μ-receptor agonists.

[clubcard]

When looking at a novel opioid, I've always used the levorphanol & fentanyl scaffold to overlay. Even things like Viminol Do overlay, or at least it follows the 4-sites identified-

1 x aromatic system
2 x hydrogen bond donors
1 x positively ionizable function

All of the high-potency opioids have 2 aromatics or similar (lipophilic) group. Fentanyl, PET-7 and etonitezine are all examples.

I have 1 single case where I cannot work out the overlay - it's tilidine. Tilidine is unusual in many ways overlaying cypenamine as it does, secondary amine being active and NOT going anywhere the Morphine Rule

1 - aromatic system
2 - quaternary carbon
3 - ethylene spacer
4 - tertiary amine

Well, it follows 1 of the rules, I guess. Has anyone with Chemoffice overlayed nortilidine & levorphanol? Anyone already solved this one, I would love to know. It's very unlikely to lead to a potent drug and precursors cost a fortune so to make this 100% clear, I'm really just interested in the overlay, not in anything else.

Cheers

 

[serotonin2A]

The phenyl ring and amine in tilidine overlay with the phenyl ring and amine in morphine.

 

[clubcard]

Not according to Chemoffice - A secondary amine is the most basic so the lone-pair so I presume that it doesn't exactly overlay (and I've tried very fine minimum-energy calculations). It's more complex. As I'm sure you know, the direction of the lone-pair of morphine is vital. Quaternary salts were used to prove which of the 2 directions was active. That's why this is unusual case.

 

[serotonin2A]

You won't get exact overlay in Chem3D. You have to dock them in the binding site using Autodock or a similar program. When you do that, the phenyl ring and the basic nitrogen in tilidine and morphine can interact with the same residues in the binding site.

Pharmacophore models (ie, trying to overlay compounds without regard to the structure of the binding site) are useful but have been superseded by models of ligand interactions with the crystal structure of MOR. The problem with pharmacophore modeling is that it doesn't take into account the fact that similar ligands can adopt very different binding poses, or may not interact with the same set of residues in the binding site.

You can see the limitations of the pharmacophore approach by trying to predict the potency of carfentanil based on the fentanyl pharmacophore. The added H-bond acceptors at C4 in carfentanil allow it to interact with residues in the binding site that do not interact with fentanyl. You would never be able to predict that unless you docked fentanyl into the crystsl structure.