It is worth considering that MXE has 3 distinct pharmacological effects:
1)NMDA antagonist
2)Dopamine transport inhibitor (DRI)
3)mu agonist (opiate)
While I would not claim to be a close friend of the MXE designer, I did know them well enough to have been involved in the testing of desoxyprpradrol & MDPV as well as MXE and relatives. I was lucky enough to have been offered (and taken) a sample of (R) ketamine. (R) ketamine has NO NMDA antagonist activity whatsoever but is a DRI. It was produced (resolved from commercial ketamine AFAIK) I am sure that readers with a rough understanding of the arylcyclohexylamine class to understand why KNOWING action of the K isomers was of material benefit.
MXE was developed from PCP via the following step (i.e. the compounds were made and tested by about 50 people so that the SUBJECTIVE QSAR could be elucidated).:
1)4-MeO PCP. Logic - PCP has too much DRI activity i.e. too stimulating and studies noted that 4-MeO PCP had significantly less DRI activity, 4-MeO PCP had been produced and tested in animal models in the 1960s but no human trials are recorded. Importantly, acute toxicity was studied. While this in now way proves a compound to be non-toxic, it does make it a MUCH safer bet than testing a totally novel compound. Even so, initial tests were carried out at 1% of the ED50 suggested by literature (which is a standard figure). From these, metabolism could be studied.
2)3-MeO-PCP Logic - papers from the 1960s showed that 3-MeO PCP & 3-OH PCP had significant mu agonist activity (while retaining NMDA antagonist and some DRI activity) and the designer suspected that the 3-MeO in particular would not only imbue opioid activity but also suppress DRI activity as the 4-MeO derivative of PCP did).
3)MXE Logic - he addition of a 2-keto moiety was known to increase opiate activity and that the N-monoehtyl (and monomethyl) homologues of PCP derivatives were of similar potenct. The ketone moiety also lowers the LogP of the compound. This means that no pharmacologically significant amounts of the compound could be stored in fat tissue (PCP was known to be stored in fat and lead to post-event symptoms).
Now, this may seem like a VERY short development pipeline but of course the designer actually read every paper, every patent and every other resource that provided information. The designer was able to (and did) provide a detailed QSAR and metabolic pathway(s) for each candidate. Those who know the designer will appreciate that they were formally educated to a level that allowed them to formally develop.
At this point, I think it's important the point out that the designer was well aware of what industrial resources were available. In some ways the 3-OH derivative would have been the more obvious option BUT the synthesis was somewhat more complex and would likely proceed via the 3-MeO. MXE can be produced by any supplier who was producing ketamine. As chemists will know, ketamine synthesis is relatively complex (compared to most RCs) and the low cost of ketamine is due to the fact that each step has been optimised. For a ketamine supplier to produce MXE requires only 1 precursor to be changed and the substitution is to an equally cheap and equally available chemical.
As a side note, QSAR data suggests that the N-methyl homologue of MXE should have significantly more opiate activity although it's NMDA activity (and so, I suppose, it's perceived overall activity) would be lower. I have never sampled it nor sought data but if anyone is in a position to comment, I would appreciate input.
A few years ago when I was datamining, I looked at the original ketamine patent and went through an exhaustive list of related patents. What I noted was that while the vast majority of patents attempt to cover as many compounds as possible, Parke Davis had patented a SINGLE compound to whit 5-methoxy ketamine i.e. 2-chloro-5-methoxy.
To be clear, ketamine's IUPAC name is:
2-(2-Chlorophenyl)-2-(methylamino)cyclohexanone
So the derivative's IUPAC name is:
2-(2-chloro-5-methoxyphenyl)-2-(methylamino)cyclohexan-1-one
It is important to note that while ketamine derivatives are chiral, all of these compounds are the racemes of all enantiomers.
This is very unusual. While the cost of a patent is not great for a large company, it does tie up a lot of resources and so the researchers must have convinced the legal department to spend the time and money seeking a patent. It would have also occupied the development team for some time. In short, they must have been convinced that said derivative had significant COMMERCIAL advantages. Never forget that whatever else, the compound was patented with a view to making a profit.
While I have to admit that I cannot recall the exact date of the latter patent, it was certainly a decade or more later and so people were producing ketamine commercially and so an improved derivative (carrying it's own patent) that could be produces with minimum retooling would be a significant commercial success.
If people wonder why this derivative did not succeed, well I have no simple answer but I suggest that it was simply because ketamine was so successful and so much work had been carried out on it's development that a new compound (regardless of similarity, the development cycle would be just as complex, time consuming and costly) was not deems cost effective.
For those seeking to understand it's NMDA antagonist activity, I suggest that people regard 8A-PDHQ which provides the active conformation of PCP) and then MK-801 (dizocilpine) which provides the exact relationship between the aromatic ring and the N: (I forget X,Y,Z but the angle is 107.5°. I Believe that further research showed that 107.5 is optimal although the research was carried out in the 1980s and so modern techniques (such as the use of radionucleotides) have reduced the granularity of such testing.
So, in short, the designer of MXE found the optimal compound. Not the most potent, not the simplest to synthesize and certainly not the cheapest BUT a safe, reliable product that even benefitted from concepts such as improvements to the synthesis of ketamine could almost certainly be applied to MXE.
But it is important to note that while PCP and ketamine and MXE are all arylcyclohexylamine derivatives, their is no necessity for them to be so. As the name of the class suggests, the only fixed moiety is the cyclohexyl and their is a body of research proving that there are bioisostesrs. Of course, the synthesis of these analogues (beyond the original research papers) has yet to be scales.
I am sure many people will suppose this to be an oversimplification but the KEY thing is that the only function of the cyclohexyl moiety is to produce a non-rotatable (rigid) bond between the aryl and the amine's lone-pair (N
. It MAY also provide some measure of space-filling which improves affinity although their are other NMDA antagonists that suggest that this is not a requirement. Conversely, the 2-keto moiety seen in ketamine, MXE and tiletamine most certainly DOES increase affinity and since simple alkyl groups (such as the methyl moiety of gacyclidine) DOES increase affinity. Singe both alkyl and keto moieties by similar amounts, it suggests space-filling rather than electronic character.
HOWEVER, the 2-keto moiety significantly increases opiate (mu) affinity of the scaffold. This is a good example of how complex balancing the 3 different activities can be. In passing it is worth mentioning that substitution of the cyclohexyl ring as it pertains to DRI activity has not been studied. Safe to say that compound such as benocyclidine are very potent and have a very steep dose-response curve (to the extant that while it is rarely encountered, it has been connected to a number of deaths). There is very little data on this subject and seemingly very little research. Even obvious homologues have not been synthesized.
So, putting aside the practical synthesis of a hypothetical compound, what size and type of training-set would be appropriate in the search for a compound that is a balances:
a)NMDA antagonist.
b)mu (opiate) agonist.
c)DRI antagonist (dopamine transport inhibitor).
That is QUITE a task. With that in mind, I offer a compound known to be a reasonably potent mu (opiate) agonist, NMDA antagonist and DRI. Sadly, it's development was something of an afterthought. The lead chemist was simply asked to produce a compound with (ideally) identically activity to an existing agent BUT that was not covered by any patents.
You will notice that I have only shown one of the 2 active (trans) isomers. The (1R,2S) isomer is a mu (opiate) agonist with a potency about x2 morphine (based on LogP & Ki). The (1S,2R) isomer is an NMDA antagonist with a potency around x2 ketamine (bases on LogP and Ki). Now, although no mention of DRI activity is stated, subjective accounts suggest DRI activity, the compound overlays camfentamine and (more elucidatory) isonortilidine overlays cypenamine.
I am quite prepared to admit that the last couple of paragraphs are based on new research but if one looks at the patent 'CYCLOALIPHATIC COMPOUNDS, ANALGESC COMPOSITIONS THEREOF AND METHOD OF USE THEREOF AS ANALGESCS' Glaxo Group 1979 (US 4,291,059) then their IS quite a lot of detail. I actually managed to contact Derek P. Reynolds, the medicinal chemist who designed the series of compounds and so I gained some valuable insights. One is that the bond-angle of the cyclohexene (nortilidine) and the cyclopentane (isonortilidine_ are very similar (but not identical). The positions of the N: are not identical (but similar).
When I say one can infer much detail from papers, an example would be that in the INT patent, one analogue listed is the m-OCH3 & m-OH. So what, I here you ask. Well, that suggests the orientation OF that ring while it overlays the A-ring of morphine. As far as I can tell, the cyclohezene/cyclopentane ring found in (nor)tilidine and INT does not overlay any of the rings found in morphine (although if I am wrong - please tell me. I would be GLAD to be wrong).
I could embark on a long diatribe but I will note two things. The amine moiety of opiates, while necessary, can appear in what initially seem to be unlikely places. In the case of tilidine, it is worth overlaying the A ring & N:. with morphine. You will discover that the cyclohexene (or cyclopentane in INT) while not overlaying a ring, do not present any biosteric bulk. In essence, it's just about the simplest chemical that will bridge the aromatic and the lone-pair.
I think cypenamine is sufficient for people to work on overlays but the NMDA activity is more complex. It is known that their are several sites within the NMDA receptor that will produce agonist and/or antagonist effects. I think it reasonable to suggest that diphenidine (and relatives) bind in a similar manner. Of course, known diphenidine homologues are both mu (opiate) agonists (e.g. lefetamine) and DRIs (pyrophenidine).
So, I hope others with more knowledge and ability will add.
1)NMDA antagonist
2)Dopamine transport inhibitor (DRI)
3)mu agonist (opiate)
While I would not claim to be a close friend of the MXE designer, I did know them well enough to have been involved in the testing of desoxyprpradrol & MDPV as well as MXE and relatives. I was lucky enough to have been offered (and taken) a sample of (R) ketamine. (R) ketamine has NO NMDA antagonist activity whatsoever but is a DRI. It was produced (resolved from commercial ketamine AFAIK) I am sure that readers with a rough understanding of the arylcyclohexylamine class to understand why KNOWING action of the K isomers was of material benefit.
MXE was developed from PCP via the following step (i.e. the compounds were made and tested by about 50 people so that the SUBJECTIVE QSAR could be elucidated).:
1)4-MeO PCP. Logic - PCP has too much DRI activity i.e. too stimulating and studies noted that 4-MeO PCP had significantly less DRI activity, 4-MeO PCP had been produced and tested in animal models in the 1960s but no human trials are recorded. Importantly, acute toxicity was studied. While this in now way proves a compound to be non-toxic, it does make it a MUCH safer bet than testing a totally novel compound. Even so, initial tests were carried out at 1% of the ED50 suggested by literature (which is a standard figure). From these, metabolism could be studied.
2)3-MeO-PCP Logic - papers from the 1960s showed that 3-MeO PCP & 3-OH PCP had significant mu agonist activity (while retaining NMDA antagonist and some DRI activity) and the designer suspected that the 3-MeO in particular would not only imbue opioid activity but also suppress DRI activity as the 4-MeO derivative of PCP did).
3)MXE Logic - he addition of a 2-keto moiety was known to increase opiate activity and that the N-monoehtyl (and monomethyl) homologues of PCP derivatives were of similar potenct. The ketone moiety also lowers the LogP of the compound. This means that no pharmacologically significant amounts of the compound could be stored in fat tissue (PCP was known to be stored in fat and lead to post-event symptoms).
Now, this may seem like a VERY short development pipeline but of course the designer actually read every paper, every patent and every other resource that provided information. The designer was able to (and did) provide a detailed QSAR and metabolic pathway(s) for each candidate. Those who know the designer will appreciate that they were formally educated to a level that allowed them to formally develop.
At this point, I think it's important the point out that the designer was well aware of what industrial resources were available. In some ways the 3-OH derivative would have been the more obvious option BUT the synthesis was somewhat more complex and would likely proceed via the 3-MeO. MXE can be produced by any supplier who was producing ketamine. As chemists will know, ketamine synthesis is relatively complex (compared to most RCs) and the low cost of ketamine is due to the fact that each step has been optimised. For a ketamine supplier to produce MXE requires only 1 precursor to be changed and the substitution is to an equally cheap and equally available chemical.
As a side note, QSAR data suggests that the N-methyl homologue of MXE should have significantly more opiate activity although it's NMDA activity (and so, I suppose, it's perceived overall activity) would be lower. I have never sampled it nor sought data but if anyone is in a position to comment, I would appreciate input.
A few years ago when I was datamining, I looked at the original ketamine patent and went through an exhaustive list of related patents. What I noted was that while the vast majority of patents attempt to cover as many compounds as possible, Parke Davis had patented a SINGLE compound to whit 5-methoxy ketamine i.e. 2-chloro-5-methoxy.
To be clear, ketamine's IUPAC name is:
2-(2-Chlorophenyl)-2-(methylamino)cyclohexanone
So the derivative's IUPAC name is:
2-(2-chloro-5-methoxyphenyl)-2-(methylamino)cyclohexan-1-one
It is important to note that while ketamine derivatives are chiral, all of these compounds are the racemes of all enantiomers.
This is very unusual. While the cost of a patent is not great for a large company, it does tie up a lot of resources and so the researchers must have convinced the legal department to spend the time and money seeking a patent. It would have also occupied the development team for some time. In short, they must have been convinced that said derivative had significant COMMERCIAL advantages. Never forget that whatever else, the compound was patented with a view to making a profit.
While I have to admit that I cannot recall the exact date of the latter patent, it was certainly a decade or more later and so people were producing ketamine commercially and so an improved derivative (carrying it's own patent) that could be produces with minimum retooling would be a significant commercial success.
If people wonder why this derivative did not succeed, well I have no simple answer but I suggest that it was simply because ketamine was so successful and so much work had been carried out on it's development that a new compound (regardless of similarity, the development cycle would be just as complex, time consuming and costly) was not deems cost effective.
For those seeking to understand it's NMDA antagonist activity, I suggest that people regard 8A-PDHQ which provides the active conformation of PCP) and then MK-801 (dizocilpine) which provides the exact relationship between the aromatic ring and the N: (I forget X,Y,Z but the angle is 107.5°. I Believe that further research showed that 107.5 is optimal although the research was carried out in the 1980s and so modern techniques (such as the use of radionucleotides) have reduced the granularity of such testing.
So, in short, the designer of MXE found the optimal compound. Not the most potent, not the simplest to synthesize and certainly not the cheapest BUT a safe, reliable product that even benefitted from concepts such as improvements to the synthesis of ketamine could almost certainly be applied to MXE.
But it is important to note that while PCP and ketamine and MXE are all arylcyclohexylamine derivatives, their is no necessity for them to be so. As the name of the class suggests, the only fixed moiety is the cyclohexyl and their is a body of research proving that there are bioisostesrs. Of course, the synthesis of these analogues (beyond the original research papers) has yet to be scales.
I am sure many people will suppose this to be an oversimplification but the KEY thing is that the only function of the cyclohexyl moiety is to produce a non-rotatable (rigid) bond between the aryl and the amine's lone-pair (N
. It MAY also provide some measure of space-filling which improves affinity although their are other NMDA antagonists that suggest that this is not a requirement. Conversely, the 2-keto moiety seen in ketamine, MXE and tiletamine most certainly DOES increase affinity and since simple alkyl groups (such as the methyl moiety of gacyclidine) DOES increase affinity. Singe both alkyl and keto moieties by similar amounts, it suggests space-filling rather than electronic character.HOWEVER, the 2-keto moiety significantly increases opiate (mu) affinity of the scaffold. This is a good example of how complex balancing the 3 different activities can be. In passing it is worth mentioning that substitution of the cyclohexyl ring as it pertains to DRI activity has not been studied. Safe to say that compound such as benocyclidine are very potent and have a very steep dose-response curve (to the extant that while it is rarely encountered, it has been connected to a number of deaths). There is very little data on this subject and seemingly very little research. Even obvious homologues have not been synthesized.
So, putting aside the practical synthesis of a hypothetical compound, what size and type of training-set would be appropriate in the search for a compound that is a balances:
a)NMDA antagonist.
b)mu (opiate) agonist.
c)DRI antagonist (dopamine transport inhibitor).
That is QUITE a task. With that in mind, I offer a compound known to be a reasonably potent mu (opiate) agonist, NMDA antagonist and DRI. Sadly, it's development was something of an afterthought. The lead chemist was simply asked to produce a compound with (ideally) identically activity to an existing agent BUT that was not covered by any patents.
You will notice that I have only shown one of the 2 active (trans) isomers. The (1R,2S) isomer is a mu (opiate) agonist with a potency about x2 morphine (based on LogP & Ki). The (1S,2R) isomer is an NMDA antagonist with a potency around x2 ketamine (bases on LogP and Ki). Now, although no mention of DRI activity is stated, subjective accounts suggest DRI activity, the compound overlays camfentamine and (more elucidatory) isonortilidine overlays cypenamine.
I am quite prepared to admit that the last couple of paragraphs are based on new research but if one looks at the patent 'CYCLOALIPHATIC COMPOUNDS, ANALGESC COMPOSITIONS THEREOF AND METHOD OF USE THEREOF AS ANALGESCS' Glaxo Group 1979 (US 4,291,059) then their IS quite a lot of detail. I actually managed to contact Derek P. Reynolds, the medicinal chemist who designed the series of compounds and so I gained some valuable insights. One is that the bond-angle of the cyclohexene (nortilidine) and the cyclopentane (isonortilidine_ are very similar (but not identical). The positions of the N: are not identical (but similar).
When I say one can infer much detail from papers, an example would be that in the INT patent, one analogue listed is the m-OCH3 & m-OH. So what, I here you ask. Well, that suggests the orientation OF that ring while it overlays the A-ring of morphine. As far as I can tell, the cyclohezene/cyclopentane ring found in (nor)tilidine and INT does not overlay any of the rings found in morphine (although if I am wrong - please tell me. I would be GLAD to be wrong).
I could embark on a long diatribe but I will note two things. The amine moiety of opiates, while necessary, can appear in what initially seem to be unlikely places. In the case of tilidine, it is worth overlaying the A ring & N:. with morphine. You will discover that the cyclohexene (or cyclopentane in INT) while not overlaying a ring, do not present any biosteric bulk. In essence, it's just about the simplest chemical that will bridge the aromatic and the lone-pair.
I think cypenamine is sufficient for people to work on overlays but the NMDA activity is more complex. It is known that their are several sites within the NMDA receptor that will produce agonist and/or antagonist effects. I think it reasonable to suggest that diphenidine (and relatives) bind in a similar manner. Of course, known diphenidine homologues are both mu (opiate) agonists (e.g. lefetamine) and DRIs (pyrophenidine).
So, I hope others with more knowledge and ability will add.
Last edited:
