Potential Roles for Microglia in Drug Addiction: Adolescent Neurodevelopment and Beyond

Abstract

Adolescence is a sensitive period for development of addiction-relevant brain circuits, and it is also when people typically start experimenting with drugs. Unfortunately, such substance use may cause lasting impacts on the brain, and might increase vulnerability to later-life addictions. Microglia are the brain’s immune cells, but their roles in shaping neural connectivity and synaptic plasticity, especially in developmental sensitive periods like adolescence, may also contribute to addiction-related phenomena. Here, we overview how drugs of abuse impact microglia, and propose that they may play poorly-understood, but important roles in addiction vulnerability and progression.

Graphical Abstract

1: Introduction

Adolescence is a sensitive period for the developing brain, where behavioral traits and salient experiences can establish life-long proclivities. Drug experimentation also typically begins during adolescence, and negative outcomes in adulthood have been linked to adolescent drug use in humans (DuPont et al., 2018). Preclinical studies also show that exposure to addictive drugs causes persistent changes in the brain and behavior, including addiction-related behaviors. In this review we discuss the possibility that microglia may contribute to the development of addiction, and perhaps especially in the potential for adolescent drug use to persistently alter addiction-related brain circuits.

Microglia, the resident macrophages of the brain’s immune system, are best known for their roles in detecting and counteracting challenges to the central nervous system (CNS), such as threats from infectious agents, injury, or damage (Aloisi, 2001). However, microglia also play critical roles in sculpting and refining the development of brain circuits, especially during these circuits’ developmental sensitive periods including adolescence (Dziabis and Bilbo, 2021a, Schalbetter et al., 2022, von Arx et al., 2023). Microglia can also retain “memories” of encounters with prior challenges, including drugs of abuse, which can shape their future responses to the same or similar challenges when re-encountered. Since microglia are directly influenced by drugs of abuse, and since they help shape dynamic processes occurring during developmental sensitive periods, it is possible that microglia may play a role in the neurodevelopmental disruptions caused by adolescent drug use. Here we overview microglia functions in general, discuss how they may be affected by drugs of abuse, and conjecture how these impacts might contribute in underappreciated ways to addiction vulnerability and progression.

2: Adolescence as a Critical Period for Development of Addiction-Relevant Brain Circuits

Adolescence is a dynamic period in the development of frontal brain circuits, and for one’s life-long cognitive, emotional, and motivational traits and proclivities. During adolescence the brain undergoes remodeling of circuits relevant to these behavioral processes, which normally leads to a stable personality and behavioral tendencies by adulthood. For example, prefrontal cortical (PFC) systems continue maturing until early adulthood, so during adolescence, ‘top-down’ control over earlier-developing ‘bottom-up’ motivational and emotional systems is still relatively weak (Casey et al., 2008, Gogtay et al., 2004, Spear, 2013). This differential maturation supports social learning, development, and exploration, essential for fostering independence and facilitating the transition to adulthood (Crone and Dahl, 2012, Spear, 2000). However, it also leaves adolescents vulnerable to emotional and impulsive decisions and actions including experimentation with addictive drugs (Spear, 2000, 2013), and unfortunately there is considerable correlational evidence that early drug users end up with worse outcomes on a range of measures including an increased propensity for later-life drug problems (Kandel, 1975, Moss et al., 2014, Palmer et al., 2009). It is likely that underlying risk factors contribute to both later-life outcomes and early onset of drug use, but it is also plausible that drugs directly alter adolescent neurodevelopmental processes, and thus permanently alter the adult brain.

The adolescent brain undergoes several developmental changes that may be disrupted by drugs of abuse, especially in late-maturing brain regions like PFC. These include increases in myelination (de Faria Jr et al., 2021, Durston et al., 2001), decreases in synapse number (Huttenlocher, 1979, 1984), and changes in neurotransmitter signaling systems including endocannabinoids and other regulators of cortical GABA and glutamate excitation/inhibition (E/I) balance (Caballero et al., 2016, Caballero et al., 2021, Ellgren et al., 2008, Freund et al., 2003, Lovinger, 2008, Luna and Sweeney, 2004, Meyer et al., 2018, Selemon, 2013). These dynamic changes are thought to underlie emerging adolescent behavioral capabilities such as cognitive control, response inhibition, and social and emotional cognition (Caballero et al., 2016, Casey and Jones, 2010, Galván, 2010, Hoops and Flores, 2017, Jentsch and Taylor, 1999).

Preclinical studies show that various drugs of abuse can alter neurodevelopmental processes occurring in the adolescent brain (Salmanzadeh et al., 2020, Spear, 2016, Steinfeld and Torregrossa, 2023), and that adolescent drug exposure can lead to long-lasting changes in biological and behavioral phenotypes. Dopamine, a neurotransmitter that is intimately linked to reward learning and motivation, is potentiated by nearly all addictive drugs and is one system that may be affected by adolescent drug use (Koob and Volkow, 2010, Pierce and Kumaresan, 2006, Wise, 2002). Ventral tegmental area (VTA) dopaminergic axons innervate PFC during adolescence in rodents (Hoops and Flores, 2017, Manitt et al., 2011, Reynolds et al., 2018, Wahlstrom et al., 2010), and since activity in these developing circuits is likely related to their mature connectivity (Naneix et al., 2012), excessive drug-induced activation of dopamine signaling might alter the functional connectivity of these dopamine systems with other concurrently-developing PFC circuits involving glutamate, GABA, and others. There are numerous mechanisms by which drugs of abuse may persistently disrupt addiction-related brain circuits that develop during adolescence, and these have been reviewed elsewhere (Chadwick et al., 2013, Chambers et al., 2003, Mooney-Leber and Gould, 2018, Salmanzadeh et al., 2020, Spear, 2016, Steinfeld and Torregrossa, 2023).

Most work on the mechanisms by which drugs alter adolescent neural circuit development has understandably focused on the plasticity-inducing influence of drugs on neurons themselves. But neurons are not the only brain cells that are directly affected by drugs of abuse, nor are they the only cells involved in sculpting the adolescent brain into its adult configuration. In the next section, we explore the potential roles of another cell type with these characteristics—microglia—in addiction-relevant brain development.

3: Overview of Microglial Functions

Microglia are long-lived (Eyo and Wu, 2019, Lawson et al., 1992, Reu et al., 2017), dynamic cells that play crucial roles in maintaining homeostasis of the CNS by surveilling for, identifying, and coordinating responses to a wide range of disruptions (Davalos et al., 2005, Nimmerjahn et al., 2005, Tremblay et al., 2010, Wake et al., 2009). Like peripheral immune cells, microglia can “learn” about previously encountered threats, a process referred to as innate immune memory (Longhi et al., 2011, Netea et al., 2020). Furthermore, microglia also interact bidirectionally with neighboring neurons and synapses, sensing their activity, and influencing it in a variety of ways (especially during development).

Microglia originate from erythromyeloid progenitor cells in the embryonic yolk sac, and a subset of these colonize the CNS during embryogenesis, eventually becoming microglia (Alliot et al., 1999, Ginhoux et al., 2010, Gomez Perdiguero et al., 2015). Microglia distribute throughout the brain in a mosaic-like pattern, with each cell overseeing a specific, non-overlapping territory, enabling them to effectively monitor and cover the entire brain (Barry-Carroll et al., 2023, Davalos et al., 2005, Nimmerjahn et al., 2005). Microglia are very long-lived—in humans they survive from four to twenty years, and in rodents they may persist throughout the entire lifespan (Eyo and Wu, 2019, Lawson et al., 1992, Reu et al., 2017). Morphologically, microglia are characterized by the highly branched, ramified processes extending from the soma, which continuously survey the local brain environment using their chemosensory capabilities, allowing microglia to migrate toward the source of homeostatic disruptions (Kettenmann et al., 2011, Paolicelli et al., 2022, Stence et al., 2001).

3.1: Immune Response and Damage Repair

The most canonical role of microglia is to detect brain homeostatic threats such as pathogens, damaged cells, or injuries, which they do through a wide range of specialized receptors (Kettenmann et al., 2011). When such homeostatic threats are detected, microglia often adopt a pro-inflammatory or “activated” state to counteract and neutralize the disruption. This process typically is accompanied by morphological changes like increased size and roundness of the soma, and retraction of processes, leaving them thicker and shorter, and potentially facilitating cell migration to a site of disruption (Kettenmann et al., 2011, Paolicelli et al., 2022). When activated, microglia often then initiate threat responses, for example by releasing chemokines that recruit other immune cells to the site, releasing pro- and anti-inflammatory mediators including cytokines, oxygen species, and nitric oxide, or secreting other substances that can help restore homeostasis (Aloisi, 2001, Kettenmann et al., 2011). Microglia also engage in phagocytosis, or engulfment and removal of cellular debris or other potentially harmful agents. This “clean up” process limits further damage to affected cells and surrounding brain areas (Cunningham, 2013, Krasemann et al., 2017, Perry et al., 2010b, Smith et al., 2012).

Notably, this role for microglia in detecting and opposing homeostatic disruptions can sometimes cause dysfunction, as in certain neurodegenerative conditions. For example, excessive microglial phagocytosis of cellular components contributes to multiple sclerosis (Beiter et al., 2024, Hickman et al., 2018, Perry et al., 2010b, Subhramanyam et al., 2019, Xu et al., 2016), and microglia can promote pathological neuroinflammation and neuron damage in neurodegenerative disorders like Alzheimer’s, Parkinson’s, and amyotrophic lateral sclerosis (Hickman et al., 2018, Perry et al., 2010b, Subhramanyam et al., 2019, Xu et al., 2016). A role for these threat-detecting and -responding mechanisms of microglia in addiction disorders is currently not yet well understood, though this could be due to lack of investigation rather than lack of involvement.

3.2: Innate Immune Memory in Microglia

The innate immune system is responsible for triggering an immediate, generalized response to challenges like pathogens or tissue damage throughout the body, and in the brain microglia play this role. The innate immune system can “learn” based on initial exposure to a threat, exhibiting either enhanced (innate immune training) or attenuated (innate immune tolerance) responsiveness to the threat upon reencounter. This learning process is more generally known as innate immune memory (Neher and Cunningham, 2019, Netea et al., 2020, Schaafsma et al., 2015, Wendeln et al., 2018, Zhang et al., 2022). Innate immune memory in microglia may involve changes in their ability to identify threats to brain homeostasis, and to coordinate the nature and intensity of responses mounted against them. Microglial ‘priming,’ in which a primary inflammatory response leads to a larger secondary response upon re-exposure to the same or similar insults, and microglial tolerance, where a primary inflammatory response results in a diminished or attenuated secondary response, may both be mechanisms by which microglia form these innate immune memories (Neher and Cunningham, 2019, Netea et al., 2020).

These experience-dependent changes involved in CNS innate immune memory largely depend upon epigenetic modifications in microglia (Neher and Cunningham, 2019, Netea et al., 2020, Zhang et al., 2022). A well-known example of this process involves maternal immune activation, in which bacterial or viral infection during fetal development leads to either blunted or augmented microglial activation in response to related threats when they are re-experienced much later in the life of the offspring. Corresponding microglial transcriptional changes suggest that such “memories” persist into later life (Hayes et al., 2022, Schaafsma et al., 2017), facilitating microglial responses to subsequent insults, at least with regard to inflammation-relevant gene expression (Kim et al., 2024, Wendeln et al., 2018, Zhang et al., 2022). We note that drugs of abuse, such as THC and morphine, can also cause epigenetic alterations in microglia that impact their responses to subsequent challenges (Lee et al., 2022, Schwarz et al., 2011), suggesting that a process analogous to innate immune memory could also impact microglial responses to later encounters with addictive drugs—a possibility we consider further below.

3.3: Microglia Modulation of Neighboring Cells and Synapses

In addition to their roles in detecting, learning about, and opposing challenges to brain homeostasis, microglia also interact bidirectionally with neighboring neurons, especially during development. Though we focus on their interactions with neurons here, we also note that microglia secrete various molecules that impact functions of astrocytes and oligodendrocytes (Auguste et al., 2022, Bezzi et al., 2001, Jay et al., 2019, Jha et al., 2019, Liddelow et al., 2017, Liddelow et al., 2020, Matejuk and Ransohoff, 2020, Pascual et al., 2012, Shinozaki et al., 2017, Vainchtein and Molofsky, 2020), and microglia also react to signals from other glia that impact their activities (Domingues et al., 2016, Jha et al., 2019, Matejuk and Ransohoff, 2020).

Microglia interact with neurons and their synaptic connections in several ways, and these interactions are important for both normal and pathological processes. First, microglia can sense neuronal chemical secretions that relate to neural activity, and in response they may adjust their motility, position, or propensity to engulf synaptic material, thereby regulating local synaptic signaling (Davalos et al., 2005, Dissing-Olesen et al., 2014, Eyo et al., 2014, Liu et al., 2019, Nimmerjahn et al., 2005, Stowell et al., 2019). For example, microglia can detect neural release of neurotransmitters via their receptors that bind GABA, glutamate, adenosine triphosphate (ATP), norepinephrine, and others (Pocock and Kettenmann, 2007). When engaged, these receptors can influence microglia activity—for instance, norepinephrine binding to microglial β₂-adrenergic receptors inhibits process motility and surveillance (Liu et al., 2019, Stowell et al., 2019), and neuron-derived ATP binding at purinergic P2YR12 receptors causes microglia to transiently extend processes towards the source of ATP (Dissing-Olesen et al., 2014, Eyo et al., 2014, Fontainhas et al., 2011, Sipe et al., 2016, Wong et al., 2011). These transient responses of microglia to neural signals may, among other things, influence ongoing neural activity; for example, neuron-secreted ATP recruits microglia in a manner that appears to decrease the ongoing activity (Badimon et al., 2020, Cserép et al., 2020, Kato et al., 2016, Li et al., 2012, Liu et al., 2021, Merlini et al., 2021, Wu et al., 2020) or increase related neural network synchronization (Akiyoshi et al., 2018).

These findings suggest that microglia may be capable of detecting synaptic activity and extending their processes toward active synapses in order to influence ongoing activity and plasticity occurring there. For example, microglial release of tumor necrosis factor alpha (TNF-α) can modulate “synaptic scaling,” or changes in the number of postsynaptic glutamate and GABA receptors in existing synapses, thereby fine-tuning synaptic signaling (Lewitus et al., 2014, Pribiag and Stellwagen, 2013, Rizzo et al., 2018, Stellwagen et al., 2005, Stellwagen and Malenka, 2006). Microglia secretion of brain derived neurotropic factor (Parkhurst et al., 2013) and interleukin-10 (Lim et al., 2013) can also promote creation of entirely new synapses. Microglia may also facilitate “tagging” of specific synapses for future potentiation via release of neuroinflammatory factors (Frey and Morris, 1997, 1998, Raghuraman et al., 2019), while microglial release of TNF-α and interleukin-1β can conversely suppress future long-term potentiation of specific synapses (Hellstrom et al., 2005, Prieto et al., 2019, Schmid et al., 2009). In addition to these microglia-derived factors, the cytokine granulocyte colony-stimulating factor influences microglial function (Chitu et al., 2021, Stanley et al., 2023), enhances synaptic plasticity in drug-reward regions, and modulates cocaine-related behaviors (Calipari et al., 2018, Hofford et al., 2019, Kutlu et al., 2018). Furthermore, microglia may remove synapses entirely by phagocytosing elements such as dendritic spines (Chen et al., 2024), and they may also even physically obstruct synapses by interposing themselves between pre- and postsynaptic elements, thus functionally inactivating them (Chen et al., 2014, Haruwaka et al., 2024, Trapp et al., 2007, Wan et al., 2020). In addition, microglia may also impact synaptic plasticity processes by helping degrade the extracellular matrix in which stable synaptic elements are embedded, allowing synaptic growth or shrinkage to take place (Crapser et al., 2021, Liu et al., 2021, Nguyen et al., 2020, Venturino et al., 2021). In sum, the roles played by microglia in bidirectional communication with neurons and in influencing synaptic plasticity may be important though underappreciated, especially after early brain development and perhaps even into adulthood.

Reciprocal microglial interactions with neurons may thus be important for normal plasticity, and they are also likely important for disease states, including addiction. For example, in Huntington’s and Alzheimer’s disease, microglia appear to be abnormal in a variety of ways, though it is not yet clear whether these abnormalities are causal of disease (i.e., abnormal microglia initiate processes that lead to degeneration and dysfunction), or whether microglia are instead responding to and opposing underlying pathological processes (Hickman et al., 2018, Perry et al., 2010a, Subhramanyam et al., 2019, Xu et al., 2016). Obviously this is an essential distinction, with major implications for how such information may be leveraged clinically. It is even less clear what role microglia play in addiction and other psychiatric disorders (Blank and Prinz, 2013, Brisch et al., 2022, Zhu et al., 2023). This said, given the integral nature of plasticity and learning in the development and progression of addiction, it stands to reason that microglia may indeed participate, at least during developmental sensitive periods like adolescence.

3.4: Microglial Roles in Brain Development

Microglial interactions with neural circuits are especially important during developmental sensitive periods, when neural circuits undergo significant synaptic remodeling in response to genetic programs, as well as to species-typical environmental experiences. These developmental processes unfold in a time-dependent manner for nearly all brain circuits, though the timing of critical or sensitive periods for each circuit differs. A classic example of environmental inputs tuning circuit development during sensitive periods is the visual system, where normal sensory input from the eye is required during a specific developmental window in which visual perception circuits in visual cortex are established. If visual input is absent during this sensitive period, normal sight cannot recover even if sensory input from the eye is restored thereafter (Hensch, 2005, Hubel and Wiesel, 1962, Wiesel and Hubel, 1963). A similar developmental process occurs in other major sensory systems (Hensch, 2005), and likely also in the brain memory, emotional, and motivational systems relevant to addiction and other psychiatric disorders (Birnie and Baram, 2022).

Microglia play a key role in the execution of this experience-dependent developmental plasticity, especially via their ability to remove excessive or unused synaptic connections. They do so by indirectly or directly eliminating synapses with physical engulfment, phagocytosis, and trogocytosis (nibbling) of synaptic elements (Lim and Ruthazer, 2021, Paolicelli et al., 2011, Schafer et al., 2012, Tremblay et al., 2010, Weinhard et al., 2018), and other mechanisms that result in synaptic destabilization (Cheadle et al., 2020). However, the role of microglia in synaptic pruning remains debated, with some arguing that direct evidence is lacking relative to synaptic stripping or trogocytosis processes that are more clearly defined (Eyo and Molofsky, 2023). In addition, microglia may also be involved in facilitating synaptogenesis, and strengthening or preserving actively used synapses during development (Gallo et al., 2022, Miyamoto et al., 2016, Weinhard et al., 2018).

Microglia appear to respond to a range of neural signals that dictate the specific synapses that should be pruned, and those which should be preserved. One of these signals involves neurotransmitters released by neurons; for example microglia respond to neural GABA release, which influences their remodeling of inhibitory synapses made by cortical interneurons (Favuzzi et al., 2021, Logiacco et al., 2021). This mechanism may be especially important for proper inhibitory synapse development during cortical developmental sensitive periods (Gesuita et al., 2022, Hensch, 2005, Le Magueresse and Monyer, 2013, Southwell et al., 2010, Wu et al., 2012), which we will argue may be especially important for development of addiction-relevant brain circuits in adolescence. In general, the neural activity relevant to this developmental role for microglia is shaped by environmental experiences that impact neural activity at specific synapses. For examples, early life stress (Ahmed et al., 2024, Bolton et al., 2022, Dayananda et al., 2023), and adolescent sleep deprivation (Tuan and Lee, 2019, Wang et al., 2023) inhibit microglial engulfment of synaptic materials during sensitive developmental periods for stress and sleep circuits. In the case of early-life stress this leads to excessive excitatory synapses on hypothalamic corticotropin-releasing hormone expressing neurons and increased glutamatergic synapses in the hippocampal stratum radiatum (Ahmed et al., 2024, Bolton et al., 2022, Dayananda et al., 2023), as well as excessive excitatory synapses on pyramidal and granule cells within hippocampal CA1 and dentate gyrus in the case of sleep deprivation (Tuan and Lee, 2019, Wang et al., 2023). Such changes likely lead to long-lasting alterations in functions of these circuits.

Mechanistically, microglia contribute to synaptic remodeling in part by responding to complement protein signaling cascades, in which neurons (with contributions from astrocytes) produce complement proteins that localize at immature or unused synapses, and act as an “eat me” signal to microglia that detect them via their complement receptor 3. This process prompts microglial remodeling of tagged dendritic spines, and elimination of the tagged synapses (Han et al., 2023, Schafer et al., 2012, Stevens et al., 2007, Vainchtein et al., 2018). Other “eat me” signals expressed by synapses are identified by microglia via receptors including TREM2, GPR56, MERTK, and CX3CR1 (Filipello et al., 2018, Gunner et al., 2019, Li et al., 2020, Li et al., 2021, Meng et al., 2024, Paolicelli et al., 2011, Park et al., 2021, Scott-Hewitt et al., 2020). Conversely, structurally mature, active dendritic spines that should be preserved may produce “don’t eat me” signals that actively inhibit microglial phagocytosis (Lehrman et al., 2018).

We note that most work examining mechanisms by which microglia shape neural development comes from studies of embryonic and early postnatal developmental periods, when the brain is massively increasing the complexity of its interconnectivity, and laying down its basic architecture. However, mounting evidence suggests that microglia also play analogous roles in later developmental periods as well, and especially in developmental sensitive periods like adolescence. For example, we previously noted that PFC circuits underlying planning, decision making, and other addiction-relevant processes mature in adolescence, and microglia seem to be important for their adolescent synaptic remodeling and potentiation (Mallya et al., 2019, Schalbetter et al., 2022, Stowell and Wang, 2024, bioRxiv, von Arx et al., 2023).

In the next section we review evidence that these complex roles for microglia may play a range of roles in the processes underlying addiction, both via influencing synaptic remodeling that occurs during adolescence, and other processes throughout life that may be subject to perturbation by drugs of abuse.

4: Microglia: roles in drug addiction?

It has long been clear that microglia are impacted by drugs of abuse in experimental animals, which alter microglia morphology, gene and protein expression, and release of signaling factors. Some drugs directly act on microglia themselves, at microglial receptors binding the drugs (Lacagnina et al., 2017, Linker et al., 2019). When this occurs, microglial structure and activity may be altered, and inflammatory or anti-inflammatory processes may be directly induced. Direct drug actions on microglia can also induce epigenetic changes that may alter their subsequent reactivity to future challenges, analogous to innate immune memory. In addition to directly influencing microglia, drugs may also indirectly impact them by inducing neural and synaptic activity to which microglia respond, potentially in support of instantiating addiction-relevant synaptic plasticity. In other words, microglial roles in the complex process of addiction are themselves complex, and may involve multiple aspects of microglia function, and multiple stages of the progressive disorder of addiction (Fig. 2A).

Figure 2: Impact of Drug Exposure on Microglial Function and Resulting Brain Circuitry.

A) Drug interactions with microglia trigger neuroinflammatory responses, genetic alterations, and drug-related behavioral changes. B) Proposed hypothetical model that drug exposure during adolescence disrupts microglia, leading to epigenetic reprogramming of microglia that either primes microglia for a heightened response or reprograms them to have a tolerant, attenuated response to future insults (like subsequent drug use). In addition, drug use during development may also disrupt active sculpting of neural circuits and networks that normally occurs during developmental sensitive periods. Drug use may disrupt this process in various ways. Drugs may therefore impact microglia-related developmental processes in various ways that could influence addiction-related behavioral and cognitive functions in a persistent manner.

While this review focuses on microglial involvement in addiction-related processes, we note that astrocytes also play crucial roles in neurodevelopment and addiction-relevant behaviors. Astrocytes aid in developmental synaptic connectivity (Ango et al., 2008, Clarke and Barres, 2013, Risher et al., 2014) and regulate synapse elimination during development and adulthood (Chung et al., 2013, Lee et al., 2021, Yang et al., 2016). Microglia-astrocyte crosstalk is bidirectional, with microglia directing astrocytic synapse engulfment (Jay et al., 2019), and astrocytes modulating microglial engulfment (Vainchtein et al., 2018). Astrocytes influence glutamatergic transmission and plasticity in reward-related circuits including nucleus accumbens, PFC, and VTA, thereby shaping drug-seeking and relapse behaviors (Harder et al., 2024, Kim et al., 2018, Kruyer and Scofield, 2021, Linker et al., 2019, Scofield and Kalivas, 2014, Wang et al., 2022). Clearly the interplay of microglia and astrocytes should be further considered for their potential roles in adolescent drug susceptibility.

4.1: Effects of Drugs on Microglia Structure and Function

Certain drugs of abuse such as ethanol, opioids, and psychostimulants cause microglia to react in an overtly similar manner to how they respond to other threats—with an “activated” morphological phenotype, recruitment of inflammation-related gene expression profiles and proteins, and secretion of various substances including pro- or anti-inflammatory factors (Catale et al., 2019, da Silva et al., 2023, Lacagnina et al., 2017, Li et al., 2024a, Linker et al., 2019, Vilca et al., 2023). This said, it is not obvious that these active microglial responses are actually helpful in mitigating drug-induced harms, as would be the case if they were launched in response to threats like infectious agents, or injured tissues and cells. One possibility is that drugs of abuse simply “hijack” the endogenous receptors and signaling cascade systems used by microglia to detect and respond to naturalistic threats, mirroring the ability of drugs to “hijack” the neural networks of reward to cause addiction (Nesse and Berridge, 1997).

Regardless of the ultimate explanation for why this is the case, several drugs—ethanol, opioids, and psychostimulants—interact directly with pro-inflammatory receptors on microglia, leading to morphological and transcriptional “activation”. These drugs all bind to toll-like receptor 4 (TLR4) and the accessory receptor myeloid differentiation factor 2 (MD2) on microglia, triggering a pro-inflammatory cascade and transcription of proinflammatory regulators, similar to responses mounted in response to bacterial infections (da Silva et al., 2023, Hutchinson et al., 2010, Kawai and Akira, 2010, Liu et al., 2022, O’Neill, 2008, Wang et al., 2012, Wang et al., 2016). Opioids can also activate TLR4/MD2 independently of classical opioid receptor signaling, as (+)-naloxone, an inactive enantiomer of naloxone that does not bind to opioid receptors, blocks TLR4/MD2 signaling and reduces opioid-induced reward behaviors, including morphine conditioned place preference and remifentanil self-administration, without affecting classical opioid receptor-mediated analgesia (Hutchinson et al., 2012, Hutchinson et al., 2010). Activation of pro-inflammatory receptors by these drugs of abuse on microglia results in an “activated” microglial morphology, upregulation of pro-inflammatory genes and proteins (Li et al., 2024b, Vilca et al., 2024, Vilca et al., 2023, Wei et al., 2023, Yan et al., 2023), and release of pro-inflammatory cytokines (da Silva et al., 2023, Marshall et al., 2016, Marshall et al., 2013, McClain et al., 2011, Peng and Nixon, 2021). Furthermore, ethanol (Erickson et al., 2019, McCarthy et al., 2018, Warden et al., 2020) and methamphetamine (Kays and Yamamoto, 2019, Li et al., 2024b, Vilca et al., 2024) acutely upregulate genes involved in innate immune responses, such as those related to type I interferon signaling, toll-like receptor signaling, cytokine production, and stress responses, presumably in service of promoting a sustained inflammatory state in the brain. Likewise, opioid drugs such as morphine and oxycodone also upregulate inflammation-related genes and inflammatory and immune signaling pathways throughout addiction-relevant brain circuits including the ventral midbrain, amygdala, striatum, and PFC (O’Sullivan et al., 2019, Seney et al., 2021, Wei et al., 2023, Yan et al., 2023, Zhang et al., 2017). At least for certain drugs, the ability to recruit pro-inflammatory responses in microglia directly may therefore be relevant to how drugs may impact addiction-relevant brain functions via microglial mechanisms.

However, not all drugs of abuse elicit pro-inflammatory responses in microglia—Δ⁹-tetrahydrocannabinol (THC), nicotine, and some psychedelic compounds may promote anti-inflammatory or neuroprotective processes as well, and perhaps particularly when microglia are already activated by other threats such as bacterial infections. These drugs may shift microglia from a pro-inflammatory phenotype to an anti-inflammatory phenotype, which reduces inflammation by releasing anti-inflammatory factors and restoring brain homeostasis, potentially having secondary effects relevant to addiction. For example, chronic adolescent THC exposure can downregulate genes related to innate immunity in adulthood in response to bacterial infection, including those encoding pro-inflammatory cytokines and toll like receptor families (Lee et al., 2022). It is likely that THC achieves such effects on microglia primarily through its agonist actions at inhibitory cannabinoid type 1 & 2 receptors, which generally oppose pro-inflammatory microglial processes (Marinelli et al., 2023, Young and Denovan-Wright, 2022). This said, other studies have found morphological changes in various brain regions suggestive of initial microglial “activation” in response to adolescent or adult exposure to THC (Cutando et al., 2013, Freels et al., 2024, Gabaglio et al., 2021, Lopez-Rodriguez et al., 2014, Zamberletti et al., 2015), so further work is needed to clarify the acute and long-term impacts of THC on microglia. Likewise, nicotine induces a neuroprotective phenotype in reactive microglia by suppressing pro-inflammatory factors, such as LPS-induced TNF release, and promoting the release of anti-inflammatory cytokines, likely via microglial α7 nicotinic acetylcholine receptor activation (Linker et al., 2019, Soares and Picciotto, 2024). Similarly, the psychedelic compounds dimethyltryptamine (DMT) and psilocin attenuate microglial pro-inflammatory responses by downregulating pro-inflammatory factors, and upregulating anti-inflammatory ones during viral infection (Kozłowska et al., 2021,bioRxiv, Kozlowska et al., 2022).

One might speculate that directly anti-inflammatory effects of drugs on microglia may contribute to individuals’ motivations for using drugs by alleviating subjective discomfort associated with inflammation, and such anti-inflammatory effects could potentially be clinically useful in some circumstances, for example in psychiatric disorders with inflammatory components such as depression or addiction. Additionally, opioids like morphine can sometimes be neuroprotective via their influence on microglial cAMP-related gene networks. During naloxone-precipitated opioid withdrawal in opioid-dependent animals, cAMP-related genes are upregulated in microglia, and suppressing this response using chemogenetic inhibition of microglial Gi signaling increased the severity of withdrawal, suggesting this pathway in microglia normally opposes withdrawal-related processes (Coffey et al., 2022, Coffey and Neumaier, 2024). Finally, microglia may also adopt neuroprotective roles during recovery from chronic drug use. For example, after extended abstinence from chronic methamphetamine, gene expression patterns in striatal microglia are suggestive of processes promoting nervous system repair, and the restoration of normal excitatory/inhibitory balance and homeostasis (Vilca et al., 2024). These findings highlight the complexity of microglial responses to drugs of abuse as they may be involved in both protective and destructive processes relevant to addiction.

It is worth noting that drugs of abuse need not necessarily recruit fully coordinated pro- or anti-inflammatory responses in microglia—sometimes only a subset of the coordinated changes typical of response to an infectious threat are induced. For example, an “activated” morphology elicited by ethanol is not accompanied by cytokine release (Marshall et al., 2013, Peng and Nixon, 2021, Sanchez-Alavez et al., 2019), and cytokine release but not “activated” morphology results from repeated, escalating doses of morphine or morphine withdrawal (Campbell et al., 2013, Coffey et al., 2022). Perhaps it should not be surprising that drugs may elicit unusual or disintegrated patterns of responses from microglia relative to more naturalistic disruptions, and this possibility should be kept in mind when interpreting findings of microglial structural and functional outcomes and their relevance to addiction.

4.2: Effects of Drugs on Microglia Innate Immune Memory

Innate immune memory is the ability of microglia to “learn” from prior exposure to a specific threat, later enabling them to more easily detect and oppose that threat, and there is some evidence that microglia also “learn” from their experiences with drugs of abuse. Specifically, drugs can induce both transient and long-lasting epigenetic changes in microglia that alter their subsequent responses to insults such as immune activation elicited by bacterial proteins, or even social stress (Lee et al., 2022). But do these changes impact their responses to the same drug when it is re-encountered later, or to other addictive drugs of similar or even distinct chemical classes? If so, this altered microglial activity would likely have implications for development of addiction, which requires repeated drug taking. Consistent with this hypothesis, Schwarz and Bilbo (2013) showed that adolescent morphine increased adulthood reinstatement of morphine conditioned place preference, and led to a long-lasting increase in TLR4 mRNA and protein in microglia. These microglia also exhibited heightened TLR4 signaling and morphological activation in response to acute morphine in adulthood, suggesting that adolescent morphine primed the microglia for a more robust response to later drug insult. Co-administration of the glial inflammatory inhibitor minocycline during adolescent morphine treatment prevented both the morphine reinstatement and the increase in TLR4 expression, further suggesting that TLR4 upregulation may play a role in the enhanced reinstatement (Schwarz and Bilbo, 2013). A similar phenomenon is also seen with ethanol, in that binge drinking by adult rats causes stronger immune responses in animals with prior alcohol experience, as evidenced by an increase in microglial number, elevated TNF-α levels, and enhanced markers of microglial activation (Marshall et al., 2016). In sum, the idea of microglial “memory” for their experiences with drugs is largely speculative, but given these intriguing initial findings it is a possibility worth further consideration when considering microglial roles in addiction.

4.3: Effects of Drug-Induced Neural Activity on Microglia

Drugs of abuse may also indirectly impact microglia by influencing neural activity and synaptic plasticity to which microglia respond. For example, many addictive drugs activate dopamine neurotransmission in nucleus accumbens, PFC, and elsewhere, which occurs via a range of mechanisms that differ for each abused drug. Psychostimulants increase synaptic and extra-synaptic dopamine by blocking dopamine transporter on forebrain axon terminals of midbrain dopamine neurons, while other drugs instead increase firing of VTA dopamine neurons themselves, for example by exciting them via their nicotinic acetylcholine receptors (nicotine) or disinhibiting them via suppression of their inhibitory inputs (THC, opioids). Each of these drug actions are likely to have distinct effects on microglia based on the types of neurons and synapses involved, and the brain regions examined. An example of this is that during adolescence frontal cortical microglia respond uniquely to dopamine signaling, where phasic stimulation of frontal dopamine axons—similar to that occurring during rewarding experiences—promotes microglial contact with dopamine axons and subsequent dopamine axonal synaptic bouton formation (Stowell and Wang, 2024, bioRxiv). One might therefore expect that cortical microglial responses to dopamine elicited by different drugs via different mechanisms might have distinct consequences in terms of microglial activity, synaptic plasticity, or neuronal function. Microglial responses to drugs may also differ across brain regions, for example, microglia in the nucleus accumbens express dopamine 1 and 2 receptors, while those in the hippocampus do not (Schwarz et al., 2013). This could reflect either regional variation in microglia themselves, or differences in microglial responses to region- or drug-specific patterns of neural activity elicited by the addictive substances (Ayata et al., 2018, De Biase and Bonci, 2019, De Biase et al., 2017, Grabert et al., 2016, Schwarz et al., 2013). Regardless, it seems likely that drug-induced neural activity recruits microglial responses that contribute to addiction-relevant processes such as synaptic plasticity, as discussed in the next section.

4.4: Establishing Necessity of Microglia for Addiction-Relevant Neural Plasticity

Addiction depends upon plasticity within brain reward circuits, and microglia reciprocally interact with neurons in a manner that may contribute to the instantiation of physical aspects of synaptic plasticity such as engulfment of synaptic material, dendritic spine remodeling, and secretion of chemical signals (Durán Laforet and Schafer, 2024). These microglial-neural interactions clearly occur during development, but the extent to which microglia importantly contribute to adulthood neural processes (Durán Laforet and Schafer, 2024, Schafer et al., 2013), including drug-induced ones like synaptic plasticity, pharmacological tolerance, sensitization, or withdrawal is more controversial.

To what extent are microglia important for modulating plasticity with relevance to these aspects of addiction? This question is difficult to answer using mere examination of microglial morphology or gene and protein expression patterns that occur after drug use or exposure. The fact that drugs of abuse can induce changes in microglia does not prove that such changes are necessary for addiction-relevant processes or behaviors.

Instead, necessity of microglia in addiction-relevant drug effects may be interrogated by perturbing or eliminating them in vivo, and determining the consequences on drug-induced brain or behavioral outcomes. This approach will be bolstered by recent advances in our ability to genetically knock out (Krauser et al., 2015, Rojo et al., 2019) or perturb the activity of microglia in an inducible fashion, for example during specific developmental stages (e.g. adolescence), or after development in adulthood (Binning et al., 2020, Buch et al., 2005, Dheer et al., 2024, Green et al., 2020, Lv et al., 2024, Ma et al., 2024, Merlini et al., 2021, Yi et al., 2021). Though exciting, these novel methods are still technically demanding and not fully characterized, so relatively few studies have yet used them to ask questions about microglial involvement in addiction-relevant processes.

A more established approach for determining the necessity of microglia in addiction-relevant processes involves temporarily depleting the brain of nearly all its microglia by administering compounds that block microglial colony-stimulating factor 1 receptor (CSF1R). Constitutive activity at this receptor is required for survival of microglia in particular, and though CSF1R is also expressed on other myeloid cell types (Patel and Player, 2009) blocking CSF1R on these cells does not seem to significantly impair their function or survival (Elmore et al., 2014, Green et al., 2020, Hou et al., 2020, Krauser et al., 2015, Spangenberg et al., 2019). Several CSF1R inhibiting compounds, such as PLX3397 and PLX5622 (hereafter referred to as PLX), have been developed (Elmore et al., 2014, Green et al., 2020, Hou et al., 2020, Krauser et al., 2015, Spangenberg et al., 2019), and microglial depletion exceeding 90% is possible using these in mice (Elmore et al., 2014, Spangenberg et al., 2019), though this seems harder to achieve in rats (Linker et al., 2020, Riquier and Sollars, 2020, Riquier and Sollars, 2022, Sharon et al., 2022, Sharon et al., 2021). Notably, when microglia are cleared in this manner in adult animals, upon cessation of CSF1R blockade new microglia emerge and repopulate the brain (Elmore et al., 2014, Huang et al., 2018, Najafi et al., 2018), and eventually attain a similar brain-wide distribution and overtly similar morphological features and gene-expression profiles as seen prior to microglial clearance (Elmore et al., 2015, Rice et al., 2017).

4.5: Addiction-Relevant Effects of Microglial Clearance and Repopulation

The strategy of removing the brain’s microglia transiently to determine behavioral and brain effects, or of depleting microglia and allowing repopulation with new, drug-naïve microglia is promising for determining how these cells might play roles in shaping neural connectivity, drug-induced plasticity, or other addiction-relevant processes and behaviors. Several studies have asked how pharmacologically clearing the brain of microglia, or “resetting” microglia by clearing them and allowing new cells to repopulate the brain impacts addiction-relevant behavioral or brain outcomes (Fig. 1).

Figure 1: Impacts of Perturbing Microglia on Behavioral Responses to Drugs of Abuse.

Studies in which the necessity of microglia for addiction-relevant behaviors was tested by pharmacologically clearing, or otherwise transiently perturbing microglia are shown. Equation sign = no effect, red up arrow = increase, and blue down arrow = decrease. (Warden et al., 2021)¹, (Soares et al., 2024, bioRxiv)², (Warden et al., 2020)³, (Linker et al., 2020)⁴, (Adeluyi et al., 2019)⁵,(Coffey et al., 2022)⁶, (Kwok et al., 2024)⁷, (El Jordi et al., 2022)⁸, (da Silva et al., 2021)⁹, (Reverte et al., 2024)¹⁰, (Wu and Lai, 2021)¹¹, (Vilca et al., 2024)¹²

Drug addiction begins with problematic substance use, but pharmacologically clearing microglia has not been found to markedly alter the acute behavioral effects of drugs of abuse. For example, microglia depletion did not alter ethanol’s acutely sedative or motor discoordination effects (Warden et al., 2021), nor does it seem to affect the acute locomotor effects of cocaine (Wu and Lai, 2021), the rewarding and place preference-inducing effects of cocaine (da Silva et al., 2021, Reverte et al., 2024), or cue-induced reinstatement of methamphetamine seeking (Vilca et al., 2024). These findings imply that the primary behavioral effects of these drugs to induce reward and impact arousal and activity do not require microglia per se, despite the fact that these same drugs can directly engage microglial receptors and alter their structure and functions. Likewise, pharmacologically clearing microglia does not consistently alter learning about the reinforcing effects of psychostimulants, since sustained microglial depletion prior to and during cocaine conditioned place preference training does not reduce subsequent expression of preference for the cocaine-paired chamber in either previously cocaine-naïve mice (da Silva et al., 2021), or in mice that were abstinent from cocaine for a period following training (Reverte et al., 2024). This is consistent with the idea that memory of cocaine’s rewarding effects does not require microglia. Likewise, clearance of microglia following prior methamphetamine self-administration failed to alter context-induced reinstatement of methamphetamine seeking, further suggesting that memories for drugs of abuse do not depend upon presence of normal microglia (Vilca et al., 2024). These findings seem to suggest that neither the acutely reinforcing effects of drugs, nor the memories previously formed about them necessarily require microglia. This said, other experiences including exposure to certain stressors can also drive drug seeking in addicted individuals (Sinha, 2001, Weiss et al., 2001), and at least in female mice, a stressful shock that normally enhances binge-like ethanol drinking fails to do so following sustained microglia clearance (Soares et al., 2024, bioRxiv). This suggests that microglia may be required for drug seeking, at least when stress initiates it.

Acute drug effects seem mostly independent of microglia integrity, but other drug effects that occur upon repeated drug exposure—and which involve plasticity processes that may be regulated by microglia—seem to be more susceptible to disruption by pharmacological microglial clearance. For example, repeated dosing with nearly all addictive drugs causes a behavioral phenomenon called locomotor sensitization, or an enhancement of the locomotor activating effects of the drug when it is repeatedly administered. Locomotor sensitization has been interpreted as a proxy for another behavioral effect that may also sensitize with repeated drug intake—incentive motivation, or the augmentation of drug craving and seeking that often occurs with repeated drug intake (Robinson and Berridge, 1993, Thomas et al., 2008, Vanderschuren and Pierce, 2010, Vezina, 2007), and thus it may be relevant to addiction. Locomotor sensitization involves marked synaptic and structural plasticity in nucleus accumbens and other brain reward regions, and it is possible these plasticity processes involve microglia (Russo et al., 2010). Two mouse studies have examined how pharmacologically clearing microglia impacts the locomotor sensitizing effects of cocaine, and they have mixed results. The first found that ablating microglia using PLX prior to and during repeated cocaine dosing (15mg/kg) suppressed cocaine’s locomotor sensitizing effects (da Silva et al., 2021), but another mouse study found that a different dosing regimen of PLX failed to impact the sensitizing effects of either 10 or 15mg/kg cocaine, including when it is administered after a period of abstinence following repeated cocaine dosing (Wu and Lai, 2021). There are several differences between these studies that could explain these conflicting results, including route of administration, dose of PLX, duration of microglia depletion, and different mouse strains, so further work should examine under which conditions microglial clearance impacts cocaine sensitization.

Another critical component of addiction that emerges with repeated drug use is tolerance, or a decrease in a drug effect after repeated administration. Tolerance is often observed to the rewarding and intoxicating effects of drugs of abuse, which can lead to escalating drug intake to overcome this diminished effect. Microglia may play a role in this process, at least for the addictive drug ethanol. One study assessed voluntary ethanol intake in mice, which normally show escalated voluntary intake following induction of dependence via chronic ethanol vapor exposure (Warden et al., 2020). When these mice were pharmacologically cleared of microglia before, during, and after induction of ethanol dependence, no such escalation of voluntary intake was seen, despite the fact that the same microglial clearance protocol failed to impact drinking in non-dependent mice (Warden et al., 2021, Warden et al., 2020). Notably, when microglia were instead cleared only after dependence and escalated drinking had already been induced, no effects on intake were seen, further supporting a special role for microglia in instantiating dependence-related plasticity and escalation of drinking behavior, rather than mediating the acute effects of ethanol per se (Warden et al., 2021).

Other aspects of the roles for microglia in addiction-relevant processes may involve more specifically their functions in neuroinflammation, which results from actions of certain drugs themselves, and which also tends to occur upon withdrawal of drugs after chronic exposure to drugs like ethanol (Calcia et al., 2016, Melbourne et al., 2021). Excessive microglial-mediated inflammation may contribute to cognitive impairments, disruptions in synaptic plasticity, and susceptibility to neurodegenerative and psychiatric disorders that are induced or worsened by chronic drug use (Cornell et al., 2022, Cunningham, 2013, Réus et al., 2015). Interestingly, even when neuroinflammation is induced by an viral mimetic (poly:IC), ostensibly unrelated to drug effects or addiction, dependence-like escalated ethanol drinking behavior is observed—an effect that failed to occur in the absence of microglia due to maintained pharmacological clearance (Warden et al., 2021). Such findings suggest that microglia-dependent inflammatory processes may indeed contribute to addiction-like behavioral profiles.

Furthermore, chronic use of drugs can lead to neuroadaptations that generate profound sickness-like behavioral and subjective states when drug use is suddenly ceased, a phenomenon called withdrawal. Avoidance of withdrawal is a powerful motivator for continuation of harmful chronic drug use, especially for drugs like opioids and ethanol. Pro-inflammatory functions of microglia are involved in both physiological and psychological aspects of drug withdrawal. Microglia depletion following chronic nicotine exposure (Adeluyi et al., 2019), or prior to binge-like ethanol exposure (Warden et al., 2020), both blunted withdrawal-induced anxiety-like behaviors (Adeluyi et al., 2019, Warden et al., 2020), while decreasing pro-inflammatory, and enhancing anti-inflammatory gene expression in brain during withdrawal (Walter and Crews, 2017). Selectively depleting microglia via toxin-mediated ablation in locus coeruleus, a key region for withdrawal-induced negative affect, also attenuates opioid (morphine) withdrawal behaviors by promoting suppression of neural activity in this noradrenergic pathway (Kwok et al., 2024). However, another study in which microglia were partially pharmacologically cleared brain-wide in rats found that naloxone-precipitated withdrawal in oxycodone-dependent rats was not impacted by the ablation of microglia (El Jordi et al., 2022) (though it is possible that the 40% reduction of microglia number achieved in this study was not sufficient to induce observable changes). Another report using a chemogenetic approach to inhibit microglial G_i/o signaling showed that this manipulation worsens naloxone-precipitated morphine withdrawal behaviors, suggesting that inhibitory G protein signaling cascades in microglia may mitigate opioid withdrawal by altering inflammatory functions (Coffey et al., 2022, Coffey and Neumaier, 2024). In sum, several lines of evidence suggest that microglial modulation of inflammatory processes implicated in drug dependence and withdrawal may play a role in certain aspects of addiction, at least for certain classes of drugs.

To summarize, strong evidence of microglia being necessary for addiction-related brain processes to occur is relatively sparse, but the few studies that exist provide some support for this idea, especially in processes resulting from repeated drug use. We note that many key questions are yet unanswered, however. Do microglia play fundamentally similar roles in addiction to different drugs, or at different stages of addiction from abuse to dependence and withdrawal? Does the presence of, or sudden withdrawal from drugs like opioids, ethanol or methamphetamine preferentially induce neuroinflammatory states involving microglia? Do microglia “imprint” their experiences with drugs of abuse epigenetically, and do these “molecular memories” impact microglial responses to subsequent drug use in a manner relevant to addiction? What is the role of microglia in plasticity processes thought to underlie aspects of the disorder, and do microglia help execute structural plasticity induced by drugs throughout the lifespan? We have much to learn still about the nature, limits, and specificity of microglial involvement in addiction, especially in the adult brain following developmental sensitive periods in which microglia play a clearer role in the remodeling of synaptic connectivity.

In the remainder of this review, we discuss a particularly interesting mechanism by which microglia may participate in addiction—by helping mediate the persistent effects of drug exposure during developmental sensitive periods like adolescence on the life-long functioning of neural circuits underlying risk of developing drug addiction.

5: Drug-induced microglia perturbation: a mechanism by which adolescent drug exposure alters later-life susceptibility to addiction?

Drugs of abuse directly and indirectly influence microglia throughout the lifespan, and the mechanisms of these effects are likely to be qualitatively similar across development. For example, opioids (Gibson et al., 2022, Jantzie et al., 2020, Mills-Huffnagle et al., 2024, Smith et al., 2022a, Smith et al., 2022b) and ethanol (Aghaie et al., 2020) seem to morphologically “activate” microglia and their inflammatory functions similarly in the prenatal and adult brain.

Yet the lasting impacts of perturbing microglia with drugs appears to be greatest during sensitive periods of development like adolescence, which may reflect the special roles for these cells during these time-locked periods when processes like synaptic pruning refine synaptic connections in a manner that gets “stamped in,” potentially for life (Dziabis and Bilbo, 2021b, Fuhrmann et al., 2015). For example, prenatal opioid exposure leads to increased dopamine D1 receptor density in nucleus accumbens of male rats once they reach adulthood, in part because prenatally opioid-experienced microglia insufficiently prune D1 receptors during adolescence (Kopec et al., 2018, Smith et al., 2022b). These changes are not just cosmetic, as they may have lifelong consequences on addiction-relevant behaviors like propensity to extinguish oxycodone conditioned place preference, which is reduced in a manner consistent with addiction-like persistent drugs seeking (Smith et al., 2022b). It is therefore possible that drug-induced perturbation of microglia could contribute to the ability of drugs to impact brain development in a manner that could lead to long-lasting susceptibility to the addictive effects of drugs.

5.1: Adolescent Drug Exposure on Microglial Functions in Neural Circuit Development

We are particularly interested in the impact of drugs on microglia during adolescence, since this is a period in which drug use is often initiated in humans, and because adolescence is a sensitive period of development for key addiction-relevant brain circuits such as PFC (Caballero et al., 2016, Kolk and Rakic, 2022). Adolescence features PFC refinement of both GABA interneurons that develop their inhibitory control of pyramidal cells, and of PFC dopamine inputs that first infiltrate frontal cortex at this time (Caballero et al., 2016, Caballero et al., 2021, Caballero and Tseng, 2016, Hoops and Flores, 2017). Both PFC local inhibition and dopamine modulation are well-known to regulate cognitive and motivational functions that mature in adolescence, and both are highly implicated in addiction vulnerability and progression (Casey et al., 2008, Spear, 2013). As in other developmental sensitive periods, PFC adolescent maturation seems to involve microglia (Favuzzi et al., 2021, Mallya et al., 2019, Stowell and Wang, 2024, bioRxiv), so understanding how drugs impact microglia neurodevelopmental functions may be key for understanding how adolescent drug use leads to long-term negative outcomes including risk of later-life substance use disorders.

A likely mechanism by which this occurs is for adolescent drug use to perturb the normal developmental remodeling of synapses by microglia that occurs specifically in the PFC in adolescence (Fig. 2B). As in other sensitive periods of development, adolescent PFC synaptic reorganization is transiently sensitive to relevant environmental stimuli, and once the adolescent sensitive window is closed, these stimuli fail to have the same lasting, organizational consequences. Therefore, it is possible that drugs of abuse perturb, via direct or indirect mechanisms, the normal synaptic remodeling that needs to occur during adolescence. Once the sensitive window closes, the re-wiring that should occur may no longer be possible, so even if further drug use is ceased, synaptic connectivity in PFC may remain immature. Indeed, at least one paper (Linker et al., 2020) has explored this possibility, finding that adolescent pre-treatment with nicotine increases the subsequent propensity of rats to self-administer cocaine. This seems to occur via a fractalkine-dependent dopaminergic signaling process initiated by nicotine during adolescence, which causes an “activated” microglia phenotype in the nucleus accumbens and basolateral amygdala, and excessive engulfment of presynaptic inputs to nucleus accumbens (Linker et al., 2020). When microglia are pharmacologically cleared during the adolescent period in which nicotine is experienced, no such augmented cocaine taking is seen, suggesting that microglia are necessary for this pro-drug phenotype to manifest. The relevance of microglia to pro-addiction phenotypes induced by adolescent exposure to other drugs is less clear, but this is a topic worthy of further investigation.

5.2: Microglial “Memory”: Lasting Effects of Adolescent Drug Exposure

Since microglia form and retain innate immune memories, we have suggested that drugs might likewise epigenetically imprint microglia and influence their subsequent responses to drugs or other challenges later in life. Indeed, there is some evidence that microglia are epigenetically altered by adolescent drug exposure in a manner that can have long-lasting effects on addiction-relevant processes. Conceptually, this means that adolescent drug exposure could prime microglia such that when addictive drugs are re-experienced later in life, these drug-primed cells may respond differently than drug-naïve microglia. This could result, for example, in potentiation of microglial involvement in drug-induced synaptic changes that underlie the transition from substance use to addiction when drugs are re-encountered. We are aware of no direct evidence for this hypothesis to date. However, a recent report suggests that adolescent exposure to THC epigenetically alters microglia, causing persistent suppression of their responses to non-drug challenges. Specifically, adolescent THC enduringly suppresses inflammatory gene expression profiles and causes hypoactive reactivity to adulthood infectious agents and social stress (Lee et al., 2022).

It is not clear to what extent such microglial “drug memories” differ in their sensitivity or importance based on the developmental stage at which they are formed (Lajqi et al., 2019, Lajqi et al., 2020)—drug-induced epigenetic alterations may or may not be induced preferentially in microglia during developmental sensitive periods like adolescence. We note that neonatal ethanol exposure persistently primes hypothalamic microglia such that they mount an excessive pro-inflammatory response to a bacterial challenge that lasts until adulthood (Chastain et al., 2019). This said, microglia activated by a prior ethanol binge in adulthood also show enhanced pro-inflammatory responses as demonstrated by potentiated immunoreactivity, increased microglial number, and elevated TNF-α concentration in response to a subsequent ethanol binge (Marshall et al., 2016). Regardless, it stands to reason that such drug-induced epigenetic imprinting, whenever it occurs, could cause lasting differences in microglial responses to subsequent drug exposures, and thereby potentially impact their addiction-relevant functions (Fig. 2B).

5.3: “Resetting” Microglia: Reversing Adverse Effects of Adolescent Drug Exposure?

Regardless of whether microglia are especially susceptible to drug-induced epigenetic imprinting during sensitive developmental periods, there is reason to predict that the prior experiences of these long-lived cells may influence their subsequent actions when drugs are re-encountered. One way to test the proposition that these microglial “memories” are necessary for augmented responses to drugs following prior drug exposure is to “reset” microglia by pharmacologically clearing them, and allowing the brain to repopulate with new, drug naïve cells. If developmental drug-induced phenotypes like vulnerability to the addictive effects of drugs are no longer present when microglia are “reset” in this way, this provides evidence that drug-induced alterations in microglia themselves, rather than just the neural circuits they prune (or fail to properly prune) during development may contribute to the progression of addiction-like behaviors. This hypothesis is admittedly speculative and is not mutually exclusive with the idea that drugs also disrupt microglia-dependent synaptic remodeling that should occur during adolescence, with both processes contributing to persistent effects of youthful drug use on the brain and behavior.

Indeed, even if “resetting” microglia in this way does reverse adolescent drug-induced phenotypes, it is not certain that this occurs directly because of the mere replacement of drug-experienced microglia with drug-naïve ones. It is also possible that clearing and repopulating the brain with new microglia using CSF1R antagonists or other methods might essentially re-open developmental sensitive periods and allow microglia to conduct synaptic refinements that normally do not occur post-development. Indeed, some evidence supports this possibility, in that some synaptic modifications normally conducted by microglia during development can occur in adulthood when microglia repopulate after clearance (Milinkeviciute et al., 2021), and “resetting” microglia in aged brains restores expression of actin cytoskeleton and synaptic function genes to levels otherwise seen only in younger brains (Elmore et al., 2018). In other words, reversal of adolescent drug-induced phenotypes by microglia “resetting” might occur due to either epigenetic or synaptic modulatory mechanisms. Going forward, the biological consequences, and potentially clinical usefulness of microglial clearance and/or resetting should be investigated further in pursuit of novel means of addressing disorders exacerbated by developmental insults like adolescent drug use.

6: Summary

Many basic principles remain to be discovered about how microglia are involved in vulnerability and progression of addiction, both during development and throughout the lifespan. Unraveling these mechanisms may lead to novel strategies for preventing and treating addiction, or other related and often comorbid disorders. We have discussed some possible ways in which microglia may contribute to development and expression of addiction, highlighting the many complexities and variables that likely mediate this role. Factors like the drug of abuse in question, its dose, the degree of and timing of exposure to it, and the sex and species of subjects under investigation make understanding this question a challenge, but the lack of widespread study of this topic might also be considered an opportunity as the field progresses.

It is clear that microglia play still underappreciated roles in brain development, as well as in adulthood brain processes. Many of these extend far beyond the classic role of microglia in fighting infection and repairing damage in the CNS. Though the scope and nature of their role in complex disorders like addiction remains largely conjecture, the possibility for them being a missing piece of the puzzle, especially when it comes to the developmental perturbations caused by drug use during adolescence, seems promising, and a topic worthy of considerable further investigation.

Highlights:

• Adolescence is a key period where the brain is highly vulnerable to drug exposure.

• Microglia regulate neural activity and synaptic plasticity.

• Microglia exhibit innate immune memory, leading to priming or tolerance to challenges.

• Adolescent drug use may disrupt microglia or their functions, increasing later drug risk.