Isatisine A

Kerr, Karadeolian. ACIEE, 2010, EarlyView. DOI: 10.1002/anie.200906632.  

It’s always going to be a tough situation for the article I blog following something like Palau’amine, but this short and sweet synthesis by Michael Kerr of the University of Western Ontario (in the other London – which I visited in ‘95) is as good as can be.  The rationale for it’s investigation is also pertinant, as it turns out that Isatisine A is a moderately active HIV isolate (EC₅₀ of 38 ?m).  A secondary interest in the isolation was that there was a little confusion, as the compound that reluctantly streaked out of the final column (it’s been one of those weeks) was the syn-acetonide.  This was initially thought to be the target, but resulted from eluting the column in acetone…

The synthesis starts with a damned neat reaction; a Johnson tetrahydrofuran synthesis, using a cyclopropane, an aldehyde and a little Lewis acid.  This reminds me of a lot of recent work with gold; the similarity clearly comes from the fact that most gold complexes are just fancy Lewis acids.  Full details on the reaction are found in a recent JACS, here.  The mechanism involves an “intimate ion pair pathway“, something I like the sound of.  I can just imagine the two halves of the cyclopropane sitting down with a nice pinot noir, a couple of steaks and a Richard Curtis film… More sersiously, read the JACS.

Kerr seems to be disappointed in the  diastereoselectivity of this reaction as the result was a 11:1 mixture of the -cis :-trans. I’d be happy enough, but no matter, as it’s improved later in the synthesis.  With two chiral centres fixed, it was time to functionalise the THF.  The suspicious looking differentiated malonate was up for it’s starring role next, as the benzyl ether was predictably ’saponified’ by hyrodgenation, and fitted up with a allyl ester.  This could then be decarboxyated using palladium, resulting in a 2,5-dihydrofuran.  As you can all guess at this point, dihydroxylation with substrate controlled conditions (methanesulfonamide, NMO, OsO4) provided a pair of stereocenters, protected up.  The tosyl group was then removed from the indole using magnesium metal in methanol – conditions I’ll have call for soon.  (For those that haven’t been there, tosyl groups, whilst easy to apply to the amine of your choice, can be a bastard to remove.  I’ve had all sorts of problems after trying sodium naphthalamide, shit-loads of neat acid, and even the new Sodium-on-silica stuff.  Grrrr.)

This left them set for the appendage of an indole, allowing with a bit of oxidation.  In two steps, they form an epoxide on the existing indole, which was attacked by the free alcohol (which looks kinda remote, but this is a cisoid THF).  The resulting aminal is perfectly set to be C3-indole alkylated, requiring only a little acid to facilitate this.  In practice, the reaction was initially favouring the incorrect epimer, but leaving it lying around in acid allowed an equilibration to occur, giving a 3:1 ratio in favour of the desired isomer after 24h.  If one is prepared to spend a bit of time reading the lit, perhaps a bit of Tot. Syn., listening to Spotify and certainly a few coffee (adding up to 42h), a more satisifying 6.3:1 ratio is to be had, with concommitant lactamisation, and delivery of the protected natural product.  Neat work, folks.