Home » Still In The RBF

Polyanthellin A   

14 July 2009 11,551 views 20 Comments

polyanthellin_a

Johnson, Campbell, Qi. JACS, 2009, ASAP. DOI: 10.1021/ja904136q. Article PDF Supporting Information Group Website

Sorry for the lack of blogging, folks – I’ve been feeling a bit under the weather, but I assure you, it’s not H1N1…  However, it seems like the res of the world is getting on with it, with several nice syntheses in JACS just now.  Lyconadin A is headed for Chemistry World, later this month, whilst plicatic acid is line behind polyanthellin A. Hopefully, sticking the latter into the Tot. Syn. search box (operated by trained hamsters) will return this post on ‘(6Z)- and (6E)-Cladiellin Diterpenes‘ by Deukjoon Kim, an acknowledged master of this family of natural products.  This sets the bar rather high – so lets look at how this fresh challenger Jeffrey Johnson deals with the target.

The target is split into two components, ready for a rather tasty unification, but first they gotta make them.  The more complex cyclopropane-containing fragment was built using some really interesting chemistry.  First up was an organocatalytic Michael addition of isovaleraldehyde into methyl vinyl ketone using a fairly standard proline-based catalyst, aided with a bit of a catechol to boost reaction rate and yield (another example is in this Baran paper).  The remaining aldehyde was then Wittig’d to bolt-on a diene, and set them up for a bit of cyclopropanation.  Initial attempts to do a direct cyclopropanation were unsuccessful, so they used an interesting Corey(and Myers)-abetted protocol, which initially diazotises the 1,3-dicarbonyl, and then using a copper(II) salen complex, does an insertion into the proximal alkene.

polyanthellin_a_1

The other fragment, an aldehyde, was built easily; asymmetry was introduced via an SAE, and the resulting epoxide opened using in-situ-developed allyl cuperate.  A few more steps including a 1-c homologation returned the desired aldehyde, ready for coupling.  This Lewis-acid mediated process is a formal [3+2] cycloaddition required a pretty exotic reagent, as more commonly used LAs resulted in aldol couplings of the aldehyde.  For success they used an aluminium based catalyst, based on some work by Yamamoto; it did the job nicely, but we’ll need to wait for the promised full-paper to the goods on it’s virtues.

polyanthellin_a_2

As any reader will see, we’re now perfectly set for a bit of RCM to close the medium ring.  This was done in high yield by using dilute conditions, and chlorinated solvents (70%), but not without the group enduring a fairly frequent problem.  Y’see, those metathesis catalysts (be they Ru or Mo based) often take a liking for other chemistry when attached to your lovely terminal alkene.  A common problem is de-methylenation – loss of methylene units – from the SM, generating a new, similar compound which can they react further.  In Johnson’s case, this lead to eight-membered product, but in other systems can result in a damn mess.  It’s be great to hear your experience of this thorny problem – including the additives to prevent this problem, such as acetic acid and titanium isopropoxide.  In Johnson’s case, simply using DCM (presumably thoroughly de-methanol’d) did the job… others are less lucky.

polyanthellin_a_3

The final intermediates are starting to look fairly similar to those encountered by Kim; looking back at that paper, Kim was able to do a rather tasty oxo-mercuration to form both the ethereal bridge and the tertiary alcohol.  Johnson hope to do a similar reaction, forming the ether from ‘the-other-side’ (if you get me), but encountered only a 10% yield.  He rescued the situation via a slightly more labourious route; iodolactonisation, followed by oxo-mercuration delivered the desired oxygenation, whilst a spot of radical tin reduction removed both the extraneous heavy atoms.  Only acetylation was required to finish the target, in a synthesis that stands up well against Kims.  Good job, sir.

polyanthellin_a_4

1 Star2 Stars3 Stars4 Stars5 Stars (12 votes, average: 4.08 out of 5)
Loading ... Loading ...

20 Comments

  • milkshake says:

    A small typo correction, the endgame sequence its properly called iodoetherification (not iodolactonization since there is no lactone).

    This is a clever scheme – but risky. I am glad that it worked out so well. The one drawback is that they needed to use an excess (3 equivs) of the aldehyde piece to get a good yield in the [3+2] step. These beta-OR substituted aldehydes are pests – they love to eliminate to conjugated aldehydes, I wonder how many iterations it must have taken them to arrive at the modified Yamamoto MAD catalyst. This is an under-used LA catalyst – (the hassle of ending up with a load of heavy substituted phenol in the product) – but it can do spectacular things with aldehydes and their equivalents because of the steric shielding. I think Johnson must have been aware of this MAD option since his thesis work with Evans.

  • cascade says:

    Are these Grubbs-catalysed isomerizations really that “common” for systems such as these (without allyl ether/amine)?

  • joel says:

    Overman tried the first RCM approach to these molecules back in ’97 using the Schrock catalyst, and got nothing but the 8-membered product.
    (Tet. Lett. 1997, 8635)
    Hoppe has also tried this method (using Grubbs II), and gets about 3:1 of the 9- and 8-membered rings. (Angew. Chem. 2008, 1654) So obviously it’s a pretty common problem.
    Crimmins has probably had the most success using RCM on these targets, but his approach is to form the 9-membered ring first, and then close up the 5 and 6 rings using a bit of Diels-Alder chemistry. (JACS 2004, 10264) So possibly these two rings introduce a bit of extra strain into the molecule which makes metathesis a bit harder.

  • European Chemist says:

    This Jeffrey Johnson guy is putting out some pretty tasty chemistry… I wonder how long NC-Chapel Hill can hold on to him.

  • Tot. Syn. says:

    Interesting stuff in Org. Lett. today – two papers on the LA catalysed synthesis of spirocyclic oxindoles from N-methyl isatin dimethylketal, back to back:

    James Panek: 10.1021/ol901202t
    John Porco: 10.1021/ol901201k

    Both at Boston Uni, and both referencing each other and thanking each other for ‘helpful discussions’.

    The actual chemistry differs slightly, but it an interesting coincidence…

  • ... says:

    A nice approach to the nine-membered ring. It seems as though a few reports for cyclononene formation via RCM have popped up recently. An interesting alternative approach has been made by the Williams group at Indiana University in a total synthesis of 4-hydroxydictyolactone. They utilize an B-Alkyl Suzuki coupling to close the medium-sized carbocycle of the natural product.

    JACS 2009, 131, 9038-9045.

  • Pilky01 says:

    The titanium-Wittig reaction is interesting and the referenced paper is quite involved. In the absense of the titanium would the normal Wittig give poorer E:Z ratios? is that why this protocol is being used? Or is it some complication when making the ylide of this substrate? Can anybody just clarify this for me please?

    • Johnny Bravo says:

      It has to do with the poor Z:E ratio obtained when doing a traditional Wittig. The reason for obtaining only the Z is pretty clearly laid out in Figure 1 of the referenced paper. The Z would be necessary for the vinyl group to end up on the concave face after cyclopropanation, which I imagine is not easy to do using other methods.

  • Kiddie Khem says:

    Anyone else see bringing in the furan with the rest of the rings bound to it and having a an alkene on the bottom of it, used for a diels alder to make the last ring and those sterocenters?? (PS i apologize for grammar and hazy explanation)

    • Atticus says:

      That would make for an awfully crappy dienophile that certainly won’t work in an intermolecular sense. You’d have to throw some EWGs on there and then play with the functional groups, at which point it would probably become a lost cause.

  • xyz says:

    I thought of checking the wisdom of the audience to see as to which one of the two is easier: the reaction/methodology development or the ‘total synthesis, considering that the resources and the size of the group aside. Which one has more impact and significance in the synthetic community? How does the (classical) organic synthetic community respond to ‘nano’thingy and survive the funding scenario?

  • Morts says:

    Nice paper!! Cycloadd ftw!

    Current Issue of Tetrahedron has an interesting author line up too!!!

  • earth23 says:

    Jeff Johnson is pretty legit, especially after this paper. Target reminds me of Vigulariol which was talked about on this site last year.

  • ZZZZZ says:

    zzzzz

    …I just woke up to attend a lit presentation on this. Good job Roger…

    zzzzz

  • stereoelectronics says:

    Very nice synthesis! The 3+2 cycloaddition looks like a powerful tool to construct the hexahydroisobenzofuran core. Their conditions for the RCM improves a lot what has been reported so far for the construction of bi- and tricyclic cyclooxanonenes (compare Jung, Tetrahedron 2003, 59, 2729) even tough the late stage RCM for this natural products has been shown first by Hoppe in the vigulariol synthesis. Grupps has reported that benzoquinone can also be used to prevent the isomerization because it quenches the Ru hydride species (J. Am. Chem. Soc., 2005, 127 (49), 17160). Congratulations!