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

17Aug/112

(–)-N-Methylwelwitindolinone C Isothiocyanate

Update: 13/09/11. I have reworded the introduction considerably as the original version of this paragraph described work by Rawal and coworkers in this area unfairly. Having reread the papers and what I'd written I couldn't leave things as they were. The original text can be found in [5] at the end of the article.

Sorry things have been so quiet around here lately - moving out of the flat I've lived in for the past three years, having no internet and applying for conferences have made blogging tricky. I was a bit worried that someone'd beat me to this as it's such a nice bit of work, and Tot. Syn. did on Sunday, but I'd mostly finished this by then, so here's my take anyway.

Total Synthesis of (–)-N-Methylwelwitindolinone C Isothiocyanate

Garg et al., J. Am. Chem. Soc., 2011, ASAP; [PDF][SI][GROUP]

DOI: 10.1021/ja206538k

There's been a lot of interest in these unusual looking natural products in the past decade with publications from Wood, Trost, Baran, Simpkins, Martin and Rawal all appearing in the literature. Credit for the first total synthesis of a member of this family goes to Rawal and coworkers, who reported a neat route to racemic N-methylwelwitindolinone D isonitrile back in March of this year, unfortunately before I'd started this blog.[1,2] This paper by Garg describes the first synthesis of N-methylwelwitindolinone C isothiocyanate to date, using a chiral pool based approach from (S)-carvone, and also contains lots of interesting chemistry.

The route began with a known five step sequence developed by Natsumi and coworkers back in 1994, converting fragrant (S)-carvone into the g-hydroxyketone shown below. LiAlH4 reduction occurred with the expected diastereoselectivity, and the resulting allyl alcohol was protected as its pivaloate ester (the yield for this second step is buried in an old Tet. Let. I can't get hold of). Next, chromium trioxide was used to perform an allylic oxidation which proceeded with the typically poor yield this transformation usually gives. I've done a few of these, and they're universally unpleasant reactions, especially on the kind of scale Garg presumable worked on here.  Finally, isomerisation of the terminal olefin into conjugation with the ketone using sodium methoxide in methanol, followed by regioselective 1,4-addition of vinyl magnesium bromide at the other enone gave the required building block in good yield and great d.r. Interestingly, if the order of these two steps was reversed then the yield was much lower, and a mixture of diastereomers was obtained.

Moving on to work conducted this millennium, the pivaloate was then removed under basic conditions and the enone then reacted with the bromoindole in the presence of iodine to give the tricyclic adduct shown. The secondary alcohol was then protected as its TBS ether (under unusually violent conditions), setting the stage for one of the coolest steps of the synthesis. Thus, treatment of the bromoindole under classic sodamide - t-BuOH  conditions in THF at room temperature regioselectively generated a 4,5-indolyne intermediate, which cyclised to give a mixture of C- and O-arylated products, in favour of the former, in a combined yield of 46%.[3] Interestingly, the group reports that the compound bearing an epimeric OTBS group doesn't undergo the reaction. The yield isn't great (amounting to 33% of the compound they actually want; the second yield in the 30s so far), but it is all kinds of cool, and it allows them to form this challenging bond without having to additionally functionalise the system. The TBS group was then removed with a bit of TBAF, and the exposed alcohol oxidised. A three step sequence was then used to convert this newly formed ketone to the corresponding vinyl chloride via the triflate and trimethyl stannane (eek!) using some known chemistry. The indole was then oxidised to the corresponding oxindole via C-2 bromination with NBS, followed by hydrolysis.

Next, one of my favourite steps of the synthesis - the functionalisation of the C-11 position. The group found that it wasn't been possible to perform the indolyne cyclisation with any substituents at this position, and intermolecular methods of functionalisation (radical, nitrene, enolate...) had been similarly unsuccessful. The group therefore turned to an intramolecular method of introducing functionality here. Thus, the ketone was diastereoselectively reduced using DIBAL, and the resulting equatorial alcohol was converted to the primary carbamate. This was then oxidatively converted to the acyl nitrene, using silver(I) triflate with a substituted phenanthroline ligand and diacetoxyiodobenzene. Intramolecular C-H insertion at the recalcitrant C-11 position ensued to form the oxazolidinone, which was then hydrolysed to give the aminoalcohol. The alcohol was reoxidised to the ketone using IBX and finally the amine was converted to the isothiocyanate using the thiophosgene equivalent shown.[4] A great end to a very interesting synthesis. At 17 steps it's not a bad length, and although the yields are a bit low in places I doubt anyone will give that much thought.

 

1. J. Am. Chem. Soc. 2011, 133, 5798–5801. The palladium catalysed intramolecular enolate arylation used in place of Garg's indolyne cyclisation is neat.

2. Yes, the N-methyl compounds are naturally occurring.

3. Garg's been into indolynes for a while - this strategy was first disclosed in an Org. Lett. back in 2009 and he then teamed up with Houk to attempt to explain the reactivity of indolynes with nucleophiles a year later. There's a nice RSC Chemistry World article (free online!) with some more references and info here.

4. It's basically the same as TCDI.

5. The first paragraph read "Here's something that's been overdue for a while - the first complete, asymmetric, synthesis of a welwitindolinone alkaloid, and not necessarily from who I would've expected. Let's not forget that Wood, Trost, Baran, Simpkins, Martin and Rawal have all published in this area; Rawal even reported the racemic synthesis of N-methylwelwitindolinone D isonitrile back in March of this year.[1] Anyway, to the victor go the spoils, and after more than a decade of work by the synthetic community Neil Garg will be remembered as the first to make one of these densely functionalised natural products.[2]"

 


Comments (2) Trackbacks (1)
  1. Welwitindolinone D is an isocyanide (RNC) rather than an isocyanate (RNCO).


Leave a comment


1 − one =