Numerous preclinical studies have reported that synthetic and naturally occurring cannabinoids effectively attenuate inflammatory and neuropathic pain, including DPN pain.[1415] Table 1 summarizes the preclinical studies that have investigated the potential benefits of modulating ECS to attenuate pain in rodents with STZ-induced DPN. An earlier study reported that oral administration of THC, a nonselective CB₁R/CB₂R agonist, to STZ-induced neuropathic rodents reduced thermal hyperalgesia.[61] Subsequent studies have found that systemic or intrathecal administration of the synthetic nonselective CB₁R/CB₂R agonist, WIN-55,212-2, decreased thermal and mechanical hyperplasia drastically in STZ-induced neuropathic rodents in a dose-dependent manner.[14466263] Similarly, the systemic or local administration of WIN-55,212-2 to either STZ-induced neuropathic rats or Zucker diabetic fatty rats (Type 2 diabetes) alleviated mechanical allodynia,[64] while pretreatment with the CB₁R-selective antagonist AM251 or the CB₂R-inverse agonist SR144528 reversed WIN 55,212-2's antinociceptive properties. This finding suggests that WIN 55,212-2's antinociceptive effects are mediated through the activation of CB₁R and CB₂R.[64] Another study reported that chronic administration of 2-Methyl-2’-F-anandamide, a CB₁R selective agonist, attenuated mechanical hyperalgesia in STZ-induced diabetic neuropathy.[62] However, the uses of nonselective cannabinoids that activate the CB₁R/CB₂R in the CNS and the PNS are associated with numerous unwanted CNS side effects, including sedation, hypothermia, catalepsy, hypolocomotion, and psychological problems that are caused by CB₁R activation present in the brain.[77] Therefore, peripherally restricted cannabinoid agonists may help avoid these adverse side effects. Activation of peripherally expressed CB₁Rs plays a vital role in cannabinoid-induced antinociceptive. Peripherally, restricted CB₁Rs agonists were found to alleviate pain effectively by activating CB₁Rs expressed on peripheral nociceptors.[34] Numerous peripherally restricted synthetic cannabinoid agonists have been synthesized and showed analgesic effects in neuropathic pain animal models.[15] Oral administration and local injection of CRA13, a peripherally acting nonselective CB₁R/CB₂R agonist, reduced both thermal hyperplasia and mechanical allodynia in rodents with neuropathic pain. CRA13 did not produce the CNS side effects observed with CB₁R agonists.[78] Further, CRA13's antinociceptive effect was diminished by the coadministration of rimonabant (SR141716), a CB₁R-selective antagonist. However, the antinociceptive effect of CRA13 was still observed when the CB₂R-inverse agonist, SR144528, was coadministered. These findings confirmed that this compound's antinociceptive action is attributable to the activation of peripheral CB₁Rs.[78] In a neuropathic pain model, the administration of the peripherally acting nonselective CB₁R/CB₂R agonist AZ11713908 reduced mechanical allodynia.[79] The analgesic effect of AZ11713908 was observed in CB₂R knockout mice but diminished in CB₁R knockout mice. This finding indicates that the activation of CB₁Rs in the periphery contributes to this compound's analgesic effects.[79] However, these compounds’ ability to attenuate DPN pain effectively has not yet been evaluated in rodent models.

Another strategy to avoid unwanted CB₁R-mediated CNS side effects is the use of CB₁R positive allosteric modulators (PAMs).[8081] CB₁R PAMs bind to the CB₁R at an allosteric site (s) on the CB₁R that are distinct from the orthosteric-binding site. The binding of a PAM to CB₁R leads to conformational changes within the receptor, which enhances the affinity and/or efficacy of the orthostatic ligand for the CB₁R. Unlike CB₁R orthosteric agonists, CB₁R PAMs do not have psychoactive effects or lead to tolerance or dependence.[8082] GAT221 (racemic mixture of GAT228 and GAT229), a CB₁R allosteric modulator characterized recently, has shown promising results in suppressing pain in chemotherapy-induced neuropathic mice and did not lead to tolerance or dependence.[83] Whether CB₁R PAMs are efficacious in attenuating neuropathic pain in a PDN pain model has not been evaluated to date.

Similar to other GPCRs, CB₁R can interact physically with other receptors, such as dopamine, adenosine, and opioid receptors.[848586] Cannabinoid dimerization can expand cell signaling diversity in response to ligands via various mechanisms of allosteric control among receptor heteromers.[8788] In diabetic mice, the coadministration of THC and morphine was found to enhance morphine's antinociceptive properties.[61] Although the CB₁R-opioid heteromer-specific mechanisms probably are responsible for these effects, more research is needed to confirm this finding. Given that CB₁R heteromers can exert effects that are unique from its constituent receptors, it may be promising to target CB₁R heteromers to attenuate DPN pain.

Emerging evidence from preclinical studies has suggested that activation of the peripheral CB₁Rs enhances the oxidative stress and inflammatory processes and results in microvascular and neuronal impairment.[154989] In diabetic rodents, CB₁R stimulation has been attributed to cannabinoids’ adverse metabolic and cardiovascular effects. These adverse effects of CB₁R agonists are the primary hindrance to their successful use to manage DPN pain.[154989] Indeed, studies have found that CB₁Rs’ inhibition may be beneficial in DPN rodent models. For example, the chronic administration of rimonabant (SR141716), a CB₁R-selective antagonist, to STZ-induced diabetic mice reduced intraepidermal nerve fiber density loss significantly. In addition, rimonabant treatment decreased skin capillary loss, improved skin blood flow, and decreased tumor necrosis factor-alpha (TNF-α) levels in STZ-induced diabetic mice's skin tissue. These findings suggest that rimonabant exerts anti-inflammatory and vasoprotective effects and might be a beneficial treatment for PDN pain.[65] Consistent with this finding, Comelli et al. reported that rimonabant treatment attenuated mechanical allodynia in a time-and dose-dependent manner.[66] Further, it reduced TNF-α overproduction in the spinal cord and oxidative stress in the peripheral nerves, while it neutralized the nerve growth factor deficit in STZ-induced diabetic rodents. These findings imply that CB₁R antagonists interfere with neuronal impairment pathways and promote nerve regeneration.[66] Rimonabant has shown promising results in clinical trials, as it improved several metabolic risk factors and reduced weight loss.[90] Although CB₁R antagonists may have benefits in treating DPN pain, their use has been hindered largely by their adverse psychiatric effects, such as anxiety and depression.[49] The development of peripherally restricted CB₁R antagonists could be a promising strategy to alleviate painful DPN without these adverse CB₁R-mediated CNS effects.

Together, the activation of peripheral CB₁Rs produces robust antinociceptive properties; however, some studies have revealed that activation of the CB₁R is associated with undesirable diabetic complications. In contrast, recent studies have reported the beneficial effects of blocking CB₁Rs to treat DPN. These conflicting findings might be attributable to the different experimental designs, animal species, and bias agonists.[9192] Further research is required to determine what type of CB₁R modulation is favorable to manage DNP.

In preclinical studies, activation of CB₂R has been shown to exert neuroprotective effects and attenuate neuropathic pain by inhibiting immune cells and microglia-driven inflammation.[15] For example, intrathecal administration of WIN-55,212-2 to STZ-induced diabetic mice reduced thermal hyperplasia significantly.[4663] The coadministration of AM630, a CB₂R antagonist, but not AM251, a CB₁R inverse agonist, diminished WIN-55,212-2's antinociceptive effects significantly. These findings suggest that the activation of CB₂R in the spinal cord mediates cannabinoids’ antinociceptive properties.[46] Notably, CB₂R's activation by its selective agonists is not associated with unwanted CB₁R-mediated CNS side effects. A study found that the selective CB₂R agonist, L-759,656 reduced thermal hyperplasia in mice with STZ-induced diabetic neuropathy considerably, and AM63 reversed this effect.[46] Similarly, the CB₂R agonists, MT178 and AM124, reduced mechanical hyperalgesia in dose-related manners in diabetic mice.[6293] Further studies have found that the intra-planter injection of the CB₂R agonist JWH-015 attenuated mechanical allodynia in mice with STZ-induced-diabetic neuropathy. A CB₂R antagonist inhibited the antinociceptive action of JWH-015, which suggests that peripheral CB₂R is involved in this compound's antinociceptive action.[6768] All of these studies have highlighted the important role that CB₂R receptors play over CB₁R in the antinociceptive effects of cannabinoids and further support the hypothesis that selective CB₂R agonists may offer an exciting approach to reduce pain in DPN.[15]

The upregulation of TRPV1 channels in the dorsal root ganglia and the myelinated primary afferent neurons have been reported to mediate diabetic thermal hyperalgesia.[94] The topical application of the TRPV1 agonist capsaicin attenuated thermal and mechanical hyperalgesia observed in STZ-induced diabetic mice in a dose-dependent manner, while pre-treatment with the TRPV1 antagonist capsazepine blocked capsaicin's effects.[69] Further, a study of STZ-induced diabetic rats showed that alpha-lipoic acid, a TRPV1 agonist, attenuated neuropathic pain and normalized TRPV1 expression in the dorsal root ganglion through the inhibition of the nuclear factor NF-κB.[70] These studies demonstrated that stimulation of the TRPV1 channels is a useful approach to attenuate painful DPN.[94]

Elevation of endocannabinoids has been observed in both the CNS and the periphery in rodents with DPN. Compounds that inhibit the endocannabinoid metabolizing enzymes, FAAH and MAGL, increase the endocannabinoid levels by inhibiting their degradation. Research based on preclinical neuropathic animal models has reported that FAAH and MAGL inhibitors produce antinociceptive effects with fewer side effects relative to CB₁R/CB₂R agonists.[1549] Systemic administration of URB597, a FAAH inhibitor, decreased both mechanical allodynia and formalin-induced cold allodynia in mice with diabetic neuropathic pain.[71] Similarly, significant anti-allodynic effects were observed following the oral administration of the reversible FAAH inhibitor, ST4070, to mice with STZ-induced neuropathic pain.[72] The coadministration of selective CB₁R and CB₂R antagonists and the PPARα antagonist diminished this compound's anti-allodynic effects.[72] Further, the systemic administration of the peripherally restricted FAAH inhibitor, URB937, attenuated STZ-induced hyperalgesia and allodynia.[73] Recently, several MAGL inhibitors have been developed and examined in neuropathic pain models. For example, the selective MAGL inhibitor, MJN110, was reported to reduce mechanical allodynia in rodents with DPN.[74] These studies have provided evidence that supports the beneficial effects of modulating endocannabinoid levels to treat pain associated with DPN.

PEA is the endogenous ligand for PPARα, and it also has an affinity for the GPR55 and GPR119 receptors. Acute and repeated PEA administration alleviated mechanical allodynia, neutralized the nerve growth factor deficit, and improved insulin levels in STZ-induced diabetic mice.[75] Importantly, PEA produces an antinociceptive effect without causing tolerance.[75] A further study by Impellizzeri et al. reported that oral administration of a micronized PEA formulation attenuated mechanical and thermal hyperalgesia and reduced mast cell and microglial activation in mice with DPN. This has been linked to an inhibition of the nuclear factor NF-κB inflammatory pathway. PEAs antinociceptive and neuroprotective properties are mediated primarily by the activation of PPARα.[7576]