Cannabinoid Pharmacology

Cannabinoid Pharmacology

Under the Radar: Synthetic Cannabinoids and Vaping-Related Lung Injuries

Escaping the Laboratory

How did a few obscure research chemicals end up spawning what is now the largest group of designer drugs?

Synthetic cannabinoids are, in many ways, a byproduct of cannabis prohibition. The proliferation of these compounds has been fostered by marijuana’s Schedule 1 status, which continues to undermine research efforts to understand the endocannabinoid system.

Scientists who want to study how plant cannabinoids impact the brain and body are barred by federal policy from accessing the most obvious research tools – cannabis extracts. Synthetic compounds that target the endocannabinoid system are okay for research purposes, according to Uncle Sam. But scientists, with few exceptions, aren’t allowed to work with oil from the plant.

So pioneers in the field had to create their own cannabinoids for research purposes. These human-designed compounds are called synthetic cannabinoids, even though their molecular structures and effects differ significantly from THC and other phytocannabinoids.1

Groundbreaking chemists – the likes of Alexandros Makriyannis, Raphael Mechoulam, and John W. Huffman – have driven cannabinoid science forward with their unique chemical contributions. They spent much of the 1990s searching for ever more powerful tools in the form of super-potent synthetic cannabinoids. Maybe they’d hit the jackpot by creating a safe and novel (read: patent-able) chemical with therapeutic potential. More likely, drug design would follow an iterative process. New chemicals would be designed with slightly different features, and their potency or lack thereof might shine a light onto hidden facets of the cannabinoid receptors.

For drug designers, there are conflicting drives towards potency and safety. With potency being a key priority, the first synthetic cannabinoids were developed as sophisticated research tools. Human safety was not the foremost goal because these chemicals weren’t meant to make it out of the lab. They were not designed for human consumption.

Several hundred sCBs, some much stronger than THC, have been utilized extensively in preclinical research. Their formulas were published in scientific journals and underground chemists were able to reproduce and market some these compounds as synthetic substitutes for cannabis. One such concoction, a potent cannabinoid (CB1) receptor stimulant called JWH-018,2 inadvertently escaped from the Clemson University lab where it originated and resurfaced later as an illicit street drug.

In 2008, JWH-018 was identified as an ingredient in Spice sold in Germany.3 Then came “CP-47,” another powerful synthetic cannabinoid. The German government banned JWH-018 and CP-47, but within weeks new analogues were sold in their place, sometimes with merely a single atom altered.5 Facilitated initially by this legal loophole, synthetic cannabinoids took on a life of their own. But modifying these chemicals also changes their risks, such that all bets are off when it comes to consuming sCBs.

Fast forward to 2019: Over 1000 sCBs have been identified in various products around the world.6

Cannabis vs. synthetic cannabinoids

Why are synthetic cannabinoids so dangerous when cannabis itself has never caused an overdose death?

Synthetic cannabinoids are worrisome, in part, because there is no historical track record indicating that they are safe. But their ultra-high potency is not necessarily what makes them risky. “Potency” is simply a number that corresponds the dose required to get an average effect from a drug. But it says nothing about what the drug actually does.

If potency was the only problem, then the deadly effects of sCBs would be replicated by a large enough dose of THC. And this has never happened. Although synthetic cannabinoids’ mechanisms of toxicity are poorly understood, it is well established that cannabis – even a high potency THC extract – bears no lethal toxicity.

The physical and psychic effects of drugs are complex, so it helps to borrow from the language of pharmacology, a field dedicated to understanding the human consequences of drug use. In pharmacological research, there are other important measures (besides potency), such as efficacy, off-target effects, and functional selectivity. These four pharmacological concepts help to clarify the differences between phytocannabinoids, endocannabinoids, and synthetic cannabinoids.

  • Potency refers to the dose of a drug needed for an effect. Synthetic cannabinoids are generally much more potent than THC. Someone who knowingly or unknowingly smokes a cannabis joint sprayed with AMB-FUBINACA might think that they can smoke the whole thing, not realizing that AMB-FUBINACA is about 50-80 times more potent than pure THC.
  • Efficacy describes the maximal effect a drug can have. THC has a lower efficacy than most endocannabinoids (the endogenous compounds that THC mimics). This means that at large doses, THC will hit a ceiling. Taking more THC won’t increase CB1 receptor activity (which mediates psychoactivity) beyond a certain point. With synthetic cannabinoids, the ceiling is much higher than with plant or endogenous endocannabinoids, so overdosing can be quite disruptive to health.
  • Off-target activity varies significantly with respect to THC and synthetic cannabinoids. Both confer euphoria (or dysphoria) through the activation of the CB1 cannabinoid receptor. But THC and sCBs each have numerous off-target actions that affect other receptors and various organs in the body, and these off-target activities can produce additional side effects or harms. There is no reason to assume that the off-target effects of THC and sCBs are similar.6 This is not well studied, partly because each individual sCB may have a different profile.
  • Functional selectivity means that two different chemicals, even when activating the same receptor, can produce widely different effects in the cell. This phenomenon is known as “biased agonism.”

A drug’s potency and efficacy at a receptor can be measured with relative ease in cellular assays. But determining off-target effects and the relevance of a drug’s functional selectivity is no easy feat, especially given how quickly old sCBs are eclipsed as newer, unrecognized sCBs emerge each year.

AMB-FUBINACA’s pharmacology

Let’s dive into the pharmacological subtleties of one particular synthetic cannabinoid to understand why, unlike THC, the sCBs can be so toxic.

AMB-FUBINACA” has been implicated in tens, if not hundreds, of deaths in New Zealand over the past few years. As a result, scientists from New Zealand have studied this sCB in meticulous detail. This effort has been spearheaded by Michelle Glass, head of the Chemistry department at the University of Otago.

AMB-FUBINACA was first identified in the product sold as “Train Wreck 2” in Louisiana in 2014, five years after the compound had been patented by Pfizer.7 In 2017, it showed up in New Zealand and quickly became one of the most common sCBs in that country.9

The intracellular effects of AMB-FUBINACA were characterized in a 2019 publication written by post-doctoral researcher David Finlay along with Dr. Glass and other scientists at the University of Otago.8 In an article in Chemical Neuroscience these researchers explored the functional selectivity of AMB-FUBINACA at the CB1 receptor, comparing the unique toxicity of this synthetic cannabinoid to the more benign effects of THC.

Pharmacologists probe these mechanisms in precise technical detail, which we’ll describe at the risk of getting lost in a flurry of acronyms.

AMB-FUBINACA is about 25 times more potent than THC in activating CB1 through one “canonical” pathway (the release of Gα; determined by cAMP formation). The synthetic cannabinoid AMB-FUBINACA outpowered THC to a similar degree in another canonical pathway (ERK phosphorylation, a measure of G-protein release). These two pathways are probably involved in the euphoria of THC and synthetic cannabinoids at low doses, as well as the anxiety that larger doses of either substance can produce.

In addition to its higher potency, a single molecule of this synthetic cannabinoid mobilized twice as much activity inside the cell as a single molecule of THC. In other words, THC is half as efficacious as AMB-FUBINACA in the pERK assay.

Cutting to the chase, the higher potency of AMB-FUBINACA means that users are more apt to accidentally use too much, while its higher efficacy means that overdosing could have more severe consequences.

CB1 desensitization and internalization

The way sCBs activate the CB1 receptor is particularly toxic. Could it be the cell’s overreaction – while attempting to recuperate from these highly powerful sCBs – that’s doing the most harm?

The body always tries to preserve balance. It’s a phenomenon known as homeostasis, and this adaptability is the hallmark of life. When a receptor is activated intensely or for a long period of time, the cell tries to fine-tune things back to normal by desensitizing its response to the receptor.

Some molecules are biased towards desensitization.9 Such compounds trick the cell into thinking that a receptor is overactive, thereby mobilizing homeostatic mechanisms to tone down the receptor’s function.

The New Zealand scientists measured desensitization by quantifying how long CB1 receptors would remain accessible to drugs on the cell surface. Within three minutes of applying AMB-FUBINACA, more than half of the desensitized CB1 receptors were brought inside the cell in a process called internalization. It took nearly 20 minutes for a high dose of THC to do the same.

In line with these results, AMB-FUBINACA recruited the classic markers of desensitization – called β-arrestins – to an incomparably greater degree than THC (even whooping doses of THC barely triggered cells’ β-arrestin pathways). β-Arrestins are one of the aforementioned mediators of homeostasis. They normally bind to a receptor after its activation, promoting balance by reducing the accessibility of overactive receptors.

Beyond this, β-arrestins appear to act as an alternate signaling outlet for receptor activity.10AMB-FUBINACA was considered to be the only compound11 with a high enough potency to signal through β-arrestin in humans. Thus, the New Zealand scientists note, the unique toxicity of this synthetic cannabinoid may be related to its functionally selectivity towards arrestin signaling.

A natural bias

How does cannabis compare to our body’s endocannabinoids, known as anandamide and 2-AG, the native activators (or agonists) of CB1? And how do endocannabinoids compare to synthetic cannabinoids as receptor agonists?

Finlay, along with Xiao Zhu and other collaborators at the University of Otago, have probed this question, employing similar methods they used to study synthetic cannabinoids.12 Initially they focused on CB1 receptor internalization in response to overstimulation.

The results for THC appear consistent between the two experiments. Their data across both studies suggest that THC is weaker than anandamide, which, in turn, is weaker than 2-AG and AMB-FUBINACA in terms of their capacity to cause CB1 receptors to withdraw into the cell.

Among the endocannabinoids, 2-AG is considered a more tonic endocannabinoid,13 whereas anandamide levels fluctuate rapidly in response to stress. The cellular study aligns with this fact: 2-AG levels will set the stage by determining the baseline rate of internalization. Meanwhile, the body can quickly regulate CB1 activity with spikes and troughs in anandamide production without causing too much desensitization.

These results have significant implications for a hot topic in drug development: inhibitors of endocannabinoid breakdown, which indirectly boost CB1 signaling. The natural “bias” of endocannabinoids suggests why new pharmaceuticals targeting enzymes that regulate the endocannabinoids system have not made it out of clinical trials.

Whether due to THC-phobia, the incentive for a molecule that can be patented, or genuine scientific interest, pharma companies are looking for new cannabinoid modulators 14. In particular, they’ve sought out inhibitors of enzymes named FAAH and MAGL, which break down anandamide and 2-AG, respectively. Such compounds are dubiously framed as being more selective therapeutically than a blunt tool like THC.15

But off-target effects ruined the chances of a FAAH inhibitor, dubbed BIA 10-2474, which made it to a phase I clinical trial. Trials of BIA 10-2474 were halted when one participant died and 5 others were hospitalized due to adverse off-target actions.16

Off-target activities aren’t the only hurdle to developing enzymatic inhibitors.17 The natural bias of both anandamide and 2-AG promote receptor internalization more potently and with greater efficacy than THC. So, it’s fair to suspect that drugs boosting endocannabinoid levels are likely to cause tolerance and require dose-escalation. And, indeed, preliminary research has highlighted that therapeutic doses of MAGL inhibitors quickly cause tolerance.1819

Interpreting pharmacology

It’s impossible to wade into this complex pharmacology and emerge with a perfect understanding of what it all means in terms of human health. When interpreting the patchwork of data and technical terminology in this report, it helps to return to the four key measures of potency, efficacy, bias, and alternate targets.

The mechanistic insights provided by various preclinical assays seem to converge on one point – that THC is a gentle modulator of the endocannabinoid system, despite what decades of fear-mongering may have led people to believe. Pharmaceutical companies will be hard pressed to create a new drug that targets the CB1 receptor with fewer risks than THC.

It’s not just the moderate potency of THC, relative to AMB-FUBINACA or other sCBs, that makes the plant cannabinoid so much safer. THC’s lower efficacy provides a built-in guardrail that limits the harms of over-use.

And THC also has off-target actions at the CB2 receptor,20 which may offset some of the potentially harmful effects of CB1 activation. For example, the pro-fibrotic effects of CB1 in the liver can be reduced by CB2 activity.21THC’s activation of CB2 could explain the surprising but welcome association between cannabis use and better insulin sensitivity.2324 By being a non-selective activator of both cannabinoid receptors, the side effects of pure CB1 activation are reduced.

THC also inhibits a pro-carcinogenic enzyme called CYP1A, which normally amplifies the toxicity of chemicals in smoke. This beneficial off-target action may explain the lack of a link between smoking cannabis and lung cancer, despite the presence of carcinogens in cannabis smoke.25

The functional selectivity of THC also seems to work in its favor compared to synthetic cannabinoids, which are much more addictive than THC and cause a severe withdrawal syndrome. It is plausible that THC’s natural bias – its weaker desensitization patterns and weaker mobilization of β-arrestin pathways – also makes THC less prone to habit-formation and harsh withdrawal, and hence safer.

Much more clinical research needs to be done before this kind of speculation can be validated or refuted. On the pharmacological (and preclinical) side of things, we would want to see the relative bias between a compound’s aptitude to confer medical benefits and its preference for desensitization.

Yet science can only guide us so far. It’s clear that synthetic cannabinoids are more acutely toxic than tobacco. But officially sCBs have caused only a handful of lethal outcomes in the United States,26 while cigarettes are responsible for some 480,000 deaths annually.

The extent of these problems is largely defined by drug policy and its failures. The scare around vaping should not be used to misdirect important changes in laws and regulations pertaining to cannabis, tobacco, and e-cigarettes.

Coming soon … Part 2 – Producing the Problem: The Vaping Crises in Context.

Adrian Devitt-Lee, Project CBD’s chief science writer, is pursuing a PhD in Mathematics at the University College of London.

Copyright, Project CBD. May not be reprinted without permission.

Footnotes and References

  1. In the context of this article, Marinol, a lab-manufactured form of THC, is not a synthetic cannabinoid. Crystalline CBD made in a lab is not an SC. “Synthetic” refers to the fact that they are designed by humans – primarily for research, often without human safety in mind – and found nowhere in nature. Sometimes science and media reports call synthetic cannabinoids “synthetic marijuana.” This is an unfortunate misnomer.
  2. JWH stands for John William Huffman.
  3. It is difficult to track exactly the emergence of synthetic cannabinoids. Spice may have been sold as early as 2002. The Research Triangle Institute, in an article written with John Huffman, states that synthetic cannabinoids were first found in Europe in 2006.4
  4. Wiley, J., Marusich, J., Huffman, J. W., Balster, R. L., & Thomas, B. (2011). Hijacking of basic research: The case of synthetic cannabinoids. Research Triangle Park, NC: RTI Press. RTI Press Publication No. OP-0007-1111
  5. Lindigkeit, Rainer, et al. “Spice: A Never Ending Story?” Forensic Science International, vol. 191, no. 1-3, 2009, pp. 58–63., doi:10.1016/j.forsciint.2009.06.008
  6. Schifano, Fabrizio, et al. “The clinical challenges of synthetic cathinones” British Journal of Pharmacology, 2019, doi:10.1111/bcp.14132
  7. One clinical trial on a FAAH inhibitor was halted because of deadly off-target actions.
  8. Buchler IP, Hayes MJ, Hegde SG, Hockerman SL, Jones DE, Kortum SW, et al. Indazole derivatives. Patent WO 2009/106980-A2. New York, NY, 2009.
  9. See… on the website…
  10. Finlay, David B., et al. “Do Toxic Synthetic Cannabinoid Receptor Agonists Have Signature in Vitro Activity Profiles? A Case Study of AMB-FUBINACA.” ACS Chemical Neuroscience, vol. 10, no. 10, 2019, pp. 4350–4360., doi:10.1021/acschemneuro.9b00429.
  11. Devitt-Lee, Adrian. “CB1 Kinetics.” Project CBD, 9 Apr. 2019,
  12. For example, it is thought that respiratory failure during opioid overdose is related to β-arrestin signaling, whereas the painkilling effect is mediated by G-proteins. However, these preclinical notions haven’t yet translated into a non-lethal opiate. No biased opioid agonists have made it through clinical trials as of now.
  13. Compared with THC and two research chemicals, CP55,940 and WIN55,212-2, which have not reportedly been used as recreational drugs.
  14. Zhu, Xiao, et al. “Model‐Free and Kinetic Modelling Approaches for Characterising Non‐Equilibrium Pharmacological Pathway Activity: Internalisation of Cannabinoid CB 1 Receptors.” British Journal of Pharmacology, vol. 176, no. 14, 2019, pp. 2593–2607., doi:10.1111/bph.14684.
  15. Endocannabinoids are typically thought to be produced and released upon demand, when the body is stressed. 2-AG is present in much greater concentrations than anandamide, so the body may be more sensitive to slight changes in anandamide levels.
  16. Lee, Martin A. “Better Than Cannabis?” Project CBD, 13 Dec. 2017,
  17. The argument is that if a drug prevents the breakdown of endocannabinoids, then activity is only amplified where the body is already releasing cannabinoid. There are a number of holes in this argument – one particular issue is that these enzymes are not selective for endocannabinoids. FAAH metabolizes a handful of biologically significant molecules, including PEA and OEA, in addition to anandamide (also called AEA).
  18. “Activity-based protein profiling reveals off-target proteins of the fatty acid amide hydrolase inhibitor BIA 10-2474”. Presented at ICRS 2017. See P8 in
  19. The difference between species has also posed a major challenge to developing FAAH inhibitors. Rats have only one kind of FAAH enzyme, while humans and mice each have two. Beyond this, rodent and human FAAH are genetically different, so chemicals developed to effectively treat a rat’s disease are unlikely to translate to the human counterpart.
  20. Ghosh, Sudeshna, et al. “The Monoacylglycerol Lipase Inhibitor JZL184 Suppresses Inflammatory Pain in the Mouse Carrageenan Model.” Life Sciences, vol. 92, no. 8-9, 2013, pp. 498–505., doi:10.1016/j.lfs.2012.06.020.
  21. Schlosburg, Joel E, et al. “Chronic Monoacylglycerol Lipase Blockade Causes Functional Antagonism of the Endocannabinoid System.” Nature Neuroscience, vol. 13, no. 9, 2010, pp. 1113–1119., doi:10.1038/nn.2616.
  22. Off-target, in some sense, depends on the point of reference. THC has a high affinity for both CB1 and CB2, and its activation of CB2 could be seen as one of its primary effects.
  23. Siegmund, Sören V., and Robert F. Schwabe. “Endocannabinoids and Liver Disease. II. Endocannabinoids in the Pathogenesis and Treatment of Liver Fibrosis.” American Journal of Physiology-Gastrointestinal and Liver Physiology, vol. 294, no. 2, 2008, doi:10.1152/ajpgi.00456.2007.
  24. Lee, Martin A. “Get High and Lose Weight?” Project CBD, 23 Nov. 2015,
  25. Devitt-Lee, Adrian. “Cannabis & Weight Loss.” Project CBD, 3 Apr. 2019,
  26. Melamede, Robert. “Cannabis and Tobacco Smoke Are Not Equally Carcinogenic.” Harm Reduction Journal, vol. 2, no. 1, 2005, p. 21., doi:10.1186/1477-7517-2-21.
  27. The death rate is small enough that it is not tabulated by epidemiological centers like the National Institute on Drug Abuse 27 or the CDC. In the UK, SCs reportedly caused 60 deaths in 2018.28
  28. National Institute on Drug Abuse. “Overdose Death Rates.” NIDA, 29 Jan. 2019,
  29. “Dataset Deaths Related to Drug Poisoning by Selected Substances.” Office for National Statistics, 2018,….