An efficient enzymatic preparation of rhinovirus protease inhibitor intermediates

Article abstract of DOI:10.1016/j.tet.2003.10.101

The development of an efficient route for the preparation of (2S)-2-[3-{[(5-methylisoxazol-3-yl)carbonyl]amino}-2-oxopyridin-1(2H)-yl] pent-4-ynoic acid (4), a key intermediate in the synthesis of a human rhinovirus (HRV) protease inhibitor, is presented. In the presence of 40% acetonitrile, the alkaline protease from Bacillus lentus can catalyze the kinetic resolution of racemic ester 7 to afford (S)-acid 4 in 49% chemical yield/per cycle with 98% ee and >98% HPLC purity. The (R)-ester can then be readily recycled via a DBU catalyzed epimerization. The enzymatic preparation described here is superior to the existing chemical resolution route, exhibiting lower costs as well as higher yields, enantioselectivity, and substrate loads. In addition, this protease displays broad substrate specificity toward this class of compounds and can be easily extended to the preparation of other tripeptide mimetics of rhinovirus protease inhibitors.

Full text of DOI:10.1016/j.tet.2003.10.101

Tetrahedron 60 (2004) 759–764
An efficient enzymatic preparation of rhinovirus protease
inhibitor intermediates
*
Carlos A. Martinez, Daniel R. Yazbeck and Junhua Tao
Pfizer Global Research and Development, La Jolla Laboratories, 4215 Sorrento Valley Boulevard, La Jolla, CA 92121, USA
Received 20 June 2003; revised 23 July 2003; accepted 17 October 2003
Abstract—The development of an efficient route for the preparation of (2S)-2-[3-{[(5-methylisoxazol-3-yl)carbonyl]amino}-2-oxopyridin-
1(2H)-yl]pent-4-ynoic acid (4), a key intermediate in the synthesis of a human rhinovirus (HRV) protease inhibitor, is presented. In the
presence of 40% acetonitrile, the alkaline protease from Bacillus lentus can catalyze the kinetic resolution of racemic ester 7 to afford (S)-acid
4 in 49% chemical yield/per cycle with 98% ee and .98% HPLC purity. The (R)-ester can then be readily recycled via a DBU catalyzed
epimerization. The enzymatic preparation described here is superior to the existing chemical resolution route, exhibiting lower costs as well
as higher yields, enantioselectivity, and substrate loads. In addition, this protease displays broad substrate specificity toward this class of
compounds and can be easily extended to the preparation of other tripeptide mimetics of rhinovirus protease inhibitors.
q 2003 Elsevier Ltd. All rights reserved.
1. Introduction
trials. All of these molecules share the same right wing piece
(P1).2 One of the most difficult issues in the synthesis of
these compounds is to install the desired chiral center in the
tripeptide mimetics [bracketed] with high optical purity.3
Currently, no effective therapy exists to directly treat the
common cold. Since the condition is mainly caused by
rhinovirus infections, the use of inhibitors that target the 3C
protease, a protein required for viral replication, has been
reported. The design and development of substrate-derived
tripeptidyl Michael acceptor-containing human rhinovirus
protease inhibitors have been extensively investigated.1 A
series of promising compounds including Rupintrivire and
other related molecules (1–3, Scheme 1) have emerged as
the lead candidates for this ailment entering human clinical
In this paper, we report an efficient and cost-effective
enzymatic preparation of a key intermediate 4 towards the
synthesis of compound 1 (Scheme 2). The method relies on
enzymatic kinetic resolution of a racemic ester precursor
using a protease isolated from a Bacillus species strain also
known as Bacillus lentus.4 The enzyme is a very
inexpensive and commercially available serine protease,
Scheme 1.
Keywords: Rhinovirus protease inhibitor; Kinetic resolution; Enzymatic hydrolysis; Process development; Solvent engineering; Bacillus lentus protease;
Substrate recycling.
*
Corresponding author. Tel.: þ1-858-622-3228; fax: þ1-858-678-8277; e-mail address: junhua.tao@pfizer.com
0040–4020/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2003.10.101

760
C. A. Martinez et al. / Tetrahedron 60 (2004) 759–764
Scheme 2.
which is used in the detergent industry to remove protein-
based stains and in the textile industry for wool finishing.
However, in view of its ease of availability and low cost, the
enzyme appears to be underused as a preparative catalyst.
The enzymatic resolution strategy presented displays broad
substrate range and could be easily extended to the
preparation of tripeptide mimetics 5 and 6 needed for the
synthesis of other rhinovirus protease inhibitors 2 and 3.
2.1. Enzyme screening
The most suitable catalyst for the kinetic resolution of
racemic ester 7 was identified by using an unbiased
screening of commercially available hydrolases. A general
high-throughput enzyme screening method was used to
screen the enzymes.6 After initial screening using 10%
dioxane, several hydrolases were identified as hits for the
enzymatic hydrolysis of the racemic substrate 7. The active
enzymes found were pig liver esterase, subtilisin carlsberg,
Candida antarctica lipase-A, Candida antarctica lipase-B,
Aspergillus oryzae lipase, Aspergillus species protease,
Bacillus lentus protease, Bacillus subtilis protease, Asper-
gillus oryzae protease, Aspergillus melleus protease and the
acylase from Aspergillus species. Most of them showed
good reactivities but unfortunately poor enantioselectivities
(E,5). Candida antarctica lipase-B carried out the reaction
with very good enantioselectivity (.98% ee at 50%
conversion), however the reaction was very slow with
only 0.5% w/v substrate loads being possible. Although
process optimization was attempted initially with this
enzyme, no major improvements in terms of reactivity
were obtained afterwards. The substrate loads and the cost
of the enzyme could not compete with the existing chemical
process, where the racemic acid from ester 7 upon chemical
hydrolysis was resolved using chiral norephedrine (see
discussions below).
Herein, we describe the screening performed in order to
identify the enzyme, as well as process optimization carried
out to improve the yield, enantioselectivity and substrate load.
The dramatic solvent effect revealed during optimization of
enzyme selectivity as well as the details on the development of
a repetitive batch process to recycle unreacted ester with
wrong stereochemistry, will be described. Finally, enzymatic
and chemical processes will be compared.
2. Results and discussion
The development of a process for the resolution of racemic
ester 75 (Scheme 3) will be used to illustrate the strategy.
The process development involved three main steps: (1)
screen for an enzyme from an internal collection containing
most commercially available lipases, proteases and
esterases, (2) perform reaction optimization primarily to
improve enantioselectivity of the enzyme hits found and
subsequently to determine the optimal pH, temperature and
substrate loads, and (3) development of a procedure for the
regeneration of racemic ester from enriched unreacted ester
(recovered after one kinetic resolution cycle) in order to
perform a repetitive batch process with yields greater than
50%.
2.2. Optimization of enantioselectivity through the use of
organic co-solvents
The use of solvent engineering to improve the enantio-
selectivity of enzymes has been extensively well docu-
mented.6,7 A comprehensive study was thus initiated to
Scheme 3.

C. A. Martinez et al. / Tetrahedron 60 (2004) 759–764
761
Table 1. Effect of dioxane contents on reactivity and enantioselectivity of proteasesa (for the hydrolysis of substrate 7)
Enzyme
[Dioxane] (%)
ee (%)
Conversion (%)
E
Bacillus lichenformis protease (subtilisin Carlsberg)
5
5
49
45
43
41
22
1.2
4.1
6.2
7.6
6.3
35
40
45
50
48
60
66
68
Aspergillus species protease
5
4
45
12
5.6
2.4
2
1.2
6.2
5.1
6.3
4.1
35
40
45
50
70
66
72
60
Bacillus lentus protease
5
30
97
98
98
98
43
49
48
43
28
2.3
.200
.200
144
35
40
45
50
143
a
Reactions were conducted at pH 7.2, 5 mg/mL substrate 7, 10% v/v enzymes, 30 8C and 5–50% solvent content. The reactions were allowed to proceed for
1 h and then quenched and analyzed by HPLC.
study the effects of solvent content on the enantioselectivity of
the enzyme hits, which were reacted under various concen-
trations of 1,4-dioxane at constant pH 7.2. Both solvents and
solvent contents showed intriguing effect on the enantioselec-
tivity of the catalyzed hydrolysis of ester 7 by the three very
inexpensive protease hits, which displayed relatively good
reactivity (Tables 1 and 2). Subtilisin carlsberg, Aspergillus
species protease and Bacillus lentus protease exhibited an
enhanced E value as the solvent contents in the reaction
mixture was increased. Moreover, Bacillus lentus protease
(BLP) exhibited a dramatic solvent content effect, having an E
value of .200 at 40% solvent content compared to an E value
of only 2.3 at 5% 1,4-dioxane. As stated previously, this
solventeffecthasbeenfoundtobequitegeneralforavarietyof
potential processes,6 further emphasizing the need for a
routine implementation of a thorough solvent study in every
enzyme screening. Once an appropriate solvent content was
identified, other solvents were screened at that content in order
to identify potentially more reactive and environmentally
friendly solvents while retain high enantioselectivity of the
enzyme. As can be seen from Table 2, although the enzyme
was still very enantioselective in most solvents at 40% solvent
content, the reactivity differed significantly, with some
reactions not proceeding at all notably in protic solvents
(glycerol, MeOH, EtOH) and nonpolar solvents (methylene
chloride, THF, toluene and hexanes). Acetone and acetonitrile
were used in further optimization studies.
2.3. Other optimization studies
Further optimization studies were performed in order to
determine parameters such as maximum substrate load,
optimum pH and temperature. Substrate load as high as 10%
(100 mg/mL) required approximately 24 h reaction time. It
was observed that proper stirring is very important since
most of the substrate remained out of solution after
saturation at 30 g/L concentration. The fact that the reaction
was performed in a heterogeneous system was in fact
advantageous since removal of the leftover ester only
involved a simple filtration once the kinetic resolution was
complete. Under these conditions, a batch process was
chosen for simplicity. BLP is an enzyme that remains active
in a broad pH range (6–11) with greater activity at higher
pH. The background chemical hydrolysis was significant
(.1%) at pH values above 8.5. Therefore a pH of 8.25 was
optimum in terms of having maximum activity in the
absence of deleterious background hydrolysis leading to low
ee’s. The enzyme-catalyzed reaction was studied at
temperatures ranging from 20–50 8C. Temperatures above
37 8C favored background hydrolysis at the optimum pH
mentioned above. Since the reaction was fast enough, there
was no need to increase temperature above 30 8C. The
amount of enzyme to be used was also studied by varying
the amount of enzyme from 0.1–50% v/v. It was found that
10% enzyme content was the optimum protein concen-
tration with minimal complication in the workup.
Table 2. Effect of solvents on reactivity and enantioselectivity of BLPa (for
the hydrolysis of substrate 7)
Solvent
Conversion (%)
ee (%)
50% Glycerol
Methanol
Ethanol
Isopropanol
Acetonitrile
Methyl-t-butyl ether
Methylene chloride
Tetrahydrofuran
1,4-Dioxane
Acetone
,5
,5
,5
29.6
41.8
27
No reaction
No reaction
48
46
,5
,5
.99
.99
.99
99.2
98.8
98.7
—
The enzyme catalyzed kinetic resolution to afford (S)-4 was
thus optimized to reach 49% chemical yield/cycle, .98%
HPLC purity and 98% ee. BLP catalyzes the selective
hydrolysis of the (S)-ester of the racemic mixture of 7 in the
presence of 40% acetonitrile as co-solvent. After the
reaction is over, the remaining ester (R)-7 is filtered from
the reaction mixture and the acid is recovered from the
aqueous mother liquor after acidification to pH 3.5 without
further purification.
—
98.8
97.7
.99
.99
Toluene
Hexane
a
Reactions were conducted at pH 7.2, 5 mg/mL substrate 7, 10 mg/mL of
enzyme, 30 8C, and 40% solvent content. The reactions were allowed to
proceed for 1 h and then quenched and analyzed by HPLC.

762
C. A. Martinez et al. / Tetrahedron 60 (2004) 759–764
Scheme 4.
2.4. Repetitive batch process
resolve 7 using BLP, the kinetic resolution of 8 and 9 to
prepare 5 and 6 was investigated under similar conditions
optimized for substrate 7 (Scheme 5). Both methyl and ethyl
esters were studied as suitable substrates. The remarkable
solvent effect on enantioselectivity was also observed at
concentrations above 30%. Both compounds were good
substrates for the enzyme with enantiopurities of the
products 5 and 6 being .97% after half of the racemic
substrate was hydrolyzed. The base catalyzed racemization
procedure outlined above for (R)-7 could also be applied to
substrates 8 and 9, thus allowing a repetitive batch process
as well. The enzyme accommodates both methyl and ethyl
esters and displays very similar reactivities and
enantioselectivities.
Due to the limitations associated with performing a kinetic
resolution, a method for the recovery or in situ conversion of
the remaining ester (R)-7 was explored. A dynamic
resolution process was attempted, however it was not
possible to find conditions in which racemization took place
efficiently under the optimum conditions catalyzed by BLP.
However, the wrong enantiomer could be easily racemized
under non-aqueous conditions (Scheme 4), thus allowing a
repetitive batch approach for the preparation of (S)-4. The
ester was epimerized by a catalytic amount of DBU
(5 mol%) in 100% acetone or acetonitrile. The second
resolution cycle continues after the addition of enzyme and
water to the mixture under optimized conditions. The small
amount of DBU did not produce any observable effect on
the activity of the enzyme in the second cycle.
3. Summary
2.5. Scope of the method
In summary, an efficient preparation of intermediate 4 has
been described using Bacillus lentus protease (BLP). As
seen from Table 3, this enzymatic preparation is superior to
Considering the attractiveness of the above process to
Scheme 5.

C. A. Martinez et al. / Tetrahedron 60 (2004) 759–764
Table 3. Comparison of chemical and enzymatic processes for the production of (S)-4
Chemical
763
Enzymatic
Substrate
Method
rac-4
Diastereomeric salt formation (norephedrin)
rac-7
BLP catalyzed hydrolysis
Substrate load (g/L)
Yield/cycle (%)
ee of acid (%)
Catalyst load
Catalyst price (per kg)
Recyclability
100
33
96
84 g/L
100
49
98
100 ml/L
,$50
Simple: filtration-DBU treatment
$315
Complicated: CDI-base treatment
the chemical resolution approach. First of all, the enzyme
BLP is significantly cheaper than chemical resolving agent
norephedrine. In addition, the enzymatic process has much
higher yields (49% vs 33% per resolution cycle). The high
yielding makes the enzymatic route far more attractive
considering it is the penultimate step of the overall synthesis
of the drug candidate. Furthermore, removal of the leftover
ester involved simple filtration once the kinetic resolution
was complete, and efficient racemization allowed for the
development of a repetitive batch process. In contrast, the
chemical recycling is much more complicated involving
activation of the undesired acid followed by epimerization.
In the BLP catalyzed process described herein, .4000 mol
of ester are converted per mole of enzyme per batch. The
enzyme has high stability under a wide range of pH,
temperature, and high organic solvent contents (40%).
Remarkably, this bacterial enzyme also displayed broad
substrate specificity and could be extended to the prep-
aration of chiral precursors 5 and 6 needed for the synthesis
of rhinovirus protease inhibitors 2 and 3. While BLP shows
similar reactivity toward both methyl and ethyl esters
ignoring the difference among propargyl (4), ethyl (5) and
difluorobenzyl (6) substituents, it differentiates each enan-
tiomer stereochemically with kinetic perfection. The
preparation of these types of densely functionalized
optically active compounds with high molecular weight is
a difficult task that still presents a major challenge in organic
synthesis in the pharmaceutical industry. Through the use of
a very robust hydrolytic enzyme such as BLP, a challenging
chemical process has been resolved in a matter of weeks
resulting in an efficient and environmentally acceptable
route with great savings.
(Gibbstown, NJ) and were of the highest purity available.
Chiral HPLC columns used in analysis were obtained from
Chiral Technologies (Exton, PA) and Phenomenex
(Torrance, CA). Commercially available Bacillus lentus
protease was purchased from Altus Biologics (Boston, MA)
as Altus 53 in crude form. A semi-purified form of the
enzyme can also be obtained from Sigma as Savinase. This
preparation showed no significant difference in the resol-
ution, both in terms of reactivity and enantioselectivity
when compared to Altus 53 preparations. HPLC methods:
chiral method using detector wavelength: 254 nm; Chiralcel
OJ-R, 3m, C18, 4.6£100 mm; flow rate 0.5 mL/min;
injection volume: 10 mL; mobile phases: (A) 25 mM
NaH2PO4 pH 2.0; (B) Acetonitrile; isocratic: 40% B for
31 min, 3 min post run. Every HPLC sample was made by
taking 5£200 mL from the reaction slurry, then combined
and diluted with 4 ml of acetonitrile. 100 mL of that solution
were further diluted with 400 mL of acetonitrile and injected
in the HPLC.
4.2. Procedure for enzyme screening
The resolution of ester intermediate 7 was carried out as
following. A 96-well plate screening kit prepared in house6
was thawed for 5 min. 80 mL of potassium phosphate buffer
(0.1 M, pH 7.2) was then dispensed into the wells using a
multi-channel pipette. 10 mL of the substrate stock solution
(50 mg of 7/mL dioxane) was then added to each well via a
multichannel pipette, and the 96 reactions were incubated at
30 8C and 750 rpm. The reactions were sampled after 1 and
16 h by the transfer of 25 mL of the reaction mixture into a
new 96-well plate, which was then quenched by the addition
of 150 mL of acetonitrile. The 96-well plate was then
centrifuged, and the organic supernatant transferred from
each well into another 96-well plate. Sampled reactions
were then sealed using a penetrable mat cover and
transferred to an HPLC system for analysis. The same
plate was used to analyze the samples for both reactivity and
enantioselectivity using alternating columns on the HPLC
simultaneously.
4. Experimental
4.1. Materials
The majority of enzymes utilized in the preparation of
screening kits were obtained from various enzyme suppliers
including Amano (Nagoya, Japan), Roche (Basel,
Switzerland), Novo Nordisk (Bagsvaerd, Denmark), Altus
Biologics Inc (Cambridge, MA), Biocatalytics (Pasadena,
CA), Toyobo (Osaka, Japan), Sigma and Fluka (see Ref. 5
for details on preparation of screening plates, including
specific enzyme sources for each enzyme, as well as a
detailed description of the screening methodology). HPLC
analysis of the screened samples was performed on an
Agilent 220 HPLC auto sampler. Reactions were performed
in an Eppendorf thermomixer-R (VWR). Solvents utilized
during optimization were obtained from EM Science
4.2.1. Procedure for the preparation of compound 4. To a
150 mL jacketed flask equipped with a pH electrode, an
overhead stirrer and a base addition line, was added the
racemic ester 7 (10 g, 29.12 mmol, 1.00 equiv.) and acetone
(40 mL). A 718 Stat Titrino-Metrohm pH titrator (Brinkman
instruments, Inc.) was used to control the pH of the reaction.
The resulting slurry was stirred at 30 8C for 5 min. A
mixture of B. lentus protease (10 mL from commercial
Altus 53 solution) and distilled water (50 mL) was added to
the reaction flask. The suspension was then stirred at the

764
C. A. Martinez et al. / Tetrahedron 60 (2004) 759–764
same temperature for 24 h maintaining the pH of the
reaction constant at 8.2 using 1 M NaOH. Both the
conversion and ee’s of the reaction was monitored by RP-
HPLC following the product, and stopped after 50% starting
material was consumed (approximately 24 h under these
conditions). The heterogeneous mixture was filtered and the
remaining ester was recovered as white crystals, washed
once with distilled water and dried under vacuum to afford
4.81 g of (R)-ester (48% yield, .98% ee). The remaining
aqueous solution was extracted once with 50 mL of ethyl
acetate to remove any traces of starting material, acidified to
pH 3.5 with 1 M HCl, and extracted three times with 50 mL
of ethyl acetate. The acid fractions were pooled, dried with
sodium sulfate and concentrated in vacuum. The crude solid
was washed once with hot water, filtered and dried
overnight. Compound 4 was obtained as white crystals
157.3, 132.9, 130.4, 128.7, 125.2, 124.3, 107.4, 101.4, 63.7,
34.8, 12.5. Mp 152–155 8C; IR (KBr) nmax/cm21: 3347,
1718, 1650, 1594, 1535, 1460, 1286, 1213; HRMS (CI) m/z:
403.0940 (403.1033 calculated for C19H15F2N305).
Acknowledgements
The authors acknowledge Lijian Chen and Ben Borer for
providing valuable synthetic precursors as well as Jason
Ewanicki for help with NMR analysis.
References and notes
1
(4.68 g, 98% ee, 49% yield, .98% UV purity). H NMR
(300 MHz, CDCl3): d 9.44 (s, 1H), 8.31 (dd, J¼1.67,
7.38 Hz, 1H), 7.54 (dd, J¼1.72, 7.02 Hz, 1H), 6.72 (d,
J¼0.91 Hz, 1H), 6.42 (t, J¼7.17 Hz, 1H), 5.28 (dd, J¼4.55,
10.68 Hz, 1H), 3.19 (ddd, J¼2.65, 10.72, 17.46 Hz, 1H),
2.97 (ddd, J¼2.70, 4.55, 17.47 Hz, 1H), 2.84 (t, J¼2.59 Hz,
1H), 2.50 (s, 3H). 13C NMR (75 MHz, CDCl3): d 172.77,
169.17, 158.83, 157.15, 156.84, 132.88, 127.64, 123.54,
105.67, 101.60, 80.03, 73.67, 60.81, 19.37, 12.17. Mp 175–
182 8C; IR (KBr) nmax/cm21: 3301, 1752, 1645, 1542, 1459,
1221, 1188; HRMS (CI) m/z: 316.0940 (316.0933 calcu-
lated for C15H14N305).
1. (a) Webber, S. E.; Okano, K.; Little, T. L.; Reich, S. H.; Xin, Y.;
Fuhrman, S. A.; Matthews, D. A.; Love, R. A.; Hendrickson,
T. F.; Patick, A. K.; Meador, J. W., 3rd; Ferre, R. A.; Brown,
E. L.; Ford, C. E.; Binford, S. L.; Worland, S. T. J. Med. Chem.
1998, 41, 2786–2805. (b) Dragovich, P. S. Expert Opin. Ther.
Patents 2001, 11, 177–184.
2. For the synthesis of P1 lactam see: Tian, Q.; Nayyar, N. K.;
Babu, S.; Chen, L.; Tao, J.; Lee, S.; Tibbetts, A.; Moran, T.;
Liou, J.; Guo, M.; Kennedy, T. P. Tetrahedron Lett. 2001, 42,
6807.
3. For an alternative synthesis of a related chiral fragment via
enzymatic reduction see: Tao, J.; McGee, K. Org. Process Res.
Dev. 2002, 6, 520–524.
4.2.2. Data for compound 5. Compound 5 was obtained as
white crystals from the resolution of racemic ester 8 with
97% ee, 48% yield, .98% UV purity. 1H NMR (700 MHz,
CDCl3): d 9.58 (s, 1H), 8.49 (dd, J¼1.7, 7.5 Hz, 1H), 7.11
(dd, J¼1.9, 7.2 Hz, 1H), 6.82 (s-broad, 1H), 6.49 (s, 1H),
6.38 (t, J¼7.4 Hz, 1H), 5.38 (dd, J¼5.1, 10.1 Hz, 1H), 2.5
(s, 3H), 2.34 (m, 1H), 2.05 (m, 1H), 0.93 (t, J¼7.7 Hz, 3H).
13C NMR (176 MHz, CDCl3): d 172.7, 171.8, 158.7, 157.9,
157.7, 129.2, 128.6, 123.8, 107.3, 101.4, 61.5, 23.7, 12.5,
10.6. Mp 158–160 8C; IR (KBr) nmax/cm21: 3345, 1744,
1699, 1643, 1534, 11459, 1202; HRMS (CI) m/z: 305.2940
(305.2933 calculated for C14H15N305).
4. (a) Kuhn, P.; Knapp, M.; Soltis, S. M.; Ganshaw, G.; Thoene,
M.; Bott, R. Biochemistry 1998, 37, 13446–13452. (b) Davis,
B. G.; Lloyd, R. C.; Jones, J. B. J. Org. Chem. 1998, 63,
9614–9615. (c) Lloyd, R. C.; Davis, B. G.; Jones, J. B. Bioorg.
Med Chem. 2000, 8, 1537–1544. (d) Plettner, E.; DeSantis, G.;
Stabile, M. R.; Jones, J. B. J. Am. Chem. Soc. 1999, 121,
4977–4981.
5. A detailed synthesis of racemic esters 7–9 will be reported in
Chen, L. In preparation.
6. Yazbeck, D. R.; Tao, J.; Martinez, C. A.; Kline, B. J.; Hu, S.
Adv. Synth. Catal. 2003, 345, 524–532.
7. (a) Chaudhary, A. K.; Kamat, S. V.; Beckman, E. J.; Nurok, D.;
Kleyle, R. M.; Hajdu, P.; Russell, A. J. J. Am. Chem. Soc. 1996,
118, 12891–12901. (b) Overbeeke, P. L. A.; Jongejan, J. A.;
Heijnen, J. J. Biotechnol. Bioengng 2000, 70, 278–290.
(c) Lundhaug, K.; Overbeeke, P. L. A.; Jongejan, J. A.;
Anthonsen, T. Tetrahedron: Asymmetry 1998, 9, 2851–2856.
(d) Hansen, T. V.; Waagen, V.; Partali, V.; Anthonsen, H. W.;
Anthonsen, T. Tetrahedron: Asymmetry 1995, 6, 499–504.
(e) Watanabe, K.; Shin-ichi, U. J. Chem. Soc., Perkin Trans. 1
2001, 1386–1390.
4.2.3. Data for compound 6. Compound 6 was obtained as
white crystals from the resolution of racemic ester 9 with
97% ee, 45% yield, .98% purity. H NMR (700 MHz,
CDCl3): d 9.50 (s, 1H), 8.45 (dd, J¼1.5, 7.5 Hz, 1H), 7.63
(s-broad, 1H), 7.00 (m, 1H), 6.92 (m, 1H), 6.82 (dd, J¼7.2,
1.5 Hz, 1H), 6.77 (m, 1H), 6.49 (s, 1H), 6.24 (t, J¼7.2 Hz,
1H), 5.28 (dd, J¼4.9, 10.2 Hz, 1H), 3.55 (dd, J¼14.8,
5.0 Hz, 1H), 3.41 (dd, J¼14.8, 10.5 Hz, 1H), 2.50 (s, 3H).
13C NMR (176 MHz, CDCl3): d 171.9, 171.2, 158.2, 157.9,
1