Enantioselective Total Synthesis of (+)-Largazole, a Potent Inhibitor of Histone Deacetylase

Abstract

An enantioselective total synthesis of cytotoxic natural product, (+)-largazole (1) is described. It is a potent histone deacetylase inhibitor. Our synthesis is convergent and involves the assembly of thiazole 3-derived carboxylic acid with amino ester 4 followed by cycloamidation of the corresponding amino acid. The synthesis features an efficient cross metathesis, an enzymatic kinetic resolution of a β-hydroxy ester, a selective removal of a Boc-protecting group, a HATU/HOAt-promoted cycloamidation reaction, and synthetic manipulations to a sensitive thioester functional group.

In January 2008, Luesch and co-workers reported the isolation of largazole, a novel 16-membered depsipeptide from Floridian marine cyanobacterium Symploca sp.¹ Largazole’s structure was elucidated by extensive NMR studies and through chemical degradation. It has shown impressive growth inhibitory activity of transformed mammary epithelial cells (MDA-MB-231) in a dose dependent manner with a GI₅₀ value of 7.7 nM. In addition, it has shown excellent selectivity over nontransformed murine mammary epithelial cells (NMuMG) with a GI₅₀ of 122 nM. More recently, Luesch, Hong, and co-workers reported the first total synthesis of largazole. Their synthesis featured a macrocyclization at C6 and a late stage addition of the thioester using cross metathesis. Subsequently, they determined that histone deacetylase (HDAC) is the molecular target for largazole.² This is very significant as HDAC inhibitors are emerging as a new and exciting class of antineoplastic agents for the treatment of solid and nematological malignancies.³ Incidentally, a number of depsipeptides are undergoing clinical trials for treatment of various cancers.⁴ Largazole’s important biological activity, its selectivity for cancer cells, and its unique structural features led to considerable interest in its chemistry and biology. To establish structure-activity relationships and design novel structural variants, we sought a convergent route to largazole. Herein, we report an enantioselective synthesis of (+)-largazole.

As shown in Figure 1, our synthetic strategy involves a late stage cycloamidation of a sterically less demanding carboxylic acid and an amine to form the 16-membered ring from the corresponding amino acid derived from 2. Ester derivative 2 could be obtained by the formation of an amide bond between the acid arising from thiazole methyl ester 3 and the 4-derived amine. Our plan is to carry out the remainder of the synthesis with the sensitive thioester functional group attached. The synthesis of thiazole 3 can be achieved from a 5-derived thiazole acid and protected (R)-2-methyl cysteine 6.⁵ Amino ester 4 could be accessed by a cross metathesis reaction between thioester 8 and optically active allylic alcohol 7 followed by a Yamaguchi esterification with the appropriately protected L-valine. Alcohol 7 could be prepared by a lipase mediated kinetic resolution of racemic β-hydroxy ester.

Figure 1.

Retrosynthetic analysis of largazole

As shown in Scheme 1, our synthesis starts with the known azido amide 9⁶ which was treated with Lawesson’s reagent in THF for 12 h to provide the corresponding thioamide in 67% yield.⁷ The resulting thioamide was then reacted with ethyl bromopyruvate in refluxing ethanol for 1 h which provided thiazole 5 in 82% yield.⁸ Saponification of ester 5 with 1 M aqueous LiOH gave the acid. The resulting acid was then coupled with trityl- protected α-methyl cysteine⁹ under EDC/HOBt conditions in the presence of diisopropylethylamine to furnish amide 10 in 96% yield over 2 steps. The conversion to thiazole-thiazoline fragment 11 was achieved following a procedure reported by Kelly and co-workers.¹⁰ Accordingly, amide 10 was reacted with 3 equiv. of triphenylphosphine oxide and 1.5 equiv. of Tf₂O in CH₂Cl₂ at 0 °C for 10 min to provide ester 11 in 89% yield. The azide group in 11 was reduced using PPh₃ in refluxing methanol¹¹ to give the amine which was then exposed to Boc₂O to furnish fragment 3 in 95% yield over 2 steps.

Scheme 1.

Synthesis of segment 3

Optically active synthesis of β-hydroxy ester and its conversion to ester 15 is shown in Scheme 2. Racemic aldol product 12 was prepared by LDA deprotonation of t-butyl acetate followed by reaction of the resulting enolate with acrolein at −78 °C to provide 12 in 81% yield. The racemic alcohol was then exposed to lipase PS-30 in pentane in the presence of excess vinyl acetate at 23 °C for 12 h to provide enantio enriched alcohol 13 and acetate derivative 14 in 45% and 42% yields respectively. Selective removal of the acetate was carried out by exposure of 14 to potassium carbonate in methanol at −30 °C to afford optically active β-hydroxy ester 7 in high enantiomeric purity (93% ee). The enantiomeric excess was determined by formation of the corresponding Mosher ester of alcohol 15 followed by analysis of ¹⁹F NMR.¹² The kinetic resolution of β-hydroxy ester has provided a convenient access to optically active esters.¹³ For preparation of alcohol 15 we planned a cross metathesis of alcohol 7 and thioester 8. The requisite thioester was prepared by reaction of 3-butene thiol¹⁴ and octanoyl chloride in the presence of DMAP. A cross metathesis reaction of alcohol 7 and thioester 8 in the presence of 3 mol% Grubbs’ 2nd generation catalyst afforded E-olefin 15 exclusively in 67% yield.¹⁵

Scheme 2.

Synthesis of thio ester 15

The final assembly of the largazole fragment is shown in Scheme 3. N-Boc-valine 16 was subjected to esterification with alcohol 15 using Yamaguchi’s protocol.¹⁶ Accordingly, reaction of 16 with 2,4,6-trichlorobenzoyl chloride in the presence of diisopropylethylamine gave the anhydride. Reaction of the resulting anhydride with alcohol 15 and DMAP furnished thioester 4 in 91% yield. Selective deprotection of the Boc group in the presence of a t-butyl ester was carried out by exposure of 4 to 30% trifluoroacetic acid in CH₂Cl₂ at 0 °C for 20 min to provide amine 17. For assembly of the largazole subunits, saponification of methyl ester 3 was carried out with 1 M aqueous LiOH to give acid 18. Coupling of acid 18 with amine 17 was accomplished by using HATU and HOAt in the presence of diisopropylethylamine to furnish the requisite protected amino ester 2 in 66% yield. Formation of the 16-membered cycloamide was carried out in a two-step sequence involving: (1) exposure of 2 to trifluoroacetic acid at 23 ^oC for 3 h to remove both the Boc and the t-butyl groups; (2) treatment of the resulting amino acid with 2 equiv. of HATU and 2 equiv. of HOAt in the presence of diisopropylethylamine under dilute conditions to provide synthetic (+)-largazole (1) in 40% isolated yield (2 steps). The spectral data (¹H- and ¹³C-NMR) of synthetic (+)-largazole (1, [□]²³_D +24, c 0.13, MeOH; lit.¹ [□]²⁰_D +22, c 0.1, MeOH) is identical with that reported for the natural (+)-largazole.¹

Scheme 3.

Synthesis of Largazole

In summary, we have accomplished an enantioselective synthesis of (+)-largazole (1). The synthesis will provide a convenient access to a variety of largazole derivatives. Structural modifications are currently in progress.¹⁷