In 1854, Rudolf Ludwig Virchow coined a term for a ubiquitous substance in the brain—myelin (Boullerne, 2016). It would take another 100 years before the central function performed by this lipid-rich sheath—saltatory conduction—was agreed upon. The rapid propagation of electrical information within the myelinated axon during saltatory conduction is possible due to the insulating properties of myelin and the nodes of Ranvier (Huxley and Stämpeli, 1949). The nodes of Ranvier are unmyelinated sections of the axon that have a high density of voltage-gated sodium (Na_v) channels that respond rapidly to alterations in charge and function to propagate signals necessary for the depolarization of the next node during an action potential (Arancibia-Carcamo and Attwell, 2014). Myelin is therefore deposited around the axon in sections, known as internodes, which creates the boundaries for the nodes of Ranvier (Figure 1). The internodes effectively insulate the axon by increasing the resistance of the axonal membrane and by reducing the capacitance of the axon (Bakiri et al., 2011; Stadelmann et al., 2019).

Mounting evidence indicates that myelination does not become fixed after development, but is experience-driven and remains adaptive well into adulthood (Young et al., 2013; Bechler et al., 2015; Fields, 2015; Ford et al., 2015; Hill et al., 2018; Hughes et al., 2018; Jünemann et al., 2022). Various structural modifications can alter the efficiency of this system. For example, the density of Na_v channels within the node and the length of the node itself can both alter conduction velocity with minimal energy expenditure (Arancibia-Cárcamo et al., 2017). The diameter of the axon, thickness of the myelin sheath, and the length of the internode can also adjust conduction velocity; with wider axons, thicker sheaths, and longer internodes increasing conduction velocity up to a certain point (Waxman, 1980; Wu et al., 2012; Chapman and Hill, 2020). A large determinant of this adaptability is experience-driven neuronal activity, which has been shown to promote myelination and contribute to the modification of established myelin sheaths (Wake et al., 2011; Gibson et al., 2014; Bechler et al., 2018; Faria et al., 2019). These modifications then alter conduction speed, translating into variations in synapse strength and, therefore, synaptic plasticity (Fields, 2015).

The structural features of the myelin sheath are also important for maintaining the synchronization of action potentials within and between neurons. This synchronization is fundamental to circuit function and proper cognition. Experience-induced changes in myelination can alter this synchronicity and synaptic plasticity to optimize processing time within the circuit contextually, contributing to complex cognitive processes like social behavior, memory, motor learning and sensory experience throughout the lifespan (Liu et al., 2012; Mount and Monje, 2017; Pan et al., 2020). However, the long-range projection neurons that make up the bulk of these white matter pathways are the most vulnerable neurons during aging, and their deterioration can result in negative effects on the aforementioned cognitive processes (Mattson and Magnus, 2006).

In the CNS, oligodendrocyte processes wrap around axons and form the myelin sheath (Boullerne, 2016). These glial cells are capable of extending many different processes to wrap multiple axons at a time, with the ability to differentially alter the parameters of each extension due to the activity of each individual neuron contacted (Chong et al., 2012; Yeung et al., 2014). Furthermore, mature oligodendrocytes are capable of facilitating remyelination after damage in cats and non-human primates, while maintaining established myelin sheaths (Duncan et al., 2018). One particularly relevant finding is that oligodendrocytes preferentially myelinate active axons in development and adulthood, as found in primary cell cultures, zebrafish and mice (Gibson et al., 2014; Hines et al., 2015; Wake et al., 2015; Mitew et al., 2018; Faria et al., 2019). Furthermore, through DTI, neuronal circuits activated during a task were found to have increased levels of myelination in seniors, showing the ability for activity-dependent myelination in the aged human brain (Scholz et al., 2009; Jünemann et al., 2022).

The myelin sheath further provides a channel for metabolic support to the axon. Oligodendrocytes transport various energy metabolites like lactate from their soma through distinct cytoplasmic channels to the innermost layer of the myelin sheath where they are deposited into the periaxonal space (Figure 1; Fünfschilling et al., 2012; Meyer et al., 2018). The subsequent uptake of metabolites supports axon function and is essential for neuronal survival (Lee et al., 2012; Meyer et al., 2018).

However, before an oligodendrocyte can mature and participate in myelination, a complex differentiation process from oligodendrocyte progenitor cell (OPC) must occur. These cells not only populate the developing brain, but remain present in the adult brain, providing a reserve for the replacement of the mature oligodendrocyte population (Bergles and Richardson, 2016). On top of this, OPCs were shown to form direct contacts with neuronal synapses, to participate in synaptic engulfment, and to engage in axon pruning during development and adulthood, thus contributing to circuit formation and modulation in mice (Bergles et al., 2000; Auguste et al., 2022; Buchanan et al., 2022). Neuronal activity conveyed through these synapses relays information to OPCs that contributes to determining their proliferation, migration and differentiation, as well as oligodendrocyte survival, and myelination as a whole (Mitew et al., 2018; Moura et al., 2022).

Astrocytes are an extremely diverse group of glial cells that contribute to the blood–brain barrier (BBB), blood flow modulation, metabolite supply, and perform modulatory roles at the synapse involved in synaptic activity and plasticity (Sofroniew and Vinters, 2010). Furthermore, astrocytes are vital for myelination. The loss of astrocytes reduces the density of oligodendrocytes, initiates demyelination, and promotes myelin abnormalities (e.g., decompaction) in the white matter of developing and adolescent mice (7 days to ~1.5-months-old; Tognatta et al., 2020). These effects are likely partly due to local increases in extracellular glutamate causing excitotoxicity, and reduced trophic support [e.g., platelet-derived growth factor (PDGF)] from astrocytes (Tognatta et al., 2020). The uptake of glutamate from the extracellular space is an essential function performed by astrocytes that prevents excitotoxicity of neurons and glial cells (Mcdonald et al., 1998; Hassel et al., 2003; Matute et al., 2007; Goursaud et al., 2009; Mahmoud et al., 2019). Additionally, the controlled release of gliotransmitters (e.g., glutamate) from astrocytic hemichannels—membrane channels between cells and the extracellular space made up of connexin proteins—directly to synapses modulates synaptic transmission and plasticity, and was even shown to be essential for behavioral outputs including fear memory consolidation in rats (~2-months-old; Ye et al., 2003; Stehberg et al., 2012; Abudara et al., 2018).

Astrocytes also maintain appropriate levels of K⁺ ions in the extracellular space and effectively disperse them throughout the pan-glial network that spans the entire brain (Beckner, 2020). The pan-glial network allows for the diffusion of ions and small metabolites between coupled cells connected through gap junctions, which are made up of adjoining hemichannels (Orthmann-Murphy et al., 2008; Stephan et al., 2021). The loss of astrocyte-oligodendrocyte gap junctions results in myelin pathology (e.g., vacuolation) and loss of astrocytes (Magnotti et al., 2011; Tress et al., 2012). Additionally, many human diseases characterized by demyelination (e.g., MS and neuromyelitis optica) show early disruption of gap junctions between astrocytes and oligodendrocytes and a decline in connexin proteins (Markoullis et al., 2012; Masaki, 2015).

Astrocytes are also a main source of cholesterol, facilitate iron transport, and directly provide oligodendrocytes with metabolic support (via gap junctions) in the adult CNS, all contributing to the processes needed to synthesize myelin (Jurevics and Morell, 2002; Schulz et al., 2012; Saher and Stumpf, 2015; Camargo et al., 2017; Cheli et al., 2020). Lastly, a wide range of soluble factors [e.g., BDNF and chemokine (C-X-C motif) ligand 1 (CXCL1)] released by astrocytes can have myriad effects on myelination, as found in rodents (Tsai et al., 2002; Fulmer et al., 2014; Kıray et al., 2016).

Microglia—the resident immune cells of the CNS—perform vital functions in all stages of life. They act as gardeners, constantly surveying their surroundings looking for debris to clear, shaping and pruning synapses, maintaining appropriate glial and neuronal population sizes, modulating neuronal activity, and releasing various trophic factors to support growth and development of glial cells and neurons (Šimončičová et al., 2022). Microglia also play a substantial role in myelination, contributing to the developmental and experience-driven process of adaptive myelination (Kalafatakis and Karagogeos, 2021; Santos and Fields, 2021). In fact, microglia are present at higher densities in human white matter compared to gray matter, highlighting their importance in this environment (Mittelbronn et al., 2001; Askew et al., 2017). However, evidence is more contradictory for microglial density in mice. One study found increased density of microglia in the white matter of the forebrain, whereas an earlier study found increased density in gray matter from the entire mouse brain (Lawson et al., 1990; Savchenko et al., 2000). Therefore, it is important to keep in mind the region analyzed, as microglial density and function can greatly vary. Additionally, it is important to note that the microglial population is not homogenous, but instead exists as a continuum of states that contribute in divergent fashions to supporting brain development, activity, plasticity and integrity (Paolicelli et al., 2022).

Although microglia have a plethora of functions, three of their activities primarily contribute to myelination. (1) Microglia release a repertoire of soluble factors [e.g., insulin growth factor-1 (IGF-1), IL-1β, TGF-β] that facilitate the promotion and prevention of myelination (Hsieh et al., 2004; Pang et al., 2007; Santos and Fields, 2021; McNamara et al., 2023). (2) Microglia phagocytose myelin debris, which is important as myelin debris can inhibit OPC differentiation, while efficient clearance of myelin debris allows for effective remyelination after experimental demyelination in rodents (Kotter, 2006; Neumann et al., 2008; Lampron et al., 2015). It was also reported that microglia are capable of removing incorrectly deposited myelin directly from the axon, contributing to the refinement of myelin sheaths, as shown in zebrafish and mice during development (Hughes and Appel, 2020; Djannatian et al., 2023). (3) Microglia dynamically contact active axons, guided by the nodal efflux of K⁺ ions. This interaction was associated with improved remyelination after experimental demyelination in mice, and may be a way by which microglia prevent neuronal damage from hyperactivity (Madry et al., 2018; Ronzano et al., 2021). Microglia respond rapidly to hyperactive neurons and wrap their processes around axons to facilitate rapid repolarization, thus preventing excitotoxicity and maintaining neuronal viability in mice (Kato et al., 2016). Therefore, microglia are emerging as essential modulators of neuronal activity that substantially contribute to determining neuronal architecture and function (Badimon et al., 2020; Cserép et al., 2021).

The endocannabinoid system extends to most regions of the body. It encompasses naturally occurring endogenous (endo)cannabinoids, the enzymes needed for their formation and degradation, and cannabinoid receptors (Lu and Mackie, 2016). Although many receptors take part in endocannabinoid signaling [e.g., peroxisome proliferator-activated receptors (PPARs), transient receptor potential cation channels (TRPs)], the two main cannabinoid receptors are cannabinoid receptor type 1 (CB₁R) and cannabinoid receptor type 2 (CB₂R; Howlett, 2002).

CB₁Rs are widespread in the CNS and are the most prevalent G-protein coupled receptor in the mammalian brain (Herkenham et al., 1990; Marsicano and Lutz, 1999). In the hippocampus and cortex, CB₁Rs have an especially high localization on inhibitory neurons, although their distribution and localization patterns differ throughout the human and rodent CNS (Glass et al., 1997; Marsicano and Lutz, 1999; Tsou et al., 1999; Fletcher-Jones et al., 2020). Interestingly, a significant number of CB₁Rs are not expressed on the cell surface, but instead localize to lysosomes and late endosomes, as shown in cell lines and primary cell cultures—possibly contributing to lysosomal integrity and function (Rozenfeld and Devi, 2008; Bilkei-Gorzo, 2012; Fletcher-Jones et al., 2020). Additionally, CB₁Rs also localize to mitochondria and can influence metabolism in neurons and glial cells, as observed in primary cell cultures and in mice (Bénard et al., 2012; Jimenez-Blasco et al., 2020).

However, the majority of neuronal CB₁Rs are found on pre-synaptic terminals, where their primary function is to suppress the release of neurotransmitters, altering the activation of post-synaptic channels and, therefore, modulating synaptic activity and plasticity (Mackie and Hille, 1992; Di Marzo et al., 2015; Zou and Kumar, 2018). The cannabinoid-mediated reduction in neurotransmitter release is achieved by the reduced influx of presynaptic Ca²⁺ due to the inhibition of voltage gated Ca²⁺ channels and of adenylyl cyclase, which downregulates cyclic adenosine monophosphate and protein kinase A, two cellular constituents involved in increasing the influx of Ca²⁺ (Castillo et al., 2012). CB₁R activation in pre-synaptic terminals is mainly facilitated through retrograde signaling of endocannabinoids released from the post-synapse (Kano et al., 2009; Castillo et al., 2012; Njoo et al., 2015).

CB₂Rs have much lower levels of expression in the CNS of humans and rodents (Lu and Mackie, 2016; Jordan and Xi, 2019). Interestingly, CB₂R mRNA expression can vastly increase during an inflammatory insult, with microglia from mice displaying as much as a 10-fold increase (Maresz et al., 2005). This finding indicated that the CB₂R likely plays a substantial role in CNS immune function, which has been subsequently supported in the literature (Turcotte et al., 2016; Komorowska-Müller and Schmöle, 2020). CBRs may attenuate pro-inflammatory cytokine secretion by interfering with the phosphorylation of mitogen activated protein kinases (MAPK), such as extracellular signal-regulated kinase (ERK), which is known to participate in pro-inflammatory pathways, as shown in microglial cell line cultures (Eljaschewitsch et al., 2006; Young and Denovan-Wright, 2022a,b). Furthermore, activation of CB₂Rs promotes IL-10 (an anti-inflammatory cytokine) secretion from primary microglia cells from mice by reducing the translocation of the transcription factor nuclear factor-κB (NF-κB) to the nucleus via reduced phosphorylation of IκB Kinase-α (IKKα)—a subunit of the IKK complex that is essential for NF-κB signaling—which subsequently prevents NF-κB formation (Solt and May, 2008; Correa et al., 2010). Of note, NF-κB-mediated inflammation is often via the inflammasome and is associated with many white matter associated diseases and is upregulated in the aged brain (Tilstra et al., 2011; Youm et al., 2013; Blank and Prinz, 2014; Rea et al., 2018). Evidence also suggests that CB₂Rs are present on post-synaptic terminals of neurons in rodents and non-human primates, although expression levels are relatively low and may depend on brain region (Brusco et al., 2008; Lanciego et al., 2011; Li and Kim, 2015; Stempel et al., 2016).

Importantly, astrocytes, oligodendrocytes, and OPCs possess CB₁Rs and CB₂Rs, highlighting the wide range of functions performed by this system (Navarrete et al., 2014; Ilyasov et al., 2018; Martinez Ramirez et al., 2023). The impact these receptors have with respect to their activation by cannabinoids on individual cell-types will be discussed in more detail below (Figure 2B).

Cannabinoids exert their influence over the endocannabinoid system mainly through CB₁Rs and CB₂Rs, which contribute to the sought after medicinal and recreational qualities of cannabis. The most common psychoactive cannabinoid, Δ-9-tetrahydrocannabinol (THC), has a relatively high affinity for the two cannabinoid receptors (Pertwee, 2008). THC is generally considered to be a partial agonist for both CBRs, although its inhibitory effect on synapses can be comparable to that of a full agonist (Laaris et al., 2010). By contrast, cannabidiol (CBD)—the most common non-psychoactive cannabinoid—does not have a particularly high affinity for either CBR, but was shown to antagonize CBR agonists (Thomas et al., 2007). CBD is a negative allosteric modulator of CB₁Rs, and is suggested to act as an inverse agonist of CB₂Rs (Thomas et al., 2007; Laprairie et al., 2015). The anti-inflammatory effects attributed to CBD may be exerted through this inverse agonism of CB₂Rs (Thomas et al., 2007; Pertwee, 2008; Yu et al., 2020). However, it is important to note that both THC and CBD have many CBR-independent or indirect signaling mechanisms that also contribute to their overall outcomes (Pertwee, 2008; Stella, 2010). Interestingly, one indirect mechanism is the ability for CBD to inhibit fatty acid amide hydrolase (FAAH)—the enzyme required for the degradation of N-arachidonoylethanolamine (anandamide; AEA), an endocannabinoid—which results in an increase in AEA (Watanabe et al., 1996; Bisogno et al., 2001; De Petrocellis et al., 2011). Of note, there is a large body of research looking at the potential benefits of inhibiting endocannabinoid metabolizing enzymes; with studies showing that increases in AEA have an immunomodulatory effect, and can potentially aid in MS, as shown in mice (Rossi et al., 2010; Vázquez et al., 2015a; Vecchiarelli et al., 2021).

The changes observed in the endocannabinoid system in the aging brain are mostly similar to those observed after chronic THC exposure (Yoo et al., 2020). The most noticeable effect observed after chronic THC exposure is the significant but reversible reduction in CB₁Rs, with cortical regions showing more extensive decreases in expression (Hirvonen et al., 2012; D’Souza et al., 2016; Augustin and Lovinger, 2022). However, it should be noted that these studies used exclusively male participants. Studies including females are lacking, which is a significant gap since sex differences with respect to the endocannabinoid system are well-described (Laurikainen et al., 2019; Levine et al., 2021).

Importantly, THC does not interact with the brain in equal measure, as found by Leishman et al. (2018). In this study, acute administration of THC [3 mg/kg; intraperitoneal injection (i.p)] differentially impacted the lipidome and transcriptome depending on the brain region and age of the subject, 2 h after administration. THC is distributed and metabolized in a region-specific manner, with highest levels in the hippocampus. Another interesting finding of this study was that adult mice (~4-months-old) displayed the largest changes after acute THC exposure compared to exposed ~1 and ~2-month-old mice, with a general downregulation of the endocannabinoid system. The effect of THC in the aged brain was not examined in this study, but the observed changes would likely be different than other time points.

Although THC and CBD are the main cannabinoids found in cannabis, it is important to note that 100 s of different cannabinoids and other biologically active compounds exist in the plant, such as terpenes, including β-caryophyllene (Kopustinskiene et al., 2022). These compounds likely work synergistically to produce the effects of cannabis through multiple different signaling pathways, creating what is called the “entourage effect” (Ferber et al., 2020; Finlay et al., 2020). However, due to a lack of literature, the next sections will focus on THC, CBD, and some synthetic cannabinoids.

One of the most essential microglial functions that promotes myelination is the clearance of myelin debris (Kotter, 2006; Lampron et al., 2015). The CB₂R has been shown to be important for phagocytosis in microglia. For instance, in CB₂R-KO primary microglia from mice, phagocytosis was significantly reduced in both steady-state conditions and following an inflammatory stimulus (i.e., TGF-β) compared to controls (Mecha et al., 2015). Similarly, activation of the CB₂R increased phagocytosis in primary cell cultures and improved the removal of amyloid-β₄₀ peptides in a mouse model of Alzheimer’s disease pathology (Ehrhart et al., 2005; Aso et al., 2016). Additionally, microglia from CB₂R-KO mice (18-months-old) had an age-dependent increase in lipofuscin granules compared to the control group, signifying reduced lysosomal degradation (Komorowska-Müller et al., 2021a). These findings indicate a significant role played by CB₂Rs in the proper removal and degradation of debris, and in the microglial response to environmental stimuli. However, other receptors likely also contribute, as CBD has been shown to promote phagocytosis through TRPs in primary microglial cell cultures from mice (Hassan et al., 2014; Yang et al., 2022). Briefly, TRPs are Ca²⁺-permeable channels known for their role in temperature sensation, but interestingly, have also been shown to play a role in the release of pro-inflammatory cytokines from microglia in mice exposed to immunogenic agents [i.e., lipopolysaccharide (LPS)] (Zhang Y. et al., 2021). The observed increase in phagocytosis due to CBD is thought to be related to the TRP-mediated increase in the influx of Ca²⁺ (Hassan et al., 2014).

Increased microglial phagocytosis of myelin debris following the application of 2-AG subcutaneously via an osmotic pump subsequently promoted remyelination in an adult mouse model of experimental demyelination (Mecha et al., 2019). This study also observed an altered secretory profile, with increases in IL-1β, TNF-α, and IL-10 in the brain after 2-AG application. This is an example of an augmented immune response, where the benefits of pro-inflammatory cytokines in tandem with anti-inflammatory cytokines work synergistically to promote remyelination. However, microglia in the aged brain are responding to chronically elevated levels of pro-inflammatory factors resulting in reduced functional capacity.

Activation of the CB₂R by AEA in primary microglial cells from mice has been shown to reduce NF-κB signaling and increase expression of IL-10 (Correa et al., 2010). The anti-inflammatory effect produced by IL-10 is partly through a negative feedback loop with astrocytes, where the binding of IL-10 causes the release of TGF-β from astrocytes, which subsequently attenuates pro-inflammatory cytokine production in primary microglia cells (Norden et al., 2014). However, astrocytes have reduced expression of IL-10 receptor-1 in aged mice (≥18-months-old), and fail to effectively diminish microglia-mediated inflammation (Norden et al., 2016; O’Neil et al., 2022). Therefore, compounds that can act directly on microglia to reduce pro-inflammatory cytokine production are of particular interest.

Selective CB₂R agonists (i.e., JWH-133; i.p.) reduce the release of pro-inflammatory cytokines from microglia in a mouse model of Alzheimer’s disease pathology (Aso et al., 2013). Similarly, activation of CB₁Rs and CB₂Rs by synthetic cannabinoids [i.e., arachidonyl-2^′-chloroethylamide (ACEA) and HU-308, respectively] reduced nitric oxide, TNF-⍺, IL-1β and IL-6 release from spontaneous immortalized microglia (SIM)-A9 cells in culture (Young and Denovan-Wright, 2022a).

CBD also has the ability to beneficially regulate the oxidative status in microglia by acting as an antioxidant, where it may directly scavenge ROS and/or inhibit the phosphorylation of upstream kinases needed for NF-κB signaling in primary microglia cells cultured from mice (van den Berg et al., 2001; dos-Santos-Pereira et al., 2020; Atalay Ekiner et al., 2022). This results in reduced levels of IL-1β and TNF-α independently of CB₁R, CB₂R or PPARγ, as tested using receptor antagonists. These effects could also be partly regulated by increased levels of endocannabinoids or through TRP channels, since CBD has a relatively high affinity for TRP vanilloid receptor 1 (TRPV1), which has been shown to modulate cytokine production in microglia (Stampanoni Bassi et al., 2019). Furthermore, in a mouse (1-month-old) model of viral-induced demyelination, CBD (5 mg/kg; i.p) attenuated morphological alterations in microglia, and reduced production of IL-1β, chemokines, and vascular cell adhesion molecule-1 (VCAM-1) through adenosine A_2A receptors (Mecha et al., 2013). Interestingly, VCAM-1—a protein expressed by endothelial cells in the BBB that is involved in peripheral immune cell recruitment—expression increases in the aged brain, and the application of anti-VCAM-1 antibodies reduces microglial reactivity and improves memory and learning in aged (19-month-old) mice (Yousef et al., 2019). Overall, it is clear that CBD acts through a number of vastly different pathways that have overlapping effects on the brain.

Although the anti-inflammatory effects of the CB₂R are well-characterized, it is also important to note that the CB₂R seems essential for many environmental-induced immune responses in microglia. Primary microglia cell and organotypic hippocampal slice cultures generated from CB₂R-KO mice showed attenuation of the microglial immune response to toll-like receptor (TLR) ligands (e.g., TLR4/3/9), preventing the pro-inflammatory cascade usually associated with TLR ligands (Reusch et al., 2022). Therefore, although the CB₂R can function to suppress inflammation, it also contributes to its initiation. These studies highlight how complex the interaction between the endocannabinoid system and microglia is.

The CB₁R is also important for the inflammatory response, as a recent study found that inflammation was dependent on microglial CB₁Rs (De Meij et al., 2021). They found decreased pro-inflammatory cytokines in mice (2–5-months-old) with CB₁R-KO microglia exposed to LPS. However, it also increased sickness behavior in male, but not female mice (De Meij et al., 2021). Similarly, ablation of the CB₁R resulted in early age-related cognitive deficits in mice (Bilkei-Gorzo et al., 2005; Albayram et al., 2011). Therefore, the activation of CB₁Rs may be beneficial for preventing sickness behavior and age-related cognitive decline. This is true with respect to THC, but only with certain doses (Sarne, 2019). Ultra-low (0.002 mg/kg; i.p) and low (3 mg/kg; i.p) doses of THC resulted in improved cognitive function in old mice (24 and 18-month-old, respectively), whereas the same dose induced cognitive impairments in adult mice (2-months-old; Bilkei-Gorzo et al., 2017; Sarne et al., 2018). However, higher doses of THC have the reverse effect (Calabrese and Rubio-Casillas, 2018). This dose-dependent alteration was also observed in microglia in the 2-month-old mouse brain, where higher doses (20 mg/kg; i.p) of THC resulted in the increased release of pro-inflammatory cytokines compared to lower doses (Cutando et al., 2013). In line with this, a recent study found that adolescent mice exposed to daily low-doses of THC (5 mg/kg; i.p) resulted in the downregulation of genes responsible for the microglial response to an immune insult (e.g., IL-1β, IL-6), which carried over into young adulthood (~1.5-months-old) but not maturity (~4-months-old; Lee et al., 2022). This blunting of the microglial response attenuated reactivity to LPS and, furthermore, altered behavior in mice, where exposed mice showed an inability to react appropriately to psychosocial stress (Lee et al., 2022). These impacts were mediated through the CB₁R, as tested using receptor antagonists. These studies highlight the differential impact cannabis can have on the developing brain and the aged brain, where low doses may be detrimental to young animals, yet beneficial to aged animals.

The reduced production of pro-inflammatory cytokines with THC administration may occur through the inhibition of NF-κB signaling in microglia and astrocytes, as observed in cell culture experiments (Kozela et al., 2010; Rizzo et al., 2019). Interestingly, the overexpression of CB₁Rs in adult mice undergoing experimental demyelination resulted in delayed onset and reduced severity of symptoms, whereas CB₁R antagonism quickened symptom onset and increased the expression of inflammatory cytokines and NF-κB proteins (Lou et al., 2016, 2018). Furthermore, an alteration in cell state and expression profile (i.e., increase in pro-inflammatory cytokines and nitric oxide) was observed in cultured BV-2 cells treated with a CB₁R antagonist (Lou et al., 2018).

It is clear that cannabinoids are involved in the microglial response to inflammation with respect to secretory profile and phagocytosis. This is the major way in which microglia may modulate myelination. However, it is important to note that there is still a paucity in in vivo experiments conducted with aged mice, emphasizing the need for increased research (Scipioni et al., 2022).

The crosstalk that exists between microglia and astrocytes is essential for proper function and plasticity of the brain and maintenance of homeostasis (Jha et al., 2019; Matejuk and Ransohoff, 2020). However, in aging, certain aspects of this communication network become exacerbated. As discussed, the aged brain environment alters microglial state including their release of soluble factors. The concomitant impact aging has on astrocytes also induces a phenotypic switch that results in altered gene expression, perpetuation of inflammation, and recruitment of peripheral immune cells, which further exacerbates inflammation (Palmer and Ousman, 2018; Jha et al., 2019).

The ability for THC to inhibit NF-κB and reduce the release of IL-1β and TNF-α from microglia also extends to monocytes/macrophages and lymphocytes in human and rodent cell lines and primary cell cultures (Shivers et al., 1994; Kozela et al., 2010; Rizzo et al., 2019; Henriquez et al., 2020). This inhibition then translates into an observed reduction in the astrocytic release of IL-6 and monocyte chemoattractant protein-1 (MCP-1) in primary human cell cultures, thereby reducing inflammatory signaling and peripheral immune cell recruitment, respectively (Rizzo et al., 2019; Henriquez et al., 2020). Although IL-6 can be beneficial for many aspects of development including myelination, the chronically increased levels that are seen in aged humans and rodents are damaging and pro-inflammatory in nature (Ferrucci et al., 1999; Godbout and Johnson, 2004; Kimura and Kishimoto, 2010; Ritzel et al., 2016; Porcher et al., 2021).

This protective effect offered by THC is thought to be facilitated by the activation of CB₂Rs, which is elevated in microglia and astrocytes within an inflammatory environment (Benito et al., 2008; Di Marzo et al., 2015; Cassano et al., 2017). However, other receptors may also play a role. Indeed, a recent study observed a marked inhibition of pro-inflammatory cytokines produced by IL-1β-stimulated primary cell cultured human astrocytes when pre−/co-treated with WIN55,212-2—a synthetic cannabinoid that displays similar effects to THC (Compton et al., 1992; Fields et al., 2022). This effect was independent of the CB₁R and PPARs (Fields et al., 2022). Conversely, through PPARγ, CBD reduced pro-inflammatory cytokine release, inhibited NF-κB, and reduced GFAP expression in primary astrocyte cells stimulated with amyloid-β_1-42 peptides, while also promoting neurogenesis in adult rats (Esposito et al., 2011). Furthermore, as discussed, CBD has the ability to diminish microglial cytokine production by scavenging ROS, diminishing NF-κB activity, reducing VCAM-1 levels and increasing the availability of endocannabinoid ligands. This may also lessen the extent to which astrocytes participate in peripheral immune cell recruitment and inflammation (Mecha et al., 2013). Indeed, a recent study found an association between CBD administration, reduced phosphorylation of NF-κB and reduced release of IL-6 from mouse primary cultured astrocytes stimulated with LPS (Wu et al., 2021).

However, cannabis includes both THC and CBD. Administration of Sativex^®—an approved oromucosal spray (Health Canada and various European health agencies) containing THC (5 mg/kg) and CBD (5 mg/kg) for the treatment of symptoms associated with MS—reduced astrocyte reactivity and decreased the expression of pro-inflammatory cytokines released by microglia in a mouse model of MS (Feliú et al., 2015). Sativex^® also preserved myelin morphology in mice exposed to virus-induced demyelination.

Inflammation can also be induced by disrupting the proper communication between astrocytes and neurons. The deletion of CB₁Rs from GABAergic neurons enhanced a phenotypic switch in astrocytes already associated with aging, including increased GFAP expression and amplified pro-inflammatory cytokine secretion in mice (Bilkei-Gorzo et al., 2018). Therefore, disruption of endocannabinoid signaling between neurons and astrocytes—which naturally occurs during aging—causes a deleterious transition in astrocytes that perpetuates cytokine-mediated damage. The application of cannabinoids may help since they directly act on astrocytes to diminish cytokine release (Sheng et al., 2005; Aguirre-Rueda et al., 2015; Rizzo et al., 2019; Fields et al., 2022).

Chronic inflammation was also identified as a major influencer of astrocytic gap junctions and hemichannels in rodents and humans (Bronzuoli et al., 2019; Peng et al., 2022). Typically, hemichannels remain mostly closed under “normal” conditions, but can be opened during pathological conditions; whereas the reverse is true for gap junctions (Peng et al., 2022). IL-1β and TNF-α released from microglia can open astrocytic hemichannels and reduce coupling between astrocytes in primary cell/slice cultures and in vivo in mice (Même et al., 2006; Retamal et al., 2007; Abudara et al., 2015; Vázquez et al., 2015b). The application of synthetic cannabinoids (e.g., WIN55,212-2) were able to reduce the microglial release of these factors and directly act on primary astrocyte cells from mice to reduce hemichannel activation, preventing astrocytic uncoupling and maintaining gap junctions (Froger et al., 2009; Gajardo-Gómez et al., 2017). Furthermore, direct activation of astrocytic CB₁Rs was found to be required for the observed decrease in hemichannel activation, as determined by CB₁R antagonism (Froger et al., 2009; Gajardo-Gómez et al., 2017). On the contrary, AEA (1 μM; topical application through a cortical cranial window) was shown to increase hemichannel activity in adult mice in vivo, resulting in a release of ATP that caused microglial process extension and migration toward the injury site (Vázquez et al., 2015b). Similarly, THC (5 mg/kg; i.p) has been shown to result in an increase in glutamate in the extracellular space by binding to astrocytic CB₁Rs in mice, contributing to long term depression at synapses and impairments in working memory (Navarrete and Araque, 2010; Han et al., 2012).

During acute inflammatory conditions, this increase in hemichannel activity may be beneficial for mounting an immune response to an insult, however, the prolonged release of many factors (i.e., glutamate) from astrocytes in chronic conditions may be associated with altered synaptic plasticity and memory impairments (Navarrete and Araque, 2010; Han et al., 2012; Vázquez et al., 2015b; Labra et al., 2018). Alternatively, if cannabinoids are able to reduce the opening of hemichannels during a chronic inflammatory event, it may subsequently prevent excitotoxicity caused by excess glutamate and reduce the release of pro-inflammatory cytokines, thus maintaining neuronal and astrocyte viability (Froger et al., 2009; Gajardo-Gómez et al., 2017). Indeed, a recent review highlights a potential signaling cascade involving NF-κB, p38 and nitric oxide in which cannabinoids prevent the release of glutamate from hemichannels (Labra et al., 2018). It can also be hypothesized that the conserved function of gap junctions would facilitate proper communication between astrocytes and oligodendrocytes, promoting myelination (Papaneophytou et al., 2019).

In summary, elevated levels of inflammatory factors partly initiated by microglia and other recruited immune cells causes a phenotypic switch in astrocytes that contributes to functionally perpetuating inflammation via the release of pro-inflammatory cytokines and recruitment of peripheral immune cells. Cannabinoids can inhibit the release of these compounds (i.e., TNF-α, IL-6, MCP-1) by reducing the activity of pro-inflammatory pathways (e.g., NF-κB) in immune cells and astrocytes and increasing the availability of endocannabinoids. The subsequent decrease in pro-inflammatory factors and direct action of (endo)cannabinoids may also modulate hemichannel activity, thereby contributing to changes in extracellular homeostasis, synaptic activity and plasticity, as well as glial support functions. Although these results from studies using younger animals can provide information on how cannabinoids impact astrocytes and what this could mean for the process of myelination, the unique environment present in the aged brain makes extrapolation conjectural.

As discussed, oligodendrocytes and OPCs are both essential for proper myelination and maintenance of myelin. Chronic increases in ROS, pro-inflammatory cytokines and other damaging compounds released from immune cells and astrocytes have the ability to damage mature oligodendrocytes and OPCs, resulting in myelination impairments (Pang et al., 2003; Jurewicz et al., 2005; French et al., 2009; Peferoen et al., 2014; Guttenplan et al., 2021; Sams, 2021). The ability for cannabinoids to reduce the release of pro-inflammatory cytokines from immune cells and astrocytes could therefore aid in the preservation of oligodendrocytes and OPCs. Furthermore, the antioxidant capacities offered by CBD could also protect these cells from oxidative stress.

However, in primary oligodendrocyte cell cultures from 12-day-old rats, CBD (100 nM–1 μM) resulted in mitochondrial dysfunction that led to increases in intracellular cytotoxic Ca²⁺ and ROS, which negatively impacted oligodendrocyte viability (Mato et al., 2010). Conversely, Mecha et al. (2012) found that CBD (1 μM) administered to inflammatory-induced primary oligodendrocyte cell cultures from the cortex of 2-day-old rats protected OPCs from oxidative stress and apoptosis. These studies found that these effects were independent of CB₁R, CB₂R, or PPARγ. The apparent discrepancy between studies is likely due to dosage and differences in age, region, and maturation of oligodendrocyte cells (Molina-Holgado et al., 2022).

Interestingly, CBD was also found to influence genes related to glycolysis—the major energy source for mature oligodendrocytes—and carbohydrate metabolism in oligodendrocytes (Rao et al., 2017; de Almeida et al., 2022). The data presented by de Almeida et al. (2022) suggests a slight downregulation of glycolysis in OPC and mature oligodendrocyte cell cultures (MO3.13). This finding is important because glycolysis produces lactate, an important energy metabolite transferred to neurons from oligodendrocytes, which has been shown to be essential for proper neuronal function in aged mice (12–24-months-old; Fünfschilling et al., 2012; Lee et al., 2012; Philips et al., 2021). Interestingly, the ablation of monocarboxylate transporter 1 (MCT1)—a lactate transporter—from OPCs resulted in hypomyelination and axonal degeneration in mature and older (18–24-months-old) mice, highlighting the understudied role played by OPCs in myelination among the adult brain (Philips et al., 2021). MCT1 expression naturally declines in the aging mouse brain, especially after 15-months of age (Ding et al., 2013; Philips et al., 2021). Therefore, further reductions in glycolysis due to CBD administration may prevent the beneficial support to neurons offered by oligodendrocytes and OPCs. Similarly, activation of mitochondrial CB₁Rs with THC resulted in reduced glycolytic activity and lactate production in primary cultured astrocytes (0–1-days-old), which was hypothesized to contribute to the subsequent impaired neuronal function and altered behavior (e.g., social interaction deficit) in young adult (8–12-weeks-old) mice (Jimenez-Blasco et al., 2020). This reduction in astrocytic lactate would also likely have an impact on the oligodendrocyte lineage cells. For example, global inhibition of lactate production prevented remyelination in the corpus callosum after experimental demyelination in mice (~3-months-old), while OPC-rich primary cell cultures displayed heightened differentiation when lactate was added to the glucose medium (Ichihara et al., 2017). Perhaps this increase in OPC differentiation due to the presence of lactate can aid in remyelination by replacing the oligodendrocyte pool, which would warrant further investigation.

Although the literature is split with respect to OPC number in the aged brain, their ability to differentiate is severely impaired in humans and rodents (Sim et al., 2002; Yeung et al., 2014; Segel et al., 2019; Luan et al., 2021; Rivera et al., 2021; Zhang X. et al., 2021). Therefore, the encouragement of differentiation and protection of OPC viability offered by certain cannabinoids is of particular relevance (Ilyasov et al., 2018; Molina-Holgado et al., 2022). 2-AG was shown to promote the differentiation of OPCs into mature myelinating oligodendrocytes in primary mixed glial cell cultures from rats through the activation of CB₁Rs and CB₂Rs (Gomez et al., 2010). OPCs also have the ability to produce and release 2-AG themselves, indicating a potential autocrine and paracrine signaling mechanism that can further promote differentiation (Gomez et al., 2010, 2011). Pathways of note that have been shown to promote OPC differentiation through cannabinoid receptors are the phosphatidylinositol 3-kinase/Akt (PI3K/Akt) and mammalian target of rapamycin (mTOR), ERK/MAPK, and Rat sarcoma (Ras) homolog family member A/Rho-associated protein kinase (RhoA/ROCK; Molina-Holgado et al., 2002; Narayanan et al., 2009; Tyler et al., 2009; Gomez et al., 2010, 2011; Giacoppo et al., 2017; Sánchez-de la Torre et al., 2022; Wang et al., 2022).

These findings have prompted further research investigating the effect of THC on OPC differentiation. A recent study found that the application of THC (3 mg/kg; i.p) to young mice (6-days-old) and organotypic cerebellar cultures promoted the differentiation of OPCs (Huerga-Gómez et al., 2021). Furthermore, the same group found that THC (3 mg/kg; i.p) induced OPC differentiation and remyelination after experimentally-induced demyelination in the corpus callosum of adult (6–8-weeks-old) mice (Aguado et al., 2021). The modulation of OPC differentiation by THC is thought to be mediated by both CBRs, but mainly CB₁Rs, since antagonism of CB₁Rs prevented the beneficial effects observed. Furthermore, OPC(Ng2/Ai6)-CB₁R-KO mice displayed impaired OPC differentiation and myelination in the corpus callosum throughout the examined lifespan (≤2-months-old; Sánchez-de la Torre et al., 2022). The impaired differentiation of OPCs lacking CB₁Rs is thought to be partly due to an increase in RhoA/ROCK signaling. For example, THC (3 mg/kg, i.p) administered to WT mice resulted in reduced RhoA/ROCK proteins compared to vehicle exposed mice, leading to increased myelin-related proteins and enhanced OPC differentiation (Sánchez-de la Torre et al., 2022). However, OPC-CB₁R-KO mice had increased levels of RhoA/ROCK proteins observed with a lower density of mature oligodendrocytes and reduced immunofluorescence against myelin-related proteins, which did not significantly change with THC administration (Sánchez-de la Torre et al., 2022). These findings highlight the importance of the CB₁R in OPC differentiation, and its relationship with the RhoA/ROCK pathway. Furthermore, these findings are supported by previous studies that have shown the importance of RhoA/ROCK signaling in OPC differentiation (Baer et al., 2009; Pedraza et al., 2014). Similarly, the application of WIN55,212-2 (i.p) resulted in improved remyelination after experimentally-induced demyelination in mice (6–7-weeks-old) when administered at a dose of 0.5 mg/kg; whereas a dose of 1 mg/kg impaired remyelination (Tomas-Roig et al., 2020). The negative effects produced by 1 mg/kg of WIN55,212-2 are hypothesized to be due to stronger reductions in Ca²⁺ influx, resulting in reduced neuronal activity, which possibly hinders activity-dependent myelination (Tomas-Roig et al., 2020). Overall, it is evident that at the right dose cannabinoids can alter myelination and promote the maturation of oligodendrocyte lineage cells.

In summary, cannabinoids at the right doses have the ability to aid in remyelination and promote OPC differentiation after pathological insults. As outlined throughout this review, cannabinoids act on a plethora of different cell types and different targets within those cells, which directly and indirectly impact myelination. Although cannabinoid-induced alterations to microglia and astrocytes can influence OPC differentiation and oligodendrocyte function indirectly, it is also important to keep in mind that cannabinoids directly act on OPCs to influence maturation and function. However, there are a lack of studies regarding the healthy aged brain. Although the aged brain does have increases in ROS, pro-inflammatory cytokines, and a general reduction in glial cell function, the extent of these changes and the dynamics at play are inherently different than those observed after the application of LPS or experimentally-induced demyelination at younger ages. Furthermore, findings from cell line and primary cell cultures need to be validated in in vivo experiments. Therefore, although studies outlined here point toward a role played by cannabinoids in myelination, more studies specifically looking at their impact in the healthy aged brain are required to confirm if these findings do indeed translate. Nonetheless, the evidence provided thus far suggests that cannabinoids may help promote myelination in the aged brain. Neuroimaging studies offer a different approach to visualize how cannabinoids impact brain communication and myelination, which allows elucidating large-scale changes that may result from the impact they have on glial cells and neuronal function.

Neuroimaging studies are valuable non-invasive techniques used to visualize large-scale changes in activity, connectivity, and structural alterations in the brain. DTI is a magnetic resonance imaging (MRI) technique that differentiates the degree and direction of water diffusion within an allotted space (Pierpaoli and Basser, 1996). Three main parameters are given by DTI for white matter: fractional anisotropy (FA), mean diffusivity (MD) and radial diffusivity (RD; Alexander et al., 2007). Generally, as described by Becker et al. (2015), increases in FA and decreases in MD and RD reflect increases in myelination.

The DTI literature encompassing cannabis use and myelin integrity is not consistent. Difficulties with respect to accurate reporting on usage and concentration of cannabis, and the ratio of cannabinoids—not to mention other confounding factors such as lifestyle and metabolism—make human cannabis studies increasingly difficult to perform. Furthermore, the DTI metrics that are used as proxies of myelin integrity do not specifically identify myelin abnormalities, but detect differences in water diffusion—a parameter that can be altered by axon packing density, axon caliber, and more confounding factors (Chang et al., 2017). However, it has been shown that FA can correlate with myelination quite accurately (Chang et al., 2017). With this in mind, some tentative directions can be ascertained.

Studies focusing on adolescent and young adults who heavily use cannabis have drawn cautious conclusions. Overall, reductions in FA and increases in RD and MD have been observed with chronic exposure to cannabis (Arnone et al., 2008; Ashtari et al., 2009; Gruber et al., 2014; Becker et al., 2015; Shollenbarger et al., 2015). However, Cousijn et al. (2022) did not find any significant differences between chronic, sporadic or control groups, while noting an association between reduced FA and the onset of regular cannabis use at younger ages (<18-years-old). This finding is generally supported in the literature, with many studies indicating that an earlier age of onset correlates with reduced myelin integrity and cognition (Gruber et al., 2012, 2014; Lisdahl et al., 2013; Orr et al., 2016).

Other studies with larger age ranges (~22–55-years of age) revealed similar results with respect to heavy cannabis use. Jakabek et al. (2016) found reductions in FA in the forceps minor across all age ranges (18–55-years of age), but RD differed depending on age, with younger users having reduced RD and older users having higher RD. Similarly, Manza et al. (2020) reported compromised myelin integrity in heavy cannabis users compared to controls within the same age range. However, Filbey et al. (2014) observed an increase in FA and a decrease in RD, suggesting improvements in white matter integrity, although protracted use did result in a reversal of these findings.

On the contrary, a recent DTI study found that adults (47.85 ± 17.42-years of age) who use medical cannabis with moderate levels of CBD and low levels of THC have significantly increased FA and reduced MD in multiple white matter regions—notably the genu of the corpus callosum—after 6 months of use (Dahlgren et al., 2022). An important distinction between this study and the previously mentioned studies is that the participants were using cannabis for medicinal purposes, not recreationally. As stated by Dahlgren et al. (2022), the medicinal use of cannabis likely involves drastically different characteristics (i.e., age of onset, ratio of cannabinoids), potentially resulting in the increases in myelin integrity found in this study. Furthermore, the medicinal use of cannabis is likely to help treat symptoms associated with increased levels of inflammation, perhaps altering cannabinoid function compared to recreational users. These findings are supported by the numerous animal studies highlighted in this review that show that lower doses of cannabinoids administered less frequently are more beneficial than the reverse.

Despite indications that the brains response to cannabis substantially changes with age, to the best of our knowledge, no DTI study looking at the integrity of myelin has yet been performed in cannabis using seniors, creating a substantial gap in the literature. Through the use of functional MRI, functional connectivity—the statistical relationship that exists between different regions of the brain that are necessary for cognitive processes—can be quantified (Gaudet et al., 2020). Importantly, these functional connections are supported by, and significantly associated with white matter pathways (Hunt et al., 2016; Meier et al., 2016; Huntenburg et al., 2017; Vandewouw et al., 2021). Again, few studies are available for adults over the age of 60. However, one study found increased functional connectivity between the anterior cerebellum and the hippocampus, and with the posterior parahippocampal cortex in cannabis users aged 60–80-years-old (Watson et al., 2022). The authors suggest this finding may indicate a potential benefit of cannabis in the aged brain, although the extent and appearance of these potential benefits are still unclear.

Due to the paucity of studies conducted in older human adults, it is difficult to identify the impact cannabis has on myelination in this population using neuroimaging techniques at this time. Regardless, this evidence forms a basis for future exploration.