Practical synthesis of potent sphingosine-1-phosphate lyase inhibitors THI and LX2931

Article abstract of DOI:10.1016/j.tet.2013.03.071

A practical and scalable synthesis of in vivo sphingosine-1-phosphate lyase inhibitor LX2931 (1) is described. The synthetic route features an improved Büchi cyclocondensation of 2-ethoxyacrylonitrile (3) with either 1-amino-1-deoxy-d-fructose acetate (4a) or d-(+)-glucosamine hydrochloride (4b) to produce 1-(4-((1R,2S,3R)-1,2,3,4-tetrahydroxybutyl)-1H-imidazol-2-yl)ethanone (2, THI), followed by oximation of THI and acid-promoted oxime isomerization to give LX2931 (1).

Full text of DOI:10.1016/j.tet.2013.03.071

Tetrahedron 69 (2013) 4041e4046
Contents lists available at SciVerse ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Practical synthesis of potent sphingosine-1-phosphate lyase
inhibitors THI and LX2931
,
a
a
b
a
Haiming Zhang a , Jie Yan , Mark S. Bednarz , Gonzalo Hernandez , Yuelie Lu ,
*
Lawrence F. Courtney a, Jason Chen c, Weifeng Hu c, Renmao Liu c, Xiaogen Yang c,
Wenxue Wu a
,
*
a Chemical Development, Lexicon Pharmaceuticals, Inc., 350 Carter Road, Princeton, NJ 08540, United States
b Analytical Chemistry, Lexicon Pharmaceuticals, Inc., 350 Carter Road, Princeton, NJ 08540, United States
c Process Research and Development, WuXi AppTec Co., Ltd, 288 Fute Zhong Road, Shanghai 200131, China
a r t i c l e i n f o
a b s t r a c t
Article history:
A practical and scalable synthesis of in vivo sphingosine-1-phosphate lyase inhibitor LX2931 (1) is de-
Received 11 January 2013
Received in revised form 14 March 2013
Accepted 19 March 2013
€
scribed. The synthetic route features an improved Buchi cyclocondensation of 2-ethoxyacrylonitrile (3)
with either 1-amino-1-deoxy-
D
-fructose acetate (4a) or
D
-(þ)-glucosamine hydrochloride (4b) to pro-
duce 1-(4-((1R,2S,3R)-1,2,3,4-tetrahydroxybutyl)-1H-imidazol-2-yl)ethanone (2, THI), followed by oxi-
mation of THI and acid-promoted oxime isomerization to give LX2931 (1).
Ó 2013 Elsevier Ltd. All rights reserved.
Available online 22 March 2013
Keywords:
Cyclocondensation
Oxime isomerization
Glucosamine
Fructosamine
Sphingosine-1-phosphate
1. Introduction
responsive reduction of circulating lymphocytes in addition to
general tolerability of this compound. Preliminary results from the
Sphingosine-1-phosphate (S1P) plays an important role in
mammalian immune systems by regulating the number of circu-
lating lymphoid cells.1 Intracellular S1P concentrations are atten-
uated by S1P lyase (S1PL), which catalyzes an irreversible retro
Aldol fragmentation of S1P to 2-hexadecenal and ethanolamine
phosphate (Scheme 1). Inhibition of S1PL therefore increases en-
dogenous S1P concentrations in lymphoid tissues, leads to lower
levels of circulating peripheral lymphocytes and mitigates the
physiologic outcome of aberrant immune function.2 Using a high-
throughput gene knockout technology platform,3 Lexicon con-
ducted a detailed phenotypic analysis of S1PL knockout mice and
identified S1PL as a potential drug target for the treatment of au-
toimmune disorders. Internal medicinal chemistry efforts resulted
in the discovery of a potent and selective in vivo S1PL inhibitor
LX2931 (1, Scheme 2). Preclinical studies showed that LX2931 de-
creased circulating lymphocyte counts in multiple species and
produced therapeutic effect in rodent models of rheumatoid ar-
thritis (RA). Clinical Phase 1 results also demonstrated dose-
recently completed Phase 2a clinical study showed initial signals of
efficacy against RA symptoms as well as a favorable safety profile.
Scheme 1. Retro Aldol fragmentation of S1P.
Herein, we wish to report a practical synthesis of LX2931 (1)
that is suitable for multi-kilogram scale production. LX2931 (1) is
an analog of 1-(4-((1R,2S,3R)-1,2,3,4-tetrahydroxybutyl)-1H-imi-
dazol-2-yl)ethanone (THI, 2), a known S1PL inhibitor.4 There were
two major challenges in devising a practical synthesis of LX2931
(1): identification of an economical process for preparing THI, and
control of oxime configuration in the target compound 1. We
* Corresponding authors. Tel.: þ1 6094666059; fax: þ1 6094663562; e-mail
address: wwu@lexpharma.com (W. Wu).
0040-4020/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tet.2013.03.071

4042
H. Zhang et al. / Tetrahedron 69 (2013) 4041e4046
€
It is noteworthy that besides the Buchi synthesis (3/2,
Scheme 3),5 THI was also synthesized by Sweeny9 using Debuse
Radziszewski imidazole synthesis10 and by Pyne11 employing
a combination of Stille coupling12 and Sharpless asymmetric
dihydroxylation.13 However, these syntheses suffered low overall
yields and required stoichiometric amounts of metal reagents and
multiple chromatographic purifications. After carefully reviewing
the literature syntheses and briefly pursuing unfruitful preliminary
in-house investigation to construct THI (2) using alternative
€
methods, we decided to focus our effort on improving the Buchi THI
synthesis.
2.1. Synthesis of 2-ethoxyacrylonitrile (3)
Scheme 2. Retrosynthetic analysis of LX2931 (1).
€
Buchi utilized Cuvigny’s protocol for the preparation of 2-
ethoxyacrylonitrile (3).6 Thus, bromination of ethyl vinyl ether
followed by treatment with cuprous cyanide and elimination of HBr
afforded 2-ethoxyacrylonitrile (3) in 49% overall yield (Scheme 4A).
In order to render the synthesis more amenable to large-scale
manufacture, we modified it to eliminate the use of hazardous
bromine and of a stoichiometric amount of cuprous cyanide. Thus,
treatment of commercially available and inexpensive 2-
bromoacetaldehyde diethylacetal (7) with trimethylsilyl cyanide
in the presence of a catalytic amount (0.5 mol %) of stannous
chloride (SnCl2), followed by diethylamine-promoted HBr elimi-
nation from the intermediate 3-bromo-2-ethoxypropanenitrile (8),
produced the desired 2-ethoxyacrylonitrile (3) in 67% overall yield
(Scheme 4B).
envisioned an overall synthetic strategy involving an improved
€
Buchi synthesis5 of THI (2) employing either fructosamine
acetate (4a) or glucosamine hydrochloride (4b) as well as an acid-
promoted isomerization of the E/Z mixture of THI oxime
(Scheme 2).
2. Results and discussion
The medicinal chemistry synthesis4a,b employed the condensa-
tion reaction of 2-ethoxyacrylonitrile (3)6 and 1-amino-1-deoxy-
D-
7
5
€
fructose acetate (4a) as originally described by Buchi. THI (2) was
then treated with a bulky O-protected hydroxylamine, O-tritylhy-
droxylamine,8 to give exclusively the E-THI O-trityloxime (6). The
trityl protecting group was then removed to give the final com-
pound 1 (Scheme 3). While this synthesis worked well on small
scale and provided rapid access to multi-gram quantities of 1 for
early investigation of SAR and toxicological evaluation, further
scale-up was problematic. Specifically, the reported yields of 2-
ethoxyacrylonitrile (49%)6 and THI (46% for 4a and 19% for 4b)5
were low and the condensation reaction suffered from a long re-
action time, tedious work-up, and inconsistent yields during scale-
up. Furthermore, the use of expensive O-tritylhydroxylamine,
though providing direct access to the E-oxime geometry, is not
atom-economical as most of the mass, the trityl group, was dis-
carded as chemical waste. Therefore, a much more practical and
efficient synthesis was desired in order to support the development
program of LX2931.
Scheme 4. Cuvigny’s and improved synthesis of 3.
2.2. Formation of imidate 5
With a robust synthesis of 2-ethoxyacrylonitrile (3) available,
€
we set out to optimize the Buchi THI synthesis. The formation of
5
€
imidate 5 from 2-ethoxyacrylonitrile (3) developed by Buchi
employed 0.1 equiv sodium methoxide and the reaction took
about 8 h at 20 ꢀC to reach a maximum of 94% conversion. As ex-
pected, the reaction time could be readily shortened by increasing
the amount of sodium methoxide used (Fig. 1). Thus, the imidate
formation reached ca. 94% conversion in 2 h using 0.6 equiv sodium
methoxide.
We also observed that higher reaction temperature resulted in
a faster reaction but in lower conversion. For example, a reaction
carried out at 60 ꢀC using 0.6 equiv sodium methoxide reached
highest conversion (72%) in minutes, as compared to a reaction
performed at 0 ꢀC, which gave 96% conversion in ca. 12 h. In-
terestingly, the conversion decreased to 70% when the 0 ꢀC reaction
was subsequently heated to 60 ꢀC. On the other hand, when the
60 ꢀC reaction was cooled to 20 ꢀC, the conversion increased to 96%
(Fig. 2). These results clearly indicated an equilibrium between 2-
ethoxyacrylonitrile (3) and imidate 5 under these reaction condi-
tions. Taking overall efficiency into account, the optimal conditions
Scheme 3. Medicinal chemistry synthesis of 1.

H. Zhang et al. / Tetrahedron 69 (2013) 4041e4046
4043
mixture at 20 ꢀC for 6 h, followed by adding another portion of
0.5 equiv sodium methoxide and stirring for an additional 16 h
completes the condensation. This protocol avoids the repeated
€
cooling/heating cycles described in the original Buchi THI synthe-
sis.5 The subsequent hydrolysis of intermediate ethyl vinyl ether 9
to THI (2) was achieved by treating the reaction mixture with acetic
acid and water at 60 ꢀC. The improved procedure reduced the re-
action time from 4 days to only 24 h and increased the overall yield
from 46% to 82% (Scheme 5).
100
90
80
70
60
50
40
30
20
10
0
0:00
1:30
3:00
4:30
6:00
7:30
Reaction Time (hours)
Fig. 1. NaOMe promoted imidate (5) formation at 20 ꢀC as monitored by 1H NMR (
0.1 equiv; 0.2 equiv; 0.3 equiv; 0.4 equiv; A 0.6 equiv).
Scheme 5. Improved synthesis of THI (2) using 4a.
This THI synthesis, though successful for multi-kilogram pro-
duction, has a drawback in that 1-amino-1-deoxy-D-fructose ace-
tate (4a) is not readily accessible commercially.7 To address this
issue, we decided to improve the synthesis using the more readily
100
90
80
70
60
50
40
30
20
10
0
available and inexpensive starting material
D
-(þ)-glucosamine hy-
drochloride (4b).16 Applying the knowledge gained during de-
velopment of the above process using 4a, we achieved the buffered
conditions by adding sodium acetate to the mixture of 5 and 4b.
Thus, imidate 5 (prepared by treating 2-ethoxyacrylonitrile (3) with
0.65 equiv sodium methoxide) was stirred with 1.5 equiv
D-
(þ)-glucosamine hydrochloride (4b) and 1.5 equiv sodium acetate
for 8 h, followed by treatment with an additional 0.65 equiv sodium
methoxide. After acetic acid promoted hydrolysis, THI (2) was iso-
lated in 66% yield (Table 1, entry 1). Without the addition of sodium
acetate to buffer the reaction, only 12% yield of THI (2) was obtained
(Table 1, entry 2). After further optimization, we were able to im-
prove the yield of THI (2) to 78% (Table 1, entry 4) by reacting
imidate 5 (generated by treating 2-ethoxyacrylonitrile (3) with
0:00
3:00
6:00
9:00
12:00
15:00
18:00
Reaction Time (hours)
0.3 equiv sodium methoxide) with 1.2 equiv
D
-(þ)-glucosamine
Fig. 2. Imidate (5) formation with 0.6 equiv NaOMe at different temperature ( 0 ꢀC;
20 ꢀC; 40 ꢀC; A 60 ꢀC).
hydrochloride (4b) and 1.2 equiv sodium acetate, followed by an-
other portion of 0.65 equiv sodium methoxide and then acetic acid
treatment (Scheme 6).
adopted for the imidate formation ultimately involved 0.6 equiv
sodium methoxide at 20 ꢀC, resulting in 94% conversion in only 2 h
(Eq. 1).
Table 1
Optimization of THI (2) formation using 4ba
Entry 3/5 (5 h)
NaOMe (equiv)
Buffer system (8 h)
5/9 (5 h)
NaOMe (equiv)
Yieldb (%)
4b (equiv) NaOAc (equiv)
ð1Þ
1
2
3
4
5
6
0.65
0.65
0.3
0.3
0.45
0.65
1.5
1.5
1.5
1.2
1.2
1.2
1.5
0
1.5
1.2
1.2
1.2
0.65
0.65
0.65
0.65
0.65
0.65
66
12
68
78(75)c
55
2.3. Synthesis of THI (2)
53
a
b
c
The reactions were carried out using 5.35 g (55.0 mmol) of 3.
Isolated yield.
The yield in parenthesis was obtained using 200 g (2.06 mol) of 3.
It is known in the literature that imidazole formation from an
imidate and an
ered system,14 possibly due to minimization of the competitive
dimerization of
-aminoketones.15 We envisaged that buffering the
reaction system by mixing the imidate solution (from Eq. 1) with
1 equiv 1-amino-1-deoxy-
-fructose acetate (4a)7 in methanol
a-aminoketone proceeds more efficiently in a buff-
a
A working mechanism is proposed for these two trans-
formations (Scheme 7). Under buffered conditions, 1-amino-1-
deoxy- -fructose acetate (4a) and
-(þ)-glucosamine hydrochlo-
ride (4b) liberate partially free 1-amino-1-deoxy-
-fructose (4a0)
D
should improve the process. Indeed, this simple change in condi-
tions gave a markedly improved reaction. Thus, stirring the reaction
D
D
D

4044
H. Zhang et al. / Tetrahedron 69 (2013) 4041e4046
Scheme 6. Synthesis of THI (2) using 4a.
Fig. 3. A view of 1$2H2O from the crystal structure. Anisotropic atomic displacement
ellipsoids for the non-hydrogen atoms are shown at the 50% probability level.
Scheme 8. Synthesis of 1 by oxime formation and acid-promoted isomerization.
of 2-ethoxyacrylonitrile (3), which obviates the use of hazardous
bromine and stoichiometric amount of cuprous cyanide; (2) an
Scheme 7. Proposed mechanism of THI (2) formation.
€
improved Buchi THI (2) synthesis using either 1-amino-1-deoxy-
D
-
fructose acetate (4a) or
D
-(þ)-glucosamine hydrochloride (4b); (3)
and
D
-(þ)-glucosamine (4b0), which then react with imidate 5 to
an oximation reaction using hydroxylamine hydrochloride to gen-
erate a mixture of 3:1 E/Z oximes; and (4) an acid-promoted
isomerization of E/Z oxime isomers to enrich the desired E-oxime.
LX2931 (1) was isolated as an anhydrous crystalline solid in ca. 60%
generate intermediates 10a and 10b. Cyclization of 10a and 10b
produces ethyl vinyl ethers 11a and 11b, which dehydrate to form
imidazole 9. Imidazole 9 then undergoes hydrolysis under acidic
conditions to afford THI (2).
yield from 1-amino-1-deoxy-
D-fructose acetate (4a) or in ca. 50%
from
D
-(þ)-glucosamine hydrochloride (4b). By employing this
process, we were able to manufacture tens of kilograms of the drug
substance with few process difficulties, enabling a rapid de-
velopment program.
2.4. Synthesis of LX2931 (1) by oxime isomerization
With the synthesis of THI under control, we investigated direct
oximation of THI using hydroxylamine hydrochloride. It was ob-
served that the reaction of THI and hydroxylamine hydrochloride in
the presence of sodium acetate in methanol at 65 ꢀC produced
a mixture of ca. 3:1 E/Z oxime isomers in a quantitative yield. Initial
attempts to effect dynamic crystallization of salts of the E-isomer
were not successful. However, we learned from these experiments
that the more stable E-isomer was readily enriched under strongly
acidic conditions.17 Thus, treatment of the above 3:1 E/Z oxime
mixture with anhydrous HCl at 65 ꢀC resulted in an improved E/Z
ratio of ca. 13:1. The reaction mixture was then neutralized and the
E-oxime (1) was isolated by crystallization from water in 81% yield
as a dihydrate. The isolated product contained only ca. 3% of the Z-
oxime. The absolute geometry of the E-oxime dihydrate was firmly
established by X-ray crystallography (Fig. 3).18 The anhydrous E-
oxime can be readily obtained by reslurrying the dihydrate in
ethanol at 75 ꢀC and then cooling to 0 ꢀC before filtration (95% yield,
Scheme 8).
4. Experimental section
4.1. General
All reactions were conducted under a nitrogen atmosphere
unless otherwise noted. All reagents, starting materials, and sol-
vents were purchased from commercial sources and used as re-
ceived. 1H and 13C NMR spectra were collected on Bruker DRX400
or DPX400 spectrometers. Chemical shifts are reported in parts per
million (ppm) in
d unit and coupling constants (J values) are given
in Hertz (Hz). Mass spectra were obtained on Waters ZQ or ZMD
LCMS systems equipped with an auto-sampler, an ELSD detector,
a UV detector measuring absorbance at 220 and 254 nm, and
a mass detector. Elemental analysis was conducted by Robertson
Microlit Laboratory, Madison, NJ.
4.2. Experimental procedures
3. Conclusion
4.2.1. 2-Ethoxyacrylonitrile (3). To a 250 mL jacketed 3-neck round
bottom flask with a magnetic stir bar and a temperature controller
was charged under nitrogen with bromoacetaldehyde diethylacetal
(35.33 g, 179.3 mmol) and tin(II) chloride (167.1 mg, 0.88 mmol,
In conclusion, an efficient and practical synthesis of an in vivo
sphingosine-1-phosphate lyase inhibitor LX2931 (1) was achieved.
The key steps of the synthesis involve: (1) an improved preparation

H. Zhang et al. / Tetrahedron 69 (2013) 4041e4046
4045
0.5 mol %). The suspension was cooled to 0e5 ꢀC. Then trime-
thylsilyl cyanide (17.69 g, 178.3 mmol) was added over 10 min at
a temperature below 15 ꢀC. The mixture was held at 10 ꢀC for
35 min and then at about 20 ꢀC for at least 1.75 h or until GC
analysis showed consumption of bromoacetaldehyde diethylacetal.
The above light yellow solution was diluted with 85 mL MTBE
and diethylamine (19.61 g, 268.1 mmol, 1.50 equiv) was added over
6 min at a temperature below 35 ꢀC. The resulting thick slurry was
diluted with an additional 25 mL of MTBE. The reaction mixture
was stirred at 25 ꢀC for at least 1.75 h or until reaction completion
by GC analysis. The mixture was filtered and the collected
solids were washed with MTBE (30 mL, ꢁ2). The combined
filtrate was concentrated under vacuum. The resulting crude 2-
ethoxyacrylonitrile (17.46 g) was filtered and vacuum distilled to
give a clear liquid (11.2 g, 67% yield). Bp: 46e49 ꢀC/31 mmHg. 1H
for 1 h. The imidate solution in R1 was transferred to R2 drop-
wise via a dropping funnel to keep the temperature at 20e25 ꢀC
and the mixture was stirred for 8 h. To the above stirred mixture
was added 25 wt % NaOMe in MeOH (279.2 g, 1.29 mol) drop-wise
via a dropping funnel at 5e10 ꢀC. The mixture was then stirred at
20e25 ꢀC for 5 h. To the stirred mixture were charged H2O (2.4 L)
and HOAc (275.0 g, 4.58 mol), and mixture was heated to
55e60 ꢀC and stirred for another 5 h. The mixture was concen-
trated to about 400 mL and cooled to 0e5 ꢀC and stirred for an
additional 1 h. The solids were collected by filtration, washed with
H2O (200 mL, ꢁ2), and dried under vacuum to afford THI (2) as an
off-white solid (365 g, 75% yield).
4.2.4. (E)-1-(4-((1R,2S,3R)-1,2,3,4-Tetrahydroxybutyl)-1H-imidazol-
2-yl)-ethanone oxime (1). To a 3-neck 3 L round bottom flask
equipped with a mechanical stirrer, a temperature controller and
a condenser were charged under nitrogen with THI (100.0 g,
434.4 mmol), hydroxylamine hydrochloride (45.2 g, 1.5 equiv), so-
dium acetate (53.4 g, 1.5 equiv), and methanol (1.0 L, 10 vol). The
above solution was heated at 65 ꢀC with stirring for 2 h.
NMR (CDCl3)
d
4.85 (d, J¼3.5 Hz, 1H), 4.75 (d, J¼3.3 Hz, 1H), 3.73
(q, J¼7.0 Hz, 2H), 1.24 (t, J¼7.0 Hz, 3H). 13C NMR (CDCl3)
115.2, 101.1, 65.5, 14.4. MHþ¼97.7.
d 136.6,
4.2.2. 1-(4-((1R,2S,3R)-1,2,3,4-Tetrahydroxybutyl)-1H-imidazol-2-yl)
ethanone (THI, 2) using 1-amino-1-deoxy- -fructose acetate (4a). To
D
To the mixture was then added a solution of HCl in isopropanol
(freshly prepared by slow addition of 92.7 mL acetyl chloride to
200 mL isopropanol at 0 ꢀC, 3.0 equiv) over 15 min and the resulting
mixture was stirred at 65 ꢀC for 3 h. The mixture was diluted with
MeOH (1.0 L, 10 vol) and cooled to room temperature and the
precipitated sodium chloride was removed by filtration. The solids
were washed with MeOH (100 mL, 1 vol) and the solution was
concentrated at 40 ꢀC under vacuum until solids started to form
(about 200 mL). Water (1.0 L, 10 vol) was then added and the re-
sidual organic solvents were removed at 40 ꢀC under vacuum.
A polish filtration was performed to afford a clear yellow solution.
To this solution was slowly added 50% aq NaOH solution at room
temperature so that the temperature of the mixture did not exceed
40 ꢀC, until the pH reached 7.2 (7.0e7.5). The resulting solution was
then heated to 65 ꢀC to form a homogeneous solution, and con-
centrated under vacuum at 65 ꢀC (60e70 ꢀC) until the solution
reached about 500 mL (5 vol) overall volume. The mixture was then
cooled to room temperature slowly, further cooled to 0 ꢀC, and
stirred at 0 ꢀC for 1 h. The solids were collected by filtration and
washed with water (0 ꢀC, 100 mL, 1 vol, ꢁ2) to afford a white
crystalline solid.
To the above wet solid was added water (400 mL) and the
resulting mixture was heated to 70e80 ꢀC until all solids dissolved.
The solution was cooled to room temperature and then stirred at
0 ꢀC for 1 h. The solids were collected by filtration and washed with
water (0 ꢀC, 100 mL, 1 vol, ꢁ2) and then dried under vacuum at
30 ꢀC overnight to afford 99.4 g of the dihydrate (KF¼12.8%, 81%
yield). 1H NMR analysis showed that the material contained about
3% of the Z-isomer.
The dihydrate of 1 was slurried with ethanol (800 mL, 8 vol) and
heated at 75 ꢀC for 1 h. The resulting mixture was cooled to 0 ꢀC and
stirred at 0 ꢀC for 1 h. The white solid was collected by filtration and
washed with ethanol (0 ꢀC, 100 mL, 1 vol, ꢁ2), and dried at 50 ꢀC
under vacuum to constant weight to give compound 1 (82.4 g, 95%).
1H NMR analysis showed about 2% of the Z-isomer. 1H NMR (D2O w/
a drop of DCl in D2O) 7.30 (s, 1H), 5.04 (s, 1H), 3.45e3.75 (m, 4H),
2.13 (s, 3H); 13C NMR (D2O w/a drop of DCl in D2O) 143.8, 140.9,
135.0, 116.9, 72.5, 70.6, 64.4, 62.7, 10.5; MHþ¼246.1. Mp:
173e177 ꢀC; Anal. Calcd for C9H15N3O5: C, 39.70; H, 6.66; N, 15.43.
Found: C, 39.52; H, 6.32; N, 15.34.
a 3-neck 12 L round bottom flask equipped with a mechanical
stirrer and a temperature controller was added under nitrogen 2-
ethoxyacrylonitrile (100.0 g, 1.03 mol) and methanol (1.0 L,
10 vol). To the above stirred solution was added sodium meth-
oxide in methanol (25 wt %, 140.6 mL, 0.60 equiv) over 15 min.
The mixture was stirred at 20 ꢀC for at least 2 h or until 1H NMR
showed at least 95% conversion of 2-ethoxyacrylonitrile to
the corresponding imidate. The above solution was transferred
to a slurry of 1-amino-1-deoxy-D
-fructose acetate7 (246.9 g,
1.0 equiv) in methanol (1.0 L, 10 vol) over 15 min and the mixture
was stirred at 20 ꢀC for 6 h. Another portion of sodium methoxide
in methanol (25 wt %, 117.3 mL, 0.50 equiv) was added to the
mixture over 10 min and the mixture was stirred at 20 ꢀC for
additional 16 h.
The mixture was then diluted with water (2.0 L, 20 vol) and
treated with acetic acid (118 mL, 2.0 equiv). After stirring at 60 ꢀC
for 1 h, the solution was concentrated under vacuum to 1.2 L total
volume. The slurry mixture was cooled to 0 ꢀC and stirred for 1 h,
then filtered and the solids were washed with water (100 mL, ꢁ2).
The solids were collected and dried under vacuum at 50 ꢀC to afford
196.8 g crude product as a pale yellow solid, which was then treated
with water (980 mL, 5 vol) and the resulting slurry was heated to
boiling for 15 min, then re-cooled to 0 ꢀC and stirred for an addi-
tional 1 h. The mixture was again filtered, and the solids were
washed with water (100 mL, ꢁ2) and dried to constant weight in
a vacuum oven at 50 ꢀC to provide 194.8 g (82% yield) of THI as
a pale yellow solid (KF¼0.4%). Mp: 233e235 ꢀC (lit. 234e236 ꢀC,5
232e233 ꢀC9); 1H NMR (D2O w/a drop of DCl in D2O) 7.48 (d,
J¼2.0 Hz, 0.9H), 7.19 (d, J¼2.0 Hz, 0.1H), 5.09 (s, 0.9H), 4.98 (s, 0.1H),
3.40e3.70 (m, 4H), 2.53 (d, J¼2.4 Hz, 3H); 13C NMR (D2O w/a drop
of DCl in D2O) 185.0, 139.4, 138.0, 119.5, 73.0, 70.9, 65.0, 63.2, 26.7;
MHþ¼231.2.
4.2.3. 1-(4-((1R,2S,3R)-1,2,3,4-Tetrahydroxybutyl)-1H-imidazol-2-yl)
ethanone (THI, 2) using
D-(þ)-glucosamine hydrochloride (4b). To
a 3-neck 3 L round bottom flask (R1) was charged under nitrogen
2-ethoxyacrylonitrile (200.0 g, 2.06 mol) and MeOH (1.5 L) and
the mixture was stirred at 10e20 ꢀC for 10 min. To the stirred
mixture was added drop-wise 25 wt % NaOMe in MeOH (145.0 g,
0.67 mol) via a dropping funnel at below 10 ꢀC. The mixture was
then stirred at 20e25 ꢀC for 7 h. To another 3-neck 3 L round
bottom flask (R2) equipped with an inlet thermometer and
Acknowledgements
a dropping funnel were charged under nitrogen
D
-(þ)-glucos-
amine hydrochloride (528.3 g, 2.46 mol), NaOAc (200.0 g,
The authors would like to thank Leonard Hargiss for his assis-
tance with LCeMS, Drs. Nicole Goodwin, Tad Jessop, and Lauren
2.44 mol), and MeOH (1.5 L). The mixture was stirred at 20e25 ꢀC

4046
H. Zhang et al. / Tetrahedron 69 (2013) 4041e4046
€
5. (a) Halweg, K. M.; Buchi, G. J. Org. Chem. 1985, 50, 1134e1136; (b) A SciFinder
search at the time showed that THI (2) was commercially available at a pro-
hibitive price of $4250/0.5 g.
Sirois for proofreading the manuscript, and Pharmorphix Ltd. for
obtaining and solving the X-ray structure of compound 1.
6. Cuvigny, T. Ann. Chim. Paris 1956, 1, 475e520.
7. Hodge, J. E.; Fisher, B. E. In Methods in Carbohydrate Chemistry; Whistler, R. L.,
Wolfrom, M. L., Bemiller, J. N., Eds.; Reactions of Carbohydrates; Academic: New
York, NY, 1963; Vol. II, pp 99e103.
Supplementary data
8. Lutz, W. B. J. Org. Chem. 1971, 36, 3835e3837.
9. Sweeny, J. G.; Ricks, E.; Estrada-Valdes, M. C.; Iacobucci, G. A. J. Org. Chem. 1985,
50, 1133e1134.
10. (a) Debus, H. Justus Liebigs Ann. Chem. 1858, 107, 199e208; (b) Radziszewski, B.
Ber. Dtsch. Chem. Ges. 1882, 15, 2706e2708.
1H NMR and 13C NMR spectra of compounds 1, 2, 3 can be found
in the online version. Supplementary data related to this article can
be found online at http://dx.doi.org/10.1016/j.tet.2013.03.071.
11. (a) Cliff, M. D.; Pyne, S. G. Tetrahedron Lett. 1995, 36, 5969e5972; (b) Cliff, M. D.;
Pyne, S. G. J. Org. Chem. 1997, 62, 1023e1032.
12. For a review on Stille coupling, see: Farina, V.; Krishnamurthy, V.; Scott, W. J.
Org. React. 1998, 50, 1e652.
References and notes
13. For a review on sharpless asymmetric dihydroxylation, see: Kolb, H. C.; Van
Nieuwenhze, M. S.; Sharpless, K. B. Chem. Rev. 1994, 94, 2483e2547.
14. (a) Magnus, N. A.; Diseroad, W. D.; Neveil, C. R., Jr.; Wepsiec, J. P. Org.
Process Res. Dev. 2006, 10, 556e560; (b) Bandarage, U.; Hare, B.; Parsons, J.;
Pham, L.; Marhefka, C.; Bemis, G.; Tang, Q.; Moody, C. S.; Rodems, S.; Shah,
S.; Adams, C.; Bravo, J.; Charonnet, E.; Savic, V.; Come, J. H.; Green, J. Bioorg.
Med. Chem. Lett. 2009, 19, 5191e5194; (c) Douty, B.; Wayland, B.; Ala, P. J.;
Bower, M. J.; Pruitt, J.; Bostrom, L. Bioorg. Med. Chem. Lett. 2008, 18, 66e71;
(d) Loksha, Y. M.; Pedersen, E. B.; El-Barbary, A. A.; El-Badawi, M. A.;
Nielsen, C. J. Heterocycl. Chem. 2003, 40, 593e599; (e) Breslin, H. J.;
Miskowski, T. A.; Kukla, M. J.; Leister, W. H.; De Winter, H. L.; Gauthier, D.
A.; Somers, M. V. F.; Peeters, D. C. G.; Roevens, P. W. M. J. Med. Chem. 2002,
45, 5303e5310.
1. (a) Rivera, J.; Proia, R. L.; Olivera, A. Nat. Rev. Immunol. 2008, 8, 753e763;
(b) Gomez-Munoz, A.; Steinbrecher, U. P. Recent Res. Dev. Lipids 2004, 7,
65e86; (c) Reiss, U.; Oskouian, B.; Zhou, J.; Gupta, V.; Sooriyakumaran, P.;
Kelly, S.; Wang, E.; Merril, A. H.; Saba, J. D. J. Biol. Chem. 2004, 279,
1281e1290; (d) Moore, A. N.; Kampfl, A. W.; Zhao, X.; Hayes, R. L.; Dask,
P. K. Neuroscience 1999, 94, 405e415.
2. Bandhuvula, P.; Saba, J. D. Trends Mol. Med. 2007, 13, 210e217.
3. Zambrowicz, B. P.; Friedrich, G. A.; Buxton, E. C.; Lilleberg, S. L.; Person, C.;
Sands, A. T. Nature 1998, 392, 608e611.
4. (a) Bagdanoff, J. T.; Donoviel, M. S.; Nouraldeen, A.; Tarver, J.; Fu, Q.; Carlsen,
M.; Jessop, T. C.; Zhang, H.; Hazelwood, J.; Nguyen, H.; Baugh, S. D. P.;
Gardyan, M.; Terranova, K. M.; Barbosa, J.; Yan, J.; Bednarz, M.; Layek, S.;
Courtney, L. F.; Taylor, J.; Digeorge-Foushee, A. M.; Gopinathan, S.; Bruce, D.;
Smith, T.; Moran, L.; O’Neill, E.; Kramer, J.; Lai, Z.; Kimball, S. D.; Liu, Q.; Sun,
W.; Yu, S.; Swaffield, J.; Wilson, A.; Main, A.; Carson, K. G.; Oravecz, T.; Augeri,
D. J. J. Med. Chem. 2009, 52, 3941e3953; (b) Bagdanoff, J. T.; Donoviel, M. S.;
Nouraldeen, A.; Carlsen, M.; Jessop, T. C.; Tarver, J.; Aleem, S.; Dong, L.; Zhang,
H.; Boteju, L.; Hazelwood, J.; Yan, J.; Bednarz, M.; Layek, S.; Owusu, I. B.;
Gopinathan, S.; Moran, L.; Lai, Z.; Kramer, J.; Kimball, S. D.; Yalamanchili, P.;
Heydorn, W. E.; Frazier, K. S.; Brooks, B.; Brown, P.; Wilson, A.; Sonnenburg,
W. K.; Main, A.; Carson, K. G.; Oravecz, T.; Augeri, D. J. J. Med. Chem. 2010, 53,
15. For dimerization of a-aminoketones under basic conditions, see: (a) Sasagi, T.;
Eguchi, S.; Okano, T. Synthesis 1980, 472e475; (b) Vallat, O.; Neels, A.;
Stoeckli-Evans, H. J. Chem. Crystallogr. 2003, 33, 39e50.
€
16. Buchi reported a low yield (19%) when
was used. See Ref. 5a.
D
-(þ)-glucosamine hydrochloride (4b)
17. For similar acid-promoted E/Z oxime isomerizations, see: (a) O’Ferral, R. A. M.;
€
O’Brien, D. J. Phys. Org. Chem. 2004, 17, 631e640; (b) Demir, A. S.; Sesenoglu, O.;
€
€
Ulku, D.; Arici, C. Helv. Chim. Acta 2003, 86, 91e105.
18. Crystallographic data have been deposited at the CCDC, 12 Union Road, Cam-
bridge CB2 1EZ, UK and copies can be obtained on request, free of charge, by
quoting the publication citation and the deposition number CCDC 921662 for
compound 1$2H2O.
€
8650e8662 For the isolation and characterization of THI, see: (c) Kroplien, U.;
Rosdorfer, J. J. Org. Chem. 1985, 50, 1131e1133 For a seminal biological study of
THI, see: (d) Schwab, S. R.; Pereira, J. P.; Matloubian, M.; Xu, Y.; Huang, Y.;
Cyster, J. G. Science 2005, 309, 1735e1739.