B.R.S.M. When all you have is a hammer everything looks like a nail


Iejimalide B

Note: I've never written a Tot. Syn. type post before, and it's come out a bit more detailed than I intended. I'll try and be a bit more brief next time...

Gram-Scale Total Synthesis of Iejimalide B

Fürstner et al., Chem. Eur. J., 2011, Early view; [PDF] [SI] [GROUP WEBSITE]

DOI: 10.1002/chem.201100178

Here's a nice modern synthesis from the Fürstner group; an impressive panoply of transition metal mediated reactions were used in this highly convergent synthesis of iejimalide B (although perhaps a little more tin than some people would be comfortable with...). I think the best way to show this is with a retrosynthesis:

Not happy with just the first total synthesis of the iejimalides (Tot. Syn. here) back in 2007 and prompted by encouraging biological results obtained using the material from the first campaign the Fürstner group set about developing a 'truly scalable' route to these compounds, a goal I'd say they'd accomplished admirably.  Apart the usual suspects (RCM, Stilles and a Heck) there's a few less common transformations which are worth a closer look.

The eastern fragment with its anti stereodiad was formed using a Marshall propargylation as the key step. I have to admit that I had to look this one up; it probably works a bit like this:

With great yield even on what most'd consider pretty big scale large amounts of the eastern fragment were easily produced. The large TIPS group is required for high diastereoselectivity, and it's easily knocked off with a bit of TBAF. Unfortunately, conversion to the vinyl iodide required a little bit more tin. The alternative, hydrozirconation with Schwartz's reagent followed by Zn - I exchange, proved problematic and also required protection of the newly formed alcohol. Hydrostannylation was easily carried out on the free alcohol (95% yield and 30:1 E:Z on >3g) using the stannyl cuprate formed from hexabutylditin, n-BuLi and copper cyanide, and ipso substitution with elemental iodine gave the desired vinyl iodide fragment.

The north fragment took its stereogenic center from the well known Roche ester after some fairly common manipulations, including a Takai-Utimoto reaction (CrCl2 and CHI3) to install the vinyl iodide. The ester side was manipulated first, requiring temporary protection of the primary alcohol. Interestingly the choice of protecting group proved critical to prevent racemisation during the Wittig reaction used to homologate the aldehyde derived from the ester. During their original route the group had employed a PMP ether for this purpose but in order minimise erosion of the ee they were forced to run the reaction at low temperature and instead use a phosphonate as the nucleophile which gave the unwanted Z isomer. They corrected this by isomerisation with PhSSPh and AIBN, but were understandably keen to improve this sequence for the large scale route. Fortunately they noticed that when Ley and coworkers had carried out a similar sequence with the chunkier TBS group that no racemisation was observed, and upon swapping the PMB to a TBS the Wittig reaction proceeded smoothly. Strange how these things can make a difference.

The western fragment was prepared in a neat four step sequence from 2-butyn-1-ol, again using the same 'lower order' cuprate employed in the synthesis of the eastern fragment in 71% overall yield. A Brown allylation (y'all remember those?) was then used to set the stereogenic center. The largest single batch carried through this sequence was 6g, and the cuprate addition to the alkyne could be carried out again without recourse to protection of the alcohol.

The southern fragment stereogenic centre was created by asymmetric reduction of the silyl ynone using a Noyori transfer hydrogenation with iso-propanol as the hydride source. The next couple of reactions really show the care taken with optimisation of this route; problems with performing ozonolysis selectively on the alkene in the presence of the alkyne were eventually solved by the use of sudan red 7B dye in the reaction mixture. Apparently, this inexpensive indicator decolorises after the ozonolysis of the olefin but before the alkyne starts to oxidise, allowing easy visual monitoring of the reaction. Methylation of the alcohol was also troublesome, but was eventually achieved by performing deprotonation with n-BuLi in THF, then changing solvent to DMSO for the alkylation with iodomethane. I bet that wasn't the first thing they tried, but you can't argue with 99% yield (or can you?).

The union of the fragments began with amide formation between the TBS protected N-formylserine and eastern fragments using HOBt and EDC. The southern stannane was then attached by copper assisted Stille coupling with good ol' Pd(PPh3)4 and CuTC in excellent yield. The north and western fragments were also Stille'd together using the same conditions, and after hydrolysis of the ester the north-west acid was coupled with the secondary alcohol in the south-east piece. Finally, Grubbs 2nd generation catalyst was used to close the macrocycle in an unbelievably good yield, given the ten double bonds present in the molecule, with only a trace of the undesired Z isomer formed. This final step was 'only' performed on 300 mg at a time, an awesome testament to the power of RCM chemistry. Finally, cleavage of the sole protecting group, the TBS ether on the serine derived fragment, gave the natural product in essentially perfect yield. Well done guys! A rare dearth-of-material solving paper that can deliver!


Comments (0) Trackbacks (1)

Leave a comment

8 × six =