Biocatalytic combinatorial synthesis

Article abstract of DOI:10.1016/S0968-0896(99)00145-5

Combinatorial biocatalysis, based on a principle of the combinatorial use of biosynthetic steps rather than the combinatorial use of reagents, offers a complementary approach to combinatorial chemistry, which, used individually or in connection with synthetic organic transformations, provides access to analogues not readily accessible by chemical synthetic means alone. The issues and strategies particular to this approach are discussed. Examples are given demonstrating these principles as well as the unique advantages of achieving chemo-, regio- and stereoselectivity under mild reaction conditions that biocatalytic methods offer.

Full text of DOI:10.1016/S0968-0896(99)00145-5

Bioorganic & Medicinal Chemistry 7 (1999) 2157±2162
Biocatalytic Combinatorial Synthesis
John L. Krstenansky* and Yuri Khmelnitsky
EnzyMed, Inc., 2501 Crosspark Road, Suite C-150, Iowa City, IA 52242, USA
Received 5 October 1998
AbstractÐCombinatorial biocatalysis, based on a principle of the combinatorial use of biosynthetic steps rather than the combi-
natorial use of reagents, o€ers a complementary approach to combinatorial chemistry, which, used individually or in connection
with synthetic organic transformations, provides access to analogues not readily accessible by chemical synthetic means alone. The
issues and strategies particular to this approach are discussed. Examples are given demonstrating these principles as well as the
unique advantages of achieving chemo-, regio- and stereoselectivity under mild reaction conditions that biocatalytic methods o€er.
#
1999 Elsevier Science Ltd. All rights reserved.
Introduction
cursors or xenobiotics are modi®ed by the action of
biocatalysts to create physiologically active materials or
their intermediates or for modi®cation to deactivate or
prepare for elimination from the body. Semisynthetic
chemical approaches to natural product derivatives
have proven useful in a number of instances, but the
approach runs into diculties with labile molecules or
ones with functional groups that require protection
from the reagents needed. Biocatalytic derivatization
o€ers a number of advantages over chemical synthesis
when working on complex molecules and o€ers a general
Combinatorial synthesis has found wide utility in a
number of industries where the screening of arrays of
compounds for desired properties has proven to be an
1,2
ecient way to discover new potential products or
improve the properties of existing materials.3 Most
commonly, combinatorial synthetic methodologies
operate from a model of total synthesis from sets of
precursors systematically combined to produce groups
of product molecules.5 Particularly powerful examples
of this approach are a number of multicomponent con-
densation reactions that have been previously described.8
,4
±7
9
,10
approach towards a synthetic derivatization strategy.
The synthetic uses of biocatalysts have been reviewed
1,12
An alternative model for combinatorial synthesis is one
based on derivatization of existing molecules rather
than their total synthesis. The di€erence inherent in this
model is that arrays of reaction sequences are per-
formed on a given molecule rather than a given chem-
istry performed on an array of precursor molecules (Fig.
extensively1
and will not be covered here. The key
advantages that biocatalysis o€ers over synthetic chem-
istry approaches are the chemo-, regio- and enantios-
electivity of enzyme catalysed reactions and the ability
to work under mild reaction conditions. Even in cases
where a simple synthetic chemical procedure exists for a
given biotransformation (e.g. acylation) the chemo- and
regioselectivity of the biocatalytic approach may elim-
inate several extra synthetic steps resulting from the
need for protection/deprotection strategies in the che-
mical route.
1). This paper will examine the issues involved with this
approach, with an emphasis on biocatalytic and che-
moenzymatic reactions, give some examples of its
implementation, and describe how it complements and
can be combined with the total synthesis strategy to
create a more powerful, integrated approach to combi-
natorial chemistry.
At the risk of oversimpli®cation, biocatalysts can be
divided into two general categories: catalysts with nar-
rower speci®city, such as many of those important for
the formation of biomolecules, and catalysts with broad
speci®city, such as many of those important for the
modi®cation or degradation of biomolecules or xeno-
biotics. The former can catalyze an impressive array of
reaction types, including carbon±carbon bond forming
The derivatization approach represents much of the
chemistry that occurs in biological systems where pre-
Key words: Biocatalysis; enzyme; microbial; synthesis; combinatorial.
*
5
Corresponding author. Tel.: +1-319-626-5400; fax: +1-319-626-
410; e-mail: jkrstenansky@enzymed.com
0968-0896/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(99)00145-5

2
158
J. L. Krstenansky, Y. Khmelnitsky / Bioorg. Med. Chem. 7 (1999) 2157±2162
ticular process, and, therefore, the enzymes with rela-
tively narrow substrate speci®city become the focus of
process development and biocatalyst screening. Once
the enzyme with the desired speci®city is found, the
question of what else this or similar enzymes can do
remains largely disregarded.
Still, the very nature of the speci®city of enzymes creates
some issues with regard to the iterative use of biocata-
lysts. One issue is that of orthogonality. Figure 3 illus-
trates this issue, The modi®cation of a substrate by one
enzyme, `X', may preclude it from being a substrate for
another enzyme, `Y', while the modi®cation provided by
the other enzyme, `Y', may not preclude it from being a
substrate for the ®rst enzyme, `X'. The more tolerant the
enzymes are of structural di€erences, the less is the issue
of access to the dimodi®ed analogues. Still, with at least
two routes to each dimodi®ed compound, the opportu-
nity to make each possible combination is reasonably
good. An example of this is the iterative enzymatic
derivatization of bergenin (1) to produce 7-chloro-11-
(a-galactosyl)bergenin (3) (Fig. 4). This product can be
potentially synthesized via two di€erent enzymatic
routes: either chlorination followed by galactosylation
or galactosylation followed by chlorination. In practice,
the ®rst route does not work, since 7-chlorobergenin (2)
is not accepted by a-galactosidase as a substrate, but the
second route, initially forming 11-(a-galactosyl)bergenin
(4), gives the desired product (3).
Figure 1. Comparison of combinatorial synthetic strategies.
reactions, but they are usually fairly speci®c for the
types of molecules that they can modify. There are
e€orts underway to harness the synthetic utility of these
biosynthetic catalysts through pathway engineering
1
3,14
strategies.
The latter type of biocatalyst has natu-
rally evolved to recognize and modify a broader sub-
strate range. These less speci®c biocatalysts have wider
utility for the synthetic modi®cation of `unnatural' sub-
strates (Fig. 2).
The ability of biocatalysis to serve as a general synthetic
methodology requires the ability to routinely transform
a large number of unnatural substrates ranging from
small molecules to complex natural products. To meet
this requirement for a given transformation, a group of
enzymes with non-overlapping speci®city is usually nee-
ded to provide a reasonable expectation that a broad
range of substrates can be e€ectively modi®ed. Clearly
there are molecules possessing functional groups that
cannot be modi®ed by any existing enzyme due to their
steric or electronic nature; however, it is often not
widely appreciated just how broadly many enzymes can
be applied. One of the reasons for this is that for the
majority of synthetic applications the most desirable
biocatalysts are those that are highly speci®c for a par-
Another approach utilized in combinatorial biocatalysis
exploits the di€erences in the speci®city of enzymes to
direct the synthesis of sets of analogues. Figure 5 depicts
an example of this application. In this case both enzyme
`X' and `Y' catalyze the same reaction, but at di€erent
positions in the substrate. Enzyme `X' speci®cally
modi®es one position and enzyme `Y' modi®es the same
position as `X' and another one. By using `X' alone, one
can speci®cally modify the one position. By using `X'
then `Y', one can obtain speci®cally dimodi®ed analo-
gues of the substrate. By using `Y' alone, one can obtain
dimodi®ed analogues as well, but if the reaction incor-
porates another reagent (e.g. an acylation reaction) then
these analogues will be symmetrically substituted;
whereas in the `X' then `Y' case dissimilar groups can be
incorporated in a controlled fashion. Finally, the dimo-
di®ed analogues from either the `X' then `Y' or the `Y'
Figure 2. Comparison of high versus broad enzyme speci®city.
Figure 3. Orthogonality issues in iterative biocatalysis.

J. L. Krstenansky, Y. Khmelnitsky / Bioorg. Med. Chem. 7 (1999) 2157±2162
2159
Figure 4. An example of the signi®cance of reaction order on the production of analogues using iterative biocatalysis.
route alone can be exposed to enzyme `X' in a reverse
mode (depicted here as `X': e.g. hydrolysis versus
acylation) to give the monomodi®ed analogues in the
position where `X' does not act. An example of this use
of di€ering biocatalyst speci®cities is given in Figure 6.
sized using the lipase catalyst was used as a substrate in
subtilisin catalyzed acylation with vinyl acetate, only
1
position 4 was acylated to give 6, where R =n-propyl
1
and R =2-propene. Therefore, subtilisin can be used as
a highly regioselective catalyst for 4-acylation of 11-
substituted bergenin. Based on these observations, the
following general three-step synthetic strategy was sug-
gested for generating a library including all possible
4,11-mono- and diacylated bergenin derivatives as dis-
crete compounds. In the ®rst step, bergenin (1) is regio-
selectively acylated at the 11-position via lipase catalysis
in dry acetonitrile. Yields were from 50 to 100% in 96 h
depending on the individual acyl donor used in the
reaction. When the reaction is complete, the immobi-
lized biocatalyst is removed by ®ltration, and the 11-
monoacylated bergenin derivative, 5, is recovered by
evaporating the acetonitrile. The clean product is then
redissolved in a dry toluene±dimethylsulfoxide mixture,
and added to immobilized subtilisin along with a second
acyl donor. Because one of the positions on bergenin
reactive to subtilisin-catalyzed acylation is already
occupied (viz. position 11), the monoacylated product is
selectively acylated at position 4, thus resulting in a
homo- or hetero-4,11-diacylated bergenin derivative, 6,
depending on the acyl donor selected for the second
step. The product can be recovered by ®ltering the solid
enzyme, evaporating the solvent and extracting excess
acyl donor. Finally, the regioselectivity of lipase for the
Bergenin (1) can be selectively acylated at di€erent
1
5
positions using di€erent enzymes. For example, it was
found that the mixture of lipases Chirazymes L-2 and L-
9
(
(Boehringer), and lipases PS30 (Amano) and FAP-15
Amano) catalyze regioselective acylation of the pri-
mary hydroxyl on the bergenin molecule giving 5. On
the other hand, subtilisin Carlsberg was found to pro-
duce a diacylated bergenin with substitutions at posi-
1
2
tions 4 and 11 (6, where R =R ). Moreover, when
1
bergenin-11-butyrate (5, where R =n-propyl) synthe-
1
1-position can be used in the hydrolysis direction by
replacing the reaction solvent with acetonitrile contain-
ing 2% (v/v) water, to give a quantitative yield of the
selectively 4-monoacylated derivative, 7. As in the pre-
vious steps, the product can be isolated by ®ltering the
solid enzyme and evaporating the solvent. Using 96-well
Figure 5. Catalyst speci®city for the directed synthesis of speci®c
analogues.

2
160
J. L. Krstenansky, Y. Khmelnitsky / Bioorg. Med. Chem. 7 (1999) 2157±2162
Figure 6. Use of di€erential speci®city of enzymes catalyzing the same reaction to regiospeci®cally synthesize combinations of substitution.
polypropylene ®lter-bottom reactors, as described in
2
matically decreased, methods have been described that
allow for the synthetic utility of enzymes in organic
solvents, often leading to reactions not possible in aqu-
eous environments.1
the Experimental, an entire library of (n +2n) deriva-
tives can be made in parallel by applying this scheme
combinatorially using `n' acyl donors. These types of
approaches, taking advantage of the di€ering speci®-
cities of biocatalysts and the ability to use them in
both `forward' and `reverse' directions, add great
versatility and control to the iterative derivatization
process.
6±21
Finally, in order to use biocatalysis in a general way for
syntheses, there are operational barriers one must over-
come. In order to eciently explore the complete spec-
trum of biocatalytic reactions possible on a given
substrate using a battery of enzymatic and microbial
systems, one must have an ecient means of running
and analysing each of the possible catalytic systems for
its ability to modify the substrate. Our approach to this
problem is through the pre-optimization of each indivi-
dual catalytic system to a common platform that can be
One key barrier to using biocatalysis in a synthetic
platform is the poor aqueous solubility of many sub-
strates that one would wish to modify. While the nat-
ural environment of most enzymatic systems is aqueous,
and their activity in organic solvents is generally dra-
Figure 7. A Biginelli/laccase oxidation chemoenzymatic synthesis.

J. L. Krstenansky, Y. Khmelnitsky / Bioorg. Med. Chem. 7 (1999) 2157±2162
2161
used for both screening and synthesis.9,10 The common
platform chosen is the familiar 96-well format used in
most high throughput screening operations. The
advantage of this system is the readily available equip-
ment designed in the 96-well format for automating a
wide variety of handling processes and its direct inter-
face with the screening process. As automation and
common platform can then be applied to any type of
problem of lead discovery and lead optimization in a
variety of industries (e.g. pharmaceutical, agrochemical,
etc.) to provide novel analogues in a timely manner.
Experimental
robotics have been important for combinatorial chem-
istry and parallel synthetic organic chemistry,2
2,23
simi-
lar bene®ts can be seen with combinatorial biocatalysis.
One operational advantage in the use of many biocata-
lysts is that enzymes that might be quite di€erent and
catalyze di€erent reactions often operate optimally
within a similar temperature range. This allows for the
running of di€erent types of reactions in the same 96-
well plate in adjacent wells allowing for ¯exibility in
strategies used in automated syntheses.
Biocatalytic chlorination. The reaction mixture con-
tained 16 mM bergenin (1) or 11-(a-galactosyl)bergenin
(4) and 1.5 mg/mL chloroperoxidase from Caldar-
iomyces fumago (Sigma) in 100 mL 0.1 M phosphate
bu€er, pH 2.8. The reaction was carried out at room
temperature by slowly adding 0.1 mL of 0.36 M hydrogen
peroxide solution in the same bu€er over the period of
2 h. The reaction was then freeze-dried and the product
was extracted from the dry residue with methanol. The
product was puri®ed by ¯ash chromatography on a
silica column using ethyl acetate: chloroform:methanol
(5:4:2) as eluent. The yield was 50% for chlorination of
bergenin (2) and 20% for chlorination of 11-(a-galactosyl)
bergenin (3).
As powerful as biocatalysis is on its own, there are still
areas of synthetic chemistry that biocatalysis does not
address. One such area is the general creation of ring
systems, particularly heterocyclic core structures. Con-
siderable work in the combinatorial chemistry area has
focused on various means of creating large numbers of
derivatives of diverse core structures, both in solution
and on solid-phase. Another method that has yet to see
extensive use in the combinatorial area is microwave
Biocatalytic ꢀ-galactosylation of 1. The reaction mix-
ture contained 25 mM bergenin (1), 25 mM p-nitro-
phenyl a-galactopyranoside, and
3
units/mL a-
2
4±26
assisted organic synthesis (MAOS).
The method
galactosidase from co€ee beans (Sigma) in 70 mL
sodium phosphate bu€er, pH 6.5, containing 25% ace-
tonitrile. The reaction was carried out at room tem-
perature for 3.5 h. The reaction was then stopped by
boiling the reaction mixture for 15 min. The product
was puri®ed by ¯ash chromatography on a silica col-
umn using isopropanol: ethyl acetate: methanol (6:12:1)
as the eluent. The yield was 20% of 11-(a-galacto-
syl)bergenin (4).
uses microwave energy to induce the reaction of
reagents absorbed onto a solid material in a solvent-free
environment. This method proves to be a convenient
means of interfacing organic synthesis with a biocataly-
tic synthetic platform. MAOS can be performed in deep
96-well plates, which can then be processed using the
same equipment that is used for the biocatalytic steps.
27
Thus, the addition of chemical steps coupled to bioca-
talytic steps can lead to unique `chemoenzymatic' pro-
ducts, that would not be easily achieved by chemical or
Regioselective acylation of bergenin (1). All reactions
were performed in 96-well (2 mL/well) glass-®lled
polypropylene ®lter plates (10 mm polypropylene
®lter, Poly®ltronics, Rockland, MA). A custom sealing
clamp and septa were used to allow sampling and pre-
vent evaporation of the organic solvent during the
reaction.
biocatalytic methods alone.2
8±32
Fig. 7 outlines such a
chemoenzymatic procedure involving a microwave-
assisted Biginelli synthesis followed by laccase-catalyzed
oxidation.
Conclusion
For the ®rst acylation step, 15 mg of the immobilized
lipase mixture (equal parts of PS30, FAP-15, Chirazyme
L-2, and Chirazyme L-9) were added to each well in the
plate. Using a Cyberlab C-200 liquid handler (Brook-
®eld, CT), 25 mM of bergenin (1) and 500 mM of the
appropriate acyl donor in acetonitrile were added to the
96-well plate (®nal reaction volume 1 mL). The sealed
96-well plate reactor was then shaken (250 rpm) at
Thus, combinatorial biocatalysis o€ers
a
com-
plementary approach to combinatorial chemistry,
which, used individually or in connection with synthetic
organic transformations, provides access to analogues
not readily accessible by chemical synthetic means
alone. With access to a collection of biocatalysts with
varying speci®city, one can apply strategies to obtain
analogues with speci®c patterns of substitution without
the need to employ complex protection/deprotection
strategies. The mild conditions typically employed make
these biocatalytic approaches compatible with mole-
cules representing a broad range of molecular size and
complexity. By further adapting these biocatalytic sys-
tems to a common automated platform that interfaces
directly with modern high throughput screening meth-
ods, one can perform the needed catalyst screening and
reaction optimization steps in an ecient manner. This
ꢀ
45 C. Periodically, samples were automatically with-
drawn and analyzed by HPLC and/or high-throughput
MS (¯ow injection at ca. 1 sample/min). After 96 h the
®rst acylation reaction was completed (yields of 50±
100% depending on the individual acyl donor), the
enzyme was removed by ®ltration through the reactor
bottom, the solvent removed under vacuum using a
Savant SpeedVac Plus centrifugal evaporator with a
microplate rotor, and the excess acyl donor removed by
washing (5x) with hexane.

2
162
J. L. Krstenansky, Y. Khmelnitsky / Bioorg. Med. Chem. 7 (1999) 2157±2162
The second acylation step was performed in an identical
ꢀ
and Dr. Douglas Clark for his thoughtful reading of the
manuscript.
fashion to the ®rst step (96 h at 45 C). Subtilisin/95%
KCl, 40 mg, was added to each well in the plate and
toluene containing 5% (v/v) dimethyl sulfoxide was
added for a total reaction volume of 1 mL; 2 mM of
bergenin derivative, 5, and 50 mM of the appropriate
acyl donor were used for the second step to give 50±
References
1
. Feceik, R. A.; Frank, K. E.; Gentry, E. J.; Menon, S. R.;
Mitscher, L. A.; Telikepalli, H. Med Res. Rev. 1998, 18, 149.
. Li, J.; Murray, C. W.; Waszkowycz, B.; Young, S. C. Drug
Disc. Today 1998, 3, 105.
. Weinberg, W. H.; Jandeleit, B.; Self, K.; Turner, H. Curr.
Opin. Solid State & Materials Science 1998, 3, 104.
. Gennari, C., Nestler, H.P., Piarulli, U., Salorn, B. Liebigs
Ann. Chem. 1997, 637.
. Ne, A.; Ostresh, J. M.; Houghten, R. A. The current
status of heterocyclic combinatorial libraries. Chem. Rev.
80% yields depending on the individual acyl donor
used.
2
3
In the third hydrolysis step the same lipase mixture as in
the ®rst step above was used as a catalyst. The lipase
mixture (50 mg) was added to 1 mL of 5±20 mM solu-
tion of 4,11-diacylated bergenin, (6), in MeCN contain-
ing 2% (v/v) water. The reaction mixture was incubated
4
5
1
6
997, 97, 449.
. Spaller, M. R.; Burger, M. T.; Fardis, M.; Bartlett, P. A.
ꢀ
under shaking (250 rpm) at 45 C for 96 h. Upon com-
pletion of the reaction, the solid enzyme was removed
by centrifugation and the products, 7, were obtained in
quantitative yield and recovered by evaporating the
solvent in vacuum.
Curr. Opin. Chem. Biol. 1997, 1, 47.
. Szostak, J. W. Chem. Rev. 1997, 97, 347.
8. Armstrong, R. W.; Combs, A. P.; Tempest, P. A.; Brown, S.
7
D.; Keating, T. A. Account. Chem. Res. 1996, 29, 123.
9
. Michels, P. C.; Khmelnitsky, Y. L.; Dordick, J. S.; Clark,
D. S. Trends Biotech. 1998, 16, 210.
0. Khmelnitsky, Y. L., Michels, P. C., Dordick, J. S., Clark,
Microwave-assisted Biginelli synthesis. Aldehyde (0.075
mmol), 1,3-dicarbonyl compound (0.13 mmol), and
urea or thiourea (0.05 mmol) were dissolved in 0.2 mL
ethanol containing 2 drops of glacial acetic acid. The
solution was mixed in a glass vial with 200 mg of mon-
tmorillonite K-10 clay and the solvent was removed
under reduced pressure. The mixture was treated in a
1
D. S. In Mol. Diversity Comb. Chem.: Libr. Drug Discovery,
Conf (1996); Chaiken, I. M., Janda, K. D., Eds., American
Chemical Society, Washington, DC. 1996; pp 144±157.
1
1. Faber, K. Biotransformations in Organic Chemistry;
Springer-Verlag: Berlin, 1997.
2. Wong, C.-H.; Whitesides, G. M. Enzymes in Synthetic
1
1
300 W domestic microwave oven for 20 min at 40%
Organic Chemistry; Pergamon: Oxford, 1994.
13. Marsden, A. F. A.; Wilkinson, B.; Corttes, J.; Dunster, N.
J.; Staunton, J.; Leadlay, P. F. Science 1998, 279, 199.
power. The product was extracted with 2 mL ethanol,
and the solution was stirred overnight at room tem-
perature with 100 mg 3-aminopropyl silicagel to remove
the excess of starting reagents. The puri®ed product was
recovered by removing the solvent under reduced pres-
sure. Isolated yields were 38% (8), 64% (9), and 43%
1
4. Jacobsen, J. R.; Hutchinson, C. R.; Cane, D. E.; Khosla,
C. Science 1997, 277, 367.
5. Mozhaev, V. V.; Budde, C. L.; Rich, J. O.; Usyatinsky, A.
1
Y.; Michels, P. C.; Khmelnitsky, Y. L.; Clark, D. S.; Dordick,
J. S. Tetrahedron 1998, 54, 3971.
(10).
1
6. Koskinen, A. M. P.; Klibanov, A. M. In Enzymatic Reac-
tions in Organic Media; Blackie Academic and Professional:
London, 1996.
17. Margolin, A. L. Trends Biotech. 1996, 14, 223.
18. Okahata, Y.; Mori, T. Trends Biotech. 1997, 15, 50.
19. Dordick, J. S.; Khmelnitsky, Y. L.; Sergeeva, M. V. Curr.
Opin. Microbiol. 1998, 1, 311.
Laccase-catalyzed oxidation. The reaction mixture was
prepared by dissolving the product of the Biginelli
reaction (2 mM) and 1-hydroxybenzotriazole (HOBT, 1
mM) in 0.5 mL benzene. The reaction was started by
adding 90 mL solution of laccase from Coriolus versico-
lor (ASA Spezialenzyme, Salzgitter, Germany) in 50
mM phosphate bu€er pH 6.5. The enzyme concentra-
tion in the ®nal mixture was 7.2 syringaldazine units/
mL. The biphasic reaction mixture was incubated at
room temperature under shaking at 250 rpm for 3 days.
The reaction mixture was then diluted with 0.6 mL of
isopropanol-DMF mixture (2:1), the enzyme removed
by centrifugation, and the product recovered by remov-
ing the solvent in vacuum. Analysis of the products by
HPLC/MS showed the following conversions: 42% for
20. Khmelnitsky, Y. L.; Welch, S. H.; Clark, D. S.; Dordick, J.
S. J. Am. Chem. Soc. 1994, 116, 2647.
2
1
2
2
2
1. Khmelnitsky, Y. L.; Rich, J. O. Curr. Opin. Chem. Biol.
999, 3, 47.
2. Cargill, J. F.; Lebl, M. Curr. Opin. Chem. Biol. 1997, 1, 67.
3. Harrison, W. J. Biomol. Screen. 1997, 2, 203.
4. Caddick, S. Tetrahedron 1995, 51, 10403.
25. Bose, A. K.; Banik, B. K.; Lavlinskaia, N.; Jayararnan,
M.; Manhas, M. S. Chemtech 1997, 27, 18.
26. Galema, S. A. Chem. Soc. Rev. 1997, 26, 233.
2
7. Cotterill, I. C.; Usyatinsky, A. Y.; Arnold, J. M.; Clark, D.
S.; Dordick, L. S.; Michels, P. C.; Khmelnitsky, Y. L. Tet.
Lett. 1998, 39, 1117.
11, 73% for 12, and 34% for 13.
2
Rev. 1996, 96, 443.
8. Gijsen, H. J. M.; Qiao, L.; Fitz, W.; Wong, C. H. Chem.
Acknowledgements
2
5
3
9. Zaks, A.; Dodds, D. R. Drug Discovery Today 1997, 2,
13.
0. Patel, R. N. Adv. Appl. Microbiol. 1997, 43, 91.
We would like to thank Cheryl Budde, Dr. Alex Usya-
tinsky, Dr. Lyudmila Mozhaeva, Dr. lan Cotterill and
Sharon Chen for their e€orts in preparing the examples
31. Johnson, C. R. Acc. Chem. Res. 1998, 31, 333.
32. Roberts, S. M. J. Chem. Soc.-Perkin Trans. 1 1998, 157.