New synthetic route to a dipeptidyl peptidase-4 inhibitor

Article abstract of DOI:10.1021/op200309z

A new synthetic route to a dipeptidyl peptidase-4 (DPP4) inhibitor was developed and demonstrated on a multigram scale. This approach takes advantage of the cheap and readily available Boc-trans-4-hydroxy-l-proline methyl ester as starting material which

Full text of DOI:10.1021/op200309z

Article
pubs.acs.org/OPRD
New Synthetic Route to a Dipeptidyl Peptidase-4 Inhibitor
Danny Lafrance* and Stephane Caron
́
Chemical R&D, Pfizer Worldwide R&D, MS-8118D/4009, Eastern Point Road, Groton Connecticut 06340, United States
ABSTRACT: A new synthetic route to a dipeptidyl peptidase-4 (DPP4) inhibitor was developed and demonstrated on a
multigram scale. This approach takes advantage of the cheap and readily available Boc-trans-4-hydroxy-L-proline methyl ester as
starting material which was derivatized through an SN2 reaction. Several leaving groups were studied, and the nosylate group
showed superiority over other derivatives. Formation of an amide using the most costly starting material, 3,3-difluoropyrrolidine,
was performed late in the synthesis to minimize its economical impact on the overall cost of the API.
that would utilize 5 as starting material is desirable from an
economical standpoint. Results related to this work are
summarized in this manuscript.
INTRODUCTION
■
PF-734200 is a dipeptidyl peptidase-4 (DPP4) inhibitor for the
treatment of diabetic neuropathy.1,2 The initial synthetic route
(Scheme 1), was successfully scaled up to support early
DISCUSSION
■
Scheme 2 presents an outline of the proposed new synthetic
route. The nucleophilic displacement of alcohol 5 is well
documented for the Mitsunobu-type reaction using DPPA.4,5
Although the displacement of an activated form (6) such as the
mesylate with sodium azide is known,6−8 direct SN2
substitutions of a sulfonyl-activated hydroxyproline by simple
amines are surprisingly absent from the literature. Some of the
rare examples include SN2 substitution with adenine9 and a
protected L-serine10 as the nucleophiles. If amine 7 could be
accessed diastereoselectively, completion of the synthesis
through carboxylic acid 8 should be straightforward. In light
of the few precedents for the key displacement to form
intermediate 7, we first sought to establish proof of concept for
the proposed bond disconnection by submitting the activated
hydroxyproline (6) to SN2 conditions with 1-(2-pyrimidyl)-
piperazine as the nucleophile. Results are summarized in Table
1.
Scheme 1. Initial route to PF-734200 (4)
development efforts. This approach begins with N-t-Boc-4-oxo-
L-proline (1) that undergoes a mixed anhydride activation with
pivaloyl chloride at 0 °C, followed by amidation with 3,3-
difluoropyrrolidine to yield the intermediate 2. Reductive
amination with 1-(2-pyrimidyl)piperazine using sodium
triacetoxyborohydride in THF/AcOH provided the desired
stereoisomer 3 in high yield and selectivity, the undesired
diastereomer being completely removed by crystallization.
Deprotection of 3 with 6 N HCl, followed by neutralization
with 50% NaOH and extraction provided PF-734200 (4) in
good yield.
One of the problems associated with this route is the absence
of a chromophore in the first step, making the in-process
analytics for reaction completion difficult to establish, for both
the activated ester intermediate and the amide product. The
introduction of the expensive 3,3-difluoropyrrolidine in the first
step is also not ideal in terms of bond disconnection.3 Finally,
the starting keto-L-proline (1) is relatively expensive ($2400/
kg), considering that it is being made from the oxidation of the
readily available and cheap Boc-trans-4-hydroxy-L-proline
methyl ester (5) ($325/kg). A synthetic route to PF-734200
It became apparent that the selection of the leaving group
was extremely important for this nucleophilic substitution to
proceed. The difference in reactivity between the mesylate
(6a), benzenesulfonate (6b), and nosylate (6c) (Table 1,
entries 1−3) was significant and prompted us to concentrate on
the more reactive nosylate. The lack of reactivity of 6a could be
one of the reasons why the nucleophilic displacement of an
activated 4-hydroxyproline derivative with a primary or
secondary amine is unprecedented, as the choice of the
nosylate is not as common as that of the mesylate, which
performed poorly. The use of a nosylate on pilot-plant scale is
precedented in the alkylation of (R)-glycidyl nosylate11 and
gave us confidence that this method of activation would be
acceptable for commercial use. A rapid solvent screen showed
that acetonitrile was superior to THF and DMF for this
reaction (entries 3−5). The addition of a crown ether had an
unfavorable effect on the reaction (entry 6). The nature of the
carbonate base was investigated, and it was found that the rate
Received: May 3, 2011
Published: January 28, 2012
© 2012 American Chemical Society
409
dx.doi.org/10.1021/op200309z | Org. Process Res. Dev. 2012, 16, 409−414

Organic Process Research & Development
Article
Scheme 2. Proposed synthetic strategy to PF-734200 (4)
Table 1. Screen of various SN2 reaction conditions
entry
substrate
base
solvent
T
% yield
1
2
3
4
5
6
7
8
9
6a
6b
6c
6c
6c
6c
6c
6c
6c
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
MeCN
MeCN
MeCN
THF
reflux
reflux
reflux
reflux
70 °C
70 °C
reflux
reflux
reflux
12
<10
92
39
DMF
62
K2CO3 18-crown-6
Na2CO3
DMF
31
MeCN
MeCN
MeCN
>95
65
Li2CO3
Cs2CO3
nd
*
1
Crude reaction mixture yield measured by H NMR.
of the reaction followed roughly the following order: Na, K ≫
Li, Cs (entries 3, 8−9). Results also suggested that powdered
solids were more efficient than granular types. Sodium
carbonate powder was eventually selected as the optimal base
due to inducing high reaction conversion.
Scheme 3. Preparation of 1,4-trans-isomer (10)
In order to confirm stereochemical integrity of the product,
the potential formation of the undesired diastereoisomer 10
through chiral inversion at C4 was ruled out. An authentic
sample of 10 was prepared using N-Boc-cis-4-hydroxy-L-proline
methyl ester (9) as the starting material (Scheme 3), and no
trace of the C4 inversion product 10 was detectable in our
experiments.
To further understand the stability of the nosylate ester 6c
towards potential undesired side reaction, a control experiment
was carried out in which 6c was treated with 1 equiv of 4-
nitrobenzenesulfonic acid, 1.5 equiv of triethylamine and
sodium carbonate under reflux for 18 h (Scheme 4). We
were pleased to observe that our starting material was very
stable under the reaction conditions employed for the
nucleophilic displacement, producing no detectable C-4
1
inversion product 11, and less than 2% (by H NMR) of the
elimination product (12). The 4,5-dehydro elimination product
13 was not detectable, in line with previous reports where the
3,4-elimination product is formed predominantly.12−14 Inter-
estingly, when the 1,4-cis-nosylate ester 11 is submitted to the
same control reaction, up to 20% of dehydro-elimination
product 12 was observed. In addition, the rate of the
nucleophilic substitution reaction to produce the undesired
1,4-trans stereoisomer 10 (Scheme 3) was much greater than
410
dx.doi.org/10.1021/op200309z | Org. Process Res. Dev. 2012, 16, 409−414

Organic Process Research & Development
Article
robust conditions. It was initially believed that the presence of
excess 1-(2-pyrimidyl)piperazine could inhibit the product
crystallization. In order to facilitate the isolation of the desired
product, 1-(2-pyrimidyl)piperazine was acetylated by acetic
anhydride to form the acetamide derivative, but this failed to
improve crystallization from the crude mixture. Alternatively,
reaction with succinic anhydride produced amide 14 bearing a
free carboxylate group. Initial attempts to keep the salt in
solution with potassium carbonate produced a difficult to stir
gel-type suspension. Instead, filtration of residue from the
nucleophilic substitution and use of triethylamine as the base
for the derivatization, leading to a more soluble ammonium
carboxylate, proved to be the solution as 14 could be removed
in the aqueous phase as part of the extraction. Furthermore,
ester 7 was found to be very soluble in slightly acidic water;
thus, the crude product was treated with aqueous HCl. This
brought 7 into the aqueous phase, allowing for the removal of
all the nonpolar impurities in the organic layer, including
unreacted starting material and the 3,4-elimination product
(12) produced in the reaction. Control of the pH between 4
and 6 in this operation was important to avoid deprotection of
the carbamate. Extraction of 7 in EtOAc was achieved by
neutralization of the acidic aqueous phase with NaOH.
Following the workup procedure described above, isolation of
7 with satisfactory purity (>95%) was achieved in 80% yield,
which also made it possible to telescope the crude solution
directly to the next step without crystallization following a
solvent swap to THF/H2O.
The initial strategy for the completion of the synthesis was to
hydrolyze the ester and isolate the resulting carboxylic acid as
the penultimate intermediate. Although the saponification
proceeds smoothly using lithium hydroxide, the isolation of
either the carboxylic acid or its salt proved very challenging due
to the high solubility in water. The strategy was modified to
telescope the crude carboxylate to the amidation step and rely
on a robust crystallization of either the API or the protected
carbamate 3. The first approach investigated involved
neutralization of the crude carboxylate with triethylamine
hydrochloride, followed by azeotropic distillation of the residual
water in toluene. Generation of the mixed anhydride with
pivaloyl chloride followed by addition of 3,3-difluoropyrrolidine
generated the amide (3) in moderate yield (45%). In order to
take advantage of the high water solubility of the carboxylate
intermediate, we turned our attention to a water-soluble amide
coupling agent, namely 1-ethyl-3-(3′-dimethylaminopropyl)-
Scheme 4. Differences in the stabilities of nosylate 6c and 11
the corresponding reaction starting with 6c.15 This is consistent
with a ring pucker stereochemistry as depicted in Scheme 4,
where 1,4-trans-substituted proline favors an exo conformation,
while 1,4-cis-substitution favors an endo form.16
Having demonstrated the key nucleophilic substitution for
diastereoselective introduction of the piperidine moiety, we
optimized the first step of the synthesis. In an initial screen,
reaction of methyl ester 5 proceeded smoothly but no further
than 80% with p-nitrosulfonyl chloride in THF using Et3N as
the base. The use of DMAP as a catalyst, use of excess base, and
selection of acetonitrile as the solvent were all effective in
bringing the reaction to completion in less than one hour at
room temperature. The final reaction condition selected
utilized a slight excess of NsCl and two equivalents of the
base in MeCN leading to full conversion of the starting material
(Scheme 5). Excess NsCl was completely removed during
aqueous workup, making it possible to telescope the crude
reaction mixture directly to the next step. The crude nosylate
(6c) was reacted under the preferred conditions described in
Table 1 to afford 7 in high yield. Although this intermediate
was found to be crystalline, attempts to develop a crystallization
method from the crude reaction mixture failed to produce
Scheme 5. Preparation and purification of ester 7
411
dx.doi.org/10.1021/op200309z | Org. Process Res. Dev. 2012, 16, 409−414

Organic Process Research & Development
Article
Scheme 6. Final 3 step approach
carbodiimide (EDAC or EDC).17 Following the hydrolysis of
the ester 7 with lithium hydroxide, the resulting carboxylate
solution was directly subjected to the amidation step (Scheme
6). The reaction mixture was simply cooled down to 10 °C
followed by addition of the hydrochloride salt of 3,3-
difluoropyrrolidine. The salt was neutralized by the one
equivalent excess of base already present in the system to
liberate the nucleophilic amine. Addition of HOBt18,19 and
EDAC led to formation of the desired amide (3). Attempts to
perform the reaction without HOBt or using a substitute (such
as DMAP), failed to generate good conversion. Using this
simple and straightforward strategy, the desired penultimate
intermediate 3 was produced in good yield (70−75%) and
isolated by crystallization. The final deprotection of the
carbamate was performed using the known methodology
from the initial route, to provide the final product PF-734200
(4). Analysis of the product generated with the new synthetic
route showed no new impurities, and the overall purity profile
was similar by HPLC to the one generated using the enabling
synthetic route shown in Scheme 1. The products of either
route proved to be equivalent through other analytical
methods.
Table 2. Cost comparison of two synthetic routes to
intermediate 3
While the new procedure requires one additional synthetic
step compared to the synthesis described in Scheme 1, it
provides a significant economical advantage upon a cost
analysis (Table 2). Since it was believed that both processes
could be reduced to a similar number of units of operation and
used common and inexpensive reagents, it was concluded that
the price of the building blocks would be the best predictor of
the cost efficiency of each synthesis. In the previous synthesis,
the ketone starting material (1) cost 7 times the quotes
obtained for the new starting material, alcohol 5, at the same
scale. The choice of either 1 or 5 as starting material between
the two syntheses represents the largest cost difference between
the two routes illustrated. In the previous synthesis, the most
expensive building block, the 3,3-difluoropyrrolidine, is used in
the second step of a telescoped process while in the new
synthesis, it is used in the fourth and last step. Since 3,3-
difluoropyrrolidine has a small molecular weight and the
telescoped last two steps of the new synthesis suffer from the
lowest reported yield, the difference between the two routes in
contributions from this fragment is minimal. However, further
optimization of the last step of the new route would represent
the biggest opportunity for further cost improvements. Finally,
contributions from the 1-(2-pyrimidyl)piperazine building
block are minimal in either synthesis since this starting material
has a low cost and fairly small molecular weight. Obviously, it is
used more efficiently in the first synthesis as it is introduced in
the last step. This analysis helped us conclude that the new
route identified offered economical advantage based on input
costs.
*
Kilograms of SM for each kilogram of 3 produced.
CONCLUSIONS
■
A new synthetic route to PF-734200 that takes advantage of the
inexpensive and readily available Boc-trans-4-hydroxy-L-proline
methyl ester (5) building block was demonstrated. The
synthesis highlights the importance of the selection of the
leaving group for a key nucleophilic substitution. In this case,
the key p-nitrobenzenesulfonyl ester (6c) provided unique
properties for the introduction of a piperidine fragment while
preserving diastereoselectivity. This new route offers multiple
advantages, namely the introduction of a chromophore early in
the synthesis, providing access to conventional HPLC analysis
for in-process control, an economic advantage due to the
selection of the starting material, and introduction of the most
412
dx.doi.org/10.1021/op200309z | Org. Process Res. Dev. 2012, 16, 409−414

Organic Process Research & Development
Article
expensive building block (3,3-difluoropyrrolidine) in the last
step. Preservation of the penultimate intermediate allowed for
the opportunity to prepare API of comparable purity. The new
route would benefit from further process development and
optimization before introduction into a manufacturing setting.
crude product was purified by chromatography using DCM
(3% MeOH) as the eluant. The product containing fractions
were pooled and concentrated in vacuo to give an oil that
crystallized upon trituration with pentane to afford 10 as an off-
white solid (0.32 g, 71% yield). Mp 135−136 °C. [α]D = −11.7
(T = 24 °C, c = 1, CHCl3) 1H NMR (400 MHz; CDCl3) δ 8.32
(m, 2H), 6.50 (m, 1H), 4.5−4.3 (m, 1H), 3.95−3.75 (m, 5H),
3.74 (s, 3H), 3.35−3.2 (m, 1H), 3.05−2.9 (m, 1H), 2.6−2.4
(m, 4H), 2.25−2.05 (m, 2H), 1.48, 1.42 (2s, 9H). 13C NMR
(100 MHz; CDCl3) δ 173.36, 173.12; 161.53; 157.73; 154.24,
153.50; 110.05; 80.20; 62.59, 61.74; 58.55, 58.11; 52.31, 52.13;
51.53, 51.48; 50.25, 49.91; 43.43; 34.31, 33.43; 28.42, 28.25. IR
(neat): 2975br, 1745m, 1695vs, 1592m, 1542m, 1398m(br),
1363m, 1186m(br), 1150m(br), 984m, 898m, 796m, 571w.
HRMS (ES, N2) Calcd for C19H29N5O4: 392.22923, found:
392.22935.
EXPERIMENTAL SECTION
■
NMR data were collected on a Bruker Ultrashield 400 plus
instrument. HRMS data were collected on a Thermo-Fisher
LTQ-Orbitrap spectrometer. IR monitoring was performed
using a Thermo Nicolet Nexus-470 FTIR instrument. Optical
rotations were measured using a Jasco P-1020 polarimeter.
( 2 S , 4 R ) - 1 - t e r t - B u t y l - 2 - m e t h y l - 4 - ( 4 -
nitrobenzenesulfonyloxy)pyrrolidine-1,2-dicarboxylate
(6c). To a solution of Et3N (1.14 mL, 8.18 mmol) and N-Boc-
trans-4-hydroxy-L-proline methyl ester (5) (1.00 g, 3.95 mmol)
in MeCN (15 mL) at 0 °C was added p-nitrobenzenesulfonyl
chloride (1.31 g, 5.93 mmol). The suspension was allowed to
warm up to rt and stirred for 18 h. Water (5 mL) was added,
and the reaction mixture was stirred at rt for 30 min and
concentrated in vacuo. EtOAc (20 mL) was added, and the
organic solution was washed sequentially with of saturated
aqueous Na2CO3 (10 mL), 1 N HCl (10 mL), saturated
aqueous Na2CO3 (10 mL), and brine (10 mL). The organic
phase was dried (Na2SO4), filtered, and concentrated in vacuo
to provide an oil that slowly crystallized upon standing. The
crystals were triturated in Et2O (10 mL) at rt, filtered, and dried
in a vacuum oven at 55 °C to afford 6c as a light-yellow solid
(1.14 g, 67% yield). Mp 95 °C. [α]D = −30.0 (T = 24 °C, c = 1,
(2S,4S)-1-tert-Butyl-2-methyl-4-(4-(pyrimidin-2-yl)-
piperazin-1-yl)pyrrolidine-1,2-dicarboxylate (7). A Boc-
trans-4-hydroxy-L-proline methyl ester (5) (10.0 g, 0.041 mol)
and Et3N (11.4 mL, 0.082 mol) solution in 150 mL of
acetonitrile was cooled to 5 °C, and 4-nitrobenzenesulfonyl
chloride (12.8 g, 0.057 mol, 1.4 equiv) was added portionwise
at 5 °C ( 5 °C) throughout the addition. The reaction mixture
was stirred at 5 °C for 1 h, warmed to 20 °C over 1 h, and
stirred at 20 °C for 1 h. The reaction mixture was quenched
with water (10 mL), stirred at rt 1 h, and concentrated to a low
stirrable volume. EtOAc (120 mL) and water (60 mL) were
added and stirred vigorously at rt for 30 min. The phases were
separated, and the organic layer was washed with 10% aq NaCl
(60 mL) and water (60 mL). The EtOAc was distilled and
replaced with MeCN (40 mL) that was distilled. To the crude
product was added MeCN (80 mL), powdered Na2CO3 (8.64
g, 0.082 mol), and 1-(2-pyrimidinyl)piperazine 8.7 g, 0.053
mol), and the suspension was stirred at reflux until reaction
completion (typically 18 h). The mixture was cooled to rt,
filtered through a pad of Celite, and washed with MeCN (40
mL). To the filtrate was added Et3N (2.84 mL (203 mmol))
and succinic anhydride (2.04 g of 204 mmol), and the mixture
was stirred at rt for 1 h, concentrated to a low stirrable volume,
and diluted with EtOAc (100 mL). The solution was washed
with 0.5 N NaOH (60 mL), and 10% aq NaCl (60 mL). The
organic layer was extracted with 0.5 N HCl (120 mL, 1.5 mol
equiv) and washed with water (20 mL). The aqueous extracts
were combined, cooled to 10 °C, and treated with 2 N NaOH
(32 mL, 1.07 equiv vs HCl). EtOAc (100 mL) was added, and
the mixture was stirred vigorously for 30 min. The layers were
separated, and the organic phase was concentrated to a low
stirrable volume. The residue was dissolved in 4:1 THF/H2O
(100 mL) and telescoped directly to the next step. Compound
7 can alternatively be isolated as a crystalline solid as described
for compound 10 (reflux 18 h instead of 6 h). Mp 109 °C. [α]D
1
CHCl3). H NMR (400 MHz; CDCl3) δ 8.35 (d, J = 8 Hz,
2H), 8.05 (d, J = 8 Hz, 2H), 5.12 (s br, 1H), 4.4−4.25 (m, 1H),
3.66 (s, 3H), 3.65−3.55 (m, 2H), 2.6−2.35 (m, 1H), 2.25−2.05
(m, 1H), 1.36, 1.32 (2s, 9H). 13C NMR (100 MHz; CDCl3) δ
172.54, 172.36, 153.72, 153.10, 150.91, 142.11, 129.10, 124.71,
81.01, 79.96, 57.24, 56.96, 52.35, 51.80, 37.26, 36.07, 28.25,
28.13. IR (neat): 2972br, 1743s, 1693vs, 1531s, 1411s, 1363s,
1347s, 1287m, 1159s(br), 1041m, 910m, 747br, 616s, 578m,
551m. HRMS (ES sodiated, N2) Calcd for C17H22N2O9SNa:
453.09382, found: 453.09377. Characterization data match
those of previous report.9
( 2 S , 4 S ) - 1 - t e r t - B u t y l - m e t h y l - 4 - ( 4 -
nitrobenzenesulfonyloxy)pyrrolidine-1,2-dicarboxylate
(11). The same procedure as for the preparation of 6c was used
starting with Boc-cis-Hyp-OMe (9, 1.00 g). This provided 11 as
an off-white solid (1.24 g, 71% yield). Mp 119 °C. [α]D = −26.2
(T = 24 °C, c = 1, CHCl3) 1H NMR (400 MHz; CDCl3) δ 8.41
(d, J = 8 Hz, 2H), 8.11 (d, J = 8 Hz, 2H), 5.22 (s br, 1H),
4.55−4.35 (m, 1H), 3.71 (s, 3H), 3.69−3.60 (m, 2H), 2.60−
2.42 (m, 2H), 1.43, 1.40 (2s, 9H). 13C NMR (100 MHz;
CDCl3) δ 171.73, 171.46; 153.60, 153.24; 150.80; 142.37;
129.03; 124.56; 80.82; 79.58; 57.21, 56.94; 52.34; 51.68; 37.08,
36.12; 28.29, 28.17. IR (neat): 2980br, 1752s, 1681vs, 1531s,
1410s, 1363br, 1188s(br), 1155s(br), 927m, 910m, 758m,
745m, 660m, 554 m. HRMS (ES sodiated, N2) Calcd for
C17H22N2O9SNa: 453.09382, found: 453.09407.
1
= −42.2 (T = 24 °C, c = 0.8, CHCl3). H NMR (400 MHz;
CDCl3) δ 8.32 (m, 2H), 6.50 (m, 1H), 4.35−4.2 (m, 1H),
3.95−3.75 (m, 5H), 3.74 (s, 3H), 3.35−3.25 (m, 1H), 2.9−2.75
(m, 1H), 2.75−2.4 (m, 5H), 2.0−1.85 (m, 1H), 1.48, 1.42 (2s,
9H). 13C NMR (100 MHz; CDCl3) δ 173.05, 172.82; 161.51;
157.73; 154.16, 153.49; 110.04; 80.30; 63.29, 62.64; 58.38,
57.88; 52.20, 52.03; 51.73, 51.59; 52.24, 49.97; 43.38; 34.92,
33.79; 28.41, 28.25. IR (neat): 2951br, 1734m, 1694vs, 1592s,
1527m, 1399m(br), 1366m, 1203m, 1164s(br), 983m, 792m,
770m, 562w. HRMS (ES, N2) Calcd for C19H29N5O4:
392.22923, found: 392.22939.
(2S,4R)-1-tert-Butyl-2-methyl-4-(4-(pyrimidin-2-yl)-
piperazin-1-yl)pyrrolidine-1,2-dicarboxylate (10). In a 20
mL pressure tube was added 11 (0.50 g, 1.16 mmol), 1-(2-
pyrimidyl)piperazine) (0.25 g, 1.52 mmol), and Na2CO3 (0.25
g, 2.36 mmol) in MeCN (2.5 mL). The tube was sealed, and
the reaction mixture was stirred at reflux for 6 h, cooled to rt,
and filtered over a frit funnel. The solid residue was washed
with MeCN and the filtrate was concentrated in vacuo. The
413
dx.doi.org/10.1021/op200309z | Org. Process Res. Dev. 2012, 16, 409−414

Organic Process Research & Development
Article
(2S,4S)-tert-Butyl-2-(3,3-difluoropyrrolidine-1-carbon-
yl)-4-(4-(pyrimidin-2-yl)piperazin-1-yl)pyrrolidine-1-car-
boxylate (3). The THF/H2O solution of 7 described above
was added LiOH (1.96 g, 0.0816 mol). The reaction mixture
was stirred at rt for 4 h, cooled to 0 °C and 3,3-
difluoropyrrolidine hydrochloride (6.17 g, 0.043 mol) was
added followed by portionwise addition of HOBT·H2O (5.81 g,
0.043) (CAUTION: Vacuum-drying of HOBT while heating
can generate a violent explosion) and EDAC (8.24 g, 0.043
mol) over 1 h. The reaction mixture was allowed to warm-up to
20 °C, stirred for 18 h, and was concentrated to a low stirrable
volume. Water (100 mL) was added, and the mixture was
extracted with EtOAc (2 × 100 mL). The combined organic
extrects were washed with brine (100 mL) and concentrated,
and the residue was triturated with 95:5 heptane/EtOAc (75
mL). The solid was filtered, washed with heptane, and dried in
a vacuum oven at 45 °C to afford 3 as an off-white solid (10.9 g,
57% yield over two steps). Mp 194 °C (decomp). [α]D = −22.9
AUTHOR INFORMATION
Corresponding Author
danny.lafrance@pfizer.com
■
ACKNOWLEDGMENTS
■
The authors thank the following Pfizer colleagues: Kyle
Leeman, Victor Soliman, and Silke Wunderwald for analytical
support and helpful discussions.
REFERENCES
■
(1) Ammirati, M. J.; Andrews, K. M.; Boyer, D. D.; Brodeur, A. M.;
Danley, D. E.; Doran, S. D.; Hulin, B.; Liu, S.; McPherson, R. K.;
Orena, S. J.; Parker, J. C.; Polivkova, J.; Qiu, X.; Soglia, C. B.;
Treadway, J. L.; VanVolkenburg, M. A.; Wilder, D. C.; Piotrowski, D.
W. Bioorg. Med. Chem. Lett. 2009, 19, 1991.
(2) Hulin, B.; Piotrowski, D. W. US Patent 2005/0256310 A1, 2005;
24 pp.
(3) Wei, L.; Makowski, T. M.; Rutherford, J. L. J. Fluorine Chem.
2011, in press.
(4) Seo, J.; Martasek, P.; Roman, L. J.; Silverman, R. B. Bioorg. Med.
Chem. 2007, 15, 1928.
1
(T = 24 °C, c = 1, CHCl3). H NMR (400 MHz; CDCl3) δ
8.32 (dd, J = 8, 4 Hz, 2H), 6.51 (t, J = 8 Hz, 1H), 4.5−4.25 (m,
1H), 4.0−3.5 (m, 9H), 3.4−3.25 (m, 1H), 2.95−2.75 (s br,
1H), 2.7−2.2 (m, 7H), 2.0−1.75 (m, 1H), 1.46, 1.41 (2s, 9H).
13C NMR (100 MHz; CDCl3) δ 170.98, 170.84; 161.51;
157.73; 154.21, 153.36; 127.35 (t, 1 J C−F = 248 Hz), 125.97
(t, 1 J C−F = 246 Hz), 110.05; 80.35, 80.20; 63.38, 62.73;
57.57, 56.75; 52.85 (m, 1C); 51.66, 51.49; 50.07; 43.98, 43.89;
43.35; 34.56 (t, 2 J C−F = 23 Hz), 32.57 (t, 2 J C−F = 25 Hz),
34.21; 28.45, 28.27. 19F NMR (377 MHz, CDCl3) δ −101.4
(m, 2F). IR (neat): 2948w, 2826w, 1698vs, 1646vs, 1587m,
1544m, 1484m, 1448s, 366s, 1267m, 1165m, 1113s, 984w,
805m, 777m, 556 m. HRMS (ES, N2) Calcd for
C22H32F2N6O3: 467.25767, found: 467.25832.
(5) Diaba, F.; Ricou, E.; Bonjoch, J. Tetrahedron: Asymmetry 2006,
17, 1437.
(6) Fisher, A.; Mann, A.; Verma, V.; Thomas, N.; Mishra, R. K.;
Johnson, R. L. J. Med. Chem. 2006, 49, 307.
(7) Marusawa, H.; Setoi, H.; Sawada, A.; Kuroda, A.; Seki, J.;
Motoyama, Y.; Tanaka, H. Bioorg. Med. Chem. 2002, 10, 1399.
(8) Tanaka, K.-I.; Sawanishi, H. Tetrahedron: Asymmetry 1998, 9, 71.
(9) Lerner, C.; Siegrist, R.; Schweizer, E.; Diederich, F.; Gramlich, V.;
Jakob-Roetne, R.; Zurcher, G.; Borroni, E. Helv. Chim. Acta 2003, 86,
1045.
(10) Pickersgill, I. F.; Rapoport, H. J. Org. Chem. 2000, 65, 4048.
(11) Barnett, C. J.; Huff, B.; Kobierski, M. E.; Letourneau, M.;
Wilson, T. M. J. Org. Chem. 2004, 69, 7653.
(12) Kim, K.-Y.; Kim, B. C.; Lee, H. B.; Shin, H. J. Org. Chem. 2008,
73, 8106.
(3,3-Difluoropyrrolidin-1-yl)-(2S,4S)-4-(4-(pyrimidin-2-
yl)piperazin-1-yl)pyrrolidin-2-yl)methanone (4). A 10 mL
water solution of 3 (5.0 g, 0.011 mol) at rt was added conc.
HCl (10 mL) over 4 h. The mixture was cooled to 5 °C and
50% NaOH (∼ 6.5 mL) was slowly added, keeping the
temperature <15 °C, until the pH reached 10−11. The mixture
was allowed to warm-up to rt and was extracted with EtOAc
(60 and 30 mL). The combined organic extracts were dried
with MgSO4 (2.5 g) and stirred at rt for 30 min. The solids
were filtered, washed with EtOAc (5 mL), and concentrated to
a volume of 10 mL, at which point the product began to
crystallize. The solids were stirred at rt for 30 min, and hexane
(10 mL) was added over a period of 1 h. The solids were
filtered, washed with hexane/EtOAc (1:1) (10 mL), and dried
in vacuum oven at 45 °C for 12 h to afford 4 (3.25 g, 83%
yield). Mp 149 °C (decomp). [α]D = −31.1 (T = 24 °C, c = 1,
CHCl3). Specific rotation of product 4 prepared using the
initial route: [α]D = −31.5 (T = 24 °C, c = 1, CHCl3). 1H NMR
(400 MHz; CDCl3) δ 8.30 (d, J = 4 Hz, 2H), 6.48 (t, J = 4 Hz,
1H), 3.95−3.6 (m, 9H), 3.25−2.85 (m, 4H), 2.6−2.25 (m,
7H), 1.75−1.6 (m, 1H). 13C NMR (100 MHz; CDCl3) δ
172.28; 161.55; 157.70; 127.22 (t, 1J C−F = 248 Hz), 126.22
(t, 1J C−F = 246 Hz), 109.95; 66.54; 58.87; 57.99; 52.71 (t, 2 J
C−F = 32 Hz); 52.00; 50.41; 43.03; 34.46, 34.37, 34.25; 19F
NMR (377 MHz, CDCl3) δ −102.1 (m, 2F). IR (neat): 2951w,
2864w, 2799w, 2759w, 1630s, 1585vs, 1547m, 1449m, 1172m,
1254m, 1129m, 982w, 923m, 796m, 638w. HRMS (ES, N2)
Calcd for C17H24F2N6O: 367.20524, found: 367.20592.
(13) Schumacher, K. K.; Jiang, J.; Joullie, M. M. Tetrahedron:
Asymmetry 1998, 9, 47.
(14) Shangguan, N.; Joullie, M. Tetrahedron Lett. 2009, 50, 6748.
(15) The reaction reaches approximately 95% conversion in 6 h with
nosylate 11 versus 18 h for 6c.
(16) Thomas, K. M.; Naduthambi, D.; Tririya, G.; Zondlo, N. J. Org.
Lett. 2005, 7, 2397.
(17) Sheehan, J. C.; Cruickshank, P. A.; Boshart, G. L. J. Org. Chem.
1961, 26, 2525.
(18) HOBT hydrate was used. CAUTION: Vacuum-drying of
HOBT while heating can generate a violent explosion! See:Urben, P.
G. In Bretherick’s Handbook of Reactive Chemical Hazards, 6th ed.;
Butterworth-Heinemann: Oxford, 1999; p 746.
(19) Carpino, L. A.; El-Faham, A.; Albericio, F. J. Org. Chem. 1995,
60, 3561.
414
dx.doi.org/10.1021/op200309z | Org. Process Res. Dev. 2012, 16, 409−414