Home » Still In The RBF


24 January 2010 25,050 views 62 Comments

Gin, Perl, Ide, Prajapati, Perfect, Duron. JACS, 2010, ASAP. DOI: 10.1021/ja910831k. Article PDF Supporting Information Group Website

When I see targets containing guanidine moieties these days, I immediately think of David Gin, which goes to show how much he owns that motif just now.  Looking back into the murky past of Tot Syn, an earlier post covered Gin’s synthesis of Batzelladine A, which was fairly guanidine-tastic.  This bad-boy only has one, but with a bit of “anticancer, anti-HIV, antifungal, and Ca2+ ion channel blocking [activity]“, that grant-form presumably filled itself, especially at MSKCC.  However, this compound does have previous, and if you had to guess, you’d guess Overman, right?  And you be correct.  And his synthesis features a Biginelli condensation, so how ya gonna beat that, Gin?

Well, through a trio of beautiful cyclisations, as it happens.  The first comes from combination of two quite elaborate starting materials, which were actually easier to come-by than one might imagine.  The skipped ene-yne was produced via a Wittig and a primary-iodide displacement, reducing the compound to two small, asymmetric building-blocks and an acetylene.  The other fragment, containing the precarious-looking carbodiimide, was produced in a second efficient route – three steps from literature material.  One step I hadn’t seen before was the synthesis of the carbodiimide by condensation of a isocyanate onto an azide using a bit of triphenylphosphine to reduce to an amine beforehand.  Neat.  Anyway, one these fragments were in-hand, combination (using a fair-old excess of the carbodiimide) resulted in a pretty efficient cyclisation onto the thioimidate, producing the guanidine-containing aminopyrimidine.

This [4+2] approach has some history, but Gin’s implementation was particularly neat as it worked well on a heavily elaborated system.  This allowed them to quickly remove the interesting silyl-ish protecting group on the imine (ammonium fluoride did that quite nicely without molesting the TBS groups).  Treatment of the now-free imine with gold (III) chloride allowed a second cyclisation onto the nearby acetylene, neatly forming the second pyrimidine, and providing an exocyclic olefin with the correct geometry.  This also stands as a somewhat rare example of metal-catalysed hydroamination of alkynes.  But they were far from done here – that unsaturated system was perfectly set for an addition reaction, which is exactly what happened after treatment with tosic-acid.  Gin doesn’t comment on the stereoselectivity here, so I guess we have to assume that the spirocyclisation was entirely selective for the desired tetrahydrooxepine

Bolting-on the greasy side-chain only took a couple more steps, rounding of a rather neat synthesis, building a complex ring-system with admirable ease.

1 Star2 Stars3 Stars4 Stars5 Stars (6 votes, average: 4.83 out of 5)


  • I am reviewing **someone's** NIH grant says:

    Nice piece of work. Pleasure to read.

  • ch3mical says:

    impressive. crambe crambe.

  • sjb says:

    A dimeric calcium, monocation – something amiss with the super/subscripting? :)

  • Hi Paul,

    Nice job. I am very happy that my isolation work still interests people after all these years (the isolation paper was published in august, 1993.


  • stork naked says:

    gin’s work is quite nice. on another note, who comes up with the taglines for ACIEE articles?

    • anniechem says:

      Yes….who DOES come up with the taglines for ACIEE articles? Sometimes, I get a slight chuckle, but often, I just feel embarrassed for the authors.

      • PotStirrer says:

        “Just another Mannich Monday” was quite good. I also liked the paper.

  • Less Than Three Palau'amine says:

    Baran coulda done it better.

  • Pilky01 says:

    Nice post Paul,

    The carbodiimide formation i think is a fairly traditional method, but it has always surprised me that you can make that disconnection. I think essentially you do a Staudinger reaction but instead of treating the iminophosphorane intermediate (which they isolate in the SI) with acid to get the amino group, you throw in a carbonyl (isocyanate here) and do an aza-wittig type reaction.

  • Ian says:

    superb total synthesis.

    The natural product is not getting left in the fridge either (Nicolaouoaa take note)

  • mahesh says:

    Can anyone please help how to make
    ” (E)-ethyl 6-bromo-5-oxohex-2-enoate ”
    I’m trying many methods… Yet no success…

  • 23l33t says:

    Did you folks see the Christina White Science paper last week? Take a look at the SI, in particular, the 2D spectra appear somewhat strange. I’ll leave it at that:


  • chantal69 says:

    I just checked out that Christina White SI and I happened to look at Page S70. At first I thought that they just erased handwritten notes and lines on the spectra, but I think something abnormal is going on.

    I am definitely making some assumptions here, but….

    There is a white splotch where the C8 to Hg/Hg’ correlation should be.
    The C12 to He correlation is also whited out (somewhat questionable).
    Around 210ppm/2.3ppm half of the signal has been deleted.
    Around 215ppm/1.1ppm part of the signal has been deleted.
    Across 1.1 ppm “noise” has been deleted which should be present due to the strong intensity of the methyl peaks (perhaps questionable).
    Is there a missing TOCSY peak for He/C9?

    Somebody PLEASE tell me that I am wrong!

    I guess they are trying to hide something – maybe if their NMRs were cleaner this would be unnecessary.


    This is a disgrace to “science.” Although, I still think that JJ LeClair was worse. Different (but same level) was NaH. IMHO.

    • Johnny Bravo says:

      There’s definitely been some cleaning up going on. No one would have noticed if they didn’t scan the spectra in by hand. What is this, 2005?

    • Martyn says:

      Move along, nothing to see here. It’s obviously just deleted pencil marks, probably at some reviewer’s behest – almost all of the whiting out is in straight lines from crosspeaks to the axes, or in blank spaces where people typically write or draw structures. It’s a mysteriously poor Microsoft Paint effort though, considering the photoshoppery on the White group’s website.

      • Buckyball says:

        Is it really that hard to go down to the NMR room and reprint the spectra? Even if they are not hiding any data, which we cannot be sure of in the first place, using whiteout like this just looks incredibly lazy and unprofessional. I can’t believe that none of the referees looked at these spectra and raised some flags.

  • wowow says:

    one word: xtina-whiteout

  • optional says:

    I cant download the SI. The file appears corrupt

    • Ian says:

      Same – the SI has been taken offline apparently

      Quick repair job

      • TMSS says:

        Just downloaded it. WOW that is a major repair job on atleast 10 spectra. I was asking myself how the paper got in to science at all though, without even reading the SI…

        Jake Yeston at Science probably handled this manuscript (I know him from my berkeley days) – he is good friends with Jacobsen. That explains alot…

        • ORLY? says:

          Jake Yeston at Science probably handled this manuscript (I know him from my berkeley days) – he is good friends with Jacobsen.

          I don’t know Jake very well (I’ve only met him once or twice), but this comment is ridiculous – lemme see if I can follow your logic: Jake is BFFs with White’s old post-doc advisor, so he decides to publish her paper as a favor to Jacobsen? Hmm… Not sure if I buy that – especially since I’ve heard that Jacobsen and White aren’t the best of friends…

  • I am reading your Blog from last few years, it is quite informative

    In one of your post  “:Sporolide B”  28 th Feb 2009 I have seen the
    structure drawn in K C Nicolaou style

    for one of my presentation i am interested in Drawing my structure in KCN

    I request you to send me necessary information / Protocol which will help
    me in drawing these structures

  • Will says:

    The SI seems to be back online – I find the same splotches when I look at S70;

    Before anyone goes flying off the handle about scientific fraud etc.. a couple of points. The cpd in S70 is 3-oxo-sclareolide, a known compound. White provides, proton, 13-C, IR, HRMS and optical rotation in addition to the HMBC; the proton spectra is reproduced at S67; assuming she has not fabricated this data, and it does match the reported values for the compound, then the compound was in fact made. Therefore, the reported transformation did in fact take place

    The deleted peaks correspond to long range HMBC couplings. Perhaps they decided to erase them for clarity of spectra. Is this the best choice? No, but maybe the grad student/postdoc made a bad decision, maybe C White made a bad decision.

    To call this “outrageous” or “disgrace” to science is absurd without more evidence. There’s not a shred of evidence to suggest the reported compound was not made, or that the reported transformations don’t work

    Methinks the original poster to this, and perhaps the kind person who provided such detailed analysis is not simply an innocent bystander, and might have an ax to grind

    • antiaromatic says:

      I agree that it doesn’t appear to be straight out fraud. It seems like for the most part, they wrote on their spectra while analyzing them, and when it came time to submit, they digitally erased those markings. I do wish they would’ve just reprinted the spectra, but I don’t know how the NMR facility is there, and since I think Chen has already graduated, it may be that they just didn’t bother trying to get the spectra again. Not that this is a good reason not to reprint the spectra, but overall, it doesn’t appear that anything was fabricated.

      • Tok says:

        They did have data limits of 20mb per user on each NMR as of 4 years ago, so it’s possible that digital backups were either not made or lost. The SI does seem to be down again so I can’t look, but I can imagine a scenario where they realized the digital copies were gone and only had heavily marked-up paper copies that they decided to remove the writing from.
        chantal69, let’s wait a bit before we break out the torches and pitchforks eh?

        • Andrei says:

          20Mb per collection of data is more than enough

          • LiqC says:

            Try to save a processed 2D spectrum in MestreC or MNova, it’s 5 MB right away, think 50 if it’s phase-sensitive like HSQC

          • Andrei says:

            I said collection of data, not processed one….do see the difference, or what was tok implying?

          • Tok says:

            It’s 20mb total, not per collection. So if you’re not johnny-on-the-spot with FTPing your data off the server regularly and backing it up somewhere safe, it can get deleted for good.

            My point is that sloppy data backup techniques do not equal academic fraud.

    • 23l33t says:

      I honestly just wanted to hear some other opinions. No vendetta against the White group here.

  • stop says:

    For a while I thought it was a dumb mistake, and I (personally) still don’t think anyone in the White lab intended fraud. However, in discussing this with lab-mates, someone asked me how many peaks I’ve digitally removed from my spectra… the answer: none. Its a form of manipulation of data, one that isn’t accepted in the community (it would be another story if this type of “NMR cleanup” was acceptable).

    The point is, it appears that in addition to pencil/pen marks, some of the “noise” was removed, and perhaps even some peaks that should be there (like chantal points out).

    Ethically speaking, even if the important data isn’t altered, it calls the whole paper in to question. It appears to me that it was rushed in the White lab, and not rigorously checked by reviewers.

    • The Next Phil McGroin says:

      And”stop” how many Science papers have you been first author on?

      Talk about much ado about nothing? The white-outs appear to block out annotations, but not actual spectra. How could anyone rationally draw the conclusion from a few white splotches on a couple of spectra that FRAUD has been committed. It is a statement on the state of our field that “pg 70 of the SI has a white splotch on it” constitutes an accusation of scientific malpractice. We are really out for our colleagues blood, huh!

      As someone who has never published in Science, but hopes to, I sincerely ask that you ignore the poorly hand drawn structures and politically incorrect caricatures representing Nobel Laureates in compromising positions I draw on my 2-D NMR’s.

      • stop says:

        …oh I lost count after my 5th Science paper…

        But honestly, I’m not claiming fraud, I’m claiming sloppiness and noting that there were a few missing peaks that probably got removed on accident (while removing hand written notes on the NMR presumably). You don’t have to have published in Science before to know those spectra look suspect though. Even a high school student wouldn’t hand in a formal, final report with white-out on it.

  • itipiti says:

    how about any control experiments in the white group…?

  • chantal69 says:

    “There is a white splotch where the C8 to Hg/Hg’ correlation should be.”

    This is only two bonds away. The same carbon can see correlations three bonds away. Riddle me this…..

    The only axe to grind I have is one with scientific integrity – not the authors in question. I think this is especially important given recent lapses in the review process.

    The field of science is regulated with due cause based upon what individuals contribute. If individuals choose to pepper the literature with sub-par data, sooner or later they will fall on their own sword.

  • sssssss says:

    what if phil baran submitted something of that sort? that would outrage a lot of people… but oh wait, christina white is no phil baran.

  • anon says:

    You guys suck, ragging on an awesome paper because some spectra deep in the SI look a little messy. I bet if we look we can find typos too then we can get the burning torches and pitchforks.

    If we all stopped nitpicking so much maybe one of us could publish something half as important as this.

    (not in the White lab…if only i was that successful)

    • Dalibor Sames Jr. says:

      Awesome paper is build on solid data. This is unacceptable. It does not look like fraud. But it IS fraud.

  • European Chemist says:

    Still cannot open the SI and have been trying since yesterday (“this file is damaged and cannot be repaired”).

    Anyway, would be VERY surprised if fraud was involved. Think about it: the real frauds usually simulate spectra (cf. JJ LaClair) or deliberately flatten the baseline, remove impurities, etc.
    I do agree however that if we are going to include reproductions of spectra in the SI it’s because there should be a purpose and they should look like untouched, clean spectra of pure compounds.

    I for one have always been against such stringent SI rules. What good is it to have a copy of the proton NMR of a compound? The baseline can be made to look flat through a myriad of clever manipulations (probably including Photoshop). The spectral window can be shortened to exclude aromatic impurities (for example). Whenever spectra have really tiny peaks I suspect that it was just so “the whole thing would look cleaner”. Their main usefulness would be for comparison purposes whenever you make a reported compound – but the usual tabular representation is more than enough.
    In the end of the day, tough requirements for SI just forces the real cheaters and frauds to work a wee little bit harder (and gives them an even better reason to do so). Which they will gladly do. My opinion.

    • ch3mical says:

      its not a corrupted file, just right click and save link

      • European Chemist says:

        Thanks. Finally checked the file.

        Is THIS what you are all getting so excited about????
        For God’s sake it’s obviously Paint or Photoshop removal of pencil markings or drawn structures on the 2D spectra. Was maybe even done/requested by the Science Editorial Office (?).


        Let’s focus on chemistry, shall we?

  • The Next Phil McGroin says:

    suppose Eurochem is right and the editors saw some penciled structures or other scribbling and asked the authors to remove it from the SI…is that still a misconduct?

    what if coffee spilled on it, and the author was embarrassed, is that a misconduct?

    seriously, there should be some sort of evidence or scientific basis for accusations of these sort, even in an anonymous setting such as this.

    • PC says:

      if an editor asks the authors to change their SI using photoshop, he shoule be fired!!!!!

  • UBChem says:

    rabble, rabble, rabble, rabble….my point, rabble rabble rabble rabble…..your point, rabble rabble rabble rabble.

  • anon2 says:

    Given the low signal-to-noise ratio in some of the spectra (e.g., the carbon spectrum on the x-axis of the HMBC on pg S70), I’m guessing the authors just ratcheted down their threshold to get larger, clearer cross-peaks. The resulting noise that accompanied those peaks might be what’s covered by white-out. It would make sense that the noise would be in horizontal and vertical alignment with the peaks we can see.

    I don’t agree with data manipulation, but I don’t think it’s maliciously fraudulent.

  • […] … but it’s interesting to see how the other half lives.  Who couldn’t love: Crambidine though it seems to have gotten them a bit riled up.  If nothing else, the sheer complexity reminds […]

  • Ian says:

    Why is anything erased from any spectra? If the spectra is not good enough run it again

    If someone has drawn all over it grab the electronic file?

    I refuse to believe that a university that hosts Scott Denmark and John Hartwig does not provide its students with the facility to save all spectra computationally.

  • European Chemist says:

    Ian has a point.

    But, as it is apparent from this discussion, NONE OF US KNOWS EXACTLY WHAT HAPPENED. And so, claiming “scientific misconduct” or that “it’s humiliating for the whole synthetic society that we can only present such low quality material to Science” is really WAAAAY over the top. Until we know what really motivated that we can only speculate. Which apparently is what (some) Organic Chemists like most.

    • Ian says:

      Some of Corey’s spectra are probably still on hard disk somewhere there

      Tippex is not a great idea

    • _ says:

      Presenting your data in as clear, concise, and direct a manner as possible is a part of scientific conduct. The goal in reporting scientific findings is to ELIMINATE speculation.

      The fact that we are even having this conversation means that there’s something wrong with the data. That doesn’t mean that it is fraud, but these questions should never be in your mind when pulling up ANY peer-reviewed SI.

      • anon says:

        Sorry to burst your bubble but ‘is this correct or not’ should always be on your mind when you read any scientific work.

        SI is there so you can draw your own conclusions about the statements in the paper.

  • tntjdrn says:

    great synthesis!
    but i dont know how they could deprotect CH2TBS group of the amine using fluoride. i think only TBS group can be deprotected.
    plz tell me the mechanism.

    • Martyn says:

      It’s a CH2OTBS – the TBS comes off leaving an NCH2O- aminal anion, which loses formaldehyde to give the deprotected amine.