Chemistry of opiates

[Ulcer]

Hi guys (and girls),

I'm a pharmacology student that has been studying opiates recently, but being a chemist at heart I noticed that none of the opioid compounds I've come across seem to contain carboxylic acids or other anion type functional groups. I've looked at the structures of maybe 100 different compounds with opioid activity (mu, kappa and delta), and none of them feature this functional group. I think every compound I've looked at contained an amine (usually tertiary), so I suppose they could form cations, but no anions are ever present.

I asked the coordinator of my class why this was, but he didn't really have an answer for it. Just a throwaway sort of line about the opioid receptors probably being hydrophobic, which makes some sense, I admit, but its hardly a conclusive answer I thought.

So...

Does anybody know of any carboxylic acids/anions in opioid compounds? If there aren't any, is there any particular reason for it?

Regards,
Ulcer

PS. I'm obviously new here, so I don't quite know where this type of thread would fit into the site, I just chose a place that seemed most likely, but I didn't dig around too deep. If its in the wrong place, could a mod or someone please move it? Thanks.

 

[anterrabae]

Your question would probably receive better, more informative answers in Advanced Drug Discussion. Also, welcome to Bluelight.

 

[Cane2theLeft]

^ yeah, let's try this over there.



OD>>>>ADD

 

[apsig]

I think the best bet would be to look on pubmed to see if you can dig up any papers with calculated docking structures of opioid drugs. I don't know the extent of your knowledge on receptor binding, but the chemical properties of these ligands are determined by the shape of the binding pocket, the dimensions of the binding pocket, and the types of amino acid residues that make up the binding pocket. As your coordinator said, the binding pocket likely contains hydrophobic amino acids and as such would be disinclined to participate in binding with polar/ionic molecules.

Let us know what you uncover. My apologies, as my knowledge of this particular topic is limited and I really should be studying for the exam I have in 55 mins rather than posting here...

 

[sekio]

Carboxylic acid-type compounds self-ionize to some extent and are rarely water soluble unless the "acid" forms a salt and has a light alkyl chain.

The motif that I've always worked with is that adding a carboxylic acid or other anionic group will prevent movement across the blood brain barrier. Even if the molecule still makes it across it usually loses activity due to greatly increased polarity. (THC-OH vs THC-COOH, methylphenidate vs ritalinic acid) This is seen with e.g. hydroxyzine to cetirizine. It is generally undesirable in centrally active drugs. Another example is the rapid loss of activity for e.g. cocaine and methylphenidate once the ester links are blown apart.

Blood pH is generally slightly alkaline, that probably explains the prevalence of cationic compounds that are "less charged" than comparable anions.

 

[Hyperthesis]

Carboxylic acid-type compounds self-ionize to some extent and are rarely water soluble unless the "acid" forms a salt and has a light alkyl chain.

Ionization actually enhances polarity of a molecule and thus water-solubility, too!


To the OP:
The requirements for a "good" opioid depend on the amino acids in the respective receptors, which mediate the contact between ligand and protein. The presence of a basic moiety in (almost) all opioids indicates that there must be one or more acidic amino acids involved, ie. glutamic acid and aspartic acid.
I don't want to make this post too lenghty, hence I'll stick with the µ-receptor for the moment: There are 2 major and distinct binding sites for agonists at the µ-receptor (well, "distinct" is relative here, as the 2 sites overlap partially). Although this is quite a gross simplification, the first site is addressed mainly by opioids fulfilling the morphine-rule*, while the second one is mainly addressed by the fentanyl-family and related structures.
The µ-receptor is a member of the G protein-coupled receptor-family, a.k.a. seven-transmembrane domain receptors. According to the literature are the transmembrane helices III, VI and VII most important for agonist ligand binding. In these morphine adresses the following residues:

• Asp147 in TMIII: salt bridge with morphine's amine-function, which is protonated at physiological pH
  Tyr148 in TMIII: π-π-stacking with morphine's A-ring and H-bond with the 3-hydroxyl
  Tyr299: π-π-stacking with morphine's A-ring
  Lys303: H-bond with the 3-hydroxyl

Ref:
Bioorganic & Medicinal Chemistry 1996, 4(12): 2151

Please note that some publications use a slightly different numbering scheme. Furthermore can the proposed binding differ between different publications, too, although the presented mode seems to meet broad acceptance.


* The morphine-rule is a several decades old but nonetheless valid, empirically established SAR of the µ-opioid-receptor. According to this rule, the following structural features make up a µ-agonist:

1. A tertiary nitrogen with a small alkyl substituent.
2. A quaternary carbon.
3. A phenyl group or its isosteric equivalent directly attached to the quaternary carbon.
4. A 2 carbon spacer between the quaternary carbon and the tertiary nitrogen.​

The majority of known µ-agonists agree more or less with this rule, the best known exception being the fentanyls (which can be explained by their different binding site).

 

[Ulcer]

There are 2 major and distinct binding sites for agonists at the µ-receptor (well, "distinct" is relative here, as the 2 sites overlap partially). Although this is quite a gross simplification, the first site is addressed mainly by opioids fulfilling the morphine-rule*, while the second one is mainly addressed by the fentanyl-family and related structures.
The µ-receptor is a member of the G protein-coupled receptor-family, a.k.a. seven-transmembrane domain receptors. According to the literature are the transmembrane helices III, VI and VII most important for agonist ligand binding.

I find this 2 binding site notion quite interesting. Is much understood about the fentanyl binding region? From what I understand making metric judgements regarding opioid receptors is exceptionally difficult because the receptors have never been successfully crystallised (nor any other GPCRs I think), so the physical structure of the receptors is not well understood. I'm most interested in the fentanyl region, because I've been reading about the drug herkinorin, a non-alkaloid mu receptor agonist.

• Asp147 in TMIII: salt bridge with morphine's amine-function, which is protonated at physiological pH
  Tyr148 in TMIII: π-π-stacking with morphine's A-ring and H-bond with the 3-hydroxyl
  Tyr299: π-π-stacking with morphine's A-ring
  Lys303: H-bond with the 3-hydroxyl

This is quite interesting. Do such amino acid role designations exist for the fentanyl region? Knowing what I know about the structure of fentanyl, I imagine the residues would be more likely to favour pi-pi stacking interactions.

Thanks for this, its been some interesting reading.

 

[Hyperthesis]

From what I understand making metric judgements regarding opioid receptors is exceptionally difficult because the receptors have never been successfully crystallised (nor any other GPCRs I think), so the physical structure of the receptors is not well understood

Since last year there were 3 or 4 (IIRC) structures of human GPCRs crystallized and published.

I look into the binding mode of fentanyl later.

 

[IHateOpiophobes]

Since last year there were 3 or 4 (IIRC) structures of human GPCRs crystallized and published.

I look into the binding mode of fentanyl later.

If I remember right G-Linked Coupled Proteins had three sub-units alpha-beta-gamma but looking at the base of all morphinan based molecules the chiral charbon with stereospecificity are bridge carbons on the the necessary tertiary amine ring. (THEBAINE)- I would think also that the outer groups would be easier to reduce making it more potent.....while a saturated ketone bond would be harder to oxidize. Too much concentrated charge. Its obvious that the OH parts of the group react with the lipid reigon of the membrane protiens (I GUESS GCP - Above) getting into the BBB. This a result of the di-acetylation of those groups to make heroin from morphine. I noticed the OH groups are beta -2-away to double bonds and therefore resonance stabilized if the molecule is deprotonated making an anion (LIKE THE ORIGINAL POSTER WAS TALKING ABOUT, making a tight fit to the polar reigons on the trans-membrane proteins.

The 3-hydroxyl anion would be very stable and therefore reactive. I dont know how exactly stereospecificity affects the outcome because G-Linked Proteins act in a cascade effect depending on the intereaction with the first alpha protein sub-unit. Maybe if the piridine ring is not fused properly it loses some possible activity. Maybe like buprenorphine. Bupe is really weird with alot of steric hindrance on one side. t-butyl and a cyclopropyl group if I remember right. So I don't see alot going on with that end. Maybe thats why it has a cieling...because so much eventually gets in it's own way.???? IDK.

Interesting though.
Love to learn more.

 

[IHateOpiophobes]

If I remember right G-Linked Coupled Proteins had three sub-units alpha-beta-gamma but looking at the base of all morphinan based molecules the chiral charbon with stereospecificity are bridge carbons on the the necessary tertiary amine ring. (THEBAINE)- I would think also that the outer groups would be easier to reduce making it more potent.....while a saturated ketone bond would be harder to oxidize. Too much concentrated charge. Its obvious that the OH parts of the group react with the lipid reigon of the membrane protiens (I GUESS GCP - Above) getting into the BBB. This a result of the di-acetylation of those groups to make heroin from morphine. I noticed the OH groups are beta -2-away to double bonds and therefore resonance stabilized if the molecule is deprotonated making an anion (LIKE THE ORIGINAL POSTER WAS TALKING ABOUT, making a tight fit to the polar reigons on the trans-membrane proteins.

The 3-hydroxyl anion would be very stable and therefore reactive. I dont know how exactly stereospecificity affects the outcome because G-Linked Proteins act in a cascade effect depending on the intereaction with the first alpha protein sub-unit. Maybe if the piridine ring is not fused properly it loses some possible activity. Maybe like buprenorphine. Bupe is really weird with alot of steric hindrance on one side. t-butyl and a cyclopropyl group if I remember right. So I don't see alot going on with that end. Maybe thats why it has a cieling...because so much eventually gets in it's own way.???? IDK.

Interesting though.
Love to learn more.

Also...
I noticed that the nitrogen substituent if more than methyl is bad. its cyclopropyl in bupe and propene in naltrexone. Both antagonists in their own way.

 

[Ulcer]

I've been reading more about herkinorin (which I find fascinating due to its lack of an amine group, tertiary or otherwise) and from one paper, it seems that herkinorin primarily exerts its activity in periphery, possibly due to instability of an ester group (which ester it refers to I'm unsure of since herkinorin technically contains 3: cyclic valerolactone, methyl and benzyl ester).
(source: Groer 2007. An Opioid Agonist that Does Not Induce mu-Opioid Receptor—Arrestin Interactions or Receptor Internalization Mol Pharmacol 71:549–557)

If they refer to the methyl ester, then it kind of describes the situation I've been talking about, because it will liberate methanol from herkinorin and leave behind a carboxylic acid with pKa ~4, which will be readily ionised at physiological pH.

I doubt they'd refer to the valerolactone ring, because destablisation of that group would drastically alter the structure of the compound and quite likely alter the activity. Also, I doubt that they're talking about the benzyl ring because removing that would just leave behind a hydroxy group on the structure, which I believe is the structure of salvinorin B.

At any rate: interesting. Any thoughts?

Also: I did a little bit of reading about GCPR crystal structures, but could only find the rhodopsin crystal structure. Which I suppose is enough to build a homology model off, but how accurate it would be is up for debate.

 

[sekio]

The methyl ester in herkinorin would be hydrolised first.

 

[Hyperthesis]

If I remember right G-Linked Coupled Proteins had three sub-units alpha-beta-gamma but looking at the base of all morphinan based molecules the chiral charbon with stereospecificity are bridge carbons on the the necessary tertiary amine ring. (THEBAINE)

What is the message of this sentence? I can't find any. The mentioned facts (a: 3 subunits in GPCRs; b: structural features of opioid morphinans) have nothing to do with each other.

I would think also that the outer groups would be easier to reduce making it more potent.....while a saturated ketone bond would be harder to oxidize.

Again: What is the supposed message? Ketones can not be oxidized. What are "outer groups"? You are speaking in riddles.

Too much concentrated charge.

Pointless statement No 3.

Its obvious that the OH parts of the group react with the lipid reigon of the membrane protiens (I GUESS GCP - Above) getting into the BBB.

Oh - my - god ...
Neither do ligands usually react with their respective receptors, nor does any ligand get into the BBB, but rather beyond it. Furthermore are ligand-binding and BBB-passage two unrelated processes.

This a result of the di-acetylation of those groups to make heroin from morphine.

???

I noticed the OH groups are beta -2-away to double bonds and therefore resonance stabilized if the molecule is deprotonated making an anion (LIKE THE ORIGINAL POSTER WAS TALKING ABOUT

No, he actually wasn't.
And what is "beta-2-away" supposed to mean?

The 3-hydroxyl anion would be very stable and therefore reactive.

General rule of thumb in chemistry: Stable ≠ reactive!

I dont know how exactly stereospecificity affects the outcome because G-Linked Proteins act in a cascade effect depending on the intereaction with the first alpha protein sub-unit.

Check the definition of "stereospecifity": http://en.wikipedia.org/wiki/Stereospecificity
It's a term used for describing reaction mechanism. As said before: Ligands usually do not react with their receptors. They simply bind in a reversible manner. The number of exceptions is incredibly small and not worth mentioning, in particular not with respect to morphinane-based opioids.

Love to learn more.

Good idea! I strongly recommend to get a textbook on basic organic chemistry.

 

[Hyperthesis]

Crystal structures of human GPCRs:

"Structure of an Agonist-Bound Human A2A Adenosine Receptor."
Science, 10 March 2011
DOI: 10.1126/science.1202793

"High-Resolution Crystal Structure of an Engineered Human β2-Adrenergic G Protein–Coupled Receptor"
Science 23 November 2007: vol. 318 no. 5854 pp. 1258-1265
​

...and there are more out there. Search at the "Protein Database" (www.pdb.org)

 

[CrimpJiggler]

A carboxyl group would turn it into a zwitterion so I'm guessing it would cross the BBB only when at its isoelectric point. The keyword there is guess, I don't really know much about the BBB but if I'm not mistaken, charged molecules don't cross it easily. I've never come across an opioid with a carboxyl group either.

 

[Hyperthesis]

There are some great pictures from a very recent article regarding opioid SAR, including agonists, partial agonists and antagonists from the morphinane-family, but also other more 'exotic' opioid classes.





Quoting the part that explains Figure 5:

Based on the presented AB´BS CSP model in combination with the large body of experimental μ-opioid SAR data, we
propose the following consensus pharmacophore for the efficacy of opioids targeting the μ receptor (Figure 5). Traditional
agonists (e.g., morphine) lacking or having short N substituents have NH⁺-receptor Asp interactions that lead to the receptor assuming an active conformation and do not interact with the B binding site (Figure 5a). In agonists with large B substituents, such as etorphine, interactions occur with both the N and B sites to maximize efficacy over that of morphine (Figure 5b). In contrast, traditional antagonists cannot interact with the essential Asp as required for the receptor to assume a conformation required for agonism due to their bulky N-substituents, and they also cannot interact with the B site (Figure 5c). Partial agonists, such as buprenorphine, have a favorable interaction with the B site, which partially overcomes the negative impact of bulky N substituents, leading to partial efficacy (Figure 5d). In this model, the shorter C19 substituent of the antagonist diprenorphine disallows interactions with the B site, thereby not being able to overcome the presence of the CPM N substituent. Thus, interactions with the B site modulate the extent of agonism associated with N-receptor Asp interactions. This can lead to enhanced efficacy when the basic N is not blocked, as with etorphine, or to partial agonism where interactions with the B site partially overcome the presence of large N substituents, yielding partial agonists, as with buprenorphine.

Source:
J. Phys. Chem. B, 2011, 115 (22), pp 7487–7496
DOI: 10.1021/jp202542g

Absolutely GREAT material and food for thought!

 

[Hyperthesis]

Binding mode of fentanyl

Results from Neural Regeration Research 2011, 6(4): 267 (employed agonistic ligands in this theoretical study were

Protonated nitrogen on piperidine interacted with negatively charged Asp147 via electrostatic and hydrogen bonding. Asp 147 was the primary binding site and served as a counterion for the protonated nitrogen of opioid ligands. The carbonyl oxygen of ligand with positively charged nitrogen of His297 was at the same electrostatic and hydrogen bonding interactions. The pyridine cycle might interact with side chains of residues Ile144 and Val143 via hydrophobic interactions. Interestingly, the two phenyl rings of ligands aligned parallel with the phenyl ring of Trp293 and Trp328 via an aromatic π-π stacking interaction. In this model, residues Phe221, Ser222, and Leu219 formed a small hydrophobic pocket, which occupied the R2 substitution position and resulted in increased hydrophobic interactions with ligands. Therefore, ethyl acyloxy or ethyl ester exhibited much higher activity than methyl acyloxy and methyl ester. The negatively charged residue Asp147 and positively charged residue His297 were identified as possible active sites. The importance of close residues in TM2(Asp114), TM3(Asp147), and TM6(His297) has been demonstrated by a modest involvement of N- and C-terminal domains in ligand-receptor interactions. Protonated nitrogen was close to Asp147 of TM3, whose HN⁺-O^- distance was 0.34 nm, although a good salt bridge HN⁺–O^- distance was 0.373 nm.

Unfortunately, the pictures from that article are crappy and one can hardly recognize anything. But it looks like the binding modes of the morphinanes-class, eg. morphine, and the fentanyl-class of compounds are not that different than initially thought.

 

[Refluxer]

Also...
I noticed that the nitrogen substituent if more than methyl is bad. its cyclopropyl in bupe and propene in naltrexone. Both antagonists in their own way.

N-Phenethylmorphine is a lot more potent than morhine, about 14x if I don't remember wrongly.

About the OP question. There two main points with the nitrogen: 1. Pharmacokinetics 2. Interaction with the receptor

Amines are very common in drugs in large because they offer good absorption. The amine is protonated in the stomach which gives good solubility, while in the intestines the pH rises and the amine gets more and more deprotonated which makes it more lipophilic and able to cross the intestinal walls into the blood. Also, in the blood pH is approximately 7.4.

 

[CrimpJiggler]

There are some great pictures from a very recent article regarding opioid SAR, including agonists, partial agonists and antagonists from the morphinane-family, but also other more 'exotic' opioid classes.

Thanks!

 

[Ulcer]

The methyl ester in herkinorin would be hydrolised first.

If this is the case, it provides an interesting answer to my question.

Fact 1: Herkinorin is 70% active in periphery (determined by observation of 70% loss of efficacy in an antagonist study).
Fact 2: Herkinorin is 30% active in CNS.

Thought 1: Herkinorin may undergo ester hydrolysis in the CNS at the methyl ester position to yield an anion product.
Thought 2: Activity of herkinorin is significantly decreased in the CNS where the anion product may be present.

Possible conclusion: An anion may be present in opioid compounds (at least in mu opioids), but it may result in a significant decrease in its binding affinity.

So... That could possibly be a reason as to why opioids don't contain anion groups, because it greatly lowers the efficacy of the compounds.

I wonder though, if an anion opioid may be effective as a weak analgesic much in the same way that codeine is used as a weak opioid analgesic (by weak, I mean in comparison to fentanyl, heroin etc.)