Lee, Kim, Choi, Ha, Kim. BOMCL, 2010, 20, 409-412. DOI: 10.1016/j.bmcl.2009.10.016. 
The Tot. Syn. bull-shit detector is at eleven just now, having just read Derek Lowe’s post on Piperkadsin C. This natural product was isolated by Kang Ro Lee of Sungkyunkwan University in a assay-guided chromatographic separation of isolates from Piper kadsura (Piperaceae). It doesn’t take much more than a glance to see that something seems very wrong with this structure, as it seems to contain a bridged cyclobutene, partially containing a dieneone. From a chemical stability aspect, this is scary enough, but the real concern is of course in the sterics. I’ve done some very simple MM2-minimisations in Chem3D:
I’ll ask the geeks chaps in our comp-chem office to try and do something a bit more meaningful tomorrow. In the meantime, have a look at the NMR data in the SI. Having trouble finding it? I did too, ’cause all we’ve got is the tabulated data in the main paper. Really, I thought all isolations came with scans of the actual NMR data?
I’m still wading through the assignments, and trying to match the 13C data with predicted data in ChemDraw (which is pretty good in 13C). But what strikes me is the fact that a search of the PDF has no hits for words like, uh, butane or butene or even frickin’ strain. Seriously, WTF?!
Part II
Okay, my colleague, James, has done a bit of proper 3D modelling for us (and also provided some tips for modelling) – I quote:
For the sake of all thing decent in comp chem you really should dump the chemdraw 3D stuff and use something decent!! These were done using omega 3D, the minimisation of all confs produced with MM3, the images are of the lowest energy conf. Aside from the method, one complaint with the chemdraw images is that you can’t see which are single and which are double bonds
One approach that is slightly preferable is to use a free demo of a commercial tool like corina (a farily industry standard 2D to 3D converter)
1. Select your mol in chemdraw and do Edit -> Copy As -> Smiles
2. Paste the smiles into the above web page, download the results into whatever free viewer you like (pymol is my favourite freeby) and generate some images.
If you get problems, put all explicit H’s in the mol in chemdraw and corina will behave better, I’m not too sure if the simple web app will handle things without H’s.
8th of February – more modeling
I’m currently not well disposed to the bloody French, but I’ll consider one chap and his colleague - Yoann Coquerel & M. Rajzmann – who have added to the modeling work above, and done some proper, AM1 (semi-empirical) and DFT (ab inito) calculations on the proposed structure of Piperkadsin C. I quote:
We have performed AM1 calculations on the proposed structure of piperkadsin C and a possible alternative structure. Note that we have not been through the reported characterization data to propose the alternative structure, and it may not be realistic at all given the reported data; we just used it as a reference point. Remarkably, the molecule looks stable, although its enthalpy of formation is much higher than the proposed alternative structure. To confirm these results, we have performed DFT calculations on a simplified analogue, and the results are similar. It may be noticed that the results obtained by both methods are very similar, but the AM1 method required less than 5 second on my laptop. If some people are interested, we can continue this study to determine the energy profile of the transformation from the proposed structure of piperkadsin C to the alternative one. Feel free to ask.
The result of their labours can be seen in the lovely graphic below:
(Click for an even bigger image)
Thanks, chaps!
No comments:
Post a Comment