Home » Still In The RBF

Isatisine A   

10 January 2010 14,613 views 17 Comments


Kerr, Karadeolian. ACIEE, 2010, EarlyView. DOI: 10.1002/anie.200906632. Article PDF Supporting Information Group Website

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.


1 Star2 Stars3 Stars4 Stars5 Stars (17 votes, average: 4.18 out of 5)


  • milkshake says:

    if you are using Mg turnings for grignard (instead of Mg powder) the Mg/MeOH reduction of tosyl goes a lot faster if you dip the Mg turnings into a diluted HgCl2 or Hg(OAc)2 methanolic solution – the amalgamation breaks the MgO layer, it may get too fast if you overdo it.

    By the way in this case they probably could have used some nucleophile+base instead, for exaple a thiolate, to remove the N-tosyl from indole: indole anion is a fairly good leaving group.

  • Industry Guy says:

    Nice write up as always. Several great reactions here – the THF formation is awesome.

    Though you say “setting two chiral centers.” I have to correct you. Chirality is a molecular property – and is not associated with any given “center.” Thus a chiral molecule has stereogenic centers – not chiral centers.

    Just being picky. It always bugs me – like everyone and his dog using “access” as a verb. You cannot gain access to a molecule, as “access” is not a verb.

    • Johnny Bravo says:

      “You cannot gain access…”

      Gain is the verb here, not access. So yes, you can “gain access” to something.

      Either way, access can be used as a verb, so the point is moot: http://wordnetweb.princeton.edu/perl/webwn?s=access

    • ; - ; says:

      Language isn’t exactly static.

      Plus, verbing weirds language!

    • antiaromatic says:

      @ Industry Guy

      I am afraid that “access” can indeed be a verb. Check any dictionary. I checked dictionary.com, and found this entry:

      verb (tr.) – to make contact with or gain access to; be able to reach, approach, enter, etc.

      Random house dictionary and webster’s have similar entries.

    • Tot. Syn. says:

      I must also add that according to the IUPAC Gold Book:

      chirality centre
      An atom holding a set of ligands in a spatial arrangement which is not superposable on its mirror image. A chirality centre is thus a generalized extension of the concept of the asymmetric carbon atom to central atoms of any element, for example N+abcd, Pabc as well as Cabcd.

      I said chiral center rather than chirality center…

      Also, I think what you mean could be expressed as an axis of asymmetry, or as the gold book puts it chirality axis.

      But lets not get overwrought with language!

  • MSG says:

    Eluting a column with acetone :O !!! Admittedly in my first year PhD I eluted my crude product mixture from a silica pad with acetone and got nasties out of it… a postdoc rightly suggested that I NEVER do that again.

    Anyone know of any other famous instances of natural products being isolated that are artifacts of the isolation procedures?

    • milkshake says:

      just ask former Corey group member who worked on synthesis “endogenous quabain-like isomeric compound” and observe them doing both-handed facepalm. That natural compound isolated from cow brains that was supposed to be an isomer of quabain with a misplaced one hydroxy group. After several years of educated guesses about its structure and several semisynthetic attempts starting from quabain (just imagine a molecule like quabain and then try to move the hydroxyls around on the steroid piece through selective protections and transformations) it has been recognized by the Japanes group who originally isolated this wonder that the mystery molecyle is indeed quabain complexed with boric acid that was most likely picked up from glassware during the isolation. How quabain got into those cows brains is stil unexplained – either the cows ate the plants or some postdoc wanted to make his Japanes boss happy and added a tip of spatula to the material…

    • Tok says:

      What’s wrong with acetone and silica? I’ve used hex:acetone to run dozens (hundreds?) of columns and have had no problems. As long as you don’t have a 1,2 diol or unprotected primary or secondary amine you should be fine. It’s actually quite nice in automated columns because it doesn’t absorb at 210nm and since it’s more polar than ethyl acetate it has a wider range of applicability.

    • MSG says:

      Just to clarify: I guess what i’m questioning is the use of acetone/silica when isolating natural products. In these cases you cannot be completely sure what sort of compounds you have in the mixture, and even if you know you don’t have alcohols in there, aldol adducts are also possible. Ethyl acetate is much more inert than acetone in regards to these possible reactions.

  • aaaa says:

    If you do not know the structure – sure you do not want to use something reactive to isolate the compound, but if you know – why not. We use in our company on daily basis – silica columns with added ammonia, acetic acid, acetone/water mixtures etc. No problem whatsoever, just think a bit before flashing. Also the BS about not using MeOH on silica – remember in my former lab people were arguing that more than 25% MeOH in the eluent dissolves silica. Yeah, books write about it. So, we took a column, filled with silica and eluted with pure MeOH in a 1L RBF. Evaporated – there was absolutely nothing! I personally have used pure MeOH with 10% added aqueous conc. ammonia on silica without any problems at all. Thinking back I clearly see that academia people have actually many myths going around in the labs, because the lab experience in a normal academi lab is no longer than 4-5 years, then people just leave. And a cler lack of industrial experience is the missing point in academic research, as I see it now.

    • milkshake says:

      silica actually does not dissolve in methanol – even when everybody (and his grandmother) swears it does. What happens instead is a strong exotherm from MeOH absorbing on the silica. You get a local hot zone at the methanol front that creates a turbulence and vapor pressure, this pushes the fine silica particulates through the sintered glass frit (or the cotton plug) on the bottom of your column so what comes out when the methanol front reaches the bottom of your column is gonna be cloudy.

      If you do not use methanol in gradient gradient and if you pack your column as a slurry with the actual solvent mix used for elution you will not have problem with silica getting through regardless of the used methanol concentration

      • SomeGuy says:

        Word. I’ve certainly seen the same thing with radical changes in an eluent’s EtOAc content. Just for completeness’ sake and all.

  • AIK says:

    I’ve run DCM/MeOH/ammonia columns like a motherfucker and usually there’s no problem with silica going through but occassionally it has occurred.

    • Billy says:

      I really don’t care what you do with you mother but there is no need for that language on a scientific blog which is gaining some repute as a place people can go and be educated. Please **** off!

  • chemist says:

    Nice to see so many chemists here with interest in total synthesis.

    Any idea whether total synthesis of FOSMIDOMYCIN (natural antibiotic) is reported. I could only find the synthesis of various analogues but not this drug.


  • ChemistryGuru says:

    I just wanted to digress for a moment to highlight a mistake in the discussion of the oxidation step. The supporting information notes the use of 2 equivalents of mCPBA. Of course, the initial equivalent is used for 2,3-epoxidation of the indole. However, the authors reference a paper by Kobayashi (Buller, M. J.; Cook, T. G.; Kobayashi, Y. Heterocycles 2007, 72, 163) as the logic for their mechanistic insight into the interesting oxidation transformation. That particular work hypothesizes that the mechanism of action following the indole nitrogen opening the epoxide is an Amadori rearrangement (relatively common to sugar chemistry) followed by N-oxidation. Elimination of water off of the protonated nitrogen at the 2 position reveals the electrophilic 2 position of the newly formed indolone, poised for intramolecular cyclization of the primary alcohol to give the 2,2-disubstituted indolone intermediate.