Totally Synthetic by Paul H. Docherty, 15 October 2007
Total Synthesis of Vinblastine
T. Miyazaki, S. Yokoshima, S. Simizu, H. Osada, H. Tokuyama, T. Fukuyama, Org. Lett. 2007, 9, 4737-4740.
I hope you’re all impressed with the target of this total synthesis already - that’s one complex beastie! In addition, Vinblastine offers some potent microtubule inhibition. This isn’t the first synthesis - there’s been a few of the natural product and of analogues, notably that of Phil Magnus back in ’92. In the current paper, the Fukuyama group document the synthesis of the top indole-containing fragment and the coupling with Vindoline, allowing completion of the synthesis. However, they state that they were supplied with Vindoline, so there’s no synthesis of that here. However, those of you with good memories will remember we covered Boger’s synthesis of that natural product last year.
The synthetic action starts with an oxidative lactonization to produce the 5,5-fused γ-lactone in great yield. I realise, this isn’t new - but it is nice!
Elaboration of this system (including a rather nice ring-expansion sequence) eventually led to a precursor to the indole. This was built in an interesting way - deprotonation of the lactone and addition into the isothiocyanate gave a thioanilide. They then did a radical-mediated cyclisation onto the styrene to complete the indole. A nice method, which reminds me much of Barton’s work in this area of heterocycle synthesis.
Opening of the lactone neatly gave them the ester and pendant alcohol required, but the macrocycle formation gave them a real headache. Protection group manipulations and trouble with function group interconversions meant that it took eighteen steps to get to the next intermediate shown! Very little of the trouble they had could have been forseen - a planned amine opening of a terminal epoxide to give the desired macrocycle failed to work reliably or in good yield - a real mid-synthesis nightmare. However, they persevered with admirable tenacity, and managed to get that top fragment complete. They chlorinated the indole with t-BuOCl to give a very reactive intermediate which coupled with Vindoline in a reaction thought to go through an iminium ion.
Perhaps the disconnection is a natural one, but that’s not the chemistry I would have planned - nice strategy! Completion of the natural product required only cyclisation of the tosylate onto a free amine to complete the skeleton of the top fragment, and deprotections, which all went smoothly.
This is an amazing accomplishment, especially given the trouble they evidently had - I take my virtual hat (I don’t wear hats) to the workers on this paper!