Beyond CB1 and CB2: Why GPR55 Could Reshape Cannabinoid Science
Photo by Anna Shvets via Pexels.
On April 9, 2026, a paper landed in the Journal of Cannabis Research that most headline writers will struggle to make sound interesting, and that's exactly why it's worth sitting with. Köfalvi and colleagues, publishing in Volume 8 as article 51, reported that a receptor called GPR55, sitting on the surface of astrocytes -- the brain's support cells, not its neurons -- appears to drive glucose metabolism in brain tissue. It's the first dataset to tie this particular receptor to brain energy handling. No clinical trial, no drug, no product. Just a mechanism, observed in mouse and rat tissue, that nobody had pinned down before.
To understand why that's notable, you have to know what cannabis science has spent the last thirty years doing, which is almost exclusively mapping two receptors: CB1, dense in the brain and responsible for most of what gets you high, and CB2, concentrated in immune tissue and the reason CBD-adjacent anti-inflammatory claims exist at all. Everything else -- GPR55, GPR18, GPR119, a scattering of other G-protein-coupled receptors that respond to cannabinoids without fitting the classical CB1/CB2 mold -- has been studied since roughly the mid-2000s and has mostly stayed a footnote, cited in review papers, rarely funded for its own sake.
This piece is going to do something the press cycle around new cannabis papers rarely bothers with: separate what's actually established from what's genuinely new from what's still a guess. Receptor distribution and basic pharmacology for GPR55 -- established, over a decade old. The glycolysis-astrocyte link -- new, specific, preclinical. Any clinical application -- speculative, full stop. The longer view, and the reason this blog is interested at all, is that the peripheral and non-canonical cannabinoid receptors scattered through bone, gut, vasculature, and now brain support cells may be where the next meaningful decade of cannabinoid drug development actually happens, precisely because the CB1/CB2 playbook has largely run its course.
The New Finding: Astrocytes, Glucose, and a 'Paracannabinoid' Receptor
Photo via Pexels.
The paper itself is a fairly careful piece of basic neuroscience. The team, whose manuscript was received August 11, 2025 and accepted February 8, 2026, ran quantitative PCR on mouse and rat brain tissue and found a clean division of labor: Cnr1, the gene coding for CB1, dominates in neurons, while Gpr55 expression is concentrated in astrocytes. That alone is a useful anatomical fact -- it tells you these two receptor systems are, at minimum, sitting in different cellular neighborhoods within the same organ.
What they did next is where it gets interesting. Delta-9-THC, the synthetic GPR55 agonist O-1602, two endogenous candidate ligands (lysophosphatidylinositol, or LPI, and palmitoylethanolamide, or PEA), and even AM251 -- a compound normally classified as a CB1 antagonist that happens to also activate GPR55 -- all stimulated glucose uptake in astrocytes at nanomolar concentrations. That's a meaningfully low dose range, the kind pharmacologists take seriously as evidence of a real receptor-mediated effect rather than some blunt-force metabolic disruption.
The control experiments are what make this more than a correlation. When the researchers blocked GPR55 specifically with the antagonist CID16020046, the glucose-uptake effect disappeared. When they chelated intracellular calcium with BAPTA-AM instead, it also disappeared. That's a one-two punch pointing to a specific, calcium-dependent GPR55 signaling pathway -- not a generic response to throwing cannabinoid-like molecules at brain cells.
The most mechanistically interesting detail is narrower still: LPI selectively boosted glycolysis and lactate release in astrocytes while leaving oxidative metabolism untouched. Neuroscientists have known for years that astrocytes can shuttle lactate to neurons as fuel -- it's called the astrocyte-neuron lactate shuttle hypothesis, and it predates this paper by decades. What's new is the suggestion that a cannabinoid-responsive receptor might sit upstream of that process, nudging astrocytes toward producing more of the fuel they hand off.
Secondary coverage from CED Clinic, in pieces dated April 11 and April 28, 2026, was unusually disciplined about stating the obvious limits: this is preclinical, rodent-tissue work, with no direct line yet to any human treatment. That caveat deserves repeating plainly here too, because it's the one detail most outlets chasing clicks on a new cannabis study tend to quietly drop.
Meet GPR55: The Receptor That Breaks the CB1/CB2 Mold
Photo by cytis via Pixabay.
GPR55 is a strange member of the cannabinoid receptor family in the sense that it barely qualifies as one. Sequence comparisons put its homology at only about 13.5% with CB1 and 14.4% with CB2 -- numbers low enough that, structurally, it doesn't share the classical endocannabinoid binding pocket CB1 and CB2 both have. It got swept into cannabinoid pharmacology mostly because it responds to some of the same molecules, not because it's a close genetic relative. That's a big part of why it sat on the sidelines for so long: researchers chasing cannabinoid targets had two well-behaved, well-characterized receptors to work with, and a structurally odd outlier that didn't fit the existing toolkit.
The bulk of GPR55's foundational characterization happened between roughly 2007 and 2017. The standard reference text, IUPHAR's review titled Cannabinoid receptors and their ligands: beyond CB1 and CB2, is still what gets cited today -- which tells you plainly that the new paper is building on old groundwork, not establishing the receptor's existence or basic behavior from scratch.
What that groundwork shows is a distribution pattern that doesn't look like either CB1 or CB2. GPR55 turns up centrally in the basal ganglia, hippocampus, thalamus, and cerebellum, but also peripherally in vasculature, the GI tract, bone, lung, spleen, liver, kidney, and uterus. CB1 is heavily brain-weighted; CB2 is heavily immune-tissue-weighted. GPR55 is scattered across both compartments in a way that's genuinely its own thing.
It doesn't travel alone, either. GPR18 and GPR119 are the other two non-canonical, cannabinoid-responsive GPCRs that have attracted the most research attention, each with a distinct tissue bias and ligand profile -- though the evidence base for both is thinner than what exists for GPR55, which has had the fifteen-plus-year head start.
And GPR55's relevance didn't start with glucose metabolism. Before this paper, it had already been implicated in intractable epilepsy, diabetes, and cancer biology in various preclinical contexts. The brain energy metabolism angle from the Köfalvi paper is the new contribution; the receptor's broader biological relevance is not.
Beyond the Brain: Bone, Gut, and the Peripheral Receptor Map
Photo via Pexels.
Brain tissue is only one stop on GPR55's map. In bone, the receptor shows up in both human and mouse osteoclasts and osteoblasts -- the cells responsible for breaking down and rebuilding bone tissue, respectively. What's notable is that expression is higher in mature human osteoclasts than in the macrophage progenitor cells they come from, which is the kind of detail that suggests a role specific to bone remodeling rather than just a byproduct of GPR55 being present generally in immune-lineage cells.
That bone connection is worth taking seriously on commercial grounds alone. Osteoporosis and declining bone density are an aging-population problem attached to an enormous existing drug market -- bisphosphonates and denosumab are both multi-billion-dollar drug classes. A GPR55-selective compound that could modulate osteoclast activity wouldn't need to invent a new disease category to find a market; it would be competing in, or potentially complementing, one that already exists and is only growing as populations age.
The gut, vasculature, liver, and kidney expression pattern tells a parallel story on the metabolic and inflammatory side. These are exactly the tissues implicated in conditions -- inflammatory bowel disease, metabolic syndrome, vascular inflammation -- where CB1- and CB2-targeted drugs have repeatedly run into trouble because hitting those receptors broadly tends to come with psychoactive effects (via CB1) or immune suppression side effects (via CB2) that limit how aggressively a drug can be dosed.
Separate from the glycolysis paper entirely, there's already a thread of GPR55 research in neurodegeneration: a chromenopyrazole compound called VCE-006.1 has been tested preclinically in Parkinson's and ALS models for its neuroprotective effects. This matters here mainly as evidence that GPR55 research has legs well beyond metabolism -- it's not a receptor that got discovered and then sat idle until this year's paper.
The through-line across bone, gut, and neurodegeneration research is the same appeal: peripheral selectivity. A compound that acts on GPR55 in bone or gut tissue without meaningfully crossing into CB1-dense brain regions could, in principle, sidestep the psychoactivity and tolerance problems that have sunk a long list of cannabinoid drug candidates before they ever reached pharmacy shelves.
Why Drug Developers Keep Circling Back to Non-Canonical Receptors
Photo via Pexels.
There's a specific historical wound driving pharma's interest in peripherally-restricted cannabinoid targets, and it's worth naming directly: rimonabant. Marketed in Europe as Acomplia, it was a CB1 antagonist developed for weight loss that worked reasonably well on that front and was pulled from the European market after being linked to serious psychiatric side effects, including suicidal ideation. Regulators in the U.S. never approved it at all, citing the same safety signal. The mechanism was almost too on-the-nose: CB1 is so densely expressed throughout the brain that blocking it systemically to affect appetite inevitably meant blocking it everywhere else too.
Rimonabant's failure didn't just kill one drug -- it reoriented an entire design philosophy in cannabinoid pharmacology. The logic of chasing peripherally-restricted targets, ones that sit outside the brain or at least outside CB1-dense regions, predates the current GPR55 interest by well over a decade. GPR55 fits neatly into a search pattern that pharma was already running.
There's also a useful precedent for how orphan and understudied GPCRs eventually mature into drug targets: once selective tool compounds -- agonists and antagonists specific enough to isolate a single receptor's effects from its relatives -- become available, research accelerates and clinical candidates start to follow. CID16020046 and O-1602, the antagonist and agonist used in the new astrocyte paper, are exactly that kind of early-stage tool compound. They're research instruments, not drug candidates, but they're the instruments that make drug candidates possible down the line.
The counter-case deserves equal weight. GPR55 pharmacology is still genuinely messy. AM251 acting on both CB1 and GPR55 is a problem, not a footnote -- it muddies interpretation of any experiment using it as a clean CB1 antagonist. Whether LPI is really the endogenous ligand GPR55 evolved to respond to is still actively debated among pharmacologists. And as of today, no GPR55-selective drug has entered human clinical trials.
Here's a reasoned extrapolation, explicitly flagged as such rather than fact: CB2-selective agents took roughly fifteen to twenty years to move from initial receptor characterization to clinical-stage compounds. If GPR55 tool compounds mature on a similar timeline from its own characterization window in the mid-to-late 2000s, a GPR55-targeted therapeutic entering human trials some years out is plausible. It is not guaranteed, and funding, not biology, is likely to be the limiting factor.
What This Could Mean for Cannabis Products and Policy, Long Term
Photo by Mikal Mc Leod via Pexels.
If peripheral receptor science like this keeps maturing, it could eventually justify a different way of thinking about cannabis formulation entirely -- products built around specific receptor targets, say a minor cannabinoid or terpenoid combination selected for GPR55 activity, rather than the blanket THC-to-CBD ratios that dominate dispensary shelves today. That's a real long-range possibility worth naming. It's also years, plausibly a decade or more, from showing up on an actual product label, and readers should treat any company claiming otherwise right now with real skepticism.
There is a useful precedent, though, for how a non-intoxicating cannabinoid-adjacent compound can reach consumers before its receptor pharmacology is fully settled: PEA, palmitoylethanolamide, one of the ligands used in the new astrocyte study, is already sold as a supplement in parts of Europe for pain and inflammation, marketed under existing novel-food and nutraceutical regulatory frameworks rather than as a prescription drug. That's a real, current example of the regulatory path a compound like this can take -- supplement shelf before pharmaceutical approval, if it gets there at all.
On the pharmaceutical side, regulatory agencies have a track record of eventually approving drugs targeting newly characterized GPCRs once selectivity and safety data accumulate. The CGRP receptor, now the target of several approved migraine drugs, took roughly twenty-five years from receptor cloning to the first approvals. A comparable multi-decade timeline for a GPR55-targeted drug is a reasonable long-range guess based on that precedent -- not a prediction anyone should attach a specific year to.
Several things could stall this entirely. Funding for orphan GPCR research is thinner than funding chasing CB1 and CB2, and most cannabis research dollars right now are still going toward THC and CBD formulation work rather than novel receptor biology. Preclinical findings in rodent tissue also fail to replicate in human tissue often enough that it's a standard caveat in pharmacology, not a cynical aside -- astrocyte biology in a mouse brain is not guaranteed to map cleanly onto a human one.
The more grounded near-term opportunity isn't a consumer product at all. It's research tooling -- companies building GPR55 assay kits, receptor-binding screening platforms, or antibody panels that pharma and academic labs can buy to study this receptor family faster and more cheaply than building those tools in-house. That's a smaller, less headline-friendly business opportunity than a wellness beverage, but it's the one with a realistic path to revenue over the next several years rather than the next several decades.
Strip away the framing and the April 2026 paper is a small, specific finding: astrocytic GPR55 appears to drive glycolysis in brain tissue. That's a real contribution to neuroscience. It is not a breakthrough that changes anything about the cannabis products on shelves today, the laws governing them, or what a patient should expect from any formulation currently for sale. Readers should be genuinely skeptical of any outlet that stretches a rodent-tissue mechanism paper into a treatment claim -- that gap between preclinical signal and clinical reality is exactly where cannabis science reporting tends to go wrong.
The real story here is structural, not clinical. Cannabis science spent thirty years building an enormous, genuinely useful body of knowledge around two receptors, CB1 and CB2, while a dozen others -- GPR55, GPR18, GPR119, and more -- sat in the periphery, understudied not because they were unimportant but because the selective tools needed to study them properly simply didn't exist yet. Those tools -- selective agonists, selective antagonists, knockout animal models -- are only now catching up to where CB1/CB2 research was a decade or two ago.
Whether GPR55 eventually becomes the next CB2-style drug target, the kind that takes fifteen or twenty years to go from characterized receptor to clinical candidate, or whether it stays a research curiosity cited in review papers indefinitely, depends far less on this one paper than on whether funding actually follows the expanding peripheral-receptor map over the next decade. History suggests it can happen -- CB2 did, CGRP did. History also suggests it's never guaranteed, and plenty of promising receptor targets never clear that bar. This is a thread worth watching, not a story that's already written.
