Catalytic Enantioselective Total Synthesis of (+)–Lycoperdic Acid

: A concise enantio-and stereocontrolled synthesis of (+)-lycoperdic acid is presented. The stereochemical control is based on iminium-catalyzed Mukaiyama–Michael reaction and enamine-catalyzed organocatalytic α -chlorination steps. The amino group was then introduced by azide displacement, affording the final stereochemistry of (+)-lycoperdic acid. Penultimate hydrogenation and saponification afforded pure (+)-lycoperdic acid in seven steps from a known silyloxyfuran. (+)-Lycoperdic acid ( 1 ) was isolated from a mushroom Lyco-perdon perlatum by Rhugenda-Banga et al . 1 It is an amino acid that shares structural similarities with both L-glutamic acid ( 2 )


Catalytic Enantioselective Total Synthesis of (+)-Lycoperdic Acid
H ABSTRACT: A concise enantio-and stereocontrolled synthesis of (+)-lycoperdic acid is presented.The stereochemical control is based on iminium-catalyzed Mukaiyama-Michael reaction and enamine-catalyzed organocatalytic α-chlorination steps.The amino group was then introduced by azide displacement, affording the final stereochemistry of (+)-lycoperdic acid.Penultimate hydrogenation and saponification afforded pure (+)-lycoperdic acid in seven steps from a known silyloxyfuran.

Synthesis
Legend: chiral pool

Figure 2. Sources of stereochemistry in the published routes
Our first challenge was to find the conditions for the desired Mukaiyama-Michael reaction with silyloxyfuran esters such as 4. The closest precedent was set by Pansare group who demonstrated that MacMillan's trimethyl imidazolidinone 5 would catalyse reactions between acrolein and silyloxyfurans with good enantioselectivities but with poor yields (Scheme 1). 18Our own previous experience with enantioselective Mukaiyama-Michael reactions with acrolein involving diphenylpyrrolidine (9) 19 and pyroglutamic-acid-derived pyrrolidine 20 catalysts suggested that systematic optimization of the catalyst might offer better results.
In contrast with the Pansare precedent, we wanted to avoid the use of any bulky ester groups, or even benzyl esters in the nucleophilic component to prevent any conflict with later operations along the route (Scheme 1).For example, benzyl esters were deemed unsuitable as they might require special precautions in the projected hydrogenation of the butenolide C=C bond.We thus selected the methyl ester 4 as the starting point for catalyst development.The work commenced by screening studies with typical iminium catalysts 9 and 13-15 (Scheme 2).These catalysts, unfortunately, gave only poor to moderate enantiomeric ratios.Nevertheless, comparison of differently substituted catalysts revealed potentially useful trends.Thus, catalysts with electron withdrawing substituents enhanced the enantioselectivities: with para-substituted diarylpyrrolidine catalysts, there was a rising trend from F (er 80:20, 16) via NO2 (er 18:82, 17) to CF3 (er 85:15, 18).In contrast, electron-donating groups (t-Bu) had a detrimental effect on er (catalyst S37, 57:43 er, see the Supporting Information, Scheme S1).Interestingly, catalysts 18 and 19, bearing either p-CF3 (18) or two m-CF3 groups (19) afforded similar enantioselectivities. Finally, diarylpyrrolidine 20 with four CF3-substituents provided a reaction with excellent level of enantioselectivity.
Studies to further enhance the er of the reaction were also carried out.For further optimization, it was clear that the diarylpyrrolidine core of catalysts 16-20 was lacking the needed modularity.Thus, we also screened with pyroglutamic-acid-derived pyrrolidine catalysts and the above trend was also observed with these catalysts (Scheme 2).Catalyst 21 gave almost a racemic product, along with a group of catalysts with electron donating substituents (see Supporting information, Scheme S1) but addition of electron-withdrawing CF3-groups improved the er of the reaction from 68:32 to 89:11 (catalysts 22 and 23).Unfortunately, the change of the phenyl-substituent of 23 to a 3,5-bis-CF3-phenyl (catalyst 24) or a pentafluorophenyl-substituent (catalyst 25) failed to elicit higher enantioselectivities.With these results, we decided to proceed with the total synthesis with our most selective catalyst 20.Scheme 2. Screening the catalysts for the Mukaiyama-Michael reaction a a) Enantiomeric ratios determined by chiral GC from the reaction mixture.
Further optimization of the reaction conditions (Table S1, entry 7, Supporting Information) revealed that TFA was the optimal counteracid.Lowering the temperature to -30 °C had no effect on er but the conversion fell dramatically.Interestingly, when water was excluded from the reaction, er of 95:5 was achieved (Table S1, entries 10 and 11, Supporting information).Unfortunately, the reaction never reached completion, thus making these conditions unpractical. 21ith these conditions at hand, we continued with the total synthesis.The entire route is shown in Scheme 3 starting from the known silyloxyfuran 4 (see also the Supporting Information). 22,23In gram-scale, the enantioselective Mukaiyama-Michael reaction afforded the aldehyde (+)-7 in 47% yield and er 94:6.(+)-7 was then reduced to 27 with an 86% yield after chromatographic purification.In both of these transformations, the sensitivity of acrolein, (+)-7 or 27 to polymerization were found to hamper the yields.
In our initial route, aldehyde 27 was first α-aminated with DBAD (28) using List's protocol 16 , and the resulting amino aldehyde 29 was oxidised to the corresponding carboxylic acid 30.Unfortunately, this relatively straightforward route to (+)lycoperdic acid had to be abandoned since the subsequent N-N-bond cleavage could not be reliably achieved (Scheme 3). 24n the alternative, ultimately successful route (Scheme 3), aldehyde 27 was subjected to an organocatalytic α-chlorination reaction, using perchlorinated quinone 31 as the Cl + -source and the MacMillan imidazolidinone TFA-salt 14 as the catalyst. 25,26nstead of 31, N-chlorosuccinimide was also tested in this reaction, but it afforded lower conversions than quinone 31. 27The intermediate α-chloroaldehyde was directly oxidised in the same pot to the corresponding carboxylic acid 32. 28The crude acid was then methylated with MeI under basic conditions, yielding the diester 33 in 71% yield over two steps.It was noteworthy that this two-step sequence could not be carried out with unsaturated aldehyde (+)-7; a complex mixture of compounds was obtained under the same reaction conditions.The diester 33 was then converted into the corresponding azide via SN2reaction with NaN3, yielding the azide 34 in 84% yield.The azide group was then converted to the Boc-protected amino group via hydrogenolysis in the presence of Boc2O.To our delight, the diastereomers were separable chromatographically at this stage, giving the desired full-protected natural product 35 in 74% yield, alongside with 4-epi-35 (9%, 86:14 diastereomeric purity).
With diastereomerically pure 35, the final stages were then explored.Saponification under basic conditions led to epimerization of the labile C2 stereocenter.In contrast, refluxing the compound 35 in 6 M HCl smoothly removed the Boc-and ester protecting groups, and after neutralization of the hydrochloride salt by an ion exchange column, crude (+)-1 was in our hands.In order to get analytically pure samples and to remove the hydroxy acid 36, the crude product was recrystallized twice from water giving us pure (+)-1 in 28% yield.It was noteworthy that 36 could not be transformed to the lactone by dehydration (e.g.benzene, reflux) since these conditions resulted in the formation of several side products.[22] and [23]   and SI

Initial synthesis route
In summary, we have developed an enantioselective organocatalytic total synthesis route for (+)-lycoperdic acid without using a chiral pool approach.As the key transformation, iminium-catalysed Mukaiyama-Michael reaction between silyloxyfuran 4 and acrolein (6) using a specifically optimized catalyst 20 successfully installed the key C4 tertiary stereogenic center.Efforts to synthesise derivatives of (+)-1 as well as wider studies of the developed Mukaiyama-Michael reaction are on their way.

ASSOCIATED CONTENT Supporting Information
Experimental details, characterisation data and copies of 1 H and Kortet, Aurélie Claraz and Petri M. Pihko* Department of Chemistry and NanoScience Center, University of Jyväskylä, Survontie 9B, 40520 Jyväskylä, Finland Supporting Information Placeholder