An efficient synthesis of 3-substituted N-glycoside indoles useful as sodium-dependent glucose transporter inhibitors

Article abstract of DOI:10.1021/op3001355

A practical synthesis of two N-glycoside indoles 1 and 2, identified as highly potent sodium-dependent glucose transporter (SGLT) inhibitors is described. Highlights of the synthetic process include a selective and quantitative Vilsmeier acylation and a h

Full text of DOI:10.1021/op3001355

Article
pubs.acs.org/OPRD
An Efficient Synthesis of 3‑Substituted N‑Glycoside Indoles Useful as
Sodium-Dependent Glucose Transporter Inhibitors
Xun Li,* Kenneth M. Wells,* Shawn Branum, Sandra Damon, Scott Youells, Derek A. Beauchamp,
David Palmer, Stephen Stefanick, Ronald K. Russell, and William Murray
Janssen Research and Development, L.L.C., Welsh and McKean Roads, Spring House, Pennsylvania 19477, United States
S
* Supporting Information
ABSTRACT: A practical synthesis of two N-glycoside indoles 1 and 2, identified as highly potent sodium-dependent glucose
transporter (SGLT) inhibitors is described. Highlights of the synthetic process include a selective and quantitative Vilsmeier
acylation and a high-yielding Grignard coupling reaction. The chemistry developed has been applied to prepare two separate
SGLT inhibitors 1 and 2 for clinical evaluation without recourse to chromatography.
INTRODUCTION
■
One and a half million new cases of diabetes mellitus are
diagnosed every year worldwide, and it is projected by the year
2050 that the number of individuals with diagnosed diabetes
mellitus will increase by 165% in the United States alone.1
Presently, there is a number of antidiabetic agents used in
clinical treatment such as biguanide compounds, sulfonylurea
compounds, insulin resistance-improving agents, and α-
glucosidase inhibitors. These antidiabetic agents, while useful,
have various serious undesired side effects.2
Figure 2. Examples of N-glycosides as SGLT2 inhibitors.
The crucial need for new diabetes mellitus treatment options
with fewer or no unfavorable side effects have enticed
pharmaceutical companies to develop novel treatment strat-
egies. For example, by inhibiting the sodium-dependent glucose
transporter 2 (SGLT2) present at the proximal convoluted
tubule of the kidney, the reabsorption of glucose in the kidney
is inhibited. The glucose is therefore excreted into the urine
thus decreasing the blood glucose level.3 Several potent SGLT2
inhibitors have entered clinical development (Figure 1).3,4 Two
RESULTS AND DISCUSSION
■
The initial discovery synthesis of 1 described by the Mitsubishi
Tanabe Pharma Corporation is shown in Scheme 1.5a Upon
review of the discovery route, we identified scale-up issues that
required synthesis improvements. For example, the Vilsmeier
formylation conditions in step 1 gave 18−20% of the 2,3-
diformyl 4a impurity, which required removal by chromatog-
raphy since recrystallization of the crude mixture could not
afford pure 3-formyl indole 4. Another issue was the poor
isolated yield of 4−bromocyclopropylbenzene (8) (Scheme 3).
Further complicating the scale-up was the chromatography
burden, in which three column chromatographies were required
to purify the intermediates 4, 6, and active pharmaceutical
ingredient (API) 1 (see Table 1). Finally the crystallization of
the 1 was not well developed, and the procedure relied on a
difficult, slow conversion of an EtOH solvate to the desired
hemihydrate, which was difficult to reproduce on large scale.
Vilsmeier Process Development. During initial observa-
tions of the Vilsmeier reaction on indole 3 with 3 equiv. of
phosphorous(III) oxychloride and 6 equiv of DMF at 90 °C, a
Figure 1. Examples of C- and O-glycosides as SGLT2 inhibitors that
have entered clinical development.
1
N-glycosides inhibitors of SGLT2, 15a and 2,5b have been
identified as candidates for further development (Figure 2).
Herein, we describe a safe, scaleable synthetic route to 1, which
has delivered several hundred grams to support early preclinical
studies. The chemistry developed for 1 was applied successfully
to the synthesis of 2 and was used to prepare material for
preclinical study.
major impurity 4a was observed which was confirmed by H
NMR and LC/MS (Scheme 2). Several reports of Vilsmeier
reactions of these indoles in the literature all used DMF as a
solvent and reaction temperatures of 90 °C which led in our
Received: May 23, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/op3001355 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 1. Synthesis of SGLT2 inhibitor 1
however, 46% and 19% of unreacted 3 remained after 12 h.
Optimization of POCl3/DMF to 2.5 equiv/1.4 equiv resulted in
complete conversion of 3 to the desired 3-formyl compound 4
as the sole product, at both 30 and 40 °C within 12 h. The
reaction mixture was safely quenched by adding the reaction
mixture over 75 min to warm aqueous sodium acetate. We were
able to safely quench the reaction without any delayed
exotherm that is typical when this reagent is used.7 It is
absolutely imperative to perform this quench between 35 °C
and 40 °C. The addition rate of the crude reaction mixture
must be adjusted accordingly to remain within this temperature
range. This quench is extremely exothermic, and the
recommended temperature range from data derived from the
reaction calorimeter (RC-1) study of the quench guarantees
instantaneous reaction of POCl3 with water.8 The instanta-
neous quenching prevents an accumulation of unreacted POCl3
and the possibility of a runaway quench reaction. These
conditions afforded 4 as a solid in high purity and quantitative
yield and thus eliminated any need for chromatography.
Cyclopropylaryl Bromide and Subsequent Grignard
Formation Process Development. Since aryl bromide 8 was
not commercially available in bulk, the synthesis from
cyclopropylbenzene was investigated. The discovery route
followed the procedure outlined by Himmelsbach et al.,
which gave four products (Scheme 3/entry 1 of Table 2) and
only a 38% yield of the desired 8 after careful fractional
distillation.9
Table 1. Summary of the discovery and scale-up reaction
yields for the preparation of SGLT2 inhibitor 1
route
step
a,b
Mitsubishi Tanabe discovery
scale-up of 1
c
1
2
3
4
column chromatography, 79% of 4, 18% of 4a
(not isolated)
100% of 4
(not isolated)
85% of 6
column chromatography 89% of 6
column chromatography, 96% of 1
98% of 1
a
Isolated yields from our laboratory using Mitsubishi Tanabe’s
b
conditions. Three chromatographic purifications were required
(steps 1, 3, and 4). 4a was isolated by this work.
c
Scheme 2. Vilsmeier impurity formation
hands to production of substantial amounts of impurity 4a.6a,b
In order to reduce or eliminate 4a, the ratio of POCl3/DMF
used, as well as the effects of time and temperature, was studied
(Figure 3).
Similar conditions were reported by Lavey et al., where
chromatography was used to purify the crude reaction mixture
and likewise gave a 21% isolated yield of 8.10 Zeolite-catalysed
bromination conditions of 7 also have been reported, but the
reported yields (30%) of 8 offered no attractive alternative.11
It was found that diformylation rate was very dependent on
the temperature of the reaction. At temperatures below 40 °C,
the diformylation product 4a was not detectable even after
extended aging. For example, treatment of indole 3 with lower
amounts of both POCl3/DMF produced no byproduct 4a;
B
dx.doi.org/10.1021/op3001355 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Figure 3. Vilsmeier optimization conditions.
Scheme 3. Original bromination conditions
Table 2. Reaction conditions and the results for selective bromination of compound 7 to 8
a
entry
7 (g)
temp (°C)
0−10
solvent
time (h)
8
crude mixture of 8/8a/8b/7/unknown (HPLC, area %)
isolated yield of 8 (g/ %)
b
c
1
5.0
1.0
HOAc
38/26/10/15/11
3.2/38.5
ND
d
e
2
DMF
ND
f
g
3
5.0
20
−62
−76
−76
−76
−76
CH2Cl2
CHCl3
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
6
36/11/37/12/4
ND
h
4
5.0
3.5
3
69/9/14/8/0
82/6/8/2/2
73/7/10/6/4
81/6/8/5/0
87/3/5/5/0
5.7/69.0
6.1/73.0
30.1/74.5
195.0/81.0
h
5
5.0
h
6
25.0
150.0
150.0
4
h
7
5
h
8
5
210.0/90.0
a
b
c
d
e
After distillation the product 8 purity was >94%. 1.0 equiv of Br2, KOAc. Mitsubishi Tanabe’s synthetic process. 1.0 equiv NBS. A complicated
f
g
h
mixture. 1.12 equiv of Br2, NaY Zeolite. Could not be separated by distillation. 1.05 equiv of Br2.
However, a report by Lvina and Gembitskii of a high-yielding
bromination of 7 using bromine at low temperatures in CHCl3
led us to further explore these conditions.12 Although these
conditions (Table 2 entry 4) showed a much improved yield
(69%) of 8 after distillation, a total of 23% of the competitive
byproducts 8a and 8b were formed. However, since the
reaction mixture in CHCl3 (mp −63 °C) became a solid mass
below −65 °C, CH2Cl2 (mp −97 °C) was chosen as an
alternative solvent. Another benefit of the solvent switch is that
the ICH guidance for residual CHCl3 is 10 times lower than for
CH2Cl2. The results (Table 2 entries 5−8) indicated that the
reaction in CH2Cl2 at −76 °C resulted in a much higher ratio of
the para versus the ortho position as well as an improved yield
of 8 (90%) after distillation.
next step. This freshly prepared Grignard reagent was added to
a THF solution of 3-formylindole 4 at 0 °C, and after quench
and workup, a quantitative recovery of the crude alcohol 5 was
obtained, which was used directly in the next reduction step
without further purification.
Benzyl Alcohol Reduction and Protecting Group
Removal. The hydroxy group of intermediate 5 was reduced
(Et3SiH/BF3OEt2) following the procedure outlined by the
discovery team and gave good reproducibility of yield and
quality of product. The product 6 was obtained in good yield
(85%) and excellent purity after recrystallization from EtOH.
Finally, the acetoxy protecting groups were removed with a
catalytic amount of NaOMe in MeOH/THF, and the resulting
crude 1 was recrystallized from aqueous EtOH to give a 97.5
LCAP of a pure EtOH/H2O solvate of 1. This solvate was
slurried in warm H2O and the solid-state form converted over
20 h to the desired hemihydrate form as a crystalline solid,
which was confirmed by pXRD. All stereogenic centers on the
N-substituted sugar moiety were confirmed by single crystal X-
ray diffraction.13
The formation of the Grignard reagent 9 in THF with
magnesium was straightforward with the aid of 1,2-dibromo-
ethane as an initiator (Scheme 4) and was used directly in the
Scheme 4. Grignard formation
CONCLUSIONS
■
A safe, reproducible, and nonchromatographic synthetic
procedure has been developed for the preparation of
multihundred grams of the SGLT2 inhibitor 1 in 73% overall
C
dx.doi.org/10.1021/op3001355 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
yield. The Vilsmeier reaction conditions were defined for the
formylation of N-glycoside indoles which produced quantitative
isolated yields of 4 with excellent chemical purities (>99
LCAP). Further to our method was the use of a safe and
convenient method for the quenching of the excess Vilsmeier
reagents using warm aqueous sodium acetate that warrants
further investigation by others. Our investigations also found an
improved preparation of 4-bromocyclopropylbenzene (8),
vastly improving the yield by lowering the reaction temperature
using a compatible solvent. In addition, the final API 1 was
prepared as a crystalline solid after recrystallization with 98.5%
HPLC purity. The absolute stereochemistry of 1 was confirmed
by X-ray diffraction study on a single crystal of 1.
mL, 1.26 mol) in CH2Cl2 (0.75 L) over 2 h, and the mixture
was stirred at −76 °C for 3 h. The mixture was warmed to −20
°C; then saturated aqueous NaHCO3 solution (2.0 L) was
added, and the mixture was stirred for 10 min. After phase
separation, the organic phase was washed with brine (1.0 L)
and concentrated under reduced pressure to afford 264.9 g of a
crude mixture which after purification by short path distillation
1
afforded 219 g (90% isolated yield) with 96 LCAP. H NMR
(DMSO-d6, 300 MHz) 7.48 (d, J = 7.9 Hz, 2 H), 6.97 (d, J =
8.0 Hz, 2H), 1.85 (m, 1H), 0.94 (m, 2H), 0.62 (m, 2H). 13C
NMR (DMSO-d6, 75.47 MHz) 143.1, 130.9 (2 C), 127.4 (2
C), 118.0, 14.6, 9.42 (2 C). LC−MS m/z MH+ = 197, 199
(MH+).
The preparation of SGLT2 inhibitor 1 was used as a
prototype for optimizing the chemistry for a second SGLT2
inhibitor 2, and the experimental details and scheme can be
found in the Supporting Information. The original synthesis of
2 by discovery relied on a Friedel−Craft acylation followed by
reduction of the resulting ketone and required several
chromatographies to purify the intermediates.14
This chemistry was previously used to prepare several
hundred grams of API for early toxicology studies but suffered
from an exothermic quench of the excess AlCl3 (2.3 equiv used
in reaction) after the Friedel−Crafts reaction. The acid-
sensitive acetoxy protecting groups were prone to cleavage
under these conditions. In contrast, the efficient and high-
yielding Grignard-based synthesis of 2 demonstrated important
process advantages over the Friedel−Crafts route. No
deprotection occurred during any of the synthetic steps, all
chromatography was eliminated, and the overall yield was
improved from 30% to 53%. Furthermore, the required
Grignard reagent, 4-ethylbenzene magnesium bromide, was
commercially available.
Therefore, taking all our experience and learning from the
scale-up campaign of 1 just described, the Vilsmeier procedure
developed for the previous compound 1 was successfully
applied to the synthesis of the required 3-formylcarboxaldehyde
intermediate. Similarly, the Grignard reaction of the 3-
formylcarboxaldehyde intermediate with the commercially
available Grignard reagent yielded the intermediate alcohol in
95% yield. The reduction of the alcohol intermediate, followed
by the acetate deprotection were straightforward reactions and
occurred without incident in a 60% yield of 2 over the two
steps. Finally, crystallization of 2 with EtOH gave the EtOH
solvate of 2 in 73% yield.
Preparation of (2R,3R,4S,5R,6R)-2-(Acetoxymethyl)-6-
(4-fluoro-3-formyl-1H-indol-1-yl)tetrahydro-2H-pyran-
3,4,5-triyl Triacetate (4). To a stirred solution of indole 3
(225.0 g, 0.459 mol) and DMF (50.2 mL, 0.643 mol) in DCE
(1.5 L) at 25 °C was added phosphoryl chloride (107.8 mL,
1.15 mol) over 75 min. After the addition was completed, the
reaction mixture was stirred for 30 min, and then it was slowly
warmed to 40 °C over 30 min. The reaction mixture was
agitated at 40 °C for 12 h. The reaction mixture was slowly
(CAUTION: this quench is very exothermic!) poured into a
rapidly stirred warm (40 °C) 3 M aqueous NaOAc (3.0 L)
solution over 45 min and was stirred for an additional 15 min
after the addition was complete. The resulting mixture was
extracted with CH2Cl2 (4.0 L), and the phases were separated.
The aqueous phase was back-extracted with CH2Cl2 (1.0 L).
The combined organic phases were washed with 0.05 M HCl
(2.0 L) and de-ionized water (2.0 L) and then were dried
(MgSO4). The solvents were removed by rotaevaporation, and
the resulting solid was dried at 40 °C to afford 230.0 g (100%
isolated yield; LCAP 99%) of pure 3-formylindole 4 as a
slightly yellow-brown solid. 1H NMR (DMSO-d6, 300 MHz) δ
10.1 (s, 1H), 8.53 (s, 1H), 7.66 (d, J = 7.3 Hz, 1H), 7.38 (m,
1H), 7.10 (dd, J = 6.7, 6.9 Hz, 1H), 6.38 (d, J = 7.5 Hz, 1H),
5.68 (dd, J = 6.5, 6.6 Hz, 1H), 5.56 (t, J = 7.1 Hz, 1H), 5.32 (t, J
= 7.2 Hz, 1H) 4.41 − 4.28 (m, 1H), 4.24 − 4.06 (m, 2H), 2.05
(s, 3H), 2.0 (s, 3H), 1.98 (s, 3H), 1.64 (s, 3H). 13C NMR
(DMSO-d6, 75.47 MHz) δ 183.8, 169.9, 169.5, 169.3, 168.4,
155.8, 139.2, 135.7, 124.8, 117.7, 113.1, 108.3, 107.9, 81.9, 73.5,
72.1, 70.3, 67.6, 61.9, 20.4, 20.3, 20.1, 19.6. LC−MS m/z MH+
= 494 (MH+), 516 [M + Na]+, 1009 [2M + Na]+. [α]25
−0.099 (c = 0.316, CHCl3).
=
D
Preparation of (2R,3R,4S,5R,6R)-2-(Acetoxymethyl)-6-
(3-((4-cyclopropylphenyl)(hydroxy)methyl)-4-fluoro-1H-
indol-1-yl)tetrahydro-2H-pyran-3,4,5-triyl Triacetate (5).
To a stirred suspension of magnesium (17.8 g, 0.725 mol) in
anhydrous THF (420 mL) was added a solution of 1,2-
dibromoethane (0.53 mL, 6.1 mmol) and 8 (142.1 g, 0.685
mol) in THF (110 mL) dropwise over 20 min at 20 °C.
(CAUTION: this experimental is designed to avoid a runaway
reaction, but a delay in initiation may cause the solvent to boil.)
When the temperature was 26 °C, the reaction was placed in a
cold-water bath, and the temperature continued to increase to
32 °C; ice was added to the water bath to maintain the
temperature between 40 and 48 °C with vigorous stirring. After
the reaction temperature declined to 16 °C, the water bath was
removed, and the mixture was stirred at 20 °C for an additional
10−20 min to give the Grignard reagent 9 as a gray-yellow
solution, which was immediately used in the next step.
EXPERIMENTAL SECTION
■
Intermediates were analyzed on an Agilent 1100 HPLC system
using the following conditions: 4.6 mm × 50 mm Luna C18
column, 5.0 μm, 254 nm, gradient elution (0.05% TFA/
CH3CN 20/80 water/0.05% TFA to 40% over 1 min, to 90%
over 3 min and hold for 6 min). Retention times are as follows:
4, 5.41 min; 5, 4.38 min; 6, 5.10 min; and 1, 3.51 min. 1H and
13C NMR data were collected on either a Bruker AC-300 at 300
MHz for proton and 75 MHz for carbon, a Bruker AVANCE
600 at 600 MHz for proton and 125 MHz for carbon, or a
Varian UNITYinova at 500 MHz for proton and 125 MHz for
carbon. Compound 3 was commercially outsourced and was
synthesized according the procedure outlined by Nomura et
al.14
Preparation of 1-Bromo-4-cyclopropylbenzene (8). To
a stirred solution of cyclopropylbenzene (7) (150.0 g, 1.23
mol) in CH2Cl2 (1.5 L) at −76 °C was added bromine (65.4
To a stirred 0 °C solution of 3-formylindole 4 (230 g, 0.457
mol) in anhydrous THF (4.2 L) was added the above freshly
D
dx.doi.org/10.1021/op3001355 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
prepared Grignard reagent 9 in THF (530 mL) over 20 min,
while the internal temperature was maintained between 0 and 8
°C by adjusting the rate of addition. The resultant mixture was
stirred at 0 °C for 30 min. The reaction was quenched with
saturated aqueous NH4Cl (5.4 L) solution, and then extracted
with EtOAc (4.0 L) twice. The combined organic phase was
washed with brine and dried (MgSO4). After filtration, the
filtrate was concentrated at 66 °C under house vacuum to
afford 334.6 g (contained 82 LCAP of the desired 5, 0.7 LCAP
of starting 4, and 2.2 LCAP of bromide 8 as a yellowish solid
was heated to 76 °C and de-ionized H2O (1800 mL) was added
in a small stream over 20 min that resulted in a slightly
yellowish clear solution, which was gradually cooled to 40 °C
over 20 min with stirring while seeded. The resulting slightly
white-yellowish suspension was stirred at 20 °C for 20 h, the
solids were collected by filtration, washed with cold (0 °C)
EtOH/H2O (1:4), and dried by air-suction for 6 h with a gentle
stream of N2. There was obtained 198.5 g (97.5% isolated yield
based on free base form; 98.8 LCAP) of 1 EtOH/H2O solvate
as an off-white crystalline solid. A slurry of the EtOH/H2O
solvate 1 (198.5 g, 0.399 mol) in de-ionized H2O (3.2 L,) was
warmed to 76 °C, and then the slurry was gradually cooled to
20 °C over 30 min. The white suspension was stirred at 20 °C
for 20 min and then at 10 °C for 1 h. The solid was collected by
filtration, washed with de-ionized H2O (100 mL × 2), dried in
an oven at 50 °C for 20 h and further at 60 °C for 3 h to afford
177.4 g, (99.8% isolated yield, 98.6 LCAP) of 1 hemihydrate as
an off-white crystalline solid, of which the 1H NMR showed no
EtOH residue and the powder X-ray diffraction (pXRD)
confirmed that it was a crystalline solid. TGA indicated it
contained 2.3% of water. Mp = 108−111 °C. 1H NMR
(DMSO-d6, 300 MHz) δ 7.36 (d, J = 8.2 Hz, 1 H), 7.22 (s, 1
H), 7.14 (d, J = 8.1, 2 H), 7.10−7.0 (m, 1 H), 6.96 (d, J = 8.1
Hz, 2 H), 6.73 (dd, J = 7.5, 7.7 Hz, 1 H), 5.38 (d, J = 7.7 Hz, 1
H), 5.21 (d, J = 6.9 Hz, 1 H), 5.18 (d, J = 6.8 Hz, 1 H), 5.10 (d,
J = 6.9 Hz, 1 H), 4.54 (t, J = 6.9, 1.8 Hz, 1 H), 4.04 (s, 2 H),
3.75−3.60 (m, 2 H), 3.52−3.30 (m, 3 H), 3.20−3.17 (m, 1 H),
1.84 (m, 1 H), 0.89 (m, 2 H), 0.61 (m, 2 H). 13C NMR
(DMSO-d6, 75.47 MHz): δ 156.2, 140.8, 139.4, 138.2, 128.2 (2
C), 125.2 (2 C), 124.4, 121.8, 115.9, 112.8, 107.4, 104.2, 84.8,
79.3, 77.4, 71.7, 69.8, 60.8, 31.3, 14.6, 8.92 (2 C). LC−MS m/z
1
which was used in next step without further purification. H
NMR (DMSO-d6, 300 MHz) δ 7.49 (dd, J = 7.9, 1.4 Hz, 1H),
7.41 (dd, J = 8.0, 1.0 Hz, 1H), 7.30 (d, J = 8.0 Hz, 1H), 7.18
(dd, J = 7.3, 1.6 Hz, 2H), 7,15 (m, 1H), 6.97 (dd, J = 7.4, 1.6
Hz, 2H), 6.78 (dd, J = 2.7, 1.6 Hz, 1H), 6.18 (dd, J = 7.8, 2.6
Hz, 1H), 5.94 (m, 1H), 5.65 (dd, J = 7.9, 1.5 Hz, 1H), 5.59−
5.45 (m, 1H), 5.24 (dd, J = 7.9, 8.3 Hz, 1H), 4.36 − 4.24 (m,
1H), 4.20−4.04 (m, 2H), 2.05 (s, 3H), 2.01 (s, 3H), 1.98 (s,
3H), 1.84 (m, 1H), 1.66 (s, 3H), 0.92 (m, 2H), 0.61 (m, 2H).
13C NMR (DMSO-d6, 75.47 MHz): δ 170.1, 170.0, 169.9,
169.3, 156.1, 140.9 139.0, 137.9, 128.0 (2 C), 125.2 (2 C),
124.2, 122.6, 116.3, 114.6, 107.4, 105.2, 81.5, 76.8, 73.0, 72.6,
70.1, 68.2, 62.0, 20.6, 20.4, 20.2, 19.8, 14.8, 8.96 (2 C) . LC−
MS m/z MH+ = 612 (MH+), 634 [M + Na]+.
Preparation of (2R,3R,4S,5R,6R)-2-(acetoxymethyl)-6-
(3-(4-cyclopropylbenzyl)-4-fluoro-1H-indol-1-yl)-
tetrahydro-2H-pyran-3,4,5-triyl Triacetate (6). To a
stirred solution of 5 (82%, 334.6 g, 0.449 mol) in DCE (1.14
L) and MeCN (2.28 L) at 0 °C was added Et3SiH (108.6 mL,
0.671 mol) followed by the addition of boron trifluoride
etherate (68.8 mL, 0.539 mol) dropwise over 10 min. After
completion of the reaction, saturated aqueous NaHCO3
solution (4.2 L) was added to the mixture, which was extracted
twice with EtOAc (4 L). The solvents were removed by
rotoevaporation to afford 315.0 g (contained 84% LCAP of the
desired product 6). The resulting crude 6 (315.0 g) was slurried
with EtOH (2.1 L) at 76 °C and cooled to 20 °C over 30 min
and stirred for 1 h. The solid was collected by filtration, washed
with cold (0 °C) EtOH (200 mL), and dried in a vacuum oven
at 60 °C. There was obtained 228.6 g (85% isolated yield, 98.4
LCAP) of pure 6 as an off-white crystalline solid. Mp 168−169
°C. 1H NMR (DMSO-d6, 300 MHz) δ 7.47 (d, J = 7.2 Hz, 1H),
7.22 (s, 1H), 7.20−7.10 (m, 1H), 7.06 (d, J = 8.1, 2H), 6.95 (d,
J = 8.1 Hz, 2H), 6.78 (dd, J = 7.3, 7.0 Hz, 1H), 6.16 (d, J = 7.1
Hz, 1H), 5.61−5.48 (m, 2H), 5.21 (t, J = 7.3, 7.1 Hz, 1H), 4.34
− 4.25 (m, 1H), 4.18−4.04 (m, 2H), 4.0 (s, 2H), 2.04 (s, 3H),
1.97 (s, 3H), 1.95 (s, 3H), 1.84 (m, 1H), 1.61 (s, 3H), 0.89 (m,
2H), 0.61 (m, 2H). 13C NMR (DMSO-d6, 75.47 MHz): δ
169.9, 169.5, 169.3, 168.3, 156.2, 140.9, 139.0, 137.9, 128.0 (2
C), 125.2 (2 C), 124.2, 122.7, 116.1, 114.1, 107.2, 105.0, 81.7,
73.0, 72.5, 69.8, 68.0, 62.0, 31.2, 20.4, 20.3, 20.2, 19.7, 14.6, 8.93
(2 C). HRMS: m/z = 596.2261 [M − 1]+. [α]25D = −0.008 (c =
0.306, CHCl3).
MH+ = 428 (MH+), 450 [M + Na]+, 877 [2M + Na]+. [α]25
=
D
−0.026 (c = 0.302, CH3OH). Anal. Calc’d for
C24H26NFO5·0.54 H2O: C, 65.93; H, 6.24; N, 3.20; F, 4.35,
H2O, 2.23. Found: C, 65.66; H, 6.16; N, 3.05; F, 4.18, H2O,
2.26.
ASSOCIATED CONTENT
* Supporting Information
■
S
Scheme and experimental data for the synthesis of 2. POCl3
RC-1 reaction quench into sodium acetate study in the RC-1.
X-ray crystallographic data for compound 1. This material is
available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
■
*(X.L.) Telephone: +1-215-628-7829. Fax: +1-215-540-4611.
E-mail: xli6@its.jnj.com; (K.M.W.) Telephone: +1-215-628-
7814. Fax: +1-215-540-4611. E-mail: kwells1@its.jnj.com.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
Preparation of (2R,3R,4S,5S,6R)-2-(3-(4-Cyclopropyl-
benzyl)-4-fluoro-1H-indol-1-yl)-6-(hydroxymethyl)-
tetrahydro-2H-pyran-3,4,5-triol (1). To a solution of
compound 6 (250 g, 0.413 mol) in MeOH (1.2 L) and THF
(2.4 L) was added sodium methoxide (2.5 mL, 0.012 mol), and
the reaction was stirred at 20 °C for 3 h. The solvent was
removed by rotaevaporation, and EtOAc (8.0 L) was added.
The resulting solution was washed with brine (800 mL × 2)
and dried (MgSO4). The filtrate was concentrated by
rotaevaporation, and EtOH (900 mL) was added. The solution
■
We are grateful to Brigitte Segmuller and Richard Dunphy for
analytical assistance. We are also grateful to Christopher Teleha
for manuscript review.
REFERENCES
■
(1) (a) Narayan, K. M.; Boyle, J. P; Thompson, T. J.; Sorensen, S.
W.; Williamson, D. F. JAMA, J. Am. Med. Assoc. 2003, 290, 1884.
(b) Boyle, J. P.; Honeycutt, A. A.; Venkat Narayan, K. M.; Hoerger, T.
J.; Geiss, L. S.; Chen, H.; Thompson, T. J. Diabetes Care 2001, 24,
E
dx.doi.org/10.1021/op3001355 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
1936−1940. (c) International Diabetes Federation. Diabetes Atlas, 3rd
ed.; International Diabetes Federation: Brussels, Belgium, 2006.
(2) Bell, D. S. H. South. Med. J. 2002, 95, 24.
(3) Oku, A.; Ueta, K.; Arakawa, K.; Ishihara, T.; Nawano, M.;
Kuronuma, Y.; Matsumoto, M.; Saito, A.; Tsujihara, K.; Anai, M.;
Asano, T.; Kanai, Y.; Endou, H. Diabetes 1999, 48, 1794.
(4) Meng, W.; Ellsworth, B. A.; Nirschl, A. A.; McCann, P. J.; Patel,
M.; Girotra, R. N.; Wu, G; Sher, P. M.; Morrison, E. P.; Biller, S. A.;
Zahler, R.; Deshpande, P. P.; Pullockaran, A.; Hagan, D. L.; Morgan,
N.; Taylor, J. R.; Obermeier, M. T.; Humphreys, W. G.; Khanna, A.;
Discenza, L.; Robertson, J. G.; Wang, A.; Han, S.; Wetterau, J. R.;
Janovitz, E. B.; Flint, O. P.; Whaley, J. M.; Washburn, W. N. J. Med.
Chem. 2008, 51, 1145.
(5) (a) Nomura, S.; Sakamaki, S. WO/2008/013322, 2008.
(b) Nomura, S.; Kawanishi, E; Ueta, K. WO/2005/012326, 2005.
(6) (a) Preobrazhenskaya, M.; Tolkachev, V.; Geling, O.;
Kostyuchenko, P. Zh. Org. Khim. 1974, 10, 1764. (b) Humenik, M;
Dzurilla, M.; Kutschy, P.; Solcaniova, E.; Kovacik, V.; Bekesova, S.
Collect. Czech. Chem. Commun. 2004, 69, 1657.
(7) Achmatowicz, M.; Thiel, O.; Colyer, J; Hu, J.; Elipe, M;
Tomaskevitch, J.; Tedrow, J.; Larsen, R. Org. Process Res. Dev. 2010,
14, 1490.
(8) RC-1 Study of the POCl3 quench is included in the Supporting
Information.
(9) Himmelsbach, F.; Ruden Von, T.; Dahmann, G.; Metz, T. WO/
1996/07657, 1996.
(10) Lavey, B. J.; Kozlowski, J. A.; Shankar, B. B.; Spitler, J. M.; Zhou,
G.; Yang, D.; Shu, Y.; Wong, M. C. K.; Wong, S.; Shih, N.; Wu, J.;
McCombie, S. W.; Rizvi, R.; Wolin, R. L.; Lunn, C. A. Bioorg. Med.
Chem. Lett. 2007, 17, 3760.
(11) Esakkidurai, T.; Kumarraja, M.; Pitchumani, K. Catal. Lett. 2004,
92, 169.
(12) Lvina, R. Y.; Gembitskii, P. A. Zh. Obshch. Khim. 1961, 31, 3480.
(13) X-ray crystallographic data is included in the Supporting
Information.
(14) Internal communications.
(15) Nomura, S.; Yamamoto, Y. WO/2006/080577, 2006.
F
dx.doi.org/10.1021/op3001355 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX