Author Topic: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth  (Read 420 times)

Dope Amine

  • Pupae
  • **
  • Posts: 59
Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« on: February 20, 2013, 02:24:30 AM »
SUBJECT OVERVIEW:

We learn from Synthesis and Biological Evaluation of 14-Alkoxymorphinans. 18.1 N-Substituted 14-Phenylpropyloxymorphinan-6-ones with Unanticipated Agonist Properties:
Extending the Scope of Common Structure-Activity Relationships

J. Med. Chem. 2003, 46, 1758-1763
  • Most excitingly, 14-phenylpropoxy-naltrexone is found surprisingly to be a potent agonist (Ki for mu receptor: 0.34, 600x morphine in TF)  BUT has poor selectivity for the µ receptor with ratios of  K/µ=1.96 and ?/µ=1.84 (so slightly less than double the preference for µ (larger numbers mean higher selectivity for µ receptor).
  • 14-phenylpropoxy-N-propyloxymorphone (swapping the N-CPM in the above structure for N-propyl) is even found to be even more potent (950x morphine in TF) and has better µ selectivity of K/µ=4.11 and ?/µ=10.3.
  • morphine Ki selectivity ratios: K/µ and ?/µ are 17.3 and 33.1 respectively.
  • Oxymorphone (known to be quite euphoric) ratios are 63 and 83.
  • 14-methyloxymorphone, ratios are 102 and 48. (Sounds like this might be better euphoria than oxymorphone  ;D)
  • 14-methoxymetopon, which is a very powerful µ agonist, ratios are 13217 and 1178.
 
It is believed by someone much more well versed in opioid pharmacology that the ideal opioid would be a ?/?-agonist and ORL1(?3)-antagonist. 

I wonder, between the two above cited 17-substituted 14-phenylpropoxynoroxymorphone structures, which one would be better to go with while keeping in mind the above provided pharacological considerations: choosing the N-propyl so that there was greater mu selectivity but at the cost of sigma affinity being less than K affinity OR N-CPM so that sigma affinity is increased to greater than half of mu but with K affinity increased to almost as much as mu?  This question is only food for thought because I will be presenting below other modifications which I believe will likely be more promising.  Also, in case anyone is not familiar with opioid receptor subtypes:
Pure µ agonists (like fentanyl) lack any appreciable euphoria and in the author's opinion not worth bothering with.
Any appreciable kappa opioid agonism is known to produce dysphoric effects such as disassociation and hallucination.  Even people who enjoy K agonists such as salvinorin A would not likely want that experience mixed in with the basking in warmth and comfort that can be found in opioid bliss.  So in this comparison, based on the negative effects associated with opioids showing a lack of significant preferance of µ over K, I'd say it would be most important to avoid those K effects, and so in this case the N-propyl structure would likely be the better candidate (and an exceedingly potent one too!)...

There is another relevant article that I would greatly appreciate if someone could kindly share:
Synthesis of N-isobutylnoroxymorphone from naltrexone by a selective cyclopropane ring opening reaction.
Med Chem Lett. 2008 Sep 15;18(18):4978-81.
doi: 10.1016/j.bmcl.2008.08.019. Epub 2008 Aug 12.

From the abstract we find that they selectively open the cyclopropyl ring of naltrexone (using Pt(IV) oxide, hydrobromic acid and H2), affording N-isobutynoroxymorphone.  They claim that, surprisingly, it produced dose dependent analgesic effects in the mouse acetic acid writhing test, yielding an ED(50) of 5.05 mg/kg, sc. (roughly equivalent to morphine).

BUT, I surmise that this article (or at least the abstract) is mentioning only part of the action.  I propose that N-isobutyl-noroxymorphone is most likely a mixed agonist/antagonist (AKA dualist).  The reasoning for this is because N-propyl-noroxymorphone, which might be considered the closest N-substituted analogue, has been reported to be both one of a number of N-sub'd dualists as well as specifically reported to be a pure antagonist  (See: Current Medicinal Chemistry. 1996;1(6):427-29 and Binding properties of the pure opioid antagonist [3H](N-propyl)-noroxymorphone in rat brain membranes., Neurobiology (Bp). 1993;1(4):327-35.).

So I think that N-isobutyl-noroxymorphone is most likely a dualist but has only been examined once, to my knowledge, and was likely only examined with regard to it's agonist properties.

Further related work is found in:
14ß-O-Cinnamoylnaltrexone and Related Dihydrocodeinones are Mu Opioid Receptor Partial Agonists with Predominant Antagonist Activity
J. Med. Chem. 2009, 52, 1553–1557

The most noteworthy compound discussed is 14-O-Cinnamoylnaltrexone which has a reported Ki for the mu receptor of 0.40 and selectivity ratios: K/µ=8.5 and sigma/µ=9.0.  "It is of interest to compare the activity of 14-cinnamoylnaltrexone with that of the phenylpropyl ether which is structurally similar in having a three-carbon chain linking the side chain aromatic ring to the C14 oxygen atom. The ether (2a) in vivo gave a full response in a battery of thermal antinociceptive assays with potency up to 400 times greater than that of morphine. In comparison, the cinnamoyl ester has much more modest in vitro and in vivo MOR agonist activity. It must be assumed that the relative conformational restraint of the alpha-beta unsaturated ester prevents an optimum interaction with MOR in the preferred agonist conformation". The optimum conformation to which they speak of, by the way, is with the aromatic ring (14-phenylpropoxy, 14-hydrocinnamoyl, etc) perpedicular and slightly below morphinan ring C.

The conformational restraint hypothesis proposed above appears reasonable considering that the combined effect of the carbonyl and alkene portions of the cinamoyl group essentially lock up all three chain carbons and allowing very limited movement besides the angled rotational freedom at the oxygen which connects the ester and the rotation of the phenyl ring. BUT if this were the complete explanation then how could one explain the significant potency of 14-cinnamoyloxycodeinone,  which is reported to be a potent agonist of roughly 178x morphine determined via the tail clip method in mice ("QSAR of Narcotic Analgetic Agents"  NIDA Research Monograph 1978 (22): 186–196. PMID 30907.).  What is even more strange is that 14-hydrocinnamoyloxycodeinone, which should have greater conformational freedom due to the reduction of the alkene bond, is actually found by the same method to be slightly weaker with a tested potency of 115x morphine.  The one part of the established SAR that does actually make sense is that the relatively weaker activity compared to the 14-phenylpropoxy ethers can be expected because of the 3-methyl ether of the codeinones compared to the morphones.  But the question that still remains is does the 7-8 double bond of the codeinones somehow assist both of these 14-esters, especially the most restricted ester (cinnamoyl) into attaining the optimal conformation?  Based on the results of 14-cinnamoylnaltrexone one would expect 14-cinnamoylcodeinone to have even less agonist activity than 14-hydrocinnamoylcodeinone, but the opposite is the case.  If all reports are true and complete then the only explanation I can imagine is that the  7-8 double bond must somehow be helping orient the esters, especially the unsaturated ester, into the optimal conformation.  I have yet to find a plausible explanation via Chemdraw 3-D.


SYNTHETIC APPROACHES USING NALTREXONE AND POTENTIAL PRODUCTS:

In the first paper that was listed above (agonist naltrexone analogues), the synthesis of 14-phenylpropoxy-naltrexone starts with 14-hydroxycodeinone being deprotonated by NaH and then alkylated with cinnamyl bromide.  Then both the cinnamyl double bond and the 7-8 double bond is reduced via Pd/C and hydrogen.

This alkylation of alcohols to form ethers is not particularly friendly to the hobby chemist as it requires a very hazardous base (NaH) and an intert atmosphere (unless someone can suggest a more accessible alternative?).  What is interesting though is that they do not alkylate with hydrocinnamyl bromide but rather reduce the double bond after alkylation.

I propose that N-isobutyl-14-hydrocinnamoyl-nornaltrexone and/or N-isobutyl-14-hydrocinnamoyl-7,8-dihydronormorphine may be strong opioids which can be attained through relatively easy steps.  First, reaction of naltrexone with 2.2 equivalents of cinnamoyl bromide in pyridine (water-sensitive) would yield 3,14-dicinnamoylnaltrexone.  Then Catalytic Transfer Hydrogenation with Pd/C 10%, formic acid and a bit of potassium hydroxide at r.t. will definitely open up the n-cyclopropylmethyl to form n-isobutyl, reduce the 14-cinnamoyl to 14-hydrocinnamoyl, most likely reduce off the 3-ester and probably will convert the 6-keto to an alcohol.

Whether or not the 6-keto group is reduced, either structure should prove to be more selective and give better effects with the n-isobutyl group (more similar to the n-propyl) than similar structures with the n-cyclopropylmethyl group.  Cyclopropyl rings are very reactive to ring-opening reductions (more reactive than alkenes to reduction) BTW.

Naltrexone is readily available.  Who has the time and the resources to check out whichever potent opioid is the actual product from these two relatively easy reactions?  Both structures are able to obtain the ideal conformation for opioid mu receptor agonism.
« Last Edit: March 06, 2013, 10:30:18 AM by Dope Amine »

RoidRage

  • Dominant Queen
  • ****
  • Posts: 386
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #1 on: February 20, 2013, 02:54:47 AM »
First 2 papers requested:

Synthesis and Biological Evaluation of 14-Alkoxymorphinans. 18.1 N-Substituted 14-Phenylpropyloxymorphinan-6-ones with Unanticipated Agonist Properties:? Extending the Scope of Common Structure?Activity Relationships
Elisabeth Greiner ,§# Mariana Spetea ,§ Roland Krassnig ,§ Falko Schüllner ,§ Mario Aceto ,† Louis S. Harris ,† John R. Traynor ,‡ James H. Woods ,‡ Andrew Coop ,? and Helmut Schmidhammer *§
J. Med. Chem., 2003, 46 (9), pp 1758–1763
DOI: 10.1021/jm021118o

Synthesis of N-isobutylnoroxymorphone from naltrexone by a selective cyclopropane ring opening reaction
Hideaki Fujiia, Yumiko Osaa, Marina Ishiharaa, Shinichi Hanamuraa, Toru Nemotoa, Mayumi Nakajimab, Ko Hasebeb, Hidenori Mochizukib, Hiroshi Nagasea
Bioorganic & Medicinal Chemistry Letters
Volume 18, Issue 18, 15 September 2008, Pages 4978–4981
DOI:10.1016/j.bmcl.2008.08.019

Assyl Fartrate

  • Subordinate Wasp
  • ***
  • Posts: 229
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #2 on: February 20, 2013, 07:29:45 AM »
Search through some of Assyl's posts and you'll find a lot of relevant information on the subject. The reason they use the codeinone, as opposed to the codone, is because you'll get alkylation at carbons 5 and 7 (enolate chemistry) otherwise - unless you ketalize it first, of course, which they do when preparing the 14-O-benzyl ether of oxycodone.

It is possible that less selectivity may result in more euphoria, and this is supported by one of Schmidhammer's papers on the matter, where 14-O-benzylmetopon is found to not only lack side effects such as constipation and respiratory depression, but also to have a "far greater" effect on the "emotive component of pain."

A promising candidate for an agonist derived from naltrexone is the 14-phenylpropanoic ester. Strong agonists tend to have a pair of aromatic rings oriented axially to each other, and this could offer such interactions. As the paper you shared states, the cinnamoyl version is a partial agonist, and this lacks the freedom of rotation that could allow access to the desired orientation.

Also, there is a paper (search Assyl's posts) that strongly suggests cinnamyl alcohol and iodine in acetonitrile (and hopefully other solvents) can be used to form cinnamyl ethers. There was an example where it worked with t-butyl alcohol, so it seems promising. Even a nominal yield would mean significant results, given the potency of said substances.
Someone Who Is Me

Dope Amine

  • Pupae
  • **
  • Posts: 59
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #3 on: February 20, 2013, 11:18:29 PM »
RoidRage: Thank you so much for the fast response with the papers! I look forward to reading through them.

Assyl:  Thanks so much for your thoughtful commentary on some of the issues I listed.  The enolate issue with alkylation is very enlightening... and disappointing. 

That is interesting that you state that less selectivity could be better.  I will have to search for the selectivity ratios of your mentioned 14-o-benzylmetopon for comparison.   I will have to fix the characters in my post- those are supposed to be kappa and sigma.  Anyway, my feeling was that too much kappa could make things "weird", at least more so than is usually desired unless departing for a trip. 

I have certainly pondered the 14-phenylpropanoic ester of naltrexone, but thought that opening up the cyclopropylmethyl to isobutyl would likely be an improvement.  Any structure in this category worth considering certainly must meet the criteria of being able to have the 14-side-chain aromatic ring orient itself perpendicular and slightly lower than ring A.  (Still waiting for somebody to be able to explain the pharmacology of 14-cinnamoylcodeinone.)

Your remarks about the enolate issue is very valuable though with consideration to possible useful products.  So, I'm guessing you meant to say "hydrocinnamyl alcohol" aka 3-phenyl-1-propanol with iodine might work?  Or is that conjugated pi system of use in the reaction?  I would guess it would not be necessary considering that the paper example uses t-butyl alcohol. 
 
So, if successful via iodine, one could stay with 14-phenylpropoxynaltrexone, or reduce to convert to the N-isobutyl-7,8-dihydronormorphine version.  Both of these sound like they could be very nice options.

Or, to avoid the enolate issue, one could reduce first via CTH to get the N-isobutyl, 6-hydroxy.  But then subsequent esterification (hydrocinnamoyl bromide) would occur at 3,6, and 14.  The 3-position could easily be removed but the final product would be N-isobutyl-6,14-hydrocinnamoyl-7,8-dihydronormorphine.  Having the 6 position acetylated doesn't sound like a terrible thing considering that it tends to help potency, at least up to benzoyl (unless this group proves too bulky and interferes with the orientation of the other crucial ring- me thinks this unlikely)

I'm going to search your posts for that paper, Assyl!  Hydrocinnamyl alcohol and iodine in AcCN seems like the most obviously easy route, assuming yields are decent.  If the yields are better for the cinnamyl alcohol then one could still CTH reduce that double bond along with converting the the N-CPM to isobutyl and possibly the 6-keto to a hydroxyl.

I should mention that all of the CTH reductions are of varying certainty because we want mild conditions so that the 4,5-ether linkage is not opened up as is known to easily happen with Pd/C 10%.  Even in ice cold conditions, the cyclopropyl ring will definitely open and cinnamyl/cinnamoyl double bonds should be easily reduced.  The question that remains is how easily the 6-keto is reduced in conditions still that help keep the 4,5 ether linkage intact...

RoidRage

  • Dominant Queen
  • ****
  • Posts: 386
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #4 on: February 21, 2013, 12:59:50 AM »
Here is the last one I have access to:

14?-O-Cinnamoylnaltrexone and Related Dihydrocodeinones are Mu Opioid Receptor Partial Agonists with Predominant Antagonist Activity
H. Moynihan †, A. R. Jales †, B. M. Greedy ‡, D. Rennison ‡, J. H. Broadbear §, L. Purington §, J. R. Traynor §, J. H. Woods §, J. W. Lewis ‡ and S. M. Husbands *‡
J. Med. Chem., 2009, 52 (6), pp 1553–1557
DOI: 10.1021/jm8012272

Assyl Fartrate

  • Subordinate Wasp
  • ***
  • Posts: 229
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #5 on: February 21, 2013, 05:09:15 PM »
The unsaturated alcohol is necessary - allylic systems are much more readily attacked by acids. The idea would be to reduce the double bond later. Really, that reaction is just an allylation (cinnamylation, specifically, and probably only) of various alcohols.

Regarding reduction of the 6-keto moiety, Meerwein-Pondorf-Verley should do the trick, and is simple, selective and OTC.

There was a paper requested by Assyl somewhere in the references section about oxycodone, which is, by many, believed to be one of the most euphoric opioids - they found that oxycodone is actually predominantly a kappa agonist. Rather surprising.
Someone Who Is Me

Dope Amine

  • Pupae
  • **
  • Posts: 59
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #6 on: February 27, 2013, 09:14:42 PM »
I'm guessing that you think that cinnamyl alcohol will form the same 1-phenylallyl cation as ?-vinylbenzyl alcohol?  That would be a lovely thing if it were the case considering that ?-vinylbenzyl alcohol is not a commonly accessible chem.

The abstract gives a pretty good outline of the procedure but maybe if there is more detail on the mechanism then we could better answer the question of substituting with cinnamyl alcohol.  If someone would be so kind as to share the full article, it would be appreciated immensely:

J Oleo Sci. 2010;59(10):549-55.
Synthesis of cinnamyl ethers from ?-vinylbenzyl alcohol using iodine as catalyst.

Abstract: Reactions of ?-vinylbenzyl alcohol with other alcohols using iodine as a catalyst were investigated. The corresponding cinnamyl ethers were obtained as products. This suggested that ?-vinylbenzyl alcohol was converted to cinnamyl ethers via 1-phenylallyl cation. Cinnamyl ethyl ether was obtained in 75% yield by the reaction of ?-vinylbenzyl alcohol and ethanol in acetonitrile with iodine under the following conditions: temperature = 50 °C, molar ratio of ?-vinylbenzyl alcohol:ethanol:iodine = 1:3.0:0.2, and time period = 6 h. Generally, the yields of the reactions using primary alcohols were higher than those using secondary and tertiary alcohols. Ether interchange also occurred by the reaction of ?-vinylbenzyl alcohol and iodine, but proceeded smoothly only when an allyl group was used as the other substituent of the starting ether.

Good find though, Assyl!

Polonium

  • Pupae
  • **
  • Posts: 70
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #7 on: February 28, 2013, 01:24:05 PM »
The article is publically available on https://www.jstage.jst.go.jp/article/jos/59/10/59_10_549/_article :). It's attached anyways for future reference.

Synthesis of Cinnamyl Ethers from ?-Vinylbenzyl Alcohol Using Iodine as Catalyst

Yoshio Kasashima, Atsushi Uzawa, Kahoko Hashimoto, Tadasuke Nishida, Keiko Murakami, Takashi Mino, Masami Sakamoto, Tsutomu Fujita
Journal of Oleo Science. Vol. 59 (2010) No. 10 P 549-555
http://dx.doi.org/10.5650/jos.59.549

Assyl Fartrate

  • Subordinate Wasp
  • ***
  • Posts: 229
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #8 on: March 01, 2013, 07:23:36 AM »
It should work with cinnamyl alcohol as well, like you said, both alcohols are pathways to the same cationic allylic species. It should also be possible to get significantly better yields on those ethers if the alcohol is treated as the limiting reagent.

Also, skip the diisopropyl ether. Use DCM or methyl tert-butyl ether instead (for volatile products). Diisopropyl ether forms peroxides very quickly relative to other ethers (and even just unstabilized THF can test hot for peroxides in a matter of days), which is already reason enough for concern - and the peroxides themselves are insoluble, so they crash out, resulting in a solvent bottle that doesn't just offer hazardous workups, but may also blow to pieces at any time, even when undisturbed.
« Last Edit: March 01, 2013, 07:36:01 AM by Assyl Fartrate »
Someone Who Is Me

Dope Amine

  • Pupae
  • **
  • Posts: 59
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #9 on: March 01, 2013, 10:25:15 AM »
Sweet. Thanks for the paper!

Assyl, what are you referring to?  I thought it was just ethanol in acetonitrile. 

I found another paper which relates positively to our 1-phenylallyl cation query- very exciting stuff indeed. ;D

Please refer to the attached article: Tetrahedron Vol. 26, pp 4743-4751. Isoprenylation of Polyphenols in Aqueous Acid Solutions

In this article, they basically use acid to protonate a conjugated alcohol in order to make it a good leaving group, which then creates the "mesomeric cation" and reacts with phenol.  The cation is attacked by the pi electrons ortho to the hydroxy of phenol (lone pair on hydroxy comes down) and a bond is formed between the two in this ortho position.  What is super exciting is that there are two possible conjugated alcohols which form the same cation intermediate!  Please see the following:

"The mesomeric cation ( 6 ), required to carry out the above condensation, could be
generated by protonation of either y,y-dimethylallyl alcohol ( 7 ) or 2-methylbut-3-
ene-2-ol ( 8 ) (see attched image)

As will be shown later, the use of either of these alcohols gives rise to identical products
in the same yield on condensation with a given phenol, indicating that the reaction
involves an SN1 type of mechanism. In contrast, Ollis et al. found that the acid catalysed
reaction of phenols with cinnamyl alcohol and 1-phenylallyl alcohol
involved a mechanism having some SN2 character since the products obtained
excluded the intermediacy of the same 1-phenylallyl cation in both cases."

I have to admit that I'm a bit perplexed with the end of that last sentence.  But in looking at the cation, I see no reason why cinnamyl alcohol should not be able to form it's similar cation equally as well as alpha-vinylbenzyl alcohol.   ;D ;D ;D  I would venture that it wouldn't be a bad idea to add a bit of acid to the reaction conditions outlined in the paper reacting alpha-vinylbenzyl alcohol.  A bit of phosphoric or sulfuric acid wouldn't be bad options but in the paper mentioned above, 5% aqueous citric acid was all that was needed.   

I think a test reaction is called for!  I hope it is possible to at least get some sort of color confirmation that a reaction is occurring.  I hope others share in my optimism about the viability of the formation of this cation.   ;)
« Last Edit: March 01, 2013, 10:31:15 AM by Dope Amine »

Assyl Fartrate

  • Subordinate Wasp
  • ***
  • Posts: 229
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #10 on: March 01, 2013, 04:12:23 PM »
The diisopropyl ether was used in the workup.

It's not a bad idea to investigate this a little further with some test substrates such as t-BuOH, if you have the ability to analyze the results. It's interesting that they didn't run these reactions neat with some of the cheaper alcohols (using the alcohol as solvent) as that would encourage the SN1 mechanism they proposed. Instead, they use MeCN, typically preferred for SN2 chemistry. It could be that this avoids mixtures of isomeric products resulting from a carbocationic intermediate. (Unfortunately, the cinnamyl ether paper has no mention of this matter.)

Quote
In contrast, Ollis et al. found that the acid catalysed
reaction of phenols with cinnamyl alcohol and 1-phenylallyl alcohol
involved a mechanism having some SN2 character since the products obtained
excluded the intermediacy of the same 1-phenylallyl cation in both cases."

We will need that reference for clarification (nice find, by the way).

Perhaps something else is going on here - substitution with iodide at the benzylic position, followed by SN2' substitution (nucleophile attacks at the allylic position, as opposed to the one with the leaving group - rearrangement follows). Typically this is done with stronger nucleophiles, but the driving force may be there, given the good leaving group - iodide - and the stability of the rearranged product (conjugation with the aromatic ring).

Again, worth investigation. Experiments with acids seem worthwhile. Any dicinnamyl ethers should be trivial to remove by A/B extractions or column chromatography (it should fly right off the silica). Try t-BuOH (limiting reagent) in cinnamyl alcohol (solvent), with iodine, and then sulfuric acid as a catalyst - and also, try replicating the same conditions mentioned in the article, except with cinnamyl alcohol instead of alpha-vinylbenzyl alcohol, and one with the I2 swapped out with sulfuric acid. If you can get a small amount of alpha-vinylbenzyl alcohol, it would be worth giving this a shot with (dry) HI or HBr, of varying concentrations, dissolved in the MeCN, to determine if I2 itself is involved. (The question is whether I2 was used merely for convenience, or if it fails with HI.)

Results from these experiments would probably answer all the relevant questions.

And one final potential issue raised by the reference you attached to your post: if there is a carbocation intermediate here with the cinnamyl chemistry, it may be necessary to 3-O-methylate the naltrexone first to avoid the formation of chromans. Given this, it may not be such a bad idea to test working conditions with cinnamyl alcohol and phenol to determine if this is a problem.

If not, the procedure really could be quite easy: cinnamylation of the 14-OH on naltrexone, diimide reduction, and cleavage of the 3-O-phenylpropyl ether by heating with HBr. Yields for similar deprotections are generally quantitative. It may, however, be preferable to leave it as a 3-MeO derivative, to retain high oral activity and reel back the potency a bit - 600x morphine is a little scary. Methylation can be done without anything so nasty as MeI or dimethyl sulfate, as reported by Nicodem:

http://www.sciencemadness.org/talk/viewthread.php?tid=15704

https://www.sciencedirect.com/science/article/pii/S0040403901883027
Biogenetic models for the formation of natural cinnamylphenols and neoflavanoids
S. Mageswaran, W.D. Ollis, R.J. Roberts, I.O. Sutherland
Tetrahedron Letters
Volume 10, Issue 34, 1969, Pages 2897–2900
DOI: 10.1016/S0040-4039(01)88302-7


Requested in the references thread.
« Last Edit: March 01, 2013, 05:59:04 PM by Assyl Fartrate »
Someone Who Is Me

Dope Amine

  • Pupae
  • **
  • Posts: 59
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #11 on: March 01, 2013, 05:49:22 PM »
I missed one of the + charges on the picture of the carbocation.  That could make it confusing so here it is with the picture fixed.

What I don't understand is how Ollis et al. found that "the products obtained excluded the intermediacy of the same 1-phenylallyl cation in both cases".  I guess we need to know what products were found to answer that.  By them saying that their reaction was Sn2, that would mean that it didn't go through the carbocation, considering that Sn2 means that everything happens at the same time and if the mechanism went through the carbocation then that would an intermediate step before nucleophyllic attack (not Sn2).

If anyone could kindly provide the reference cited in Assyl's post above it would be greatly appreciated.

Assyl Fartrate

  • Subordinate Wasp
  • ***
  • Posts: 229
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #12 on: March 01, 2013, 06:00:57 PM »
Assyl would assume that isomeric product mixtures were obtained, but that ones derived from alpha-vinylbenzyl alcohol and cinnamyl alcohol gave different ratios, biased toward one product or the other.
Someone Who Is Me

Assyl Fartrate

  • Subordinate Wasp
  • ***
  • Posts: 229
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #13 on: March 08, 2013, 06:51:30 PM »
Naloxone isn't as readily available, but this is interesting enough to warrant sharing. Quantative N-deallylation of nalaxone using Wilkinson's catalyst (RhCl(PPh3)3) in water with microwave radiation at 200C for 60 minutes.
Someone Who Is Me

Dope Amine

  • Pupae
  • **
  • Posts: 59
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #14 on: March 10, 2013, 07:16:31 AM »
(reaction of 14-OH opioids with cinnamyl-OH and I2 or H+)

What I have found to be very disappointing and not a good sign is that a google search for "cinnamyl (or cinnamic) cation (or ion, alcohol) doesn't yield much of anything relevant or useful.   :'( Maybe somebody else can be more resourceful in this search but what I found seemed all crap.

I am surprised that with both the potential for an allylic cation and aromatic stabilization (at the benzylic position) that cinnamyl alcohol (esp. protonated) isn't known to be an particularly reactive with nucleophiles at the 1 and 3 positions. 

Naltrexone is so very easily obtained in good quantity.  Noloxone is a challenge to get more than 10 mg at a time.  So I feel that the most useful work and development is with naltrexone.  The N-cyclopropylmethyl group is highly reactive to reductive ring opening yielding the N-isobutyl (N-butyl could also be made via bromination).


BTW, cinnamyl bromide can be made from cinnamyl alcohol via HBr and acetic acid: J. von Braun and Z. Kohler, Ber., 51, 79 (1918).
« Last Edit: March 10, 2013, 08:57:59 AM by Dope Amine »

Assyl Fartrate

  • Subordinate Wasp
  • ***
  • Posts: 229
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #15 on: March 10, 2013, 08:24:16 AM »
Cinnamyl halides are quite effective for alkylations, comparable to benzyllic and allylic species. The problem is probably just alkylation of an enolate species, though if it's 600x morphine crappy yields may be meaningless.

Check out this paper. http://127.0.0.1/talk/index.php?action=dlattach;topic=3547.0;attach=6230 (Thanks again to POSEIDON for fulfilling the request.) Maybe that could be used on naltrexone, with a vast excess of cinnamaldehyde. The question there is whether it'd also lead to forming a 6-phenylpropoxy ether of unknown (and probably significantly lower) potency. But really, all you need is 1x morphine for it to be useable, and you're starting from 600x morphine, so there's a lot of leeway. At some point, it's like debating whether you'd prefer getting shot in the face with a 12 gauge or a Navy railgun - it's going to do the job regardless.

Attacking the CPM seems to be the more difficult approach, as the chemistry involved is more complicated and less likely to yield something very potent. Do you have any pharmacological data on N-isobutylenenoroxymorphone or anything similar? Not entirely certain this applies to morphones as well, but with morphine, N-propyl, N-isobutyl, N-allyl, and N-CPM are all antagonists (you're obviously familiar with the latter two).

hxxp://pubs.acs.org/doi/abs/10.1021/ja01116a024 (Data on this if you're curious.)

What is a possibility is finding a way to break open that CPM to yield an allylic amine (probably with H+), akin to that found in naloxone, and tearing it off in a similar manner. The primary issue with that procedure is the temperature would require a vessel that could withstand a lot of pressure (200C and aqueous). There are surely other methods that could be found, as there are many ways of isomerizing allylic species.

Ultimately, some of these ideas will just have to be screened. The iodine/cinnamyl alcohol procedure could be easily tested with t-BuOH. If it works on that, it's likely to work on naltrexone, too. And don't forget about the 14-O-phenylpropanoate ester idea. There's a very real chance that'd yield a potent pure agonist, and the chemistry is already in the bag.

Always bear in mind as well that studies and procedures will not exist without a motive. Often there's workable chemistry out there that was never published simply because pharma never had a use for it. How often do you need to dealkylate a N-CPM amine? It's hard to find any procedures for that, but there's probably an easy way. It's just not interesting to the journals, and the PIs are worried about damaging their reputation publishing clown papers. 

« Last Edit: March 10, 2013, 09:03:27 AM by Assyl Fartrate »
Someone Who Is Me

Assyl Fartrate

  • Subordinate Wasp
  • ***
  • Posts: 229
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #16 on: March 10, 2013, 09:30:46 AM »
Revisiting the topic of the 14-phenylpropanoate ester of naltrexone is worthwhile here. Here is the paper that made Assyl wonder about this in the first place, and having just given it another read, if Assyl were a betting man (which he's not, unless there's an undercover in the ICU fighting for his life, perforated with bullet wounds) he would put down good money on this being a full agonist of reasonable to high potency.

Quote
The unsubstituted cinnamoyl ester (7a) exhibited partial agonist activity of modest potency for all three opioid receptors...

7a is 3-MeO-14-cinnamoylnaltrexone.

Quote
It is of interest to compare the activity of 14-cinnamoylnaltrexone (6a) with that of the phenylpropyl ether (2a) which is structurally similar in having a three-carbon chain linking the side chain aromatic ring to the C14 oxygen atom. The ether (2a) in vivo gave a full response in a battery of thermal antinociceptive assays with potency up to 400 times greater than that of morphine.7 In comparison, the cinnamoyl ester has much more modest in vitro and in vivo MOR agonist activity. It must be assumed that the relative conformational restraint of the alpha,beta-unsaturated cinnamoyl ester prevents an optimum interaction with MOR in the preferred agonist conformation.

This is where the interest of the topic comes into play. There's no real incentive to set out to find N-CPM agonists, so the trail ends here - with a report of a curious accident. Odds are all that needs to be done is hydrogenate that double bond, ideally before the acylation unless the reducer is selective enough to not hit the carbonyl in naltrexone.

Wouldn't it be funny if the next drug epidemic were fueled by a substance used to treat addiction?
Someone Who Is Me

Assyl Fartrate

  • Subordinate Wasp
  • ***
  • Posts: 229
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #17 on: March 11, 2013, 01:56:13 PM »
Alright, owe you an apology Dope Amine - funny story - was about to post that same paper about naltrexone being hit with PtO2 to yield an active compound, to say that it turns out N-isobutyl is an agonist - decided to search first to make sure it hadn't already been posted. Guess where that led Assyl? This thread.

Need to stop posting fucked up. Assyl doesn't sleep much. ;)

Still think though that you should give that ester a shot, as it's likely to be much more potent.

What's strange about that paper is that they use the 3-MeO ether for the PtO2 reduction of N-CPM. Wonder why.
« Last Edit: March 11, 2013, 01:58:27 PM by Assyl Fartrate »
Someone Who Is Me

Dope Amine

  • Pupae
  • **
  • Posts: 59
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #18 on: March 13, 2013, 02:12:56 AM »
I meaning to address that when I finally got the time to post a proper reply.   ;D

Yes, N-isobutylnornaltrexone is definitely an agonist (although it may possess some antagonism as well).  The only pharmacological data I know of for that molecule is in that paper where they convert to N-isobutyl with PtO2.  Even is N-isobutyl does possess some antagonism, it certainly is a better N substituent than CPM!  The closest thing to N-isobutyl-14-phenylpropoxynor-N that I could find pharmacological data is the N-propyl variant which happens to be even stronger than the N-CPM variant.  So I would argue that N-isobutyl should also be stronger than N-CPM (even though increasing strength isn't of great importance here). 

Hydrogenolysis of cyclopropanes can be done easily using typical reductive metal and metal oxide catalysts (at STP).  As a rule, cyclopropanes are more reactive than alkenes, both to hydrogenation and bromination. 

Another potential reaction that can be done is isomerization of the cyclopropane ring to an olefin via silica or alumina catalysts. This could yield something similar to naloxone except that we would have the terminal alkene on a four carbon chain instead of the three carbon chain of naloxone (not sure if that would still make it useful?).

N-isobutyl-14-methoxy/ethoxy-nornaltrexone could also possibly be of value but would require protecting the 3 position with a benzyloxy and reaction with dimethyl or diethyl sulfate.  But that seems like a lot of work.

In the end I agree that 14-phenylpropanoyl is probably the best option but with the N-isobutyl variant as well!  That way, one could protect (or not) the 3-position with a benzyloxy, react with either cinnamoyl bromide or hydrocinnamoyl bromide, and then reduce to open the cyclopropyl group and reduce off the possible 3-O-benzyl.  If enol ethers form at the 6 position then they can be removed under mildly acidic conditions.

That paper you posted looked interesting but I wonder if cinnamaldehyde would work or not?

Assyl Fartrate

  • Subordinate Wasp
  • ***
  • Posts: 229
Re: Potent N,14-disub'd Opioid Agonists from Naltrexone (in 2 steps!), Pharm & Synth
« Reply #19 on: March 13, 2013, 03:42:19 AM »
Actually, in this case by phenylpropanoyl Assyl is referring to the ester. That one seems highly likely to be active, and would be very easy to prepare. One step - phenylpropanoyl chloride, done. Acylation at the 3 position shouldn't be an issue as it'll just hydrolyze anyway.

The 3 position wouldn't need to be protected regardless, as the 3-MeO is cleaved with quantitative yields with HBr. This only works when there's a 14-OR (R=H,alkyl), otherwise it tends to trash the structure.

Your goal may be slightly different than Assyl's, which is to find a way to prepare something active in one single step with the most readily available materials possible. Assyl's ultimate objective is to make the Drug War even more unsustainable than it already is. ;)
Someone Who Is Me