B.R.S.M. The road to Tet. Lett. Is paved with good intentions


This week in alkaloids part 1: (-)-Leuconicine A and B

Two nice alkaloid syntheses published last week; I'm not sure which I like more, so I'll cover them both. Look out for the second one in part two in a day or so. Also, I don't have any internet access at home for the next three weeks, so posts this month may be less well referenced, or just less existent.


Total Synthesis of (-)-Leuconicine A and B

Andrade et al., Org. Lett., 2011, ASAP; [PDF] [SI] [GROUP]

DOI: 10.1021/ol202056w

The leuconicines, like their more popular brethren strychnine and akuammicine, are members of the family of Strychnos alkaloids, and bear a strong resemblance to these natural products. They've got two fewer stereocentres and one less ring than strychnine itself, but they still look pretty formidable to me. The Andrade group accomplished syntheses of (±)-strychnine and (±)-akuammicine last year using a neat one pot spirocyclisation-Baylis-Hillman sequence to add the C and E rings to the indole core, but this paper is the first report of the use of their methodology to access enantioenriched targets.[1]

The work began with the formation of a chiral N-tert-butanesulfinimine from the aldehyde shown and readily available (R)-tert-butanesulfinamide, which was then allylated stereoselectively using an organoindium species. Treatment of the product with 4 M HCl followed by magnesium in methanol cleaved off the chiral auxiliary and the tosyl group to give the chiral homoallylic amine. This was alkylated with known (Z)-2-iodobutenyl bromide and then acylated with bromoacetyl chloride. Cross metathesis with methyl acrylate was (unsurprisingly) selective for the unhalogenated double bond, and the a,b-unsaturated ester required for the cascade sequence was obtained in good yield. Treatment of this compound with silver triflate and 2,6-di-tert-butyl-4-methylpyridine (DTBMP) effected the anticipated cyclisation to the spiroindolenine, and subsequent addition of DBU caused formation of the E ring by a Baylis-Hillman reaction. I've never seen this reaction done with DBU, which is notoriously non-nucleophilic, but it seems to work well here.

The Raucher protocol (Tetrahedron Lett., 1980, 21, 4061) for the reduction of amides to amines was then used as a milder and more chemoselective alternative to just bashing the molecule with LiAlH4. This involves conversion to the thioamide, S-alkylation with Meerwein's salt (Et3OBF4) and finally, elimination of ethanethiol from the resulting thioimidate with reduction by sodium borohydride. Athough this sounds like a lot of work, the reaction only required two flasks (simple concentration and a change of solvent is all that was required after the alkylation), and the yield over the two steps was very good. Reduction of the ester to the aldehyde worked best in two steps via the Weinreb amide. Methyl malonyl chloride was then used in a one pot N-acylation-Knoevenagel sequence to install the fifth ring of the natural product. Finally, the two-decade old Heck protocol establised by Rawal for the formation of Strychnos alkaloid D-rings was applied, using triethylamine as solvent and forming the hexacyclic skeleton of the natural product in good yield. Reduction of the ethylidene unit using Raney nickel (as platinum and palladium catalysts proved ineffective) gave (-)-leuconicine A, which was converted to (-)-leuconicine B using the Weinreb method for amide formation with trimethylaluminium.

All in all, a nice short chiral auxilliary based route to (-)-leuconicines A (14 steps, 9% yield) and B (13 steps, 10% yield), which could be easily adapted to make more of these popular alkaloids.


1. I also didn't have a blog back then. See J. Org. Chem., 2010, 75, 3529. (free Syn. Arch. here)


Comments (9) Trackbacks (0)
  1. If leuconicine B was made from A how can the synthesis of B have fewer steps and a higher yield than A?
    Wouldn’t that starting material be just (R)-tert-butanesulfinamide and wouldn’t the resulting homoallylic amine have an e.e. rather than d.r.?
    Idle curiosity – is there any way they could have used something like dibromoethane instead of bromoacetyl chloride to avoid having to reduce the amide?

  2. Ah, I’ve got them the wrong way round, that’s how! A should be the amide and B the ester. I’m not sure I follow you on the sulfinamide – yes, you could regard that as the SM, I mention the sulfinimine formed from that and the aldehyde for people who don’t know what’s going on here. You’re right about the amine – the d.r. refers to the allylated product before cleavage of the auxilliary. Originally I drew this as the second structure but then changed my mind (and forgot to move the d.r.). All better now.

    Regarding your curiosity – Alkyl halides are a lot less reactive than alpha-halocarbonyl compounds. I think I remember being told that bromoethane reacts with nucleophiles more than 100 times slower than bromoacetone (I’ll check this later when I have more time). Presumably the alpha-bromoamide they use is on the limit as they need a whole equivalent of silver triflate to help the reaction go. It might be that the yield was actually lower when dibromoethane was used instead of bromoacetyl chloride/amide reduction (80% overall to reduce the amide isn’t so bad). Or maybe they just didn’t think of it.

  3. Actually, now that I think a bit more, it’s probably also because beta-bromoamines just aren’t stable – I imagine they’d react to give aziridines very quickly, even here where that means quaternisation of a tertiary amine. Isn’t that how the nitrogen mustards work, and they’re tertiary amines with beta-chlorides (bis(2-chloroethyl)ethylamine was a popular war gas)? Maybe having an sp2 carbon in the way (and a much less reactive amide nitrogen) prevents this.

    • I just used the dibromo as an example since bromine was used in this synthesis. My thought was that given the nearly unlimited number of possible leaving groups there might be a pair that would work here.
      Incidentally, the war “gas” was bis(2-chloroethyl)sulfide.

      • The bis(chloroethyl)amines are called nitrogen mustards – they’re used as anticancer agents. I think they have some of the same effects as mustard gas and are pretty effective carcinogens.

  4. Nice review! I’m also very surprised they went with a DBU as a BH catalyst over the typical phosphines or DABCO. I decided to do some digging and figured out why. Andrade did in fact try the normal BH catalysts are a variety of loading and none of them worked. However, when the tried using DBU they got 90% yield of their test system. They noted that the mechanism could be that of gamma deprotonation but deuterium studies suggested that it was in fact just a intramolcular aza-BH. My bet is in both cases, the unsaturated ester is positioned so properly all you need is a tap the olefin to get the reaction to go and because DBU is so poor for BHs, it works perfectly. I thought this was a pretty effective and succinct total synth, I liked it!

    • Cheers for looking that up and letting us know! I did wonder if it was a gamma deprotonation, but it seems not. It’s always nice when these kind of unusual conditions are actually investigated by the authors, rather than just left for use to all wonder about, as in a lot of total syntheses.

Leave a comment

No trackbacks yet.