About this Author
DBL%20Hendrix%20small.png College chemistry, 1983

Derek Lowe The 2002 Model

Dbl%20new%20portrait%20B%26W.png After 10 years of blogging. . .

Derek Lowe, an Arkansan by birth, got his BA from Hendrix College and his PhD in organic chemistry from Duke before spending time in Germany on a Humboldt Fellowship on his post-doc. He's worked for several major pharmaceutical companies since 1989 on drug discovery projects against schizophrenia, Alzheimer's, diabetes, osteoporosis and other diseases. To contact Derek email him directly: Twitter: Dereklowe

Chemistry and Drug Data: Drugbank
Chempedia Lab
Synthetic Pages
Organic Chemistry Portal
Not Voodoo

Chemistry and Pharma Blogs:
Org Prep Daily
The Haystack
A New Merck, Reviewed
Liberal Arts Chemistry
Electron Pusher
All Things Metathesis
C&E News Blogs
Chemiotics II
Chemical Space
Noel O'Blog
In Vivo Blog
Terra Sigilatta
BBSRC/Douglas Kell
Realizations in Biostatistics
ChemSpider Blog
Organic Chem - Education & Industry
Pharma Strategy Blog
No Name No Slogan
Practical Fragments
The Curious Wavefunction
Natural Product Man
Fragment Literature
Chemistry World Blog
Synthetic Nature
Chemistry Blog
Synthesizing Ideas
Eye on FDA
Chemical Forums
Symyx Blog
Sceptical Chymist
Lamentations on Chemistry
Computational Organic Chemistry
Mining Drugs
Henry Rzepa

Science Blogs and News:
Bad Science
The Loom
Uncertain Principles
Fierce Biotech
Blogs for Industry
Omics! Omics!
Young Female Scientist
Notional Slurry
Nobel Intent
SciTech Daily
Science Blog
Gene Expression (I)
Gene Expression (II)
Adventures in Ethics and Science
Transterrestrial Musings
Slashdot Science
Cosmic Variance
Biology News Net

Medical Blogs
DB's Medical Rants
Science-Based Medicine
Respectful Insolence
Diabetes Mine

Economics and Business
Marginal Revolution
The Volokh Conspiracy
Knowledge Problem

Politics / Current Events
Virginia Postrel
Belmont Club
Mickey Kaus

Belles Lettres
Uncouth Reflections
Arts and Letters Daily
In the Pipeline: Don't miss Derek Lowe's excellent commentary on drug discovery and the pharma industry in general at In the Pipeline

In the Pipeline

« Nobel Season 2010 | Main | Open Thread »

September 28, 2010

Wisterone: A Structure I Can't Believe

Email This Entry

Posted by Derek

You don't see an awful lot of chemistry publications from Vietnam. So in a way, I'm reluctant to call attention to this one, in the way that I'm about to. But it's in the preprint section of Bioorganic and Medicinal Chemistry Letters, and some of my far-flung correspondents have already picked up on it. So it's a bit too late to let it pass, I suppose.

The authors isolate a number of natural products from Wisteria (yep, the flowering woody vine one), and most of them are perfectly fine, if unremarkable. But their compound 1 (wisterone) is something else again.
Man, is that thing strained. Nothing with that carbon skeleton has ever been reported before (I just checked), outside of things that you can draw as part of the walls of fullerenes. I have a lot of trouble believing that this compound exists as shown - and if it does, then it deserves a lot more publicity than being tossed into a list inside a BOMCL paper - even though that journal is now getting a reputation for. . .interesting structural assignments.

This thing could get you into Angewandte Chemie or JACS, no problem. But the authors don't make much of it, just calling it a new compound, and presenting mass spec and NMR evidence for it. The 13C spectrum is perfectly reasonable for some sort of para-substituted aryl ring, but this compound would not give a perfectly reasonable spectrum, I would think. Surely all that strain would show up in some funny chemical shifts? Another oddity must be a misprint - they have the carbon shift of the carbonyl as 190.8, which is OK, I suppose, but they assign the methylenes as 190.8, which can't be right. (The protons come at 4.48).

No, I really think something is wrong here. I don't have a structure to propose, off the top of my head (not without resolving that weirdo methylene carbon shift), but I don't think it's this. Anyone?

Update: just noticed that this is said to be a crystalline compound, melting point of 226-228. I find it hard to imagine any structure like this taking that much heat, but. . .it's a crystal! Get an X-ray structure. No one's going to believe it without one, and BOMCL should never have let this paper through without someone asking for at least that. . .

Comments (46) + TrackBacks (0) | Category: Analytical Chemistry | Chemical News


1. partial agonist on September 28, 2010 10:57 AM writes...

"Dreiding model cleanup in aisle 3"... there's bent tubes everywhere

Maybe it could be like a dimer or trimer of that, where it closes to a big ring we can call "doughnuttium" with biphenyl, CO-CH2CH2-CO, biphenyl, CO-CH2CH2-CO, biphenyl, CO-CH2CH2-CO, the back to the first biphenyl.

But to my knowledge a plant never made anything so wacky, and why?

My best guess is that it's a total drylab

Permalink to Comment

2. Anonymous on September 28, 2010 11:00 AM writes...

Without looking at the paper and the spectrum: wouldnt a cyclophane-structure be an option? the two benzenerings parallel to each other bridged over a -COCH2CH2- group?

Permalink to Comment

3. Scott on September 28, 2010 11:01 AM writes...

Looks like the Flying Spaghetti Monster to me, but what do I know, I'm just a molecular biologist.

Permalink to Comment

4. Anonymous on September 28, 2010 11:11 AM writes...

get a sample of it and take an X-Ray

Permalink to Comment

5. Wavefunction on September 28, 2010 11:20 AM writes...

I can't believe it too. I just built the structure and minimized it in a couple of molecular mechanics force fields, and the benzene ring is bent out of plane by a freaking 45 degrees.

Next, I deleted the bond between the two rings and minimized the structure. The rings snap away from each other and the distance between the two ring carbons increases to 5.2 A (previously it was 1.5 A!), way too long for any bond. Plus, the energy drops by a whooping 57 kcal/mol.

Finally, I calculated the energy difference between a single bent, strained benzene ring and the unstrained planar one, and- surprise- it's about 35 kcal/mol.

That baby's strained.

Permalink to Comment

6. HappyDog on September 28, 2010 11:25 AM writes...


The calculated potential energy of this monster is 171 kcal/mol!

I don't think so...

Permalink to Comment

7. anchor on September 28, 2010 11:36 AM writes...

#4 spot on! The fact that they have isolated lot of lignan based compound, allows me to speculate that perhaps the product may a dimeric compound derived from p-hydroxy-dihydrocinnamic acid (an essential precursor to many lignans). The large membered bis-lactone is a possibility

Permalink to Comment

8. Zoltan on September 28, 2010 11:44 AM writes...

(At least for the NMR-Data) it must be an error in the text. Table 1 cites 30,7 for the methylene group and 190,9 for the carbonyl.

Still, would love to see the crystal structure to this one!

Permalink to Comment

9. mas3cf on September 28, 2010 11:52 AM writes...

I second the cyclophane structure... just to be clear, in my mind it would be two phenyl rings parallel, and then two (C=O)CH2 bridges connecting them, with each phenyl connected to a (C=O) of one bridge and the CH2 of the other.

Permalink to Comment

10. none on September 28, 2010 11:56 AM writes...

Is it just me or does it look like a bra? A joke from the authors maybe?

Permalink to Comment

11. Hap on September 28, 2010 11:59 AM writes...

Those methylenes should have 1H shifts < 0 because they're poised over the benzene rings and stuck in the shielding area. Even with the shifts from the carbonyl, the protons should be notably upfield. I don't know where the 13C methylene peaks should be. In addition, the thing should probably show an awful lot of hindered rotation in the linker, which should show up in the 1H NMR.

The alternative cyclophane in comment 8 makes more sense, although there should still be some funny things with diastereomers and those carbonyls should be pretty strained and probably further downfield than normal.

How did this get by the review process again?

Permalink to Comment

12. anchor on September 28, 2010 12:08 PM writes...

#9 correct and the model (chemdraw etc.) looks OK too!

Permalink to Comment

13. JAB on September 28, 2010 12:43 PM writes...

A couple of notes: the high res mass spec value measured is pretty far off from the calculated, so I don't think you can trust the molecular formula, also, note that the corresponding author is not from Vietnam, they are Korean, for what that's worth (better chem in general in Korea, for sure), and ref 12 that they cite is a Clardy crystal structure of a twice-bridged biphenyl with different macrolide attachment. The paper doesn't reflect well on the BOMCL review process, that's for sure.

Permalink to Comment

14. Stiv on September 28, 2010 12:50 PM writes...

"Another oddity must be a misprint - they have the carbon shift of the carbonyl as 190.8, which is OK, I suppose, but they assign the methylenes as 190.8, which can't be right. (The protons come at 4.48). "

It is a typo. The listing in Table 1 further down in the paper shows the carbon shift of the methylene carbon as 30.7

Permalink to Comment

15. at on September 28, 2010 12:52 PM writes...

I wonder if the authors have the correct mass. All that I've seen come out of ESI+ are [M-H/Na/K]+ molecular ions, not [M+].

What I'd also like to see, is journals requiring the actual NMR data as additional material.

Permalink to Comment

16. Stiv on September 28, 2010 1:06 PM writes...

#9 mas3cf is right, it is likely the cyclophane - prepared by Dewhirst and Cram back in 1958.

See structure XX from JACS Volume 80, Pages 3115-25, 1958

Permalink to Comment

17. mad on September 28, 2010 1:11 PM writes...

I can only see the paper preview but Derek drew 2 rings but the pic I see in theo paper has one ring open.

Permalink to Comment

18. Hap on September 28, 2010 1:29 PM writes...

The reported mp for their compound is at least 40C off that reported for XX.

Permalink to Comment

19. Anon on September 28, 2010 1:54 PM writes...

My model kit done broke.

Permalink to Comment

20. Zoltan on September 28, 2010 1:56 PM writes...


mad, the picture does indeed look strange in the preview, but the paper itself has the one Derek drew.


Let's see if anybody else reads the comments before posting? ;)

Permalink to Comment

21. Org Lett Reader on September 28, 2010 1:57 PM writes...

My head a-splode.

Permalink to Comment

22. emmitt smith on September 28, 2010 2:15 PM writes...

That there molecule done got blowed up. And don't be makin' them Korean & Vietnamese chemists no escape goats.

Permalink to Comment

23. anon the II on September 28, 2010 2:56 PM writes...

Sometimes I feel bad because I don't keep up with the literature as well as I should. Every once in a while, an article comes along that makes me feel better about it. This is such an article.

Permalink to Comment

24. Gummy on September 28, 2010 3:45 PM writes...

A comment on the BMCL review process:

I just received an invitation to review a BMCL paper due to my "expertise in the area."

Problem is, although I have worked briefly on the general target class, that information is still confidential. In addition, I have published neither a single paper nor a single patent application in the target area, so my expertise is quite limited or non-existent to the rest of the scientific community.

I don't have a concern with my overall ability to effectively review the paper, but I'm not arrogant enough to think that there are not plenty of other med chemists with more expertise in the area who are more appropriate reviewers. And I'll let BMCL know that.

My concern is that there is either too large of a load for more appropriate reviewers, not enough experts in the area are doing their scientific duty to review papers, there is just too much material coming out of BMCL which effectively reduces the quality overall to the journal, or their review system is broken much more than I thought.

I haven't seen as much of an issue with J Med Chem with regards to this. Has anyone else? Is the whole system in general broken?

Permalink to Comment

25. Aspirin on September 28, 2010 4:45 PM writes...

I wonder if BMCL outranks Tet Lett in abysmal editorial and reviewing policies.

Permalink to Comment

26. A on September 28, 2010 7:02 PM writes...

Maybe they got the phenyl substitution pattern wrong? Imagine the same structure, except that both phenyls are ortho (rather than para) substituted. The C(O)-CH2-CH2-C(O) chain would be part of an 8-member ring.

Permalink to Comment

27. AlchemistOrganique on September 28, 2010 7:10 PM writes...

Hey know, perhaps bond angles & lengths are different in Vietnam. Maybe botanists working for the NVA were trying to engineer a hazardous plant that could incapacitate enemy troops with the release of ring strain energy. As we continue bashing BMCL & Tet Lett, let us not forget the thorough peer review of La Clair's Hexacyclinol synthesis by the ACIE editors!

Permalink to Comment

28. barry on September 28, 2010 8:53 PM writes...

a structure with the phenyl rings ortho-substituted rather than para- would have more lines in the 13C NMR. That sort of symmetry element should be hard to botch.

Permalink to Comment

29. PlatoMolloy on September 28, 2010 9:45 PM writes...

Given the structure reported, I would not doubt the authors ability to botch anything.

Permalink to Comment

30. InfMP on September 28, 2010 11:44 PM writes...

Tetrahedron Lett. and BOM-Cl should abolished from this earth, maybe even taking Synlett with them.

I don't care if you're in industry and you didn't bother to take 13C or HRMS. Cry a river.
And no one should be publishing anything from a sketchy postdoc who vanished from the lab back to a faraway land, leaving only his notebooks and his reputation for belittling others.

Many young chemists will agree that IR and mp are a waste of time and should not be required anymore.
Clean 1H, 13C and HRMS plus experimentals. 2D spectra should be required when needed. Old people like EA so maybe that should be required. That's cool - I've never done it though.

Permalink to Comment

31. John on September 29, 2010 12:40 AM writes...

In Vietnam they have a superstition that the first customer through the door determines how the rest will go that day. It may apply to their papers as well.

Permalink to Comment

32. Pallas Renatus on September 29, 2010 1:15 AM writes...

Hah! All I know is, if this stuff exists, I want a vial of it to play with. Too bad it most likely doesn't and probably never will.

Permalink to Comment

33. Petros on September 29, 2010 4:52 AM writes...

I wonder if the problem with TL and BMCL is the use of regional editors/reviewers.

I don't object to the journals per se but TL has been steadily getting poorer and inclusion of this sort of thing, with only micromolar activities reported, in BMCL isn't my idea of med chem.

I'm an occasional reviewer in areas I know enough about

Permalink to Comment

34. Hap on September 29, 2010 11:38 AM writes...

On large scale (not natural product synthesis late stage), mp can be useful - see above for example. It should be a decennt indicator of purity. IR is pretty easy to get, and is indicative of several functional groups. It's not like most groups don't have an IR lying around. UV is less useful except in certain circumstances.

This would have been one of the times a 2D-NMR of some (or multiple) sorts would have been useful, though the limited data for what they do have probably should have given somebody pause.

27 - ACIEE, JACS, and BMCL have all been whacked here with some frequency (between Hexacyclinol and the diaza[12]annuleneXXXXXXXXXXXXXXXZincke pyridine debacles, ACIEE got rode pretty hard.)

Permalink to Comment

35. Norepi on September 29, 2010 5:45 PM writes...

Aye freakin' caramba on this one.

@ #11: I was thinking the same thing about the shift anisotropy. 4.48 is wayyyyyy too far downfield. I'd be surprised if it were even 2.48. I've had alkyl chains (no carbonyls, sitting in the shielding cone) come out at ca. 0.

Even for an acetophenone-type structure, it's too far downfield.

Permalink to Comment

36. Anonymous BMS Researcher on September 29, 2010 9:05 PM writes...

I'm on the Bio side so wouldn't know a plausible structure from an implausible one, but in my own field I've certainly seen stuff that should not have been published. My grad school prof showed me a paper some journal had just sent him, and his review: "I have been asked to review this manuscript by six different journals now, and I do not wish to see it a seventh time." He also used showing the impossibility of this paper's premise as a homework exercise for one of his undergraduate classes.

Permalink to Comment

37. InfMP on September 29, 2010 11:47 PM writes...


Our dept (industry) does not even have a mp or IR! But we have the equipment to separate 30 g of racemic into the ents!
It's such trouble to try to find one to use everytime we want to publish that we usually take the risk. I don't like taking that risk, but besides third world countries, wouldn't you agree that NOBODY reading SI checks the IR or mp??
People read the NMR and mass and I know people who just typed random crap for their IR stretches throughout grad school. People with >5 JACS!

Permalink to Comment

38. Petros on September 30, 2010 8:40 AM writes...

Could someone list the NMRs we don't all have access to the article

Permalink to Comment

39. partial agonist on September 30, 2010 12:48 PM writes...


proton: 4.48 (s), 7.76 (d, J= 8.2), 8.11 (d, J= 8.2)

13C: 30.7, 127.8, 129.8, 133.7, 144.9, 190.9

A confusing diagram also claimed some NOEs in a COSY experiment, but it is hard to describe.

Permalink to Comment

40. partial agonist on September 30, 2010 1:45 PM writes...

Norepi (#35), yeah that 4.48 shift makes no sense for an acetophenone methylene, and shielding will only move it up.

The proton spectrum is consistent with a cyclophane with ether oxygen atoms inserted into the bridge,



with the two indicated fragments joined at the para positions of the phenyls, if you follow me.

If you don't, paste this name into chemdraw and imagine what you could get if you treat it with base:


The problem with my structure is how to get their MS fragment, and how to explain the 30.7 in the carbon spectrum, when the ether in the bridge would make the CH2 fall into the 70-80 range. Shielding cone effects could move it to the right, but I can't see 30.7

Permalink to Comment

41. Dr Van Nostrand on October 1, 2010 7:03 AM writes...

Looks like a mansiere to me. Giddyup!

Permalink to Comment

42. Petros on October 1, 2010 7:06 AM writes...

Those chemical shifts do seem odd.

Surely any strained structure would see vicinal coupling between the methylene protons?

Would an oligomer of 4-hydroxyphenylacetic acid give the sort of shifts reported?

Permalink to Comment

43. stiv on October 1, 2010 12:19 PM writes...

"42. Petros on October 1, 2010 7:06 AM writes...

Would an oligomer of 4-hydroxyphenylacetic acid give the sort of shifts reported?"

No. Among other things, the shifts of the aromatic carbons are all wrong (C with an oxygen on it would be 150 or further downfield, the two ortho carbons would be shifted significantly upfield to low 120's or even more)

For those who have commented on the H-shift of the methylene, note that the protons in phenylbenzyl ketone are published as 4.30ppm - the "wisterone" compound methylene is about the same.

I still think that the cyclophane XX referred to in comment 16 above is most likely but as pointed out the mp doesn't match up with the published value (the Dewhirst/Cram article referred to in comment #16 is an interesting read if for no other reason than to look at the hoops chemists had to leap through just 50 years ago to determine which structure was XX and which was XXI - today that would take a few minutes of NMR time).

I also find the reported UV max of 307nm to be odd - I didn't think a simple acetophenone structure will have a lambda max out that far.

Permalink to Comment

44. Norepi on October 1, 2010 6:39 PM writes...

@ #40: Your structure makes perfect sense save the 30.7. I don't even know what kind of magnetic anisotropy would be necessary for an upfield shift of 50 PPM. Either it's a typo, or that looks like a methyl group...

Permalink to Comment

45. AlchemistOrganique on October 3, 2010 7:41 PM writes...

Okay, can't somebody analyze possible structures using the Rychnovsky Algorithim? If it deduced the true isomer of hexacyclinol then it should have no problems tacking this POS polyketide.

Permalink to Comment

46. annette bak on October 13, 2010 8:33 PM writes...

This molecule would be an excellent candidate to form a co-crystal.

Permalink to Comment


Remember Me?


Email this entry to:

Your email address:

Message (optional):

The Last Post
The GSK Layoffs Continue, By Proxy
The Move is Nigh
Another Alzheimer's IPO
Cutbacks at C&E News
Sanofi Pays to Get Back Into Oncology
An Irresponsible Statement About Curing Cancer
Oliver Sacks on Turning Back to Chemistry