Reversible Michael reaction - Enzymatic hydrolysis: A new variant of dynamic resolution [6]

Full text of DOI:10.1021/ja011811l

J. Am. Chem. Soc. 2001, 123, 11075-11076
11075
Reversible Michael Reaction-Enzymatic Hydrolysis:
A New Variant of Dynamic Resolution
Scheme 1. Kinetic Enzymatic Resolution of (R/S)-1
†
‡
Jaan A. Pesti,* Jianguo Yin, Lin-hua Zhang, and
Luigi Anzalone
Chemical Process Research and DeVelopment
Bristol-Myers Squibb Co.
Chambers Works, Deepwater, New Jersey 08023
ReceiVed July 25, 2001
Enzymatic resolution of racemates into enantiomerically en-
riched compounds is a valuable and popular technique whose
value extends from the efficient preparation of complex com-
pounds to the preparative production of pure enantiomers.1,2
However, simple kinetic enzymatic resolutions are restricted to
a maximum yield of 50% per enantiomer. More useful is the
coupling of racemization with resolution, known as a dynamic
resolution.3 The benefits are two-fold: the need to remove or
recycle (after racemization) the undesired isomer is eliminated,
and if both enantiomers are substrates for the enzyme, enanti-
oselectivity remains constant due to continuous racemization of
Scheme 2. Dynamic Enzymatic Resolution Route to (R)-2
,4
the less reactive enantiomer.5,6 Since racemization conditions are
infrequently compatible with enzyme activity, dynamic resolutions
as sodium bicarbonate in 1:1 methanol:acetonitrile. In addition,
when 1 or 3b was stirred in methanol-d with catalytic methoxide,
only the side-chain methylene group was deuterated. The simplest
explanation for this observation, as well as for the facile
racemization,15 would be equilibrium between the enolate and
are uncommon.7,8 We report here the combination of enzymatic
9
resolution with a Michael-retro-Michael tandem to achieve
racemization, an unprecedented means to produce dynamic kinetic
resolution.
16
the enone/oxime anion structures, (see Scheme 1 and further
discussion in Supporting Information). Unfortunately, reaction
conditions permitting racemization and enzymatic resolution for
1 were incompatible, precluding a dynamic resolution and
requiring the recycling of (S)-1 for efficiency.
Single enantiomers of 3-aryl-4,5-dihydroisoxazol-5-ylacetic
acid derivatives are important cores of a series of non-peptide
10
platelet GPIIb/IIIa antagonists. This work in this area has led
to roxifiban,1
1,12
our leading candidate in development as therapy
Drueckhammer and co-workers have recently demonstrated that
thioesters enhance the acidity of the R-protons when compared
to corresponding oxoesters, sometimes increasing the racemization
rate sufficiently to lead to dynamic enzymatic resolution under
mild reaction conditions.1 As part of a study of new syntheses
of roxifiban, we discovered that the conversion of 1 to a wide
variety of thioesters 3 is efficient.19 In turn, some of these
thioesters, in the presence of phosphate buffer, amine, lipase PS-
for a range of cardiovascular disorders arising from undesired
platelet adhesion. The aryl isoxazoline 1 (Scheme 1) is resolved
by the lipase Pseudomonas cepacia (Amano PS-30) in pH 8
phosphate buffer to produce (R)-2 in 93% available yield and
7,18
13,14
9
5% enantiomeric excess (ee
p
).
Of note is that the unreactive
S isomer can be subsequently racemized by conditions as mild
*
Corresponding author. Current address: Bristol-Myers Squibb Co.,
Experimental Station, E336, Wilmington, DE 19880. Telephone: 302-695-
3
0, and surfactant, could be hydrolyzed to the acid (R)-2 in >90%
p
3
189. Fax: 302-695-3167.
†
Schering-Plough Research Institute, Union, New Jersey.
Gilead Sciences Inc., Foster City, California.
ee and yields as high as 89%, clear evidence of a dynamic
‡
resolution (Scheme 2).
(
1) Sheldon, R. A. Chirotechnology: Industrial Synthesis of Optically ActiVe
To optimize this unique resolution, we examined the reaction
of a set of common lipases upon various thioesters 3. Thioesters
are rarely used as substrates for enzymatic resolution, and the
best choice of R was not predictable.20 Of the enzymes and
thioesters screened, only the combination of PS-30 with the
Compounds; Marcel Dekker: New York, 1993; Chapter 7.
(
2) Wong, C.-H.; Whitesides, G. M. In Enzymes in Synthetic Organic
Chemistry; Baldwin, J. E., Magnus, P. D., Eds.; Tetrahedron Organic Chemistry
Series, Vol. 12; Pergamon Press: Oxford, 1994.
(
3) Caddick, S.; Jenkins, K. Chem. Soc. ReV. 1996, 25, 447-456.
(
4) Derivation of the kinetics of dynamic kinetic resolution: Kitamura, M.;
Tokunaga, M.; Noyori, R. J. Am. Chem. Soc. 1993, 115, 144-152.
(
(
5) Fulling, G.; Sih, C. J. J. Am. Chem. Soc. 1987, 109, 2845-2846.
6) Chen, C. S.; Fujimoto, Y.; Girdaukas, G.; Sih, C. J. J. Am. Chem. Soc.
(15) For examples of Michael-retro-Michael equilibria with the capacity
for racemization, see ref 9 and: (a) Shieh, W.-C.; Carlson, J. A. J. Org. Chem.
1994, 59, 5463-5465. (b) Reddy, S. M.; Walborshy, H. M. J. Org. Chem.
1986, 51, 2605-2607. (c) Yamagishi, M.; Tamada, Y.; Oszki, K.; Da-te, T.;
Okamura, K.; Suzuki, M.; Matsumoto, K. J. Org. Chem. 1992, 57, 1568-
1571. (d) Harrison, J. R.; O’Brien, P.; Porter, D. W.; Smith, N. M. J. Chem.
Soc., Perkin Trans. 1 1999, 3623-3631. (e) Slosse, P.; Hootele, C. Tetrahedron
1981, 37, 4287-4294. (f) Morley, C.; Knight, D. W.; Share, A. C. J. Chem.
Soc., Perkin Trans. 1 1994, 2903-2907. (g) Kiehlmann, E.; Li, E. P. M. J.
Nat. Prod. 1995, 58, 450-455. (h) Michael, J. P.; Gravestock, D. S. Afr. J.
Chem. 1998, 51, 146-157. (i) Foo, L. Y. Phytochemistry 1987, 26, 813-
817. (j) Kolarovic, A.; Berke sˇ , D.; Baran, P.; Povazanec, F. Tetrahedron Lett.
2001, 42, 2579-2582.
1
982, 104, 7294-7299.
(
7) Ward, R. S. Tetrahedron: Asymmetry 1995, 6, 1475-1490.
(
8) Noyori, R.; Tokunaga, M.; Kitamure, M. Bull. Chem. Soc. Jpn. 1995,
6
8, 36-56.
(
9) Ho, T.-L Tandem Organic Reactions; Wiley-Interscience: New York,
1
992; pp 40-41.
(
10) Wityak, J.; Sielecki, T. M.; Pinto, D. J.; Emmett, G.; Sze, J. Y.; Liu,
J.; Tobin, E.; Wang, S.; Jiang, B.; Ma, P.; Mousa, S. A.; Wexler, R. A.; Olson,
R. E. J. Med. Chem. 1997, 40, 50-60.
(
11) Xue, C.-B.; Mousa, S. A. Drugs Future 1998, 23, 707-711.
(
12) Xue, C.-B.; Wityak, J.; Sielecki, T. M.; Pinto, D. J.; Batt, D. G.; Cain,
G. A.; Sworin, M.; Rockwell, A. L.; Rodercik, J. J.; Wang, S.; Orwat, M. J.;
Frietze, W. E.; Bostrom, L. L.; Liu, J.; Higley, C. A.; Rankin, F. W.; Tobin,
A. E.; Emmett, G.; Lalka, G. K.; Sze, J. Y.; Di Meo, S. V.; Mousa, S. A.;
Thoolen, M. J.; Racanelli, A. L.; Hausner, E. A.; Reilly, T. M.; DeGrado, W.
F.; Wexler, R. R.; Olson, R. E. J. Med. Chem. 1997, 40, 2064-2084.
(16) Similar base-induced cleavages of tetrahydro-2-furanacetic acid deriva-
tives are known: (a) Brion, F. Tetrahedron Lett. 1982, 23, 5299-5302. (b)
Keay, B. A.; Rodrigo, R. Can. J. Chem. 1983, 61, 637-639. (c) Stille, J. R.;
Grubbs, R. H. J. Org. Chem. 1989, 54, 434-444.
(17) Tan, D. S.; Gunter, M. M.; Drueckhammer, D. G. J. Am. Chem. Soc.
1995, 117, 9093-9094.
(
13) Zhang, L.-H.; Anzalone, L.; Ma, P.; Kauffman, G. S.; Storace, L.;
Ward, R. Tetrahedron Lett. 1996, 37, 4455-4458.
14) Zhang, L.-H.; Chung, J. C.; Costello, T. D.; Valvis, I.; Ma, P.;
Kauffman, G. S.; Ward, R. J. Org. Chem. 1997, 62, 2466-2470.
(18) Um, P.-J.; Drueckhammer, D. G. J. Am. Chem. Soc. 1998, 120, 5605-
(
5610.
(19) Pesti, J. A.; Yin, J.; Chung, J. Synth. Commun. 1999, 2, 3811-3820.
1
0.1021/ja011811l CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/11/2001

11076 J. Am. Chem. Soc., Vol. 123, No. 44, 2001
Communications to the Editor
Table 1. Resolution of Thioesters with P. cepacia Lipase under
the Optimized Reaction Conditions
of the unreacted thioester and analysis of the product 2 indicated
a
2
2.7% conversion (c) and 96.5% ee
formula E ) (ln[1 - c(1 + ee )])/(ln[1 - c(1 - ee
determine E as 74.1 under these nonracemizing conditions, a
maximum ee
of 97.3% is predicted28 under dynamic resolution
p
, respectively. Using the
6
p
p
)]) to
thioester
time (h)
% conversion
p
ee % (R)
3a
3b
3b
3c
45.3
22.1
70.2
47.5
85
>99
94
95.5
97.6
96.8
97.3
p
b
conditions. This is in good agreement with the value of 97.6%
ee measured at >99% hydrolysis after 22.1 h when PS-30 was
added to an otherwise identical reaction mixture.24
99
a
0
.18 M in 0.65 N sodium dihydrogen phosphate buffer, 40 °C, pH
9.2 ( 0.3 (maintained by the periodic addition of 6 N NaOH), 2.0
equiv of 25% trimethylamine in H O, Triton X-100 as surfactant and
Several aspects of this dynamic resolution underscore the
efficiency of the reaction. The solubility of 3b is measured as
)
2
-
3
7
.0 × 10 g/mL after 5 h of reaction, (the particles of 3b are
lipase PS-30 (the last two charged at 0.1 of the substrate weight).
slow to establish equilibrium with solution). The combination of
this low solubility with the 90.5 h racemization half-life would
suggest that at least several half-lives would be required to achieve
reaction completion. Yet rapid resolution to (R)-2 (sometimes as
n-alkyl thioesters: ethyl, n-propyl and n-butyl, produced an
acceptable enantioselectivity and reaction rate. In contrast to the
13
O-isobutyl ester 1, R- or â-branched thioesters, such as i-propyl,
p
brief as 20 h) in high ee still occurs. The heterogeneity of the
s-butyl, i-butyl, etc, were either not hydrolyzed by this enzyme,
reaction complicates measurements, but the fast and highly
enantioselective reaction suggests the lipase expeditiously hy-
drolyzes any solubilized (R)-3b. Another indication of a rapid
enantiospecific reaction is that the ratio of (S) to (R) thioester 3b
gradually increased to 7-8:1 once the reaction was ∼80%
complete, suggesting a significantly higher enzyme-catalyzed
hydrolysis rate versus racemization rate (otherwise comparable
rates would have soon equilibrated the 3b enantiomers). Typical
dynamic resolutions require racemization rates to at least ap-
p
or in low conversion and poor ee . Evidently, the sulfur atom
exerts influence upon the hydrolysis beyond electronic effects.
We found no racemization of the n-propyl thioester 3b in 40
°C phosphate buffer of pH 9.25 with surfactant until amines were
added. In contrast to the frequently selected base trioctylam-
ine,1
7,18,20d-f
trimethylamine was superior; only 15.2% racemiza-
tion occurred after 161 h for the former compared to 49.7%
racemization after 90.5 h for trimethylamine. This may be a
21
function of trioctylamine’s lower water solubility. This system
only minimally hydrolyzed any of the thioesters 3 (<1%) over 2
days, but once lipase was also added, this combination possessed
high enantioselectivity (Table 1) and, furthermore, would not
racemize the product (R)-2. When conducted with 3b at 0.64 mol
scale, the reaction was >99.0% complete after 47 h in 98.8% ee.
Ethanol recrystallization raised this to 99.7% ee in 89% overall
yield. The addition of 10% v/v toluene dramatically increased
the racemization rate for all thioesters. For example, 3b was 49%
racemized after only 24 h. However, with PS-30 present, the lipase
was inhibited under these biphasic conditions and after 70 h, only
proximate that of the enzymatic reaction of the fast reacting isomer
5
to attain acceptable ee
p
values. These examples demonstrate, that
if the lipase’s enantioselectivity is high and the reaction conditions
properly selected, the optical purity does not degrade even when
the desired rate order of racemization and hydrolysis is reversed.
Enzymatic dynamic resolution has been conducted using a
fundamentally different racemization mechanism (retro-Michael/
Michael reactions) from those previously known for resolutions.
Any system capable of facile, reversible cleavage two bonds distal
from an ester may be a candidate for resolution by this
combination of reactions. Much as the use of thioesters has
increased the scope of dynamic resolutions, this new combination
of racemization with enzymatic resolution adds a new dimension.
p
a 94% conversion in 96.8% ee had occurred.
A useful gauge for a resolution is the “enantiomeric ratio” (E),
a measure of the intrinsic enantioselectivity of a particular
enzymatic resolution.6 For nondynamic resolutions with large
,22
Acknowledgment. We thank Drs. Philip Ma and Robert E. Walter-
mire for helpful discussions; Drs. Robert DiCosimo, Joerg Deerberg, and
Rafael Shapiro and Professor Romas Kazlauskas for critical review of
the manuscript; Drs. Charlotte Silverman and John Blackwell for chiral
analytical methods; Mr. Peter Mattei for literature searches and Dr. Sridhar
Desikan for solid-state research.
E, the ee
p
remains high only at low conversions. In contrast,
resolution combined with racemization decouples optical purity
p
from the extent of the reaction; the ee is only dependent upon
the value of E.23 To determine E for 3b, it was reacted under the
optimized conditions but without trimethylamine so as to stop
racemization, and the reaction was halted after 7.0 h. Isolation
Supporting Information Available: Expanded discussion of isox-
azoline racemization mechanisms, the chiral LC methods, and detailed
experimental for conducting the enzymatic resolution of 3b (PDF). This
material is available free of charge via the Internet at http://pubs.acs.org.
(
20) (a) Reference 18 and refs 9 and 19-22 therein. (b) Kumar, I.; Jolly,
R. S. Org. Lett. 1999, 1, 207-209. (c) Chang, C. S.; Tsai, S. W.; Lin, C. N.
Tetrahedron: Asymmetry 1998, 9, 2799-2807. (d) Chang, C.-S.; Tsai, S,-
W.; Kuo, J. Biotechnol. Bioeng. 1999, 64, 120-126. (e) Chang, C,-S.; Tsai,
S.-W. Biochem. Eng. J. 1999, 3, 239-242. (f) Lin, C.,-N.; Tsai, S.-W.
Biotechnol. Bioeng. 2000, 69, 31-38.
JA011811L
(
21) The aqueous solubility of trioctylamine is 7.5 mg/L: Nikolaev, A.
V.; Grishin, G. M.; Kolesnikov, A. A.; Pogadaev, G. I. IzV. Sib. Otd. Akad.
p
(24) We observed that ee as well as the reaction rate varied, depending
Nauk SSSR, Ser. Khim. Nauk 1969, 2, 145-8.
upon particle size of the substrate, stir rate, vessel size, and other physical
factors. In particular, the reaction time was variable; the time to achieve >99%
conversion to (R)-2 varied from 20 to 50 h.
(
22) Stecher, H.; Faber, K. Synthesis 1997, 1-16.
(
23) Measurement of E for incompletely dissolved solid substrates is usually
complicated by mass-transfer limitations. While E for solid-to-solid kinetic
resolutions can be cautiously derived, no similar treatment for the analogous
dynamic resolution is known to us. As much of the literature of solid-to-solid
resolutions does not make allowances for such limitations, the E derived above
is still useful for comparison to its peers and indicates the efficiency of this
example. Measurements of E for heterogeneous reactions are discussed in
refs 25-27.
(25) Straathof, A. J. J.; Wolff, A.; Heijnen, J. J. J. Mol. Catal. B: Enzym.
1998, 5, 55-61.
(26) Wolff, A.; von Asperen, V.; Straathof, A. J. J.; Heijnen, J. J. Biotechnol.
Prog. 1999, 15, 216-227.
(27) Straathof, A. J. J.; Jongejan, J. A. Enzyme Microb. Technol. 1997,
21, 559-571.
(28) ee ) (E - 1)/(E + 1)
p