Catalytic enantioselective synthesis of key intermediates for triazole antifungal agents

Article abstract of DOI:10.1021/ol050398+

(Chemical Equation Presented) A short-step synthesis of versatile chiral building blocks for triazole antifungal agents such as ZD0870 and Sch45450 was developed via catalytic enantioselective cyanosilylation of electron-deficient ketones as the key step.

Full text of DOI:10.1021/ol050398+

ORGANIC
LETTERS
2005
Vol. 7, No. 13
2527-2530
Catalytic Enantioselective Synthesis of
Key Intermediates for Triazole
Antifungal Agents
Masato Suzuki,† Nobuki Kato,‡ Motomu Kanai,*,†,‡ and Masakatsu Shibasaki*,†
Graduate School of Pharmaceutical Sciences, The UniVersity of Tokyo, Tokyo
113-0033, Japan, and PRESTO, Japan Science and Technology Corporation, Japan
mshibasa@mol.f.u-tokyo.ac.jp
Received February 23, 2005 (Revised Manuscript Received May 21, 2005)
ABSTRACT
A short-step synthesis of versatile chiral building blocks for triazole antifungal agents such as ZD0870 and Sch45450 was developed via
catalytic enantioselective cyanosilylation of electron-deficient ketones as the key step. High enantioselectivity was produced using a catalyst
prepared from Gd(HMDS)3 and ligand 5 in a 2:3 ratio. This new catalyst preparation method was superior to the previous method using
Gd(OiPr)3 as a metal source. A rationale for the difference is proposed on the basis of structural studies of the catalyst complexes using
ESI-MS.
Systemic fungal infections can be fatal for patients with
immune deficiency or suppression such as occurs in AIDS,
cancer, or following organ transplants. Triazole antifungal
agents (Figure 1) demonstrate great potential to treat these
infectious diseases because of their broad antifungal spectrum
and low toxicity.1 Fluconazole, itraconazole, and voricon-
azole are clinically used representative antifungal drugs of
this family, and at least several compounds are currently
undergoing clinical trials. Therefore, the development of an
efficient and convergent synthetic route that can be applied
to large-scale synthesis of these antifungals is an important
goal.
Although excellent enantioselectivity was obtained in these
cases, multiple steps were required for the synthesis of
substrates of the key catalytic enantioselective reactions,
which decreases the total synthetic efficiency. In this
communication, we describe a short synthetic route to 1 using
a catalytic, enantioselective cyanosilylation of commercially
available ketone 2a as the key step. During the optimization
of the key cyanosilylation reaction, we developed a new
catalyst preparation protocol using Gd(HMDS)3 as a metal
source. The catalyst prepared through this new method
produced significantly higher enantioselectivity than the
previous method using Gd(OiPr)3.
Many triazole antifungal agents share a common synthetic
intermediate, 1, containing a chiral tertiary alcohol. Previous
synthesis of 1 utilized either Sharpless asymmetric dihy-
droxylation,2 Sharpless asymmetric epoxidation,3 or enzy-
matic transformations4 as a key enantioinduction step.
We developed an enantioselective cyanosilylation of
ketones catalyzed by a gadolinium complex generated from
Gd(OiPr)3 and ligand 4 in a 1:2 ratio.5,6 This reaction is
practical and applicable to the catalytic asymmetric synthesis
of key intermediates of important pharmaceuticals such as
† The University of Tokyo.
(3) Saksena, A. K.; Girijiavallabhan, V. M.; Lovey, R. G.; Pike, R. E.;
Desai, J. A.; Ganguly, A. K.; Hare, R. S.; Loebenberg, D.; Cacciapuoti,
A.; Parmegiani, R. M. Bioorg. Med. Chem. Lett. 1994, 4, 2023.
(4) (a) Lovey, R. G.; Saksena, A. K.; Girijiavallabhan, V. M. Tetrahedron
Lett. 1994, 35, 6047-6050. (b) Yasohara, Y.; Miyamoto, K.; Kizaki, N.;
Hasegawa, J.; Ohashi, T. Tetrahedron Lett. 2001, 42, 3331.
‡ PRESTO.
(1) For review, see: (a) Como, J. A.; Dismukes, W. J. New Engl. J.
Med. 1994, 330, 263. (b) ZD0870: Peng, T.; Galgiani, J. N. Antimicrob.
Agents Chemother. 1993, 37, 2126. (c) Sch45450: Murakami, K.; Mochi-
zuki, H.; Boyle, F.; Wardleworth, J. M. Eur. Pat. Appl. 472392, 1992.
(2) Blundell, P.; Ganguly, A. K.; Girijiavallabhan, V. M. Synlett 1994,
263.
(5) Yabu, K.; Masumoto, S.; Yamasaki, S.; Hamashima, Y.; Kanai, M.;
Du, W.; Curran, D. P.; Shibasaki, M. J. Am. Chem. Soc. 2001, 123, 9908.
10.1021/ol050398+ CCC: $30.25
© 2005 American Chemical Society
Published on Web 06/02/2005

Scheme 1. Synthetic Plan
On the basis of the above plan, we started our studies with
optimization of the catalytic enantioselective cyanosilylation
of reactive ketone 2a. When previously optimized conditions
were applied to 2a using the catalyst derived from Gd(OiPr)3
and ligand 4 in propionitrile at -40 °C, product 3a was
obtained in good yield, but with only 24% ee (Table 1, entry
1). Using ligand 5 containing difluorocatechol, the enantio-
selectivity was significantly improved to 68% ee (entry
2).9-11 Efforts to further improve the enantioselectivity
through optimization of the fundamental reaction conditions
(temperature, concentration, solvent, lanthanide metals, and
metal/ligand ratio), ligand structure tuning, or additive effects
were not successful.
Figure 1. Examples of triazole antifungal agents.
camptothecins7 and oxybutynin.8 On the basis of preliminary
studies of the catalyst structure using ESI-MS, the active
catalyst was proposed to be a 2:3 complex 6 of gadolinium
cyanides and the ligands. Because slightly higher enantiose-
lectivity was obtained when the catalyst was prepared with
a 1:2 gadolinium/ligand ratio than with a 2:3 ratio, we
determined this 1:2 ratio to be optimal for simple ketones.5
On the basis of this background, as well as the increasing
demands for new triazole antifungal agents, we planned to
develop an alternative synthetic route to 1 using the catalytic
enantioselective cyanosilylation of ketones.
Delightful results were obtained when switching the metal
source from Gd(OiPr)3 to Gd(HMDS)3. Thus, when the
catalyst was prepared from Gd(HMDS)3 and 5 in a 1:2 ratio,
product 3a was obtained with 80% ee (Table 1, entry 4).12
Screening of the gadolinium/ligand ratio using 2 mol %
catalyst at -30 °C revealed that the highest enantioselectivity
was produced with Gd(HMDS)3/5 ) 2:3 (entry 6), and
product 3a was obtained with a synthetically useful enan-
(6) For examples of catalytic enantioselective cyanation of ketones from
other groups, see: (a) Belokon’, Y. N.; Green, B.; Ikonnikov, N. S.; North,
M.; Tararov, V. I. Tetrahedron Lett. 1999, 38, 6669. (b) Belokon’, Y. N.;
Caveda-Cepas, S.; Green, B.; Ikonnikov, N. S.; Khrustalev, V. N.; Larichev,
V. S.; Moscalenko, M. A.; North, M.; Orizu, C.; Tararov, V. I.; Tasinazzo,
M.; Timofeeva, G. I.; Yashkina, L. V. J. Am. Chem. Soc. 1999, 121, 3968.
(c) Belokon’, Y. N.; Green, B.; Ikonnikov, N. S.; Larichev, V. S.; Lokshin,
B. V.; Moscalenko, M. A.; North, M.; Orizu, C.; Peregudov, A. S.;
Timofeeva, G. I. Eur. J. Org. Chem. 2000, 2655. (d) Belokon’, Y. N.; Green,
B.; Ikonnikov, N. S.; North, M.; Parsons, T.; Tararov, V. I. Tetrahedron
2001, 57, 771. (e) Tian, S.-K.; Deng, L. J. Am. Chem. Soc. 2001, 123,
6195. (f) Tian, S.-K.; Deng, L. J. Am. Chem. Soc. 2003, 125, 9900. (g)
Deng, H.; Isler, M. P.; Snapper, M. L.; Hoveyda, A. H. Angew. Chem., Int.
Ed. 2002, 41, 1009. (h) Chen, F.; Feng, Z.; Qin, B.; Zhang, G.; Jiang, Y.
Org. Lett. 2003, 5, 949. (i) Shen, Y.; Feng, X.; Li, Y.; Zhang, G.; Jiang, Y.
Eur. J. Org. Chem. 2004, 129.
(7) (a) Yabu, K.; Masumoto, S.; Kanai, M.; Curran, D. P.; Shibasaki,
M. Tetrahedron Lett. 2002. 43, 2923. (b) Yabu, K.; Masumoto, S.; Kanai,
M.; Shibasaki, M. Heterocycles 2003, 59, 369.
(8) (a) Masumoto, S.; Suzuki, M.; Kanai, M.; Shibasaki, M. Tetrahedron
Lett. 2002. 43, 8647. (b) Masumoto, S.; Suzuki, M.; Kanai, M.; Shibasaki,
M. Tetrahedron 2004, 60, 10497.
Figure 2. Chiral ligands and proposed catalyst structure for
cyanosilylation of ketones.
(9) Chiral ligands 4 and 5 are commercially available from Junsei
Chemical Co., Ltd., Tokyo, Japan. Fax: +81-3-3270-5461.
(10) Ligands 4 and 5 produced almost comparable enantioselectivity from
simple ketones (e.g., acetophenone or 2-heptanone) as a substrate. Therefore,
the significant difference in enantioselectivity between 4 and 5 is specific
for the electron-deficient substrate 2a.
(11) A marked advantage of 5 to 4 was observed in the Strecker reaction
of ketoimines and the conjugate addition of cyanide: (a) Masumoto, S.;
Usuda, H.; Suzuki, M.; Kanai, M.; Shibasaki, M. J. Am. Chem. Soc. 2003,
125, 5634 (Strecker reaction). (b) Mita, T.; Sasaki, K.; Kanai, M.; Shibasaki,
M. J. Am. Chem. Soc. 2005, 127, 514 (conjugate addition).
(12) Addition of a small amount of HMDS (1 equiv to gadolinium) to a
catalyst solution prepared from Gd(OiPr)3 (2 mol %) did not improve the
enantioselectivity, which suggested that the improvement is not due to
remaining trace amount HMDS.
Our synthetic plan is summarized in Scheme 1. Because
our preliminary studies suggested that synthetically useful
enantioinduction from the corresponding triazole-containing
ketone is difficult, we selected commercially available ketone
2a as a substrate for the key catalytic enantioselective
cyanosilylation. Once 3a is obtained with high enantiose-
lectivity, reduction of the cyanide into the hydroxymethyl
group followed by substitution of the chloride by triazole
should give the versatile intermediate 1.
2528
Org. Lett., Vol. 7, No. 13, 2005

Table 1. Catalytic Enantioselective Cyanosilylation of Ketones
for the Synthesis of Triazole Antifungals
Gd sourceb
(mol %)
ligand
(mol %) ketone
time yieldc eed
entrya
(h)
(%)
(%)
1
2
3
4
5
6
7
8
A (5)
A (5)
A (5)
B (5)
B (2)
B (2)
A (2)
B (2)
4 (10)
5 (10)
5 (7.5)
5 (10)
5 (4)
5 (3)
5 (4)
5 (3)
2a
2a
2a
2a
2a
2a
2b
2b
14
3
3
3
3
3
3
3
74
85
83
96
81
93
99
95
24
68
52
80
80
83
91
96
a Reaction temperature was -40 °C for entries 1-4 and -30 °C for
entries 5-8. b A ) Gd(OiPr)3. B ) Gd(HMDS)3. c Isolated yield. d Enan-
tiomeric excess (ee) was determined by chiral GC. See Supporting
Information for details.
tioselectivity of 83% ee (entry 6).13 This optimized catalyst
preparation protocol also produced better enantioselectivity
in the case of another electron-deficient ketone, 2b, and
product 3b was obtained with 96% ee (entry 8).14 3b should
be a useful chiral building block for the synthesis of new
antifungal triazole derivatives.15
Having established synthetically useful catalytic enantio-
selective cyanosilylation for electron-deficient ketone 2a,
conversion of 3a to the versatile, enantiomerically pure
intermediate 1 was relatively straightforward (Scheme 2).
Scheme 2. Conversion to Intermediate 1
Figure 3. ESI-MS observations of the catalyst complexes.
Reduction of the aldehyde with NaBH4, followed by
substitution of the chloride with triazole anion, gave the target
1 with reasonable yield. Enantiomerically pure 1 was
obtained through recrystallization from acetonitrile. This is
the shortest catalytic enantioselective synthesis of the
versatile intermediate 1 from a commercially available
starting material.
To elucidate a possible origin of the enantioselectivity
difference between using Gd(OiPr)3 or Gd(HMDS)3 as a
metal source, we studied catalyst compositions and molecular
weights by ESI-MS. After Gd(OiPr)3 and 5 were mixed in a
1:2 ratio, the liberated iPrOH was evaporated under vacuum
Reduction of the cyanide with DIBAL-H and treatment with
aqueous HCl provided the hydroxyaldehyde 7a in high yield.
(13) Representative Procedure Using the Catalyst Prepared from
Gd(HMDS)3. Gd(HMDS)3 (0.17 M stock solution in THF, 1.3 mL, 0.22
mmol) was added to a solution of 5 (164 mg, 0.35 mmol) in THF (7 mL)
in an ice bath, and the mixture was warmed to 45 °C for 30 min. After
cooling to room temperature, volatiles were evaporated, and the resulting
white powder was dried under vacuum (2 mmHg) at 45 °C for 3 h.
Propionitrile (7 mL) was added, followed by the addition of TMSCN (2.1
mL, 15.8 mmol) at -30 °C. After 10 min, substrate ketone 2a (2.00 g,
10.5 mmol) was added to start the reaction.
(14) In contrast to the results obtained from electron-deficient ketones
2a and 2b, comparably high enantioselectivity was obtained from a simple
ketone (acetophenone), either using Gd(OiPr)3 or Gd(HMDS)3 as a
gadolinium source (95 and 96% ee, respectively, using 2.5 mol catalyst at
-40 °C). The different tendency might be attributed to the susceptibility
of the reactive ketones to the undesired reaction pathways promoted by the
contaminating complex 8 (see text).
(15) Itoh, H.; Furukawa, Y.; Tsuda, M.; Takeshiba, H. Bioorg. Med.
Chem. 2004, 12, 3561.
Org. Lett., Vol. 7, No. 13, 2005
2529

(ca. 2 mmHg) for 1 h. The resulting alkoxide complex was
dissolved in CH3CN, and TMSCN (10 equiv) was added to
generate the active gadolinium cyanide catalyst. Two main
species were observed in this solution by ESI-MS measure-
ment (Figure 2, A); one major peak corresponded to the 2:3
complex 6 (observed MW ) 1868, calcd for [M - 2CN +
OMe]+ ) 1867), which is consistent with previous observa-
tions.16 Another higher molecular weight peak corresponded
to a µ-oxo-containing 4:5 complex 8 (observed MW ) 3010,
calcd ) 3011).17 On the other hand, when the active catalyst
was prepared from Gd(HMDS)3 and 5 in a 2:3 ratio, the
peak corresponding to complex 8 was not observed (Figure
2, B).18 On the basis of this comparison, we proposed that
the difference in enantioselectivity is due to the presence of
the µ-oxo 4:5 complex.19 This complex might promote the
cyanosilylation of reactive ketones such as 2a and 2b with
lower enantioselectivity than the desired 2:3 complex. The
activity of the µ-oxo 4:5 complex, however, should be lower
than that of the 2:3 complex because enantioselectivity from
simple ketones such as acetophenone is not affected by the
gadolinium source. These results also suggested a reason the
optimum metal/ligand ratio is different depending on the
gadolinium source; more chiral ligand might be necessary
to maximize the 6/8 ratio when the catalyst is prepared from
Gd(OiPr)3.
In summary, the enantioselectivity of catalytic cyanosi-
lylation of reactive ketones can be improved using a catalyst
prepared from Gd(HMDS)3 and difluorocatechol-containing
ligand 5 in a 2:3 ratio. This finding allowed for a short,
efficient enantioselective synthesis of a versatile chiral
intermediate 1 for triazole antifungal agents. Preliminary
studies of the catalyst compositions using ESI-MS suggested
that the improvement in enantioselectivity is due to the
absence of contamination by the µ-oxo 4:5 complex. This
study should facilitate further structural studies of the active
and enantioselective 2:3 complex. Studies are ongoing in the
due course.
(16) ESI-MS provided valuable information about the compositions of
the multimetallic chiral gadolinium complexes. See: Kato, N.; Suzuki, M.;
Kanai, M.; Shibasaki, M. Tetrahedron Lett. 2004, 45, 3147. See also ref 5.
The methoxide ligand might come from remaining trace amounts of MeOH
in the MS apparatus. Observation of the peak corresponding to the
methoxide-incorporated 2:3 complex is to be considered an exception only
in cases where ligand 5 was used. Using ligand 4, the 2:3 complex 6′ (calcd
for [M - CN]+ ) 1754) without methoxide incorporation was observed,
concomitant with the µ-oxo 4:5 complex 8′ (calcd for [M - CN]+ ) 2831;
see ESI-MS chart below).
Acknowledgment. Financial support was provided by a
Grant-in-Aid for Specially Promoted Research of MEXT and
PRESTO of the Japan Science and Technology Corporation.
We thank Mr. Tsuyoshi Mita for ESI-MS measurement.
Supporting Information Available: Experimental pro-
cedures and characterization of the products. This material
is available free of charge via the Internet at http://pubs.acs.org.
OL050398+
(18) Only a single peak corresponding to the 2:3 complex was observed,
even when the catalyst was prepared from Gd(HMDS)3 and 5 in a 1:2 ratio,
which excluded the possibility that generation of the µ-oxo 4:5 complex is
due to the different metal/ligand ratio.
(19) Studies to selectively generate the µ-oxo 4:5 complex and to
elucidate the catalyst activity and enantioselectivity of this complex are
ongoing.
(17) Rare earth metal isopropoxides can generate the µ-oxo complex even
under strictly anhydrous conditions with liberation of diisopropyl ether at
temperatures higher than the boiling point of 2-propanol. See: Bradley, D.
C.; Chudzynska, H.; Frigo, D. M.; Hammond, M. E.; Hursthouse, M. B.;
Mazid, M. A. Polyhedron 1990, 9, 719.
2530
Org. Lett., Vol. 7, No. 13, 2005