Asymmetric aldol reactions on a soluble polymeric support

Article abstract of DOI:10.1021/ol991383c

(Formula presented) The combinatorial synthesis of small, nonpeptidic compounds is of increasing interest in current medicinal chemistry. To meet this demand, efficient entries, preferentially on polymeric supports, to pharmacologically interesting classes of compounds such as polyketides are necessary. Therefore, we have developed a synthetic protocol allowing for the asymmetric synthesis of diketides on the soluble support MeOPEG-5000. The strategy employed mainly allows for repeated aldolizations, thus providing access to functionalized polyketides of varying degree of oligomerization.

Full text of DOI:10.1021/ol991383c

ORGANIC
LETTERS
2000
Vol. 2, No. 4
531-533
Asymmetric Aldol Reactions on a
Soluble Polymeric Support
M. Reggelin,* V. Brenig, and C. Zur
Institut fu¨r Organische Chemie, Marie Curie Strasse 11, UniVersita¨t Frankfurt/Main,
60439 Frankfurt/Main, Germany
re@cortex.chemie.uni-mainz.de
Received December 21, 1999
ABSTRACT
The combinatorial synthesis of small, nonpeptidic compounds is of increasing interest in current medicinal chemistry. To meet this demand,
efficient entries, preferentially on polymeric supports, to pharmacologically interesting classes of compounds such as polyketides are necessary.
Therefore, we have developed a synthetic protocol allowing for the asymmetric synthesis of diketides on the soluble support MeOPEG-5000.
The strategy employed mainly allows for repeated aldolizations, thus providing access to functionalized polyketides of varying degree of
oligomerization.
In 1996, as part of an endeavor to prepare polyketide
libraries, we described the first asymmetric aldol reaction
on a solid support using a resin-bound aldehyde as a starter
unit.1 Our approach was based on Evans’ chiral enol
borinates2 and on the reduction of Weinreb amides3 as the
key transformation to regenerate the aldehyde functionality.
Although feasable in principle, we encountered some dif-
ficulties originating from the absence of a suitable linker and
the incomplete and sometimes sluggish reduction of the
polymer-bound Weinreb amide. In 1998 we introduced a new
silicon-based linker and a much more reliable, albeit longer,
protocol to regenerate the aldehyde involving thioester 1 as
key intermediate.4 Using this methodology, we were able to
synthesize chiral, nonracemic δ-lactones and a number of
O-silylated di- and triketides.
Despite this success, some severe drawbacks related to
the insoluble polymeric matrix became more and more
obvious:5(1) The heterogeneous reaction conditions entail
reduced reaction rates. (2) Both the necessity for swelling
and the chemical structure of the polymer are often incom-
patible with reaction temperature and reagents.6(3) The
insoluble polymer hampers routine reaction monitoring.
(1) Reggelin, M; Brenig, V. Tetrahedron Lett. 1996, 37, 6851.
(2) Evans, D. A.; Bartroli, J. A.; Shih, T. L. J. Am. Chem. Soc. 1981,
103, 2127.
(3) Nahm, S.; Weinreb, S. M. Tetrahedron Lett. 1981, 22, 3815.
(4) Reggelin, M; Brenig, V.; Welcker, R. Tetrahedron Lett. 1998, 39,
4801.
(5) Reggelin, M. Nachr. Chem. Tech. Lab. 1997, 45, 1002.
10.1021/ol991383c CCC: $19.00 © 2000 American Chemical Society
Published on Web 01/29/2000

Because of these disadvantages, we often had to use
tremendous excesses of reagents (50 equiv of TIPSOTf is
needed to drive a silyl protection of the â-hydroxy group of
the diketide product to completion!) and unacceptably long
reaction times. Reactions involving heterogeneous catalysis
(e.g., hydrogenations) or reactions producing insoluble
byproducts during workup (e.g., DIBAH reductions) pro-
ceeded, if at all, only very unsatisfactorily. Thus, the polymer
imposes restrictions on the choice of reagents and reaction
conditions and renders development times high. Therefore,
we decided to use a soluble polymer that can be precipitated
in a solvent different from the one used in the reaction, which
combines the best of two worlds.5,7 The favorable properties
of solution phase chemistry, particularly the higher reaction
rates,8 and the ease of reaction monitoring are combined with
the operative advantages (easy separation of excess reagents
by simple filtration of the reaction mixture) of an insoluble
support.
In the 1H NMR, even without presaturation of the polymer-
related intense resonances around 3.65 ppm, the signal of
the terminal methoxy group appears as a baseline-separated
singlet (Figure 1). Its integral serves as an internal standard
to calculate the polymer loading.
For our purposes we found it advantageous to use the
popular monomethylated poly(ethylene glycol), MeOPEG,
as the polymer of choice, although there are other alternatives
such as non-cross-linked polystyrenes evolving, leading to
interesting alternatives.9 Poly(ethylene glycol) and its monom-
ethylated derivative, introduced by Bayer and Mutter in the
early 1970s10 as a polymeric support for peptide synthesis,
is commercially available at different degrees of polymeri-
zation (up to 20 000 g/mol), are soluble in THF and
dichloromethane and can be precipitated by addition of
diethyl ether or tert-butyl methyl ether.
1
Figure 1. Part of the H NMR of aldehyde 3 (400 MHz, 2.5 mg
in 0.4 mL of CDCl3). The terminal methoxy group of the polymer
serves as an internal standard.
Polymeric aldehyde 3 was now dried by azeotropic
distillation with toluene (100 mL/mmol),12 dissolved in
dichloromethane (80 mL/mmol), and cooled to -60 °C. To
this solution were added three different enol borinates and
their respective enantiomers from a separate flask at -78
°C using PTFE tubing (Scheme 2, Table 1). The enol
To attach the starter aldehyde onto the soluble support,
we reacted MeOPEG-5000 (MW ) 5000 g/mol) with 1,4-
bis(chloromethyl)benzene following literature procedures
(Scheme 1).11 After precipitation of the polymeric aldehyde
Scheme 1. Preparation of the MeOPEG-Bound Aldehydea
Scheme 2. Asymmetric Aldol Reactions on MeOPEGa
a Reagents and conditions: (a) 6 equiv of NaH, 2 equiv of NaI,
12 equiv of 1,4-bis(dichloromethyl)benzene, THF, rt, 5 d; (b) 6
equiv of p-hydroxybenzaldehyde, sodium salt, DMF, 50 °C, 24 h.
a Reagents and conditions: (a) X ) Cl; 1.01 equiv of n-BuLi,
THF, -78 °C to rt, 1 h; (b) X ) OH; 1.03 equiv of 4, 2.5 equiv of
Et3N, 1.03 equiv of Piv-Cl, 25 °C, 2 h, 1.0 equiv of 5, 1.13 equiv
of LiCl, rt, 7h; (c) 1.15 equiv of n-Bu2BOTf (2.3 equiv, R ) CH2-
ind), 3.0 equiv of Et3N, CH2Cl2, 0 °C, 30 min; (d) 0.16 equiv of 3,
CH2Cl2, -78 to -20 °C, 16 h. For R decoding, see Table 1.
in diethyl ether (300 mL/mmol) and recrystallization from
cold ethanol, we obtained a polymer loaded with 90% of
the theoretical maximum load of 0.2 mmol/g.
(6) These problems have also been observed by others. See, for
example: Harris, J. M.; Liu, Y.; Chai, S.; Andrews, M. D.; Vederas, J. C.
J. Org. Chem. 1998, 63, 2407.
(7) For recent reviews on the use of soluble polymers in other applications
of organic synthesis, see: (a) Wentworth, P., Jr.; Janda, K. D. Chem.
Commun. 1999, 1917. (b) Gravert, D. J.; Janda, K. D. Chem. ReV. 1997,
97, 489. (c) Geckeler, K. E. AdV. Polym. Sci. 1995, 121, 31.
borinates were prepared from the corresponding acylated
oxazolidinones using standard procedures.2 Only in the case
of the novel 3-(3-indolyl)propyl imides 6c and ent-6c was it
532
Org. Lett., Vol. 2, No. 4, 2000

Scheme 3. Cleavage by Heterogeneous Hydrogenationa
Table 1. Asymmetric Aldol Reactions on MeOPEG 5000
educt
aldol adducts
R
yield [%]
6a
ent-6a
6b
ent-6b
6ca
ent-6ca
8a
ent-8a
8b
ent-8b
8c
ent-8c
CH3
CH3
CH2Ph
CH2Ph
CH2-(3-ind)
CH2-(3-ind)
56
83
76
88
57
71
a Two equivalents of Bu2BOTf were used.
a Reagents and conditions: (a) H2 (1 atm), Pd black, MeOH, rt,
48 h.
necessary to use an additional equivalent of Bu2BOTf to
temporarily protect the indole NH.
the supernatant liquid delivered the product as a pure isomer
in 85% yield.
After precipitation of the polymeric aldol adducts as
described above, their diastereomeric purity (g97% ds) was
confirmed by 1H NMR spectroscopy. Following this protocol
we were able to synthesize three pairs of enantiomeric
polymer-bound diketides as pure isomers in unoptimized
yields between 57% and 88%.
To validate the assumption that the soluble polymeric
support should permit the removal of the benzyl linker by
heterogeneous hydrogenation, we treated ent-8a with hy-
drogen (1 atm) over Pd black for 48 h in methanol at room
temperature (Scheme 3).
In summary we have shown that asymmetric aldol reac-
tions employing chiral enol borinates are feasable on the
soluble polymeric support MeOPEG-5000 without deteriora-
tion of stereoselectivity. 3-Indolylmethyl-substituted diketides
8c and ent-8c have been prepared for the first time, and their
potential as open chain scaffolds in combinatorial di- and
triketide libraries is of current interest in our laboratories.
To our delight and in contrast to the results of a similar
experiment using an unsoluble support,13 cleavage of the
desired aldol product 9 occurred smoothly. After filtration
and precipitation of the unloaded polymer, evaporation of
(8) Kinetic studies on the aminolysis of glycine p-nitrophenyl esters with
polymeric or monomeric glycine esters have shown that the reaction rate
is not significantly affected by the presence of the polymer. Bayer, E.;
Mutter, M.; Uhmann, R.; Polster, J.; Mauser, H. J. Am. Chem. Soc. 1974,
96, 7333.
(9) First application of non-cross-linked polystyrene in organic synthe-
sis: (a) Shemyakin, M. M.; Ovchinnikov, Y. A.; Kinyushkin, A. A.
Tetrahedron Lett. 1965, 6, 2323. (b) Peptide synthesis: Narita, M. Bull.
Chem. Soc. Jpn. 1978, 51, 1477. (c) Recent application in prostanoid
chemistry: Lee, K. J.; Angulo, A.; Ghazal, P.; Janda, K. D. Org. Lett. 1999,
1, 1859.
(10) (a) Mutter, M.; Hagenmaier, H.; Bayer, E. Angew. Chem., Int. Ed.
Engl. 1971, 10, 811. (b) Bayer, E.; Mutter, M. Nature 1972, 237, 512. (c)
Small molecule library synthesis on MeOPEG: Han, H.; Wolfe, M. M.;
Brenner, S.; Janda, K. D. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 6419.
(11) Douglas, S. P.; Whitfield, D. M.; Krepinsky, J. J. J. Am. Chem.
Soc. 1995, 117, 2116.
(12) This treatment of the polymer prior to the aldol reactions proved to
be an unavoidable necessity. As expected the polyether structure retains a
considerable amount of water which has to be removed thoroughly.
(13) Even the application of a Pd(OAc)2 solution prior to hydrogenation
(Schlattner, J. M.; Mazur, R. H.; Goodmonson, O. Tetrahedron Lett. 1977,
33, 2851) was not very successful. Reggelin, M.; Brenig,V., unpublished
results.
Furthermore, it has been shown that catalytic hydrogena-
tion using insoluble catalysts is compatible with the poly-
meric nature of the substrate, thus considerably widening
the scope of applicable reagents in polymer-supported
organic synthesis. Encouraged by these results we are
currently exploring the possibility of returning to our initial
concept for the synthesis of combinatorial libraries of
decorated polyketides employing reactions which proved to
be difficult to perform on an insoluble support.1,3,4
Acknowledgment. We gratefully acknowledge the Solvay
Pharmaceuticals (Hannover, Germany) for their generous
support of this work. Furthermore, we thank Prof. Dr. C.
Griesinger for continuous support.
OL991383C
Org. Lett., Vol. 2, No. 4, 2000
533