Development of Large-Scale Routes to Potent GPR119 Receptor Agonists

Article abstract of DOI:10.1021/acs.oprd.6b00157

Practical and scalable syntheses were developed that were used to prepare multikilogram batches of GSK1292263A (1) and GSK2041706A (15), two potent G protein-coupled receptor 119 (GPR119) agonists. Both syntheses employed relatively cheap and readily avai

Full text of DOI:10.1021/acs.oprd.6b00157

Subscriber access provided by United Arab Emirates University | Libraries Deanship
Full Paper
Development of Large-Scale Routes to Potent GPR119 Receptor Agonists
Richard T Matsuoka, Eric E Boros, Andrew D Brown, Kae Miyake Bullock, Will L. Canoy, Andrew J. Carpenter,
Jeremy D. Cobb, Shannon E. Condon, Nicole M. Deschamps, Vassil I. Elitzin, Greg Erickson, Jing M. Fang,
David H. Igo, Biren K. Joshi, Istvan W. Kaldor, Gregory E. Peckham, Daniel W. Reynolds, Matthew C.
Salmon, Matthew J Sharp, Elie A. Tabet, Jennifer Fell Toczko, Lianming Michael Wu, and Xiao-ming M. Zhou
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.6b00157 • Publication Date (Web): 13 Jul 2016
Downloaded from http://pubs.acs.org on July 14, 2016
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 free 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 accessible to all
readers and 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.
Organic Process Research & Development 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 9
Organic Process Research & Development
1
2
3
4
5
6
Development of Large-Scale Routes to Potent GPR119 Receptor
Agonists
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
,
†
Richard T. Matsuoka,* Eric E. Boros, Andrew D. Brown, Kae M. Bullock, Will L. Canoy, Andrew
J. Carpenter, Jeremy D. Cobb, Shannon E. Condon, Nicole M. Deschamps, Vassil I. Elitzin, Greg
†
Erickson, Jing M. Fang, David H. Igo, Biren K. Joshi, Istvan W. Kaldor, Gregory E. Peckham,
Daniel W. Reynolds, Matthew C. Salmon, Matthew J. Sharp, Elie A. Tabet, Jennifer F. Toczko,
‡
†
Lianming Michael Wu, and Xiao-ming M. Zhou.
API Chemistry Department, Analytical Science & Development Department, Particle Sciences and Engineering De-
partment, GlaxoSmithKline, 709 Swedeland Road King of Prussia, Pennsylvania 19406, United States
ABSTRACT: Practical and scalable syntheses were developed that were used to prepare multi-kilogram batches of
GSK1292263A (1) and GSK2041706A (15), two potent G protein-coupled receptor 119 (GPR119) agonists. Both syntheses
employed relatively cheap and readily available starting materials and both took advantage of an S Ar synthetic strategy.
N
INTRODUCTION
the methyl aryl sulfone portion of 1 was the last step of
the molecule’s synthesis, namely the conversion of aryl
sulfide 9 with Oxone to sulfone 1. Several API-related
GSK1292263A (1), a G protein-coupled receptor 119
TM
(
GPR119) agonist, has been explored as a treatment for
1
hydroperoxide impurities were also found to form during
the final stage of the synthesis where a re-crystallization
procedure was used to control the polymorphic form of
type 2 diabetes (Figure 1). Numerous articles describing
GPR119 agonists have been disclosed, and a number of
these compounds have been progressed into clinical tri-
5
2-3
als.
the API. These hydroperoxide impurities showed up, for
This particular compound was on a rapid timeline
example, when 1 was re-crystallized in methyl ethyl ke-
tone. Three major API-related hydroperoxide impurities;
namely 10, 11, and 12; were identified in some of the clini-
cal batches made from the initial route (see Figure 2).
The presence of these hydroperoxide impurities did not at
first appear to be a problem. Unfortunately, when two of
the three hydroperoxides identified, namely 11 and 12,
were shown to be mutagenic via Ames testing, the pres-
ence of these hydroperoxides became a critically im-
portant issue in regards to this route of synthesizing API
for development when the project entered our process
chemistry department; this required quick identification
of a synthetic route which offered the versatility of being
robust and safe for both immediate lab-scale synthesis to
produce gram quantities of drug substance and for subse-
quent rapid scale-up to pilot-plant kilogram quantities.
6,7
1
.
1
1
0: R = OOH, R = H, R = H
1 2 3
O
O
1: R = H, R = OOH, R = H
1
2
3
S
12: R = H, R = H, R = OOH
1 2 3
H C
3
N
Figure 1. Structure of GSK1292263A (1)
R1
R2
O
N
N
Original Process. The initial route of synthesis, em-
ployed for the early deliveries of 1, was convergent and
consisted of six steps (see Scheme 1). The overall 43%
R3
O
N
4
Figure 2. Structure of the three mutagenic hydroperox-
ide impurities found in the API.
yield of this route, from 6-bromopyridin-3-ol (2), was very
good. However, the active pharmaceutical ingredient
(API) 1 contained relatively high levels of various hydrop-
eroxide impurities. These hydroperoxide impurities arose
partly because the oxidation procedure used to synthesize
ACS Paragon Plus Environment

Organic Process Research & Development
Page 2 of 9
a
Scheme 1. Initial Route to GSK1292263A (1)
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Reagents and conditions: a) 5% Pd(OAc) , 15% PPh , K CO , IPA/water, isolated as the HCl salt in 60-72% yield; b) ace-
2
3
2
3
TM
tonitrile, 75-85% yield; c) MsCl, Et N, DCM; d) K CO , DMSO, 75-80% yield; e) Oxone , EtOH-water, H SO , 95% yield;
f) methyl ethyl ketone re-crystallization to correct form (polymorph), 85-90% yield.
3
2
3
2
4
RESULTS AND DISCUSSION
cially when the solvent utilized was methyl ethyl ketone
MEK). An extensive solvent screen was undertaken to
(
At first, methods to remove the hydroperoxides from
API 1 synthesized via the initial route were studied in or-
der to find a way to obtain 1 with acceptable hydroperox-
ide levels. The sum of hydroperoxides 11 and 12 in the API
had to be controlled to <5 ppm. These experiments re-
vealed that the hydroperoxides related to 1 react rapidly at
room temperature with triphenylphosphine (TPP) to af-
ford triphenyl-phosphine oxide (TPPO) and presumably
the corresponding alcohols. For example, treating a THF
solution of 1 containing hydroperoxides at a total level of
6
divinylbenzene polymer resin, containing ~3 mmol of
covalently bound TPP per gram of resin, for just an hour
at room temperature eliminated approximately 98% of
the hydroperoxides. Nevertheless, identifying either a
modified route of synthesizing 1 without forming any hy-
droperoxides or a reliable re-crystallization procedure to
remove potential hydro-peroxides to safe levels was felt to
be a more ideal long term strategy going forward.
find an alternative re-crystallization solvent to MEK. A
seeded re-crystallization of the API 1 from anhydrous
2-MeTHF afforded the desired polymorph with minimal
hydroperoxide formation in comparison to the previous
solvent. The presence of the stabilizer butylated hydroxy-
toluene (BHT) in the anhydrous 2-MeTHF solvent was
hypothesized as playing a role in minimizing the hydrop-
eroxide formation. However, this modified route still
included the use of potentially genotoxic methanesul-
fonate 8 which needed to be controlled at low levels in
the drug substance. Primarily for this reason, an alterna-
tive synthesis was desired.
8
9
ppm with commercially available styrene-
9
The second GPR119 agonist molecule. At about this
time, our department began working on the
GSK2041706A (15) API molecule (see Figure 3). This
Modified Route. The second route of synthesizing 1
reduced the levels of these hydroperoxide impurities by
moving the sulfide-to-sulfone transformation one step
earlier in the synthetic sequence (see Scheme 2); this
strategy provided an additional opportunity to purge the
unwanted hydroperoxides via an additional crystalliza-
tion. Unfortunately, the hydroperoxide issue did not
completely disappear as hydroperoxide impurities were
still found to form in the final re-crystallization procedure
used to control the polymorphic form of the API; espe-
Figure 3. Structure of GSK2041706A (15)
ACS Paragon Plus Environment

Page 3 of 9
Organic Process Research & Development
a
Scheme 2. Second Route to GSK1292263A (1)
1
2
3
4
5
6
7
8
9
a
TM
Reagents and conditions: a) 1.2 eq Oxone , EtOH/water; b) K CO , DMSO; c) DCM/aq. 5% NaCl solution extraction;
2
3
d) 2-MeTHF crystallization; e) seeded 2-MeTHF re-crystallization to correct form (polymorph).
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
molecule represented a second GPR119 agonist candi-
date with the company. The convergent synthesis of 15,
which our department initially received for scale-up, effi-
ciently employed an asymmetric hydrogenation to syn-
thesize the stereogenic center of (R)-1-(pyridin-4-
yl)ethanol (18) from the corresponding achiral 1-(pyridin-
4
Figure 5. Structure of 5-bromopyrazin-2-ol (25) and the
unwanted Mistunobu reaction side-product (26)
4
-yl)ethanone (16) (see Scheme 3). The strategy em-
ployed in this asymmetric transformation using RuCl(p-
1
0
cymene)[(R,R)-Ts-DPEN] (24) had been reported earlier
1
1
(
Figure 4). This initial synthesis of API 15, however, suf-
Mitsunobu reaction (see Figure 6). The yield of the
Mitsunobu reaction was also low due to the inherent dif-
ficulties of isolating 15 away from the triphenylphosphine
oxide produced in the reaction.
fered from a couple of challenges that made further scale-
up difficult. These challenges included a low yielding
Mitsunobu reaction between 19 and 20 and the genera-
tion of a potentially dangerous aryl diazonium salt inter-
mediate during the conversion of pyrazinamine 23 to py-
razinol 20.
Figure 6. Structure of mesylate 27
Alternative strategies for synthesizing 15 were quickly
examined. The low yielding Mitsunobu reaction was suc-
cessfully replaced with a nucleophilic aromatic substitu-
Figure 4. Structure of chiral catalyst 24
tion reaction (S Ar) where alcohol 31, the enantiomer of
N
alcohol 19, was reacted with aryl chloride 29 under basic
conditions to afford 15 in very high yield (see Scheme 4).
Chiral alcohol 31 was synthesized in the same manner as
19 with the exception that the ruthenium-based asym-
metric reduction reaction with 16 was carried out using a
catalyst containing the ligand of the opposite enantiomer
of that in 24, namely RuCl(p-cymene)[(S,S)-Ts-DPEN].
Fortunately, the cost of this second catalyst was identical
to its enantiomer. A Suzuki-Miyaura reaction was used to
synthesize 29 from boronic acid 21 and 2-bromo-5-
chloropyrazine 28 in very good yield. The modified syn-
Attempts to synthesize pyrazinol 20 directly from cou-
pling boronic acid 21 and 5-bromopyrazin-2-ol (25) were
surprisingly disappointing (see Figure 5); even after ex-
tensive screening of a variety of reaction conditions, Su-
zuki-Miyaura conditions to afford this transformation in
high yield could not be found. The chemical nature of 25
in regards to oxidative addition must be fairly different
from 6-bromo-3-pyridinol (2) since the latter was success-
fully used in an analogous Suzuki-Miyaura reaction in the
synthesis of 1.
thesis of 15 using the S Ar strategy affords the GPR119
N
Improving the yield of the low-yielding Mitsunobu re-
action was challenging primarily because of a concomi
tant olefin side-product formation; this olefin was theo
rized to be 26 although its precise chemical structure was
not conclusively identified (see Figure 5). Under a variety
of different Mitsunobu reaction conditions examined,
olefin 26 was consistently generated at approximately 16%
conversion from the non-isolated alcohol 19. Olefin 26
target molecule in five convergent steps in approximately
3% overall yield. This represents an increase in the over-
all yield of about 10% in comparison to the initial scale-up
3
route of synthesis. More importantly, the S Ar strategy
N
eliminated the need to employ potentially hazardous dia-
zonium salt chemistry to synthesize intermediate hydroxy
aromatic molecules such as 20. All stages of the modified
synthesis were successfully scaled-up to
was generated in similar quantities even when S 2 condi-
N
tions were employed using mesylate 27 in lieu of the
ACS Paragon Plus Environment

Organic Process Research & Development
Page 4 of 9
a
Scheme 3. Initial Route to GSK2041706A (15)
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Reagents and conditions: a) 0.3 mol% Chiral Catalyst 24, formic acid, triethylamine, 80% yield (92% ee); b) 15 wt%
Pt(O) , hydrogen (1 atm), acetic acid, methanol; c) aqueous sodium hydroxide/MTBE, d) MTBE recrystallization to in-
crease enantiomeric purity to >95% ee (47% yield over two steps); e) 15 mol% DBU, methanol, 1-2 days, >90% yield; f)
2
Ph P, DIAD, THF (crystallized from aqueous methanol), 50% yield (>95% ee); g) re-crystallization in aqueous acetonitrile
3
to correct form; h) Na CO , Pd(PPh ) , p-dioxane/MeOH, reflux, 98% yield; i) NaNO , H SO , water, 90% yield.
2
3
3
4
2
2
4
b
Scheme 4. Final Route to GSK2041706A (15)
b
Reagents and conditions: a) DBU, MeOH; b) 0.4-0.5 mol% Pd(dppf)Cl · DCM, aq. 2M Na CO solution, DME, 70-90%
2
2
3
yield; c) 2 eq NaOtBu, THF, Darco G-60 (activated carbon), 70-85% yield; d) seeded aqueous acetonitrile re-slurry to cor-
rect form (polymorph), 95-99% yield.
c
Scheme 5. Final Route to GSK1292263A (1)
c
Reagents and conditions a) 0.25 mol% Pd(t-Bu P) , triethylamine, EtOH/water, 85-95% yield; (b) KOtBu, DMPU/THF,
3
2
85-95% yield; (c) Darco G-60 (activated carbon), 2-MeTHF, seeded re-crystallization to correct form (polymorph) 75-85%
yield.
ACS Paragon Plus Environment

Page 5 of 9
Organic Process Research & Development
synthesize about 5 kilograms of 15. Hydroperoxide
Semi-solid 2-bromo-5-chloropyrazine (28) (4.06 kg,
21.0 mol), solid (4-(methanesulfonyl)phenyl)boronic acid
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
formation was not as much of an issue for 15 versus the
GPR119 agonist 1; the recrystallization of 15 in aqueous
acetonitrile was employed to ensure that the correct form
(21) (4.96 kg, 24.8 mol), and solid PdCl (dppf) [1,1'-
2
bis(diphenylphosphino)ferrocene palladium (II) chloride]
1
2
was synthesized.
as
a
complex with dichloromethane (0.0857 kg,
0
.105 mol), and finally DME (20.3 L) were charged to a
Final Route. Based upon the success of the new 15 syn-
reaction vessel at room temperature. The resulting slurry
was stirred at 25 °C and then treated with an aqueous
thesis, an analogous S Ar approach to the synthesis of 1
N
was examined (see Scheme 5). The Suzuki reaction be-
tween boronic acid 21 and 2-bromo-5-fluoropyridine 32
afforded the coupled product 33 in very good yield. This
Suzuki-Miyaura reaction was initially carried out using in
2.0 M sodium carbonate solution (28.4 L). The resulting
biphasic reaction mixture was heated to 65-70 °C with
stirring and stirred at 65-70 °C for 7 h. The reaction mix-
ture was cooled to 25 °C and treated with water (40.6 L).
The reaction mixture was stirred for 1 h and filtered. The
reaction vessel was rinsed with water (2 X 20.3 L) and this
rinse was added to the wet cake. The wet cake was
charged to a clean vessel. The wet cake was treated an
aqueous 1.0 N sodium hydroxide solution (81.2 L) with
stirring at 25 °C. The resulting slurry was stirred at 25 °C
for 2 h and then filtered. The wet cake was washed with
an aqueous 10 wt% ammonium chloride solution (40.6 L)
and with water (40.6 L). The wet cake was air dried for
2 h. The partially dried wet cake was then charged to a
clean vessel. The wet cake was treated with dichloro-
methane (101.5 L) at 25 °C. The resulting slurry was
stirred at 25 °C for 2 h and then filtered through a Celite
impregnated filter pad (ErtelAlsop Micro-Media XL Series
Filter Pad) to remove the unwanted boric acid salts. The
pad of Celite was rinsed with dichloromethane (2 X
81.2 L) to retrieve any desired product 29 that may have
stuck to the pad. The filtrates were combined and con-
centrated about 40.6 L. The resulting slurry was treated
with MTBE (81.2 L) and concentrated to approximately
40.6 L. The slurry was treated with MTBE (81.2 L) and
concentrated to about 40.6 L. The slurry was treated one
final time with MTBE (81.2 L) and concentrated to ap-
proximately 60.9 L. The slurry was heated to approxi-
mately 50 °C and stir at 50 °C for 1 h. The slurry was
cooled to 25 °C. The slurry was stirred at 25 °C for 1 h and
then filtered. The wet cake was rinsed with MTBE (2 X
20.3 L). The wet cake was dried in a vacuum oven at 45 °C
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
situ
prepared
tetrakis-(triphenylphosphine)-palla-
dium (0). Subsequent experimentation showed that the
palladium loading could be reduced from 0.05 eq to
.0025 eq. by switching to Pd(tert-Bu3P)2. The strategi-
0
cally important S Ar reaction between 33 and alcohol 7
N
also proceeded smoothly and in very good yield. Initially
carried out using sodium tert-butoxide in a mixed solvent
system of DMSO/THF, the ultimate S Ar reaction condi-
N
tions employed potassium tert-butoxide as base in a
mixed solvent system of DMPU/THF. All stages of the
synthesis were later successfully scaled-up to synthesize
about 71 kilograms of API 1 in multiple 20-25 kilogram
batches.
CONCLUSIONS
The process development and synthesis for two GPR119
receptor agonists, namely GSK1292263A (1) and
GSK2041706A (15), has been described. An S Ar strategy
N
was found to be applicable to the synthesis of both API
targets; such a strategy enabled the employment of rela-
tively cheap and commercially available starting materi-
als. Highly cost-effective and robust processes for the
large-scale manufacture of both molecules were subse-
quently developed.
EXPERIMENTAL SECTION
General Procedures. All materials were purchased from
commercial suppliers and used without further purifica-
tion. All non-aqueous reactions were performed in dry
glassware or glass-line reactors under an atmosphere of
dry nitrogen. Organic solutions were concentrated via
rotary evaporation at about 30 mmHg at less than 55 °C
except where noted.
to afford 4.18 kg (15.6 mol; 74.2% yield) of 29 as a cream-
1
colored solid. H NMR (DMSO-d , 400 MHz) δ 9.24 (d, J
6
= 1.5 Hz, 1H), 8.93 (d, J = 1.2 Hz, 1H), 8.38 (d, J = 8.5 Hz,
1
3
2H), 8.09 (d, J = 8.5 Hz, 2H), 3.29 (s, 3H). C NMR (100.6
MHz, DMSO-d ) 149.0, 148.5, 144.6, 142.7, 142.4, 139.8,
6
128.2, 128.1, 43.9. HRMS calcd for C H ClN O S (M + H)+
11 10
2
2
269.0146, found, 269.0143.
NMR spectra were measured on a Bruker DRX400 op-
erated at 400 and 100 MHz, for 1H and 13C, respectively,
with data reported as follows: chemical shift (ppm), mul-
tiplicity (s, singlet; d, doublet; t, triplet; q, quartet; m,
multiplet), integration and coupling constant (J, Hz).
Elemental analysis, high resolution mass spectra, melt
solvates, infrared spectra (IR), and thermogravimetric
analyses (TGA) were obtained from GlaxoSmithKline
Analytical Chemistry Department.
Preparation of (S)-1-(1-(3-isopropyl-1,2,4-oxadiazol-5-
yl)piperidin-4-yl)ethanol (31).
Solid (S)-1-(piperidin-4-yl)ethanol (30) (6.99 kg, 54.1
mol), 3-isopropyl-5-(trichloromethyl)-1,2,4-oxadiazole (6)
(
14.96 kg, 65.2 mol), and MeOH (7.69 L) were charged to
a reaction vessel at room temperature. The resulting so-
lution was stirred at 25 °C. The reaction mixture was then
treated with DBU (1.24 kg, 8.1 mol) and MeOH (7.69 L).
The reaction mixture was stirred at 25 °C for 72 h. The
reaction mixture was cooled to 20 °C and then slowly
treated with aqueous 2.0 N sodium hydroxide solution
Preparation
of
2-chloro-5-(4-(methylsulfonyl)phe-
nyl)pyrazine (29).
ACS Paragon Plus Environment

Organic Process Research & Development
Page 6 of 9
(
15.4 L) at a rate such that the process temperature re- 25 °C overnight. The contents of the reactor were filtered.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
mained below 30 °C. The reaction mixture was then
stirred at 25 °C for 2 h during which time all of the re-
maining starting material 6 decomposed to various non-
chromophoric ring-opened molecules. The reaction mix-
ture was diluted with MTBE (69.9 L). The resulting two-
phase solution was vigorously stirred before the layers
were allowed to settle and separate. The lower aqueous
layer was drained and saved. The organic layer was
drained and saved. The vessel was charged with the
aqueous layer. The aqueous layer was extracted with ad-
ditional MTBE (2 X 35.0 L). The organic layers were com-
bined and charged into a vessel. The organic layer was
washed with with an aqueous 10% w/w sodium dihy-
drogenphosphate solution (2 X 30.8 L) followed with an
aqueous 10% w/w sodium carbonate solution (2 X 30.8 L).
The organic layer containing product 31 was concentrated
in vacuo to a minimum stir volume. The concentrated
solution was treated with THF (139.8 L) and concentrated
again in vacuo to about 24.5 L. The product 31 as a solu-
tion in THF was used in the following step without fur-
ther purification.
The wet cake was washed with with an aqueous methanol
solution (12 L) (3:2 ratio of MeOH/water). The wet cake
was dried in a vacuum oven at 60 °C to afford 3.84 kg (8.13
mol; 72.8% yield) of 15 as an off-white colored solid.
Conversion of GSK2041706A (15) as a mixture of poly-
morphic forms to a single form
Solid 15 (1.90 kg, 4.02 mol) was charged to a vessel and
then treated with an aqueous 33.2 %v/v acetonitrile solu-
tion (47.5 L) at room temperature with stirring. The re-
sulting slurry was treated with crystalline seed material
and then stirred at room temperature overnight. The
slurry was filtered and the wet cake was washed with an
aqueous 33.2 %v/v acetonitrile solution (9.5 L). The crys-
tals were dried in a vacuum oven at 60 °C to afford 1.82 kg
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(3.85 mol; 95.8% yield) of 15 as an off-white colored solid.
1
H NMR (400 MHz, DMSO-d ) δ 8.91 (bs, 1H), 8.40 (bs, 1
6
H), 8.28 (d, J = 8.5 Hz, 2H), 8.02 (d, J = 8.5 Hz, 2H), 5.17-
5
.09 (m, 1H), 4.09-3.95 (m, 2H), 3.27 (s, 3H), 3.16-2.99 (m,
2H), 2.80 (q, J = 6.9 Hz, 1H), 1.98-1.85 (m, 2H), 1.83-1.70
(
=
m, 1H), 1.47-1.33 (m, 2H), 1.31 (d, J = 6.3 Hz, 3H), 1.17 (d, J
1
3
6.8 Hz, 6H). C NMR (100.6 MHz, DMSO-d ) 175.3,
6
Preparation of GSK2041706A (15) as a mixture of poly-
1
4
70.9, 159.8, 142.6, 141.2, 141.0, 139.1, 135.7, 128.1, 126.9, 75.7,
6.0, 45.9, 44.0, 40.2, 27.1, 27.0, 26.7, 20.7, 16.9. HRMS
morphic forms
+
Solid 29 (3.00 kg, 11.16 mol) (1 wt, 1 eq), 31 [2.81 kg, 11.74
mol; a solution in THF (about 12 L)], and then THF (4.5 L)
were charge to a reactor. The resulting reaction mixture
was cooled to 0 °C and then slowly treated with sodium
tert-butoxide (2.14 kg, 22.27 mol) as a solution in THF (15
L) at a rate such that the process temperature remained
below 10 °C. After the addition of the base was complete,
the process temperature was increased to about 25 °C.
The reaction mixture was stirred at about 25 °C for 2 h.
The reaction mixture was treated with acetic acid (0.670
kg, 11.2 mol) and then with water (9 L). The resulting bi-
phasic reaction mixture was stirred at about 25 °C for 1 h
and then treated with aqueous 20% NaCl solution (3 L)
and stirred for about 15 min. The layers were separated
and the organic layer was treated with aqueous 20% w/w
sodium bisulfite solution (12 L). The process temperature
was increased to 60 °C and the contents of the reactor
were stirred for 1 h at 60 °C. The process temperature was
deceased to 25 °C. The layers were separated and the or-
ganic layer was treated with Darco G-60 powder. The
process temperature was increased to 60 °C and the con-
tents of the reactor were stirred for 1 h. The hot solution
was filtered through a Celite impregnated filter pad (Er-
telAlsop Micro-Media XL Series Filter Pad) into multiple
calcd for C H N O S (M + H) 472.2013, found, 472.2009.
23 30 5 4
Preparation of (1-(3-isopropyl-1,2,4-oxadiazol-5-yl)-
piperidin-4-yl)methanol (7)
Liquid 6 (73.0 kg, 318.10 mol), piperidin-4-ylmethanol
(5) (73.0 kg, 633.85 mol) and acetonitrile (54.8 L) were
charged to a vessel at ambient temperature. The resulting
slurry was then heated to 60 °C; during the heating pro-
cess the reaction mixture became a homogeneous solu-
tion. The reaction mixture was stirred at 60 °C for 16 h.
The reaction mixture was cooled to 20 °C, treated with
aqueous 1 M hydrochloric acid solution (350.4 L) and then
stirred at that temperature for 30 min. The vessel was
charged with toluene (292 L) and the resulting biphasic
solution was stirred for 30 min. The layers were allowed
to settle and separate. The bottom aqueous layer was
separated and saved. The organic layer was saved. The
aqueous layer was placed into a vessel and extracted with
toluene (2 X 146 L). The organic layers were combined
and concentrated in vacuo to afford 7 as a brown-colored
solvent-free solution (219 L). The crude product 7 was
used in the following step without further purification.
2
0-L containers. The contents of the container were fil-
Preparation
of
5-fluoro-2-(4-(methylsulfonyl)-
tered through a clarifying filter cartridge into a second
clean reactor. The organic layer was concentrated in vac-
uo to about 9 L. The contents of the reactor were diluted
with isopropanol (18 L). The reaction mixture was con-
centrated in vacuo to about 15 L. The reactor was charged
with water (18 L). The process temperature was increased
to 70 °C over 45 min. The contents of the reactor were
stirred at 70 °C for about 1 h and then slowly cooled to
phenyl)pyridine (33)
Solid (4-(methanesulfonyl)phenyl)boronic acid (21)
34.4 kg, 171.98 mol), bis(tri-t-butyl phosphine)palladium
0) (0.2025 kg, 0.396 mol), ethanol (139.5 L), solid 2-
(
(
bromo-5-fluoropyridine (32) (27.9 kg, 158.53 mol), and
water (139.5 L) were charged into a reactor. The reaction
was heated with stirring to 25 °C. Triethylamine (24.1 kg,
238.17 mol) was then charged to the reactor at such a rate
2
5 °C over 3 h and the resulting slurry was left to stir at
ACS Paragon Plus Environment

Page 7 of 9
Organic Process Research & Development
L) and then subjected to atmospheric distillation to re-
that the reaction temperature remained between 25 °C
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
and 35 °C. After the addition of triethylamine was com-
plete, the reaction mixture was heated to about 55 °C and
stirred at that temperature until complete by HPLC (reac-
tion was deemed complete when <1.5% 32 was seen). The
reaction was cooled to about 20 °C, treated with water
move the THF solvent. Once the distillation was com-
plete, the solution temperature was adjusted with stirring
to 70 °C. The solution temperature was then slowly de-
creased with stirring from 70 °C to 50 °C at a rate of 0.1 °C
per minute and then from 50 °C to 0-5 °C at a rate of
0.25 °C per minute. The resulting slurry was filtered; the
wet cake was dried in a vacuum oven at 65 °C to afford
(
279 L), and the mixture was stirred for 2 h. The resulting
solids were filtered and the wet cake was washed with
water (2 X 69.8 L). The wet cake 33 was charged into a
vessel and dissolved in THF (399.0 L). Solid L-cysteine
50.4 kg (110.4 mol; 92.5% yield) of 1 as a white colored
1
solid. H NMR (400 MHz, DMSO-d ) δ 8.44 (d, J = 3.0 Hz,
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
(
8.09 kg) (a palladium scavenger) was dissolved in water
198.1 L) and added to the 33 solution. The resulting mix-
1H), 8.28 (d, J = 8.8 Hz, 2H), 8.06 (d, J = 8.8 Hz, 1H), 7.99
(bd, J = 8.5 Hz, 2H), 7.54 (dd, J = 8.8, 3.0 Hz, 1H), 4.03 (d,
J=6.3 Hz, 2H), 4.03-3.97 (m, 2H), 3.25 (s, 3H), 3.20-3.09
(m, 2H), 2.81 (q, J = 6.7 Hz, 1H), 2.13-2.00 (m, 1H), 1.88 (bd,
J = 12.8 H, 2H), 1.42-1.29 (m, 2H), 1.18 (d, J = 7.0 Hz, 6H).
ture was stirred at about 60 °C for 1 h. The reaction mix-
ture was filtered and transferred to a clean vessel. The
layers were allowed to settle and separate. The aqueous
bottom layer was separated and transferred to a vessel
and back-extracted with DCM (139.5 L). The biphasic
solution was stirred for 15 min; the stirring was halted and
the layers were allowed to separate. The DCM-based or-
ganic layer, after filtration, was combined with the first
THF-base organic layer with stirring. After the stirring
was halted and the layers separated, the aqueous layer
was discarded. The DCM/THF based organic layer con-
taining 33 was then distilled down to about 279 L. Etha-
nol (390.6 L) was added and the solution was concentrat-
ed to about 279 L. The temperature of the ethanol solu-
tion was adjusted to about 55 °C and then slowly cooled
to about 0 °C at a rate of 0.5 °C/min. The resulting slurry
was then stirred for at least 2 hours at 0 °C. The slurry
was then filtered. The wet cake was washed with cold
EtOH (2 X 139.5 L). The wet cake was dried under vacuum
at about 50 °C to afford 35.8 kg (142.5 mol; 89.9% yield) of
1
3
C NMR (100.6 MHz, DMSO-d ) 175.3, 170.9, 155.5, 147.0,
6
143.5, 140.5, 138.6, 127.9, 127.0, 122.4, 122.3, 72.5, 45.7, 44.1,
35.0, 28.0, 26.7, 20.8. HRMS calcd for C H N O S (M +
2
3
29
4
4
+
H) 457.1904, found, 457.1900.
Anal. Calcd for
C H N O S: C, 60.51; H, 6.18; N, 12.27. Found: C, 60.64;
23
28
4
4
H, 6.16; N, 12.24.
Preparation of API grade GSK1292263A (1)
In a nitrogen purged vessel, a stirring suspension of 1
(
30.0 kg, 65.71 mol) in 2-methyl tetrahydrofuran (630 L )
initially at ambient temperature was heated to 70-80 °C.
The resulting homogeneous solution was transferred
through an appropriate in-line filter containing activated
carbon (Darco G-60, 5.01 kg) sandwiched between two
layers of filtering agent (Celite 545, 2 X 20.1 kg into a se-
cond nitrogen purged and pre-heated reactor vessel. The
filterpad was rinsed with additional methyl tetrahydrofu-
ran (60 L) and the filtered solution was concentrated in
vacuo to about 510 L. The concentrate was reheated to 70-
1
33 as a white colored solid. H NMR (400 MHz, DMSO-
d6) δ 8.73 (d, J = 2.8 Hz, 1H), 8.31 (bd, J = 8.8 Hz, 2H), 8.20
(dd, J = 8.8, 4.3 Hz, 1H), 8.04 (bd, J = 8.5 Hz, 2H), 7.91 (td,
1
3
80 °C. Once complete dissolution was established, the
J = 8.8, 3.0 Hz, 1H), 3.27 (s, 3H). C (100.6 MHz, DMSO-
batch temperature was reduced to 55 °C over approxi-
mately 30 minutes and seeded with GSK1292263A (0.15
kg) suspended in a minimal amount of filtered 2-methyl
tetrahydrofuran (0.18 L) (0.006 vol.). The mixture was
then cooled to 35 °C at a rate of 0.1 °C/min and then
cooled to 0 °C at a rate of 0.25 °C/min. The resulting slur-
ry was then stirred at about 0 °C for 1 h. The product was
filtered. The wet cake was rinsed with pre-chilled 2-
methyl tetrahydrofuran (2 X 60 L). The material was
dried at 60-65 °C under vacuum to afford 22.51 kg (49.30
mol; 75.0% yield) of 1 as a white colored solid.
d6) 158.3, 151.3, 142.7, 141.3, 138.4, 128.0, 127.8, 125.0, 124.8,
+
123.3, 123.2, 44.0. HRMS calcd for C H FNO S (M + H)
1
2
11
2
252.0489, found, 252.0485.
Preparation of intermediate grade GSK1292263A (1)
Potassium t-butoxide (28.1 kg, 250.42 mol) and THF
148.4 L) were added to a vessel. The resulting slurry was
(
cooled to 0-5 °C. The vessel was then charged with a so-
lution of crude 7 (29.58 kg, 131.33 mol) dissolved in DMPU
(
59.2 L) at a rate such that the temperature remained be-
low 15 °C. The reaction mixture was stirred for 30
minutes at 0-5 °C. The vessel was treated with a solution
of 33 (30.0 kg, 119.39 mol) in THF (121.5 L) and DMPU
ASSOCIATED CONTENT
Supporting Information
(
29.6 L) at a rate such that the reaction temperature was
kept below 15 °C. Once the addition was complete, the
reaction was held for about 30 min at 0-5 °C and then
warmed to 18-22 °C for approx-imately one hour. The
reaction was quenched with an aqueous sodium sulphite
solution [60.0 kg (476.03 mol) of solid sodium sulphite
dissolved in 360 L of water]. The resulting biphasic solu-
tion was heated with stirring to about 45 °C and then al-
lowed to settle and separate at 45 °C. The layers were
separated. The organic layer was treated with water (150
The Supporting Information is available free of charge on the
ACS Publications website.
Full characterization data for all new compounds (PDF)
AUTHOR INFORMATION
Corresponding Author
*E mail: richard.t.matsuoka@gsk.com.
ACS Paragon Plus Environment

Organic Process Research & Development
Page 8 of 9
†
API Chemistry Department
Analytical Sciences & Development Department
‡
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ACKNOWLEDGMENTS
We thank our GlaxoSmithKline colleagues in the former API
Chemistry, Analytical Science, Particle Science and Engineer-
ing Departments located in Research Triangle Park, North
Carolina for their contributions to the development and im-
plementation of this chemistry. The authors wish to thank
Svetlana A. Burova, Lynda A. Jones, Cecilia S. Koble, Mark B.
Mitchell, and Jim B. Thompson for their scientific contribu-
tions to the work reported in this manuscript. We also thank
Brenda Willoughby, the GlaxoSmithKline GPR119 Project
Manager, for her support and encouragement throughout
the writing of this paper.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
REFERENCES
(1) Ohishi, T.; Yoshida, S. Expert Opin. Invest. Drugs 2012, 21, 321.
(2) Dai, X.; Stamford, A.; Liu, H.; Neustadt, B.; Hao, J.; Kowalski
T.; Hawes, B.; Xu, X.; Baker, H.; O’Neill, K.; Woods, M.; Tang, H.;
Greenlee, W. Bioorg. Med. Chem. Lett. 2015, 25, 5291.
(3) For a review see Shah, U.; Edmondson, S.; Szewczyk, J. W. In
New Therapeutic Strategies for Type 2 Diabetes; Jones, Rob M.,
Eds.; RSC Drug Discovery Series; Royal Society of Chemistry:
Thomas Graham House, Science Park, Milton Road, Cambridge
CB4 0WF, 2012; Vol. 27, pp 177.
(
4) Fang, J.; Tang, J.; Carpenter, A. J.; Peckham, G.; Conlee, C. R.;
Du, K. S.; Katamreddy, S. R. US 2012/0077812 A1, 2012.
5) comprehensive crystal-form screening study of
(
A
GSK1292263A (1) indicated that four anhydrous polymorphic
forms can be produced.
(6) (a) ICH M7: Assessment and Control of DNA Reactive (Mu-
tagenic) Impurities in Pharmaceuticals to Limit Potential Car-
cinogenic Risk, Business Plan, 2014. (b) Jacobson-Kram, D.;
McGovern, T. Advanced Drug Delivery Reviews 2007, 59, 38.
(
7) Interestingly, hydroperoxide 10 was not found to be muta-
genic via an Ames test. Mortelmans, K.; Zeiger, E. Mutat. Res.
000, 455, 29-60.
8) Kinetic experiments described in the literature on the reac-
2
(
tions of TPP with various hydroperoxides have shown that the
“instantaneous” reaction is first order in TPP and hydroperoxide
and not inhibited by radical scavengers. See for example Hiatt,
R.; Smythe, R. J.; McColeman, C. Can. J. Chem. 1971, 49, 1707.
(9) This resin is often used for Wittig chemistry and then simply
removed by filtration. See for example Bernard ,M.; Ford, W. T.
J. Org. Chem., 1983, 48, 326.
(10) RuCl(p-cymene)[(R,R)-Ts-DPEN] is an abbreviation for
RuCl[(R,R)-N-(p-toluenesulfonyl)-1,2-diphenyl-ethylene-
diamine](p-cymeme).
(11) Okano, K.; Murata, K.; Ikariya, T. Tetrahedron Lett. 2000, 41,
9277.
(12)
A
comprehensive crystal-form screening study of
GSK2041706A (15) indicated that only two anhydrous polymor-
phic forms can be produced with melting points of 145.0 °C
(Form 1) and 129.3 °C (Form 2). Interestingly, Form 2 was found
to be more stable than Form 1 at temperatures below 60 °C as
the two forms are enantiotropically related. Hence, Form 2 be-
came the desired form for the API.
ACS Paragon Plus Environment

Page 9 of 9
Organic Process Research & Development
Table of Contents Graphic
1
2
3
4
5
6
7
8
9
O
O
O
O
S
S
H C
H C
3
3
N
N
H
CH3
O
O
(S)
N
N
N
N
O
O
N
N
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
GSK1292263A
GSK2041706A
9
ACS Paragon Plus Environment