Saturday, June 23, 2012

Trauner's Synthesis of the Hydrindane Substructure of Retigeranic Acid A

We have previously discussed synthetic strategies to access functionalized hydrindane structures in the context of the preparation of building blocks for steroid total synthesis. Hydrindanes also constitute substructural skeletal fragments of related natural products including limonoids and sesterpenoids. A popular synthetic target within the latter class of natural products is retigeranic acid (structure shown below). This complex pentacyclic terpenoid has been synthesized previously by the laboratories of Corey, Paquette, Hudlicky and Wender. Quite recently (JOC ASAP), Dirk Trauner's group in Germany described a new synthetic protocol to generate hydrindanes that might be useful for the synthesis of various sesterpenoids (including retigeranic acid) and derivatives thereof. The route exhibits a high level of stereocontrol and features a fascinating hydrogenation step that results in an unexpected stereochemical outcome.
The forward synthesis begins from the Hajos-Parrish diketone (2) which is accessed in enantiopure form by implementation of the Hajos-Parrish-Eder-Sauer-Wiechert (H-P-E-S-W) reaction. Nine subsequent operations are then required to advance to the silyl enol ether (3). The highly linear nature of the conversion of 2 into 3 is a shortcoming of the Trauner hydrindane synthesis. Nonetheless, 3 undergoes Saegusa oxidation and conjugate addition to secure the hydrindanone 4 in good yield and as a single diastereomer.
Interestingly, hydrogenation of the olefin 4 under standard palladium on carbon conditions afforded the ketone 5 containing the (R)-configuration at the isopropyl-substituted cyclopentanone carbon position. The authors postulate that the epimerization is the result of palladium-catalyzed isomerization of the isopropenyl group to the more highly substituted olefin prior to reduction. The unexpected result was confirmed by an X-ray crystal structure analysis. Deoxygenation of the ketone 5 is then achieved by palladium-catalyzed reduction of the corresponding enol triflate to deliver the cyclopentene 6. Finally, straightforward hydrogenation and acetal cleavage generates the trans-hydrindane building block 7 which contains the intact substructure of the sesterpenoid natural product retigeranic acid A. Notably, intermediate 7 has been converted into retigeranic acid A twice previously (here and here). The authors comment, with regard to the 'stereochemical serendipity' in the conversion of 4 into 5, that this work "emphasizes the importance of meticulous product analysis, even with seemingly straightforward reactions."

Sunday, June 17, 2012

E. J. Corey's Synthetic Approach to Germanicol

In 2008, E. J. Corey's laboratory disclosed a formal synthesis of germanicol (see previous post) that was enantioselective. The work was characterized by the authors as 'a major advance over the original route to [germanicol] in terms of brevity, efficiency, and enantiocontrol.' The Corey route is initiated by the Corey-Zhang dihydroxylation of 2,6-E,E-farnesyl acetate (1), a reaction that, to its credit, delivers multigram quantities of the chiral diol 2, but that necessitates a complex cinchona catalyst that is not commercially available (catalyst A). This rationally designed ligand is capable of molecular recognition of a polyolefin substrate and delivers the terminal dihydroxylation product (2) in good yield and with outstanding stereoinduction. Three subsequent operations then secured the acyl silane 3.
Nucleophilic addition of 2-propenyllithium to 3 generated an intermediate (Int-I) that readily underwent Brook rearrangement under the specified conditions. In situ benzylation of the resultant allylic lithio species then provided 4. This triene (4) was the substrate for an epoxide-initiated pi-cation/olefin polyannulation reaction that occurred upon exposure to Lewis acidic conditions at low temperature. O-Silylation of the cation-olefin polyannulation product delivered 5, which cyclized further to give 6 under conditions that were reported by Ireland and co-workers during their germanicol campaign. The pentacycle 6 intersects with an Ireland-Johnson synthetic intermediate en route to germanicol. It is noteworthy that the laboratory of Corey has provided access to 6 as a single stereoisomer. However, no less than eight additional steps are required (as per the Ireland-Johnson protocol) to convert 6 into germanicol. Additionally, upon inspection of the supporting information of the Corey JACS publication (JACS, 2008), it was found that the conversion of 4 into 6 was conducted on 100 milligram scale, providing only 25 milligrams of 6 over the course of three synthetic steps. It is not clear from the data provided by Corey and co-workers that this sequence would work well on large (gram) scale. Therefore, while the Corey protocol is indeed an advancement with regard to 'brevity,' if I were tasked with the synthesis of one gram (arbitrary quantity) of germanicol, I would probably start reproducing Bob Ireland's total synthesis, which, to reiterate, was published in 1970. Clearly, unmet challenges to present day synthetic science persist.

Wednesday, June 13, 2012

R. E. Ireland's Total Synthesis of Racemic Germanicol

In celebration of the contributions of Robert Ellsworth Ireland to the field of organic synthesis, we will examine Ireland's classic total synthesis of germanicol, conducted in collaboration with the laboratory of W. S. Johnson and disclosed in JACS in 1970. This work constituted the first synthesis of a complex and unsymmetrical pentacyclic triterpene and its efficiency remained unrivaled until 2008 when Corey described a formal synthesis of germanicol that was significantly shorter than the 32 step campaign of Ireland/Johnson (0.1% yield). Ireland's preparation of germanicol constructs rings A-C in the early stages and appends the final two rings by means of a conjugate addition of a benzylic Grignard reagent to a key exocyclic enone intermediate.
The early tricyclic intermediate 4 is first prepared in seven steps from the Wieland-Miescher ketone derivative 3. Functionalization and subsequent annulation of the pro-C13 and C14 carbons was then required to introduce the terpenoidal D-ring. This was accomplished by the initial formation of an olefinic ketal (5) through standard functional group interconversions. Oxidation of the pro-C13 position to the requisite ketone is then achieved by exposure of the olefin to singlet oxygen (generated in situ from molecular oxygen in the presence of a sensitizer with irradiation) in combination with a reducing agent. Under these conditions, a Schenck ene reaction ensues and the intermediate alkyl peroxide is reduced by LAH to give a secondary alcohol. Oxidation of this alcohol with hexavalent chromium in the subsequent step secures the key enone intermediate 6 that is poised to undergo smooth conjugate addition with meta-methoxybenzylmagnesium chloride to give 2.
Intermediate 2 is advance to 7 in two straight-forward operations and cyclization of 7 occurs upon treatment with polyphosphoric acid for 30 minutes at room temperature. This efficient protocol affords the pentacyclic intermediate 8 in 90% yield. Nine additional steps are then required to install one angular methyl group, adjust the oxidation states of several carbon positions across rings C-E and, finally, to install one unit of unsaturation in the E-ring (enone C-C double bond). In the endgame of the germanicol synthesis, deconjugation of the E-ring double bond precedes a base-mediated double-methylation reaction. Finally, a Wolff-Kishner reduction of the extraneous E-ring carbonyl results in totally synthetic (+/-)-germanicol. Ireland's laboratory subsequently developed a general convergent synthesis of pentacyclic triterpenes that was used to make additional triterpenoid natural products including alnusenone and friedelin.