And as usual, the easiest answer to "Why does this compound in particular fits resp. activates receptor XYZ so much better than its derivatives" is: Because it just fits so damn good.
that is a simplistic answer and doesn't withstand even cursory scrutiny
Yes yes, I admit, my choice of words was of course a bit oversimplifying...
With respect to the Ki-value of 0.08 nM, well, I doubt it somehow. It's more a gut-feeling, but looking for example at the closely related compounds
AM-2225 (that is AM-694 + a methyl-group at the indole's pos. 2) with a Ki at CB1 of 5.57 nM (
80 times higher!), or
AM-679 (i.e. AM-694 but with a H for the F) with a Ki at CB1 of 13.5 nM, I doubt that picomolar affinity is for real.
Apart from the differing Ki-values for close relatives, I doubt information from current patents in general, if is not backed up by a reliable source. Patents are notoriously badly phrased and often cryptically written. Nobody checked these values in any way, and usually nobody puts them in doubt if they may vary so drastically from other measured values.
What does a single fluorine-atom offer in contrast to a hydrogen in the same place? Suggestions:
1.
Metabolic stability, but as was mentioned by Vecktor, this is irrelevant when talking about
in vitro data.
2. The
van der Waals-radius of H vs. F are quite similar; fluorine's is just ~20% larger. As hydrophobic residues in that position are well tolerated within a consirable range of substitutions (propyl, butyl, pentyl, hexyl; all straight or branched; morpholino-ethyl; piperidinylmethyl, ...) without changing CB1-affinity this drastically, I don't think that the difference in sizes is the deciding parameter.
3.
Different hydrophobicity maybe? Substituting a hydrogen for a fluorine leads to an increase in overall hydrophobicity, but when substituting just a single proton in a C5-chain, the difference is quite small. I think this can be disregarded as deciding reason, too.
4. In contrast to a hydrogen, fluorine can act as
hydrogen-bond acceptor. Maybe this makes the difference. What are the proposed amino acids that surround the hydrophobic binding pocket (ie. the one where the N-alkyl residue of the AAIs goes)? Mayb interaction occurs not with a amino acid sidechain but with the backbone, thus forming a hydrogen bond of the type N-H•••F, which is indeed possible...
Again, I must emphasize that I doubt the 0.08 nM-value somehow...
Peace! -
Murphy
Edit: I'd like to add one point to the short list above:
5. Fluorine is capable of providing so-called
orthogonal multipolar interactions with target proteins. These should not be confused with hydrogen bonds (atom distances are too large). This can
indeed make a difference between Ki-values of 1-2 orders of magnitude. An example was published in
Org Biomol Chem 2006,
4, pp.2364 – 2375, where thrombin inhibitors were studied. The Ki-values of one compound vs. a derivative where a proton was changed for a fluorine got down from 310 nM to 57 nM.
I think that this property is unique for fluorine and not shared by related atoms (ie. halogens); for example with a chlorine-substituent in the same position the Ki was just about 190 nM. Could this be the answer?