Since its first mention ∼4000 BCE, Cannabis sativa has evolved through selective cultivation from being a source of durable fiber (hemp) to a plant enriched in bioactive ingredients. Currently, >100 potentially bioactive phytocannabinoids from Cannabis spp. have been cataloged, yet their precise structure-function relationships are mostly unclear (1). Δ⁹-Tetrahydrocannabinol (THC) and cannabidiol (CBD) are primarily studied, particularly because high-grade Cannabis subspecies can produce over 20% yield of either compound. The variety of bioactive constituents in C. sativa, together with their defined ratios, suggests that they have potential application in many illnesses (1). Possibly due to many phytocannabinoids producing similar pharmacological effects through different mechanisms, selecting which to study for a disease remains a formidable challenge.

THC action in humans is dependent on the CB₁ cannabinoid receptor (CB₁R) (2), which is the most abundant G protein–coupled receptor (GPCR) in the brain, as well as its ortholog, CB₂R. When activated, CB₁R in the plasma membrane signals through G_(i/o) proteins to inhibit either Gα_i-mediated SRC–signal transducer and activator of transcription (STAT) or Gβγ-mediated adenylyl cyclase, AKT, or extracellular signal–regulated kinase (ERK) cascades. For excitable cells, such as neurons, activated CB₁R inhibits Ca²⁺ influx through voltage-gated Ca²⁺ channels, thus reducing neurotransmitter release.

Although most studies focus on THC because of its psychostimulant effects, CBD is another abundant (up to 40 to 50% of total phytocannabinoid content) and yet nonpsychotropic Cannabis component. CBD is thought to have anti-inflammatory and tissue-protective effects. This is because CBD action is putatively mediated by more than one receptor, including transient receptor potential cation channel V1 (TRPV1), GPCR 55 (GPR55), and peroxisome proliferator-activated receptor–γ (PPARγ). Additionally, CBD enhances antioxidant cellular defenses by scavenging hydroxyl radicals and can counteract THC action intracellularly, through CB₁Rs on mitochondrial membranes (3). Indeed, CBD opposes the THC-induced disruption of oxidative phosphorylation, at the level of complex I (3), thereby protecting from the deleterious consequences of THCinduced reduction in cellular respiration. Notably, these potential cellular and molecular sites of action are not limited to the brain but apply to, e.g., pancreas, muscle, liver, and gut—suggesting that cannabinoids may have applications in diverse settings.

It is important to recognize that C. sativa is more than just THC and CBD. Many phytocannabinoids that exist in lesser amounts (e.g., cannabivarin, cannabigerol, and THC acid) (4) in plant preparations could individually be biologically powerful and even supersede or modify THC and/or CBD action. Thus, the continued structure-function study of phytocannabinoids is warranted. Accordingly, combinations of phytocannabinoids might deliver relief to disorders with complex etiology (1).

Receptor-mediated actions of phytocannabinoids, particularly THC, center on displacing high-affinity endocannabinoids, innate ligands that bind CB₁R, CB₂R, and alternative receptors including TRPV1, GPR55, and PPARγ. Two such molecules, 2-arachidonoylglycerol (2-AG) and arachidonoylethanolamide (anandamide), with their respective biosynthetic and catabolic enzyme machineries, form the molecular backbone of the endocannabinoid system (5, 6). 2-AG and anandamide are functionally redundant. Yet, endocannabinoid signaling can adopt cell-type–specific configurations to support cell-autonomous (e.g., self-inhibition in neurons), intercellular (e.g., metabolic interplay), and intracellular (e.g., cellular respiration through CB₁R on mitochondria) signaling in the brain and in peripheral tissues. Such functional flexibility is made possible by the different half-lives (minutes versus hours) and tissue distributions of 2-AG and anandamide, together with the diversity of available receptors. These arrangements allow for systemic actions of endocannabinoids and their modulation by THC and CBD (1).

An expanding catalog of endocannabinoidlike molecules are being recognized for their modulation of TRP channels, PPARs, and orphan GPCRs. This “endocannabinoidome” offers a more comprehensive physiological substrate for phytocannabinoid action than the core endocannabinoid system. Moreover, the array of endogenous CB₁R and CB₂R ligands now includes allosteric modulators derived from steroids and hemoglobin fragments, which can antagonize THC intoxication (7) and reduce neuropathic pain (8), respectively.

Functional interrogation of 2-AG action on CB₁Rs in the brain revealed that endocannabinoids modulate synaptic plasticity by limiting Ca²⁺-dependent neurotransmitter release (9). This mechanism relies on CB₁Rs partitioned at presynaptic termini with endocannabinoid ligand production in postsynaptic neurons. Because endocannabinoids are eicosanoid lipids that travel by nonvesicular diffusion, their activity-dependent production and fast degradation ensure phasic availability and short-lived action. Accordingly, neuropsychiatric effects of THC-containing Cannabis preparations (e.g., euphoria, hypomotility, and amnesia) occur through CB₁R hyperactivity at the plasma membrane and intracellularly, which disrupts temporal precision and prolongs endocannabinoid-dependent synaptic inhibition. In the adult brain, the near-complete recovery of endocannabinoid action (and reinstatement of synaptic neurotransmission) may take hours to weeks after THC consumption. Yet, given that only a small fraction of first-time consumption progresses to chronic heavy use, Cannabis is classified as moderately addictive in adults.

A prominent clinical niche being explored for cannabinoid medicines is treatment-resistant epilepsy (10). Phytocannabinoids that trigger CB₁R-mediated signaling at glutamatergic synapses dampen neuronal hyperexcitability in epileptic foci in rodents and humans. Repeated epileptic seizures provoke neurodegeneration in, e.g., the hippocampus. Therefore, although still speculative, rescuing neurons from oxidative damage and mitochondrial dysfunction by, e.g., CBD, could partly underpin the substantial reduction in seizure frequency seen in clinical trials (10).

Phytocannabinoids may also have applications in diseases associated with aging. This is largely because aging tissues contain increased amounts of immune cells, which remove cellular debris or extracellular (proteinaceous) deposits. THC-induced activation of CB₂Rs during aging may offer relief in neurological disorders (Alzheimer’s, Parkinson’s, and Huntington’s diseases) by limiting inflammatory cytokine release from activated microglia, which can be harmful to neurons (see the figure). Alternatively, THC can dampen errant synaptic neurotransmission by CB₁R activation, which might improve cognition. Long-term, low-dose THC treatment in a mouse model of Alzheimer’s disease rescued memory deficits, neuronal morphology, and aberrant gene transcription (11). This duality of correcting synaptic neurotransmission and protecting neurons from degeneration is a typical example of cannabinoid polypharmacology: beneficial effects at more than one cellular target (1). This also exemplifies biphasic responses to cannabinoids, manifesting as beneficial effects at low concentrations and harmful effects at high concentrations. Apart from the multitarget nature of these compounds, these biphasic effects may depend on biased cannabinoid action at different CB₁R populations and, hence, on the baseline (patho) physiological substrate these molecules act on. A further obstacle to the therapeutic use of THC per se is that CB₁Rs, unlike CB₂Rs, often exacerbate inflammation (1).

Cannabinoid therapy is also potentially applicable for the management of neuropathic pain, induced by a lesion or disease of the somatosensory nervous system, because dampening excitatory neurotransmission at the level of spinal neurocircuits can reduce hyperalgesia in rodents (12). Another druggable target emerges in inflammatory pain because of peripheral sensitization in the skin, where accumulation of anandamide facilitates pain processing and proinflammatory signaling through TRPV1 on primary sensory afferents (Aδ and C fibers). TRPV1 activation increases the transcription of nerve growth factor receptors [tropomyosin receptor kinase A (TrkA) and p75], whose activation augments pathological touch sensitivity. Thus, inactivating TRPV1 by sequential activation and desensitization, as some phytocannabinoids do (1), might be medically relevant.

In contrast to the transient and reversible effects of THC on synaptic neurotransmission at mature synapses, the situation is different during pre- and postnatal brain development. The precisely timed activation of CB₁R, CB₂R, and likely GPR55 is critical for cell fate decisions during organ development. In the brain, cell-autonomous signaling contributes to neurite outgrowth (13). Alternatively, intercellular endocannabinoid action determines the size of neural stem cell pools and lineage commitment of daughter cells, their migration, synaptogenesis, and synapse maintenance in vertebrates. Thus, the coincidence of THC exposure with these processes, which occur in the human brain from the second trimester during pregnancy until late adolescence, can imprint adverse and life-long modifications on the structural integrity of the brain. Accordingly, exposure of (pre-) adolescent children to Cannabis is associated with an increased number of hospitalizations for neurological complications (14).

Like CB₂R on neural stem cells, the activation of CB₁R in adipocytes, TRPV1 in pancreas, or GPR55 in skin and salivary gland defines tissue size by modulating the rate of cell proliferation during organ development. Thereafter, CB₁R (or TRPV1) activation in differentiating progeny also determines cell survival. These findings support the exploration of phytocannabinoid-based therapy for disorders in which errant cell-cycle regulation is pathogenic, including cancer. A leading concept is that possible antitumoral effects of THC (and likely CBD) may involve increased ceramide production, triggering autophagy-mediated cancer cell death (15).

Increasing knowledge of endocannabinoid mechanisms and Cannabis constituents has led to the development of synthetic cannabinoids, encompassing THC analogs and structurally unrelated compounds such as high-affinity and selective CB₁R antagonists and inhibitors of a variety of lipases and hydrolases that catalyze endocannabinoid synthesis and inactivation. Some of these synthetic ligands have already entered clinical practice (e.g., rimonabant, nabilone, orlistat) or are in clinical trials (e.g., ABX-1431). Thus, the expanding repertoire of drugs targeting the endocannabinoid system and the endocannabinoidome is of broad therapeutic appeal.

Given the abundance and subcellular partitioning of CB₁R and CB₂R and the endocannabinoid enzymatic machinery, organspecific targeting by synthetic ligands remains a key pharmacological challenge. Dissecting how subcellular pools of CB₁R bring about differential drug action and define functional outcome will be key to improving disease-specific applications. Similarly, delivering single components versus phytocannabinoid mixtures requires decisions on targeting specific signaling pathways rather than harnessing the pleiotropy and synergy of coadministered phytocannabinoids (1). Nonetheless, the expanding knowledge of the endocannabinoid system is lifting phytocannabinoids from fringe utilization to potentially safe and effective medicines in adults.