Enantioselective synthesis of the novel chiral sulfoxide derivative as a glycogen synthase kinase 3β inhibitor

Article abstract of DOI:10.1248/cpb.58.1252

Glycogen synthase kinase 3β (GSK-3β) inhibitors are expected to be attractive therapeutic agents for the treatment of Alzheimer's disease (AD). Recently we discovered sulfoxides (S)-1 as a novel GSK-3β inhibitor having in vivo efficacy. We investigated practical asymmetric preparation methods for the scale-up synthesis of (S)-1. The highly enantioselective synthesis of (S)-1 (94% ee) was achieved by titanium-mediated oxidation with D-(-)-diethyl tartrate on gram scale.

Full text of DOI:10.1248/cpb.58.1252

1252
Note
Chem. Pharm. Bull. 58(9) 1252—1254 (2010)
Vol. 58, No. 9
Enantioselective Synthesis of the Novel Chiral Sulfoxide Derivative as a
b
Glycogen Synthase Kinase 3 Inhibitor
Morihisa SAITOH,* Jun KUNITOMO, Eiji KIMURA, Toru YAMANO, Fumio ITOH, and Masakuni KORI
Medicinal Chemistry Research Laboratories, Pharmaceutical Research Division, Takeda Pharmaceutical Co., Ltd.;
2–17–85 Jusohonmachi, Yodogawa-ku, Osaka 532–8686, Japan.
Received May 1, 2010; accepted June 18, 2010; published online June 28, 2010
Glycogen synthase kinase 3b (GSK-3b) inhibitors are expected to be attractive therapeutic agents for the
treatment of Alzheimer’s disease (AD). Recently we discovered sulfoxides (S)-1 as a novel GSK-3b inhibitor hav-
ing in vivo efficacy. We investigated practical asymmetric preparation methods for the scale-up synthesis of (S)-1.
The highly enantioselective synthesis of (S)-1 (94% ee) was achieved by titanium-mediated oxidation with D-(؊)-
diethyl tartrate on gram scale.
Key words glycogen synthase kinase 3b; enantioselective oxidation; Alzheimer’s disease
Alzheimer’s disease (AD) is a neurodegenerative disorder
manifested by cognitive and memory deterioration, progres-
sive impairment of the activities involved in daily living, and
behavioral disturbances. Glycogen synthase kinase 3b (GSK-
3b) inhibitors are thought to be attractive agents for the treat-
ment of AD.1,2) Recently we reported a novel GSK-3b
inhibitor (Fig. 1) and this compound significantly reduced tau
phosphorylation in a mouse model.3)
Fig. 1. Structure of (S)-1
Compound (S)-1 initially was obtained by the resolution of
racemic-1 using chiral high performance liquid chromatogra-
phy (HPLC). Although separation by HPLC was a sufficient
method for the preparation of small quantities of (S)-1, large
scale production was expected to be problematic for this
compound. Thus, we investigated asymmetric synthesis
routes for (S)-1. Among the various methods of asymmetric
oxidation for sulfides,4) the method using a chiral titanium
Chart 1. Strategies for Preparation of Sulfoxides (S)-1
complex derived from a Sharpless reagent reported by Kagan
and colleagues5) is attractive. Our colleagues have previously
applied this methodology to the synthesis of optically active (R)-1, we changed the ligand to (Ϫ)-DET. It is known that
lansoprazole.6) Therefore, we embarked upon optimizing re- the amount of water significantly affects the selectivity of
action conditions for preparing (S)-1 from prochiral sulfides this reaction.5) Although combination of Ti(O-i-Pr)4/H2O
2 (Chart 1).
(1 : 1) is effective in enantioselective oxidation,7) 0.5 eq of
water ensured a good enantiomeric excess for this particular
substrate (entries 3—5). Lower temperature was preferred in
Results and Discussion
The starting material, methyl sulfide 2 was prepared by the terms of selectivity and afforded good enantiomeric excess
6-step protocol previously reported.3) Spectroscopic data and (entries 6—8). However, the reaction rate tended to be slow
X-ray crystal structures confirmed the configuration of (S)- at low temperatures and reaction was not completed within
1.3) First, we attempted conditions similar to those used for 24 h at Ϫ40 °C (entries 9, 10). Enantioselectivity was not
the synthesis of chiral lansoprazole,6) which used a stoichio- affected by the use of molecular sieves (entry 11).7) Other
metric amount of Ti(O-i-Pr)4, (ϩ)-diethyl tartrate (DET) as ligands such as (Ϫ)-dimethyl tartrate (DMT), (Ϫ)-diiso-
ligand, cumene hydroperoxide (CHP) as oxidant, and toluene propyl tartrate (DIPT) and (R)-(ϩ)-1,1Ј-bi-naphthol
or dichloromethane as solvent, in the presence of water. Be- (BINOL) were less effective than (Ϫ)-DET (entries 12—14).
cause of the difficulty in separating the sulfoxide 1 and sul- Because a large excess of water and the use of toluene as sol-
fone 3, we used substoichiometric CHP (0.83 eq) relative to vent enhance the enantioselectivity of oxidations with the
sulfide. The condition using toluene as solvent resulted in titanium complex of BINOL,8) optimized conditions for
production of undesired enantiomer with low enantiomeric tartrate derivatives were not considered suitable for the
excess (Table 1, entry 1). We assumed the low enantioselec- BINOL methodology.
tivity was caused by low solubility of sulfide 2 in toluene.
We also tried asymmetric oxidation using vanadyl acetyl-
Therefore, we used dichloromethane as solvent and reaction acetonate (VO(acac)2) and Schiff base ligand 4,9) but enan-
proceeded in a homogenous solution to show improved enan- tioselectivity was found to be low (Chart 2). Since Schiff
tiomeric excess of 66% ee (entry 2). From this result, we bases are known to significantly affect the enantioselectivity
used dichloromethane in further investigations. Since (ϩ)- of vanadium mediated oxidations,9) screening of Schiff bases
DET resulted in the conversion of sulfide 2 to the undesired and optimization of reaction conditions were considered to
∗ To whom correspondence should be addressed. e-mail: Saitoh_Morihisa@takeda.co.jp
© 2010 Pharmaceutical Society of Japan

September 2010
1253
Table 1. Enantioselective Oxidation of Sulfide 2
Yieldc)
Time
(h)
Entrya)
Ligandb)
H2O
Additive
Solvent
Temp.
ee of (S)-1c)
1
2
3
1
2
3
4
5
6
7
8
9
10
11
12
13
14
L-(ϩ)-DET
L-(ϩ)-DET
D-(Ϫ)-DET
D-(Ϫ)-DET
D-(Ϫ)-DET
D-(Ϫ)-DET
D-(Ϫ)-DET
D-(Ϫ)-DET
D-(Ϫ)-DET
D-(Ϫ)-DET
D-(Ϫ)-DET
D-(Ϫ)-DET
D-(Ϫ)-DIPT
(R)-(ϩ)-BINOL
1.1
1.1
0.5
2
—
—
—
—
—
—
—
—
—
—
MS3A
—
—
—
Toluene
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
rt
0 °C
0 °C
0 °C
0 °C
24
24
5
5
5
5
5
24
5
24
5
—
—
—
—
2.9
—
—
32.1d)
66.0d)
86.5
41.0
85.0
90.6
92.6
93.5
92.2
94.0
91.5
57.9
88.3
30.6
65.6
39.4
70.3
66.9
83.5
80.3
36.2
60.4
57.0
56.0
64.8
59.2
3.2
1.8
2.8
1.3
0.5
1.6
1.5
0.6
1.3
0.4
1.8
0.7
12.1
23.9
16.0
36.0
11.9
58.5
23.0
14.3
43.6
25.0
32.1
0
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
Ϫ20 °C
Ϫ30 °C
Ϫ30 °C
Ϫ40 °C
Ϫ40 °C
Ϫ20 °C
Ϫ20 °C
Ϫ20 °C
Ϫ20 °C
5
5
5
a) Cumene hydroperoxide solution (entry 1—13: 0.83 eq, entry 14: 1.05 eq). b) DET: diethyl tartrate, DMT: dimethyl tartrate, DIPT: diisopropyl tartrate, BINOL: 1,1Ј-bi-
2-naphthol. c) Determined by chiral HPLC. d) ee of (R)-1.
Chart 2. VO(acac)2-Catalyzed Enantioselective Oxidation of Sulfide 2
Chart 3. Scale-Up Synthesis of (S)-1
take enormous efforts, thus we stopped further investigation
of using vanadium catalysts.
On the basis of these results obtained above, we applied
doublet; dd, doublet of doublet; m, multiplet. Coupling constants (J values)
are given in hertz (Hz). Enantiomeric excess and the conversion of yields
were all determined by HPLC analysis on a CHIRALPAK AS column (0.5
the best conditions of titanium mediate asymmetric oxidation
(Table 1, entry 8, Ti(O-i-Pr)4/H2O (1 : 0.5), Ϫ20 to Ϫ30 °C,
24 h) on a larger scale, and the reaction of 40 g of sulfide 2
proceeded well to afford 33 g of (S)-1 in 78% yield and 94%
ee (Chart 3).
cmϫ50 cm, EtOH, 0.4 ml/min) for the conversion of (S)-1.
General Procedure for Asymmetric Oxidation of Sulfide To a mix-
ture of sulfide 2 (1.0 eq) and H2O was added a tartrate ester or (R)-(ϩ)-1,1Ј-
bi-naphthol (2.6 eq) in solvent (5 ml) and Ti(O-i-Pr)4 (1.0 eq), and the solu-
tion was stirred for 1 h at room temperature. Then the mixture was cooled,
CHP (0.83 eq or 1.05 eq) was added to the solution and was stirred at the
same temperature. The reaction mixture was quenched with saturated aque-
ous Na2SO3 solution, the organic layer was washed with H2O and brine,
dried over Na2SO4, filtered and concentrated in vacuo. The yield and enan-
tiomeric excess of sulfoxide (S)-1 were determined by chiral HPLC.
Vanadium Catalyzed Asymmetric Synthesis (Chart 2) A mixture of
VO(acac)2 (1.64 mg, 6.20 mmol), 4 (2.97 mg, 9.30 mmol) in dichloromethane
(5 ml) was stirred for 30 min at room temperature, then 2 (0.20 g, 0.62
mmol) was added. The mixture was cooled to 0 °C and was added H2O2
(30%, 76 ml, 0.74 mmol) at the same temperature. After stirring for 28 h at
0 °C, the reaction mixture was partitioned between EtOAc and H2O. The
combined organic layers were washed with brine, dried over Na2SO4, filtered
and concentrated in vacuo. The yield of sulfoxide was 66% and enantiomeric
Conclusion
We developed a practical enantioselective synthesis for the
preparation of the novel GSK-3b inhibitor (S)-1. The opti-
mized amount of water in Kagan’s enantioselective oxidation
of sulfide enhanced enantioselectivity in the synthesis of (S)-
1. This methodology was also applied for the large scale syn-
thesis and 33 g of (S)-1 with 94% ee was obtained. This
scale-up synthesis facilitated further pharmacological and
toxicological studies of these compounds. Catalytic asym-
metric oxidation is a significant challenge in the future of excess was 35% ee (determined by chiral HPLC).
Large Scale Asymmetric Synthesis of (S)-1 To a mixture of sulfide 2
organic synthesis.
(40.0 g, 124 mmol) and H2O (1.12 ml, 62.1 mmol) was added (Ϫ)-DET
(53.6 ml, 273 mmol) in dichloromethane (1.2 l) and Ti(O-i-Pr)4 (36.7 ml,
124 mmol), and the solution was stirred for 1 h at room temperature. The
Experimental
Melting points were determined on a Buchi melting point apparatus and
were not corrected. All commercially available reagents and solvent were mixture was cooled to Ϫ24 °C, CHP (80%, 19 ml, 103 mmol) was added to
employed without prior purification. Proton nuclear magnetic resonance (1H-
NMR) spectra were recorded on a Bruker DPX300 (300 MHz) instrument.
the solution, and stirring was continued at Ϫ24 °C for 24 h. The reaction
mixture was quenched with saturated aqueous Na2SO3 solution, the organic
Chemical shifts are reported as d values (ppm) downfield from internal layer was washed with H2O and brine, dried over Na2SO4, filtered and con-
tetramethylsilane referenced to the residual solvent peak. Peak multiplicities centrated in vacuo. The residue was purified by silica gel (400 g) using
are expressed as follows. Abbreviations are used as follows: s, singlet; d, EtOAc–MeOH (1/0—9/1) and NH silica gel (500 g) using hexane– tetrahy-

1254
Vol. 58, No. 9
1
tor G. P., Snell G. P., Takizawa M., Itoh F., Kori M., J. Med. Chem., 52,
6270—6286 (2009).
drofuran (THF)(1/2) to give (S)-1 (80%, 94% ee) as a colorless solid. H-
NMR (300 MHz, CDCl3) d: 2.65 (3H, s), 2.81 (3H, s), 7.69 (1H, d, Jϭ8.7
Hz), 7.78—7.86 (4H, m), 7.94 (1H, s), 8.08 (1H, dd, Jϭ8.7, 1.9 Hz), 8.48—
8.51 (1H, m). [a]D25 Ϫ89.9 (cϭ0.49, MeOH). mp 170—171 °C. tRϭ31.8 min.
4) Kagan H. B., “Organosulfur Chemistry in Asymmetric Synthesis,” ed.
by Toru T., Bolm C., Wiley InterScience, New York, 2008, pp. 1—26.
5) Pitchen P., Duñach E., Deshmukh M. N., Kagan H. B., J. Am. Chem.
Soc., 106, 8188—8193 (1984).
6) Kawada M., Yamano T., Taya N., Jpn. Kokai Tokyo Koho JP2001-
131172.
Acknowledgements The authors thank Mr. Katsuhiko Miwa for HPLC
analyses. We also thank Mr. Yuichiro Kondo and Mr. Masayuki Yamashita
for discussions about experimental procedures.
7) Zhao S. H., Samuel O., Kagan H. B., Tetrahedron, 43, 5135—5144
(1987).
References
8) Komatsu N., Hashizume M., Sugita T., Uemura S., J. Org. Chem., 63,
7624—7626 (1993).
9) Skarz˙ewski J., Ostrycharz E., Siedlecka R., Tetrahedron: Asymmetry,
10, 3457—3461 (1999).
1) Cohen P., Goedert M., Nat. Rev. Drug Discov., 3, 479—487 (2004).
2) Martinez A., Med. Res. Rev., 28, 773—796 (2008).
3) Saitoh M., Kunitomo J., Kimura E., Iwashita H., Uno Y., Onishi T.,
Uchiyama N., Kawamoto T., Tanaka T., Mol C. D., Dougan D. R., Tex-