Fukuyama, Toma, Kita. JACS, 2010, ASAP. DOI: 10.1021/ja103721s
Ah – now this is a bit of a blast-from-the-past. I knew I’d written something about this molecule before, so I had a quick rummage in the Tot. Syn. folder on my PC – and found nothing. So I had a look in the other Tot. Syn folder, and another place I keep stuff for this website, and then four or five other places… (yeah, great filing system here) …and it turned up in this presentation I wrote a life-time ago. You can download the presentation here – it’s all based on a Nicolaou review that was published in Angewandte in 2005, with some mechanisms and added bits thrown in. I quite like the idea of sharing these kind of presentations, so perhaps I’ll try to add a few more. Anyway, I managed to find the structure of Manzamine A (no simple sketch!), along with mention of prior syntheses by Martin and Winkler.
I’m actually kinda surprised that there haven’t been many more efforts in this direction, as this is one tasty target. As usual, I’ll quote the authors with respect to biological activity: ‘cytotoxic, antibacterial, antimalarial, insecticidal, anti-inflammatory and anti-HIV’ activity are the headline – you need any more?! And that ring system is something special – two medium rings and a strained 6,6,5 core, with a tryptamine-derived unit bolted-on for good measure. Yeah, there’s a lot to be getting on with!
Going back to your first class in retrosynthesis, you’ll remember that cyclohexene = Diels Alder, and Fukuyama doesn’t disappoint. Working with a molecule related to Danishefsky’s diene (with an alkynyl pendant chain), a Diels-Alder with a chiral-butenolide (that is allegedly readily avaliable) completed the cyclohexene, providing two new stereocenters in excellent yield. Unfortunately, the endo/exo selectivity wasn’t amazing, with a 2:1 ratio in their favour, but a 65% yield of their desired product is still pretty respectable. Over the next few steps, the acetoxylactone was broken apart to reveal a pair of alcohols – one remained protected, whereas the other was used to build the fifteen-member ring.
To build-on the remaining 5,8-system, the group needed to append a small chiral fragment first, which they constructed in great yield by an asymmetric zinc addition. The protocol referenced is a sole-author paper by Nugent from back in ’99. Having done this, the group completed this small fragment by first introducing a carbamate group, and then displacing the primary alcohol with iodine.
With both fragments in hand, they did a pair of deprotonation / alkylation sequences to produce a quaternary centre with great control of stereochemistry. Considering that there are quite a few functional groups lying around, the control of reaction is really quite impressive. Even more impressive is that they apparently isolated only one stereoisomer; control derives entirely (and unsurprisingly) from facial selectivity – which is also presumably what controls the third step in this sequence, a nice little epoxidation.
With this done, it was the work of a moment to turn the allylic carbonate into a dihydro-pyrrole – though I’m simplifying the protocol by saying that. Firstly, the acid and amine perform a dehydration, creating a intermediate which does a [3,3] sigmatropic rearrangement to give the product, neatly transferring the carbonate stereochemistry across the double bond, forming an intermediate isocyanate. After a bit of experimentation, the group then found conditions to close the C-ring and form the required imine.
Another alkylation appended a greasy chain for creation of the D-ring – no prizes for guessing that this will be a bit of RCM to complete that eight-member ring. However, the A ring is still incomplete, so it’s interesting to see how Fukuyama went about it. Firstly, a bit of oxidation takes the ethanol group up to the aldehyde. Then, in a protocol that must smell worse than the salad-drawer of my fridge, a little thiophenol and caesium carbonate were used to remove the nosyl protection group, forming the imine concomitantly, which was reduced with a bit of hydride. And in a very nice yield!
The RCM did the business, and the remaining silyl group was removed the alcohol oxidised to the aldehyde. This allowed for coupling of the tryptamine group using a modified Pictet-Spengler reaction, and completion of the target in only a few more steps.
It’s hardly a short synthesis, but the target is a phenomenal challange, so hats off to the intrepid trio who managed this.