Asymmetric Synthesis of Vitamin D3 Analogues: Organocatalytic
Desymmetrization Approach toward the A‑Ring Precursor of
Haifeng Wang, Linjie Yan, Yan Wu, Yipei Lu, and Fener Chen*
Department of Chemistry, Fudan University, Shanghai 200433, P. R. China *S Supporting Information
ABSTRACT: A novel asymmetric synthesis has been developed for the construction of the A-ring of a chiral precursor to calcifediol. The highlights of this synthesis include (i) the introduction of the stereochemistry at the C5-position of the A-ring through the organocatalytic enantioselective desymmetrization of a prochiral cyclic anhydride using a bifunctional urea catalyst and (ii) the introduction of the exo-cyclic (Z)-dienol side chain by a tandem Claisen rearrangement/sulfoxide thermolysis of an allylic alcohol.
Calcifediol (25-hydroxyvitamin D3, 1) is a biologicallyactive metabolite of vitamin D3 and represents the major circulating form of vitamin D3 present in human plasma. 1 The medicinal importance of 1 for the treatment of various metabolic diseases as well as renal failure, rickets, and osteoporosis has attracted considerable interest from researchers working in a variety of different fields, including synthetic organic chemistry.2 The chiral dienol 2 is a precursor of the Aring building block required for the preparation of 1 (Scheme 1). Several studies have been reported to date describing the development of elegant synthetic processes for the synthesis of 2, including Lythgoe’s partial approach starting from vitamin D2 or vitamin D3. Several stereoselective total syntheses have also been reported on the basis of Lythgoe’s chiral pool approach and William’s catalytic asymmetric strategy.3 Despite significant progress in this area, the development of an efficient and practical process for the preparation of 2 has not yet been achieved and is still highly desired. Herein, we report a novel catalytic asymmetric synthesis of 2 from commercially available cyclic anhydride 3 using an organocatalytic anhydride desymmetrization strategy.
Our retrosynthetic analysis of 2 is depicted in Scheme 2. It was envisaged that the chiral dienol 2 could be assembled from the allylic alcohol 14 through a one-pot tandem Claisen [3,3]sigmatropic rearrangement/sulfoxide thermolysis reaction followed by the reduction of the resulting ester. Compound 14 could be synthesized from 12 via the E2 elimination and the reduction of the benzyl ester to give the required allylic alcohol 14. In turn, compound 12 could be prepared from hemiester 4 via a series of transformations. Finally, it was envisaged that the chiral hemiester 4 could be synthesized by the organocatalytic enantioselective alcoholysis of the meso-cyclic anhydride 3.
To allow for the introduction of the required stereochemical information into the key intermediate 4, we directed our
Received: September 29, 2015
Scheme 1. Structures of 25-Hydroxyvitamin D3 (1) and the
A-Ring Allyl Alcohol 2
Scheme 2. Retrosynthetic Analysis
Letter pubs.acs.org/OrgLett © XXXX American Chemical Society A DOI: 10.1021/acs.orglett.5b02813
Org. Lett. XXXX, XXX, XXX−XXX research efforts toward the development of an efficient process for the asymmetric alcoholysis of the prochiral cyclic anhydride 3. In this way, we investigated the synthesis of (1S,6R)hemiester 4 from 3 using a series of chiral bifunctional urea catalysts I−V, which were developed by our group (Table 1).4
The result of our initial experiment showed that the exposure of 3 to benzyl alcohol in the presence of catalyst I (5 mol %) in
MTBE (0.1 M) at room temperature gave the desired hemiester 4 in 92% yield with 69% ee (Table 1, entries 1).
The subsequent screening of a wide range of catalysts (i.e., catalysts II−V) and solvents, including dichloromethane, toluene, acetonitrile, and tetrahydrofuran, failed to afford any improvement in the enantioselectivity of the desymmetrization process (Table 1, entries 2−9). It is noteworthy that the reaction showed a very strong dilution effect. For example, decreasing the concentration of the anhydride from 0.1 to 0.025 mol/L in the presence of catalyst I (5 mol %) led to a significant increase in the enantioselectivity for the formation of the desired product 4 inform 69 to 88% ee (Table 1, entries 1,10−11). Increasing the loading of the catalyst to 10 mol % led to a significant increase in the catalytic activity, as well as the enantioselectivity (90% ee, entries 12). Pleasingly, the enantioselectivity of 4 (90% ee) was further increased to 96% ee in 87% yield by a single recrystallization from methyl tertbutyl ether.
The iodolactonization of 4 under the conditions established by Van Tarnelen et al. (i.e., NaHCO3, I2/KI, rt) furnished the corresponding iodolactone 5 in 86% yield, although a long reaction time (3 days) was needed.5 Pleasingly, the treatment of 4 with NIS in dichloromethane at room temperature gave 5 in 91% yield following a much shorter reaction time (only 1 h) (Scheme 3). The absolute configuration of 5 was determined by X-ray crystallographic analysis.6 The subsequent reductive deiodination of 5 was performed with 10% Pd/C in methanol under an atmosphere of H2 in the presence of sodium acetate to give the corresponding lactone 7. Bromolactone 6 was prepared in 88% yield from 4 using NBS instead of NIS under conditions similar to those used for the formation of 5.
However, the reductive debromination of 6 failed to provide 7 under the same hydrogenation conditions as those used for 5.
The subsequent treatment of 7 with MeOH in the presence of Na2CO3 at room temperature gave diester 8, which was immediately protected with TBSCl in the presence of imidazole in DMF to give tert-butyldimethylsilyl ether 9 in 80% yield over the two steps (Scheme 4). The hydrolysis of 9 with lithium hydroxide (1.8 equiv) in a 2 mL/2 mL/2 mL mixture of THF/
MeOH/H2O at 40 °C led to the formation of a 7:1 (m/m) mixture of hemiester 10 and the corresponding dicarboxylic acid, which was purified by silica gel chromatography to afford pure 10 in 71% yield.