20

[clubcard]

I'm looking for 20 or so opioids for a training set. the compounds should fulfil the following:

2 aromatic groups
2 hydrogen-bond acceptors
1 positively ionizable function
1 alkene group.

Cinnamyloxycodeine is number 1 in the list.

 

[Nagelfar]

Cinnamyloxycodeine is number 1 in the list.

Would that be this one? Do you have the SMILES on it?

 

[clubcard]

The work is an updated version of 'Insights into subtype selectivity of opioid agonists by ligand-based and structure-based methods' J Mol ModelDOI 10.1007/s00894-010-0745-1

If you can't find a free copy on the net, it's one of 3 articles I would recommend everyone gets. Then we are all on the same page and so far more productive. A good idea is a good idea whomever has the idea. I have NO ISSUE whoever has the idea and I know there are at least 3 other chemists who frequent this forum. I don't know how many medicinal chemists we have. Lipinski's rule of five with a Log upto 5 and a pKa so the so most of the compound is unprotonated in the blood. ADME is the most important acronym in med. chem. Of course, my education was 27 years ago.

CCC(=O)N([C@H]1CCN(C[C@@H](O)c2ccc(F)cc2)C[C@H]1C)c3ccccc3 - I strongly suspect that altering the fluorobenzene tp a 2-Thiophene would have more affinity
CN1CC[C@]23[C@@H]4C(=O)C=C[C@]2([C@H]1CC5=C3C(=C(C=C5)OC)O4)OC(=O)/C=C/C6=CC=CC=C6 - chosen for affinity & increase potency due to alkene
BrC1=CC=C(C=C1)[C@]2(N(C)C)CC[C@@](CC2)(O)CCC3=CC=CC=C3 - chosen because it breaks the morphine rule.
O=N(C1=CC=C(CCN2CCCC/C2=N\S(=O)(C3=CC=C(Cl)C=C3)=O)C=C1)=O - note use of sulfonamide. They tried it on fentanyl analogues - it didn't work
C[C@]12C(=O)CC[C@@]3([C@]14CCN([C@@H]3CC5=C4C(=C(C=C5)O)O2)C)OC - chosen for tertiary ether - like the Bentley compounds
Oc6ccc4c1c6O[C@H]3[C@]5(OC)[C@H](C[C@@]2([C@H](N(CC[C@@]123)C)C4)CC5)[C@@](O)(C)CCC - interesting that it's found the second aromatic AND because the duration is much longer than most high affinity compounds
CCC(N1[C@@H]2CC[C@H]1CN(C2)C/C=C/C3=CC=CC=C3)=O - alkene
CCOC1=CC=C(CC2=NC3=C(N2CCN(CC)CC)C=CC(N(=O)=O)=C3)C=C1 - high affinity, 2 aromatics. Could add the carboxamide for verisimilitude (in mammals it was 240 x M).
CN([C@H]1CCCC[C@@]11CCCN1C)C(=O)c1ccc(Br)c(Br)c1 - The logical conclusion of benzamides
O=C(CC)OC1(CCN(C)CC1CC=C)C2=CC=CC=C2 - for the allyl group which vastly increases potency.
C[C@]12C(=O)CC[C@@]3([C@]14CCN([C@@H]3CC5=C4C(=C(C=C5)O)O2)C)OCCCC6=CC=CC=C6 - strongest of the Schmidhammer modification.
CCC(=O)N([C@H](C)CN1CCC(c2ccccc2)C(C)(C)C1)c3ccccc3 - proof that phenapromide can be potent (x60 morphine)
COc1cc2CCN(C)C(CCS(=O)c3ccc(cc3)N(=O)=O)c2cc1OC - proof that metofoline isn't just an alternative to codeine.
CCC(=O)C(CCN1CCC2(CCCC3=CC=CC=C32)CC1)(C4=CC=CC=C4)C5=CC=CC=C5 - semi-rigid methadone derivative.
CC1Cc2cc(O)ccc2CCN1CCc3ccc(cc3)N(=O)=O - benzazepines CAN be potent.
CC1C2CCC[C@]1(CCC2Cc3ccccc3)c4cccc(O)c4 - rigid derivative of ketobemidone with a phenethyl group which may be substituted or the whole ring substitution to 2-thienyl.
O=C1CC(CN1)c2ccccc2.C(N3CCCCC3)c4ccccc4c5cccs5 - discovered when Paul Janssen was looking for neuroleptics. It has been noted that for many neuroleptics, replacing a =0 with a benzene = opioid
C=C1CC[C@@]6(O)[C@H]3Cc5ccc(O)c4O[C@@H]1[C@]6(CCN3CCc2ccccc2)c45 N-phenylethyl analogue of nalmefenev -well spotted of a phenanthacine with alkene. Thanks
CC[C@@H](C)N(C[C@H](O)[C@@H]1CCC(=O)N1Cc2ccccc2CC(F)(F)F)[C@H](F)CC - Viminol analogue x434 morphine
COc1cccc(c1)[C@@]23CCC[C@@H](N(CCc4ccccc4)CC2)C3=C benzomorphan with allyl & N-phenylethyl. Ki 0.19

And there are the 20. 20 totally different opioids and I wish to expand on the paper I mentioned from:

1-aromatic system
2-hydrogen-bond acceptors
1-positively ionizable function.

So only 4 sites have been found BUT it's clear a second aromatic is vital to high-potency. I will be running them through the software on the paper with the hops of adding -

2-aromatic systems
2-hydrogen-bond acceptors
1-alkene displaying van der waals forces
1-positively ionizable function.

So, please use a viewer to see the 20 totally unlike opioids in 2D but think in 3D and minimum-energy alignment. If you can think of more examples of scaffolds I haven't used, please speak up, the more the merrier. As I said, I intend to write a paper on the results so it WILL get read by hundreds if not 1000s. Anyone who finds another scaffold is, therefore, named in the paper. This is very useful if you haven't been published before. If you E-mail medicinal chemists, the first thing they do is look what papers you are on. Once you have 1, more people talk to you. Right now I'm involved in patenting a new class of benzodiazepines. As soon as the patent is given, I will run through how I got to them. Lesson 1 - Think like Paul Janssen!

Of course - I've discussed no synthesis - that would be Eunoia.

 

[aced126]

I'm guessing you require opioids with known receptor affinities? What is a training set BTW? I think I have a rough idea but just wanna be sure

 

[aced126]

Post might've been deleted, not sure it might've glitched, in the case that it has I'll just suggest again: n-phenethylmorphine (14x morphine) and derivatives. Got the 7,8 double bond although not sure that's what you're looking for.

 

[serotonin2A]

The point is, to produce a set of 20 high-mu-affinity agonists to find the key structural features. The paper I wrote was essential took a training set and the result was:

1 - Aromatic system
2 - hydrogen bond acceptors
1 - positively ionizable group.

So, it's my belief that there are 2 aromatics and a van der waals force between 2 -CH=CH- groups so I'm trying to work out a training set of the most potent scaffolds and, just as importantly, as many different scaffolds as possible.

Two questions:

(1) The crystal structure of the mu receptor has been published. Wouldn't it ultimately be more useful to run a docking study?

(2) Are you sure the second site has to be aromatic, as opposed to just lipophilic? There is at least one example that has been published showing that the N-phenethyl group in a potent opioid can be replaced with a very lipophilic substituent (with suitable length) without loss of affinity or potency. The results of substitution were very sensitive to chain length, branching, saturation, and stereochemistry (as might be expected due to steric constraints).

 

[serotonin2A]

The potency if these drugs are known - we are above design, we are just suggesting candidates for 3D mapping. Who would think that a viminol analogue overlays fentanyl? Hence my effort to find 20 different scaffolds to find the common features. This isn't about making drugs - it's analysis of how high mu affinity agonists are arranged in space.

What I wrote wasn't about designing drugs, it is about understanding why the arrangement of the atoms in space affords the molecules high affinity. Pharmacophore mapping will show possible points of overlap but will never allow you to understand why that is happening. It is also based on the assumption that all ligands assume the same binding pose and interact with the same parts of the receptor, which often isn't true, especially across different scaffolds.

 

[Nagelfar]

So are you looking for morphinans specifically or as many different diverging scaffolds as possible? If it's just a spread of morphinans I think of some of the simple, but better than those in heavy currency; 6-MDDM, heterocodeine, etc.

 

[clubcard]

I keep asking. Copy 1 line of SMILES and paste it in here for the image, stereocentres and IUPAC name. I'm not going to sit around drawing pictures when you can look yourselves in 10 seconds. I HAD hoped people would have been interested in 20 compounds they have never seen - the huge range of opioids. There isn't a single class that doesn't spawn a > x100 morphine derivative. I did spend 3 years looking.

https://www.chemaxon.com/marvin/sketch/index.php