Application of biocatalysis towards asymmetric reduction and hydrolytic desymmetrisation in the synthesis of a β-3 receptor agonist

Article abstract of DOI:10.1039/c1gc15694b

Chemoenzymatic syntheses of two key intermediates in the preparation of a potent β-3 receptor agonist 1 are described. A lipase-catalysed hydrolytic desymmetrisation is employed in a new synthesis of intermediate 7, which avoids the use of alkyl-tin reagents. A second biotransformation delivers chiral chlorohydrin 5 from its parent ketone in greater enantiomeric excess than the previously-described Noyori-reduction process. A brief discussion of the enantioselectivity of a set of single-point mutants of Sporobolomyces salmonicolor aldehyde reductase in this bioreduction is also presented.

Full text of DOI:10.1039/c1gc15694b

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Green Chemistry
Cite this: Green Chem., 2011, 13, 2888
www.rsc.org/greenchem
PAPER
Application of biocatalysis towards asymmetric reduction and hydrolytic
desymmetrisation in the synthesis of a b-3 receptor agonist
Matthew Badland,a Michael P. Burns,b Robert J. Carroll,*a Roger M. Howard,a Daniel Laitya and
Nathan J. Wymerb
Received 14th June 2011, Accepted 4th August 2011
DOI: 10.1039/c1gc15694b
Chemoenzymatic syntheses of two key intermediates in the preparation of a potent b-3 receptor
agonist 1 are described. A lipase-catalysed hydrolytic desymmetrisation is employed in a new
synthesis of intermediate 7, which avoids the use of alkyl-tin reagents. A second biotransformation
delivers chiral chlorohydrin 5 from its parent ketone in greater enantiomeric excess than the
previously-described Noyori-reduction process. A brief discussion of the enantioselectivity of a set
of single-point mutants of Sporobolomyces salmonicolor aldehyde reductase in this bioreduction is
also presented.
methyl Grignard afforded 7 in high yield. Tertiary alcohol 7 was
Introduction
then coupled with the commercially available amino isoxazole 8,
The use of b-3 agonists for the treatment of a variety of diseases
prior to conversion to amine 3 by means of a Ritter reaction (see
and disorders has been of interest to medicinal chemistry for
Scheme 1). Whilst this sequence was high yielding, it relied on
many years. In particular, many of the larger pharmaceutical
a tin reagent in the second step, with the associated difficulties
companies have progressed candidates into development for
of handling, cleaning and waste disposal when utilizing such
obesity and Type 2 diabetes.1 More recently, groups have targeted
reagents.
b-3 agonists for the treatment of urinary incontinence.2 We now
wish to report an alternative method of preparation of 1, a potent
In order to tackle these limitations, we took the oppor-
tunity to apply biocatalysis to the synthesis in the form
of a ketoreductase-catalyzed asymmetric ketone reduction of
b-3 agonist for the treatment of overactive bladder (OAB).
The synthesis of 1 has been described elsewhere.3 In our
ketone 4 and a lipase-catalysed desymmetrisation-incorporating
original synthesis, the target compound 1 was synthesised
chemoenzymatic route to alcohol 7, with our results described
by coupling epoxide 2 and amine 3, which were themselves
herein.
generated by Noyori asymmetric ketone reduction and Ritter
chemistries, respectively. The route suffered from two key issues.
The first of these was the relatively low e.e. (77%) achieved
for the Noyori asymmetric reduction4 of 4 to 5. Due to the
sub-optimal enantiomeric excess, preparative reversed-phase
chiral HPLC was required for purification of chlorohydrin 5
to 99.6% e.e resulting in an overall yield of only 42%. Alterna-
tive reduction conditions, including homogeneous asymmetric
hydrogenations, CBS-type and DIP-Cl asymmetric reductions,
were attempted without any improvement.5
Results and discussion
Bioreduction of ketone 4 to (R)-chlorohydrin 5
A collection of 260 ketoreductases was screened for the biore-
duction of ketone 4 to chlorohydrin 5. Many enzymes were
identified as being highly selective for the generation of either the
desired (R)-chlorohydrin 5 or for the undesired (S)-chlorohydrin
6 (Table 1). The best candidate ketoreductases based upon high
conversion to product (>70%) and e.e. (generating >90% e.e.
(R)-5 or >99% e.e. (S)-5) included biocatalysts from commercial
sources (Codexis, Julich and Daicel) as well as ketoreductases
that had been cloned into E. coli previously by our group.
Our in-house candidates included Candida parapsilosis conju-
gated polyketone reductase C2,6,7 Candida magnoliae carbonyl
reductase S1 8,9 and single-point mutants of the Sporobolomyces
salmonicolor aldehyde reductase (SSAR),10,11 as prepared by the
reported method.12
The second issue was the use of undesirable tin chemistry in
the synthesis of alcohol 7, as detailed in Scheme 2. The original
synthesis involved a Stille coupling of isopropenyl acetate with
11 to provide ketoester, 12. Treatment of acid 13 with excess
aPfizer Worldwide Research and Development, Sandwich Laboratories,
Ramsgate Road, Sandwich, Kent, United Kingdom, CT13 9NJ.
E-mail: Robert.carroll@pfizer.com
bPfizer Worldwide Research and Development, Chemical Research and
Development, Eastern Point Road, Groton, CT, 06340, USA
2888 | Green Chem., 2011, 13, 2888–2894
This journal is
The Royal Society of Chemistry 2011
©

View Online
Scheme 1 Chemical synthesis of 1. Reagents and conditions: (a) 3-Amino-5-methylisoxazole, EtOAc, CDI, 50–55 ◦C, 80%; (b) ClCH2CN, TFA,
50 ◦C, 83%, then thiourea, AcOH, MeOH, 60 ◦C, 81%; (c) (RuCl2(Cymene)2)2, DMF, HCO2H, TEA, TBME, 25 ◦C, 42%, 99.6% e.e. (post chiral
chromatography); (d) NaOH, 2-MeTHF, 10 min, 92%; (e) 2 + 3, H2O, 70 ◦C, 48%; (f) Zn, AcOH/H2O/MeOH 10 : 10 : 1, 25 ◦C, 6 h, 93%.
enantiopurity; when the amino acid residue at position 245 was
changed from the wild-type’s glutamine (Q) to either proline
(P) or lysine (K) the enzyme’s enantioselectivity completely
reversed, switching from (S)- to (R)-selectivity, as illustrated in
Fig. 1.
Following screening, the development of a bioreduction
process employing either Julich ADH-T or a Sporobolomyces
salmonicolor aldehyde reductase mutant (either Q245P or
Q245K in particular) was attractive to us due to the availability,
Scheme 2 Initial preparation of the tertiary alcohol 21 via organotin
reagents. Reagents and conditions: (a) AcCl, MeOH, rt, 16 h, 100%; (b)
(Bu)3SnOMe, isoprenyl acetate, Pd(OAc)2, (o-tolyl)3P, PhMe, 100 ◦C,
98%; (c) LiOH (2 M), THF, 0 ◦C, 89%; (d) MeMgBr, Et2O, THF, rt,
73%.
high enantioselectivity and low cost of these biocatalysts. Our
initial investigations into a preparative bioreduction using
Julich ADH-T and NADPH co-factor recycling with glucose
dehydrogenase (Julich CDX-901) were encouraging (data not
shown). However, on preparative scale the reaction stalled at
60% conversion, despite being operated at the relatively modest
substrate loading of 10 g L-1. The use of this biocatalyst was not
pursued further.
Table 1 Ketoreductase (KRED) catalyzed reduction of 2-chloro-1-(6-
chloropyridin-3-yl)ethanone 4
Entry source
enzyme
% conversion % e.e.
Our focus shifted to Codexis KRED-130, which had shown
higher conversion to product in screening, albeit in modest
enantiomeric excess (Table 1, entry 7). A KRED-130-catalysed
reduction was carried out on 1.0 g substrate at 50 g L-1 con-
centration, employing a glucose dehydrogenase (Julich CDX-
901) co-factor recycling system and complete conversion was
observed within 5 h. The resulting (R)-chlorohydrin 5 was
isolated in a yield of 97% with a better-than-expected e.e. of
95% (Scheme 3).
1
2
3
4
5
6
7
8
Codexis Julich ADH-T
Diacel Diacel E048
83
98
92
80
74
95
95
87
98 (R)
97 (R)
95 (R)
95 (R)
95 (R)
94 (R)
90 (R)
99 (S)
99 (S)
99 (S)
99 (S)
99 (S)
99 (S)
99 (S)
99 (S)
99 (S)
99 (S)
Codexis KRED-142
Pfizer
Pfizer
SSAR Q245Pa
SSAR Q245Ka
Codexis KRED-132
Codexis KRED-130
Codexis EXP-A1E
9
Pfizer
Candida parapsilosis cpr-C1 87
10
11
12
13
14
15
16
17
Codexis EXP-B1K
Codexis KRED-107
Codexis EXP-A11
86
85
88
92
87
85
86
73
Pfizer
Candida magnoliae CRS1
Codexis EXP-B1E
Codexis EXP-B1H
Codexis EXP-A1F
Codexis Julich CDX013
a Mutants of Sporobolomyces salmonicolor aldehyde reductase. Reagents
and conditions: 10 mg mL-1 4; 40 mg mL-1 cofactor (NADH or
NADPH); ~10 mg mL-1 biocatalyst; 100 mM phosphate buffer, pH
7.0, 30 ◦C; 24 h. Reactions run in 96-well plate format with reaction
volume of 0.1 mL.
Scheme 3 Formation of the (R)-chlorohydrin 5.
Interestingly, the results obtained from our collection of
SSAR single-point mutants demonstrate the power of enzyme
engineering to affect the stereoselectivity of an enzyme. Not
only did the mutants show significant variation in product
Furthermore, we repeated the bioreduction on 1.0 g scale
(100 mL volume), this time employing our Sporobolomyces
salmonicolor aldehyde reductase mutant Q245K (1.8 g of wet
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 2888–2894 | 2889
©

View Online
Fig. 1 Selectivity of SSAR mutants at position 245 (positive values for % ee denote (R)-selectivity).
Scheme 4 Preparation of alcohol 7 by a chemoenzymatic route. Reagents and conditions: (a) H2O, HCl, 100 ◦C, 88%; (b) EtOH, H2SO4, 100 ◦C,
99%; (c) Lipozyme TL 100 L, pH 6.0 calcium acetate buffer (0.2 M), MeCN, 35 ◦C, 94%; (d) THF, MeMgBr, 0 ◦C, 54%.
cell paste, NADPH and glucose dehydrogenase co-factor recycle
system). The transformation was successful with the desired
(R)-chlorohydrin recovered in 85% yield, 98% chemical purity
(as determined by HPLC) and 97% enantiomeric excess, the
highest achieved to date for a preparative synthesis of this
compound.
We believe that the proof-of-concept scale procedures de-
scribed herein constitute a process suitable for the delivery of
at least multi-hundred gram quantities of (R)-chlorohydrin 5
upon scale-up, thus eliminating the need for a Noyori reduction
with chromatographic purification.
lases capable of selectively hydrolysing 17 to 18.15 It was noted
that small amounts of diacid 16 (<10%) were generated by some
of these hydrolases.
One of the most promising hits, Meito lipase MY (Can-
dida cylindracea lipase), was investigated at preparative scale.
Desymmetrisation of diester 17 with Meito lipase MY (100
wt% loading) was attempted under pH control (reaction main-
tained at pH 6.5 by automatic titration with 1 M NaOH).
After 21 h, analysis of the reaction liquors by HPLC showed
94% conversion to mono-acid 18. Unfortunately, work-up by
extraction with ethyl acetate at pH <4.0 provided the isolated
product in only 49% yield. The loss of material was tenta-
tively attributed to product entrainment within the biocatalyst
residues and no further investigations were pursued with this
lipase.
Novozymes Lipozyme TL 100 L (Thermomyces lanuginosus
lipase) had also been identified as a potential biocatalyst for
the desymmetrisation and milligram-scale studies confirmed
that it could be employed without product entrainment during
work up. A screen of organic co-solvents with pH 6.0 calcium
acetate buffer (0.2 M) was conducted to optimize the degree
of conversion to mono-acid 18 with suppression of di-acid 16
formation.16 A range of co-solvents were identified as being
suitable for the reaction, with acetonitrile (10–30 vol%) giving
excellent conversion as well as selectivity towards mono- over
di-acid (Fig. 2). The optimized desymmetrisation reaction in
aqueous acetonitrile was successfully carried out on 200 g scale
generating mono-acid 18 in 94% yield (Scheme 4). The route
to 18 is operationally-simple and does not require any chro-
matographic purifications or highly toxic reagents, making it
an obvious choice to replace the original tin-mediated synthetic
route.
Chemoenzymatic synthesis of alcohol 7
As an alternative to the route shown in Scheme 2, it was
envisaged that commercially available dinitrile 15 could be
the starting point for a chemoenzymatic synthesis of alcohol
7 (Scheme 4). Hydrolysis of the dinitrile to di-acid 16 was
easily achieved and subsequent esterification in the same vessel
furnished the corresponding diester 17 in 88% yield over two
steps. Successful hydrolytic desymmetrisation of 17 would pro-
vide monoester 18 which, upon treatment with excess Grignard
reagent, could be converted into dimethyl-hydroxy acid 7 in 54%
yield.
Literature precedent for a chemical desymmetrisation to 18 by
reacting the d◦iacid 16 and diester 17 together with hydrochloric
acid at 140 C was noted.13 However, a milder and more
environmentally friendly synthesis was considered desirable and
for this reason the use of a hydrolase-catalysed process was
investigated.
A general high-throughput screening protocol14 was em-
ployed, which identified fifteen commercially available hydro-
2890 | Green Chem., 2011, 13, 2888–2894
This journal is
The Royal Society of Chemistry 2011
©

View Online
Fig. 2 Co-solvent screen in the enzymatic desymmetrisation of diester 17. Reagents and conditions: 40 mg mL-1 17; 40 mg mL-1 Lipozyme TL 100L;
0.2 M Ca(OAc)2 or pH 7.0 0.1 M potassium phosphate buffer; organic co-solvent as specified, 40 ◦C, 900 RPM, 24 h.
Table 2 PMI values comparing chemical and enzymatic synthesis. All
values are kg input to kg product
Conclusions
In summary, to overcome the poor selectivity of a Noyori
asymmetric hydrogenation of 4, a highly selective and high-
yielding bioreduction was demonstrated with both a commer-
cially available and a custom-made ketoreductase. Elimination
of difficult-to-handle tin reagents through the development of
an operationally-simple chemoenzymatic route to 7 was also
achieved, employing Thermomyces lanuginosus lipase in the key
enzymatic desymmetrisation step. Process Mass Intensity (PMI)
is a common green chemistry metric, widely applied, to enable
benchmarking of processes. It is a measure of the mass of raw
materials, reagents, solvents, etc. used to provide the mass of
API synthesized, i.e. kg input/kg output. Calculation of PMI17
for the epoxide 5 via our previous chemical route3 gives a value
of 3051 (impacted by column chromatography) vs. 395 using
the above enzymatic reduction (individual reagent, solvent and
water PMI contributions are given in Table 2). The synthesis
of acid 7 under the organotin route3 results in a PMI of 214
vs. 300 using the hydrolytic desymmetrisation approach. This
surprisingly high PMI for the bio-catalytic synthesis of 7 is
predominantly the result of using relatively large amounts of
aqueous media.18 No doubt this could be optimized to improve
the efficiency of the process.
Synthesis
Reagent
Solvent
Water
Total PMI
Enzymatic route to 5
Chemical route to 5
Enzymatic route to 7
Chemical route to 7
18
33
32
35
265
2983
101
112
35
167
52
395
3051
300
127
214
Ketoreductases cloned into E. coli
Candida parapsilosis conjugated polyketone reductase C2 and
Candida magnoliae carbonyl reductase S1 were cloned into E.
coli BL21 Gold(DE3) cells (Agilent 230132) and expressed using
Overnight Express Instant TB medium (Novagen 71491) at
30 ◦C.
SSAR-Q245x mutants
SSAR-Q245x mutant preparation. The gene for the S.
salmonicolor AR2 enzyme was optimized for E coli expression
and synthesized de novo (Blue Heron, Bothell, WA, USA).
This gene was then cloned into a pET28b expression plasmid
(Stratagene, Santa Clara, CA, USA) between the 5¢-NdeI
and 3¢-EcoRI restriction sites. The mutants were prepared
using the QuickChange mutagenesis protocol (Stratagene). The
complementary DNA primers listed in the appendix were used
to prepare the mutants individually.
Both of these processes should be capable of providing
multi-hundred gram to kilogram quantities of the desired
materials.
Into each PCR tube, SSAR-pET28 plasmid (1 mL, ~50 ng mL-1
stock), primers (1 mL each, 10 mM stock), dNTP’s (1 mL,
10 mM combination stock), Pfu buffer (5 mL, 10¥ stock), Pfu
Turbo polymerase (1 mL, Stratagene) and water (40 mL). The
following PCR profile was used: initial (1 cycle): 2 min at 95 ◦C;
heat cycles (16 cycles): 30 s at 95 ◦C, 30 s at 50 ◦C, 6 min
at 68 ◦C; polish (1 cycle): 10 min at 68 ◦C; hold at 4 ◦C.
To each reaction, DpnI restriction enzyme (1 mL, Stratagene)
was added and the reaction mixture incubated at 37 ◦C for 1
h. The digested samples were then electroporated into Stbl4
Experimental
General experimental
Commercial grade reagents and solvents were purchased from
Sigma-Aldrich and used without further purification. The
commercially available ketoreductase enzymes were purchased
from Codexis (Redwood City, CA, USA) and Daicel Chemical
Industries (Japan).
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 2888–2894 | 2891
©

View Online
Fig. 3 The chiral HPLC spectrum of 5.
cells (Invitrogen, Carlsbad, CA) following normal procedures,
plated onto LB medium containing kanamycin and incubated
overnight at 37 ◦C. Single colon◦ies were picked for each mutant
and incubated overnight at 37 C/210 RPM. These overnight
colonies were submitted to plasmid DNA isolation (miniprep,
Qiagen, Hilden, Germany) and submitted for DNA sequencing.
Only those plasmids containing the proper mutations were
carried forward for expression and activity testing.
previously frozen 200 mg mL-1 suspension in 100 mM potassium
phosphate buffer, pH 7.0; and a solution of ketone 4 (1.0 g) in
DMSO (2 mL). The flask was covered with Parafilm and placed
on a rotary shaker with a 50 mm throw and shaken at 210
RPM at 30 ◦C for 4 h. The reaction mixture was extracted
with 3 volumes of MTBE and the combined organic layers
filtered through Celite. The organic layers were then dried under
reduced pressure to yield crude chlorohydrin 5 (0.85 g, 85%).
The product’s analytical data was consistent with the desired
structure and literature precedent.3
SSAR-Q245x mutant expression. The SSAR-Q245x library
plasmids were transformed using normal procedures into
chemically competent E. coli strain BL21 Gold(DE3) cells
(Stratagene) for protein expression, plated onto LB medium
containing kanamycin and incubated overnight at 37 ◦C. Single
colonies were picked for each mutant into 2 mL LB medium
containing kanamycin and incubated overnight at 37 ◦C/210
RPM. Samples (500 mL) of overnight culture were dispensed
into autoinduction medium (25 mL, Overnight Express TB
from Novagen, Gibbstown, NJ, USA) containing kanamycin
in 125 mL baffled flasks (4116-0125, Nalgene, Rochester, NY,
USA). The cultures were incubated overnight at 30 ◦C/210
RPM. To analyze protein expression, overnight culture (1 mL)
was pelleted with centrifugation and the medium decanted. The
cell pellet was lysed with BugBuster (1 mL, Novagen) and the
cell debris pelleted. The soluble protein expression was analyzed
using standard SDS-PAGE techniques with Coomassie staining.
All of the SSAR-Q245x mutants showed consistent, high-level
soluble expression. The remaining overnight expression cultures
were pelleted with centrifugation, the medium decanted and the
wet cell pellets stored at -80 ◦C until needed.
Chiral HPLC for the analysis of ketone 4 chlorohydrin 5
The bioreduction of ketone 4 to chlorohydrin 5 was followed
by chiral HPLC using a ChiralPak AS-RH, 4.6 ¥ 150 mm,
5 micron column (Chiral Technologies, Japan) eluted with
water : acetonitrile:TFA (80 : 20 : 0.1, v/v/v). Detection was by
photodiode array with quantitation using 270 nm. A rep-
resentative chromatogram showing a bioreduction using S.
salmonicolor AR Q245K is shown above (Fig. 3).
Preparation of 16
To a suitably sized vessel was charged 15 (1.25 mol, 196.5 g,
1.0 eq) followed by H2O (786 mL, 4 mL g-1) and the resulting
slurry stirred at room temperature under N2(g). To the slurry was
charged conc. HCl (~36%, 4 mL g-1, 786 mL). A small exotherm
was noted (internal temperature rise to 35 ◦C). The resulting
slurry was heated to reflux (100 ◦C) for 12 h. The reaction slurry
was cooled to room temperature and isolated by filtration. The
filter cake was washed with water (3 mL g-1, 588 mL), the solid
collected and dried in vacuum oven for 8 h at 45 ◦C to yield 16
(1.1 mol, 214.2 g, 88% yield). The product’s analytical data was
consistent with the desired structure and literature precedent.19
Preparation of 5
To a 250 mL Erlenmeyer flask was charged 80 mL of 100 mM
potassium phosphate buffer; pH 7.5, 9.5 mL of 20% aqueous
glucose solution; 210 mg NADP+ disodium (Oriental Yeast
Company, Japan, cat # 44300900); 200 mg glucose dehydroge-
nase (Codexis, Redwood City, CA, USA, cat # CDX-901); 9.0
mL of Sporobolomyces salmonicolor Q245K wet cell paste as a
Preparation of 17
To a suitably sized vessel was charged 16 (1.02 mol, 200.0 g)
followed by ethanol (5 mL g-1, 1.0 L) and the resulting slurry
stirred at room temperature under N2(g). To the slurry was added
2892 | Green Chem., 2011, 13, 2888–2894
This journal is
The Royal Society of Chemistry 2011
©

View Online
conc. H2SO4 (0.5 mL g-1, 100 mL) and an exotherm was observed
(internal temperature rise to 45 ◦C). The resulting slurry was
heated to reflux (80 ◦C) to give a clear solution. The solution was
maintained at 80 ◦C for 12 h before cooling to room temperature
and removing the volatiles in vacuo. The resulting residue was
dissolved in toluene (6 mL g-1; 1.2 L) and NaHCO3(aq) (30%
solution) was slowly added (off-gassing observed) to achieve
pH 8.0 in the aqueous layer. The solution was transferred
to a separating funnel, the aqueous layer removed and the
organic layer washed with water (4 mL g-1, 800 mL) and again
separated. The organic phase was then dried over anhydrous
MgSO4, filtered and concentrated to dryness to provide 17,
which crystallized upon standing (1.02 mol, 255.8 g, 99% yield).
The product’s analytical data was consistent with the desired
structure and literature precedent.20
filtration to yield 7 (0.31 mol, 65.2 g 54% yield). The product’s
analytical data was consistent with the desired structure and
literature precedent.20
Appendix
Complementary DNA primers used for mutant creation
Q245A
Q245Ar
Q245R
Q245Rr
Q245N
Q245Nr
Q245D
Q245Dr
Q245C
Q245Cr
Q245E
Q245Er
Q245G
Q245Gr
Q245H
Q245Hr
Q245I
CCT GCC CTT GCC TTA ATG CCC CCA GCG TAT
TAT GTC AGC GCC GTA GAT ATT GG
CCA ATA TCT ACG GCG CTG ACA TAA TAC GCT
GGG GGC ATT AAG GCA AGG GCA GG
CCT GCC CTT GCC TTA ATG CCC CCA CGT TAT
TAT GTC AGC GCC GTA GAT ATT GG
CCA ATA TCT ACG GCG CTG ACA TAA TAA CGT
GGG GGC ATT AAG GCA AGG GCA GG
CCT GCC CTT GCC TTA ATG CCC CCA AAC TAT
TAT GTC AGC GCC GTA GAT ATT GG
CCA ATA TCT ACG GCG CTG ACA TAA TAG TTT
GGG GGC ATT AAG GCA AGG GCA GG
CCT GCC CTT GCC TTA ATG CCC CCA GAC TAT
TAT GTC AGC GCC GTA GAT ATT GG
CCA ATA TCT ACG GCG CTG ACA TAA TAG TCT
GGG GGC ATT AAG GCA AGG GCA GG
CCT GCC CTT GCC TTA ATG CCC CCA TGC TAT
TAT GTC AGC GCC GTA GAT ATT GG
CCA ATA TCT ACG GCG CTG ACA TAA TAG CAT
GGG GGC ATT AAG GCA AGG GCA GG
CCT GCC CTT GCC TTA ATG CCC CCA GAG TAT
TAT GTC AGC GCC GTA GAT ATT GG
CCA ATA TCT ACG GCG CTG ACA TAA TAC TCT
GGG GGC ATT AAG GCA AGG GCA GG
CCT GCC CTT GCC TTA ATG CCC CCA GGT TAT
TAT GTC AGC GCC GTA GAT ATT GG
CCA ATA TCT ACG GCG CTG ACA TAA TAA CCT
GGG GGC ATT AAG GCA AGG GCA GG
CCT GCC CTT GCC TTA ATG CCC CCA CAC TAT
TAT GTC AGC GCC GTA GAT ATT GG
Preparation of 18
To a suitably sized vessel was charged Ca(OAc)2·H2O (0.91 mol,
160 g) followed by water (4.50 L) and stirred until a solution
was obtained. To this was added Lipozyme TL 100 L (100 mL,
Thermomyces lanuginosus lipase from Novozymes, USA) and
the resulting mixture stirred at 35 ◦C. A pre-formed solution
of 17 (0.79 mol, 200 g) in MeCN (2.5 mL g-1, 500 mL) was
added to the reaction mixture to give a hazy suspension and an
autotitrator was employed to maintain pH 6.5–6.6 with addition
of 1.0 M NaOH (~800 mL added over 12 h). Once full conversion
to mono-acid 18 was achieved, the reaction was acidified to pH
< 2.0 with the addition of 2 M HCl (~1.30 L). The reaction
mixture was extracted with EtOAc (10 mL g-1, 2.0 L) and the
resulting biphasic mixture filtered through a bed of Celite before
being transferred to a separating funnel and the lower aqueous
layer removed. The organic phase was washed with water (5 mL
g-1, 1.0 L) and then brine (5 mL g-1, 20% w/w, 1.0 L). Finally,
the aqueous layers were discarded and the organic phase was
concentrated to dryness to yield 18 (0.74 mol, 164.5 g, 94%).
The product’s analytical data was consistent with the desired
structure and literature precedent.20
CCA ATA TCT ACG GCG CTG ACA TAA TAG TGT
GGG GGC ATT AAG GCA AGG GCA GG
CCT GCC CTT GCC TTA ATG CCC CCA ATC TAT
TAT GTC AGC GCC GTA GAT ATT GG
Q245Ir
Q245L
Q245Lr
Q245K
Q245Kr
Q245M
Q245Mr
Q245F
Q245Fr
Q245P
CCA ATA TCT ACG GCG CTG ACA TAA TAG ATT
GGG GGC ATT AAG GCA AGG GCA GG
CCT GCC CTT GCC TTA ATG CCC CCA CTG TAT
TAT GTC AGC GCC GTA GAT ATT GG
CCA ATA TCT ACG GCG CTG ACA TAA TAC AGT
GGG GGC ATT AAG GCA AGG GCA GG
CCT GCC CTT GCC TTA ATG CCC CCA AAA TAT
TAT GTC AGC GCC GTA GAT ATT GG
CCA ATA TCT ACG GCG CTG ACA TAA TAT TTT
GGG GGC ATT AAG GCA AGG GCA GG
CCT GCC CTT GCC TTA ATG CCC CCA ATG TAT
TAT GTC AGC GCC GTA GAT ATT GG
CCA ATA TCT ACG GCG CTG ACA TAA TAC ATT
GGG GGC ATT AAG GCA AGG GCA GG
CCT GCC CTT GCC TTA ATG CCC CCA TTC TAT
TAT GTC AGC GCC GTA GAT ATT GG
Preparation of 7
To a suitably sized vessel was charged 18 (0.67 mol, 150.0 g)
followed by anhydrous THF (10 mL g-1, 1.5 L) and the resultant
mixture stirred until a solution was obtained. The solution
was cooled to 0 ◦C and MeMgBr in 75% toluene–25%THF
(1.4 M, 2.69 mol, 4 eq.) added dropwise whilst maintaining
the temperature below 10 ◦C. During the addition the solution
became a thick suspension and an exotherm and gas evolution
were observed. After complete addition of the Grignard reagent,
the reaction was stirred for a further 30 min, maintaining the
reaction temperature below 10 ◦C. The reaction was quenched
by slow addition of HCl (2.0 M, 10 mL g-1, 1.5 L) maintaining
a reaction temperature of less than 10 ◦C (exotherm and gas
CCA ATA TCT ACG GCG CTG ACA TAA TAG AAT
GGG GGC ATT AAG GCA AGG GCA GG
CCT GCC CTT GCC TTA ATG CCC CCA CCG TAT
TAT GTC AGC GCC GTA GAT ATT GG
◦
Q245Pr
Q245S
CCA ATA TCT ACG GCG CTG ACA TAA TAC GGT
GGG GGC ATT AAG GCA AGG GCA GG
CCT GCC CTT GCC TTA ATG CCC CCA TCT TAT
TAT GTC AGC GCC GTA GAT ATT GG
evolution observed). The reaction was then warmed to 25 C,
stirred for 15 min and the resultant clear biphasic solution
was transferred to a separating funnel. The aqueous layer was
discarded and the organic layer washed with water (5 mL g-1,
750 mL) and then brine (5 mL g-1, 750 mL). The organic phase
was concentrated to dryness and the crude residue granulated in
TBME (7 mL g-1, 1 L) for 2 h prior to isolation of the solid by
Q245Sr
Q245T
CCA ATA TCT ACG GCG CTG ACA TAA TAA GAT
GGG GGC ATT AAG GCA AGG GCA GG
CCT GCC CTT GCC TTA ATG CCC CCA ACC TAT
TAT GTC AGC GCC GTA GAT ATT GG
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 2888–2894 | 2893
©

View Online
(a) S. M. Nettles, K. Matos, E. R. Burkhardt, D. R. Rouda and J. A.
Corella, J. Org. Chem., 2002, 67, 2970; (b) X. Fu, T. L. McAllister, T.
K. Thiruvengadam, C. H. Tann and D. Su, Tetrahedron Lett., 2003,
44, 801; (c) J. Duquette, M. Zhang, L. Zhu and R. S. Reeves, Org.
Process Res. Dev., 2003, 7, 285.
6 A. Yamamura, S. Maruoka, J. Ohtsuka, T. Miyakawa, K. Nagata,
M. Kataoka, N. Kitamura, S. Shimizu and M. Tanokura, Acta
Crystallogr., Sect. F: Struct. Biol. Cryst. Commun., 2009, F65, 1145.
7 A.-R. G. D. Hidalgo, M. A. Akond, K. Kita, M. Kataoka and S.
Shimizu, Biosci., Biotechnol., Biochem., 2001, 65, 2785.
8 Y. Yasohara, N. Kizaki, J. Hasegawa, M. Wada, M. Kataoka and S.
Shimizu, Biosci., Biotechnol., Biochem., 2000, 64, 1430.
9 N. Kizaki, Y. Yasohara, J. Hasegawa, M. Wada, M. Kataoka and S.
Shimizu, Appl. Microbiol. Biotechnol., 2001, 55, 590.
Q245Tr
Q245W
Q245Wr
Q245Y
Q245Yr
Q245V
Q245Vr
CCA ATA TCT ACG GCG CTG ACA TAA TAG GTT
GGG GGC ATT AAG GCA AGG GCA GG
CCT GCC CTT GCC TTA ATG CCC CCA TGG TAT
TAT GTC AGC GCC GTA GAT ATT GG
CCA ATA TCT ACG GCG CTG ACA TAA TAC CAT
GGG GGC ATT AAG GCA AGG GCA GG
CCT GCC CTT GCC TTA ATG CCC CCA TAC TAT
TAT GTC AGC GCC GTA GAT ATT GG
CCA ATA TCT ACG GCG CTG ACA TAA TAG TAT
GGG GGC ATT AAG GCA AGG GCA GG
CCT GCC CTT GCC TTA ATG CCC CCA GTT TAT
TAT GTC AGC GCC GTA GAT ATT GG
CCA ATA TCT ACG GCG CTG ACA TAA TAA ACT
GGG GGC ATT AAG GCA AGG GCA GG
10 A. Shimizu, M. Kataoka, E. Yanase, K. Kita, T. Morikawa and
T. Myoshi, Japanese Patent Application JP0810369 A,, date of
publication 23 April 1996.
11 K. Kita, M. Kataoka and S. Shimizu, J. Biosci. Bioeng., 1999, 88,
Notes and references
591.
1 (a) R. L. Dow, Expert Opinions on Investigational Drugs, 1997, p.
1354; (b) A. E. Weber, Annu. Rep. Med. Chem., 1998, 33, 193; (c) J.
A. Lafontaine, R. F. Day, J. Dibrino, J. R. Hadcock, D. M. Hargrove,
M. Linhares, K. A. Martin, T. S. Maurer, N. A. Nardone, D. A. Tess
and P. DaSilva-Jardine, Bioorg. Med. Chem. Lett., 2007, 17, 5245.
2 (a) M. Imanishi, Y. Tomishima, S. Itou, H. Hamashima, Y. Nakajima,
K. Washizuka, M. Sakurai, S. Matsui, E. Imamura, K. Ueshima, T.
Yamamoto, N. Yamamoto, H. Ishikawa, K. Nakano, K. Hamada, Y.
Matsumura, F. Takamura and K. Hattori, J. Med. Chem., 2008, 51,
1925; (b) N. Tanaka, T. Tamai, H. Mukaiyama, A. Hirabayashi, H.
Muranaka, T. Ishikawa, J. Kobayashi, S. Akahane and M. Akahane,
J. Med. Chem., 2003, 46, 105.
12 H. Li, D. Zhu, L. Hua and E. R. Biehl, Adv. Synth. Catal., 2009, 351,
583.
13 P. Vander Elst, D. Gondol, C. Wynants, D. Tourwe and G. Van Binst,
Int. J. Pept. Protein Res., 1987, 29, 331–46.
14 D. R. Yazbeck, J. Tao, C. Martinez, B. J. Kline and S. Hu, Adv. Synth.
Catal., 2003, 345, 524.
15 These were NS81052, NS81071, NS81037, NS81038, Savinase,
Lipozyme TL 100 L and Lecitase Novo (all from Novozymes),
Aspergillus oryzae protease and Candida rugosa lipase (Protease M
and Lipase AY 30, respectively, both from Amano), two Candida
cylindracea lipases (Lipases OF and MY from Meito), Esterase 003
from Codexis, Protex 6 L from Genencor, Mucor miehei esterase from
Fluka and Cholesterol esterase from Roche.
16 Co-solvent screen of enzymatic desymmetrisation of diester (17).
Conditions: diester 17 (20 mg) was charged to a set of reaction vials
and co-solvent (0–250 mL) added, followed by pH 6.0 calcium acetate
buffer (0.2 M) to make the total solvent volume up to 0.5 mL. Finally,
a solution of Lipozyme TL 100 L (20 mL) was charged to each
vial, the reaction vials were crimped and placed into an Eppendorf
thermomixer orbital shaker and shaken at 900 RPM at 40 ◦C over
24 h.
3 P. A. Bradley, R. J. Carroll, Y. C. Lecouturier, R. Moore, P. Noeureuil,
B. Patel, J. Snow and S. Wheeler, Org. Process Res. Dev., 2010, 14,
1326.
4 R. Noyori and T. Ohkuma, Angew. Chem., Int. Ed., 2001, 40, 40.
5 Some of the poor selectivity in these alternative reductions can be
attributed to the reducing agent used. Depending on the supplier,
borane-tetrahydrofuran complex is stabilized by different additives.
Thus, for example, Aldrich adds <0.005 M NaBH4 as stabilizer.
Experiments have shown that even small amounts of the NaBH4
have a dramatic impact on the selectivity of the asymmetric reduction
with (R)-Me-CBS-oxazaborolidine, and hence when using stabilized
batches from Aldrich, the proportion of unwanted (S)-enantiomer
was 17%–20%. Under the same conditions using BH3/THF from
Fluka (contains no NaBH4 stabilizer) only 3%–4% of the unwanted
(S)-enantiomer was obtained, however, the yield was low (~60%)
17 C. Jimenez-Gonzalez, C. S. Ponder, Q. B. Broxterman and J. B.
Manley, Org. Process Res. Dev., 2011, 15, 912.
18 R. A. Sheldon, Green Chem., 2007, 9, 1273.
19 Radau, Monatshefte fu¨r Chemie, 2003, 134, 1159.
20 A. D. Brown, K. James, C. A. L. Lane, I. B. Moses, N. M. Thomson,
US2005/171147 A1, 2005.
2894 | Green Chem., 2011, 13, 2888–2894
This journal is
The Royal Society of Chemistry 2011
©