Dynamic kinetic asymmetric transformations of conduritol B tetracarboxylates: An asymmetric synthesis of D-myo-inositol 1,4,5- trisphosphate [4]

Full text of DOI:10.1021/ja992960v

10834
J. Am. Chem. Soc. 1999, 121, 10834-10835
Scheme 1. Dynamic Kinetic Asymmetric Transformation of
Conduritol B
Dynamic Kinetic Asymmetric Transformations of
Conduritol B Tetracarboxylates: An Asymmetric
Synthesis of D-myo-Inositol 1,4,5-Trisphosphate
Barry M. Trost,* Daniel E. Patterson, and Erik J. Hembre
Department of Chemistry, Stanford UniVersity
Stanford, California 94305-5080
ReceiVed August 16, 1999
The importance of the cyclohexene tetraol conduritol B and
its derivatives as synthetic building blocks is amply demonstrated
by its extensive use in natural product synthesis.1-7 The synthesis
of derivatives of conduritol B in enantiomerically enriched form
has proven difficult, requiring resolutions8-10 or synthesis from
chiral pool starting materials.11-17 The enantioselective palladium-
catalyzed allylic alkylation18 of conduritol B derivatives 1 is
somewhat complex due to their C2-axis of symmetry (see Scheme
1). The ionization of either enantiomer of 1 leads to the same
meso π-allylpalladium complex 2, however, at different rates with
chiral catalysts (matched vs mismatched ionization). In the
presence of the chiral ligand (R,R) 3, matched attack of a
nucleophile at one of the enantiotopic termini of the π-allyl 2
gives the monosubstitution product (-)-4. Mismatched attack
would lead to the opposite enantiomer (+) 4. Furthermore, the
monoalkylated product 4 can be a substrate for a second allylic
alkylation. Ionization of (-)-4 would lead to π-allylpalladium
complex 5, which can go on to the 1,4 disubstituted product 6
since both processes involve matched events, or, less likely, to
the 1,2 disubstituted product 7 via a mismatched nucleophilic
attack. On the other hand, both formation and further reactions
of (+)-4 should be disfavored (i.e., mismatched) under these
conditions.
Since the initial ionization of racemic 1 involves both a matched
and mismatched ionization, a kinetic resolution of conduritol B
tetracarboxylate should be possible and was demonstrated with
oxygen nucleophiles.19 On the other hand, it is more desirable to
perform a dynamic kinetic asymmetric transformation (henceforth
abbreviated DYKAT) on the conduritol B system to provide
complete conversion of the racemic starting material to enantio-
merically enriched product. To achieve this goal, the chiral
palladium-ligand complex must be able to ionize both enanti-
omers of the starting material 1, and then convert the resulting
meso π-allyl into a single enantiopure product. This goal is
complicated by the fact that, because the ionization of (-)-4 is a
matched event, its ionization leading to polysubstitution may be
competitive or even faster than the mismatched ionization of (+)-
1sa circumstance that can lead to mixtures of mono- and
disubstituted products. We wish to report that synthetically useful
DYKAT reactions leading selectively to either mono- or di-
substitution can be accomplished. Further, the latter has been
applied to a short asymmetric synthesis of the cellular signal
transducer D-myo-inositol-1,4,5-trisphosphate, and formal syn-
theses of the 1,2,4,5-tetrakisphosphate and cyclophellitol, an
inhibitor of HIV.
Racemic conduritol B tetraacetate 8 was synthesized in three
steps from benzoquinone by a simple modification of the method
of Guo.20 When 8 was used as a substrate for the palladium-
catalyzed allylic alkylation, a kinetic resolution was observed with
oxygen nucleophiles, and no reaction was observed with carbon
and nitrogen nucleophiles. In hopes of increasing the reactivity,
tetraacetate 8 was converted into the tetramethyl carbonate 9 or
the tetratrichloroethyl carbonate (TROC) 10 by a two-step, one-
pot procedure. While low conversion was still observed in the
(1) Park, T. K.; Danishefsky, S. J. Tetrahedron Lett. 1994, 35, 2667.
(2) Chida, N.; Ohtsuka, M.; Nakazawa, K.; Ogawa, S. J. Org. Chem. 1991,
56, 2976.
(3) Arjona, O.; de Dios, A.; Plumet, J.; Saez, B. Tetrahedron Lett. 1995,
36, 1319.
(4) Lalegerie, P.; Legler, G.; Yon, J. M. Biochemie 1982, 64, 977.
(5) Lee, K. J.; Boyd, S. A.; Radin, N. S. Carbohydr. Res. 1985, 144, 148.
(6) Sato, K. I.; Sakuma, S.; Muramatsa, S.; Bokura, M. Chem. Lett. 1991,
1473.
(7) Trost, B. M.; Pulley, S. R. J. Am. Chem. Soc. 1995, 117, 10143.
(8) Le Drian, C.; Vionnet, J. P.; Vogel, P. HelV. Chim. Acta 1990, 73,
161.
(9) Innes, J. E.; Edwards, P. J.; Ley, S. V. J. Chem. Soc., Perkin Trans. 1
1997, 795.
(10) Yu, L.; Cabrera, R.; Ramerez, J.; Malinovskii, V. A.; Brew, K.; Wang,
P. G. Tetrahedron Lett. 1995, 36, 2897.
(11) Watanabe, Y.; Mitani, M.; Ozaki, S. Chem. Lett. 1987, 123.
(12) Akiyama, T.; Shima, H.; Ozaki, S. Tetrahedron Lett. 1991, 32, 5593.
(13) Akiyama, T.; Takechi, N.; Shoichiro, O. Tetrahedron Lett. 1990, 31,
1433.
(14) Akiyama, T.; Shima, H.; Ohnari, M.; Okazaki, T.; Ozaki, S. Bull.
Chem. Soc. Jpn. 1993, 66, 3760.
reactions of tetracarbonate 9 with carbon nucleophiles (such as
Meldrum’s acid), switching to the more reactive tetratrichloroethyl
carbonate 10 led to 95% yield of monosubstituted product 11b
with an enantiomeric excess of 96%.21 Acetoxy Meldrum’s acid
behaved similarly whereby 11c was obtained in 89% yield and
89% ee (99% after one recrystallization). The fact that complete
conversion is observed indicates that a DYKAT has been
achieved, that is, both enantiomers of starting material are being
ionized and converted to the same product. None of the disub-
stituted product was observed in alkylations using carbon nu-
cleophiles. Surprisingly, when the steric bulk of the carbon
nucleophile was increased by using methyl Meldrum’s acid, the
(20) Guo, Z.-X.; Haines, A. H.; Pyke, S. M. Pyke, S. G.; Taylor, R. J. K.
Carbohydr. Res. 1994, 264, 147.
(15) Paulsen, H.; von Deyn, W. Liebigs Ann. Chem. 1987, 133.
(16) Paulsen, H.; Roben, W.; Heiker, F. R. Chem. Ber. 1981, 114, 3242.
(17) Paulsen, H.; Heiker, F. R. Liebigs Ann. Chem. 1981, 2180.
(18) Trost, B. M.; Van Vranken, D. L. Chem. ReV, 1996, 96, 395.
(19) Trost, B. M.; Hembre, E. J. Tetrahedron Lett. 1999, 40, 219.
(21) The enantiomeric excess of 11b was determined by conversion to the
corresponding dimethyl ester, and chiral HPLC analysis of this ester (see
Supporting Information). The enantiomeric excess of the recovered starting
material 10 was determined by conversion to the tertaacetate 8 and chiral
HPLC on 8.
10.1021/ja992960v CCC: $18.00 © 1999 American Chemical Society
Published on Web 11/09/1999

Communications to the Editor
J. Am. Chem. Soc., Vol. 121, No. 46, 1999 10835
ions from the non-mitochondrial intracellular store. Because of
its importance in signal transduction in animal cells and low
availibility from natural sources, 1,4,5-IP3 has been the focus of
synthetic interest.23 To date there are no syntheses of 1,4,5-IP3
that generate chirality using a catalytic asymmetric transformation.
Dibenzoate 14, available in >99% enantiomeric excess as
described above, is dihydroxylated with osmium tetroxide/NMO
at 0 °C (temperature important) to give the desired diol 16 in
83% yield as a single diastereomer. Because of the slow rate of
reaction became mainly a simple kinetic resolution giving 11d
of 99% ee in 42% (71% brsm) yield. Increasing the palladium
loading to 10% increased the yield to 55% (99% ee) which
demonstrates the beginning of the DYKAT.
When phthalimide was used in the substitution reaction of 10,
the result paralleled those with methyl Meldrum’s acid. Using
2.5 mol % of (dba)3Pd2‚CHCl3, the monosubstituted product 12b
was obtained in 37% yield (maximum 50% yield) and 97%
enantiomeric excess in a kinetic resolution. When the catalyst
the dihydroxylation, however, a 5% loading of osmium was
required to ensure a reasonable reaction time.
Chemoselective removal of the TROC groups with zinc dust
in THF containing acetic acid gave an 88% yield of the tetraol
18. Completion of the synthesis in two steps from tetraol 18 as
reported by Meek et al.24 gave pure 1,4,5-IP3 (-24.4° (c 0.2, H2O,
pH adjusted to ∼10 with cyclohexylamine), lit25 -25.2° (c 1.03,
H2O, pH adjusted to 10.5 with cyclohexylamine) whose spectral
and physical properties fully agree with the natural product. The
synthesis of intermediate 18 also represents a formal asymmetric
synthesis of D-myo-1,2,4,5-tetrakisphosphate which has shown
activity as a phosphatase inhibitor.26
DYKAT reactions of conduritol B tetracarboxylates provide
ready access to a variety of enantiomerically pure conduritol B
derivatives. Chemoselective mono- and disubstitution may be
achieved, depending upon the steric demands of the nucleophile.
This strategy allows access to either enantiomer of the product,
depending only on the choice of chiral ligand. The utility of these
intermediates becomes immediately apparent in that the previously
reported racemic synthesis of cyclophellitol from the triacetate
analogue of 11b now becomes an asymmetric one.27 Herein, an
efficient synthesis of the natural product D-myo-inositol-1,4,5-
trisphosphate in five steps from racemic conduritol B tetratrichlo-
roethyl carbonate is reported.
loading was increased to 5 mol %, the reaction proceeded to give
a 61% yield (95% ee) in a process that is, at least in part, a
DYKAT. When dibenzylamine was used, a DYKAT was ob-
served since complete conversion to 12b in 89% yield (95% ee)
could be achieved.
While only the monosubstituted products were observed in the
asymmetric alkylation of conduritol B tetracarbonates with carbon
and nitrogen nucleophiles, disubstitution was the major process
when carboxylate nucleophiles were employed under conditions
for DYKAT reactions. Treatment of 9 with π-allylpalladium
chloride dimer and either enantiomer of ligand 3 led to nearly
complete conversion to the 1,4-disubstituted product 13a in >99%
enantiomeric excess. Pivalate also was an effective nucleophile,
giving a good yield 64% (>99% ee) of the 1,4-disubstituted
product 13b, along with a significant amount (17%, 91% ee) of
the 1,2-disubstituted product, presumably because of the steric
bulk of the pivalate and how it interacts with the walls of the
chiral pocket. The enantiomeric series is easily accessed simply
by switching the chirality of the ligand. Thus, dicarbonate 10 gave
the mirror image product 14 in 80% yield and >99% ee with the
S,S enantiomer of ligand 3. With oxygen nucleophiles, therefore,
Acknowledgment. We thank the National Science Foundation and
the National Institutes of Health General Medical Sciences Institute for
their generous support of our programs. D.E.P. was supported in part by
a fellowship from Roche Biosciences. E.J.H. was supported in part by a
NIH postdoctoral fellowship and the Job & Gertrud Tamaki Foundation.
Mass spectra were provided by the Mass Spectrometry Facility at the
University of California-San Francisco supported by the NIH Division
of Research Resources.
Supporting Information Available: Experimental details (PDF). This
material is available free of charge via the Internet at http://pubs.acs.org.
JA992960V
(22) For a review of the biological activity of I-1,4,5P3, see: Berrige, M.
J.; Irvine R. F. Nature 1989, 341, 197.
a DYKAT can be achieved, and racemic starting material can be
selectively converted into chiral enantiomerically pure disubsti-
tuted products. This reaction constitutes a deracemization of
racemic conduritol B tetracarboxylate with the added benefit of
differentiation of the esters. For example, treatment of 14 with
zinc in THF containing acetic acid gave a 77% yield of diol
dibenzoate 15.
(23) For a review on the synthesis of inositol phosphates, see: (a) Potter,
B. V. L.; Lampe, D. Angew. Chem., Int. Ed. Engl. 1995, 34, 1933 and
references therein. (b) Clive, D. L. J.; He, X.; Postema, M. H. D.; Mashimbye,
M. J. Tetrahedron Lett. 1998, 39, 4231. (c) Clive, D. L. J.; He, X.; Postema,
M. H. D.; Mashimbye, M. J. J. Org. Chem. 1999, 64, 4397.
(24) Meek, J. L.; Davidson, F.; Hobbs, F. W., Jr. J. Am. Chem. Soc. 1988,
110, 2317.
(25) Ozaki, S.; Kondo, Y.; Shiotani, N.; Ogasawara, T.; Watanabe, Y. J.
Chem. Soc., Perkin Trans. 1 1992, 729.
(26) (a) Mills, S. J.; Potter, B. V. L. J. Chem. Soc., Perkin Trans. 1 1997,
1279. (b) Mills, S. J.; Safrany, S. T.; Wilcox, R. A.; Nahorski, S. R.; Potter,
B. V. L. Bioorg. Med. Chem. Lett. 1993, 3, 1505. (c) Hirata, N.; Narumoto,
N.; Watanabe, T.; Kanematsu, T.; Koga, T.; Ozaki, S. Mol. Pharmacol. 1994,
45, 271. (d) Kozikowski, A. P.; Fauq, A. H.; Wilcox, R. A.; Nahorski, S. R.
Bioorg. Med. Chem. Lett. 1993, 5, 1295. (e) Chen, J.; Dorman, G.; Prestwich,
G. D. J. Org. Chem. 1996, 61, 393.
The efficiency of the deracemization induced us to consider a
synthesis of D-myo-inositol-1,4,5-trisphosphate (1,4,5-IP3) from
achiral starting materials and requiring no resolution. (1,4,5-IP3)
16 has been recognized as a second messenger to mobilize
intracellular calcium ions.22 When cell surface receptors are
activated, 1,4,5-IP3 is released inside the cell and binds to receptors
on the endoplasmic reticulum, stimulating the release of calcium
(27) Barton, D. H. R.; Dalko, P.; Gero, S. D. Tetrahedron Lett. 1991, 32,
2 471.