Identification and Implementation of Biocatalytic Transformations in Route Discovery: Synthesis of Chiral 1,3-Substituted Cyclohexanone Building Blocks

Article abstract of DOI:10.1021/acs.oprd.8b00139

Several biocatalytic approaches for the preparation of optically pure methyl 3-oxocyclohexanecarboxylates (S)-, (R)-1 and 3-oxocyclohexanecarbonitriles (S)-, (R)-2 have been successfully demonstrated. Screening of reaction-focused enzyme collections was used to identify initial hits using three enzymatic strategies. Reaction optimization and scale-up enabled the production of chiral intermediates for route scouting efforts on scales of up to 100 g. The enzymes applied in these processes (lipases, enoate reductases, and nitrilases) have been shown to be robust catalysts for drug manufacturing and represent a green alternative to conventional methods to access these chiral cyclohexanone building blocks.

Full text of DOI:10.1021/acs.oprd.8b00139

Subscriber access provided by UNIVERSITY OF THE SUNSHINE COAST
Full Paper
Identification and Implementation of Biocatalytic Transformations in Route
Discovery: Synthesis of Chiral 1,3-Substituted Cyclohexanone Building Blocks
Timin Hadi, Alba Diaz-Rodriguez, Diluar Khan, James P Morrison, Justin M
Kaplan, Kathleen T Gallagher, Markus Schober, Michael R Webb, Kristin
Brown, Douglas Fuerst, Radka Snajdrova, and Gheorghe-Doru Roiban
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.8b00139 • Publication Date (Web): 22 Jun 2018
Downloaded from http://pubs.acs.org on June 25, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 36
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Identification and Implementation of Biocatalytic
Transformations in Route Discovery: Synthesis of
Chiral 1,3ꢀSubstituted Cyclohexanone Building
Blocks
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Timin Hadi,a Alba Diaz-Rodriguez,b Diluar Khan,b James P. Morrison,a Justin M. Kaplan,c
Kathleen T. Gallagher,a Markus Schober,d Michael R. Webb,b Kristin K. Brown,e Douglas
Fuerst,a Radka Snajdrova,b, † Gheorghe-Doru Roiban*d
a Advanced Manufacturing Technologies, GlaxoSmithKline, 709 Swedeland Road, King of
Prussia, Pennsylvania 19406, United States.
b API Chemistry, GlaxoSmithKline, Medicines Research Centre, Gunnels Wood Road,
Stevenage SG1 2NY, United Kingdom.
c API Chemistry, GlaxoSmithKline, 709 Swedeland Road, King of Prussia, 19406, PA, United
States.
d Advanced Manufacturing Technologies, GlaxoSmithKline, Medicines Research Centre,
Gunnels Wood Road, Stevenage SG1 2NY, United Kingdom. Email:
gheorghe.d.roiban@gsk.com
ACS Paragon Plus Environment
1

Organic Process Research & Development
Page 2 of 36
1
2
3
4
5
e Molecular Design, Computational and Modeling Sciences, GlaxoSmithKline, 1250 S.
Collegeville Road, Collegeville, Pennsylvania 19426, United States.
6
7
8
9
Abstract
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Several biocatalytic approaches for the preparation of optically pure methyl 3ꢀ
oxocyclohexanecarboxylates (S)ꢀ, (R)ꢀ1 and 3ꢀoxocyclohexanecarbonitriles (S)ꢀ, (R)ꢀ2 have been
successfully demonstrated. Screening of reactionꢀfocused enzyme collections was used to
identify initial hits using three enzymatic strategies. Reaction optimization and scaleꢀup enabled
the production of chiral intermediates for route scouting efforts on scales of up to 100 g. The
enzymes applied in these processes (lipases, enoate reductases and nitrilases) have been shown to
be robust catalysts for drug manufacturing and represent a green alternative to conventional
methods to access these chiral cyclohexanone building blocks.
Keywords: Biocatalysis, Enoate Reductases, Nitrilases, Lipases, Active Pharmaceutical
Ingredients.
1. INTRODUCTION
The use of biocatalysis in the synthesis of pharmaceuticallyꢀrelevant intermediates has been an
increasingly valuable methodology in the chemist’s synthetic toolbox.1ꢀ3 Enzymeꢀcatalyzed
reactions allow for potentially greener processes by increasing atom efficiency, simplifying
chemical routes and reducing pollution and cost of goods.1, 4, 5 They can especially offer
significant advantages in the production of chiral intermediates where the regioꢀ and
stereoselectivity of the enzyme can be harnessed to generate enantiopure products as an
alternative to conventional chiral auxiliaries or inorganic catalysts.6 The increasing number of
ACS Paragon Plus Environment
2

Page 3 of 36
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
characterized and, in many cases, commercially available collections of biocatalysts has enabled
hundreds of unique enzyme variants to be screened for a desired transformation in a rapid
fashion.7 Despite some of the advantages that biocatalysis has to offer, widespread adoption in
process development for active pharmaceutical ingredients (APIs) has yet to occur. Frequently
cited concerns include: the inability to provide biocatalysts on a multigram scale affordably and
promptly, lack of access to enzyme screening collections, and difficulties encountered during
downstream processing of biocatalytic reactions.8 We have been actively working to overcome
these limitations by increasing access to “readyꢀtoꢀuse” enzyme panels generated from
unpurified bacterial cell lysates for common chemical transformations, thereby expanding the
biocatalytic toolbox available to synthetic chemists.9 In addition, we have invested in
infrastructure improvements that facilitate the progression of an enzyme hit in a microtiter plate
to gramꢀscale process development within a matter of weeks. These improvements have enabled
the evaluation of enzymatic reactions in synthetic route scouting in a manner that can be directly
compared to more traditional chemical processes. Cyclohexanes with different substitution
patterns and absolute configurations can be found in a number of pharmaceutically relevant
compounds including Tacrolimus (Prograf),10 Desvenlafaxine,11 Tramadol,12 peroxisome
proliferatorꢀactivated receptor ligands,13 soluble epoxide inhibitor,14 Oxaliplatin,15 and
Gabapentin16 (Fig. 1).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
3

Organic Process Research & Development
Page 4 of 36
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Fig. 1 Selected API structures containing cyclohexane motif.
During recent API route scouting at GSK, optically pure methyl 3ꢀ
oxocyclohexanecarboxylates ((S)ꢀ, (R)ꢀ1) and 3ꢀoxocyclohexanecarbonitriles ((S)ꢀ, (R)ꢀ2) were
identified as key chiral building blocks. The ester and nitrile groups offered different reactive
handles towards the synthesis of more complex structures.
ACS Paragon Plus Environment
4

Page 5 of 36
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Herein we describe efforts to access the cyclohexanone derivatives (S)ꢀ and (R)ꢀmethyl 3ꢀ
oxocyclohexanecarboxylate (1) and (S)ꢀ and (R)ꢀ3ꢀoxocyclohexanecarbonitrile (2) on an
industriallyꢀrelevant scale using three different enzymatic strategies: a) kinetic resolution of racꢀ
1 using lipases, b) enantioselective reduction of prochiral enones using enoate reductases
(EREDs) and c) kinetic resolution of racꢀ2 using nitrilases. Although both enantiomers of 1 and
2 were desired for route scouting, the (S)ꢀenantiomers of the nitrile and methyl ester were
ultimately prioritized as they contained the appropriate configuration for downstream chemistry.
The aim of this article is also to encourage chemists to consider enzymatic approaches by first
intent and to provide several easyꢀtoꢀuse protocols for the preparation of chiral cyclohexanone
derivatives.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2. RESULTS AND DISCUSSION
Few reports describe the preparation of enantiomerically enriched methyl 3ꢀ
oxocyclohexanecarboxylate (1) and 3ꢀoxocyclohexanecarbonitrile (2). (R)ꢀ1 has been prepared
using asymmetric phase transfer catalysis and expensive ruthenium catalysts but result in low
product enantioselectivity (47% ee).17 The synthesis of (S)ꢀ2 has been reported utilising a
catalytic conjugate addition of cyanide to enones at low temperatures, resulting in products with
moderate enantiomeric excess (81% ee).18 In the biocatalysis space, syntheses of enantiopure
compounds 1 and 2 have been described starting from the corresponding enones using enoate
20
reductases (EREDs)19,
or through enzymatic cascades starting from the corresponding
cyclohexenes.21, 22 Although there have been a number of published examples of EREDꢀ
mediated catalysis over the last few years reporting high product enantioselectivity (>99%),23ꢀ25
EREDs have yet to gain widespread adoption in the field of industrial biocatalysis. Many studies
ACS Paragon Plus Environment
5

Organic Process Research & Development
Page 6 of 36
1
2
3
4
5
have been limited to milligram scale syntheses26ꢀ29 with only a few reporting enzymatic reactions
at gram scales.30ꢀ33
6
7
8
9
Several biocatalytic strategies were considered in order to access chirallyꢀenriched compounds
1 and 2 (Scheme 1). Two resolutionꢀbased strategies using lipases and nitrilases were explored
as well as an EREDꢀbased asymmetric synthesis using a chiral reduction. For the synthesis of
chirallyꢀenriched 1, a lipaseꢀmediated resolution of a racemic mixture of 1 was investigated
while an analogous nitrilaseꢀmediated resolution of racemic 2 was also pursued. An alternative
strategy that was investigated for the syntheses of both 1 and 2 employed a set of enoate
reductases to catalyze the chiral reduction of the unsaturated cyclohexenones methyl 3ꢀ
oxocyclohexꢀ1ꢀenecarboxylate (3) and 3ꢀoxocyclohexꢀ1ꢀenecarbonitrile (4).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Synthetic strategies to access chiral cyclohexanones 1 and 2
2.1 Preparation of (R)ꢀ and (S)ꢀmethyl 3ꢀoxocyclohexanecarboxylate (1)
2.1.1 Lipaseꢀmediated resolution of racemic methyl 3ꢀoxocyclohexanecarboxylate (rac-1)
Lipases are robust and readily available catalysts that have been widely utilized in organic
syntheses.34, 35 Due to their widespread use, collections of commercially available lipases are
ACS Paragon Plus Environment
6

Page 7 of 36
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
easy to access on small scales for screening and on larger scales for process development. We
envisioned employing a lipase to kinetically resolve racꢀ1 to access enantiopure (S)ꢀ or (R)ꢀ1.
Although racꢀ1 is expensive to purchase, it can be made cheaply from methyl 3ꢀ
hydroxybenzoate in two steps36 making this a potentially viable route even though the theoretical
yield is only 50%. In order to achieve proof of concept for this biocatalytic approach, various
lipases were screened for their activity and selectivity on racꢀ1 (Table 1), in the hopes of
identifying unique enzymes able to resolve both (R)ꢀ1 and (S)ꢀ1.37 Three enzymes that were
tested displayed low selectivity for hydrolysis of the (R)ꢀenantiomer (Table 1 entries 3, 5ꢀ6), with
lipase CES Lꢀ1 giving the best ee for remaining (S)ꢀ1 (26% ee) (E= 6). A few lipases showed
promising selectivity for the hydrolysis of (S)ꢀ1, with Novocor ADL producing (R)ꢀ1 with
>99.5% ee (Table 1, entry 1, E= 18).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. Screening results using commercial lipases in the resolution of racꢀ1
Entry
Lipase a
Conv (%) b
ee (%) 1 c
>99.5 (R)
10 (R)
1
2
3
4
5
Novocor ADL
ENZAꢀ06
ENZAꢀ04
ENZAꢀ43
CES Lꢀ1
63
15
22
57
28
8 (S)
74 (R)
26 (S)
ACS Paragon Plus Environment
7

Organic Process Research & Development
Page 8 of 36
1
2
3
4
5
6
6
7
CES Lꢀ2
CES Lꢀ4
10
33
3 (S)
26 (R)
7
8
9
a Entries 1ꢀ7 represent selected hits from several commercial lipases. Reaction conditions: racꢀ
1 (10 mg) in DMSO (50 µL), potassium phosphate buffer 100 mM (1 mL), pH 7.0, 50% wt
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
b
c
enzyme, 15°C, 22 h with orbital shaking at 1000 rpm. Conversion was determined by HPLC.
Absolute configuration determined after comparison with authentic samples.
A 13 g reaction was performed using Novocor ADL to generate (R)ꢀ1. The product ee at this
scale decreased to 88%, with an isolated yield of 38%. Although ee and yield could be improved
through reaction optimization and further process development, the moderate product ee
observed on scaleꢀup and lack of a complementary lipase that would enable resolution of the (S)ꢀ
enantiomer led us to draw our attention to an EREDꢀmediated approach.
2.1.2 EREDꢀcatalyzed reduction of methyl 3ꢀoxocyclohexꢀ1ꢀenecarboxylate (3)
An initial screen for the stereoselective reduction of 3 was carried out using an internal
collection of ERED enzymes. Reactions were screened under aqueous conditions in the presence
of a glucose dehydrogenase (GDH) cofactor recycling system. A number of ERED enzymes
displayed conversion of the starting material 3 as well as high product ee for either enantiomer of
compound 1 (Table 2). For many enzyme variants, conversion to the desired product was
accompanied by the production of the carbonyl reduction products of both 1 and 3, as well as the
disproportionation reaction product27 methyl 3ꢀhydroxybenzoate (Fig. 2). Formation of carbonyl
reduction products was more pronounced in the case of less active ERED variants. In addition,
negative controls (produced from cells harboring plasmids lacking the ERED gene) displayed a
background conversion of 3 preferentially to (R)ꢀ1, suggesting the presence of endogenous
EREDꢀlike activity in the E. coli expression strain. This endogenous background activity is a
ACS Paragon Plus Environment
8

Page 9 of 36
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
possible cause for the lower enantioselectivity of the C26GꢀYqjM variant under these screening
conditions. C26GꢀYqjM was reported to give (S)ꢀ1 in ee >99% when assayed as a purified
enzyme preparation.19 The observed ee of the identical variant screened in our hands as a crude
bacterial cell lysate was found to be 96%.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 2. EREDꢀcatalyzed reduction of 3 → 1
Entry
1
ERED a
Conv (%) b, c
ee (%) 1 d
92 (R) e
>99 (R)
96 (S)
wtꢀYqjM
85
80
91
85
80
59
81
72
64
67
78
77
2
C26DꢀYqjM
C26GꢀYqjM
EREDꢀB3
EREDꢀC5
EREDꢀC12
EREDꢀD3
EREDꢀE8
EREDꢀE11
EREDꢀF7
EREDꢀG3
EREDꢀG5
3
4
98 (S)
5
99 (S)
6
97 (R)
98 (S)
7
8
78 (R)
50 (R)
61 (R)
99 (S)
9
10
11
12
99 (S)
ACS Paragon Plus Environment
9

Organic Process Research & Development
Page 10 of 36
1
2
3
4
5
6
7
8
13
14
EREDꢀH6
EREDꢀ207
80
99
99 (S)
99 S)
82 (R)
15
Negative control f
<3
9
a
b
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Entries 4ꢀ13 represent selected hits obtained from the Codexis ERED screening plate 1.
Reaction conditions: enone 3, potassium phosphate buffer 100 mM, pH 7.4, entries 1ꢀ3: whole
cells, time 3 h, 10 g/L substrate loading, entries 4ꢀ14: clarified lysate, 3 g/L substrate loading. c
Conversion was calculated by GC, % a/a of 1 versus sum a/a of 1, starting material 3, methyl 3ꢀ
d
e
hydroxybenzoate, and other carbonyl reduction side products. ee values measured by GC.
f
Absolute configuration determined after comparison with authentic samples. Negative control
produced from cells harboring plasmids lacking the ERED gene.
100
O
80
COOMe
60
40
20
0
(S)-
1
Methyl
3-hydroxybenzoate
pH 7
pH 7.5
pH 8
Fig. 2. pH effect on the conversion of 3 and formation of disproportionation product methyl
3ꢀhydroxybenzoate. Reaction conditions: EREDꢀ207 (5% wt/wt), 3 (10 g/L) in 200 mM
potassium phosphate buffer (pH = 7.0, 7.5 and 8.0), NADP+ (0.1 equiv), GDH recycling
system, 30 °C.
ACS Paragon Plus Environment
10

Page 11 of 36
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Three enzyme hits (C26DꢀYqjM, C26GꢀYqjM and EREDꢀ207) were chosen for further
development based on both their selectivity and impurity profiles. A number of factors were
quickly examined including pH, buffer strength and reaction volumes to maximize both product
yield and ee while minimising the production of unwanted side products.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Reaction pH was varied from 7 to 8 and was found to play an important role in the rate of
formation of the disproportionation sideꢀproduct methyl 3ꢀhydroxybenzoate (Fig. 2). Methyl 3ꢀ
hydroxybenzoate production increased with pH (pH 7.0 – 0%; pH 7.5 – 1.5%; pH 8.0 – 2.6%)
which corresponds to previously reported pH studies for cyclohexꢀ2ꢀenone disproportionation.38
The overall reaction conversion was also negatively impacted. When using the GDH cofactor
recycling system, a molar equivalent of gluconic acid is produced, necessitating buffering to
maintain constant pH throughout the reaction. Increasing the buffer concentration up to 200 mM
was a practical solution on small scales with low substrate loadings (< 50 mM), but at higher
reaction volumes pH was controlled via addition of aqueous sodium hydroxide (5M).
Larger scale reactions were performed using both YqjM mutants (C26D and C26G) in order to
generate representative material for downstream chemistry and evaluate the scalability of the
EREDꢀcatalyzed reactions. The two EREDs were expressed using a previously established
protocol on both the shake flask (1L) and fermenter (50L) scale.39 Initial gramꢀscale reactions
were performed using whole cells expressing either the C26DꢀYqjM or C26GꢀYqjM variants,
respectively. The C26GꢀYqjM ERED reaction, carried out with a whole cell enzyme loading of
300% wt, gave the desired product (S)ꢀ1 in 72% isolated yield with 96% ee (Table 3, entry 1).
The C26DꢀYqjM variant displayed lower activity than the C26G and thus required a higher cell
loading (640% wt loading of whole cells), but still gave the opposite enantiomer (R)ꢀ1 in 60%
isolated yield with 99% ee (Table 3, entry 2).
ACS Paragon Plus Environment
11

Organic Process Research & Development
Page 12 of 36
1
2
3
4
5
6
7
8
9
An extractive workup using ethyl acetate was employed for both of these whole cell reactions,
but this resulted in the formation of emulsions, complicating the isolation of product and
negatively impacting yields. Improved phase separation was achieved through addition of Celite
in addition to ethyl acetate, but this is a nonꢀpreferred option on industrial scales due to long
filtration times. Vessel cleaning and Celite detection is laborious, increasing time and cost and
also producing large amounts of waste. No other filtration aids were investigated at that time.
After examination of a number of different solvent extraction systems, a mixture of isopropanol
(IPA) and isopropyl acetate (iPrOAc) was identified as an improved system for product isolation.
Addition of 1 reaction volume (1.9 L) of a 1:1 IPA : iPrOAc mixture to the completed reaction
mixture gave excellent phase separation and improved isolated yield (Figure 3).40 A large scale
reaction (100 g, Table 3, entry 3) using C26GꢀYqjM was performed using the improvements to
this whole cell process and gave good isolated product yield of (S)ꢀ1 (85%), albeit with a slightly
lower ee (94%) than observed in smaller scale reactions. The scale up of this reaction enabled us
to supply gram amounts of (S)ꢀ1 to the project team for downstream chemistry.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For the production of (S)ꢀ1, EREDꢀ207 was examined in more detail as it was observed to be
more selective than the C26DꢀYqjM enzyme variant. EREDꢀ207 was produced as a lyophilized
powder from clarified cell lysate as opposed to the previously described whole cells for the two
YqjM variants. Supplying enzyme catalysts as lyophilized powders greatly simplifies reaction
setup since they do not require any specialized storage facilities, can be charged in reactors in a
similar manner to other chemical reagents, and generally contain more concentrated catalyst per
unit mass.41
Using lyophilized EREDꢀ207 (5% wt), (S)ꢀ1 was successfully isolated in 89% yield and 98%
ee on a 1 g scale (Table 3, entry 4). An additional reaction was successfully performed on 55 g
ACS Paragon Plus Environment
12

Page 13 of 36
Organic Process Research & Development
1
2
3
4
scale (Table 3, entry 5) in order to provide larger amounts of (S)ꢀ1 for downstream chemistry,
5
6
with emulsion formation being slightly reduced compared to the wholeꢀcell reaction scaleꢀup.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 3. Scale up experiments for the preparation of (R)ꢀ and (S)ꢀ1 starting from enone 3
Substrate
loading (g/L)
41.4
Enzyme
Yield 1, (%)
Entry
3 (g)
ee (%) 1 c
b
1
2
3
4
5
4
C26GꢀYqjM
C26DꢀYqjM
C26GꢀYqjM
EREDꢀ207
EREDꢀ207
72
60
85
89
90
96 (S)
99 (R)
94 (S)
98 (S)
97 (S)
5
20.0
100
1
34.5
15.0
55
14.8
a Reaction conditions enone 3, potassium phosphate buffer (200 mM, pH 7.0), GDH cofactor
recycling system. b Isolated yield. c ee values measured by GC.
ACS Paragon Plus Environment
13

Organic Process Research & Development
Page 14 of 36
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Fig. 3 100 g scale reaction performed in 5 L controlled laboratory reactor (CLR) using E.
coli whole cells expressing C26GꢀYqjM. Reaction conditions: 3 (34.5 g/L), potassium
phosphate buffer (200 mM, pH 7.0), 250 rpm, 2.5 h, 30 °C, (left) and workꢀup with
IPA/iPrOAc (right).
2.2 Preparation of (R)ꢀ and (S)ꢀ3ꢀoxocyclohexanecarbonitrile (2)
2.2.1 EREDꢀcatalyzed reduction of 3ꢀoxocyclohexꢀ1ꢀenecarbonitrile (4)
3ꢀOxocyclohexanecarbonitrile (2) was identified as another potential starting building block
for the synthesis of APIs incorporating the cyclohexane core. A nitrile group at the 3ꢀposition
allows for access to chiral acid, amine and amide derivatives. To begin, the EREDꢀcatalyzed
reduction of 3ꢀoxocyclohexꢀ1ꢀenecarbonitrile (4) was investigated. Syntheses of 3ꢀoxocyclohexꢀ
1ꢀenecarbonitrile (4) have been reported in high yield by oxidation,42ꢀ48 or functional group
interconversion.49, 50, 51, 52 However, due to irreproducible yields, toxicity and scalability concerns
with the afforementioned methods, we rapidly prepared multigram quantities of 4 from 3ꢀ
ACS Paragon Plus Environment
14

Page 15 of 36
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
chlorocyclohexꢀ2ꢀenone (see supporting information). Our ERED collection was screened
against 3ꢀoxocyclohexꢀ1ꢀenecarbonitrile (4) at 24 g/L substrate loading and a number of hits
were identified that catalyzed the formation of the reduced product 2 (Table 4). Most of the
variants identified were very selective (>99% ee) for the (S)ꢀ2, while a single enzyme variant
catalyzed the formation of (R)ꢀ2 in 46% yield and 97% ee (Table 4, entry 12). Interestingly, the
two YqjM variants C26DꢀYqjM and C26GꢀYqjM, which gave high conversion and ee with 3
(Table 2, entries 2,3), gave comparably poor results for the reduction of 4, with < 20%
conversion and low product ee values (Table 4, entries 1ꢀ2).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 4. Screening results using GSKꢀERED collection for 3ꢀoxocyclohexꢀ1ꢀenecarbonitrile
(4).
Entry
ERED a
% Conv b,c
ee (%) 2 c,d
64 (R)
1
2
3
4
5
6
7
C26DꢀYqjM
C26GꢀYqjM
EREDꢀA1
EREDꢀA2
EREDꢀA9
EREDꢀA10
EREDꢀA12
15
13
95
98
94
96
99
38 (S)
>99 (S)
>99 (S)
>99 (S)
>99 (S)
>99 (S)
ACS Paragon Plus Environment
15

Organic Process Research & Development
Page 16 of 36
1
2
3
4
5
6
7
8
9
8
EREDꢀB1
EREDꢀB7
EREDꢀD1
EREDꢀG7
EREDꢀH7
99
98
97
99
46
>99 (S)
>99 (S)
>99 (S)
>99 (S)
97 (R)
9
10
11
12
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
b
Entries 3ꢀ12 represent selected hits obtained from the Codexis ERED Screening plate 1.
Reaction conditions: a) entries 1ꢀ2: enone 4, 5 g/L, whole cells; potassium phosphate buffer 100
c
mM, pH 7.4, entries 3ꢀ12 enone 4, 24 g/L, clarified lysate, (Codexis plate). Conversion and ee
d
values measured by GC. Absolute configuration was determined after comparison with
authentic standards.
Additional scouting experiments identified EREDꢀB1 as the most active variant. EREDꢀB1
was overꢀexpressed and produced as a lyophilized powder for reaction scale up. An initial 1 g
reaction using 10% wt of EREDꢀB1 produced (S)ꢀ2 in 80% isolated yield with excellent ee
(99.5%). The reaction was then scaled up to 40 g using 10% wt lyophilized enzyme loading at a
substrate concentration of 18.2 g/L. GC monitoring showed that the reaction was complete after
just 1 h and (S)ꢀ2 was isolated in 87% yield with 99.5% ee after extractive workꢀup.
2.2.2 Nitrilaseꢀmediated resolution of racemic 3ꢀoxocyclohexanecarbonitrile (racꢀ2)
During the course of the ERED studies, beside the lack of a commercial supplier, additionally
it was found that substrate 4 appeared less stable than 2 during distillative purification at higher
temperatures. Instability of 4 posed a significant risk that could limit the applicability of the
EREDꢀcatalyzed approach in a manufacturing process. To mitigate this risk, we decided to
ACS Paragon Plus Environment
16

Page 17 of 36
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
investigate the use of nitrilases to resolve a racemic mixture of the easily available nitrile, racꢀ
2.53ꢀ55 Although this resolution only has a theoretical yield of 50%, extractive separation of the
desired nitrile from the undesired carboxylic acid (5) should be straight forward leading to
isolated yields close to the theoretical maximum.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A panel of nitrilases comprised of a set of wildꢀtype enzymes identified from literature
reports56 and through protein homology modelling was screened. Initial screening was
performed using chiral GC analysis to identify nitrilases that were selective for hydrolysis of
either (S)ꢀ2 or (R)ꢀ2. Nitrilase NITꢀE6 allowed for the resolution of (R)ꢀ2 (up to 38 %ee), while
other nitrilases had modest to very good selectivity for the resolution of (S)ꢀ2. A representative
set of selective nitrilases is summarized in Table 5, with NITꢀG6 (97% (S)ꢀ2 ee, entry 4) showing
the most promising selectivity for the resolution of (S)ꢀ2.
Table 5. Nitrilase mediated resolution of racemic 3ꢀoxocyclohexanecarbonitrile (2).
Entry
Nitrilase a, b
NITꢀD9
NITꢀE6
% Conv c
ee (%) 2 c, d
3 (S)
1
2
3
4
5
23
31
28
51
21
38 (R)
11 (S)
NITꢀE8
NITꢀG6
NITꢀH6
97 (S)
13 (S)
ACS Paragon Plus Environment
17

Organic Process Research & Development
Page 18 of 36
1
2
3
4
a
b
Entries 1ꢀ5 are aquired in house wildꢀtype nitrilases. Reaction conditions: 2, 5 g/L,
5
6
7
8
c
potassium phosphate buffer 100 mM pH 7.0, 30 °C, 800 rpm, 20 h. Conversion and ee were
determined by chiral GC.d Absolute configuration was determined after comparison with
authentic samples.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Nitrilase G6 was selected for further development work and expressed as a crude clarified cell
lysate and used for reaction scaleꢀup and optimization. An achiral GC method was used to
monitor formation of the hydrolysis product 5 in conjunction with the chiral GC method for
analysis of remaining compound 2. Initial reactions performed on a 0.5 and 2.6 g scale of racꢀ2,
generated (S)ꢀ2 in 99.4 and 98.4% ee, respectively (Table 6, entries 1ꢀ2). pH adjustment of the
reaction mixture and extractive separation of the acid product from the desired nitrile proceeded
efficiently, giving isolated yields of 36% and 42%, respectively.
A 5 g reaction was then conducted in a jacketed reactor using clarified lysate at 5% wt lysate
loading and 26.5 g/L substrate loading. The reaction proceeded to completion after 3h and
afforded (S)ꢀ2 in 27% yield (99% ee, Table 6, entry 3).
Table 6. Scale up experiments for the preparation of (S)ꢀ2 starting from racemic 3ꢀ
oxocyclohexanecarbonitrile (2) using NITꢀG6.
Entry
racꢀ2 (g)
0.5
Isolated yield 2, (%)
ee (%), 2
>99 (S)
98 (S)
1
2
3
4
36
42
27
28
2.6
5.0
99 (S)
20.0
> 99 (S)
ACS Paragon Plus Environment
18

Page 19 of 36
Organic Process Research & Development
1
2
3
4
a Reaction conditions: 2, entries 1ꢀ3: 26.5g/L; entry 4: 50g/L; potassium phosphate buffer 200
5
6
mM, pH 7.0, 30 min, 30 °C.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To evaluate NITꢀG6 at a higher substrate loading (50 g/L), a 20 g scale reaction was performed
and promisingly showed similar results in terms of yield and selectivity to the previous reactions
(Table 6, entry 4).
This unoptimized process using NITꢀG6 was deemed suitable for producing (S)ꢀ2 and provided
an alternative to the EREDꢀcatalyzed reduction of 4. Additional enzyme and process
optimization is likely to improve reaction profiles and increase isolated yield.
CONCLUSIONS
Several biocatalytic approaches have been used on multigram scales to access chiral 3ꢀ
substituted cyclohexanones under mild reaction conditions. These approaches represent a useful
alternative to traditional chemical methods and produced the desired compounds with high
isolated yields and ee. Three different enzyme type (lipases, enoate reductases and nitrilases)
were screened from internal enzyme collections. The majority of these enzymes were produced
as crude cell lysates, removing the need for laborious purification and making them ideal
catalysts from a costꢀofꢀgoods standpoint. Proof of concept experiments on gram scales were
conducted rapidly to generate material of sufficient quality (ee, purity) and quantity for
downstream chemistry applications. Further reaction optimization and process development
allowed for the successful demonstration of these enzymatic reactions on a scale of up to 100 g
in a rapid fashion, highlighting the robustness and scalability of these biocatalytic processes.
When comparing these strategies it is important to note that the EREDꢀmediated reduction can
produce the desired product in 100% yield, while the theoretical yield of a lipaseꢀ or nitrilaseꢀ
ACS Paragon Plus Environment
19

Organic Process Research & Development
Page 20 of 36
1
2
3
4
5
6
7
8
9
mediated resolution would be limited to 50%. This makes the EREDꢀmediated reduction an
attractive strategy for synthesizing chiral cyclohexanone building blocks. However, resolution
approaches were also pursued as factors including alternative methods of substrate synthesis,
biocatalyst availability, substrate instability and downstream processing challenges came to light
during our work.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
We have demonstrated that enzymatic approaches can be rapidly (1ꢀ3 months) evaluated and
implemented in chemical route development on scales that truly compete with other more
traditional synthetic approaches. Biocatalysis is an increasingly valuable tool for chemists in the
pursuit of cheaper, greener, and more efficient routes for API manufacture.
4. EXPERIMENTAL
Biology
The genes encoding EREDs and nitrilases were codon optimised for expression in E. coli
BL21(DE3), synthesised, and cloned into pET28a at GenScript. Transformation, shake flask
enzyme expression and 50 L fermentation trials were performed as described previously.39
Escherichia coli BOU730 cells [an E. coli strain derived from BL21(DE3) that contains a copy
of GDH gene from B. megaterium under the control of T7 promoter]57 were used during this
study for the wtꢀYqjM, C26D and C26G variants. Plasmids containing wtꢀYqjM and mutants
C26D and C26G were received from Reetz’s group and were published previously.19 ERED
plate 1 and EREDꢀ207 were acquired from Codexis. The nitrilase panel was designed and
produced inꢀhouse, and is comprised of gene sequences encoding putative nitrilases assembled
using an extensive homology search. Sequences of enzyme NITꢀD9, E6, E8, G6 and H6 are
given in supporting information. Lipases were acquired from Amano, and Enzagen. GDHꢀCDXꢀ
ACS Paragon Plus Environment
20

Page 21 of 36
Organic Process Research & Development
1
2
3
4
5
901 was acquired from Codexis, NADP+ was bought from Bontac BioꢀEngineering and used
without further purification.
6
7
8
9
Chemistry
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Racꢀ2 was either acquired from FCH group (Ukraine) or prepared in house (see supporting
information), while racꢀ1, enone 3, and the other chemical reagents were purchased from the
following commercial sources: Manchester Organics, Acros Organics, Sigma Aldrich and Alfa
Aesar. Enone 4 was prepared in house according to the procedure described in the supplementary
information. All chemicals and solvents were used without any additional purification. NMR
spectra were recorded on a Bruker Advance 400 (1HꢀNMR 400 MHz, 13CꢀNMR 101 MHz)
spectrometer using TMS as internal standard (δ=0). Details on plateꢀbased screening as well as
analytical methods are given in the supporting information. Scaleꢀup reactions were performed in
conical centrifuge tubes (50 mL), and controlled laboratory reactors (CLR) (volumes from
0.25 ꢀ 5L). Small amounts of cell paste were lysed using a Fisher Scientific sonicator model
FB705 equipped with a specific probe depending on the lysis volume (up to 100 mL). Higher
amounts of cell paste were disrupted using a microfluidiser M110Y from analytikLtd
Cambridge. Lyophilisation was performed using a BPS VirTris SP Scientific Advantage Pro
lyophiliser. For the enzyme fermentation a 50 L (35L working volume) fermenter was used as
previously described.39
Preparation of (R)ꢀmethyl 3ꢀoxocyclohexanecarboxylate (1) using C26DꢀYqjM ERED (5
g scale)
ERED C26DꢀYqjM was expressed as previously reported.19 Cell pellets were stored in the
freezer at ꢀ20°C. To a 500 mL Erlenmeyer flask potassium phosphate buffer was added (KH2PO4
/ K2HPO4, 250 mL, pH 8.0, 100 mM). Dꢀglucose (7.5 g, 1.25 equiv), GDHꢀCDXꢀ901 (0.2 g) and
ACS Paragon Plus Environment
21

Organic Process Research & Development
Page 22 of 36
1
2
3
4
5
6
7
8
9
NADP+ (0.256 g, 0.01 equiv) were then added. Reaction was started by addition of methyl 3ꢀ
oxocyclohexꢀ1ꢀenecarboxylate (3) (5.0 g, 32.4 mmol final conc 17.7 g/L) and ERED C26Dꢀ
YqjM (32 g, cell wet mass) and shaken in an Kühner shaker at 200 rpm and 30°C. Reaction
samples were extracted with ethyl acetate and monitored by GC. After full conversion was
reached, the reaction mixture was extracted with ethyl acetate (2 x 250 mL). The organic extracts
were combined and dried over Na2SO4 before being concentrated in vacuo to afford (R)ꢀ1 as a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
pale yellow oil (3.0 g, 60% yield, 77% purity by GC, 99.4 ee%). HꢀNMR (400 MHz, CDCl3) δ
ppm 3.7 (s, 3 H) 2.8 (m, 1 H) 2.6 (m, 2 H) 2.4 ꢀ 2.3 (m, 2 H) 2.2 ꢀ 2.0 (m, 2 H) 1.9 ꢀ 1.7 (m, 2 H).
13CꢀNMR (101 MHz, CDCl3) δ ppm 209.1 (1 C) 174.0 (1 C) 51.9 (1 C) 43.0 (1 C) 42.9 (1 C)
40.8 (1 C) 27.6 (1 C) 24.3 (1 C).
Preparation of (S)ꢀmethyl 3ꢀoxocyclohexanecarboxylate (1) using C26GꢀYqjM ERED
(100 g scale)
To a controlled laboratory reactor (5 L) potassium phosphate buffer (KH2PO4 / K2HPO4, 2.0 L,
200 mM, pH 7.0) was charged and the temperature was adjusted to 30°C. Dꢀglucose (175 g, 1.28
equiv) and NADP+ (0.5 g) were then added to the vessel and agitation adjusted to 200 rpm. At
this point pH monitoring indicated a value of 7.0. When all components dissolved after 5 min,
frozen whole cells (180 g) harbouring ERED variant C26GꢀYqjM in cells coexpressing GDH for
cofactor regeneration57 were added and the mixture was stirred for 30 min during which time all
the cells resuspended with no lumps being observed. The reaction was started by the addition of
methyl 3ꢀoxocyclohexꢀ1ꢀenecarboxylate (3) (100 g, 649 mmol, final conc 34.5 g/L). Additional
potassium phosphate buffer (KH2PO4 / K2HPO4, 0.3 L, 200 mM, pH 7.0) was added to rinse the
CLR. The reaction mixture was then stirred at 30°C and 200 rpm for 50 min at which point
GDHꢀCDXꢀ901 (0.2 g) and additional frozen cell paste (53 g) were added. After 1 h 30 min
ACS Paragon Plus Environment
22

Page 23 of 36
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
further amounts of NADP+ (1.0 g) and cell paste (70 g) were added and the reaction was allowed
to stir for additional 30 min, after which GC analysis showed complete conversion. The solution
pH was maintained during the course of reaction between pH 6.7 ꢀ 8.0 by addition of NaOH (5
M).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The crude reaction mixture (2.9 L) was then divided in two batches. The first batch (1L) was
quenched with acetonitrile and the second batch (1.9 L) with isopropanol (IPA) / isopropyl
acetate (iPrOAc) 1:1 v/v. Both workꢀup approaches resulted in no emulsion formation,
acetonitrile delivering a thin interface in between the organic and aqueous layer in comparison
with IPA/ iPrOAc extraction where no interface was formed.
a) Acetonitrile workꢀup
1L of reaction mixture was quenched with acetonitrile (1 L) and stirred for 30 min at 444 rpm
to ensure proper mixing and then stirring was stopped for 6 min to allow phase separation. The
organic layer (0.65 L) was separated and a second acetonitrile extraction (1 L) of the remaining
aqueous phase proceeded with no emulsion formation, allowing the recovery of a total of 1.65 L
of organic phase. The remaining aqueous phase from the acetonitrile wash was extracted with
ethyl acetate (0.5 L) to determine extraction efficiency;mass balance indicated that <1% (S)ꢀ1
was present in the aqueous phase. The acetonitrile washes were then combined and evaporated
to remove most of the acetonitrile and then extracted twice with ethyl acetate (2 x 0.6 L). The
combined organic layers (1.8 L) were washed with brine (0.3 L) and then concentrated in vacuo
to yield (S)ꢀ1 (29.3 g, 83% yield, 93% GC purity, 94% ee).
b) IPA / iPrOAc workꢀup
1.9 L of reaction crude was quenched with IPA/iPrOAc (1:1 v/v, 1.9 L) and stirred for 30 min
at 444 rpm to ensure proper mixing. Stirring was stopped and phase separation occurred after 20
ACS Paragon Plus Environment
23

Organic Process Research & Development
Page 24 of 36
1
2
3
4
5
6
7
8
9
min. The organic layer (1.7 L) was separated and recovered and a second IPA/iPrOAc (1:1)
extraction (1.9 L) of the remaining aqueous phase proceeded smoothly with no emulsion
formation, allowing the recovery of 2.5 L of organic phase. The organic phases were combined
and concentrated in vacuo to give (S)ꢀ1 as a pale yellow oil (56.6 g, 86% yield, 89% GC purity,
94% ee).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Preparation of (R)ꢀmethyl 3ꢀoxocyclohexanecarboxylate (1) via kinetic resolution using
lipase Novocor ADL (13 g scale)
To a 300 mL CLR was added potassium phosphate buffer (200 mM, 79 mL) followed by
methyl 3ꢀoxocyclohexanecarboxylate (1) (13.0 g, 83 mmol) in dimethyl sulfoxide (DMSO) (1.9
mL). The reactor was cooled to 5°C and NovoCor ADL (2.6 g) was added. The reaction progress
was monitored using chiral HPLC and was deemed complete after 48 h of stirring. The reaction
mixture was removed from the reactor and the reactor was rinsed with dichloromethane (50 mL).
The material was filtered over Celite and the layers separated. A significant emulsion formed so
the entire contents were once again filtered over celite. Upon filtration, the layers were once
again separated. The aqueous layer was extracted with dichloromethane (150 mL) and the layers
separated. The organics were combined, dried over Na2SO4 and concentrated in vacuo. The
residue was purified via flash chromatography (0ꢀ40% EtOAc/Hexanes; 80 g silica gel column).
The desired fractions were collected to yield (R)ꢀmethyl 3ꢀoxocyclohexanecarboxylate (1) as a
clear oil (5.7 g, 38 % yield).
Preparation of (S)ꢀ3ꢀoxocyclohexanecarbonitrile (2) using EREDꢀB1 (40 g scale)
To a 5L CLR potassium phosphate buffer (KH2PO4 / K2HPO4, 1.5 L, 200 mM, pH 7.0) was
charged and the temperature was adjusted to 30°C. Dꢀglucose (64.6 g, 1.25 equiv), GDHꢀCDXꢀ
901 (3.0 g), and NADP+ (2.2 g, 0.01 equiv) followed by EREDꢀB1 lyophilised powder (4.0 g,
ACS Paragon Plus Environment
24

Page 25 of 36
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
10% wt) were then added to the vessel and the stirꢀrate was adjusted to 170 rpm. At this point pH
monitoring indicated a value of 7.0. When all components were suspended (in 10 min), the
reaction was started by the addition of 3ꢀoxocyclohexꢀ1ꢀenecarbonitrile (4) (40.0 g, 327 mmol,
final conc 18.2 g/L). Additional potassium phosphate buffer (KH2PO4 / K2HPO4, 0.5 L, 200 mM,
pH 7.0) was added to rinse the CLR and the stirꢀrate was increased to 250 rpm. The reaction
mixture was then stirred at 30°C and 250 rpm for 2 h after which GC analysis showed total
conversion to (S)ꢀ2. During the course of the reaction, the solution pH was maintained between
pH 6.7 ꢀ 7.0 by the addition of NaOH (5 M). After the reaction was completed, the solution was
quenched by adding ethyl acetate (2 L) and the resultant emulsion was then filtered over cotton
wool to remove the cell debris. Additional ethyl acetate (2 x 0.2 L) was used to wash the cotton
wool. The filtrate was then transferred back to the vessel and stirred at 64 rpm. After partial
separation of the organic and aqueous layers occurred, both the clear organic layer and the
interface emulsion (1.2 L) were collected. The aqueous phase was then extracted with ethyl
acetate (2 L) and mixture stirred at 130 rpm for 10 min. After stirring was stopped, the mixture
was allowed to settle and the organic layer (1.6 L) was then collected. The interface emulsion
from this extraction (0.6 L) was combined with the interface layer from the first extraction (1.2
L). This mixture was washed with brine (0.9 L) to help break the emulsion and extracted with
additional ethyl acetate (0.9 L). The organic layers were combined and concentrated in vacuo to
give the desired 3ꢀoxocyclohexanecarbonitrile (S)ꢀ2 (34.9 g, 87% yield, 96% GC purity, 99.5%
ee). 1HꢀNMR (400 MHz, CDCl3) δ ppm 2.9 ꢀ 3.1 (m, 1 H) 2.5 ꢀ 2.7 (m, 2 H) 2.4 (t, J=6.52 Hz, 2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13
H) 2.1 ꢀ 2.2 (m, 2 H) 1.9 ꢀ 2.1 (m, 1 H) 1.7 ꢀ 1.9 (m, 1 H); CꢀNMR (101 MHz, CDCl3) δ ppm
205.4 (1 C) 120.1 (1 C) 43.1 (1 C) 40.6 (1 C) 28.6 (1 C) 28.0 (1 C) 23.7 (1 C).
Preparation of (S)ꢀ3ꢀoxocyclohexanecarbonitrile (2) using nitrilase NITꢀG6 (20 g scale)
ACS Paragon Plus Environment
25

Organic Process Research & Development
Page 26 of 36
1
2
3
4
5
6
7
8
9
Preparation of clarified lysate: NITꢀG6 cell paste (31.6 g) was resuspended in potassium
phosphate buffer (120 mL, KH2PO4 / K2HPO4, 200 mM, pH 7.0) and lysed for 10 min using a
sonicator (40% amplitude, 9 sec on/ 20 sec off). The lysed cell suspension was spun down at
11000 rpm for 40 min until a clear solution was obtained. Lysate was stored at 4°C overnight. To
a 1L CLR, clarified lysate (80 mL) was charged followed by potassium phosphate buffer
(KH2PO4 / K2HPO4, 400 mL, 200 mM, pH 7.0). The reaction was started by the addition of racꢀ2
(20.0 g, 162 mmol, final conc 50 g/L). The reaction progress was monitored by GC analysis of
the remaining nitrile 2 and was stopped after 3 h 15 min when the ee of 2 was >98%. The
reaction mixture was quenched by adjusting the pH (7.1) to 8.1 by addition of NaOH (5 M) and
subsequently adding EtOAc (400 mL). The mixture was stirred for 10 min (200 rpm), then left
without stirring for 30 min to allow separation of the organic and aqueous layers. A partial
emulsion was observed and this was partially broken by filtration through cotton wool. A second
extraction of the aqueous layer with EtOAc (400 mL) resulted in no further emulsion formation.
The organic layers were combined and washed with brine twice (300 mL) followed by a
saturated potassium bicarbonate wash (200 mL). The combined organic layers were dried over
MgSO4 and subsequently concentrated in vacuo to give (S)ꢀ2 as a yellow oil (5.6 g, 28% yield,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
13
99% ee). H and C spectra of the product were found to be identical with the spectra reported
for the ERED catalysed reduction of enone 4.
ASSOCIATED CONTENT
Supporting Information.
The Supporting Information is available free of charge on the ACS Publications website at
DOI:
ACS Paragon Plus Environment
26

Page 27 of 36
Organic Process Research & Development
1
2
3
4
NMR spectra and GC chromatograms (file type, PDF)
5
6
AUTHOR INFORMATION
7
8
9
Corresponding Author
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
*Eꢀmail: gheorghe.d.roiban@gsk.com.
ORCID
GheorgheꢀDoru Roiban: 0000ꢀ0002ꢀ5006ꢀ3240
Present Addresses
† Current address: Global Discovery Chemistry, Novartis Institute of Biomedical Research,
Novartis Pharma AG, Lichtstrasse, Basel CHꢀ4056 Basel, Switzerland.
Author Contributions
The manuscript was written through contributions of all authors. All authors have given approval
to the final version of the manuscript.
ACKNOWLEDGMENT
This research was supported by the Innovative Medicines Initiative (IMI) joint undertaking
project CHEM21 under grant agreement no. 115360, resources of which are composed of
financial contribution from the European Union’s Seventh Framework Programme (FP7/2007ꢀ
2013) and EFPIA companies in kind contribution. We thank Prof. Manfred Reetz and Dr.
Adriana Ilie for providing ERED variants C96DꢀYqjM and C96GꢀYqjM.
REFERENCES
ACS Paragon Plus Environment
27

Organic Process Research & Development
Page 28 of 36
1
2
3
4
5
6
7
8
9
(1) Patel, R. N., Biocatalytic Routes to Chiral Intermediates for Development of Drugs. In
Biocatalysis for Green Chemistry and Chemical Process Development, John Wiley & Sons, Inc.:
2011; pp 89ꢀ149.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(2) McLachlan, M. J.; Sullivan, R. P.; Zhao, H., Directed Enzyme Evolution and Highꢀ
Throughput Screening. In Biocatalysis for the Pharmaceutical Industry, John Wiley & Sons,
Ltd: 2009; pp 45ꢀ64.
(3) Science of Synthesis: Biocatalysis in Organic Synthesis, Thieme Verlag, 2015.
(4) Sheldon, R. A., Biocatalysis and Green Chemistry. In Green Biocatalysis, John Wiley &
Sons, Inc: 2016; pp 1ꢀ15.
(5) Bommarius, A. S.; Riebel, B. R., Introduction to Biocatalysis. In Biocatalysis, WileyꢀVCH
Verlag GmbH & Co. KGaA: 2005; pp 1ꢀ18.
(6) Gröger, H.; Asano, Y., Introduction – Principles and Historical Landmarks of Enzyme
Catalysis in Organic Synthesis. In Enzyme Catalysis in Organic Synthesis, WileyꢀVCH Verlag
GmbH & Co. KGaA: 2012; pp 1ꢀ42.
(7) Whittall, J.; Sutton, P. W.; Kroutil, W. ed. Practical Methods for Biocatalysis and
Biotransformations 3, John Wiley & Sons, Ltd, 2016.
(8) Sheldon, R. A.; Woodley, J. M., Role of Biocatalysis in Sustainable Chemistry. Chem.
Rev., 2018, 118, 801–838.
(9) Kern, M.; Roiban, G.ꢀD.; Fosberry, A.; and Snajdrova, R., GSK: Biocatalyst Discovery and
Optimisation. in Biocatalysis: An Industrial Perspective, Dominguez de Maria, P. and de
Gonzalo, G. ed., RSC, 2018, pp. 257ꢀ275.
ACS Paragon Plus Environment
28

Page 29 of 36
Organic Process Research & Development
1
2
3
4
5
6
(10) Gordon, R. D.; Holm, P.; Lademann, A.ꢀM.; Norling, T. Tacrolimus for improved
treatment of transplant patients. US. Pat. US8664239B2, 2014.
7
8
9
(11) Pospisilik, K.; Thijs, L. Process for making desvenlafaxine. US. Pat. US7491848B2,
2009.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(12) Ansari, S. A.; Kulkarni, A. K.; Ravikumar, B. Process for preparing tramadol
hydrochloride and/or tramadol monohydrate. EU. Pat., EP1346978A1, 2003.
(13) Glombik, H.; Stapper, C.; Falk, E.; Keil, S.; Schaefer, H.ꢀL.; Wendler, W.; Knieps, S. 2ꢀ{ꢀ
3 2ꢀ(phenyl)ꢀoxazolꢀ4ꢀylmethoxylꢀcyclohexylmethoxy}ꢀpropionic acid derivatives used as ppar
ligands (peroxisome proliferatorꢀactivated receptors) for the treatment of hyperlipidaemie and
diabetes, WO 2006/018115A1, 2006.
(14) Ding, Y.; Thalji, R.K.; Marino, Jr., J. P. Novel SeH inhibitors and their use. US. Pat.
US2010249137A1, 2010.
(15) Kysilka, V.; Mengler, J.; Kacer, P.; Cerveny, L. Process for the preparation of an
oxaliplatin. US. Pat. US20100174102A1, 2011.
(16) Wijemanne, S.; Jankovic, J. Restless legs syndrome: clinical presentation diagnosis and
treatment. Sleep Medicine, 2015, 16, 678ꢀ690.
(17) Ceccarelli, R.; Insogna, S.; Bella, M., Organocatalytic regioselective Michael additions of
cyclic enones via asymmetric phase transfer catalysis. Org. Biomol. Chem., 2006, 4, 4281ꢀ4284.
(18) Tanaka, Y.; Kanai, M.; Shibasaki, M., A Catalytic Enantioselective Conjugate Addition of
Cyanide to Enones. J. Am. Chem. Soc., 2008, 130, 6072ꢀ6073.
ACS Paragon Plus Environment
29