About two years ago a paper came out in JBC that shook the tylenol world to its foundations. This paper demonstrated that tylenol was converted to a compound called AM404. So what is the big deal? And why write about this today? First some background on AM404…
AM404 was first described as an anandamide (an endogenous cannabinoid receptor agonist) uptake inhibitor and was useful in demonstrating the effects of this uptake mechanism in several cell lines and in the CNS. AM404 increases anandamide concentrations when administered in vivo. AM404 was subsequently found to be a cannabinoid receptor agonist and a capsaicin (TRPV1) receptor agonist. The effects on TRPV1 likely contribute to the vasodilatory effects of the compound. In other words, AM404 was a dirty compound, but it did show analgesic effects in a number of models even though it was difficult to understand if this effect was mediated by cannabinoid or TRPV1 receptors.
Tylenol has been around for a long time and is a pretty good analgesic and antipyretic. The interesting thing about Tylenol is that it is a poor antiinflammatory even though it is capable of inhibiting cyclooxegenase (COX) enzymes, albeit rather weakly. There was a big fuss some time ago when the supposed COX3 was cloned and it was shown that COX3 was strongly inhibited by Tylenol and COX3 was mostly localized to the CNS perhaps explaining the MOA of Tylenol. Some of these findings have come under scrutiny, for a variety of reasons (which is beyond the scope of this here post), but the point is that it remained an open question in terms of the MOA of Tylenol.
So what does the JBC study show? They demonstrate that when Tylenol is administered to rats or mice it is converted to an intermediate compound p-aminophenol that is then conjugated with arachidonic acid to yield AM404. This reaction is absent in fatty acid amide hydrolase (FAAH) knockout mice (and inhibited by FAAH inhibitors) indicating that AM404 is a product of p-aminophenol and arachidonic acid conjugation by FAAH. Finally they demonstrate that AM404 is capable of inhibiting COX enzymes and that AM404, but not reaction intermediates, activates TRPV1.
So what does this mean for the MOA of Tylenol? That remains to be seen, but the results of the JBC study indicate that this might be a very complex story. Several actions of AM404 would be capable of causing analgesia. 1) Cannabinoid receptor agonists are well known analgesics and have central, spinal (which is still central but you get the point) and peripheral actions. 2) COX inhibitors are analgesic and are among the most commonly used drugs on the market today. 3) TRPV1 agonists are also analgesic, likely through a desensitizing action, and AM404 is capable of inducing TRPV1 receptor desensitization. One potential problem with hypotheses 1-3 is that Tylenol is not a very good antiinflammatory and that list of MOAs I just proposed should all lead to an antiinflammatory effect either through immune or neurogenic mechanisms. 4) It could be that the expression of FAAH limits the antiinflammatory action while achieving analgesia through other mechanisms mostly involving the CNS.
It so happens that a paper came online today at Pain (the flagship journal of the pain field) that indicates that hypothesis (4) above is at least a major part of the story. Mallett et al., demonstrate that the analgesic effects of tylenol (in several pain tests) are blocked by cannabinoid receptor antagonists and are absent in CB1 receptor knockout mice. Moreover, these analgesic effects are inhibited by blockade of FAAH (the enzyme that converts tylenol to AM404) and are further absent when RVM projections to the spinal cord are lesioned. Hence, the analgesic effects of tylenol are dependent on FAAH and CB1 receptors.
In my post yesterday I told you about how the RVM and PAG are an important site for cannabinoid analgesia. The lesion studies used by Mallett et al., indicate that this analgesic circuit likely plays a key role in mediating the analgesic effects of tylenol and that these effects are similar, if not identical, to the analgesic circuit originally described by Ian Meng, Howard Fields and J Michael Walker. Pretty interesting stuff, if you ask me.
Another point that should be noted here is that these studies indicate another route to engage cannabinoid analgesia without eliciting abuse or psychotropic effects. In my mind at least, these studies suggest that we can utilize some clever tricks to engage certain aspects of the endogenous cannabinoid system to achieve analgesia (in the CNS) while avoiding side effects that no one wants in an analgesic. We still have a lot to learn on the MOA of these compounds and that is a GoodThing for people like me because at least 3 specific aims immediately pop into my mind on this one…
One final note, where is TRPV1 in this story? While TRPV1 is mainly expressed in peripheral pain sensing neurons, several groups have demonstrated that TRPV1 is also expressed in the RVM and PAG and there are some pharmacological effects there (also leading to analgesia). Wow, there you have it, 3 more specific aims!