Targeting the Endocannabinoid System to Treat Human Disease

People have used cannabis plants as pain relievers at least since the dawn of recorded history, some five thousand years or so ago. But the endogenous receptors mediating cannabis’ multiple effects, CB1 and CB2, and their endogenous ligands weren’t isolated and identified until the 1980s. Today, we know that the endocannabinoid system plays a vital role in maintaining homeostasis throughout the body; it is involved in physiological processes ranging from neurotransmission, mood, and energy balance to immune responses and wound healing. On January 27, 2021, the New York Academy of Sciences hosted a webinar on Targeting the Endocannabinoid System for Treatment of Human Diseases. Researchers described their work leveraging the endocannabinoid system to treat obesity and metabolic disorders and different types of inflammation and pain. Their ultimate goal is to promote the beneficial actions of signaling through the endocannabinoid system while eliminating the psychoactive side effects.

Symposium Highlights

• Endocannabinoid receptor agonists, antagonists, and inhibitors of the enzymes that metabolize endocannabinoids are being developed to harness the endocannabinoid system’s vast effects for pharmacotherapeutic purposes. >
• “Cannabis” is  not one thing; it contains over a hundred cannabinoids. But many health care providers, bombarded with information from the cannabis industry, lack an adequate understanding of phytocannabinoids and the various ways they can stimulate the endocannabinoid system.>
• An inhibitor of MAG Lipase, the enzyme that metabolizes the endocannabinoid 2-AG, can alleviate chemotherapy induced peripheral neuropathy and is currently in Phase 2 clinical trials. >
• CBD induces apoptosis in melanoma cells by inducing the Endoplasmic Reticulum stress response. >
• CB1 is found primarily in the brain, while CB2 is expressed primarily on immune cells. Thus molecules that act through CB2, or those that act through CB1 but don’t cross the blood-brain barrier, might provide therapeutic benefits without eliciting psychotropic side effects. >
• A positive allosteric modulator of CB1, a CB2 agonist, and FAAH inhibitors all suppress pathological pain. Unlike THC and morphine, they do so without inducing tolerance, dependence, or reward. >
• A CB2 agonist reduced proinflammatory mediators of gout and alleviated swelling in a rat model of the disease. >

Biology and Pharmacology of the Endocannabinoid System 

An exogenous ligand of the endocannabinoid receptors was identified before the receptors themselves were; Δ⁹-Tetrahydrocannabinol, popularly known as THC, was isolated from confiscated cannabis in Israel in 1965. It took twenty years for a synthetic version of THC to be granted FDA approval, first to treat nausea in cancer patients getting chemotherapy and then to treat fatal wasting syndrome in AIDS patients. At that point it and its analogs were orphan drugs. Then the two endocannabinoid receptors, CB1 and CB2, were identified; and their endogenous ligands, the lipid transmitters N-arachidonoylethanolamine (AEA, or anandamide) and 2-arachidonoylglycerol (2-AG), followed. Allyn Howlett gave an overview of what we know about the whole system today.

CB1 and CB2 are 7-transmembrane G protein coupled receptors. CB1 receptors are expressed all over the brain: in the prefrontal cortex, the seat of judgment and executive function; in the amygdala, which mediates the reward system; in the hippocampus, where memories are made; and even in the cerebellum. But they are not limited to the brain. Cells in other organs, including the liver, skeletal muscle, pancreas, GI tract, and adipose tissue, upregulate their expression of CB1 in response to a high-fat diet. CB2 receptors are found primarily in immune cells.

CB1 receptor stimulation returns a hyperactive cell to its normal state. If excessive amounts of a neurotransmitter are released from a neuron--either an excitatory neurotransmitter like glutamate or an inhibitory neurotransmitter like GABA—the postsynaptic cell makes endocannabinoids and sends them back across the synaptic cleft. When these bind and activate CB1, the CB1 receptors’ G proteins dissociate into their component subunits. The 𝛂 subunit reduces cAMP production through adenylyl cyclase, which reduces protein kinase A activity and activates certain types of potassium channels. The ꞵ/γ subunit inhibits some types of calcium channels. The ultimate result is membrane hyperpolarization and a reduction of neurotransmitter release. Hydrolases at both the presynaptic and postsynaptic sites are on hand to cleave the endocannabinoids to ensure that their effects are fleeting.

Signaling through CB1 receptors.

The most straightforward way to clinically interfere with the endocannabinoid system is through selective agonists or antagonists of CB1 or CB2. But there are also drugs in the pipeline that inhibit MAG Lipase and FAAH, the enzymes that metabolize 2-AG and anandamide, respectively. Both positive and negative allosteric modulators of the receptors have also been developed. And of course there are plenty of proteins associated with CB1 and CB2 that may have regulatory functions. All of these may be appropriate targets to harness the beneficial actions of the endocannabinoid system—treatment of pain, anxiety, depression, and metabolic disorders—without eliciting the psychoactive side effects of THC.

Regulation of Appetitive and Addictive Behavior by the Peripheral Endocannabinoid/CB1 receptor System

George Kunos and his lab hypothesized that antagonizing CB1 receptors in the periphery, but not in the brain, might be a way to treat metabolic syndrome without eliciting neuropsychiatric effects. But first, they had to show that blocking these peripheral receptors could affect appetite and addiction, behaviors that are mediated by the CNS.

To do this, his lab employed a peripherally restricted CB1 inverse agonist, JD5037. Surprisingly, they found that it reduced food intake, body weight, and insulin resistance in mice fed a high-fat diet; moreover, it did so to the same extent as its parent compound, a CB1 inverse agonist that is brain penetrant. Both compounds also dropped the highly elevated leptin levels in these obese mice back down to normal. A series of experiments amply demonstrated that that the appetite suppressive effects of peripheral CB1 blockade are in fact mediated through downregulating endogenous leptin, which in turn achieves its effects by activating anorectic POMC/CART neurons. Leptin is made peripherally in adipose cells but acts centrally. It could thus be the very link the Kunos lab set out to find connecting peripheral CB1 receptors with CNS pathways controlling appetite.

Similar results emerged from studies of a mouse model for alcoholism. The Kunos lab used two methods to model alcoholic behavior in male C57Bl6/J mice: the two-bottle free-choice paradigm and the drinking-in-the-dark paradigm. In the former, mice were free to choose to drink from a bottle containing water or one containing 15% ethanol; they preferred the ethanol. In the latter, mice were given access to 20% ethanol while they were in the dark. This is meant to be a model for binge drinking. In both cases, treatment with JD5037 reduced drinking just as much as the globally acting CB1 inverse agonist. The treatment did not alter the mice’ preference for sucrose, saccharine, or quinine, so it was specific to alcohol and not to taste or calorie count.

Interplay between the endocannabinoid system and the brainstem neurocircuit comprising vagovagal reflexes

As with their studies with obese mice, the Kunos lab wanted to find the link between peripheral CB1 receptors and the addictive behavior mediated by the CNS. But this time, it wasn’t humoral, like leptin; it was neural. They found it in ghrelin signaling through receptors located on vagal afferent terminals. Ghrelin is made in the stomach mucosa, which also expresses CB1 receptors and high levels of endocannabinoids. Alcohol induces endocannabinoid expression is specifically induced by alcohol in these stomach cells but not in brain or plasma cells. Ghrelin is known to promote drinking alcohol. CB1 blockade reduces alcohol consumption by preventing the acylation of ghrelin into its active form. It does this by promoting the degradation of ghrelin’s substrate, octanoic acid.

Paradoxically, there is evidence that THC—the original CB1 agonist, the one that makes people hungry—can reduce obesity. Kunos suggested that perhaps heavy THC use desensitizes CB1 receptors so that endocannabinoids can no longer signal through them, making THC act almost like these inverse agonists rather than a standard one.

When Phytocannabinoids Meet the Endocannabinoid System

The mission of Greenwich Biosciences, a subsidiary of GW Pharmaceuticals, is to develop plant-based cannabinoid medicines for neurologic and psychiatric conditions. As their Senior Director of Medical Affairs, Kathryn Nichol oversees the translational medicine and drug development.

Nichol explained that phytocannabinoids differ from endocannabinoids in several clinically significant ways. Whereas only five endocannabinoids have been isolated to date, about 120 different chemicals have been found in Cannabis sativa. Their chemical structures are different; endocannabinoids are polyunsaturated fatty acids with a polar head group, while phytocannabinoids are terpenophenolic compounds derived from geranyl pyrophosphate. Endocannabinoids act and are degraded locally; phytocannabinoids are metabolized in the liver, and their site of action depends on how they are ingested. Both may reduce neuronal hyperexcitability. Yet research findings on endocannabinoids are often presented out of context, especially online, and presumed to apply to cannabis as well. And of course “cannabis” is not one monolithic entity; it contains hundreds of cannabinoid molecules in varying amounts, all of which exist in active and inactive forms and have active and inactive metabolites.

Phytocannabinoids have varying binding affinities and directions of action for CB1 and CB2. THC is a partial agonist of both; CBD is an inverse agonist, albeit a weak one, of both; and others have agonistic and antagonistic effects of varying levels. They can also bind to and exert their varying effects on other receptors, like TRP channels. This diversity makes drug discovery extraordinarily challenging.

Varying effects of various phytocannabinoids.

Nichol provided a few real-world examples of this challenge. In animal models of epilepsy, glaucoma, neuropathic pain, and spasticity, CBD and THC have conflicting effects. But regardless of their different and sometimes opposing activities, health care practitioners still want CBD because “it’s the non-psychoactive one,” and because their knowledge of cannabis and its myriad effects is generally outdated and misguided.

Findings on endocannabinoids cannot be extrapolated to phytocannabinoids, Nichol stressed, and endocannabinoid researchers “can’t pretend the twenty-billion-dollar commercial cannabis industry does not exist.” She said they “must be aware that marketers can and will use your basic research. But for medical utility and advancing drug development, risk/benefit analyses, preclinical molecular and animal studies, and clinical studies must be done before products are given to patients.” Nichol noted that perhaps rushing opioids to market without those safeguards contributed to the current opioid crisis, and we certainly don’t want to repeat that mistake with phytocannabinoids. Greenwich has thus proactively sponsored independent medical education to keep physicians updated on research into the endocannabinoid system, cannabis, and FDA regulations regarding cannabis use.

Cannabidiol (CBD) Inhibits Cancer Survival Through Upregulating the Endoplasmic Reticulum (ER) Stress Response

CBD has been pressed into service to treat numerous cancers: glioblastomas, leukemias and lymphomas, and breast, prostate, colon, and gastric cancers. It induces the apoptosis of cancer cell lines in a time and dose dependent manner. In melanoma cells, the most sensitive cell lines Andrea Watters and her labmates assayed, CBD upregulated the expression of proteins involved in the ER stress response, especially CHOP, IRE1𝛂, and XBP1. Overexpression of CHOP in particular is known to lead to apoptosis. They found that melanoma cells were almost three times more sensitive to CBD’s apoptosis inducing effects than normal skin fibroblasts, suggesting that CBD could be used clinically with minimal side effects. Watters and her colleagues think that in melanoma, CBD is acting as an agonist on TRPV1 receptors in the plasma and ER membranes to increase intracellular calcium, which leads to ER stress and induces CHOP. CHOP overexpression causes the generation of reactive oxygen species and apoptosis. They hope to confirm this proposed mechanism in mouse models of melanoma.

Panel Discussion One

The first panel discussion focused on and reiterated how much is still unknown about cannabis and the endocannabinoid system. Nichol was asked how CBD halts seizures and if its effects are limited to certain seizure disorders; she said “the truest answer I can give you is that we don’t know,” since all of the studies done thus far have been in culture or in mouse models. When asked about CB1 long term desensitization, Mackie noted that “heavy users of THC still report the munchies, so that aspect of CB1 signaling hasn’t desensitized.” There was some speculation that CB1 antagonists could be used to help respiratory distress, like that seen in COVID-19, since vapers who inhale synthetic CB1 agonists wound up in the hospital with respiratory distress. One question regarded THC’s negative effects on psychotic disorders, bipolar disorder, and depression, which Howlett labeled “a very complex issue.”

The panelists concurred that cannabis does not induce psychosis in most people, but has been shown to exacerbate anxiety. And its effects, like individuals, are wildly heterogeneous; different cannabinoids at different doses can have very different effects on different people, which makes extrapolating clinical utility from any data extremely complicated. Moreover, we don't know exactly which cannabinoids people are getting when they use cannabis to self-medicate, nor do we know their underlying physiology. Nichol summed up the state of the field by noting that “the data in the literature is so murky.” The panel concluded by referencing anecdotal data about cannabis improving patients’ quality of life--getting them back to work, allowing them to sleep better--even though it may not be as effective at pain relief as opioids, and collectively bemoaned that clinical trials of pain relievers don’t use these secondary but perhaps more functional endpoints.          

New Insights into the Resolution of Acute Inflammation in Humans Provided by Lenabasum, a Synthetic Analogue of delta -8- tetrahydrocannabinol (THC)-11-oic Acid

Derek Gilroy is interested in the resolution of acute inflammation. Within minutes of an infection or injury, neutrophils migrate to the affected site, controlled by cytokines and chemokines. Once the infection has been cleared, the inflammation begins to resolve: the neutrophils die of apoptosis and are eaten by macrophages, which then release anti-inflammatory molecules. His lab would like to harness these internal pro-resolution pathways to switch off ongoing inflammation.

Lenabasum, or ajulemic acid, is a synthetic analog of the terminal metabolite of THC, so it should not have psychotropic effects. It is a selective CB2 agonist that has been shown to be anti-inflammatory in various animal models. There is even evidence that it induces pro-resolution pathways. Anti-inflammatories dampen the immune response, while a pro-resolution drug would allow the same level of acute inflammation but then clear it more quickly.

Gilroy’s lab set out to determine if Lenabasum is anti-inflammatory or pro-resolution in humans. They injected UV-killed E. coli into the arms of volunteers to induce local inflammation in their skin. The skin gets red within a day, but the redness resolves two days later. When the volunteers got Lenabasum for three days before their bacterial injection, their neutrophil infiltration levels were severely diminished, as were levels of the proinflammatory cytokine IL-8 and lipid mediators of inflammation like LTB₄ and PGE₂. Lowering PGE₂ levels is known to help neutrophils phagocytose bacteria. So Lenabasum acts as an anti-inflammatory.

UV-killed E. coli injected into the arms of volunteers to induce local inflammation in the skin.

“Inflammation will never go away or resolve until the irritant that caused it has been removed,” said Gilroy. Lenabasum also reduced endotoxin levels, so it helped the immune system clear the bacteria while diminishing inflammation.

Even though it dampened molecular markers of inflammation, the higher dose of Lenabasum (20mg) seemed to draw more blood to the site and make the skin look redder. The redness was not associated with a reinfiltration of neutrophils. Gilroy thought it could be due to CO, a byproduct of heme oxygenase activity, which is elevated as inflammation is resolved. His lab found that resolution phase macrophages have increased expression of both 2-AG and its receptor CB2, revealing that the endocannabinoid system plays a role in our endogenous resolution pathway. He speculates that Lenabasum might trigger resolution by acting through CB2 to induce heme oxygenase activity. People with rheumatoid arthritis and other chronic and inflammatory diseases don’t exhibit this enhanced redness; perhaps Lenabasum can help stimulate their resolution pathways if their endogenous mechanisms to do so have flagged.

Medical Marijuana and the Endocannabinoid System: Therapeutic Potential for Treating Neuropathic Pain

The opioid epidemic has renewed interest in using cannabis for pain relief. But Andrea Hohmann noted  that even though cannabis has been used for millennia to combat pain, the current state of the field is chaotic. Most of the studies done are small, short term, and of poor quality, without adequate blinding or appropriate placebos; disparate kinds of chronic pain and different cannabinoid-based medicines are often mixed; safety assessments are scarce; and as in other fields, only positive results tend to get published.

Her lab looks at chemotherapy-induced peripheral neuropathy using paclitaxel treated rats as their model system. The rats have pain in their extremities that manifests as hypersensitivity to cold and increased brain functional connectivity, especially in regions involved in motivational and emotional responses to chronic pain. Exposing them to vapor from cannabis that was 10.3% THC and 0.05% CBD brought their resting state brain connectivity back down to normal levels and eliminated their cold hypersensitivity. But it had the undesired side effect of reduced locomotor activity.

Hohmann aimed to retain THC’s nociceptive effects while eliminating side effects like getting high. To start, her lab tried a positive allosteric modulator of CB1 called GAT211 that would increase the affinity and efficacy of endogenous cannabinoids or CB1 agonists. In a mouse model, GAT211 worked through CB1 to eliminate paclitaxel-induced pain. Unlike mice treated with THC, mice treated with GAT211 did not develop tolerance to it over time or experience withdrawal symptoms like paw tremors, head shakes, or scratching. It also did not seem to induce reward pathways. In contrast, mice given morphine or a MAG lipase inhibitor developed tolerance to these substances; their pain returned over time, even with treatment.

Next, Hohmann’s group investigated inhibitors of Fatty Acid Amide Hydrolase (FAAH), the enzyme that metabolizes anandamide. They found that in mice, each inhibitor suppressed both the development of paclitaxel-induced neuropathic pain and also alleviated established pain. Neither inhibitor produced tolerance or physical dependence. Promisingly, both FAAH inhibitors synergized with morphine’s ability to relieve pain (but not with morphine’s ability to cause constipation or seek reward).

Cartoon depiction of CB2’s therapeutic potential.

Lastly, they looked at CB2 activation as a means of treating pain and circumventing the side effects elicited by signaling through CB1. Like the previous agents they examined, their CB2 agonist reversed established paclitaxel neuropathy without inducing tolerance or symptoms of physical withdrawal. It did so even in CB1 knockouts, suggesting that therapeutic exploitation of CB2 receptors can be accomplished without any input from CB1 receptors. The treatment decreased mRNA levels of the proinflammatory cytokine TNF-α and the chemokine MCP1 that were elevated by paclitaxel treatment in the lumbar spinal cord of mice. This may begin to explain the mechanism by which it relieves pain. Another CB2 agonist they tried also had antinociceptive effects and was able to block the undesirable side effects of morphine like tolerance, dependence, reward seeking, and respiratory depression.

Two clinical trials of CB2 agonists for pain have already failed because of lack of efficacy, one for third molar extraction and the other for osteoarthritis pain. So although the pathway has therapeutic potential, Hohmann stressed that the proper drug must be matched to the proper indication to achieve a clinically meaningful effect.

Targeting CB1 and CB2 for Human Therapeutics

“Without a doubt, there are therapeutic uses for CB1 agonists and cannabidiol,” said Barbara White, chief medical officer and head of research at Corbus Pharmaceuticals. CB1 inverse agonists have been investigated as treatments for metabolic and fibrotic diseases and various neurologic problems. One, rimonabant, was approved for weight loss in Europe but had to be pulled because of serious side effects: anxiety and suicide. So Corbus is looking for inverse CB1 agonists that don’t cross the blood-brain barrier so readily and might therefore be safer. They currently have five candidates in development that, because of size, or polarity, or lipophilicity, remain in the plasma but don’t get into the brain as much. However, for some of them, the little that gets into the brain remains there for days; these may not be optimal for clinical use. But those that were tested in a mouse model of diet induced obesity precipitated weight loss and improved glucose tolerance and, they are rapidly cleared from the brain. Dr. White is hopeful that they will be safe and effective in humans.

Corbus’ CB1 inverse agonists have limited penetration into the brain.

Corbus has primarily focused on CB2 agonists, which have been reported to be beneficial in cancers, fibrotic diseases, autoimmune diseases, and CNS diseases. Specifically, they are examining the synthetic selective CB2 agonist Lenabasum as a treatment for chronic inflammation and fibrotic diseases because it reduces the production of proinflammatory cytokines as well as pro-fibrotic growth factors and collagen. It also reduces myofibroblast transformation. CB2 receptors are expressed on activated immune cells and fibroblasts. While Lenabasum does have some CB1 agonist activity, it has a limited ability to cross the blood-brain barrier. White and her team tested its use in systemic sclerosis, or scleroderma. This  rare, debilitating, and life-threatening autoimmune disease is characterized by inflammation and fibrosis in the skin, joints, and lungs.

CB2 knockout mice develop scleroderma-like symptoms when their innate immune system is challenged, linking the endocannabinoid system to the disorder. In a Phase 2 study of people with sclerosis, Lenabasum improved inflammation and fibrosis in skin biopsies after twelve weeks. In the RESOLVE-1 Phase 3 study, Lenabasum stabilized lung function in sclerotic patients, while those given placebos showed declining lung function over time. Similar results were found in a Phase 2 study of people with cystic fibrosis: Lenabasum lowered their rate of pulmonary exacerbations. Both studies showed a favorable safety profile, but the primary efficacy endpoint was not met in either. In the RESOLVE-1 study, this could be because many of the participants were already  taking immunosuppressants.

In preliminary studies, Corbus’ novel CB2 agonists also inhibited the growth of breast cancer cells in culture to varying degrees, and at least one of them inhibited Her2 phosphorylation in a breast cancer cell line and Her2+ tumor growth in a xenograft model. The agonists also inhibit the growth of non small-cell lung cancer and a glioblastoma cell line at nontoxic levels. Dr. White plans to use animal models to investigate the mechanisms responsible for this growth inhibition.

CRB-317, a Selective Cannabinoid Receptor Type 2 (CB2) Agonist, Inhibits NLRP3 Inflammasome Activation, Cytokine Production, and Has Activity in a Model of Gouty Arthritis

Ping Zhang tested another of Corbus Pharmaceuticals’ novel synthetic selective CB2 agonists against gouty arthritis. Gouty arthritis is one of the most common forms of inflammatory arthritis, affecting about 4% of adults in the US. She found that the CB2 agonist inhibited activation of the inflammasomes in human macrophages known to be activated in acute gout. This inhibition manifested as a reduction in caspase-1, IL-1β, and IL-18. It also inhibited the production of a number of other proinflammatory cytokines—IL-6, IL-23, MCP-1, and TNF𝛂—in PBMCs, and attenuated knee inflammation in a rat model of arthritis in a dose dependent manner.

Panel Discussion Two

This panel discussion started with Kunos asking Gilroy which disease his skin inflammation model best recapitulates. Gilroy replied, “this is something we philosophize about a lot.” He went on to explain that “the model is very useful for understanding basic immune responses to infection and injury,” as well as immune senescence in aging and the mechanistic basis for multimorbidities. But White warned against over interpreting any data because we so desperately want accurate models. Next, Kunos asked Hohmann if she has looked at any biased CB2 agonists—like one that worked only through G proteins—that might synergize with opiates to prevent the development of long term tolerance and respiratory depression. She said that the agonist she used was biased in that exact way-it is G protein based and does not recruit arrestins at all. But it did fail in a clinical trial for osteoarthritis pain due to lack of efficacy. She is currently trying to unravel which pathways are necessary and which are dispensable for therapeutic effects and for side effects. Howlett wondered if the agonist might be working in the brain stem, an area of the brain that incontrovertibly expresses CB2 receptors. Lastly, Kunos asked White if she’d examined the expression levels of MDR ABC transporters, given that brain tumors often express them at high levels. Do they shunt her CB2 agonist out of the tumor? White answered that they’ve tested some of their compounds in vitro, but not yet in tumors. They vary.