Synthesis of polyketides via diastereoselective acetalization

Article abstract of DOI:10.1021/ol027538p

(Matrix presented) Diastereoselective acetalization of pseudo-C ₂-symmetric 1,3,5-triol systems is a general strategy for the rapid generation of polyketides. The oxidative acetalization reaction shown above was studied under both kinetic and t

Full text of DOI:10.1021/ol027538p

ORGANIC
LETTERS
2003
Vol. 5, No. 7
1027-1030
Synthesis of Polyketides via
Diastereoselective Acetalization
Jennifer N. Shepherd*,†,‡ and David C. Myles*‡,§
Department of Chemistry, Gonzaga UniVersity, Spokane, Washington 99258,
Department of Chemistry and Biochemistry, UCLA, Los Angeles, California 90095,
and Kosan Biosciences, Hayward, California 94545
shepherd@gonzaga.edu; myles@kosan.com
Received December 23, 2002
ABSTRACT
Diastereoselective acetalization of pseudo-C -symmetric 1,3,5-triol systems is a general strategy for the rapid generation of polyketides. The
2
oxidative acetalization reaction shown above was studied under both kinetic and thermodynamic conditions, using synthetic 1,3,5-triol units.
In addition, all possible stereochemical variants of a 1,3,5-triol system were prepared from the corresponding acetal to expand the synthetic
versatility of this method.
Polyketide natural products are a structurally complex class
of compounds. Many of these compounds show important
biological activity. For example, the recently synthesized
polyene macrolides dermostatin A (1) and B1 (2) exhibit
potent antifungal activity, while the polypropionate disco-
dermolide2 (3) exhibits potent antimitotic activity.
these compounds poses a synthetic challenge. There are many
reiterative strategies designed for 1,3-polyol fragment syn-
thesis. One common theme is based on the stereoselective
functionalization of allylic and homoallylic alcohols. For
example, the Sharpless asymmetric epoxidation of allylic
alcohols is often used for preparation of syn- or anti-1,3-
polyols.3 However, a typical sequence of this method requires
about 8 linear steps. Homoallylic alcohols may be function-
alized with use of cyclic iodocarbonates.4 This strategy
requires a minimum of 4 steps per iteration, but it is limited
to the formation of only syn-1,3-diols. Other strategies
include the [1,2]-Wittig rearrangement of â-alkoxyalkyl allyl
ethers5 and the selective reduction of acyclic â-hydroxy
ketones.6
We have reported a highly general approach to the
synthesis of polyacetate and polypropionate structures based
(1) Sinz, C. J.; Rychnovsky, S. D. Tetrahedron 2002, 58, 6561-6576.
(2) Chen, J. G.; Horowitz, S. B. Cancer Res 2002, 7, 1935-8.
(3) (a) Gao, Y.; Hanson, R. M.; Klunder, J. M.; Ko, S. Y.; Masamune,
H.; Sharpless, K. B. J. Am. Chem. Soc. 1987, 109, 5765. (b) Finan, J. M.;
Kishi, Y. Tetrahedron Lett. 1982, 23, 2719.
(4) (a) Lipshutz, B. H.; Kozlowski, J. A. J. Org. Chem. 1984, 49, 1149.
(b) Lipshutz, B. H.; Barton, J. C. J. Org. Chem. 1988, 53, 4495.
Efficient synthetic strategies for the preparation of
polyketides are necessary; however, the 1,3-polyol unit of
† Gonzaga University.
‡ UCLA.
§ Kosan Biosciences.
10.1021/ol027538p CCC: $25.00 © 2003 American Chemical Society
Published on Web 03/04/2003

on the oxidative acetalization of p-methoxybenzyl (PMB)
ethers of pseudo-C2-symmetric 1,3,5-triols (Figure 1).7
The irreversibility of this reaction was tested by resubmit-
ting acetal mixtures (5 and 6, using both the 1,3-polyol and
the polypropionate models) to the same DDQ reaction
conditions. This led only to over-oxidation products. No
equilibration of the diastereomers was observable. The
similarity of the experimental ∆∆Gq values to the ∆E values
from MM2 calculations10 suggests that the transition state
of cyclization is late. Because we have hypothesized that
the DDQ-mediated oxidative acetalization of C-3 PMB ethers
of pseudo-C2-symmetric 1,3,5-triols is a kinetically controlled
process, we also wanted to investigate the thermodynamic
selectivity of PMP-acetal formation.
A simplified polyol model was synthesized for these
thermodynamic studies. The polyol model 9 (Scheme 1) was
Scheme 1a
Figure 1. Oxidative acetalization.
Pseudo-C2-symmetric models of the 1,3-polyol and polypro-
pionate systems were previously prepared to study the
oxidative acetalization process.7 The oxidative acetalization
was achieved via anhydrous DDQ in CH2Cl2 at room
temperature. The 1,3-polyol model prepared contained R )
(S)-CH(OH)CO2Et and R′ ) H. This compound gave a
selectivity of 95:5 (5:6) in the oxidative cyclization reaction.
The polypropionate model contained R ) H and R′ ) CH3.
A selectivity of 75:25 (5:6) was achieved in this acetalization.
A low-temperature study showed increasing selectivity in
the oxidative acetalization process with lower temperature.
For example, the 1,3-polyol model gave a diastereoselectivity
of >99:1 at -30 °C and the polypropionate model gave an
improved diastereoselectivity of 82:18 at -30 °C. With use
of anhydrous DDQ conditions, the oxidative acetalization
process is presumed to be irreversible and therefore under
kinetic control.
a Reagents and conditions: (a) (i) O3, MeOH/CH2Cl2, Sudan Red
III, pyr, -78 °C; (ii) DMS.(b) (i) [(+)-ipc]2allylborane, THF, -100
°C; (ii) H2O2, NaOH. (c) (i) [(-)-ipc]2allylborane, THF, -100 °C;
(ii) H2O2, NaOH. (d) DDQ, rt, CH2Cl2, 4 Å molecular sieves.
prepared from the PMB-protected diene 7. Treatment of 7
with ozone, followed by homologation with B-(+)-allyldi-
isopinocampheylborane, using conditions developed by
Brown,8 gave compound 9. An oxidative acetalization
performed on 9, using standard kinetic conditions (DDQ,
CH2Cl2, rt), furnished acetals 11 and 12 (90:10, 77% yield,
ee >99%). MM2 calculations predicted an energy difference,
∆E ) 2.6 kcal/mol, between 11 and 12.9 These acetals were
more susceptible to over-oxidation than the previous two
examples. After only 5-10 min of exposure to the acetal-
ization conditions, 15-20% of the mass balance cor-
responded to a mixture of p-methoxybenzoate regioisomers.
The enantiomeric 1,3,5-triol was also prepared with B-(-)-
allyldiisopinocampheylborane.
(5) Schreiber, S. L.; Goulet, M. T. Tetrahedron Lett. 1987, 28, 1043.
(6) (a) Evans, D. A.; Hoveyda, A. H. J. Am. Chem. Soc. 1990, 112, 6447.
(b) Evans, D. A.; Chapman, K. T. Tetrahedron Lett. 1986, 27, 5939. (c)
Evans, D. A.; Chapman, K. T.; Carreira, E. M. J. Am. Chem. Soc. 1988,
110, 3560. (d) Evans, D. A.; Hoveyda, A. H. J. Org. Chem. 1990, 55, 5190.
(7) Shepherd, J. N.; Na, J.; Myles, D. C. J. Org. Chem. 1997, 62, 4558.
(8) Ramachandran, P. V.; Chen, G.; Brown, H. C. Tetrahedron Lett. 1997,
38, 2417.
(9) Molecular modeling was carried out on a Silicon Graphics workstation
using MacroModel V 5.0 (Richards, N. G. J.; Guida, W. C.; Liskamp. R.;
Lipton, M.; Caulfield, C.; Chang, G.; Hendrickson, T.; Still, W. C. J.
Comput. Chem. 1990, 11, 440). For compounds 7 and 8, Monte Carlo
conformational searches were carried out using Batchmin (500 MC steps,
MM2, energy window ) 25 kJ/mol, 2000 maximum iterations). Calculated
energies are global minima.
(10) Smith, M.; Rammler, D. H.; Goldberg, I. H.; Khorana, H. G. J.
Am. Chem. Soc. 1962, 84, 430.
1028
Org. Lett., Vol. 5, No. 7, 2003

To determine if the ratio of products obtained by direct
oxidation was the kinetic or thermodynamic ratio, the acetal
mixture (11:12) was allowed to equilibrate for 2 h in benzene,
using 20% TFA and 4 Å molecular sieves. The ratio of
acetals 11 and 12 changed from 90:10 to >99:1. No further
changes in the >99:1 ratio were observed after 24 h.
To investigate further the thermodynamic control in the
formation of 11 and 12, we sought to compare the product
ratios from these equilibration experiments with those
obtained from an acid-catalyzed acetalization of the 1,3,5-
triol with PMP-aldehyde.
The 1,3,5-triol (15, eq 1) was prepared by basic hydrolysis
(K2CO3, MeOH) of the p-methoxybenzoates (obtained from
DDQ over-oxidation of 11 and 12). The thermodynamically
controlled acetalization (p-anisaldehyde, TFA, 4 Å molecular
sieves in benzene)10 of 15 gave acetals 11 and 12, again in
a >99:1 ratio.
perature and by cooling to -30 °C, using the irreversible
DDQ conditions.
The synthetic utility of simple polypropionate intermedi-
ates such as 5 (eq 2) is limited. However, better diastereo-
selectivity could perhaps be achieved with other substrates
where R ) alkyl instead of H. This extra substitution may
stabilize the chairlike transition state A (see transition state
model, Figure 1) and further destabilize transition state B.
The optimization of the diastereoselective acetalization with
polypropionate systems must be investigated.
The diastereoselective acetalization examples presented for
1,3-polyol synthesis gave selectively protected 1,3,5-triol
units with anti-syn stereochemistry. To generalize our
approach further, we prepared all stereochemical variants of
the 1,3,5-triol unit. The syn-syn 1,3,5-triol 17 was prepared
via a Mitsonobu inversion of the alcohol 11 (Scheme 2).12
Scheme 2a
Finally, a low-temperature experiment was performed.
Compound 9 was subjected to irreversible DDQ conditions
at -30 °C. A >99:1 ratio of 11:12 was obtained. From these
results, we have concluded that the formation of the syn-
acetal 11 is favored under both kinetic and thermodynamic
conditions. However, the diastereoselectivity is the highest
under thermodynamic control at room temperature or kinetic
control at -30 °C.
a Reagents and conditions: (a) (i) DEAD, PPh3, benzoic acid,
THF; (ii) K2CO3, MeOH. (b) TIPSCI, NaH, TBAI, THF. (c)
DIBAL-H, toluene, 0 °C. (d) (i) DEAD, PPh3, benzoic acid, THF;
(ii) K2CO3, MeOH.
The same two thermodynamic experiments were per-
formed with the polypropionate model system (4, R ) H
and R′ ) CH3, Figure 1). The kinetic mixture of acetal
products 5 and 6 (R ) H and R′ ) CH3) was subjected to
equilibration conditions. After exposure to TFA for 2 h, the
acetal ratio of 5 to 6 changed from 75:25 to 82:18. No further
fluctuations in this ratio were observed over 48 h.
For the second thermodynamic-based experiment, the
polypropionate-derived 1,3,5-triol 16 was prepared by depro-
tection of 4 (CAN, CH3CN, and H2O).11 Acetalization of 16
under thermodynamic conditions (p-anisaldehyde, TFA, 4
Å molecular sieves in benzene) gave the same 82:18 ratio
of 5 and 6 as in the previous experiment (eq 2). From these
The anti-anti triol 18 was prepared in a three-step sequence,
also starting from 11. First, 11 was protected with a
triisopropylsilyl group. A regioselective reduction of the
acetal was then achieved with DIBAL-H.13 Finally, the anti-
anti triol 18 was generated by a Mitsonobu inversion.
Diastereoselective acetalization of pseudo-C2-symmetric
chains is a powerful strategy for the rapid synthesis of
stereochemically complex structures. The results presented
here demonstrate that this strategy can be used to differentiate
the end-groups in 1,3,5-triol systems having pseudo-C2-
symmetry. These reactions can be used to synthesize highly
disymmetric precursors for the preparation of complex
polyoxygenated natural products. The generality of this
approach makes it highly applicable for the synthesis of all
stereochemical variants of the 1,3,5-triol unit. This methodol-
ogy can now be utilized in the context of polyacetate and
polypropionate synthesis. The development and application
(11) Johansson, R.; Samuelsson, J. J. Chem. Soc., Chem. Commun. 1984,
201.
(12) (a) Mitsonobu, O. Synthesis 1981, 1. (b) Schmidt, U.; Utz, R. Angew.
Chem., Int. Ed. Engl. 1984, 23, 723. (c) Hsu, C.-T.; Wang, N.-Y.; Latimer,
L. H.; Sih, C. J. J. Am. Chem. Soc. 1983, 105, 593.
(13) Johansson, R.; Samuelsson, J. J. Chem. Soc., Chem. Commun. 1984,
201.
results, we propose that the polypropionate-derived acetal 5
is also favored under both kinetic and thermodynamic
control. The best selectivity (82:18) can be achieved both
by using reversible equilibration conditions at room tem-
Org. Lett., Vol. 5, No. 7, 2003
1029

of this approach toward the synthesis of long-chain polyols
will be forthcoming.
Year fellowship. This work was supported in part by the
Alfred P. Sloan Foundation.
Supporting Information Available: Experimental pro-
cedures and analytical data for new compounds. This material
is available free of charge via the Internet at http://pubs.acs.org.
Acknowledgment. The authors would like to acknowl-
edge Michele Lee and Jim Na for assistance at various stages
of this research. J. N. Shepherd acknowledges the Department
of Chemistry and Biochemistry at UCLA for a Dissertation
OL027538P
1030
Org. Lett., Vol. 5, No. 7, 2003