I Like to Draw Pictures of Random Molecules

[DotChem]

... you can't invent a molecule and put it into a model, and be sure it fits or have electronic properties as you think it does. With good imagination you can imagine how a close structural analogue might bind, but even for a not so distantly related molecule it becomes too little. I'm wondering if making a model of a given receptor to be able to use it in computations would be a costly thing to do if you had access to a biochemical lab.

I am not sure if you're familiar with protein X-ray crystallography technique used to solve drug-receptor structure! If you ever been to a dentist, you have probably been subjected to the same technique! usually he'll order X-ray done to get a "picture" of teeth and gum, is that right? He'll put you in an X-ray machine and get a "picture" of your teeth and gum (ie the relative electrons density of different atoms of the organs and their relative locations). Hopefully no caries holes to fill!!

That's the EXACT same technique used to solve the structure of drug-receptor complexes by X-ray crystallography.
Same technique, same principles, same machines (may be more powerful) a dentist use to see patient teeth (the picture he gets IS NOT A COMPUTER predicted model. It is EXACTLY the location of teeth in real life ..the relative electronic density of bones, teeth, gums, other tissues..etc. Otherwise dentists would long be out of business if they were relying only on computer imagined model of the patient teeth to work on him/her.

Now take a drug (real compound! real powdered highly pure sample of a compound NOT A COMPUTER imagined model, real powder!) and a sample of a protein it binds to (like cocaine + DAT protein or meth + DAT) real protein, such as the powdered protein you can buy in health stores only highly purified. Not a computer model of the protein, real protein !! Mix the drug and the protein.. hopefully, you'll get a crystalline complex of protein+drug bound to it. filter the mix, recrystallize if necessary to purify and put the crystal thus obtained in a X-ray machine. Put that on a X-ray machine and record the picture of the complex (showing the electrons distribution and hence the position of different atoms in the complex including that of the drug.. that's what people are talking about when you see papers saying "X-ray structure of drug receptor complex at xyz Armgstrom resolution blablablabla.." Just like the picture the dentist use to see teeth and treat any dental pbs..The only thing computer does is rendered the picture. You don't have to use computer! They used to develop the X-ray thus obtained on films.. Some doctors still prefer the image on the film to that on computer! Here is the basic steps

The structure of psychostimulants in complex with DAT refer to are X-Ray Structures. What the Nature paper is telling you is the X-ray they've taken of cocaine molecule bound to drosophila dopamine transporter DAT protein ...Not a computer "imagined" model! Here is methamphetamine bound to DAT (will post later...)

Now having this "picture" in hand just like a dentist, you can see where there are holes .. what part of the drug is interacting with what parts of the receptor..etc With that information in hand, you can then predict what to add, subtracts or replace..etc in the original drug molecule so as to increase binding or selectivity ...etc etc.. You then go the lab, synthesize the molecule you've predicted, test it and use the information to design better drugs.. more later....

 

[Incunabula]

I like to draw molecules of random pictures

lol, me too.

 

[adder]

I am not sure if you're familiar with protein X-ray crystallography technique used to solve drug-receptor structure! If you ever been to a dentist, you have probably been subjected to the same technique! usually he'll order X-ray done to get a "picture" of teeth and gum, is that right? He'll put you in an X-ray machine and get a "picture" of your teeth and gum (ie the relative electrons density of different atoms of the organs and their relative locations). Hopefully no caries holes to fill!!

That's the EXACT same technique used to solve the structure of drug-receptor complexes by X-ray crystallography.
Same technique, same principles, same machines (may be more powerful) a dentist use to see patient teeth (the picture he gets IS NOT A COMPUTER predicted model. It is EXACTLY the location of teeth in real life ..the relative electronic density of bones, teeth, gums, other tissues..etc. Otherwise dentists would long be out of business if they were relying only on computer imagined model of the patient teeth to work on him/her.

Now take a drug (real compound! real powdered highly pure sample of a compound NOT A COMPUTER imagined model, real powder!) and a sample of a protein it binds to (like cocaine + DAT protein or meth + DAT) real protein, such as the powdered protein you can buy in health stores only highly purified. Not a computer model of the protein, real protein !! Mix the drug and the protein.. hopefully, you'll get a crystalline complex of protein+drug bound to it. filter the mix, recrystallize if necessary to purify and put the crystal thus obtained in a X-ray machine. Put that on a X-ray machine and record the picture of the complex (showing the electrons distribution and hence the position of different atoms in the complex including that of the drug.. that's what people are talking about when you see papers saying "X-ray structure of drug receptor complex at xyz Armgstrom resolution blablablabla.." Just like the picture the dentist use to see teeth and treat any dental pbs..The only thing computer does is rendered the picture. You don't have to use computer! They used to develop the X-ray thus obtained on films.. Some doctors still prefer the image on the film to that on computer! Here is the basic steps

(...)

The structure of psychostimulants in complex with DAT refer to are X-Ray Structures. What the Nature paper is telling you is the X-ray they've taken of cocaine molecule bound to drosophila dopamine transporter DAT protein ...Not a computer "imagined" model! Here is methamphetamine bound to DAT (will post later...)

Now having this "picture" in hand just like a dentist, you can see where there are holes .. what part of the drug is interacting with what parts of the receptor..etc With that information in hand, you can then predict what to add, subtracts or replace..etc in the original drug molecule so as to increase binding or selectivity ...etc etc.. You then go the lab, synthesize the molecule you've predicted, test it and use the information to design better drugs.. more later....

I'm sorry, I guess, I didn't make myself clear. Yes, I know that those are X-ray crystallographic images of real complexes, this technique is probably the only method that can show real bond lengths, positions of atoms in relation one to another, electron density, and whatnot as opposed to calculated values, and for this reason it is definitely indispensable for many purposes like for example studying how already known ligands exactly bind to a receptor. But what I meant was to take such an X-ray image of a real ligand-bound protein complex and use it to make a virtual model to be able to virtually bind newly designed ligands. I suppose it's in many ways cheaper than making a complex of a protein and each new potentially good ligand, you don't have to synthesize all the ligands many of which will inevitably turn out useless and you're not wasting protein material which is probably even more expensive. That way using computations you can sort of strain the more promising ligands from useless ones at the very beginning of a research. Given how long the process of producing a new effective drug is, I guess that's more or less how pharmaceutical companies look for new drugs.

 

[Nagelfar]

lol, me too.

Now; this thrread is starting to make memeingful sense.

 

[aced126]

Yes, I know you can find those in articles, and not only for DAT, I came across such images of ligands bound to opioid and serotonin receptors, they are fun, but still you can't invent a molecule and put it into a model, and be sure it fits or have electronic properties as you think it does. With good imagination you can imagine how a close structural analogue might bind, but even for a not so distantly related molecule it becomes too little. I'm wondering if making a model of a given receptor to be able to use it in computations would be a costly thing to do if you had access to a biochemical lab. Certainly, a fast computer would be necessary in the latter phase, but it's interesting if simply experiments necessary to build a model would be expensive provided that you had the know-how to do it. I would guess the most expensive phase of new-drug research is the synthesis of potential ligands if it gets complex and requires expensive reagents.

Quite far from the current topic discussed, but here's an interesting article comparing affinity of intrinsic activity of analogues of buprenorphine, you could say, with t-butyl group swapped for different alkyl and cycloalkyl groups with different stereochemistry at C20.

This is quite true.

There is a good deal of research currently going on in Pharma to develop softwares to predict which molecules are going to hit the target well, a sort of computer-simulated version of high-throughput screening. This means they can chop down the number of molecules they need to actually synthesise by orders of magnitude, which as you mentioned is the main area of expense.

 

[aced126]

I am not sure if you're familiar with protein X-ray crystallography technique used to solve drug-receptor structure! If you ever been to a dentist, you have probably been subjected to the same technique! usually he'll order X-ray done to get a "picture" of teeth and gum, is that right? He'll put you in an X-ray machine and get a "picture" of your teeth and gum (ie the relative electrons density of different atoms of the organs and their relative locations). Hopefully no caries holes to fill!!

That's the EXACT same technique used to solve the structure of drug-receptor complexes by X-ray crystallography.
Same technique, same principles, same machines (may be more powerful) a dentist use to see patient teeth (the picture he gets IS NOT A COMPUTER predicted model. It is EXACTLY the location of teeth in real life ..the relative electronic density of bones, teeth, gums, other tissues..etc. Otherwise dentists would long be out of business if they were relying only on computer imagined model of the patient teeth to work on him/her.

Now take a drug (real compound! real powdered highly pure sample of a compound NOT A COMPUTER imagined model, real powder!) and a sample of a protein it binds to (like cocaine + DAT protein or meth + DAT) real protein, such as the powdered protein you can buy in health stores only highly purified. Not a computer model of the protein, real protein !! Mix the drug and the protein.. hopefully, you'll get a crystalline complex of protein+drug bound to it. filter the mix, recrystallize if necessary to purify and put the crystal thus obtained in a X-ray machine. Put that on a X-ray machine and record the picture of the complex (showing the electrons distribution and hence the position of different atoms in the complex including that of the drug.. that's what people are talking about when you see papers saying "X-ray structure of drug receptor complex at xyz Armgstrom resolution blablablabla.." Just like the picture the dentist use to see teeth and treat any dental pbs..The only thing computer does is rendered the picture. You don't have to use computer! They used to develop the X-ray thus obtained on films.. Some doctors still prefer the image on the film to that on computer! Here is the basic steps

The structure of psychostimulants in complex with DAT refer to are X-Ray Structures. What the Nature paper is telling you is the X-ray they've taken of cocaine molecule bound to drosophila dopamine transporter DAT protein ...Not a computer "imagined" model! Here is methamphetamine bound to DAT (will post later...)

Now having this "picture" in hand just like a dentist, you can see where there are holes .. what part of the drug is interacting with what parts of the receptor..etc With that information in hand, you can then predict what to add, subtracts or replace..etc in the original drug molecule so as to increase binding or selectivity ...etc etc.. You then go the lab, synthesize the molecule you've predicted, test it and use the information to design better drugs.. more later....

What do you take adder for? Of course he, and most people in this forum, know what X-ray crystallography is.

You've made the bolded part sound quite trivial, but in reality this is far from being an easy job.

 

[Solipsis]

lol, me too.

Haha ^ sorry for fucking the thread but this cracked me up

 

[Nagelfar]

Is this co-ordination as inane as the cat pic or? I know nothing about co-ordination chemistry; but I am trying to get a DAT allosteric modulator to work as a DRI

 

[sekio]

Only metal centres (e.g. Cr(CO)3) will pi-coordinate like that, nags.

 

[Nagelfar]

Only metal centres (e.g. Cr(CO)3) will pi-coordinate like that, nags.

I took this from a patent in my decade long sifting of phenyltropanes; almost looks sulfur "co-ordinated" but not with any connected placement of anything else; perhaps the proper terminology for what this exact methodology is, is unknown to me in this case, any clue?

 

[Solipsis]

I took this from a patent in my decade long sifting of phenyltropanes; almost looks sulfur "co-ordinated" but not with any connected placement of anything else; perhaps the proper terminology for what this exact methodology is, is unknown to me in this case, any clue?

That is a cyclohexyl, not an aromatic ring so no pi-interaction. Are you sure it's not a different sort of context that explains the "S" ? What does the patent say referring that image?
(Even if it was sulfur, I don't think that behaves in any such way in coordination chemistry - as said you need something like iron for that.. sulfide would be 2- and you need something that is cationic without real covalent binding tendencies to coordinate I think)

 

[RandomClean3]

They are structurally similar.

If you look at both molecules, yeah, it's hard to find similarities. But we know the ester bond on cocaine can be replaced entirely by a carbon bond and it retains activity, therefore, it's not part of the pharmacore. Also, we can ignore levo-methylphenidate only focus on dexmethylphenidate since it's the active enantiomer. So if you draw cocaine with that change (which I believe it's some analogue, I know the chlorinated version is) you get this.

Now, if you look at the connection between the carbonyl carbon and the amine, you notice it's a perfect match, they both go through 2 chiral carbons with the same chilarity leading to the same functional groups. The benzene is not static given that in both cases it's attached to a secondary carbon, and you can see a 6 member ring in the cocaine analogue analogous to the piperidine in the phenidates. Sure, it's not a perfect match, that's why cocaine is a sigma agonist and a serotonin reuptake inhibitor and methylphenidate isn't. But the similarity is there, and it's pretty considerable given both tropanes and piperidines are found constantly in stimulants and I suspect they may be bioisosteres. So yeah, they are pretty similar, the phenetylamine skeleton of MTP is deceiving since it's activity is not related to that. More similarities can be found in other ''dirty drugs'' like pethidine (serotoninergic and very weak dopaminergic) and even ibogaine if you wanna get creative

 

[RandomClean3]

Dude, that's old, the first is nocaine. The entire family is being investigated as less addictive analogues for addiction, depression, obesity, etc. But they don't seem very promising to be honest, they pretty much follow the same pharmacore in a more planar form, and oxazoles are bioisosteres to esters, so it makes sense they tried it out.

 

[RandomClean3]

The S is sometimes used as substituted. They leave the ring alone with no double bonds since it's possible to replace with an oxygen double bond.

 

[Solipsis]

Ah I only know that as a diagonal line crossing the bond of an aromatic ring, but that makes perfect sense!

I think the name 'nocaine' is very pessimistic

 

[Nagelfar]

Are you sure it's not a different sort of context that explains the "S" ? What does the patent say referring that image?

US6479509 B1, Frank I. Carroll, Research Triangle Institute 4229 (phenyltropane) series

Sift through those images; that is the only one with an "S" like that, the rest are all arene/aryl, definitely aromatic and why would this be a cyclohexyl in context is beyond me.

The S is sometimes used as substituted. They leave the ring alone with no double bonds since it's possible to replace with an oxygen double bond.

Thanks RC3; probably the explanation, but it's weird how they throw that in there (but they also have circles/double bonds interchangeably to show the aromaticity)

 

[Nagelfar]

So yeah, they are pretty similar, the phenetylamine skeleton of MTP is deceiving since it's activity is not related to that.

A lot of people overlook that for sure, so confusing and aggravating to explain to some who are "sure they know" what they're talking about; I have even argued with licensed prescribers that MPH is *not an amphetamine* (they were insisting it was)

and thanks for that elucidation it was wonderful.

 

[aced126]

Also another thing to note is that the stereochemistry at the alpha carbon of MPH is akin to that of levo-amphetamine.

 

[DotChem]

What do you take adder for? Of course he, and most people in this forum, know what X-ray crystallography is.

No the intent is not to imply whether adder (on anybody on this forum) is or is not aware of x-ray crystallography technique. The reason I posted is that lots of people easily make no distinction between computer predicted structural models of drug-protein and X-ray determined structures. The 2 are not the same!! I guess it is because the representation as computer images are the same! So it is very important to make difference.

You've made the bolded part sound quite trivial, but in reality this is far from being an easy job.

No! actually it is quite trivial once the X-ray of the target protein is solved. Even easier if the X-ray of the complex target protein+bound drug is solved. The hardest part is to get the X-Ray of the protein and the MOST DIFFICULT part of this is to get pure crystals of protein or protein+drug complex! The reason beiing that most of of the interesting proteins targeted by drugs are trans-membrane proteins, like G-protein coupled receptors that bind most neurotransmitters or transporters lite DAT SERT, or ion channels like GABA, NMDA..etc etc.

They are all transmembrane proteins in their native form in the cells and it is very difficult to crystallize them without destroying them!! Unlike small organic molecules. Why is that? because they fall apart the minute you take them outside their natural environment ie cells membrane lipid bilayer. I mean the 3-D conformation of your protein collapse. The original 3-D tertiary structure the one of the native protein in the cell. You need a lot (i mean really a LOTS of trial and error of conditions, salts, detergents, temperature...etc etc) to get the right conditions if at all. Lots of interesting proteins like GABA channels are still near impossible to solve

It is like you want to have your cake and eat too: you have to use detergent (aka soap) to free them from the cell membrane but soap is exactly what you use to destroy their 3D and solubilize them in the first place! That's how you wash away food craps! So this is the hardest part of all that business: isolate your protein (or protein-drug complex) with its 3D native conformation intact. Depending on the protein, that can take a graduate student anywhere from 3 to 5 years to get good crystals of his target protein suitable for X-Ray IF AT ALL!!!

Oh I forgot another hard part: get the MONEY to fund your project (good luck asking gov give you money to make better cocaine!!!)

While we were talking on the DAT-cocaine complex X-ray structure, here is paper published 2 days ago by a Pharma in Danmark who use exactly the same technique I was talking about but with Serotonin transporter to design SSRI starting with SERT-citalopram X-ray structure, ($$paying article$$).. enjoy the read

The same technique can be used to design DRI starting with DAT X-ray...

 

[DotChem]

Haha ^ sorry for fucking the thread but this cracked me up

Have you met cathy noon? I bet you have Metcathinone..