Home » Still In The RBF

Hygromycin A   

3 August 2009 11,745 views 11 Comments


Donohoe, Flores, Bataille, Churruca. ACIEE, 2009, ASAP. DOI: 10.1002/anie.200902840. Article PDF Supporting Information Group Website

Sugar bashing time!  It’s been ages since I glanced at sugar synthesis, probably because I developed an irrational phobia whilst at Oxford.  Not because of the intricate syntheses, the illogical reactivity patterns, or the ridiculous names (psychose psicose, anyone??) – but for the shear quantities of potassium cyanide used by the Fleet group in their Kiliani Ascension chemistry.  Probably time I got over that, so we’re staying with the folks at the old place, and a bit of tethered aminohydroxylation from Tim Donohoe (hi Tim!).

The synthesis starts with beating the crap out of D-arabinose, differentially protecting, oxidising and then Grignarding.  There’s nothing particularly new, except for an interesting removal of an allyl protecting group.  Normally, I would reach for the palladium cupboard, but they did something pretty different here.  Firstly, a bit of G2 isomerises the allylic group, and then NBS removes it; Donohoe notes that the G2 works by decomposing to the ruthenium hydride species, which does the dirty work.  I mentioned this as a curse of metathesis chemistry a few months ago, but it’s pretty neat when you want it to work.


Next up is the real meat of the chemistry; the groups own aminohydroxylation methodology.  They’ve made some improvements here, certainly since they first examined this chemistry in an Org. Lett. back in 2005, reducing the loading of osmium to 1%.  A good thing too – I hate that stuff even more than cyanide.  However, it’s notable that they’re using potassium osmate, which is a hell-of-a-lot easier to work with, as it’s not volatile. The result is a pretty dandy amino-hydroxylation in high yield.


Next they removed the ‘chiral auxiliary’ (except it’s not really) by reducing opening the THF, and doing a retro-Diels-Alder.  However, you’re gonna have to look in that Org. Lett. to see more about that chemistry.


Lastly, we’ve got the crucial stitching together of the molecule; one third was coupled using standard peptide chemistry, whilst they had something far more interesting planned for the other side.  The idea was to use a Mitsunobu displacement of the hemiacetal hydroxyl, but as usual they’re running into that anomer-selectivity problem.  Their bet was that by choosing the correct protecting group in the C-2 position, and the right conditions, they might get the correct stereochemistry.  However, I’m not sure I understand the reasons for this selectivity. Any sugar-bashers want to enlighten me?


Neat work, folks…

1 Star2 Stars3 Stars4 Stars5 Stars (4 votes, average: 4.50 out of 5)


  • Alexander says:

    There is an extra oxygen in the oxazolidone ring, Scheme 2.

  • Woertink says:

    Seems like was waste of Grubbs2 to do the isomerization. Ru tristriphenylphosphine would have done the trick.

  • milkshake says:

    beta vs alpha selectivity of glycosidation: normally one is inclined to worry about the anomeric effect first but I think in this case of bis-TIPS-protected piece simple sterics prevailed: all three substituents probably wanted to be pseudoequatorial so the easiest access was from the beta face

  • Jim says:

    Ahem…that’s psicose to us sugar chemists.

    And without reading the paper; other than TIPS’s lack of anchimeric assistance I can’t see another reason for the selectivity. I vaguely remember Woerpel did some work on this type of system, but I don’t have the reference to hand.

  • mt says:

    Regarding glycosylation stereoselectivity: The hemiacetal donor was designed so that the alpha-anomer was favored (the beta-anomer would be disfavored due to the bulky TIPS group at the 2-O-position). The authors then hoped that an SN2 mechanism would prevail during the Mitsunobu glycosylation, resulting in inversion at the anomeric center and therefore affording the beta-linked product.

    They mention that the correlation between the hemiacetal alpha/beta ratio and glycosylation stereoselectivity is not absolute (see Table 1), so SN1 (formation of the sp2 oxocarbenium ion, then attack of the acceptor) or anomerization-SN2 could have competed to give alpha-linked product. I suppose the conditions of DIAD/PPh3/Tol/60 deg C favored the desired SN2 attack in this system.

  • Matt says:

    Jim, I believe the paper you were refering to is the following: J. Am. Chem. Soc., 2005, 127 (31), pp 10879–10884.

    As it turns out, nucleophilic additions to the arabinose-derived oxocarbenium ion are essentially unselective (when benzyl protected). However, as milkshake mentioned, the bulk of the substituents used in this synthesis would bias the oxocarbenium ion towards the all-equatorial conformer. “inside-attack” of thIs conformer would lead selectively to the desired beta-product. I would have predicted this selectivity from a purely Sn1 mechanism, however, the addition of additives to promote the Sn2 pathway muddles the analysis…

    • mt says:

      It would’ve been interesting for the authors to discuss whether they did try any other glycosylation methods. With the hemiacetal, a Schmidt glycosylation is one step away and typically proceeds by SN1 via theoxocarbenium cation, so maybe they tried this type of reaction and didn’t get good beta selectivity…?

      • Jim says:

        There are plenty of others as well. The N-phenyl trifluoroacetimidates are great and the Gin dehydrative glycosylations can be excellent. They could make the phosphite, phosphate, phosphordiamidates from Wong/Schmidt (among others). And with a nitrile solvent they could boost the B selectivity.

        With those TIPS groups the glycosyl donor would be nice and reactive (though hindered) and the phenol should be a strong acceptor. I assume they tried traditional glycosylations and didn’t have much luck so fell back on Mitsunobu. Or, they were lucky and the Mitsunobu was the first thing they tried and it worked great.

        Either way, a nice synthesis – paricularly the amino inositol route.

  • will says:

    Are we sure the furanose wants to be pseudo-equatorial? you’ll have substantial 1,2 strain between the two TIPS, perhaps there is greater steric interaction there than the 1,3’s that are present in the pseudo axial conformation…In the pseudoaxial, the alpha anomer (in this case, the hydroxy is in the equatorial position) is more available to the activated phosphonium, which, after Sn2 displacement, leads to the observed product.

    I can’t access the paper, did they investigate small protecting groups, if so, did the selectivity flip?

  • mt says:

    They only discussed the Mitsunobu-type glycosylation on the substrates shown. Maybe we’ll see a full paper or some comments to provide some more info on the glycosylation.

  • Liquidcarbon says:

    I’d suggest you post more often on sugars. :)
    Sugars are dirt-cheap, but many people are afraid of lengthy protection-deprotection sequences (which are often quite ingenious). I’m quite convinced there’s a whole lot more to explore there. Then, there’s plenty that could be done in sugar-based asymmetric reagents. A potenital drawback here is that unnatural sugars are not as available as unnatural amino acids. As brilliantly demonstrated by Shi, accessing the other enantiomer is not a big issue if you know your sugar chemistry:


    I highly recommend this book:

    “You should read your Bible, sirs. You’ll find all kinds of weird shit there. Like did you know Jesus was a Jew?” – Clerks II