User Controls
structure-activity relationships
-
2018-02-19 at 9:16 AM UTCfor hypothetical psychoactive drugs
does anyone know some good introductory materials on the leddit or something so i can learn about this?
does understanding structure-activity relationships require a strong background in chemistry, or can basic principles be used to estimate a molecule's activity on receptors (for example is a certain element or substructure enough to make some basic predictions?)
starting questions:
a) what traits would a molecule need to have to suggest affinity to the 5ht2a receptor, and to what degree
b) what traits would a molecule need to activate the 5ht2a receptor, while also activating the phosphor pathway or whatever within the 5ht2a receptor that is responsible for psychedelia
c) what traits would a molecule that is 5ht2a receptor-friendly have that could suggest toxicity? or at least some examples of this, of what would suggest a chemical should be disregarded -
2018-02-19 at 9:25 AM UTCMy understanding (from non-psychoactive drug design) is that tertiary structure is integral to drug action makes it exceedingly to predict binding affinity. It's difficult to even predict what the tertiary structure will be, and is almost always done by computational approximation followed by high throughput screening. Drug discovery entails billion dollar startup costs for a reason.
-
2018-02-19 at 9:32 AM UTCwell gunlover was predicting a certain quaalude analog would be a good recreational drug based on its structure.
on places like bluelight people make assumptions about fentalogs based on what makes up the molecule.
im wondering what logic gun lover (and others) uses to make these predictions. i also noticed a lot of these predictions seem to be made with 2-d representations
i get what you're saying that if you're just planning on mixing random chemicals together to make something you can patent, you need to use specialized software to deal with algorithms to see what you're dealing with
what i mean in the most basic sense is if you had a chemical like tetrahydrocannabinol, and made a slight change like adding 3 methoxy groups for 3-meo-thc, could you predict if it would be more or less potent, or if it is a likely toxin? i dont know if this (3-meo-thc) is even possible, i'm not trying to predict how to synthesize things, just hypothetically if they existed, what their pharmacology would be -
2018-02-19 at 9:35 AM UTCor like with the 5ht2a receptor, the receptor has a certain molecular makeup, it is like a lock and the chemical needs to have certain qualities to open the receptor like a key. so wouldnt all chemicals sharing a certain mechanism of action need to have specific qualities to catalyze this chemical reaction?
-
2018-02-19 at 12:03 PM UTC
Originally posted by lightray well gunlover was predicting a certain quaalude analog would be a good recreational drug based on its structure.
on places like bluelight people make assumptions about fentalogs based on what makes up the molecule.
im wondering what logic gun lover (and others) uses to make these predictions. i also noticed a lot of these predictions seem to be made with 2-d representations
i get what you're saying that if you're just planning on mixing random chemicals together to make something you can patent, you need to use specialized software to deal with algorithms to see what you're dealing with
what i mean in the most basic sense is if you had a chemical like tetrahydrocannabinol, and made a slight change like adding 3 methoxy groups for 3-meo-thc, could you predict if it would be more or less potent, or if it is a likely toxin? i dont know if this (3-meo-thc) is even possible, i'm not trying to predict how to synthesize things, just hypothetically if they existed, what their pharmacology would be
Yeah, I'm not sure, I see people talking about anticipated effects from 2D primary structure diagrams on drug forums pretty often too. I'm pretty skeptical of it but they're much simpler molecules than I ever see from the computational perspective so it could be the 3D structure isn't as important there. Either from fear of chemistry or natural complexity of the domain I don't know but CS people tend to be of the opinion that pharmacology is totally unknowable from primary structure. There's plenty of examples of where it clearly is, but again, not sure if that's globally true or just for a certain class of molecules. I think gun lover mentioned starting a PhD in something pharma related so I'd be interested to hear from someone who presumably knows more about the subject. -
2018-02-19 at 11:46 PM UTCBeing able to predict such things isn't really a matter of "will the chemical modification retain activity" and more of "is the chemical modification a bad idea?" Knowing the basics of organic chemistry and molecular biology is essential so that you can ask yourself questions like:
Will the substitution be too polar or too fatty?
How will that effect absorption, distribution, metabolism & excretion?
Will it play nice with your metabolic enzymes?
Will it remain a similar steric size?
Do similar substitutions within the same drug scaffold lead to desirable drug?
You can have the best, most accurate answer to all of these and still wind up with a dud. It happens all the time in pharma research where there is basically unlimited manpower and resources.
In the case of serotonin 2A, you should peruse pihkal & tihkal. You can infer what positions can be substituted at all, what substitutions seem most safe and effective, then play around from there.
What you won't be able to ascertain without synthetic organic chemistry knowledge is how difficult your proposed drug will be to make. It's easy to draw exotic things that look good on paper. It's an entirely different ordeal to actually obtain the stuff.
Originally posted by Lanny Yeah, I'm not sure, I see people talking about anticipated effects from 2D primary structure diagrams on drug forums pretty often too. I'm pretty skeptical of it but they're much simpler molecules than I ever see from the computational perspective so it could be the 3D structure isn't as important there. Either from fear of chemistry or natural complexity of the domain I don't know but CS people tend to be of the opinion that pharmacology is totally unknowable from primary structure. There's plenty of examples of where it clearly is, but again, not sure if that's globally true or just for a certain class of molecules. I think gun lover mentioned starting a PhD in something pharma related so I'd be interested to hear from someone who presumably knows more about the subject.
The reason chemists use 2D skeleton drawings is because they are simple & convenient. They are idealized structures existing in total isolation; they are succinct and easy to compare. But chemists understand that molecules are 3-D and that the preferred orientation of the molecule is always going to depend on its chemical environment. So when we're talking about all of these structures it is understood that their preferred conformation dynamic and varied upon things like temperature, pH, buffer effects, solvent effects, etc. You will pay tens of thousands of dollars in supercomputer costs in order to get a 10 nanosecond window into "realistic" drug-receptor docking. But then your initial parameters have to be impeccable, or else any result one way or the other is effectively null.
I actually quit pursuing this research, because, aside from a few scaffolds to play with, no one really knows what is a viable psychoactive or not. The best minds and supercomputers struggle to pump out safe and effective drugs. The experiments suck to actually run.The following users say it would be alright if the author of this post didn't die in a fire! -
2018-02-20 at 12:02 AM UTCThe thing about serotonin 2A is that even if you know the drug is safe & shows high receptor affinity, you still have to deal with the enigma of functional selectivity.
You can have a drug that displays affinity & efficacy at 5-HT2A but does not induce psychedelia. This has to do with the receptor inducing two different signaling pathways -- one makes you trip but the other doesn't. I have never heard of a way to predict the drug's preference for one pathway or another. There isn't a way to know without measuring it, and there is no obvious trend to me. Even then that answers only one question -- it says nothing of the pharmacokinetics.
Serotonergic psychedelics really are a crapshoot. You can have fun thinking up stuff, but there's no good way to know its merits as a drug without tasting it. -
2018-02-20 at 12:18 AM UTCI don't have a sophisticated understanding of pharmacokinetics, so this may be a silly question, but I'm curious: is there a way to block the nonpsychoactive serotonin pathway, allowing it to bind to the other?
-
2018-02-20 at 12:28 AM UTChttps://en.wikipedia.org/wiki/5-HT2A_receptor#Agonists
"Newer findings reveal that psychoactive effects of classic psychedelics are mediated by the receptor heterodimer 5-HT2A–mGlu2 and not by monomeric 5-HT2A receptors"
https://en.wikipedia.org/wiki/Metabotropic_glutamate_receptor_2#Role_in_hallucinogenesis
"Many psychedelic drugs (e.g. LSD-25) produce their effects by binding to the oligomerized complexes of the 5HT2A and mGlu2 receptors"
I would assume not. If a chemical can bind to to 5-ht2a-mglu2, it will do so. Blocking the regular serotonin pathway won't make it hop over to the heterodimers, and blocking the regular serotonin pathway will make it impossible for psychedelia to even occur, since for psychedelia all receptors in the heterodimer need to be activated
in simpler terms: for a drug to be a psychedelic, it has to bind to both a certain serotonin subunit and a certain glutamate unit. Which I think would imply that the mechanism of action of psychedelics overlaps with dissociatives? I dunno -
2018-02-20 at 1:04 AM UTC
Originally posted by lightray https://en.wikipedia.org/wiki/5-HT2A_receptor#Agonists
"Newer findings reveal that psychoactive effects of classic psychedelics are mediated by the receptor heterodimer 5-HT2A–mGlu2 and not by monomeric 5-HT2A receptors"
https://en.wikipedia.org/wiki/Metabotropic_glutamate_receptor_2#Role_in_hallucinogenesis
"Many psychedelic drugs (e.g. LSD-25) produce their effects by binding to the oligomerized complexes of the 5HT2A and mGlu2 receptors"
I would assume not. If a chemical can bind to to 5-ht2a-mglu2, it will do so. Blocking the regular serotonin pathway won't make it hop over to the heterodimers, and blocking the regular serotonin pathway will make it impossible for psychedelia to even occur, since for psychedelia all receptors in the heterodimer need to be activated
Looks like the science has progressed a little bit since I last looked. This helps explain functional selectivity, but I don't think it does anything for SAR predictions.
Be careful with trying to quantify how many sites must be occupied before the receptor is activated. It is not always necessary for every site to be occupied, and largely depends on the stage of the receptor's life (extent of phosphorylation at that receptor at that time).
Originally posted by lightray in simpler terms: for a drug to be a psychedelic, it has to bind to both a certain serotonin subunit and a certain glutamate unit. Which I think would imply that the mechanism of action of psychedelics overlaps with dissociatives? I dunno
It only has to bind to the serotonin unit. The glutamate receptor is activated by glutamate naturally present throughout your body.
I wouldn't be surprised at all if this receptor dimer thing is tied to the psychedelic effects of ketamine and other NMDA antagonists. I'd go so far as to bet that it does. -
2018-02-20 at 1:10 AM UTC
Originally posted by Zanick I don't have a sophisticated understanding of pharmacokinetics, so this may be a silly question, but I'm curious: is there a way to block the nonpsychoactive serotonin pathway, allowing it to bind to the other?
You could block the nonpsychoactive serotonin pathway, but then you'd probably die. It won't just go to the other path. -
2018-02-20 at 1:27 AM UTC
Originally posted by Gun Lover It only has to bind to the serotonin unit. The glutamate receptor is activated by glutamate naturally present throughout your body.
but wouldn't the 5ht2a unit in a homodimer and heterodimer be the same? the only difference is that the heterodimer also has a glutamate receptor
in that case all 5h2a agonists would be activating both the 5ht2a on homodimers and heterodimers. so the only way to activate this psychedelic pathway would be an abnormal amount of glutamate, or else there would be no specific pathway for psychedelia beyonf 5ht2a -
2018-02-20 at 2:21 AM UTC
Originally posted by lightray but wouldn't the 5ht2a unit in a homodimer and heterodimer be the same? the only difference is that the heterodimer also has a glutamate receptor
Absolutely not. You've gotta realize that these dimers have an inherent allosteric effect upon one another. They are essentially locked together at a junction at the intracellular space. It is not at all accurate to say that receptor = homodimer = heterodimer. When one part changes shape, the other changes too.An allosteric receptor–receptor interaction exists in this complex since mGlu2 receptor activation produces an increase in the affinity of hallucinogenic 5-HT2A agonists for the 5-HT2A protomer binding site (Gonzalez-Maeso et al., 2008). A bidirectional receptor–receptor interaction is present since 5-HT2A agonists reduce the affinity of mGlu2 agonists for the mGlu2 protomer binding site. It is of substantial interest that the allosteric receptor–receptor interactions in these heteroreceptor complexes are dysregulated in postmortem brains from schizophrenia subjects
It seems to me that endogenous glutamate binds to the dimer, which increases the binding affinity of some, but not all 5-HT2A agonists. A ligand like serotonin binds more poorly after glutamate activation while drugs like LSD bind with great efficacy. But just because a drug has high affinity does not mean it gives efficacy. Lisuride is an example that binds strongly to the dimer without activating the second messenger pathway that makes you trip.The following users say it would be alright if the author of this post didn't die in a fire!
