kocovsky protocol.[4e, 16] A chemoselective
reduction using DIBAL-H at 788C and subsequent Dess ±
Martin oxidation produced the enal 5.
The final coupling step between methyl ketone 2 and
aldehyde 5 required the correct introduction of the C7-
stereocenter. Model studies had revealed that boron-medi-
ated aldol reactions with 2 and g-chiral Z-enals, related
structurally to 5, gave unexpectedly high levels of 1,4-stereo-
induction from the aldehyde component in the undesired
sense.[17] Gratifyingly, the inherent facial bias of the Z-enal 5
was overturned by carrying out this boron aldol coupling
under reagent control using (‡)-Ipc2BCl to give 77%
diastereoselectivity in favor of the desired (7S)-adduct 29
(52%).[18] Notably, this is the first case in which this reagent
system has been used successfully in a triple asymmetric
induction situation to achieve the desired stereocontrol in a
mismatched aldol coupling.[19] A hydroxyl-directed reduc-
tion[20] of 29 with Me4NBH(OAc)3 provided the 1,3-anti-diol
30 together with pre-lactonized 31 in quantitative combined
yield. Finally, global deprotection and concomitant lactoniza-
tion was achieved in one pot by reaction of 30 and 31 with
HF ´ py in THF to provide (‡)-discodermolide (1) in 85%
yield. The spectroscopic data of the synthetic material were in
Scheme 5. Subunit coupling and total synthesis of (‡)-discodermolide (1). a) 3, LiTMP, LiBr, THF, 1008C; 4, 7 min; b) LiAlH4, THF, 308C, 3 h; c) 2,4,6-
Me3(C6H2)SO2Cl, Et3N, CH2Cl2, 208C, 24 h; d) LiAlH4, THF, 108C, 3 h; e) TBSOTf, Et3N, CH2Cl2, 208C, 24 h; f) DDQ, CH2Cl2/pH7 buffer, 208C, 16 h;
g) cat. TEMPO, BAIB, CH2Cl2, 208C, 4h; h) (CF3CH2O)2P(O)CH2CO2Me, [18]c-6, K2CO3, PhMe, 20 !08C, 2 h; i) 1) Cl3CC(O)NCO, CH2Cl2, 208C, 1 h;
2) K2CO3, MeOH, 208C, 1 h; j) DIBAL-H, CH2Cl2, 788C, 1.5 h; k) DMP, CH2Cl2, 208C, 1 h; l) 1) 2, (‡)-Ipc2BCl, Et3N, Et2O, 1.5 h; 5, 78 ! 278C, 16 h;
2) H2O2(30%aq)/MeOH, pH7 buffer, 08C, 1 h; m) Me4NBH(OAc)3, MeCN/AcOH, 25 !208C, 16 h; n) HF ´ py, THF, 208C, 16 h. LiTMP ˆ lithium
2,2,6,6-tetramethylpiperidide, DDQ ˆ 2,3-dichloro-5,6-dicyano-1,4-benzoquinone, TEMPO ˆ 2,2,6,6-tetramethyl-1-piperidinoxyl, BAIB ˆ [bis(acetoxy)io-
do]benzene, [18]c-6 ˆ [18]crown-6, KHMDS ˆ potassium bis(trimethylsilyl)amide, DIBAL-H ˆ diisobutylaluminium hydride, Ipc ˆ isopinocampheyl, py ˆ
pyridine.
Angew. Chem. Int. Ed. 2000, 39, No. 2
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0570-0833/00/3902-0379 $ 17.50+.50/0
379

COMMUNICATIONS
excellent agreement to those of a natural sample (1H and
13C NMR, IR, TLC, [a]2D0 ˆ ‡13.0 (c ˆ 1.09, MeOH).
In conclusion, this total synthesis of (‡)-discodermolide
proceeds in 27 steps and 7.7% overall yield for the longest
linear sequence starting from commercial methyl (S)-3-
hydroxy-2-methylpropionate. The three key subunits were
synthesized efficiently using boron-mediated anti-selective
aldol reactions of chiral ketones (S)-6, (S)-11, and (S)-17. This
synthesis has the potential to provide useful quantities of (‡)-
discodermolide, which will allow detailed biological evalua-
tion, as well as offering a variety of options for analogue
chemistry.
[17] I. Paterson, G. J. Florence, unpublished results.
[18] a) I. Paterson, J. M. Goodman, M. A. Lister, R. C. Schumann, C. K.
McClure, R. D. Norcross, Tetrahedron 1990, 46, 4663; b) I. Paterson,
R. M. Oballa, R. D. Norcross, Tetrahedron Lett. 1996, 37, 8581.
[19] For a review of asymmetric aldol reactions using boron enolates, see
C. J. Cowden, I. Paterson, Org. React. 1997, 51, 1.
[20] a) D. A. Evans, K. T. Chapman, E. M. Carreira, J. Am. Chem. Soc.
1988, 110, 3560; b) A. K. Saksena, P. Mangiaracina, Tetrahedron Lett.
1983, 24, 273.
Received: September 13, 1999 [Z14006]
Cofactor-Bound Cross-Linked Enzyme Crystals
(CLEC) of Alcohol Dehydrogenase**
[1] S. P. Gunasekera, M. Gunasekera, R. E. Longley, G. K. Schulte, J. Org.
Chem. 1990, 55, 4912 (Corrigendum: J. Org. Chem. 1991, 56, 1346).
[2] a) E. ter Haar, R. J. Kowalski, E. Hamel, C. M. Lin, R. E. Longley,
S. P. Gunasekera, H. S. Rosenkranz, B. W. Day, Biochemistry 1996, 35,
243; b) S. L. Schreiber, J. Chen, D. T. Hung, Chem. Biol. 1996, 3, 287.
[3] a) R. J. Kowalski, P. Giannakakou, S. P. Gunasekera, R. E. Longley,
B. W. Day, E. Hamel, Mol. Pharmacol. 1997, 52, 613; b) R. Balachan-
dran, E. ter Haar, M. J. Welsh, S. G. Grant, B. W. Day, Anti-Cancer
Drugs 1998, 9, 67.
[4] a) J. B. Nerenberg, D. T. Hung, P. K. Somers, S. L. Schreiber, J. Am.
Chem. Soc. 1993, 115, 12621; b) D. T. Hung, J. B. Nerenberg, S. L.
Schreiber, J. Am. Chem. Soc. 1996, 118, 11054; c) A. B. Smith, III, Y.
Qui, D. R. Jones, K. Kobayashi, J. Am. Chem. Soc. 1995, 117, 12011;
d) S. S. Harried, G. Yang, M. A. Strawn, D. C. Myles, J. Org. Chem.
1997, 62, 6098; e) J. A. Marshall, B. A. Johns, J. Org. Chem. 1998, 63,
7885.
[5] a) D. L. Clark, C. H. Heathcock, J. Org. Chem. 1993, 58, 5878; b) I.
Paterson, S. P. Wren, J. Chem. Soc. Chem. Commun. 1993, 1790;
c) J. M. C. Golec, S. D. Jones, Tetrahedron Lett. 1993, 34, 8159; d) P. L.
Evans, J. M. C. Golec, R. J. Gillespie, Tetrahedron Lett. 1993, 34, 8163;
e) J. M. C. Golec, R. J. Gillespie, Tetrahedron Lett. 1993, 34, 8167; f) I.
Paterson, A. Schlapbach, Synlett 1995, 498; g) M. Miyazawa, S.
Oonuma, K. Maruyama, M. Miyashita, Chem. Lett. 1997, 1191; h) M.
Miyazawa, S. Oonuma, K. Maruyama, M. Miyashita, Chem. Lett. 1997,
1193; i) J. A. Marshall, Z.-H. Lu, B. A. Johns, J. Org. Chem. 1998, 63,
817; j) D. A. Evans, D. P. Halstead, B. D. Allison, Tetrahedron Lett.
1999, 40, 4461; k) S. A. Filla, J. J. Song, L. Chen, S. Masamune,
Tetrahedron Lett. 1999, 40, 5449; l) A. M. Misske, H. M. R. Hoffmann,
Tetrahedron 1999, 55, 4315.
[6] a) I. Paterson, M. V. Perkins, Tetrahedron Lett. 1992, 33, 801; b) I.
Paterson, J. M. Goodman, M. Isaka, Tetrahedron Lett. 1989, 30, 7121;
c) I. Paterson, R. D. Norcross, R. A. Ward, P. Romea, M. A. Lister, J.
Am. Chem. Soc. 1994, 116, 11287.
[7] J. R. Harrison, A. B. Holmes, I. Collins, Synlett 1999, 972.
[8] I. Paterson, E. A. Arnott, Tetrahedron Lett. 1998, 39, 7185.
[9] D. A. Evans, A. Hoveyda, J. Am. Chem. Soc. 1990, 112, 6447.
[10] a) M. Petrzilka, Helv. Chim. Acta 1978, 61, 3075; b) J. W. Burton, J. S.
Clark, S. Derrer, T. C. Stork, J. G. Bendall, A. B. Holmes, J. Am.
Chem. Soc. 1997, 119, 7483.
[11] a) B. Neises, W. Steglich, Angew. Chem. 1978, 90, 556; Angew. Chem.
Int. Ed. Engl. 1978, 17, 522; b) G. Höfle, W. Steglich, H. Vorbrüggen,
Angew. Chem. 1978, 90, 602; Angew. Chem. Int. Ed. Engl. 1978, 17,
569.
[12] I. Paterson, D. J. Wallace, C. J. Cowden, Synthesis 1998, 639.
[13] It was crucial to lower the temperature to 1008C in order to suppress
partial decomposition of the lithium enolate. a) S. H. Montgomery,
M. C. Pirrung, C. H. Heathcock, Org. Synth. Coll. Vol. 1990, 7, 190;
b) M. C. Pirrung, C. H. Heathcock, J. Org. Chem. 1980, 45, 1727; c) I.
Paterson, Tetrahedron Lett. 1983, 24, 1311.
Nancy St. Clair, Yi-Fong Wang, and
Alexey L. Margolin*
The use of dehydrogenases in organic synthesis is often
limited by the intrinsic instability of enzymes and their
nicotinamide cofactors.[1] The protein part of the molecule can
be efficiently stabilized by several techniques such as directed
evolution,[2] immobilization,[3] and protein crystallization and
cross-linking.[4] The latter approach has turned out to be
especially efficient in producing robust and productive
biocatalysts for chemical synthesis.[5] Here, we expand this
approach to the stabilization of the cofactor part of the
dehydrogenase molecule. Horse liver alcohol dehydrogenase
(HLADH) was crystallized in the presence of reduced
nicotinamide adenine dinucleotide (NADH), and the result-
ing crystals were treated with glutaraldehyde to yield the
cross-linked enzyme crystals (CLECs). The crystallized and
cross-linked HLADH was first introduced by Lee et al., and it
demonstrated good activity (26% of that in solution) and an
increased stability of the cross-linked crystals in the presence
of zinc salts.[6] In this work, we use this system to address two
main questions: 1) Is a cofactor more stable when bound
inside the enzyme crystal? 2) Is it possible to regenerate a
cofactor using a coupled substrate system, thus making
HLADH-NADH-CLEC a useful catalyst for organic syn-
thesis?
The activity of soluble enzyme and various HLADH-CLEC
preparations was compared in the reduction of 6-methyl-5-
hepten-2-one (1) in the presence of isopropanol for cofactor
regeneration (Scheme 1). The results, presented in Table 1,
afford several conclusions. The HLADH-NADH-CLECs
exhibit higher activity when HLADH is cocrystallized with
a cofactor and an inhibitor, DMSO. In this case, the resulting
complex exhibited 64% of the activity of the soluble enzyme
in the absence of an exogenous cofactor. DMSO seems to be
[*] Dr. A. L. Margolin, N. St. Clair, Dr. Y.-F. Wang
Altus Biologics Inc.
625 Putnam Ave., Cambridge, MA 02139-4807 (USA)
Fax: (‡1)617-577-6502
[14] A. De Mico, R. Margarita, L. Parlanti, A. Vescovi, G. Piancatelli, J.
Org. Chem. 1997, 62, 6974.
[15] W. C. Still, C. Gennari, Tetrahedron Lett. 1983, 24, 4405.
[16] P. Kocovsky, Tetrahedron Lett. 1986, 27, 5521.
E-mail: margolin@altus.com
[**] This work was funded by the NIH (grant 5R44GM51781).
380
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0570-0833/00/3902-0380 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2000, 39, No. 2