Home » Still In The RBF

Guanacastapene N   

28 September 2006 8,443 views 28 Comments


Overman, Limura, Paulini and Zakarian. JACS, 2006, ASAP. DOI: 10.1021/ja0650504.

I seemed to miss this one somewhere, probably in the move from my placement back to Cambridge. Anyway, this is a really nice synthesis of the popular guanacastapene targets (see this smart synthesis by Sorensen), using a unusual and funky Heck cyclisation. This reaction forms the main unifying transformation of the separate fragments, along with a 1,4 organocuparate addition. On with the synthesis!


I really liked this smart use of a Ireland claisen to set the stereochemistry about the quaternary centre in this transformation. The chiral acetate was recieved from a resolution, and with three neat steps they’ve already set a stereochemistry of a complex centre. Conversion of the alcohol to an iodide left them with a coupling partner for the 5-member ring in 69% over four steps, and into the raison d’etre:


Doing an unusal 7-endo-heck, through an eclipsed insertion topography, allowed then to access the 7-member ring in a good yield. The 6-exo process is stereochemically impossible, as the required geometry forces the methyl group into the five-member ring, so the desired product and process prevails completely, completing the carbocyclic skeleton of the natural product. Interestingly, they acheived a better yield of the methyl ester analogue, but were unable to cleave the ester, so had to resort to the benzyl ester shown.

Elaboration of this structure was completed in a fashion similar to that used by Danishefsky in his synthesis of guanacastepene A, leaving the group with a intermediate which could lead to many mambers of the guanacastapene family.

1 Star2 Stars3 Stars4 Stars5 Stars (No Ratings Yet)


  • HOMO-LUMO says:

    Guanacastepene N. And no comments!!. Hello, somebody there?!

    Thanks for the post Tot.Synt.

    Lovely molecule, lovely chemistry, and interesting way to make the molecule. The only issue which can be criticised is the complete lack of diastereoselectivity in the introduction of the isopropyl group which is the main synthetic issue of this target.

  • Chinese chemist says:

    nice work

  • European Chemist says:

    Am on holidays right now and so couldn’t read the .pdf file. Anyway, the steps you highlighted are definitely very interesting and original, particularly the 7-endo Heck cyclisation. The Claisen manoeuver is sort of “conventional” chemistry.

    Thank you Tot. Syn. for keeping me posted on chemical literature even during my “vacaciones”! Great post once again.

    HOMO-LUMO mentions something which is not detailed and which seems to be a drawback. Could you add something about this in the post or at least insert a small comment? It is true that from reading the description one gets the idea that the rest of the synthesis is even better than the steps you showed… no criticism intended, just trying to raise some discussion. ;-)

  • WillisWill says:

    I can’t access the article in order to read Larry’s sale of the Heck, but is the 7-endo pathway really that unusual in this case? If it proceeds via 6-exo, the resulting intermediate can’t beta-hydride eliminate. Maybe I’m missing something

    Overman = King of the 3,3!

  • Mitch says:

    What is DMPU?

  • Klug says:

    Has anyone read the Nicolaou’s synthesis of platensimycin? Interesting stuff, but the last two paragraphs of the paper aren’t the hottest.

  • HOMO-LUMO says:

    Ufff, European chemist,

    Is quite difficult to explain you that issue if you dont have the paper on hand.

    In order to get the 4-isopropyl-3-methylcyclopent-2-en-1-one (5 membered ring with the enone, the methyl group and the isopropyl group with the correct stereochemistry), they started from the a avalaible 3-hydroxycyclopentenone (2 ene, 1-one) then they condensated it with (+)-menthol (chiral auxiliary which creates a chiral enviroment to allow “stereoselective functionalization” in the other alfa position to the enone). That functionalization was a Stork-Danheiser alkylation via formation of a zinc enolate (instead of the traditional lithium enolates)by using LDA and subsequently Et2Zn then treatment with 2-iodopropane as electrophile allowed the attachment of the isopropyl group. (This is probably the most challenging C-C bond formation reaction in the Guanacastepene family and hence probably one of the most interesting.) At the end 71% of a diastereomeric mixture but just with a 1.5:1 ratio. (meaning that the real yield of the desired diasteromer is circa 45%).

    DMPU, (1,3-Dimethyl-3,4,5,6-tetrahydro-2(1H)-pyrimidinone), solvent used commonly as a replacement of the highly toxic, aprotic polar solvent HMPA (HMPT), used in many reaction to enhance the reactivity of some species, i.e. in LDA alkylations makes the lithium enolates more reactive maybe via deagregation of n-BuLi from hexamers (if in hexane to dimers or monomers).
    In theory DMPU is a expensive sh..which does not make the trick at all. If your reaction works without HMPA, great, if it doesnt, then try HMPA, if not, you are in trouble (try something else),DMPU is expensive and doesnt do anything.

  • Disillusioned Hamster says:

    I’ve seen the platensimycin synth – and I was betting on it showing up here fairly soon. It’s slightly disappointing in parts; nice use of Trost’s Ru cat. but the low yield and selectivity in the SmI2 radical cyclisation lets it down.
    It’s quite a hot molecule at the moment, so I expect the idea was to get the first synthesis in print before anyone else gets the chance. It sounds like they’re still working on (?asymmetric) synthesis, maybe the full paper will be better.

  • SJB says:

    DMPU = DiMethylPropyleneUrea (or something like that) – cyclohexanone, with N-Me groups instead of the alpha carbons.

  • European Chemist says:


    I just read the paper, as I am back to the lab (snif). Guess holidays can’t last long when you’re a graduate student…
    And you’re absolutely right about the synthesis of that building block. Being, as you correctly state, one of the most challenging issues (the isopropyl group!) in the Guanacastepenes, it’s rather disappointing to get such a bad outcome at such an early stage of the synthesis!


    You’re right, the resulting alkylpalladium species in a 6-exo would not be able to eliminate. But it could still do all sorts of stuff, like engage in a 3-exo with the carbonyl followed by ring-opening. The authors seemed to plan something like that in the first place and cite some references.

  • Klug says:

    Hamster: I agree about getting it out first. Matter of fact, the last two paragraphs were, IMHO, a note that said: We tried some stuff, it hasn’t worked yet, but we’re first.

  • Tot. Syn. says:

    Hamster, Klug et al,

    I’ll have a post on platensimycin tomorrow. I’m having a bit of operational difficulty just now – no internet access at home, and I don’t like writing the blog at work. Generally, because I tend to post at 11pm or later. Anyway, should be up tomorrow!

  • dtb says:

    Re: platensimycin synthesis:

    “No one in 2006 and beyond can doubt that, in principle,
    virtually every structure within reason can be synthesized if
    enough resources (including time!) are applied to the problem…

    If this obvious point is accepted, the emphasis on finishing first…should be “down-regulated.” Since the feasibility, in principle, of complex molecule total synthesis is no longer under challenge, the imperative to finish first, although in keeping with human nature, is correspondingly diminished. The real issues of contemporary total synthesis are more subtle and more sophisticated than sheer demonstrations of feasibility and order of crossing the finish line, however exhausted. The major determinants of contemporary total synthesis may well be in problem selection, synthetic style, and teaching potential – in summary, *quality.*”

    – Danishefsky

  • Canadian Chromatographer says:

    I do not have the reference with me, but in Nature there was an editorial that lauded organic chemists for their constantly renewed approaches to problems. It’s definitely worth a good read!

  • HOMO-LUMO says:

    dtb, is that from a JOC 2006, Wilson, R.M.; and Ose sensei Danishefsky?

  • yepyep says:

    I’ll be commenting on platensimycin when the synthesis is posted but I have a question for you all. Which one do you value more, the first synthesis or the first practical synthesis? And by “practical” I mean things like: optimized to the bone, no overly expensive reagents, upscalable (is that a word?), etc.

  • HOMO-LUMO says:

    Good question.

    The one with more value to me is the better one (the one which would be more dificcult to superseed in terms of efficiency,elegance and innovation), no matter when is reported. i.e Corey-lactacystin. Corey-salinosporamide A. Evans (-)-FR182877. Hard to say with the CP-molecules, etc.

  • jimbo says:

    yepyep: depends on the purpose/length of the route

    Is it for SAR studies? A quick and dirty unoptimized route is fine. (The so-called “med chem route” that process guys love to complain about)

    Is it to scale up to put in something living? You had better optimize the hell out of it and get rid of all of your beloved toxic transition metals. Oh yeah… no columns either.

    In my opinion, your question is a matter of personal preference. What if KCN’s somewhat unoptimized route gave birth to a new, more active analog of the natural product. (I doubt it, but you never know.) Is this of more or less value than the highly optimized route of another group? What if another group sees this work and realizes that they can intercept an intermediate with their pet methodology and this leads to a much more efficient route? I say, if you have a synthesis, put it out there… more knowledge is good.

    It’s odd that KC’s group catches flak for publishing quick and dirty routes to natural products, but nobody calls out the plethora of groups who optimize the snot out of fragments of molecules but are never able to finish anything.

    Anyway, arguing this point is silly… if this molecule were actually going to be used, it would be obtained by fermentation. Despite the fact that we justify our work with biological arguments, the truth of the matter is that complex molecule total synthesis is academic masturbation… it’s practice for the real thing and when you’re done you’re really tired.

  • HOMO-LUMO says:

    I just saw the paper this morning (no idea what was platensimycin). Impressive molecule, some good and some modest chemistry.(All my credit to them)

    Knowing one of the authors, no surprise that one of the key steps is the SmI2 radical cyclization but this time in a substrate 100 time more complex than the ones he used to work with. :).
    Other points are for example I quickly notice, if you are using a Saegusa oxidation to regenerate an enone double, is not quite an accomplishment to use 1.1 equiv of Pd(OAc)2.

  • scale-uppable is the word you were searching for, yepyep :)

  • Chemist of Sorts says:

    I look for elegance and novelty in synthesis. Since its accepted that most compounds can be made if you throw enough postdocs at the problem (ala KC), the punch comes from doing things smart. Process type syntheses are interesting and can be extremely impressive, but I like to see novel chemistry. I think the platensimycin is one of KCs best in a while.

  • rb says:


    where is that danishefsky quote from?

  • dtb says:

    rb, HOMO-LUMO:

    The Danishefsky quote is indeed from an essay currently on JOC ASAP: http://dx.doi.org/10.1021/jo0610053

    (page M, right-hand column)

    The whole essay is a good and rather provocative read (I think it’s been featured on this blog, actually)

  • regular chem says:

    To European chemsit,
    “You’re right, the resulting alkylpalladium species in a 6-exo would not be able to eliminate. But it could still do all sorts of stuff, like engage in a 3-exo with the carbonyl followed by ring-opening”
    I consider myself quite knowledgable of Pd-chem, but I can not recall any example of what you suggested. I know what you suggested has been observed in Ti-chem (see cyclopane synthesis)and Zn-chem. I can understand if the organopalladium intermediate can further react with aldehydes, esters, undergo CH-activation,etc. Can you give me example of what you suggested.
    Another thing to consider here is whether the interpetation of 7-endocyclization is correct or not. Contrary to a wishful thinking that the reaction proceed in a 7-endo-heck fashion, there is anothe possibility. After the first oxidative insertion, the resultant organopalladium species can undergo 6-exo-cyclization, followed by anothe carbopalladation to the double bond conjugated to the ester to generate a cyclopropane and Pd-enolate. The Pd-enolate can beta-eliminate and open the cyclopoprane ring to generate another Pd-enolate, this time next to the ketone. Beta-hydride elimination complete the Heck catalytic cycle.
    The author discount the possibility by steric argument. I think detailed mechanistic investigation is required.

  • WillisWill says:

    That’s a pretty clever proposition regular chemist…I wonder though – the cyclopropane containing tetracycle you propose is going to be very geometrically rigid. My models were long ago “donated” to the lab – can that ring system adopt anti-conformation about the bonds that will be involved in the retro-michael?

    We’ll have to see how sharp occum’s blade is…the inital heck may go 6-exo 98% of the time, but with no energetically favorable path in front of it, reverts back to SM – every now and then when it does go 7-endo, it can continue forward to a nice thermodynamically stable product.

    How do the polarization of the exo double bond affect the regiochemistry of attack by vinyl-pd species in general (not just in this case where other factors are in play as well) ie, how does the reaction of 1-palladium cyclohexene with cyclohexenone proceed?

  • regular chem says:

    I, myself, being a Pd-chemist, orgnally also had the notion that carbopalladation follow the usaual pathway, i.e.,carbon, being partially nagatively charged should add to conjugate acceptor in 1,4 -fashion. However, over the years, I have learned that is not neccessary the case. For example, in overman’s asymmetric heck addition to conjugated amides which was applied in a number of natural product synthesis, carbon add to the alfa-position istead of beta-position. This is couterintuitive to common notion that negatvie carbon should add to positive charged carbon. I guessed carbon-Palladium bond is not too polarized.

  • cjdquest says:

    WillisWill: #25 – the beta carbon elimination you propose (reversible 6-exo) has only been observed in strained systems. Take for example Uemura’s Pd(II) alkoxide beta-carbon elimination of cyclobutanol substrates JACS 2003, 125, 8862. Typically, olefin insertion into a Pd(II)-R bond is not reversible.

    That said, the same precedent makes Regular Chem’s beta carbon elimination (or cyclopropane fragmentation) proposal look at least permissible. However, I am more inclined to agree with Overman’s proposal that steric constraints do not allow syn-pari-planar 6-exo migratory insertion. The two quaternary carbons in the tether bump into each other if one builds a model of that transition state.

    How would you propose to test your hypothesis, Regular Chem? If a substrate in which the benzyl ester functionality was replaced with say, a methyl group, and the ‘7-endo’ still goes. would that satisfy?

  • reverse mortgage for seniors illinois…

    sixteens eject steelmaker,enamel.nymphomania …