Mini-review: The therapeutic role of cannabinoids in neuroHIV

Abstract

In the era of combined antiretroviral therapy (cART), human immunodeficiency virus type 1 (HIV-1) is considered a chronic disease with an inflammatory component that specifically targets the brain and causes a high prevalence of HIV-1-associated neurocognitive disorders (HAND). The endocannabinoid (eCB) system has attracted interest as a target for treatment of neurodegenerative disorders, due to the potential anti-inflammatory and neuroprotective properties of cannabinoids, including its potential therapeutic use in HIV-1 neuropathogenesis. In this review, we summarize what is currently known about the structural and functional changes of the eCB system under conditions of HAND. This will be followed by summarizing the current clinical and preclinical findings on the effects of cannabis use and cannabinoids in the context of HIV-1 infection, with specifically focusing on viral load, cognition, inflammation, and neuroprotection. Lastly, we present some potential future directions to better understand the involvement of the eCB system and the role that cannabis use and cannabinoids play in neuroHIV.

Graphical Abstract

1. Background

In the context of human immunodeficiency virus type 1 (HIV-1) infection, cannabis use is an important topic and is one of the most commonly used drugs among people living with HIV-1 (PWH). Cannabis use has been reported to be higher in PWH compared to the general population [178], potentially to manage HIV-1 symptoms such as pain, nausea, and appetite loss, despite the negative effects [20, 178, 240, see also Table 1]. Additionally, certain cannabinoids are emerging as therapeutically promising neuroprotective agents in several neurodegenerative diseases, including Parkinson’s disease, Alzheimer’s disease, and multiple sclerosis due to their anti-inflammatory, anti-oxidative, and anti-excitotoxic properties [156, 226]. To enhance our understanding about the role of the endocannabinoid (eCB) system in neuroHIV, the current review focuses on how the eCB system is altered by neuroHIV and how cannabinoids affect HIV-1 infection and specifically HIV-1-associated neurocognitive disorders (HAND).

Table 1:

Human clinical findings

Major effects	Species	HIV pathogen	ART	Target	Ligand	Effect	Reference
Neuroinflammation	Human (postmortem tissue)	HIV, encephalitis	Yes	CB1R CB2R	Anti- CB1R and anti- CB2R antibodies	↑ CB1R in white matter microglia and perivascular cells ↑ CB2R microglia, astrocytes and perivascular macrophages	[52]
	Human	HIV-1	Yes	pDC	Δ9-THC	↓ IFN-α by pDC	[99]
	Human	HIV-1	Yes	T-cell	Δ9-THC	IFN-α ↑ IL-7R-α expression in T cells IFN-α ↑ IL7-induced phosphorylation of STAT5 in CD4+ and CD8+ cells CD3/CD28/IFN-α-induced proliferation was ↑ by IL-7 and ↓ by THC	[98]
	Human and primary leukocytes (in vitro)	HIV-1		CD16 and IP-10 levels	Cannabis	HIV+ users – ↓ circulating CD16 monocytes and plasma IP-10 than HIV- nonusers HIV+ users – no CD16 expression when treated with in vitro IFNa THC treatment of PBMC and purified monocytes ↓ IP-10 levels	[197]
Cognitive performance	Human	HIV-1	Yes	Effect on neurocognitive impairment	Cannabis	Lower likelihood of neurocognitive impairment	[233]
	Human	HIV-1	No info	Effect on cognitive performance, CD4 count and viral loads	Cannabis	HIV+ patients – lower neurocognitive performance than control Moderate-to-heavy HIV+ users – low learning/memory performance than moderate-to-heavy HIV- users HIV+ light users – more verbal fluency than HIV- light users HIV+ cannabis users had lower viral loads and higher CD4 count than non-users	[228]
	Humans	HIV-1	No info	Effect on brain structure and cognitive performance	Cannabis	Heavy users – smaller volumes in the entorhinal cortex and fusiform gyrus HIV+/- smaller thickness of the cingulate HIV- light-users had better cognitive performance than HIV+	[227]
	Humans	HIV-1	Some subjects on ART	Effects on cognition and brain metabolites	Cannabis	No effect on cognition HIV+ non-users – ↓ N-acetyl aspartate in parietal white matter and ↑ choline compound in basal ganglia Cannabis users (HIV+ and HIV-) – ↓ basal ganglia N-acetyl aspartate, choline compound, and glutamate, ↑ thalamic creatine HIV+ cannabis users – ↓ glutamate in frontal white matter	[35]
	Humans	HIV-1	No info	Effects on cognitive function	Cannabis	Frequent users reported more symptoms of depression and anxiety No significant difference effects of marijuana on CD4 levels Impact of marijuana was greater on delayed memory in severe HIV disease No difference in attention, learning or memory due to marijuana use	[58]
Viral load / Immune cells	Human	HIV-1	Yes (Indinavir or Nelfinavir)	HIV-1 RNA levels, CD4+ and CD8+ cells subset, PK analyses of protease inhibitor	Δ9-THC	Does not elevate viral load in patients on stable antiretroviral regimens No effect on CD4+ or CD8+ cell counts No clinical interaction of cannabinoid with protease inhibitors	[1]
	Human	HIV-1	Yes	HIV-1 RNA levels, CD4+ and CD8+ cells	Dronabinol or cannabis	No negative changes No changes in CD4+ and CD8+ cell levels	[29]
	Human	HIV-1	No	HIV-1 RNA viral loads	Cannabis	↓ Plasma HIV-1 RNA viral loads	[158]
	Human	HIV-1	Yes	HIV-1 viral suppression	Cannabis	No viral suppression in daily or less than daily cannabis users	[175]
	Human	HIV-1	Yes	HIV-1 RNA levels in blood and semen	Cannabis	↑ HIV-1 RNA levels in semen	[83]
	Human	HIV-1	Yes	HIV-1 viral load	Cannabis	Daily and nearly daily cannabis users show viral load suppression	[218]
	Human	HIV-1	Yes	Inflammatory immune cell frequency	Δ9-THC	↓ Frequency of HLA-DR+, CD38+, CD4+, and CD8+ cells ↓ Monocytes subset ↓ IL-23 and TNF-α	[146]
	Human	HIV-1	Yes (Azidothymidine and/or Dideoxyinosine	Effects of marinol on HIV-1 progression	Marinol	↓ Clinical indicators, amylase, lipase, ALT and AST (not significant)	[236]
	Human	HIV-1	Yes	Effects of cannabis on inflammatory and circulating monocytes	Cannabis	↓ Inflammatory, nonclassical, activated classical and activated-inflammatory monocytes	[33]
	Human	HIV-1	Yes	Effects of cannabis use on BMI, CD4+ cells and HIV-1 RNA suppression	Cannabis	No changes in BMI and CD4+ cell count Cannabis users had detectable viral loads	[133]
	Human	HIV-1	No info	CD4+ and CD8+ cell counts	Δ9-THC	↑ CD4+ and CD8+	[120]
ART adherence	Human	HIV-1	Yes	ART adherence and HIV-1 symptom	Cannabis	Cannabis dependent group: Had low adherence than non-users and non-dependent users Had higher viral loads Had frequent and severe HIV symptoms/ ART side effects	[26]
	Human	HIV-1	Yes	ART adherence	Cannabis	No relationship between cannabis use and adherence Cannabis use for reducing nausea resulted in ART adherence	[62]
	Human	HIV-1	Yes	ART adherence	Cannabis	Recreational users showed low ART adherence Therapeutic users showed no association with ART adherence	[145]
	Human	HIV-1	Yes	Retention outcomes	Cannabis	Not associated with IOM retention outcome Associated with missing next appointment	[125]
	Human	HIV-1	Yes	ART adherence	Cannabis	Use led to nonadherence	[231]
Appetite and/or Mood	Human	HIV-1	Yes	Effects on caloric intake	Dronabinol, Δ9-THC	↑ Caloric intake Minor effects on cognitive performance	[90]
	Human	HIV-1	No info	AIDS-related anorexia	Dronabinol	↑ Appetite above baseline Mood improvement ↓ Nausea	[16]
	Human	HIV-1	Yes	Effects of caloric intake, mood, and sleep	Dronabinol, Δ9-THC	↑ Caloric intake No cognitive impairment Only Δ9-THC improved sleep	[89]
	Human	HIV-1	No info	Effect on nutritional status	Dronabinol	↑ Percent body fat ↑ Weight gain ↑ Prealbumin ↓ Symptom distress Improved appetite	[222]
	Human	HIV-1	Yes	Effects of high dose	Dronabinol	↑ Food cravings Improved sleep Mood improvement	[18]
	Human	HIV-1	Yes	Effect on appetite hormones	Δ9-THC	↑ Plasma levels of ghrelin, leptin ↓ Plasma levels of PYY No effect on insulin	[196]
	Humans	HIV-1	No info	Over all effects	Δ9-THC	↓ Anxiety and/or depression Improved appetite Pain relief	[189]
	Human	HIV-1	Yes	Long-term effects of dronabinol	Dronabinol	Safe to use for anorexia associated weight loss in patients with AIDS	[17]
	Human	HIV-1	Yes	HIV-1 wasting syndrome with anorexia	Dronabinol Megestrol acetate	Dronabinol alone did not affect weight High dose of megestrol acetate + dronabinol ↑ weight	[229]
	Human	HIV-1	Yes	Effect of appetite and weight gain	Dronabinol	Improves appetite Reverses weight loss	[63]
	Human	HIV-1	Zidovudine in 6 patients	Effects on weight	Dronabinol	↑ Body weight	[86]
	Human	HIV-1	N/A	Effects of HIV-1 symptoms	Δ9-THC	↑ Appetite ↓ Muscle pain ↓ Nausea ↓ Anxiety ↓ Nerve pain ↓ Depression ↓ Paresthesia	[240]
Neuropathic pain	Human	HIV-1 and symptomatic HIV-SN	Yes	Effect of smoked cannabis on HIV-associated neuropathy	Δ9-THC	↓ Chronic neuropathic pain from HIV-associated sensory neuropathy	[2]
	Human	HIV-DSPN	Yes	Effect of smoked cannabis on HIV-associated neuropathy	Δ9-THC	↓ Pain Improved mood and daily functioning	[73]

Abbreviations: AIDS, acquired immunodeficiency syndrome; ALT, alanine transaminase; ART, antiretroviral therapy; AST, aspartate transaminase; BMI, body mass index; CB₁R, cannabinoid type 1 receptor; CB₂R, cannabinoid type 2 receptor; CBR, cannabinoid receptor; Δ⁹-THC, delta-9-tetrahydrocannabinol; HIV-SN, HIV-associated sensory neuropathy; HIV-DSPN, HIV-associated distal sensory predominant polyneuropathy; HLA-DR+, human leukocyte antigen – DR isotope; IFN-α, Interferon alpha; IL-7, interleukin 7; IL-23, interleukin 23; IL-7R-α, IL-7R-α receptor; IOM, Institute of Medicine; IP-10, IFN-γ-inducible protein 10; PBMC, peripheral blood mononuclear cells; pDC, plasmacytoid dendritic cells; PK, pharmacokinetics; PYY, peptide YY; STAT5, signal transducer and activator of transcription 5; TNF-α, tumor necrosis factor alpha

Criteria for exclusion from this Table: (1) Studies on cannabinoids and HIV effects not directly related to the central nervous system. (2) Studies on the effects of cannabinoids on other diseases/disease pathogens.

1.1. HIV-1 associated neurocognitive disorders (HAND)

HIV-1 associated neurocognitive disorders (HAND) was introduced in 2007 [9] and is an umbrella term for a group of neurocognitive disorders that include three subtypes; HIV-associated dementia (HAD), minor neurocognitive disorder (MND), and asymptomatic neurocognitive impairment (ANI) [72]. Before the availability of HIV-1 therapy, more than 15% of infected patients developed the more severe form HAD and autopsy usually revealed pathological and inflammatory changes to the brain, also known as HIV encephalitis (HIVE) [47, 72]. With the introduction of combined antiretroviral therapy (cART), which is very effective in suppressing HIV-1 replication and restoring the immune system [12], HAD has significantly declined (< 5%) and hardly any HIVE cases are reported at autopsy [72, 96]. However, as ART medication does not eradicate the virus, low levels of viral replication and chronic immune activation still linger, specifically in the brain due to low brain penetration of cART [148]. The difficulty of efficient delivery of cART to the central nervous system (CNS) [24, 177, 195] results in the prevalence of the milder forms of HAND to remain high. Up to 50% of cART treated PWH exhibit MND or ANI that can interfere with daily life [9, 69, 72, 80, 96, 206], involving problems in executive function, memory consolidation, decision-making, attention [60, 82, 91, 96, 214], and/or mood [25, 149, 176].

The brain mechanisms underlying HAND involve two pathways, including HIV-1 induced neuroinflammation within the brain that indirectly affect neuronal health and continued production of neurotoxic HIV-1 proteins that can target neurons directly.

Chronic neuroinflammation within the brain appears to predominate and significantly contribute to the onset of HIV-1 associated neuronal injury and thus, HAND [81, 92, 118]. Shortly after infection, HIV-1 can enter the brain within infected macrophages, monocytes, and T cells [75, 103, 116, 212, 238] and as cell-free virus that establish central reservoirs by infecting microglia, brain endothelial cells or astrocytes [8, 28, 44, 95, 127, 131, 134]. As the virus itself is not able to infect neurons the release and production of neurotoxic factors such as inflammatory mediators from HIV-1 infected cells contribute indirectly to neuronal dysfunction and injury [3, 44, 84, 104]. HIV-1 has been demonstrated to cause neurotoxicity by stimulating the production of proinflammatory cytokines and chemokines in the brain, inducing the release of TNF-α, RANTES/CCL5, and MCP-1/CCL2 from infected microglia and macrophages [71, 217, 235] and IL-8, IL-1ß, and TNF-α from infected astrocytes [37].

Additionally, HIV-1 contributes to HAND through the continued production of neurotoxic HIV-1 proteins from cellular reservoirs within the CNS that can target neurons directly [76, 77, 137, 155, 188]. HIV-1 proteins, such as the transactivator of transcription (Tat) and the envelope glycoprotein 120 (gp120) are likely agents of the observed neuronal loss in PWH and have been measured in the CNS of PWH under cART [97, 115, 153]. Besides their indirect effects on neurons via actions on microglia and astrocytes [38, 70, 119, 123, 130, 136], Tat and gp120 have direct effects on neurons by activating glutamatergic NMDA receptors [76, 77, 94, 142, 151], altering chemokine receptor signaling [gp120; 101, 154, 155], and interacting with the lipoprotein receptor-related protein [Tat; 137]. The effects of Tat on NMDA receptors in neuronal cultures has been demonstrated to potentiate glutamate-induced excitotoxicity [139], leading to increases in neuronal intracellular calcium levels, dendritic damage, and synapse loss [76, 94, 151, 215]. Similarly, blocking chemokine receptor signaling has been shown to prevent gp120-induced neuronal apoptosis in the absence of non-neuronal cells [155]. Thus, even though HIV-1 itself is not able to infect neurons, the release of HIV-1 proteins from HIV-1 infected cells can contribute to neurotoxicity via direct mechanisms on neurons.

The persistence of HAND in the era of cART has raised questions about the causes and treatment of HIV-1-related brain disorders and about the extent to which HAND and its underlying structural changes are reversible.

1.2. The endocannabinoid (eCB) system

As HAND is a group of neurodegenerative cognitive disorders with an inflammatory component, the eCB system, which regulates both cognition and immune function, presents a promising therapeutic target for treating the consequences of HIV-1 infection on the CNS.

The eCB system, constituted of endogenous cannabinoids (‘endocannabinoids’), cannabinoid receptors, and enzymes which synthesize or degrade cannabinoids, has attracted interest as a target for treatment of neurodegenerative disorders, due to the potential anti-inflammatory and neuroprotective properties of cannabinoids. A schematic presentation of the eCB system in a healthy individual is depicted in Figure 1A. The two main eCB ligands are N-arachidonoylethanolamine (AEA, also known as anandamide) and 2-arachidonoylglycerol (2-AG). These ligands as well as exogenous cannabinoids, such as delta-9-tetrahydrocannabinol (Δ⁹-THC) found in cannabis, act predominantly via cannabinoid type 1 and/or cannabinoid type 2 receptors (CB₁R and CB₂R, respectively), but can also activate the transient receptor potential vanilloid (TRPV) ion channels [102, 117], peroxisome proliferator-activated receptors (PPARs) [174, 224], and/or other G-protein-coupled receptors, including GPR55 and GPR18 [42, 50, 93, 202].

Figure 1.

A schematic presentation of the endocannabinoid (eCB) system. (A) Represents the eCB system in a healthy individual. CB₁Rs are present on presynaptic neurons. The influx of Ca²⁺ into the presynaptic neuron causes release of glutamate in the synapse and interacts with postsynaptic receptors (i.e. NMDR, mGluR). Excess of glutamate is taken up by the glutamate transporter present on astrocytes. CB₁R agonists block the release of glutamate and decrease excitotoxicity. CB₂Rs are predominantly expressed on microglia and their activation by CB₂R agonists decreases neuroinflammation by blocking the production of proinflammatory cytokines. It has been shown that neurons, but also glial cells, produce eCBs AEA and 2-AG which are hydrolyzed by the enzymes MAGL and FAAH, respectively. (B) Shows the possible mechanism of action of Tat and its effects on the eCB system in PWH. (1) Tat causes an excess of glutamate release into the synapse and (2) abnormally increases Ca²⁺ influx by acting on the NMDR, VGCC, and TRPC. (3) Tat blocks the glutamate transporter which further increases glutamate concentration in the synapse. TRPC channel is a non-selective cation channel that is also permeable for Ca²⁺. (4) Tat increases the IP₃ concentration which activates the TRPC and leads to Ca²⁺. Additionally, the increased IP₃ levels cause intracellular Ca²⁺ release from the ER. This excess of intracellular Ca²⁺ concentration causes Ca²⁺ overload, Ca²⁺ dysregulation, oxidative stress which leads to neuronal injury and eventually neuronal death. Tat also leads to the upregulation of CB₁Rs and CB₂Rs. Even though the CB₁R is upregulated in the presynaptic neuron, it is currently debated whether its inhibitory function is impaired or enhanced (represented by the broken inhibitor line). The negative effects of Tat are counteracted at the microglia as the overexpression of CB₂R blocks proinflammatory cytokines more effectively. Whether the endogenous ligands AEA and 2-AG are affected by Tat and their levels are upregulated in the brain of PWH is still not known (represented by ?). Abbreviations; 2-AG, 2-arachidonoylglycerol; AA, arachidonic acid; AEA, arachidonoyl ethanolamine (anandamide); Ca²⁺, calcium; DAG, diacylglycerol; DGL, diacylglycerol lipase; eCB, endocannabinoid; EA, ethanolamine; ER, endoplasmic reticulum; FAAH, fatty acid amide hydrolase; IP₃, inositol triphosphate; mGluR, metabotropic glutamate receptor; MAGL, monoacylglycerol lipase; NAPE-PLD, N-arachidonoyl phosphatidylethanolamine phospholipase D; NAPE, N-arachidonoyl phosphatidylethanolamine; NMDR, N-methyl-D-aspartate receptor; PE, phosphatidylethanolamine; PIP₂, phosphatidylinositol 4,5-bisphosphate; PLC, phospholipase C; PWH, people living with HIV; Tat, transactivator of transcription; TRPC, transient receptor potential cation channel; VGCC, voltage-gated calcium channel. Created with BioRender.com.

The CB₁R is the most abundant G-protein-coupled receptor in the CNS, mainly expressed on neurons [150] and is responsible for the psychoactive effects of Δ⁹-THC, which is a primary compound of cannabis [65, 220]. CB₁R agonists have demonstrated promising protective effects, such as inhibiting excitotoxic neurotransmission by blunting presynaptic glutamate release [43, 107, 147] and decreasing intracellular calcium [173]. However, therapeutic use of direct CB₁R agonists is limited due to the psychoactive side effects associated with activation of CB₁Rs, including sensorimotor, affective and cognitive disturbances [65, 162, 220].

In turn, CB₂Rs are predominantly expressed by cells of the immune system [30, 163, 225] but can also be found in the CNS on immune-activated glia [163, 221]. CB₂Rs represent a promising therapeutic target as their activation has been shown to induce anti-inflammatory signaling in astrocytes [219], regulate microglial migration and cytokine production [4, 11, 144], and reduce oxidative stress and apoptosis in neurons [232].

Another line of research has focused on the development of drugs targeting enzymes regulating the biosynthesis and degradation of AEA and 2-AG [6, 135, 187]. Because eCBs are neuromodulators that are synthetized locally on demand, the inhibition of their degradation is a therapeutic strategy that will cause their elevation only in locations where they are being actively produced to evoke their local neuroprotective effects, e.g. at the site of injury. Thus, in contrast to CB₁R or CB₂R agonists that are associated with side effects resulting from lack of site specificity and affecting receptors in the entire body, eCB catabolic enzyme inhibitors have high therapeutic potential as they are targeting ‘on site’ produced eCBs and inhibit eCB degradation [67, 184]. There is strong preclinical evidence that selective inhibitors of the main AEA-metabolizing enzyme, fatty acid amide hydrolase (FAAH), and of the main 2-AG enzyme, monoacylglycerol lipase (MAGL) can ameliorate the unwanted effects in a variety of different laboratory animal models of neurodegenerative diseases [165, 184]. Hydrolytic enzyme inhibitors of AEA (i.e. AM5206) and 2-AG (i.e. URB602, JZL184) have demonstrated to produce neuroprotective effects in vitro [41, 165, 209] and in vivo [124, 165, 166]. Additionally, the new generation of hydrolytic enzyme inhibitors, such as FAAH inhibitor PF3845 and MAGL inhibitors MJN110 and Abx-1431 show highly improved selectivity, potency and produce less side effects compared to previously available compounds [5, 27, 46, 64, 110, 111, 114, 171, 182].

Overall, targeting the eCB system appears to present a promising strategy to alleviate inflammatory and neurodegenerative consequences of HIV-1 infection on the CNS, which is reviewed in detail below.

2. The endocannabinoid (eCB) system in HAND

Most of what is known about the effects of HIV-1 infection on the eCB system is derived from protein expression studies for cannabinoid receptors, eCB ligands, and their enzymes, but little is known about the extent to which HIV-1 might disrupt their function. A schematic presentation of how the eCB system is potentially altered by HIV-1 Tat in PWH is depicted in Figure 1B.

Effects on CB₁R expression in the context of neuroHIV have been variable, with reports ranging from no effects to upregulating effects. Brain tissue analysis from frontal cortex of simian immunodeficiency virus (SIV) rhesus macaques [21] or whole brain samples from HIV humanized mice [85] demonstrated no alterations in CB₁R protein or mRNA expression levels, respectively. However, when assessing cell-type specific changes, CB₁R upregulation levels have been reported [52, 112, 239]. CB₁R upregulation was noted in perivascular cells and white matter macrophages in brains of PWH with HIVE [52]. Further, CB₁R levels were reported to be increased in neurons in the infralimibic region in a HIV-1 Tat transgenic mouse model and was shown to be associated with behavioral deficits in an inhibitory control task [112]. The upregulation of CB₁R expression levels upon Tat exposure was also supported in mouse primary prefrontal cortex neuronal cultures in vitro, which demonstrated a time-course dependent linear increase of CB₁R protein expression over a 24 h time period [244]. Whether the upregulation of CB₁R expression levels observed in the female Tat(+) mice is a compensatory response to the Tat-induced observed behavioral deficits or contributes to the seen deficits in the behavioral Go-No-Go task needs to be further investigated [112]. Modified expression of CB₁R levels in other diseases has been negatively correlated with the prognosis of the symptoms [157]. For example, in neuropathic pain and multiple sclerosis, increases in CB₁R expression is associated with reductions of symptoms and/or dampened disease progression, suggesting a neuroprotective role [185], which is also confirmed in psychiatric disorders [169].

Most of the findings for changes of CB₂R expression levels provide evidence for increased expression levels of CB₂Rs in the context of neuroHIV [21, 52, 193]. Clinical postmortem studies demonstrated CB₂Rs upregulation in white matter microglia, astrocytes, and perivascular macrophages [52, 193], which was specifically high in HIVE tissue and differed from HIV+ brains without HIVE [52]. This finding was confirmed in brain cortical tissue of SIV-infected rhesus macaques, demonstrating cell-type specific upregulation of CB₂Rs in perivascular monocytes/macrophages and microglia [21]. The increase of CB₂R expression levels in HIV-1 is suggested to be due to the inducible nature of CB₂R upon microglial cell activation under pathological conditions [10, 15, 23, 49, 157] and has been associated with an anti-inflammatory function in various disease models [10, 113, 179]. Interestingly, the upregulation of CB₂R expression appears to be specific to inflammation-driven neurodegeneration [48], as substantially more pronounced increases in CB₂R expression levels have been noted in response to HIV-1 infection or other bacterial/viral inflammatory mediators compared to direct neurotoxins that cause neuronal injury from within the cell, e.g. via oxidative stress [23, 49].

Much less is known about changes of protein expression levels of eCB ligands and their degrading enzymes in the context of neuroHIV. Fatty acid amide hydrolase (FAAH) was found to be overexpressed in perivascular astrocytes and astrocytic processes of cortical SIV tissue samples [21]. It has been previously shown that the FAAH protein is selectively overexpressed in neuritic plaque-associated astrocytes in Alzheimer’s disease brains [22]. FAAH upregulation in astrocytes appears to contribute to proinflammatory effects, as they are involved in converting AEA to arachidonic acid [56, 152], thus, providing a potential source of inflammatory processes.

Further, it is not known if endogenous ligands such as AEA or 2-AG are upregulated in the brain of PWH. A recent in vitro study, in primary prefrontal cortex neuronal cultures was unable to demonstrate a significant upregulation of AEA upon Tat exposure [100] but more clinical and preclinical studies are necessary to assess whether HIV-1 significantly alters eCB levels in the CNS.

Lastly, hardly any studies have investigated the effects of HIV-1 on eCB’s regulatory function. It is not quite clear whether eCB signaling is diminished or enhanced in the context of neuroHIV. A recent study reported Tat-induced reduction of the inhibiting effects exerted on glutamatergic neurotransmission by cannabinoids in hippocampal cultures, including Δ⁹-THC and 2-AG, due to impaired CB₁R-mediated presynaptic inhibition of glutamate release [241]. In turn, another study has shown that the downregulating effects of eCBs on glutamatergic transmission are enhanced in the context of neuroHIV [112]. Upregulating AEA via the FAAH enzyme inhibitor PF3845 demonstrated increased inhibition of glutamate release in prefrontal cortex brain slices of male Tat(+) transgenic mice compared to their control counterparts [112]. The discrepancy between results may be due to brain region-specific differences as the prefrontal cortex and hippocampus have been shown to display differential sensitivities to Tat [45].

Overall, additional studies are necessary to understand the effects of HIV-1 on the eCB system in more detail, specifically eCB signaling, and the underlying mechanisms involved.

3. Clinical and preclinical evidence of therapeutic properties of cannabinoids in HAND

3.1. Effects of cannabinoids on viral load

The effects of cannabinoids on viral load, HIV-1 replication, and CD4⁺ cell count have been assessed over the last years [1, 29, 83, 133, 160, 175, 186, 193, 198, 203, 228, 239, see also Tables 1 and 2].

Table 2:

Preclinical animal studies (in vivo)

Major Effects	Species	HIV Pathogen	ART	Target	Ligand	Effect	Receptor Involved	Reference
Neuronal activity	Mice	HIV-1IIIB Tat1–86	No		PF3845	Tat(+) female mice – inhibitory control deficits, ↑ CB1R in infralimbic cortex Negative correlation between inhibitory control and infralimbic CB1R expression ↑ sEPSC in Tat(+) mice PF3845 ↓ sEPSC	CB1R	[112]
Neuroinflammation and Immune cells	Rhesus macaques	SIVmac251, encephalitis	No info	CB1R, CB2R, FAAH	Anti- CB1R & anti- CB2R antibodies	↑ CB2R microglia, perivascular macrophages and T-lymphocytes ↑ FAAH in perivascular astrocytes and astrocytic processes	CB1R, CB2R	[21]
	Mouse	pVRCgp120	No info	Immune cells	Δ9-THC	↑ or ↓ gp120 specific T cell responses depending on magnitude of IFN-γ response	No	[40]
	Mouse	pVRCgp120	No info	Immune cells	Δ9-THC	↑ gp120 specific INF-γ and IL-2 response with gp120 derived peptide 81 Δ9-THC ↑ gp120-specific T cell activation in WT but not CB1-/- and CB2-/- mice	CB1R, CB2R	[39]
	Rhesus macaques	SIV	No info	CD4+ and CD8+ T lymphocytes	Δ9-THC	Chronic administration: No difference in lymphocyte subtypes, proliferation or apoptosis ↑ T lymphocyte CXCR4 expression of both CD4+ and CD8+ cells	No	[132]
	Rhesus macaques	SIVmac251	No info	miR	Δ9-THC	No differences in plasma viral loads ↑ Striatal BDNF ↓ TNF-α mRNA expression in THC/SIV group miRs modulation	No	[216]
Neurogenesis and Neuroinflammation	Mouse	GFAP/ gp120	No	Deletion of FAAH gene	None	↑ Neurogenesis by ↑ expression of COX-2 and PGE-2 ↓ Astrogliosis	No	[13]
	Mouse	GFAP/ gp120	No	CB2R	AM1241	↑ Neurogenesis in hippocampus ↓ Astrogliosis and gliogenesis	CB2R	[14]
Viral load and disease progression	Rhesus macaques	SIVmac251	No	HIV-1 RNA levels CD4+ and CD8+ cells	Δ9-THC	No effect on disease progression, morbidity, and mortality ↓ Plasma SIV-RNA viral load and lengthened survival ↓ Classic markers of SIV disease	No	[161]
	Rhesus macaques	SIVmac251	No	Effect of chronic Δ9-THC on viral load and inflammation	Δ9-THC	In lymph nodes and spleen: ↓ Viral replication ↓ Viral gag RNA ↓ INF-γ and IL-6	No	[160]
	Rhesus macaques	SIVmac251	No	Effect of chronic Δ9-THC on plasma viral load	Δ9-THC	Tolerance to disruptive effects of Δ9-THC ↓ CB1R and CB2R levels in the hippocampus No effect on viral load in the plasma, CSF or brain tissue ↓ Neuropathology and opportunistic infections Lower expression of inflammatory cytokine MCP-1	No	[239]
	Mice (huPBL-SCID)	HIV-1NL4–3	No	Effect of Δ9-THC on HIV-1 progression	Δ9-THC	↑ HIV-infected peripheral blood leukocytes 50-fold ↑ viral load Upregulation of CCR5 and CXCR4	No	[203]
	Mice (huPBL/HIVE)	HIV-1ADA	No	Effect of CB2R agonist	Gp1a	↑ CB2R expression in perivascular microglial cells and lymphocytes Gp1a ↓ infiltration of human cells in the mouse brain and HLA DQ activation Gp1a ↓ CCR5 expression on human cells in spleen ↑ Fas ligand expression	CB2R	[85]
	Rhesus macaques	SIVmac251	No	Viral load, CD4+ and CD8+ cells, IgE+ B cells	Δ9-THC	No difference in plasma viral load ↓ CD4+/CD8+ ratio ↓ IgE+ B cells	No	[234]
BBB impairment	HBMEC and human astrocyte cocultures (in vitro), mouse (in vivo)	gp120MN	No	TJ ZO-1, Claudin-5 expression	CP55,940, ACEA, URB597	In vitro – CP55,940 and ACEA prevented BBB permeability and prevented ZO-1 and claudin-5 downregulation in HBMEC In vivo – ACAE inhibited BBB permeability and prevented ZO-1 and claudin downregulation	CB1R	[140]
Nociception	Rat	gp120MN	No	FAAH	URB597, PF3845	↓ Nociception in rat HIV neuropathy model ↓ Cold and tactile allodynia	CB1R, CB2R	[167]
	Rat	gp120IIIB	No info	None	WIN55,212–2, AMD3100	↓ Analgesic effectiveness AMD3100 restores the analgesic effects of WIN55,212–2	No	[180]

Abbreviations: BBB, blood-brain barrier; BDNF, brain derived neurotropic factor; CB₁R, cannabinoid type 1 receptor; CB₂R, cannabinoid type 2 receptor; CBR, cannabinoid receptor; COX-2, cyclooxygenase-2; CSF, cerebrospinal fluid; Δ⁹-THC, delta-9-tetrahydrocannabinol; FAAH, fatty acid amide hydrolase; GFAP, glial fibrillary acidic protein; HBMEC, human brain microvascular endothelial cells; IFN-γ, Interferon gamma; IgE, immunoglobulin E; IL-2, interleukin 2; ; IL-6, interleukin 6; miR, microRNA; PGE-2, prostaglandin E2; SIV, simian immunodeficiency virus; sEPSC, spontaneous excitatory postsynaptic current; TJ ZO-1, tight junction zonula occludens-1; TNF-α, tumor necrosis factor alpha; WT, wild-type; ZO-1, zonula occludens-1

Criteria for exclusion from this Table: (1) Studies on cannabinoids and HIV effects not directly related to the central nervous system. (2) Studies on the effects of cannabinoids on other diseases/disease pathogens.

Enhancing effects of cannabis use on HIV-1 replication and viral load are more observational in nature [26, 83] and have been reported to be potentially due to low ART adherence [133], which has been demonstrated in a number of studies [61, 141]; even though others have found no association with cannabis use [201] or even enhanced ART adherence if cannabis use was associated with medical use [62, 145].

On the other hand, previous short-term randomized placebo-controlled studies reported no effects of cannabis, smoked or taken orally, on viral load, CD4⁺, and/or CD8⁺ cell counts in PWH [1, 29], which has also been confirmed in some studies for recreational cannabis use in PWH [36, 175, 236].

Nevertheless, the vast majority of literature reports a downregulation of viral load and HIV-1 replication by cannabinoids, including clinical studies [158, 228], and preclinical studies in vivo [160] and in vitro [186, 193, 198]. Clinical studies have found anti-viral effects of cannabinoids, with PWH cannabis users demonstrating lower viral load and higher CD4⁺ counts than PWH non-users [228]. This has also been confirmed in PWH injection drug users that use cannabis at high intensities [158] or daily and near-daily use [218], as well as in an underrepresented group of black PWH [120]. Preclinical studies in SIV-infected rhesus macaques further support clinical findings [160]. Prolonged period of Δ⁹-THC treatment prior to and during SIV infection resulted in lower viral load in lymph nodes and spleen as well as decreased inflammation (i.e. lower INF-γ and IL-6 protein in lymph nodes and spleen) [160, 239].

Direct effects of cannabinoids on viral replication have been demonstrated in vitro by the reduced numbers of SIV-infected cells in culture incubated with Δ⁹-THC prior to infection [161]. Interestingly, various preclinical in vitro studies have found CB₂Rs to be involved in inhibiting viral expression [186, 193, 198]. Specifically it has been shown that CB₂R activation with the agonists JWH144 and O-1966 decreased viral replication by affecting the long term repeat (LTR) of HIV-1 [193], potentially due to the CB₂R-induced inhibition of cellular transcription factors involved in the transactivation of the HIV-1 LTR [193].

Another possible mechanism for suppression of viral replication includes the interaction of cannabinoids with HIV-1 coreceptors, CXCR4 and CCR5 [53, 85, 198], that are upregulated upon HIV-1 infection [132, 203]. Whereas, CB₂R agonists JWH144 and O-1966 did not alter surface protein or gene expression of CXCR4 or CCR5 in HIV-1 infected monocyte-derived macrophages [193], other studies have shown an inhibitory effect of Δ⁹-THC on CCR5 and CXCR4 in HIV-1 infected monocyte-derived macrophages [237], WIN55,212–2 inhibitory effects on CCR5 in microglia [198], and CB₂R agonist inhibitory effects in CD4⁺ T cells on CCR5 [85] or CXCR4 [53]. Specifically, it has been suggested that CB₂R activation in CD4⁺ T cells can inhibit actin reorganization with decreasing F-actin levels, which is an important contributor to productive HIV-1 infection [53].

Overall, findings indicate that the activation of CB₂Rs appear to potentially downregulate HIV-1 infection, with proposed mechanisms including CB₂R-induced inhibition of cellular transcription factors involved in the transactivation of the HIV-1 LTR or CB₂R cross-regulation of HIV-1 coreceptors, CXCR4 and CCR5, via inhibitory crosstalk.

3.2. Cannabinoid effects on cognition

A number of clinical studies have assessed cannabinoid effects in PWH. Specifically, cannabis use has been investigated and found to affect various biological processes, including cognitive performance [58, 228, 233], appetite [63, 90, 222], mood [16, 18], and neuropathic pain [2, 73, 167]. More detailed information about these studies is provided in Table 1, with this subsection specifically focusing on cannabis effects on cognition. It is known that cannabis use, in general, has negative (psychoactive) effects on memory and executive functions in healthy individuals [57, 194]; however, its effect in PWH remains unclear and divided.

The relationship between HIV-1 infection and cannabis use and their interactive effects on neurocognitive functioning is complex and various variables have been considered, including disease progression, the amount of cannabis use, and the cognitive domain assessed.

Some studies show that the effects of cannabis use in PWH depend upon disease stage [35, 58]. Cannabis use resulted in cognitive dysfunction in PWH with an advanced stage of infection (symptomatic phase) but no effects were noted for HIV negative cannabis users or asymptomatic PWH cannabis users [58]. Additionally, frequency of cannabis use was associated with greater cognitive impairment among symptomatic PWH, which appeared to be primarily related to performance on memory tasks [58]. The lack of synergistic or interactive effects of HIV-1 and cannabis use on cognitive performance for the asymptomatic stage was confirmed in another study, which did not observe any significant abnormalities for neuroasymptomatic PWH cannabis users compared to PWH non-users on standard neuropsychological tests when correcting for group differences in age, education, or depression scores [35]. Since both, HIV-1 infection and cannabis use, have effects on the immune system, there is a likelihood that their interactions are specifically exacerbated in PWH with greater immune suppression.

Other studies have demonstrated that the amount of cannabis use determines what type of effects cannabis exhibits on cognition in PWH [26, 227, 228]. For example, significant negative effects for HIV-1 symptoms and ART side effects have been reported for cannabis dependent use, whereas no differences were noted between non-cannabis use and non-dependent cannabis use [26]. Similarly, a recent study found that independent of HIV-1 status moderate-to-heavy users consistently performed worse than light users on cognitive functioning tasks (global score) [228]. In a follow up study, the finding that the amount of cannabis use has significant effects on global cognition in PWH was not confirmed and based on their results the authors concluded that cannabis use beyond 1.43 g/week had more adverse effects on neurocognitive performance in HIV negative users compared to PWH users [227]. However, caution should be exercised with this conclusion as a shortcoming of the study was that HIV negative non-users started out with significant higher cognitive performance compared to PWH non-users but then were significantly negatively affected by higher cannabis use (>1.43 g/week), which was not seen for PWH cannabis users [227]. Thus, baseline differences might have contributed to the seen adverse effects of high cannabis use in HIV negative individuals compared to PWH.

In addition to HIV-1 disease stage and the amount of cannabis use, the interactive effects of HIV-1 and cannabis also seem to depend on the cognitive domain assessed. Whereas moderate-to-heavy cannabis use significantly worsened overall cognitive performance independent of HIV-1 status (i.e. HIV negative individuals and PWH) PWH were most severely affected by moderate-to-heavy use for the learning and memory domain, displaying significantly lower scores in learning and memory performance compared to all other comparison groups [228]. It is known that the psychoactive effects of cannabis can be attributed to Δ⁹-THC, which is known to negatively impact memory function [87, 213] specifically at higher doses [109]. On the other hand, the study further reported that when the effects of light cannabis use were compared in PWH and HIV negative healthy controls, PWH light users outperformed HIV negative light users on verbal fluency [228]. This is an important finding as it indicates that depending on amount of cannabis use and the cognitive domain assessed, cannabis can have detrimental or beneficial effects on disease progression. The beneficial effects of cannabis use on verbal fluency find support in a recent study, that demonstrated cannabis use, defined as history of cannabis substance use disorder and cannabis use in the past year, lowered odds of neurocognitive impairment in PWH regardless of age and viral levels or disease stage, with PWH cannabis use being specifically associated with higher performance in verbal fluency [233]. The differential cannabinoid receptor and/or ligand distribution across brain regions could contribute to the distinct effects seen for cannabis use depending on the cognitive domain assessed [79, 181, 210].

Overall, based on the studies outlined above interactive effects of cannabis use and HIV-1 infection on cognition appear to be specifically seen in PWH with more advanced symptomatic stages of HIV-1 infection, and further seem to depend on the amount of cannabis use and the cognitive domain assessed; with high cannabis use exhibiting more negative effects, specifically in the learning and memory domain, whereas light cannabis use might have some beneficial effects for domains such as verbal fluency. However, additional research is needed to investigate more in detail what role different cannabis types play on cognition in PWH, including cannabis use with high versus low Δ⁹-THC content, as well as the components of cannabis that are non-psychoactive, including cannabinol or cannabidiol.

3.3. Anti-inflammatory properties of cannabinoids with neuroprotective benefits

The anti-inflammatory and immune modulatory properties of cannabinoids are well known [7, 59, 126, 164, 170] and have been reviewed in the context of HIV-1 infection [19, 54, 106, 242, see also Tables 1–3]. Cannabis use has been reported to lower inflammatory responses in HIV-1 infection from immune cells [33, 146, 197] and the peripheral nervous system in PWH and animal models [34, 160, 239], with this section focusing on the effects of cannabinoids on neuroinflammatory processes and its consequence on neuronal health.

Table 3.

Preclinical in vitro findings

Major Effects	Species	Sample	HIV Pathogen	ART	Target	Ligand / Antibody	Effect	Receptor Involved	Reference
Neuronal activity	Rat	Hippocampal neurons	Tat1–86 (clade B)	No	Effects of Tat on eCB system	WIN55,212–2, 2-AG, JZL184, Δ9-THC	Tat ↓ DSE 2-AG ↓ EPSCs JZL184 did not affect 2-AG mediated EPSCs WIN55,212–2 did not affect EPSC Δ9-THC ↓ EPSC	CB1R	[241]
Neuronal damage and neuroinflammation	Mouse	PFC neuronal cultures	Tat1–86 (clade B)	No	FAAH	PF3845	↑ Neuronal survival ↓ [Ca2+]i ↑ Dendritic volume ↑ AEA, PEA	CB1R	[100]
	Mouse	PFC neuronal cultures	Tat1–86 (clade B)	No	Neurons	AEA, 2-AG	↑ Neuronal survival ↓ [Ca2+]i ↓ Dendritic injury ↓ sEPSCs	CB1R	[244]
	Human	Mesencephalic neuronal/glial culture	gp120LAV/IIIB (clade B)	No	Dopaminergic neurons	WIN55,212–2	↓ Neuronal damage ↓ Microglial damage ↓ Superoxide production ↓ Chemokine and cytokine production	CB2R	[105]
	Human	Primary Müller cell culture	Tat (clade B)	No	Müller glia	AEA, 2-AG	AEA and 2-AG: Suppress Müller cell activation by ↓ inflammatory cytokines Control Tat-induced proinflammatory cytokines through MAPK phosphorylation Inhibit NF-KB signalosome AEA induces MKP- independent of MEK necessary for ↑ anti-inflammatory and ↓ pro-inflammatory cytokines	CB1R, CB2R	[129]
	Human	Primary Müller cell culture	Tat (clade B) and Tat (clade C)	No	Müller glia	AEA	HIV-1 clade Tat B and C act differently Tat B suppresses through MKP-1 and Tat C through MEK-1 ↑ PBMC attachment	CB1R, CB2R	[128]
GABA	Mouse	Mouse brain slices	Tat1–86 (clade B)	No	GABA	WIN55,212–2, AEA	AEA ↓ GABAergic neurotransmission (mIPSCs) in PFC	CB1R	[243]
Cyto / Neurotoxicity	Rat	C6 glioma cells	Tat1–86	No	NO synthase	WIN55,212–2, AEA	↓ Cytotoxicity	CB1R	[74]
	Human and murine	Human and murine NPCs	gp120IIIB (X4 strain) and gp120Ba-L (R5 strain)	No	CB2R	AM1241	↓ Neurotoxicity and apoptosis ↑ Differentiation of NPCs to neuronal cells ↑ Neurogenesis in vivo	CB2R	[14]
Synapse loss and neuroinflammation	Rat	Primary neuronal cultures	gp120IIIB	No	MAGL	JZL184	↓ Synapse loss ↓ Prostaglandins signaling Blocks potentiation of NMDARs	CB2R	[245]
	Rat	Hippocampal neuronal culture	gp120IIIB	No	Synapse	WIN55,212–2	Inhibits synapse loss Blocks IL-1β release in microglia	CB2R	[122]
Cell migratory and/or adhesion response	Human	Leukemic monocyte lymphoma cell line	Tat1–86	No	Migration of U937 towards Tat	Δ9-THC, CP55,940, O-2137	↓ Migration of U937 microphage-like cells towards Tat	CB2R	[191]
	Human	Leukemic monocyte lymphoma cell line	Tat1–86		Tat enhanced monocyte-like cell adhesion	Δ9-THC, CP55,940	↓ Attachment of U937 cells to ECM proteins by altering b-integrin expression and distribution of polymerized actin	CB2R	[192]
	Mouse	BV-2 microglial-like cells	Tat1–86	No	Migration of BV-2	Δ9-THC, CP55,940, 2-AG,	↓ Migration of BV-2 cells towards Tat	CB2R	[78]
Inhibition of viral expression	Human	Microglial culture	None	No	HIV-1	WIN55,212–2	↓ HIV-1 viral expression	CB2R	[198]
	Human	Primary monocytes	None	No	HIV-1	JWH133, GP1a, O-1966	↓ Activity of HIV-1 LTR Partially ↓ expression of HIV-1 pol	CB2R	[193]
	Human	HIV-1 infected CD4+ lymphocyte and microglial cultures	None	No	HIV-1	WIN55,212–2	↓ Viral expression in both CD4+ lymphocyte and microglial cultures		[186]
HIV infection	Human	CD4+ T cells	HIVNL-GI	No	CD4+ T cells	JWH-133, JWH-150, AM630	CB2R activation in CD4+ cells inhibit actin reorganization which prevents infection of CXCR4-tropic HIV-1 in CD4+ T cells	CB2R	[53]
	Human	Primary human monocyte cell lines	None	No	HIV-1	Δ9-THC	↓ HIV-1 infection of macrophages ↓ Cell surface receptors CD4, CCR5, and CXCR4 which ↓ viral entry	CB2R	[237]
	Human	MT-2 cells	None	No	Syncytia formation	CP-55,940, Δ9-THC, WIN-55,212,2, WIN-55,212,3	Cannabimimetic drugs ↑ HIV-1 infection	CB1R, CB2R	[172]

Abbreviations: AEA, N-arachidonoylethanolamine; 2-AG, 2-Arachidonoylglycerol; ART, antiretroviral therapy; [Ca²⁺]_i, intracellular calcium concentration; CB₁, cannabinoid type 1 receptor; CB₂, cannabinoid type 2 receptor; CBR, cannabinoid receptor; Δ⁹-THC, delta-9-tetrahydrocannabinol; DSE, depolarization-induced suppression of excitation; ECM, extracellular matrix; eCB, endocannabinoid system; EPSC, excitatory postsynaptic currents; FAAH, fatty acid amide hydrolase; GABA, gamma aminobutyric acid; LTR, long terminal repeat; MAGL, monoacylglycerol lipase; MAPK, mitogen-activated protein kinase; MEK, mitogen-activated protein kinase kinase; MKP-1, mitogen-activated protein kinase phosphatase-1; NF-KB, nuclear factor kappa-light-chain-enhancer of activated B cells; NMDAR, N-Methyl-D-aspartic acid receptor; NO, nitric oxide; NPCs, neuronal progenitor cells; mIPSC, miniature inhibitory postsynaptic current; PBMC, peripheral blood mononuclear cells; PEA, palmitoylethanolamide; PFC, prefrontal cortex; sEPSC, spontaneous excitatory postsynaptic current

Criteria for exclusion from this Table: (1) Studies on cannabinoids and HIV effects not directly related to the central nervous system. (2) Studies on the effects of cannabinoids on other diseases/disease pathogens.

Various studies in the context of HIV-1 have demonstrated that the anti-inflammatory effects of cannabinoids in the CNS are related to the specific activation of CB₂R [14, 32, 53, 105] which also has been reviewed previously [183, 190]. A preclinical mouse study reported downregulation of astrogliosis and gliogenesis in the hippocampus of GFAP/Gp120 transgenic mice when treated with the CB₂R agonist AM1241 which was further accompanied by improved neurogenesis in the hippocampus [14]. In vitro studies demonstrated that cannabinoids, such as Δ ⁹-THC, CP55,940, and 2-AG, can inhibit the migration of mouse BV-2 microglial-like cells to viral products, such as HIV-1 Tat, which was linked functionally to the CB₂R potentially due to reducing CCR3 levels and altering its intracellular compartmentation [78]. Interestingly, recent studies have shown that CB₂R and chemokine CXCR4 are able to form heterodimers and display negative-crosstalk and cross-antagonism, thus, resulting in decreased CXCR4-mediated cell migration, invasion, and adhesion [168, 211].

The downregulation of inflammatory responses derived from microglia by cannabinoids is specifically relevant as chronic activation of brain microglia is a major contributing factor for HIV-associated brain disease [159, 199]. Using a culture model from human primary microglia, WIN55,212–2 inhibited the migration of gp120-activated microglia, thus suppressing the toxic activity of gp120 on the CNS [105]. Further, CB₂R activation inhibited gp120-induced superoxide production in purified human microglial cells and reduced gp120-induced production of chemokine and cytokine (CCL2, CX3CL1, IL-1β, CXCL10) in the human mesencephalic neuronal/glial cultures [105]. The protective effects of CB₂R activation are further supported in a neuronal/glial hippocampal culture model that reported inhibition of gp120-induced IL-1β production by WIN55,212–2, which subsequently let to reduced loss of synapses and was reversed by a CB₂R antagonist [122].

Cannabinoid-induced attenuation of neuroinflammation has also been demonstrated by upregulating eCB levels via the direct application of 2-AG or AEA or via enzyme inhibitors such as MAGL or FAAH [13, 129, 245]. The genetic deletion of the FAAH enzyme in GFAP/gp120 transgenic mice, and thus upregulation of AEA, but not 2-AG in whole brain, demonstrated downregulation of astrogliosis as well as enhanced neurogenesis [13]. As FAAH inhibition also upregulates levels of non-eCB-related lipids (i.e. palmitoylethanolamide, PEA; oleoylethanolamide, OEA) additional mechanisms beside AEA might be involved in the anti-inflammatory effects of FAAH inhibition, including PEA and OEA’s effects on AEA metabolism by binding to PPAR-α or to TRPV1 [66, 88, 138]. In a different study the upregulation of 2-AG via the MAGL inhibitor JZL184 reduced gp120-induced prostaglandin E2 and IL-1β production, which prevented synapse loss and was attributed to CB₂R activation [245]. In contrast to AEA, which acts only as a weak partial agonist toward CB₁R and CB₂R, 2-AG has been shown to act as a full agonist toward both receptors, CB₁Rs and CB₂Rs [223].

I sum, the findings suggest that cannabinoids have anti-inflammatory effects with neuroprotective benefits in the context of neuroHIV, potentially via CB₂R activation, but additional mechanisms might be involved. Further, enzyme inhibitors targeting MAGL and FAAH have great therapeutic potential as they allow for selective elevation of eCB signaling, which enables investigation of physiological actions of particular eCBs as well as reveal therapeutic potential of such precise modulation [67, 184].

3.4. Neuroprotective effects of cannabinoids via presynaptic mechanisms

Even though the anti-inflammatory properties of cannabinoids appear to be the predominant mechanism for the displayed neuroprotective benefits of cannabinoids in neuroHIV, some studies have demonstrated cannabinoids-regulating effects against HIV-1 protein toxicity directly on neurons via presynaptic mechanisms [100, 244].

Direct neuroprotective effects of cannabinoids have been demonstrated in primary prefrontal cortex neuronal cultures that assessed the effects of eCB ligands against HIV-1 Tat toxicity [244]. Findings indicated that the two endogenous cannabinoid ligands 2-AG and AEA significantly decreased Tat-induced intracellular calcium, neuronal excitability, dendritic injury, and neuronal death [244]. The protective effects of both eCBs were attributed to CB₁R synaptic function with CB₁R but not CB₂R protein levels being significantly upregulated [244]. Similarly, a different study demonstrated that enhanced AEA levels via the FAAH enzyme inhibitor PF3845 displayed protective effects against Tat-induced intracellular calcium, dendritic injury, and neuronal death in prefrontal cortex neurons in vitro [100]. Importantly, whereas CB₁R played a role for downregulating the immediate effects of Tat on intracellular calcium production, CB₂R appeared to be involved in the more long-term PF3845-induced protective effects on dendritic changes and neuronal survival, potentially due to the 10% astrocyte contribution in the neuronal culture model [100].

Additional cannabinoid-regulating effects on neuronal function in neuroHIV have been demonstrated on glutamatergic neurotransmission, including spontaneous, miniature, and evoked excitatory postsynaptic currents (sEPSCs, mEPSCs, and eEPSCs, respectively) [112, 241, 244]. The 2-AG- and AEA-induced downregulation of EPSCs has been shown to be related to a presynaptic CB₁R-related mechanism [241, 244] with neuroprotective effects against excitations an neuronal injury [244]. It has been demonstrated that the mechanisms involved in Tat-induced synapse loss include calcium influx via NMDARs [121]. Interestingly, cannabinoid-induced glutamatergic hypofunction has been shown to involve coupling of CB₁Rs with NMDAR NR1 subunit by forming heterodimers [200, 207, 208]. Note however, whether CB₁R-related signaling is impaired or enhanced during neuroHIV is currently debated [112, 241] and needs to be investigated in more detail.

For inhibitory neurotransmission, not much is known about the effect of eCBs on HIV-1/HIV-1 protein-induced alterations. In one study HIV-1 Tat-induced decreases in inhibitory gamma-aminobutyric acid (GABA)ergic neurotransmission in prefrontal cortex mouse brain slices were occluded by cannabinoids, such as WIN55,212–2 and AEA, via a presynaptic CB₁R-related mechanism [243]. Further, a recent study demonstrated normal CB₁R signaling at inhibitory synapses in the presence of Tat [241].

Overall, there is evidence that presynaptic CB₁R-related mechanisms contribute to neuroprotective effects against HIV-1 protein-induced toxicity. Additional studies are necessary to understand the underlying mechanisms involved in eCB’s potential ability to attenuate or reverse HIV-1-induced synapse decline and altered neurotransmission on glutamatergic and potentially GABAergic neurons.

4. Conclusion and future directions

Here we have reviewed the specific involvement of the eCB system in HIV-1 disease progression and its potential use as a therapeutic target to decrease HAND pathology. The review shows the specific involvement of CB₂R in neuroHIV due to the inflammatory nature of the disease and indicates to be a promising therapeutic target via its anti-inflammatory effects on the immune and CNS systems, with additional beneficial effects on viral load potentially due to CB₂R and CXCR4 heterodimerization. Further alternative cannabinoid receptors, including GPR55 and GPR18, are worth investigating in more detail to understand their regulating effects on the CNS and how they may contribute to or attenuate HAND pathogenesis.

In contrast, CB₁R involvement in neuroHIV is less clear. Besides the known psychoactive effects via CB₁R, and thus limited therapeutic use, potential protective effects via inhibition of HIV-induced excitatory neurotransmission by pre-synaptic CB₁R mechanisms have been reported but more studies are necessary to assess the role of CB₁R and/or cannabinoids on the inhibitory system in neuroHIV, and whether CB₁R signaling is enhanced or inhibited. Additionally, as viral load and the distribution of the eCB system vary across brain regions, systematic studies are necessary to assess brain-region specific differences.

In turn, the use of eCB degrading enzyme inhibitors as a tool to alter eCB tone in neuroHIV seems a promising future avenue. Enhancing AEA or 2-AG via FAAH or MAGL enzyme inhibitors, respectively, has high therapeutic potential as they are targeting ‘on site’ produced eCBs and inhibit eCB degradation [67, 184]. While there is debate about the safety of FAAH inhibition [68, 108] the first-class MAGL inhibitor Abx-1431 [46, 64, 114] has entered clinical phase 2 for the study of Tourette syndrome or chronic motor tic disorder and indicates positive effects in these patients (www.clinicaltrials.gov). Moreover, the combination of different mechanisms of action to enhance eCB tone has started to receive attraction in other diseases [230] and is worth investigating in more detail in the context of neuroHIV.

Noteworthy, an important point to consider is the interaction of cannabinoids and cART medication. Not a lot of information is currently available but clinical studies are being conducted to investigate their interactive effects [55]. A past study has reported no clinical interaction of cannabinoid with protease inhibitors [1]. If these findings are confirmed in future studies cannabinoid-based treatment will have high therapeutic potential in the context of neuroHIV.

Further, evidence indicates that the eCB system appears to be altered in neuroHIV but more research is necessary to evaluate its impact on the CNS in PWH. Specific factors that should be taken into account are differences in HIV strains/variants, human/genetic variability, pharmacokinetics, and sex, which all could contribute to disease progression and degree of response to cannabinoid treatment. For example, inhibitory crosstalk between HIV-1 coreceptor CXCR4 and CB₂R has been reported to play a potential role in viral suppression, whereas CB₂R cross-regulation of CCR5 is less clear. Further, sex differences have been reported for the eCB system [51, 205] as well as for the severity of neurocognitive impairments in HIV [31, 143, 204], thus indicating the importance of including sex as a biological variable in clinical and preclinical studies. Lastly, when thinking about identifying components of the eCB system as useful biomarkers the use of in vivo neuroimaging studies, such as PET and fMRI, is highly encouraged and would provide very relevant data on eCB-related alteration in neuroHIV as well as cannabinoid treatment opportunities.

Highlights.

• HIV neuropathogenesis alters the endocannabinoid system

• Effects of cannabis use on cognition in people living with HIV-1 depend on various factors, including disease progression, the amount of cannabis use, and the cognitive domain assessed

• Cannabinoid type 2 receptors and endocannabinoid catabolic enzymes are promising therapeutic targets in people living with HIV-1