KinasePro

Kinase Chemistry – Just a year and a half behind the times.

S – – – – O

Posted by kinasepro on July 31, 2007

I thought this was pretty well known, but looking at some recent literature along with Ashutosh, I guess its worth highlighting. The short of it is that if you need it in plane maybe you aughta see what it looks like with a sulfur atom.

Dasatinib is one such example where the carbonyl is in plane with a heteroaromatic Sulfur atom, but there are plenty of more where that came from…

MM2 doesn’t handle the phenomena well so your minimizations are likely going to be off. The thiadiazoles are perhaps the best described in the literature as far back as ’85 but more recently in a ’98 JACS.

While I didn’t look that hard I couldn’t find a single example where any of these kinds of acyl – thiophene, thiazole, or thiadiazoles were other then adjacent and in plane forming a 4, 5, or 6 membered pseudo-ring system.

The recent 2PZI is an example as is the BI P38 reference at The Curious Wavefunction. There’s also a recent Chem. Pharm. Bull. describing a series of IKK-b inhibitors where they suggest the phenomena is observed as well:

Energetically this may well not be worth much, but with all these examples it’s quite clearly enough to bias a ligands conformational preference. Some sort of π*(s) through space effect I presume.

Advertisements

8 Responses to “S – – – – O”

  1. Thanx for this interesting point; I was not aware of this pattern yet! I was told one or two years ago that sulfur is the most difficult atom to model theoretically, so am not surprised that modeling methods to not favor this conformation.

    BTW, does anyone have a Jmol script to visualize this in 3D?

  2. Wow…that’s a lot of useful structures. Thanks for the references. Now, while it’s true that MM2 does not locate the cis structure at a global energy minimum, MMFF locates it as the second lowest energy structure, separated from the minimum energy trans by only 2 kJ/mol.
    But interestingly, OPLS does locates the cis as the minimum energy structure, separated from the next lowest trans by a considerable 13 kJ/mol.
    I am in the process of doing conformational searches on some of the other structures with different FFs. Will let you know if something interesting comes up.

  3. weirdo said

    Wait a minute, before considering through-space effects and pi-* interactions, etc. etc., let’s just do a little crude sophomore conformational analysis.

    For any molecule where the amino group is not at the 2-position of a thiazole (or thiadiazole), this conformation is exactly what simple plastic models would predict. The amides are in an S-cis conformation — and in plane with the heteroaromatic ring. In this conformation, you have the choice of placing the N-H next to the the S or the C=O (180-degree rotations). A simple steric argument places C-R away from the carbonyl to avoid A1,3 strain.

    What about the 2-amido thiazoles? Placing the carbonyl next to the nitrogen atom of a thiadiazole leads to lone-pair interactions that are certainly unfavorable.

    Thus, rather than invoking a favorable interaction of the carbonyl with the sulfur atom, my simple synthetically-based mind would focus on the repulsive interaction of the carbonyl moiety interacting with what is on the other side — whether C-R or N-lone pair.

    Now, show me some structures of N-alkylated amides with the same conformational bias, and my reasoning pretty much flies out the door . . . but I’ve done a little looking for this and cannot find any. I yield to your superior database-mining skills.

    {I wouldn’t mind you proving me wrong. I’ve seen this affect before, and didn’t find it remarkable for the reasons stated above. }

  4. Derek Lowe said

    This might well account for some of those times when thiophene and phenyl are not so equivalent, eh?

  5. milkshake said

    or thiophene and furan (we had this thing juste recently and the difference was more thn 200 fold)

  6. Kinase amateur said

    Similar thing seen with AZ’s thiazolylaminoquinazolines (Aurora), only the quinzoline nitrogen acts as a psuedo carbonyl…or have I got it completely wrong, again.

  7. kinasepro said

    Don’t remember the reference, but I have indeed seen N—S described similarly, only a bit weaker. Something I think the distances in the dasatinib image above illustrate nicely.

    On Weirdo’s point: Before considering this trend my sixth synthetic chemistry sense led me down a different path, one that is similar to what Ashutosh describes from the BI J-Med-Chem. Namely that sulfur has a lone pair. A couple of them in fact and even though they may not be as tightly held as their furan or aza cousins, they are still there…

    Couple that with structures like the amide seen in 2p54 (second image in the list above) which has a methyl group on the other side… I would expect both of these factors to push the amide out of plane, as is routinely seen on phenyl rings, but it too is in plane with the oxygen 2.83A from the sulfur.

    I’m not interested in proving anyone wrong and Weirdo’s point is certainly valid. I just fail to see how in not just one structure, but entire lists of structures all are found to be in plane unless there is some sort attractive force at play?

    This observation extends passed amides. A structure like 1PGE shows a ketone displaying a similar preference. Nothing is ever quite so tidy though, and I did manage to find one counterpoint in 2D1J: a tertiary amide which is out of plane and 3.15A away (the carbonyl is still on the same side though :p)

  8. tweaker said

    I reckon there is an electrostatic component too, those sulphurs often
    carry a slightly positive charge and you do occasionally find sulphur (or halogen)
    to carbonyl hydrogen bond like interactions in crystal structures…

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

 
%d bloggers like this: