Role of Microglia in Psychostimulant Addiction

Abstract

The use of psychostimulant drugs can modify brain function by inducing changes in the reward system, mainly due to alterations in dopaminergic and glutamatergic transmissions in the mesocorticolimbic pathway. However, the etiopathogenesis of addiction is a much more complex process. Previous data have suggested that microglia and other immune cells are involved in events associated with neuroplasticity and memory, which are phenomena that also occur in addiction. Nevertheless, how dependent is the development of addiction on the activity of these cells? Although the mechanisms are not known, some pathways may be involved. Recent data have shown psychoactive substances may act directly on immune cells, alter their functions and induce various inflammatory mediators that modulate synaptic activity. These could, in turn, be involved in the pathological alterations that occur in substance use disorder. Here, we extensively review the studies demonstrating how cocaine and amphetamines modulate microglial number, morphology, and function. We also describe the effect of these substances in the production of inflammatory mediators and a possible involvement of some molecular signaling pathways, such as the toll-like receptor 4. Although the literature in this field is scarce, this review compiles the knowledge on the neuroimmune axis that is involved in the pathogenesis of addiction, and suggests some pharmacological targets for the development of pharmacotherapy.

1. INTRODUCTION

Psychostimulants induce neuroplastic adaptations in memory and learning processes related to the positive reinforcement in the reward system, “hijacking” the neural adaptive motivational mechanisms and causing the dysfunctional pattern of behavior that characterizes drug addiction [1, 2]. Importantly, these neuroplastic changes that occur after drug exposure are robust; even after extended periods of abstinence, the reinforcing effects are still present, leading to high relapse rates [3]. Consequently, despite their deleterious effects and difficulty in treatment, psychostimulants remain extremely widely used [4]. This review will begin by discussing important dopaminergic and glutamatergic alterations involved in neuroplasticity induced by psychostimulant abuse. In addition, we will review the neurobiology of microglia and how these immune cells are important for synaptic plasticity. We will discuss the literature that shows alterations in microglial cells induced by psychostimulants. Finally, we will outline possible pathways through which microglia can mediate the neuroplasticity induced by psychostimulants.

2. GENERAL EFFECTS OF PSYCHOSTIMULANTS IN THE DOPAMINERGIC AND GLUTAMATERGIC PATHWAYS

Drug addiction is a chronic disease characterized by a compulsive drug-seeking behavior, lack of self-control during use, and negative physiological/psychological changes such as irritability, anxiety, and dysphoria in the absence of the substance [3, 5]. According to DSM-5, substance use disorder encompasses multiple distinct classes of drugs. Despite being divided into different categories and presenting different neuropharmacological properties, psychoactive substances can increase dopamine levels in the reward system associated with the development of addiction [5, 6].

Psychostimulants, such as cocaine, amphetamine, methamphetamine (METH), and methylene-dioxy-methamphet-amine (MDMA; ‘ecstasy’) acutely increase the monoaminergic neurotransmission in this circuit by either blocking the membrane transporters involved in the reuptake of dopamine, norepinephrine, and serotonin, or competing with the uptake of monoamines albeit with different affinity [7]. Additionally, amphetamines can disrupt neurotransmitter storage by interacting with the vesicular monoamine transporters (VMAT2) [8] that underlie enhanced monoamine release via a reverse efflux transport mechanism [9-11]. Importantly, amphetamines increase dopamine transporter-induced sodium conductance, leading to neuronal depolarization and increased dopamine transmission [9, 10, 12-14].

The adaptive cellular responses in the CNS induced by prolonged exposure to psychoactive substances may occur due to the signaling pathways recruited by the dopamine receptors. For instance, activation of D₁ increases the cAMP levels that further activate protein kinase A (PKA) and extracellular signal-regulated kinase (ERK) [15]. This leads to the activation of transcription factors such as delta FosB (ΔFosB), transcription factor regulated by cAMP response element-binding protein (CREB), myocyte enhancing factor 2 (MEF2), and nuclear factor kappa B (NF-κB), which are related to changes in genes and proteins expression [16], receptor expression, cell morphology and neuronal excitability [17-20]. Thus, psychostimulants can modify signaling cascades related to neuroplasticity, which is crucial for psychostimulants' craving, drug seeking, and relapse [21, 22].

Apart from dopamine, acute and chronic psychostimulants exposure can also dysregulate glutamatergic synaptic activities that involve alterations in glutamate levels and their receptors expression, long-term potentiation (LTP), long-term depression (LTD), and modulation of synaptic connectivity strength in various brain regions involved in reward and reinforcement [23, 24]. These substances also alter the glutamatergic system of medium spiny neurons (MSN) in the nucleus accumbens (NAc), whose functions are driven by excitatory synapses arising from several cortical and subcortical projections [25].

Recently, it has been shown that some of these molecular mechanisms could be regulated by microglial cells. For example, microglia can alter the morphology and functionality of dopaminergic neurons by modifying the expression of the receptors, dopamine transporter (DAT) and tyrosine hydroxylase (TH) [26-29]. These cells are also involved in glutamatergic-induced synaptic alterations by changing AMPA receptor expression, AMPAR/NMDAR ratio, and glutamate release [30-33]. Finally, both dopaminergic and glutamatergic neurotransmissions also modulate microglial activity [34, 35], including in the context of psychostimulant abuse [36, 37]. Thus, this places these cells as an essential factor in the pathogenesis of substance use disorder. In the following sections, we will revise the current knowledge on the effects of microglia and inflammatory mediators in the brain alterations induced by cocaine and amphetamines.

3. THE INVOLVEMENT OF MICROGLIA IN DRUG ADDICTION

3.1. Overview of the History and Neurobiology of Microglial Cells

Microglia were first studied by Nissl in 1899, who described them as a distinct cellular entity with migratory and phagocytic capacity, called Stäbchenzellen, or rod cells, due to the shape of their nuclei. In 1913, Rámon y Cajal named microglia, together with oligodendrocytes, “third element”, with neurons and astrocytes being the first and second elements. However, microglia were still unknown, and only in 1932 Pío del Rio-Hortega introduced the term “microglia” and characterized them phenotypically [38, 39]. He described the microglia as cells with ameboid morphology from the mesoderm, which enter all brain regions during early development. In the mature brain, they present a ramified morphological phenotype in the homeostatic state but acquire an ameboid morphology during pathological events, with an increased phagocytic activity [38-40].

Microglia are resident cells of the immune system that consist of about 5-10% of brain cells. They are present in the cerebral parenchyma, interacting with neurons, astrocytes, and oligodendrocytes, and together with perivascular macrophages, they occupy strategic niches, covering the entire central nervous system (CNS) [40]. Additionally, microglia numbers and their spatial distribution are heterogeneous under physiological conditions, with main differences in some regions such as the striatum, hippocampus, and cortex [41].

Microglia detect changes in the brain mainly through pattern recognition receptors (PRRs), which recognize pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs), triggering an inflammatory response [42]. In this process, microglia exert a neuroprotective function to reestablish homeostasis [43, 44]. Their homeostatic state is characterized by a highly branched morphology and a small cell body. However, once exposed to pathogens or tissue debris, these cells become polarized in a reactive phenotype with high phagocytic activity, characterized by a larger cell body and a rounded shape, shorter and thicker branches [45, 46]. In addition, there is an increase in molecules such as cytokines and the major histocompatibility complex (MHC) [43]. These morphological characteristics and specific secretory profiles occur due to the microglia's ability to perform functions such as phagocytosis of apoptotic neurons and cellular debris, protection of the brain against infections, elimination of the pathogen and tissue repair [47, 48].

These cells produce a variety of mediators depending on the stimuli they are exposed to. For example, when stimulated with lipopolysaccharide (LPS) or interferon-gamma (IFN-γ), microglia produce more pro-inflammatory or “toxic” mediators, such as tumor necrosis factor (TNF), interleukin (IL)-1β and nitric oxide (NO). On the other hand, their stimulation with IL-4 or IL-13 leads to the production of anti-inflammatory or “protective” mediators like IL-10 and arginase-1 [49-52]. However, these stimuli lead to different alterations in gene expression, electrophysiological properties, neurotransmitter receptors and microglial cell morphology and motility, allowing functional changes according to stimuli and environmental conditions. Thus, this broad spectrum of microglia activation gives them the versatility to participate in different brain processes [53] such as neural network development, neuronal survival, neurogenesis, axonal outgrowth, and synaptogenesis [54].

3.2. Microglia Regulate Synaptic Plasticity

Microglia were initially considered static cells in healthy environments. However, this view has evolved, with studies increasingly focusing on their physiological roles in the healthy brain [55, 56]. In this context, in addition to its neuroinflammatory function as a response to tissue damage, microglia may modulate the establishment and remodeling of neuronal circuits, actively participating in processes such as neurogenesis, programmed cell death, cell survival, and proliferation. Therefore, since microglia may be involved in modeling the central nervous system throughout life, microglia malfunction may be involved in neurodevelopment and CNS diseases [57-59].

Microglia have also been recognized as important regulators of synaptic plasticity due to their ability to change pre-and postsynaptic elements that alter the synaptic morphology. This is made through the release of trophic factors or microglia-neuron contact.

3.2.1. Cell-to-Cell Communication

Microglia-neuron communication through the CX₃CL₁, released by neurons, and CX₃CR₁ receptors, expressed by microglia, are important for neuronal excitability [60, 61]. IL-1β release by stimulation of CX₃CR₁ increases glutamate release, leading to postsynaptic NMDA receptor activation. TNF, another important cytokine released by microglia, regulates homeostatic synaptic scaling of AMPA receptors and may boost glutamatergic neurotransmission. Inhibitory neurotransmission in the cortex is reduced by activated microglia, which displaces GABAergic synapses [62-64] Thus, these cells can either remodel neuronal networks or control the synaptic properties.

The synaptic remodeling mediated by microglia involves the CX₃CR₁ (fractalkine receptor) and other receptors like the CR₃ complement and purinergic receptors. Microglia phagocytize dendritic spines and trogocyte small portions of axons by these different pathways. In fact, both pre- and postsynaptic (SNAP25 and PDS95, respectively) components were found inside the microglia lysosomes [65, 66]. The attraction of the microglial process to dendrites to provide contact with neurons is mediated by P2Y₁₂ receptors. In this sense, complement C1q, C3, and C4 components which are expressed by neurons, function as targets for microglia. Indeed, synapses may be eliminated by recognizing synaptical C3 by the microglial CR₃. Depleting these glial cells or inhibiting the complement system in the adult mouse brain affects memory removal in the contextual fear test [67]. On the other hand, CD47 and its receptor SIRP mediate the “do not eat me” signaling, since their deletion causes an increase in synaptic phagocytosis by microglia and a decrease in functional synapses [68]. Furthermore, modulation of neuronal connection by these cells can also occur by forming functional synapses. Microglial processes contacting the dendrite can increase calcium concentrations and actin accumulation. They can engulf the extracellular matrix and trogocyte presynaptic structures and induce filopodia formation and extension [55].

The clusters of differentiation (CD) are important signaling that control microglia homeostasis [69]. The CD200 is a neuronal glycoprotein that binds to the microglial receptor CD200r and controls the inflammatory activity of the cells by maintaining them in an homeostatic state [70, 71]. Disruption of this signaling can lead to synaptic deficits and cognitive impairment [71]. The CD22-CD45 pathway also works as a “negative regulator” of the neuroinflammatory processes. CD45 is a transmembrane protein-tyrosine phosphatase expressed by microglial cells, while neurons express their ligand CD22 [72, 73]. Upregulation of CD22 pathway is related to impaired microglia phagocytic activity and surveillance, as well as cognitive impairment in aged mice [72, 74]. Finally, the CD95 (Fas) and CD95 (Fas ligand) is a pathway that can drive homeostatic microglia to a more activated state [75] and mediate synaptic plasticity [76, 77].

3.2.2. Soluble Factors (Non-Contact Dependent)

Synaptic remodeling can also occur by microglia release of trophic factors that induce cell proliferation and survival, involving IL-9, IL-10, BDNF, and insulin-like growth factor type 1 (IGF-1) [78-80], among others. In addition, signals that involve TNF-α and IFN-γ can regulate synaptic pruning, eliminating the excess of immature synapses [81, 82].

Various substances secreted from neurons, such as neurotransmitters, may control the microglia's ability to change the synapses' morphology. This is achieved by different receptors expressed on glial cells, including metabotropic and ionotropic receptors. For example, IL-33 released from neurons acts on microglia and is important for dendritic spine plasticity and, possibly, memory consolidation [55]. Furthermore, microglial P2Y₁₂ purinergic and β-adrenergic receptors control cytokine release and alter the visual cortex's synaptic plasticity [83, 84].

Finally, it's important to emphasize that microglia release factors that modulate the activity of other glial cells, which could exert roles in synaptic plasticity. For example, ATP and TNF released by microglia increase glutamate release by astrocytes via P2Y₁ and TNF receptors, which alters the synaptic transmission [85]. The release of ATP by astrocytes may also attract microglia processes to these regions. Additionally, microglial BDNF can bind on TrkB in OPCs, which is important for activity-regulated myelination and memory [86]. Thus, glia-glia interaction is an essential communication for regulating neuroinflammation and acting as sculptors of synaptic plasticity.

3.2.3. Microglia as a Source of BDNF for Synaptic Plasticity

BDNF is the most abundant neurotrophic factor in the brain and, together with its receptor TrkB, is expressed in brain regions involved in learning and memory processes. In the CNS, BDNF is stored and released from axonal and dendritic compartments and other neural cells, such as the microglia and the astrocytes [87].

Recently, studies have shown the importance of the BDNF released from microglial cells in physiological and pathological conditions. Parkhurst et al. (2013) demonstrated that animals with a specific depletion of the microglial BDNF had impaired spine formation and, consequently, impaired motor task learning [78]. On the other hand, although the specific genetic ablation of microglial BDNF had no effects on the morphology of the dendritic spine of adult-born neurons under physiological conditions, it impacted microglial proliferation and hippocampal neurogenesis [88]. Besides, in neuroinflammatory conditions induced by LPS or traumatic brain injury, the specific microglial BDNF ablation increased the generation and survival of immature neurons, and also reduced the activation of these glial cells [88]. Finally, specific microglial BDNF modulation was also important in other conditions such as neuropathic pain [89], withdrawal, and hyperalgesia induced by morphine [90, 91], and after cocaine exposure [92]. Recent studies showed the role of the microglia in synaptic plasticity, besides the importance of the BDNF-TrkB pathway in conditions where microglial cells are activated [93, 94]. Activation of the phosphatidylinositol 3-kinase (PI3K)/Akt or Ca²⁺/calmodulin dependent kinase 2a (CAMK2A)/CREB in BV-2 microglial cells increased the expression of BDNF and led to alterations in neuronal LTP and synaptic plasticity [95, 96]. In this sense, microglia activation is necessary for functional and structural plasticity in a model of chronic pain induced by high-frequency stimulus (HFS). Besides, specific ablation of the microglia or ablation of the microglial BDNF prevented the LTP induced by the HFS [97]. In another model of chronic pain, activation of the microglial receptor P2X₄ induced BDNF release from microglia that, consequently, mediated the plastic changes that occur in central sensitization induced by nitroglycerin (NTG) [98], and NTG associated with nicotine [99, 100]. Besides, increased BDNF levels induced by P2X₄ activation were also observed after alcohol exposure, which was prevented by the 5-BDBD - a P2X₄ antagonist - suggesting the P2X₄ signaling as an essential modulator of the microglia in neuroplastic functions in the context of alcohol [101-103]. In sum, given the interaction of the microglia and BDNF in synaptic plasticity in different conditions, both could act as key players in neural plasticity induced by psychostimulants.

4. RESPONSES OF MICROGLIA TO PSYCHOSTIMULANTS

Although alterations in the dopaminergic and glutamatergic systems are considered the key to neurobiological changes that regulate motivated behavior, it is well known that psychoactive substances can also alter other neurotransmitter systems such as glutamatergic, serotonergic, GABAergic, and also modulate different molecular pathways, including the neurotrophic factor and cytokines levels [104-106]. In addition, microglial cells express ion channels and neurotransmitter receptors that are also expressed in neurons; therefore, they can respond to neurochemical changes induced by psychoactive substances. Interestingly, microglial activation has been shown to modulate the reward system [29, 107], mainly because dopaminergic neurons present high liability to neuroinflammatory signals [108]. For example, in the mesolimbic pathway, D₃ knockout or D₃ inhibition showed increased BDNF levels, microglial activation, and synaptic density, which was prevented by the microglia inhibitor minocycline [109, 110].

4.1. Microglia and Cocaine

Preclinical studies have demonstrated that cocaine increases the expression of the classical microglial marker Iba-1 in the hippocampus, frontal cortex, and NAc [36, 92, 111, 112], as well as in sections encompassing the entire brain [112, 113] of mice and rats. In addition, cocaine increased the CD11 expression in the ventral tegmental area (VTA), striatum, and cortical regions [114-116]. Interestingly, Brown et al. (2018) found that acute cocaine injection only increases CD11 expression in mice with self-administration history [115]. Furthermore, cocaine also increased the expression of CD68 - a transmembrane glycoprotein expressed by the microglia that suggests increased phagocytic activity [117, 118]. In addition to altering protein expression in microglial cells, cocaine also alters their number and morphology. For example, Thangaraj et al. (2020) found an increased number of cells expressing AIF1 after cocaine treatment for seven consecutive days [119]. Moreover, Lewitus et al. (2016), Cotto et al. (2018), Silva et al. (2020), and Burkovetskaya et al. (2021) demonstrated that cocaine exposure increases the microglial cell body size and reduces the microglial branches [36, 92, 118, 120]. In other words, cocaine induces morphologic alterations, which may be related to increased microglial activation.

In non-human primates, striatal sections from monkeys subjected to the cocaine self-administration paradigm for approximately 15 months presented a greater binding of [3H]PK11195 - a marker for the microglial 18-kDa translocator protein (TSPO) in the autoradiography [121]. A study that evaluated cocaine users during abstinence did not observe any alteration in TSPO, as assessed by the PET imaging with [11C]PBR28 [122]. On the other hand, immunohistochemistry of brain slices from cocaine users showed higher ricinus communis agglutinin-1 (RCA-1) - a lectin that specifically binds to microglia - and CD68 expression than matched controls [123].

Microglial cell activity can modify the molecular mechanisms and behavioral alterations induced by psychostimulants [36, 118, 124, 125]. Inhibition of microglial activation by minocycline, a tetracycline antibiotic widely used as a microglial inhibitor, reverses the behavioral sensitization, CPP, and the NAc dopamine release induced by cocaine [125, 126]. Besides, minocycline also alters cocaine reward after nicotine and morphine exposure [91, 124]. In addition, ibudilast, a PDE4 inhibitor and glial modulator, reduces cocaine-seeking and craving in mice and humans [127-129]. Moreover, microglia depletion with the CSF-1 inhibitor PLX3397 decreased behavioral alterations induced by cocaine [118, 124]. However, the role of microglia in cocaine addiction needs further investigation as some results are controversial. For example, reduced microglial number and activity induced by minocycline, PLX3397, GW2580, or Mac-1-saporin had no effect on the CPP induced by cocaine [29, 91].

Psychoactive substances are recognized as exogenous substances by microglial cells, leading to a neuroinflammatory response [36, 125]. Regarding the psychostimulants, most studies have focused on the direct activation of toll-like receptors (TLR) - a family of PPRs present in the microglia - that, when activated, can trigger the translocation of transcriptions factors such as NF-κB and, consequently, activation of these cells and the release of cytokines. Inhibition of this transcription factor reduces both CPP and synaptic plasticity in the NAc induced by cocaine [18].

Northcutt et al. (2015) showed in silico docking and competitive binding assay that cocaine can directly bind in the classical myeloid differentiation factor 2 (MD-2) domain in TLR4 [125]. In microglia cell cultures and animal models, cocaine up-regulates the expression of TLR4 and other components of this receptor's pathway - MyD88, IRAK1, TRAF6, and NF-ĸB. Besides mediating microglial activation and cytokine release after cocaine treatment, activation of TLR4 by cocaine is also important for the increased dopamine levels in the NAc induced by this psychostimulant [115, 125, 130, 131]. Thus, modulation of TLR4 signaling appears to be important for addictive-like behaviors induced by cocaine.

The antagonism of TLR4 receptors by using (+)-naloxone and (+)-naltrexone inhibited the development of CPP and cocaine self-administration, respectively [125]. An intra-VTA injection of LPS from Rhodobacter sphaeroides, an antagonist of TLR4, also diminished cocaine seeking in a model of relapse-like behavior [115, 132] . Moreover, depending on the dose, TLR4^-/- mice do not develop CPP and hyperlocomotion induced by cocaine [115, 132]. On the other hand, while activation of this receptor in the VTA by LPS reinstates cocaine-seeking behavior, stimulation with MPLA, a weak TLR4 activator, decreases the behavioral sensitization by this drug [36]. Finally, decreased TLR4-mediated neuroinflammation appears to be the mechanism of activation of other substances to protect against behavioral alterations induced by this psychostimulant. For example, exendin-4, an analog of glucagon-like peptide 1, decreased the TLR4, TNF-α, and IL-1β, as well as the CPP induced by cocaine [131]. In addition, engineered extracellular vesicles loaded with miR-124 inhibited cocaine-mediated TLR4 signaling and microglial activation [113].

Other receptors from the TLR family may be associated with cocaine-mediated microglia activation. Exposure of microglial cells to cocaine resulted in increased expression of TLR2 with concomitant induction of microglial activation [117]. It has been shown that cocaine induces the formation of reactive oxygen species, leading to endoplasmic reticulum stress and subsequent translocation of the transcription factor ATF4 to the nucleus, which induces TLR2 expression. Interestingly, the signaling activated by this receptor could be further stimulated by cocaine, which would induce microglial activation [117]. Pharmacological or genetic depletion of TLR3 reduces behavioral alterations induced by cocaine, such as CPP, self-administration, locomotor activity, and pathways necessary for NF-κB activation [133].

Finally, recent studies provide the involvement of another family of PPRs in microglial activation induced by cocaine, the nucleotide-binding and oligomerization domain (NOD)-like receptors (NLRs), most specifically, the pyrin domain-containing 3 (NLRP3). Cocaine also induces NLRP3 expression and IL-1β formation in microglia. The reduction of NLRP3 activity reduced both effects induced by cocaine in the brain of mice [116]. In addition, KO mice for the CX3CR1, a receptor expressed solely by microglial cells, presented higher hyperlocomotion activity and CPP induced by cocaine and higher activity of NLRP3 [134]. Moreover, inhibition of cocaine-mediated increase in IL-1β in the VTA blocks the CPP and self-administration [125]. Thus, activation of this neuroinflammatory signaling in microglial cells appears to be involved in behavioral abnormalities induced by cocaine.

4.2. Microglia and Amphetamines

Amphetamine is one of the phenethylamine derivatives that includes d-amphetamine (d-AMPH), METH, and MDMA/ecstasy [135]. METH has been the most investigated for its properties to interfere with microglia activity. In vitro experiments using different lineages of microglial cells, such as N9, BV2, and highly aggressively proliferating immortalized cells (HAPI), have shown METH-induced microglial activation and up-regulation of IL-1β, IL-6, and TNF-α [136-140]. This also occurred in primary microglial cells [139, 141, 142] and mixed neuron/microglia culture [143]. Notably, the regulation of the cytokines may vary depending on the amphetamine class [144]. In BV2 cells, METH exposure also increases intracellular proteins related to inflammatory pathways such as the kinases p-p38 and MAPK [145], and the pro-apoptotic protein Bax, caspase-3, and caspase-9 [146]. This drug also increased the expression of iNOS and decreased arginase [147]. In the human microglial cell line (HTHU), METH also increased caspase-3/7 and -9 expressions and decreased mitochondrial (mt)DNA content [148]. Moreover, METH also upregulated CD11b in different hippocampal subregions of organotypic slices [149]. On the other hand, incubating microglia with METH did not alter the expression of genes involved in inflammatory processes, albeit it reduced IL-1β protein levels. In addition, only a high concentration of METH reduced cell viability [150].

METH can also alter microglial-like cell morphology. For example, N9 cells, which typically present ramifications indicative of surveillance, retract their filopodia and turn into a round shape, characteristic of activation, after being exposed to METH. This psychostimulant also reduced cell volume of NR-9460 microglia [151]. METH also altered microglia in neuron/glia cocultures, which authors describe as reactive cells [152]. However, it is worth noting that the evaluation of microglia morphology in cultures is often a topic of criticism due to the absence of other cells and environment necessary for a plain occurrence of this phenomenon.

In vivo, repeated administration of METH in mice or rats led to increased Iba-1 in the medial prefrontal cortex (mPFC), striatum (CPu), and hippocampus (Hip, including the CA1 and DG) [37, 136, 147, 153-160], parietal cortex [161] and other regions [162, 163]. Interestingly, METH increased Iba-1 expression that started 30 minutes after drug exposure and lasted seven days in the striatum [164]. On the other hand, other studies demonstrated that METH decreases Iba-1 expression in the hippocampus [165]. In addition, acute METH exposure can increase the other classical microglial marker CD11b in mice striatum [166] and hippocampus [167]. Besides, CD11b was upregulated in the VTA, striatum, and hippocampus following repeated METH administration [168-171]. Interestingly, repeated administration of METH led to an accumulation of activated microglia, evaluated by CD11b immunoreactivity, in the striatum that peaked 3 days after the initial dose [172]. Moreover, METH increased the expression of CD68, a marker of phagocytic cells, in the striatum, hypothalamus, hippocampus, and cerebellum [37, 157, 160]. Furthermore, several works have used the isolectin B₄ (ILB₄) to evaluate microglial activation in the context of METH abuse. Repeated or acute METH injection increased the ILB₄ in the striatum [173-177]. Importantly, increased ILB₄ induced by METH was dose-dependent and coincided with the damage of dopamine nerve endings [178]. On the other hand, Thomas et al. (2009) found no difference in ILB₄ in the SNc or the VTA [179].

Microgliosis was also accessed by the saturation binding of the peripheral benzodiazepine receptor (PBR) ligand 1-(2-chlorophenyl)-N-methyl-N-(1-methylpropyl)-3-isoquinoline carboxamide ([3H]PK 11195). Chronic METH exposure increased the [3H]PK 11195 binding in the striatum, NAc, hippocampus, frontal cortical areas, the rhinal cortices, and cerebellum [180-182]. Finally, METH also increased other microglial markers such as OX-42 and F4/80 in the striatum [183], and integrin-αM in the PFC and hippocampus [184], while decreased AF1 expression in the hippocampus [165]. Lloyd et al. (2017) also described an increased microglia activation in the striatum after 10 days of METH treatment, but decreased in the nucleus arcuate [185]. Moreover, Friend et al. (2013) also found a significant thickening of branches and more intense microglia cell-body staining in the striatum after acute toxicity induced by METH [166]. Finally, Canedo et al. (2021) found that METH induces a reduction in microglial total process length, territorial area and territorial volume, while increasing the number of branching points [37].

Other amphetamines also can alter microglial marker expression. Repeated amphetamine administration increased CD11b immunoreactivity in the hippocampus, in the prelimbic PFC but not infra-limbic PFC [186], as well as ILB₄ in the striatum and piriform cortex [187, 188]. Furthermore, para-methoxymethamphetamine (PMMA) - a para-ring-substituted amphetamine derivative - and methiopropamine - a METH analog - significantly increased Iba-1 in the striatum [189, 190]. Importantly, methiopropamine also increased the CD16, CD32, and CD86 mRNA expression [190]. In addition, reactive Iba-1 and ILB₄ expressing cells were also increased by MDMA in the hilus and striatum [187, 191]. At the same time, MDMA exposure increased OX-42 immunoreactivity in the hippocampus [192], frontal cortex, and hypothalamus [193, 194]. Regarding the CD11b, some studies have shown that this microglial marker was increased in the striatum [169, 195-199], motor cortex, NAc, VTA [198], CPu [200], frontal cortex, hypothalamus [201], SNc [199] and hippocampus [169]. However, some studies found no difference in CD11b in the striatum, SNc [202], hippocampus, and the mPFC [203] after MDMA exposure.

Amphetamines also alter the microglia number and morphology. It has been shown that these cells can exhibit increased soma size and thicker processes after METH exposure in the striatum [155, 204-207], hippocampus [147, 208], thalamus, dorsal raphe/central gray area, inferior colliculus [208], somatosensory and piriform cortices, periaqueductal gray [209] and substantia nigra pars compacta [210]. On the other hand, the total dendritic length and the cell soma area in the hilus were not different after the reinstatement of METH self-administration [211]. In addition, microglia with a morphology resembling activation were found in the parietal cortex after MDMA exposure [212].

In human studies, western blotting and immunohistochemistry assay from chronic METH users showed increased Iba-1 expression in the hippocampus [213]. Moreover, by using immunohistochemistry, Kitamura et al. (2010) found increased human glucose transporter 5 (hGLUT), a useful marker of microglia, in the striatum from METH abusers, but no difference in the CR3.43 - a major histocompatibility complex class II antigen specific for activated microglia [214]. In addition, positron emission tomography using the radiotracer ([3H]PK 11195) to evaluate activated microglia was higher in METH abusers in the striatum, midbrain, thalamus, and orbitofrontal and insular cortices. Importantly, the binding ([3H]PK 11195) was inversely associated with the duration of METH abstinence [215]. However, human studies appear to be contradictory. For example, a positron emission tomography study with the 18 kDa translocator protein radioligand [F-18]FEPPA to evaluate microgliosis found no difference between METH abusers and controls [216]. Further, human brain tissue from death cases of lethal intoxications by METH found no difference in TMEM119 [217]. Thus, more studies are necessary to understand how METH abuse can alter microglia activity since microglial markers themselves could be affected by METH exposure. For example, while METH exposure did not affect TMEM119 expression in mice striatum, other unique microglial markers, such as P2ry12 and Olfml3, were changed [218].

In silico modulations, molecular dynamics simulations and titrations showed that similar to cocaine, METH and AMPH are capable of binding in the TLR4/MD-2 complex [171, 219]. In the microglial cell line BV2, METH-induced microglial activation and cytokine release was mediated through the TLR4/NF-κB pathway [137, 138]. In this sense, METH exposure increases TLR4 expression and other signaling components such as NF-κB, MAPK, p-p65, p-p38 and Peli1 [137, 138, 145]. In addition, Wires et al. (2012) found that METH can increase NF-κB expression in the CHME-5 cell line [220]. Moreover, apart from altering NR-9460 cell morphology, METH can also alter the TLR4’s distribution on the surfaces of these microglia-like cells [151].

Increased TLR4 signaling activation after METH exposure was also found in vivo. Yang et al. (2020) found that repeated METH administration leads to microglial activation and upregulation of TLR4, TRIF, Peli1, and IL-6 and TNF-α expression in the mouse brain [137]. Similarly, Xie et al. (2018) found increased TLR4, MyD88, TRAF6, NF-κB, PARP-1, IL-1β, IL-6 and TNF-α in the rat striatum [221]. Significantly, blocking TLR4 expression inhibited METH-induced neuroinflammation, which could be related to behavioral alterations induced by METH [222]. For example, the non-opioid TLR4 antagonist (+)-naloxone blocked METH-induced microglial activation and IL-6 expression and increased NAc dopamine and CPP induced by METH [171]. Injection of the classic TLR4 antagonist LPS-RS also blocked the increase in dopamine release in the NAc induced by METH [171]. Furthermore, minocycline blocked the microglial activation and the increased TLR4, MyD88, and phosphorylation of NF-κB and TNF-α, improving recognition memory and spatial learning and memory impairments in the Novel object recognition and Morris water maze, respectively, induced by METH [223] [183]. In addition, ibudilast, its amino analog AV1013, and minocycline - which block microglial activity - reduced METH self-administration [224]. In fact, both ibudilast and AV1013 also blocked the development of behavioral sensitization induced by METH [225]. Interestingly, minocycline blocked the development of locomotor hypersensitivity to amphetamines in other neuroinflammatory models [226, 227]. Regarding the other amphetamine drug - the MDMA - minocycline prevented the MDMA-induced increase in microglial activation, NF-kB activation, and IL-1β release in the hippocampus, striatum, and frontal cortex [169, 201].

Postmortem analyses demonstrated an increased Iba-1 expression in the CA1 region of the hippocampus of chronic METH users. By performing a protein-protein interaction network analysis, the authors suggested an essential role of the TLR4 signaling in the effects of this psychoactive substance [213]. However, contradictory observations are reported. For example, in METH dependent volunteers, the anti-inflammatory ibudilast decreased some reward-related subjective effects [228] and peripheral inflammatory markers [229]. In addition, minocycline improved the psychotic symptoms of METH use disorder in a female patient [230], and there was no difference between ibudilast and the placebo group in METH abstinence in a randomized placebo-controlled trial, suggesting that ibudilast did not affect METH use frequency [231].

Apart from the TLR4 pathway, this class of psychostimulants can also induce microglia activation through other signaling mechanisms. For example, METH altered the expression of different cytokines in human macrophages by interfering with the MyD88-dependent TLR9 signaling pathway [232]. Furthermore, the drug also increased NLRP3 expression and caspase-1 cleavage in mice hippocampus and the microglia-like cell culture [233]. METH can also increase NLRP3, caspase-1, and IL-1β in the brain of METH users [213]. Importantly, blocking TLR4 expression also blocked the increase in caspase-1 induced by METH [222]. In addition, METH and MDMA increased P2X₇ expression in microglial cells in vitro and in vivo [192, 234]. Silencing the microglial P2X₇ decreases cytokine release induced by METH [234], while P2X₇ deficiency mice or P2X₇ pharmacological blockade protects against sensitization induced by amphetamine [235]. Moreover, microglial migration towards CX₃CL₁ is enhanced by METH, an effect that is reduced by P2X₇ blockage [234]. Despite microglial morphological alterations induced by METH occurring to the same extent among WT and CX₃CR₁ KO mice [176], the CX₃CR₁ KO mice showed minor locomotor alteration induced by amphetamine in comparison with the WT mice [236]. Further, by activating the CD200R, expressed specifically in microglia - there is a reduction in microglial activation after METH exposure [143]. Finally, microglia depletion induced by CD11b-TK^mt-30 transgene attenuates wake/sleeping abnormalities induced by METH [237]. Thus, different neuroinflammatory signaling appears to be involved in microglial activation induced by amphetamines.

Although psychostimulants activate microglia, it is not yet known whether this process occurs in a direct or indirect way. It has been shown that METH induces different inflammatory mediators via activation of the NF-κB and other pathways in N9 and BV2 microglial cell lines [136-138]. On the other hand, Canedo et al. (2016), Frank et al. (2021) and Bravo et al. (2022) have showed that METH can not directly activate microglial cells in primary cell culture models [37, 150, 238]. Instead, glutamate and TNF-α released by astrocytes mediated METH-induced microglial activation. Interestingly, neurons can modulate this astrocyte-microglia activation through the CD200/CD200r signaling [37, 238]. This would suggest that only microglia cell lines appear to directly respond to METH, but not primary cells. Thus, although different neuroinflammatory signaling appears to be involved in microglial activation induced by amphetamines, the mechanisms are unclear.

4.3. Effects of Psychostimulants on Inflammatory Mediators

Psychostimulants alter the activity of microglial cells, significant producers of cytokines, eicosanoids, and others, in the CNS. Besides, psychostimulants can also induce the release of inflammatory mediators by the peripheral immune system. Therefore, these drugs could influence microglia by acting in the CNS and in the periphery.

4.3.1. Cytokines

Although most of these studies were based on correlations between cytokine levels and behavioral outcomes, a few studies presented causal evidence for an involvement of neuroimmune responses in the central effects of psychostimulants (Tables 1 and 2). For instance, in IL-6 knockout mice, several effects of METH were attenuated in the HPC, cortex, and the dorsal striatum, such as decreases in serotonin levels, dopamine transporter depletion, serotonin transporter depletion, and gliosis [239]. Repeated administration of IL-6 induces long-term increased sensitivity to locomotor effects of AMPH [240]. IL-6 is increased in METH users compared with the controls. In addition, the levels of this cytokine are positively associated with the corticostriatal pathway in the users [241]. On the other hand, it was suggested that IL-6 plays a role in the protective effect of memantine in the CPP induced by cocaine [242].

Table 1.

Cytokines modulation by cocaine.

Name	Effect	Structure	Model	Refs.
IL-1β	No effect	PFC NAc	Chronic treatment in rodents	[120, 125, 283]
-	Increase	NAc Piriform cortex Somatosensory cortex HPC VTA	Plasma from patients Pericytes Acute or repeated treatment in rodents	[115, 120, 125, 251, 259, 284-286]
IL-2	Increase	-	Plasma from patients	[283]
IL-4	No effect	-	Plasma from patients	[287]
IL-6	No effect	NAc PFC Striatum	Plasma from patients after use Chronic treatment in rodents	[120, 286, 288]
-	Increase	-	Serum from patients Pericytes	[256, 285, 289]
-	Decrease	-	Chronic treatment in rodents	[283]
IL-8	No effect	-	Plasma from patients	[287]
IL-10	Increase	-	Plasma from patients Acute treatment in rodents	[285, 290, 291]
-	Decrease	Striatum	Plasma from patients Plasma from patients after use	[286, 289, 292, 293]
IL-17α	Decrease	-	Plasma from patients	[287]
IL-20	Decrease	-	Plasma from patients	[105]
IL-29	Decrease	-	Plasma from patients	[105]
INF-γ	No effect	-	Plasma from patients Acute treatment rodents	[287, 290]
-	Increase	-	Plasma from patients after use	[291]
-	Decrease	-	Plasma from patients	[291]
TNF-α	No effect	VTA NAc PFC Striatum	Repeated treatment in rodents	[115, 120, 286]
-	Increase	Striatum	Pericytes Repeated treatment in rodents	[36, 251]
-	Decreased	PFC NAc	Chronic and repeated treatment in rodents Plasma from patients	[242, 258, 283]

Table 2.

Cytokines modulation by amphetamines.

Name	Effect	Structure	Model	Refs.
Amphetamine and D-Amphetamine
IL-1β	Increase	-	Serum from patients	[294]
IL-2	Increase	-	Serum from patients
-	Decrease	-	Lymphocytes	[295]
IL-4	No effect	-	Lymphocytes
-	Increase	PFC Striatum	Chronic treatment in rodents Serum from patients	[294, 296-298]
IL-5	Increase	-	Serum from patients	[294]
IL-6	Increase	PFC Striatum	Chronic treatment in rodents Serum from patients	[294, 297, 298]
IL-8	Decreased	-	Serum from abstinent patients	[294]
IL-10	Increase	PFC Striatum	Chronic treatment in rodents Serum from patients	[294, 296-298]
-	Decreased	-	Serum from abstinent patients	[294]
TNF-α	Increase	PFC Striatum	Chronic treatment in rodents Macrophages Human endothelial cells	[296-298] [299, 300]
Methamphetamine
IL-1β	No effect	HPC Striatum PFC VTA Nac	Acute treatment in rodents Serum from abstinent patients Hippocampal neurons	[171, 238, 301, 302]
-	Decrease	-	Serum from abstinent patients	[302]
-	Increase	HPC PFC Nac Striatum VTA HPT	Repeated and acute treatment in rodents Highly aggressively proliferating immortalized (HAPI)	[140, 150, 184, 223, 244, 252, 264, 303-305]
IL-2	No effect	-	Serum from abstinent patients	[302]
-	Increase	-	Serum from abstinent patients
IL-4	No effect	-	Serum from abstinent patients
-	Decrease	-	Serum from abstinent patients
-	Increase	-	Serum from abstinent patients
IL-5	No effect	-	Serum from abstinent patients
-	Increase	-	Serum from abstinent patients
IL-6	No effect	Nac PFC	BV-2 cells Hippocampal neurons Serum from abstinent patients Acute treatment in rodents	[139, 171, 238, 302]
-	Increase	HPC PFC VTA Nac Striatum	Repeated and acute treatment in rodents Neuroblastoma (SY5Y) Astrocytes Highly aggressively proliferating immortalized (HAPI) Serum from abstinent patients Serum from rodents after acute or chronic treatment N9 cells	[136, 140, 150, 171, 183, 184, 223, 252, 264, 301, 302, 306-310]
IL-7	Increase	-	Macrophages Serum from abstinent patients	[232, 302]
IL-8	Increase	-	Macrophages Astrocytes	[232, 308]
IL-9	Decrease	-	Serum from abstinent patients	[302]
-	Increase	-	Serum from abstinent patients	[238, 302]
IL-10	No effect	-	Hippocampal neurons	-
-	Decrease	-	Serum from abstinent patients	[302]
-	Increase	-	Macrophages Serum from rodents after acute treatment	[232, 310]
IL-12p70	Increase	-	Serum from abstinent patients	[302]
IL-13	No effect	-	Serum from abstinent patients
-	Increase	-	Serum from abstinent patients
IL-15	No effect	-	Serum from abstinent patients
-	Decrease	-	Serum from abstinent patients
IL-17a	No effect	-	Serum from abstinent patients
IL-18	Increase	-	Serum from abstinent patients
IFN-γ	No effect	-	Serum from abstinent patients
	Increase	-	Serum from abstinent patients
TNF-α	No effect	Striatum VTA NAc PFC	BV-2 cells Hippocampal neurons Acute treatment in rodents	[139, 171, 238, 301]
-	Increase	HPC PFC VTA NAc Striatum Serum	Repeated and acute treatment in rodents Neuroblastoma (SY5Y) Macrophages Highly aggressively proliferating immortalized (HAPI) Human microvascular endothelial cells Serum from abstinent patients Serum from rodents after chronic treatment N9 cells	[136, 140, 150, 167, 171, 183, 184, 223, 243, 252, 264, 299, 301, 302, 306, 307, 309, 311-313]
3, 4-Methylenedioxymethamphetamine (MDMA)
IL-1β	Increase	HPC Frontal cortex HPT	Sub chronic or acute treatment in rodents	[194, 201, 314, 315]
IL2	Increase	-	Acute treatment in rodents Macrophages	[316, 317]
IL-4	No effect	-	Macrophages	[317]
IL-10	Increase	-	Serum from patients after use	[318]
TNF-α	Increase	Striatum	Acute treatment in rodents	[319]
-	Decrease	-	Macrophages	[317]

Similarly, the TNF-α knockout presented enhanced responses to reward, locomotor, and neurotoxic effects of METH. Also, exogenous TNF-α prevented these effects while potentiating dopamine reuptake [243]. Intriguingly, repeated cocaine administration activated microglia and enhanced TNF-α levels in the striatum. This was related to a decrease in synaptic strength in the glutamatergic neurons of the NAc core [36].

METH administration induces spatial memory deficits in the HPC, the main structure responsible for this task, and increases IL-1β levels [244]. More importantly, blocking IL-1β signaling prevented the deficits induced by METH [244]. However, neither IL-1β antagonists attenuated MDMA-induced hyperthermia [194], nor did their agonists impair locomotor sensitization induced by AMPH [245]. Notably, the increase in dopamine signaling in the NAc induced by cocaine is regulated by IL-1β/TLR4 in the VTA [125], suggesting an important role of this cytokine in modulating the reward properties of cocaine. In agreement with these reports, using the self-administration model, it was found that IL-1β mRNA levels increased in the VTA, and the antagonism of the IL-1β receptor in this structure reduced drug seeking [115]. More importantly, inflammatory responses to psychostimulants can also be part of a homeostatic response to the neurotoxic effect of these drugs. In this sense, exogenous IL-6 and TNF-α prevented METH-induced microglia death through the JAK-STAT3 pathway [246].

There may also be crosstalk between the cytokines in the context of drug use. For example, TGF-β is increased by cocaine in endothelial cells, which, in turn, down-regulates IL-8 [247]. In rats, cocaine withdrawal increased TGF-β expression in the NAc [248] and the dorsal HIP [249].

In addition to potential central mechanisms involved in the release of cytokines, it is important to notice that psychostimulants can decrease the integrity of the BBB. Cocaine affects the BBB integrity by increasing CXCL10 secretion by pericytes through c-Src kinase, activation of NF-kB pathway [250], the release of TNF-α, IL-1β, and IL-6 [251], and the induction of endoplasmic reticulum stress and autophagy mechanism [251]. Moreover, METH reduces tight junctions on the BBB [252, 253], induces the release of TNF-α, and increases vesicular transport across endothelial cells [254]. In addition, METH enhanced LPS-induced infiltration of macrophages, neutrophils, and leukocytes in the brain, and this was accompanied by higher levels of proinflammatory cytokines [255]. The hypothesis of peripheral immune infiltration is also supported by the correlation between circulating cytokines and the central effects of psychostimulants. For instance, IL-6 levels correlated with cognitive symptoms [256], IL-8 with the intensity of withdrawal symptoms [257], and IL-1β with the severity of the cocaine use disorders [258].

4.3.2. Chemokines

Psychostimulants modulate chemokine production. Human macrophages exposed to METH upregulate CXCL16, CXCL2, CCL20, CXCL1, CCL24, and CXCL13 and downregulate CCL7 levels, which return to basal levels by 24h [232]. Certain chemokines correlate with substance use disorder symptoms. In cocaine users, CX₃CL₁ levels correlate with symptom severity [258]. In knock-out mice for the receptor of this cytokine, Cx3Cr1-KO, impairments were observed in the extinction but not in the acquisition or reinstatement of cocaine-induced CPP. In addition, after acute cocaine exposure, CX₃CL₁ is enhanced in the HPC [259] and in the PFC [260]. The release of this chemokine was accompanied by increased phosphorylation of ERK1/2 and p65 NF-kB, and after repeated treatment, cocaine increased the phosphorylation of p38, MAPK, NF-kB and CREB [259].

Similar to CX₃CR₁-KO, CCR2-KO also presented impairments in cocaine-induced locomotor sensitization and reductions in ERK phosphorylation and dopamine transporter regulation in the striatum [261]. Besides, the antagonism of CCR2 prevents METH-induced CPP [262, 263]. The ligand of this receptor, CCL2, was also altered after acute cocaine administration in the PFC, together with the chemokine CCL7 [260]. The increase in CCL2 was also observed in the PFC [263] and the striatum [262-264] after administration of METH. In addition, D1R receptor antagonism prevents the increase in the striatum [262]. CCL7, which can act as a microglia inhibitor, modulates the CPP induced by METH through the D₁ receptor signaling [265]. However, the role of chemokine mediating psychostimulants' effects seems to depend on the brain structure. In this sense, icv administration of CXCL12 potentiates ambulatory and stereotypic activity induced by cocaine. However, the administration of CXCL12 into the VTA potentiated cocaine-induced ambulatory activity, which is abolished when the same cytokine is administered into the NAc shell. No effect was found when administered into the core. Finally, when CXCL12 microinjections were into the caudate putamen, it increased stereotypic behaviors induced by cocaine [266].

4.3.3. Cyclo-Oxygenase and Eicosanoids

Microglia are an important source of cyclo-oxygenase (COX) and eicosanoids, which can be increased during various pathological conditions [267]. The role of these mediators in addiction has been suggested by evaluating the expression of the enzyme and its products and investigating the effects of its inhibitors in animal models. For instance, chronic consumption of METH decreases COX-2 [268]. After 24h of METH treatment, there is a reduction in striatal COX-2, but no changes were identified in COX-1 levels [269]. On the other hand, AMPH increases cortical COX-2 expression [270, 271] and a COX-2 inhibitor was able to attenuate this increase as well as the hyperlocomotion induced by AMPH [270]. Similarly, systemic or intra-VTA treatment with a COX inhibitor, but not a lipo-oxygenase inhibitor, impairs cocaine- and AMPH-induced locomotor sensitization [272]. Conversely, a dose of COX inhibitor reduced hyperlocomotion, but it did not affect AMPH self-administration [273], suggesting a COX-specific involvement in the psychomotor effects of AMPH. Some evidence indicates that COX is implicated in abstinence. Notably, 72h after METH consumption, COX-2 was upregulated, and their levels correlated with prostaglandin (PG) E₂ [269]. Consistently, diclofenac, a COX-2 inhibitor, attenuated the cue-dependent drug reinstatement [274]. Contrary to these preclinical data, a clinical trial with celecoxib, a COX-2 selective inhibitor, showed poor improvements in cocaine use and craving [275].

The described alterations in COX and PLA pathways due to psychostimulants are consequently linked to modification in eicosanoid levels. Thromboxanes [276] and prostaglandin [277] were altered in vessels due to psychostimulant exposure, albeit another study did not observe these alterations [278]. This may rely on differences in experimental models. It has also been suggested that eicosanoids alter hyperactivity, stereotypy and other behaviors induced by AMPH [279-282].

CONCLUSION

The mechanisms by which psychoactive substances induce addiction are complex and depend on various factors like the type of the drug, environment, genetics, and others. In the case of psychostimulants, these substances may first increase dopamine, norepinephrine, and serotonin levels as they alter the reuptake of these neurotransmitters. In addition to monoamines, glutamatergic neurotransmission is also increased, affecting various cells. In neurons, a series of intracellular modifications could happen after activating dopamine receptors, including the recruitment of the cAMP/PKA/ERK pathways that activate different transcription factors, such as CREB, ΔFosB, MEF2, and NF-κB. Thus, genes involved in cell morphology, plasticity, and excitability could be expressed. As a result, alterations in the cell activity pattern, as well as its morphological aspects, could occur and be associated with the addiction process. Although acute exposure to psychoactive substances could lead to cell modifications, sustained exposure could alter neuronal networks that could be important for the onset and maintenance of drug dependence.

Various studies have shown that microglia may be important in this phenomenon as inhibiting these cells reduces some addiction-related behavior. Acutely, psychostimulants may activate microglia to produce inflammatory mediators that regulate neurotransmission and the activity of other cell types. It is still not clear the outcome of these initial events. However, chronic reexposure to the drug may contribute to the long-term alterations observed in the context of addiction.

Psychostimulants can, directly or indirectly, activate microglia. A direct activation could occur via the binding of psychostimulants to a microglial receptor, such as the TLR family, producing an inflammatory milieu in the CNS and neurotrophin release. In this sense, TLR4 activation by psychostimulants could induce the NF-κB translocation to the nucleus and promote the release of different cytokines, which, in turn, can modulate both synaptic plasticity and neuronal activity. For example, TNF-α release by microglia modulates induced behavioral sensitization by cocaine through AMPA/NMDA ratio alterations [36].

On the other hand, microglia possess neurotransmitter receptors, including D₁ and D₂, as well as glutamate receptors, which could be stimulated during the use of psychoactive substances. Activation of these receptors in microglia may be important not only for controlling the production of inflammatory and neurotrophic factors but also for synaptic changes. For example, the pattern of D₁ expression in the NAc can be mediated by the microglia [26]. In addition, psychostimulants can also modulate neuron expression and release of molecules related to synaptic pruning or synaptic formation by microglia. Collectively, these mechanisms could reorganize brain networks, cellular responses, neuronal communications, and neuro-immune communications, leading to addiction (Fig. 1).

Fig. (1).

Role of microglia in synaptic plasticity induced by psychostimulant abuse. Psychostimulants increase dopaminergic and glutamatergic neurotransmission, leading to neural plasticity. In addition, psychostimulants induce the neuronal release of ligands that mediate microglial cell activation, such as CX3CL1. In parallel, psychostimulants may directly bind to microglial TLR4 leading to cytokine and neurotrophic factors release, which, in turn, modulate neurotransmission and intracellular pathways that mediate synaptic plasticity. Created with BioRender.com.

In conclusion, increasing evidence indicates that microglia are involved in the pathogenesis of addiction, albeit the mechanisms are unclear. It is also essential to determine the role of these cells in either acute or chronic exposure to psychostimulants. This knowledge would contribute to the development of pharmacological therapies for the treatment of substance use disorder.

LIMITATIONS

Although some studies point to a potential role of microglia in drug addiction, which would suggest these cells as pharmacological targets in this psychiatric condition, the literature in this field is still scarce. Fewer studies have evaluated the psychostimulants effects in microglial cells number and morphology. In addition, very few studies have tried to understand the molecular mechanism that psychoactive substances can modulate microglia’s activity and how these cells can contribute or protect against drug-induced alterations. Although we described the alterations in microglia morphology induced by psychostimulants in in vitro studies, evaluation of these cell morphology in cell culture may be considered unreliable nowadays. Moreover, some papers have shown that alterations in microglia number or morphology in vivo can occur in acute or chronic exposure, but it is still unclear the timeline for these events. The molecular effects associated with inflammation induced by the drugs may not be dependent only on microglia, since there is an important communication between them and other glial or immune cells that release inflammatory factors. Finally, some data found in the current literature are also contradictory, which raises questions about how the psychostimulants activate these immune cells. In this sense, further studies are necessary for a better understanding of the role of microglia in drug addiction.

LIST OF ABBREVIATIONS

CD

Clusters of Differentiation

CNS

Central Nervous System

COX

Cyclo-Oxygenase

DAMPs

Damage-Associated Molecular Patterns

DAT

Dopamine Transporter

ERK

Extracellular Signal-Regulated Kinase

hGLUT

Human Glucose Transporter 5

LTP

Long-Term Potentiation

MEF2

Myocyte Enhancing Factor 2

MHC

Major Histocompatibility Complex

mPFC

Medial Prefrontal Cortex

MSN

Medium Spiny Neurons

NAc

Nucleus Accumbens

NO

Nitric Oxide

PAMPs

Pathogen-Associated Molecular Patterns

PG

Prostaglandin

PKA

Protein Kinase A

PRRs

Pattern Recognition Receptors

TH

Tyrosine Hydroxylase

TNF

Tumor Necrosis Factor

VMAT2

Vesicular Monoamine Transporters

CONSENT FOR PUBLICATION

Not applicable.

FUNDING

This work was supported by FAPEMIG, CNPq (Grant no. 310347/ 2018-1) and Coordenação de Aperfeiçoamento de Pessoal de Nivel Superior, Brazil (CAPES - Finance Code 001).

CONFLICT OF INTEREST

The authors declare no conflict of interest, financial or otherwise.