Keywords

FormalPara Bullet Points

  • Immune surveillance, antigen recognition, and phagocytosis are key functions of microglia during the initiation, propagation, and/or resolution of inflammation.

  • Innate immune activation of microglia can be triggered by numerous pathways including pattern recognition receptors and chemokine receptors.

  • Microglial cells participate actively during innate immune responses, but also activate pathways that contribute to adaptive immunity.

  • The physiological importance of microglia in inflammasome activation during central nervous system (CNS) infection and neurodegeneration highlights potential venues for CNS pathogen clearance.

  • Functions of microglia do not occur in isolation but are influenced by neighboring cells including astrocytes and neurons, and the blood–brain barrier.

1 Regulation of CNS Innate Immune Reactions

Microglia are the resident mononuclear phagocytes of the CNS, comprising approximately 10–15 % of the total cell population within the parenchyma. These cells share many phenotypic and functional characteristics with macrophages and, as such, are major players in the brain’s innate immune responses. Developmental studies in mice have shown that microglia are derived from primitive yolk sac myeloid progenitors seeding the brain parenchyma between embryonic days 9 and 10 (Alliot et al. 1999; Ginhoux et al. 2010), indicating that microglia constitute a population of cells ontogenically distinct from tissue macrophages, which originate from bone marrow-derived monocytes (Fogg et al. 2006; Parwaresch and Wacker 1984).

1.1 Immune Surveillance

Historically, the CNS was considered an immune-privileged site based on its lack of lymphatic drainage, low levels of major histocompatibility complex (MHC) expression, which is essential for antigen presentation (see below), and restricted diffusion of molecules and cells from the periphery by the blood–brain barrier (BBB). However, the healthy CNS is now recognized as a site where immune surveillance occurs (Greenwood et al. 2011). Microglia share several markers with bone marrow-derived macrophages (Prinz et al. 2011), including the fractalkine receptor CX3CR1 and the lectin Siglec-H (Gautier et al. 2012; Harrison et al. 1998). It has been proposed that the expression of CX3CR1 and its ligand CX3CL1 (fractalkine or FKN) on microglia and neurons, respectively, assists in chemotactic migration of microglia to neurons during neuronal injury or inflammatory conditions (Chapman et al. 2000; Harrison et al. 1998; Mizutani et al. 2012; Streit et al. 2005) (see Chap. 9 for further reading on their roles in the healthy brain). This highlights the use of traditional immune chemotactic signals to facilitate specific interactions between CNS cellular constituents.

Immune surveillance within the CNS occurs during normal physiological conditions and microglia play a critical role in this process. Although microglia were previously considered to exist in a “resting” state characterized by a ramified morphology, elegant studies have demonstrated that in the intact adult brain, microglia are highly active and continuously surveying their microenvironment by extending and retracting their motile processes (Davalos et al. 2005; Nimmerjahn et al. 2005) (see Chaps. 2 and 4 for further reading). This attribute may be essential to elicit immediate responses to neuronal injury or invading pathogens.

Microglia exist in at least two morphologically distinct states based exclusively on their appearance; namely, amoeboid and ramified. The amoeboid state refers to cells with larger cell bodies as well as shorter and thicker processes, a phenotype which has been correlated with microglial “activation” in response to neuronal injury and inflammatory conditions. Conversely, “quiescent” microglia are generally described as ramified, exhibiting longer and thinner processes. Interestingly, microglia possess an amoeboid morphology during early stages of CNS development, transitioning into a ramified form during the late fetal and early neonatal periods (Boya et al. 1979). Ramified microglia transform again into the amoeboid-like morphology after CNS insults, such as injury or inflammation (Streit 2002; Xiang et al. 2006) (Fig. 5.1). However, in some models of inflammation the correlation between microglial morphology and function is not as strong. For instance, low doses of systemic lipopolysaccharide (LPS, a major component of the outer membrane of gram-negative bacteria) can induce the production of pro-inflammatory cytokines without any apparent changes in morphology (Sierra et al. 2007) (Table 5.1). Among other mechanisms, microglial remodeling, swelling, and process extension were shown to be controlled by chloride channels (Ducharme et al. 2007; Zierler et al. 2008) and the phosphorylation/dephosphorylation of cofilin, a member of the actin depolymerizing factor (ADF) family (Hadas et al. 2012), while the inflammatory response is regulated by the transcription factor NFκB (Wilms et al. 2003). Although it remains unclear whether these two pathways interact, disruption of the actin cytoskeleton has been shown to activate NFκB in cultured human intestinal epithelial cells, which supports the notion that cytoskeletal disruption may occur upstream of NFκB (Nemeth et al. 2004).

Fig. 5.1
figure 1

Microglia and innate immunity. Antigen recognition, phagocytosis, inflammatory mediator release, and antigen presentation are among the most recognized functions of microglia in health and disease. Their Toll-like receptors (TLR) and danger-associated molecular patterns (DAMPs) receptors play an important role in recognizing antigens such as myelin byproducts, debris, or amyloid-beta (Aβ), making their engagement critical for sensing cellular damage. Membrane receptors such as CD68 and T cell immunoglobulin mucin-4 (TIM-4) are involved in the processes of phagocytosis and clearance of damaged cells. In general, antigens presented through MHC class II lead to adaptive responses that elicit Th1, Th2, or Th17 responses. In parallel, antigen recognition activates the inflammasome. The pro-inflammatory cytokine interleukin (IL)-1β is implicated in the pathophysiology of numerous neurodegenerative diseases, including multiple sclerosis/experimental autoimmune encephalomyelitis (EAE) and Alzheimer’s disease (AD), and a number of central nervous system (CNS) infections ranging from bacterial meningitis, brain abscess, and human immunodeficiency virus (HIV)-associated dementia. The initial response upon TLR activation (signal 1) in certain pathologies with further inflammatory cues leads to IL-1β release and inflammasome activation (signal 2) that in turns activates TLR components creating a positive feedback regulation of the inflammatory cascade. Although the roles of inflammasomes are becoming increasingly well-defined, insights into inflammasome biology in the CNS are only just emerging. Interactions between TLRs and the inflammasome mediate responses to both pathogen-associated and endogenous triggers during CNS pathology

Full size image
Table 5.1 Microglial cytokinesa
Full size table

1.2 Antigen Recognition in the CNS

Microglial responses are often measured based on the array of cytokines, chemokines, and complement components produced during specific pathologies (Tables 5.1 and 5.2). Cytokines are small signaling molecules that circulate within and among organs, affecting nearly every biological process by orchestrating the degree of inflammatory reaction. Cytokines are broadly categorized as pro-inflammatory (e.g., interleukin (IL)-1β, IL-6, and tumor necrosis factor α (TNFα)) and anti-inflammatory (e.g., transforming growth factor β (TGFβ) and IL-4), and further stratified based on their capacity to promote (1) cellular immunity (i.e., the activation of phagocytes and antigen-specific T lymphocytes, such as helper (Th)1 cells, which produce mainly interferon γ (IFNγ) and IL-12); (2) antibody-mediated responses (mediated by Th2 cells and associated with cytokines such as IL-4, IL-10, and IL-13); and (3) responses mediated by Th17 cells (distinguished by their release of IL-17, associated with pathogen clearance and tissue inflammation in autoimmune diseases) (Dinarello 2007) (see Chap. 16 for further reading on their roles in multiple sclerosis). Chemotactic cytokines (or chemokines) participate in cellular activation by arresting and positioning relevant cells with spatiotemporal precision (Ransohoff 2009). The complement system also constitutes one of the first defense mechanisms by coating the surface of cells and pathogens, thereby assisting in the process of phagocytosis (see Chap. 9 for more information on its involvement with synaptic pruning). Therefore, the coordinated control of soluble mediators by microglial cells is considered a key determinant in the processes of antigen recognition in the CNS (Tables 5.1 and 5.2).

Table 5.2 Microglial chemokinesa
Full size table

Innate immune activation in the CNS can be triggered by numerous pathways upon recognition of invading pathogens and/or tissue damage by pattern recognition receptors (PRRs). Within the last 10 years, much attention has been focused on Toll-like receptor (TLR) activation in several models of CNS infection, neurodegenerative disease, and injury (Hanamsagar et al. 2012). TLRs are stimulated by pathogen- or danger-associated molecular patterns of pathogen or host origin (PAMPs and DAMPs), respectively, leading to NFκB and MAPK activation, and the subsequent release of pro-inflammatory mediators. A total of 13 TLRs have been identified in human, all of which are expressed at the cell surface except TLRs 3, 7, and 9, which are exclusive to the endosomal compartments.

Microglia express a wide range of TLRs to varying degrees depending on their phenotype (Olson and Miller 2004). In the resting state, many TLRs cannot be detected in microglia; however, TLRs are rapidly induced upon cellular activation (Bsibsi et al. 2002) (Fig. 5.1). In microglia, TLRs are considered a crucial first line of defense against bacteria and viruses, as well as recognition of endogenous danger signals (Kielian et al. 2005; Ravindran et al. 2010; Zhou et al. 2006). Resultant pro-inflammatory mediator and reactive oxygen species (ROS) production triggered by TLR ligands can be either beneficial or detrimental to the host. For example, TLR9 activation can enhance microglial phagocytosis of amyloid-beta (Aβ) as reported in primary neuron-microglia cocultures (Doi et al. 2009), and TLR2 signaling is pivotal for nitric oxide (NO) production and anti-bacterial responses following Streptococcus challenge, leading to neuronal cell death (Lehnardt et al. 2006). Conversely, persistent TLR activation may cause exaggerated immune responses, resulting in damage to neurons and oligodendrocytes (Lehnardt et al. 2002; Lehnardt et al. 2003). Microglial activation is evident following a single systemic LPS injection, which also triggers the expression of numerous pro-inflammatory cytokines and chemokines (Chen et al. 2012; Lehnardt et al. 2002; Rivest 2009). Although microglial activation was originally thought to result from the action of systemically induced cytokines elicited by TLR4 engagement, strong evidence suggests that TLR4 expression within the CNS is essential for eliciting central inflammation either via direct activation of resident microglia or indirectly through cerebral vascular endothelial cells (Chen et al. 2012; Lehnardt et al. 2002). In addition, injection of LPS directly into the brain induces robust and transient expression of numerous inflammatory mediators, including IL-1, TNFα, IL-6, the subunit p40 of IL-12 (IL-12p40), and TGFβ through TLR4 and its adapter, MyD88 (Glezer et al. 2007b). Although astrocytes, endothelial cells, and neurons have been shown to express TLR4 (Bowman et al. 2003; Liu and Kielian 2011; Rolls et al. 2007; Zhou et al. 2006), their role in immunity in response to TLR4 ligands is considered to be limited as compared to microglia (Glezer et al. 2006; Glezer et al. 2007a).

Chronic neurodegeneration also leads to microglial activation (Dheen et al. 2007; Nakajima and Kohsaka 2001) (see Chap. 18 about microglial involvement in neurodegenerative diseases). Aβ, the principal component of Alzheimer’s disease (AD)-associated senile plaques, has been shown to induce production of IL-1β, IL-6, TNFα, and IL-18 in white and gray matter microglia cultured from human autopsy tissue (Lue et al. 2001a; Rogers and Lue 2001). IL-1β and IL-18 are produced after a series of signaling events that involve Nod-Like receptors (NLRs), inducing the formation of a multi-cellular complex known as the inflammasome (see below). It was shown that Aβ can activate microglia through the NLRP3 inflammasome pathway (Blasko et al. 2004; Halle et al. 2008) and the subsequent production of IL-1β has been proposed to worsen neurodegeneration (Lindberg et al. 2005). In Parkinson’s disease (PD), α-synuclein aggregates are considered central to the pathology, associated with a progressive neuronal loss in the substantia nigra pars compacta (SNpc). In fact, it has been shown that microglial activation is widespread in the SNpc (McGeer et al. 1988), but also in the hippocampus, putamen, and cingulate cortex of PD patient brains (Imamura et al. 2003). A correlation has also been demonstrated between microglial activation and the loss of dopaminergic terminals in the postmortem midbrain of PD patients (Imamura et al. 2003). Indeed, microglia exposed to α-synuclein in vitro express high levels of IL-1β, TNFα, and IL-6 (Su et al. 2008) and a recent study has reported that α-synuclein can also activate the inflammasome (Codolo et al. 2013).

1.3 Phagocytosis: Scavenging Functions/Clearance

Phagocytosis is an evolutionarily conserved process serving to remove cellular debris and dying cells or aggregated proteins that can be toxic and trigger exaggerated immune responses. It refers to the recognition, engulfment, and degradation of different types of cargo including dead cells, debris, bacteria, Aβ, dendritic spines, etc. (Sierra et al. 2013). From Drosophila (Kurant 2011) to higher eukaryotes, efforts have been directed at understanding the molecular basis underlying the phagocytosis of endogenous debris and pathogens. Understanding antigen uptake and the pathways that lead to phagocytosis and antigen presentation for proper T-cell activation has also been a topic of study for many years. Microglia are considered a first line of defense against invading pathogens and they respond rapidly to inflammatory stimuli. A primary function of microglia is to eliminate harmful exogenous insults as well as endogenous proteins or cellular debris not only during CNS injury, but also during development and adult neurogenesis (Sierra et al. 2013). There is ample evidence that microglia can phagocytose neurons during neuroinflammatory conditions. Generally, neuronal debris and degenerating axons are readily engulfed by microglia utilizing the scavenging receptor CD68 (Fraser et al. 2010; Tanaka et al. 2009). It is thought that the nucleotide ATP released from dying cells is not only an important trigger for microglial migration to sites of injury, via purinergic receptor and inflammasome activation, but also regulates the phagocytic process itself (Koizumi et al. 2007). Upon axonal degeneration, it was also shown that microglial release of type I interferons (IFNα and IFNβ, which are important for protection against many viral infections) can provide additional inflammatory feedback to maintain phagocytic activity, while creating a more permissive environment for axonal outgrowth in vitro (Hosmane et al. 2012). Interestingly, activated microglia have been shown to phagocytose viable neurons (Fricker et al. 2012), an in vitro observed phenomenon termed “phagoptosis” which is triggered by inflammation and the transient and reversible exposure of membrane phosphatidylserine (PS) on stressed cells. However, the physiological implications of this process remain unclear (Brown and Neher 2012). Other work has shown that microglia play a neuroprotective role during CNS injury by their rapid engulfment of apoptotic polymorphonuclear neutrophils (PMN) (Neumann et al. 2008), in addition to inducing T-cell apoptosis (Magnus et al. 2001)—the latter representing a potential mechanism to subvert autoimmunity. Besides the role of microglia in the phagocytosis of bacterial and viral particles during CNS infections, several studies have highlighted the importance of microglial phagocytosis in response to Aβ deposition in the brain. For example, microglia have been shown to clear Aβ plaques through Fc receptor-mediated phagocytosis and peptide degradation (Bard et al. 2000), although they do not seem to reduce overall the Aβ load in the AD brain. Other surface receptors implicated in fibrillar Aβ engulfment include the scavenger receptor CD36, α6β1 integrin, and the integrin and thrombospondin-1 receptor CD47 (Koenigsknecht and Landreth 2004).

Another interesting area in the process of phagocytosis relates to the pathways activated during physiological versus inflammatory conditions. Initial approaches using opsonized beads (i.e., polystyrene microspheres) in vivo proposed two mechanisms of phagocytosis. Inflammatory microglial phagocytosis is accompanied by the release of TNFα, IL-1β, ROS, and NO. In contrast, homeostatic phagocytosis, which is important to clear apoptotic cells or myelin debris, correlates with the induction of anti-inflammatory factors such as IL-10, TGF-β2, prostaglandin E2, and platelet-activating factor (PAF) (Ryu et al. 2012). Moreover, the uptake of apoptotic cells by microglia can downregulate pro-inflammatory cytokines, including TNFα, IL-12, and IL-1β (Ryu et al. 2012). Of relevance are the differences in lysosomal attributes between various cell types, including microglia, macrophages, monocyte-derived dendritic cells (DCs), and neutrophils. For example, a comparison of primary mouse microglia and J774 macrophages revealed that microglia contain higher levels of many lysosomal proteases than macrophages (Majumdar et al. 2007). However, the microglial lysosomes were less acidic with a pH ~ 6, compared to pH ~ 5 in macrophages, suggesting a decreased lysosomal enzymatic activity. Interestingly, treatment with macrophage colony-stimulating factor (M-CSF) and IL-6 correlated with lysosomal acidification, enabling microglial cells to degrade Aβ fibrils effectively (Majumdar et al. 2007). Immunohistochemical analyses of primary microglia also showed that the CIC-7 chloride transporter, which acidifies lysosomes via cotransporting protons and chloride ions, is not delivered efficiently to microglial lysosomes, appearing mistargeted and mobilized to the ER for degradation. In this study, M-CSF was also shown to induce CIC-7 trafficking to the lysosomes, revealing an important pathway for the phagocytosis of different types of cargo upon microglial activation through M-CSF (Majumdar et al. 2011). As mentioned above, the relationship between M-CSF microglial activation and lysosome acidification suggests a protective role in AD, by promoting degradation of Aβ fibrils and hence reducing plaques deposition. However, these in vitro findings may need to be interpreted with caution, as conflicting results that microglial phagocytosis is the primary cause of neuronal cell death in the presence of low levels of extracellular Aβ were also obtained in vitro (Neniskyte et al. 2011). In mouse models of AD it was additionally shown that CX3CR1 deletion is associated with an increased phagocytic ability, leading to greater Aβ contents inside of microglial phagolysosomes (Liu et al. 2010b). Therefore, cell–cell interactions between microglia and neurons, notably through CX3CR1-CX3CL1 communication, and the cytokine environment may collectively determine whether microglia are equipped to initiate phagocytosis.

Microglia express a wide range of TLRs, which makes them effective defenders against invading pathogens. For example, peptidoglycan (PGN) recognition is mediated by TLR2 in microglia (Kielian et al. 2005), TLR4 is necessary for Citrobacter koseri-mediated microglial activation and response to LPS (Lehnardt et al. 2002; Liu and Kielian 2009; Olson and Miller 2004), while TLR stimulation is important for eliciting the release of inflammatory mediators following Streptococcus pneumonia (Ribes et al. 2010) and Staphylococcus aureus challenge (Kochan et al. 2012). Studies with cultured macrophages have demonstrated that several TLRs localize to phagosomes (i.e., the vesicles formed around bacteria once engulfed by phagocytosis) following bacterial uptake (Husebye et al. 2010; Underhill et al. 1999), although similar findings have not yet been reported in microglia.

Recognition of dying cells is a key mechanism in the process of cellular digestion. It is clear that the classical “eat me” signal PS, apparently specific to apoptotic cells, is recognized by the phagocytic receptor TIM-4 (T cell immunoglobulin mucin-4) (Kobayashi et al. 2007; Santiago et al. 2007; Savill and Gregory 2007). The phagocytic “synapse” is now considered as a complex structure, which involves PS and the receptors TIM-1 and TIM-4 on phagocytic cells. The seven transmembrane G-protein-coupled receptor BAL1 (B aggressive lymphoma 1) mediates PS binding on the extracellular surface, and their intracytoplasmic portions form complexes with modular proteins engulfment and motility (ELMO)-Dock180-Rac involved in phagocytic signaling (Kobayashi et al. 2007; Santiago et al. 2007). This information brings us closer to understanding the mechanisms by which cellular debris in the form of exosomes regulate immune responses (Sokolowski and Mandell 2011).

1.4 Microglia and Inflammasome Activation

Microglia are capable of recognizing a diverse array of infectious agents, including bacteria, fungi, and viruses (Esen and Kielian 2006; Kaushik et al. 2011; Liu and Kielian 2009; Rambach et al. 2010), by expressing various PRRs, namely TLRs and NLRs (Hanamsagar et al. 2011; Jack et al. 2005b; Lee et al. 2013; Regnier-Vigouroux 2003). The recognition of PAMPs by TLRs makes it nearly impossible for a pathogen to go undetected, as it is difficult to evade the immune system by mutating an essential pathogenic motif. Microglia respond to pathogens by producing a wide range of pro-inflammatory mediators, including ROS and nitrogen species, cytokines, and chemokines (Aloisi 2001; Tambuyzer et al. 2009). In a similar manner, microglia are activated in response to endogenous danger signals such as the nucleotide ATP or misfolded proteins, which also culminate into pro-inflammatory cytokine secretion. It is thought that recognition of these DAMPs occurs via the activation of NLRs within the cytoplasm of host cells. NLR inflammasome activation leads to the processing and release of pro-inflammatory cytokines, including IL-1β and IL-18 which have been implicated in the pathophysiology of numerous CNS neurodegenerative diseases, including AD and PD (Benzing et al. 1999; Koprich et al. 2008). Acute brain injuries such as stroke, trauma, and hemorrhage are also characterized by neuroinflammation and IL-1 release, a process which is particularly linked to disease exacerbation (Denes et al. 2011; Lu et al. 2005; Masada et al. 2003). A role for IL-1β and IL-18 has also been described in several CNS infection models, including bacterial meningitis, brain abscess, and human immunodeficiency virus (HIV)-associated dementia (Ghorpade et al. 2003; Iannello et al. 2010; Kielian et al. 2004; Saukkonen et al. 1990; Xiong et al. 2012) (see Chap. 15 for further reading on HIV). IL-1β and its receptor IL-1R play a crucial role in the amplification of cytokine/chemokine networks, and bacterial clearance during acute S. aureus infection (Kielian et al. 2004; Xiong et al. 2012). IL-1β and IL-18 also regulate the induction of adaptive immunity, and both cytokines have been shown to influence disease development in experimental autoimmune encephalomyelitis (EAE), an animal model for multiple sclerosis (MS), an inflammatory-mediated demyelinating disease of the human CNS (Gris et al. 2010; Sutton et al. 2006) (Fig. 5.1) (see Chap. 16 for further reading on MS). The clinical course of MS is variable but usually initiates with reversible episodes of neurological disability that transform into a disease of continuous and irreversible neurological decline. Axonal damage and neurodegeneration are a major cause of irreversible neurological disability in MS. Inflammation and demyelination are additional hallmarks of the disease that manifest in clinical symptoms such as numbness, muscle spasms, optic neuritis, neuropathic pain, and paralysis (Dutta and Trapp 2011; Trapp et al. 1999; Trapp and Nave 2008).

In AD, extracellular Aβ triggers activation of the NLRP3 inflammasome in microglia, due to lysosomal damage and cathepsin B release following fibril uptake (Halle et al. 2008). Enhanced caspase-1 expression has been reported in the brains of AD patients, while double transgenic APP/PS1 mice harboring the human amyloid precursor protein (APP) and the presenilin 1 (PS1) mutation for familial AD demonstrated reduced caspase-1, IL-1β, and Aβ burdens when lacking NLRP3 or caspase-1 (Heneka et al. 2013). NLRP3 deficiency has been associated with a skewing of microglial activation towards an M2-like anti-inflammatory state, suggesting an important regulatory role for the inflammasome in microglial-induced inflammation. It is not clear whether other types of NLR inflammasomes could serve as platforms for IL-1β processing in response to Aβ, as their contribution has not yet been tested in AD animal models. However, a recent study that examined NLRP1 gene polymorphisms in a cohort of AD patients revealed an association of four non-synonymous polymorphisms in NLRP1 with AD, suggesting that these mutations may be involved in the predisposition to AD (Pontillo et al. 2012). Nevertheless, a direct cause-and-effect relationship remains to be defined.

The EAE animal model commonly requires injection of Mycobacterium along with the myelin peptides, to render the endogenous protein immunogenic. It is thought that the initial detection of danger signals (which remain ill-defined) within the CNS is mediated by microglia through their ability to present antigen and secrete immune molecules that can recruit peripheral immune cells into the CNS. Infiltrating macrophages and T cells can damage neurons, which provides signals to perpetuate microglial activation. If not tightly regulated, this can lead to progressive and irreversible tissue damage (Jack et al. 2005a). Microglia can phagocytose myelin debris, which triggers IL-1β release (Williams et al. 1994); however, additional studies are needed to better define the role of microglial inflammasomes during MS development and/or progression. In general, NLRP3 inflammasome activation has been reported as detrimental to the host in the context of EAE. For example, NLRP3-deficient mice displayed reduced neuroinflammation and delayed myelin loss in a model of demyelination induced by cuprizone, i.e., a dietary chelator commonly used to induce demyelinating lesion in the corpus callosum of rodents as a model of MS (Jha et al. 2010). Similarly, in myelin-oligodendrocyte glycoprotein (MOG)-induced EAE, NLRP3 deficiency reduced disease severity by inhibiting inflammatory cell infiltrates into the spinal cord (Gris et al. 2010). There is also evidence that the inflammasome adaptor protein Apoptosis-associated Speck-like protein containing CARD (ASC) contributes to the progression of EAE, independently from its inflammasome activity; although no association with NLRP3 expression was found in that study (Shaw et al. 2010). Nonetheless, in all cases, the cellular basis of inflammasome activation remains to be determined.

Recently, it has been demonstrated that activated microglia are critical for the maintenance of chronic pain following traumatic spinal cord injury (SCI) (Hains and Waxman 2006). Additionally, microglia are among the first responders to neuronal damage and injury and produce a vast range of pro-inflammatory mediators, including IL-1β, within hours following SCI (Pineau and Lacroix 2007). The NLRP1 inflammasome may be involved in IL-1β and IL-18 processing following SCI through the activation of P2X4 receptor in neurons (de Rivero Vaccari et al. 2008). Currently, it is thought that the majority of IL-1β release that follows SCI originates from neurons (Bernier 2012), and additional studies are needed to determine the role of microglia in inflammasome activation during SCI and chronic pain (see Chaps. 11 and 19 for further reading on these topics).

Although the CNS is usually well-protected from microbial invasion, in many instances are bacterial, fungal, viral, parasitic, and prion infections known to be life-threatening. The role of microglia in the rapid recognition and responses to pathogens has been extensively studied (Cosenza et al. 2002; Hanisch et al. 2001; Kielian et al. 2002; Lee et al. 1995). However, the physiological importance of inflammasome activation in the context of infectious diseases has only recently begun to be investigated. In particular, it was shown that cultured microglia respond to live S. aureus by producing IL-1β in a NLRP3- and ASC-dependent manner (Hanamsagar et al. 2011). The same study demonstrated the existence of redundant mechanisms for microglial IL-1β release, including bacterial pore-forming toxins and extracellular ATP, as well as caspase-1 and cathepsin B activity. A recent study with Legionella pneumophila demonstrated that flagellin recognition by the NLRC4 inflammasome may be important in restricting bacterial replication within microglia (Jamilloux et al. 2013). Neurotoxin prion infection can also lead to NLRP3 inflammasome activation in microglia (Shi et al. 2012), defining a novel trigger for the NLRP3 inflammasome apart from the traditional stimuli identified to date. Viral infections of the CNS were also shown to activate microglia, and a recent study further suggested that ROS production and subsequent NLRP3 inflammasome activation are critical during West Nile virus infection (Kaushik et al. 2012). Other inflammasomes such as HIN200 or AIM2, which are known to respond directly to viral nucleic acids, may also be involved; however, their role in microglia remains to be investigated.

2 Microglia in the Regulation of Adaptive Immunity

As demonstrated in the previous section, microglia, the main immune responsive glial cell population of the CNS, do not only produce a myriad of inflammatory cytokines and chemokines, but also receive inflammatory signals from nearby astrocytes, neurons, endothelial cells, and infiltrating leukocytes. Although chemokines and chemokine receptors are key players in the innate immune response, there are also pathways that directly link chemokine receptor signaling on microglia to the generation of effector T cells. In addition, alternative molecular mechanisms further highlight the role of microglial phagocytosis and antigen processing in effector T-cell responses in the CNS. Therefore, the type of CNS insult, the local inflammatory response, and the molecular program of microglial cells altogether contribute to the contextual regulation of neuroprotective versus neurotoxic pathways.

2.1 Antigen Presentation

T cells can recognize antigens only in the form of a peptide bound to MHC on the surface of antigen-presenting cells (APCs), in a process called antigen presentation. Professional APCs (e.g., macrophages, DCs identified by the expression of CD11c, and B lymphocytes) express MHC class II, while non-professional APCs (e.g., fibroblast, thymic epithelial cells, glial cells, pancreatic beta cells, and vascular endothelial cells) express MHC class I. Only professional APCs can activate a helper T cell that has never encountered the antigen before. Therefore, it is well established that adaptive immune responses depend on the function of professional APCs in the CNS. In addition to the peptide-MHC molecule interactions, contacts between the T cell and APCs via costimulatory molecules (such as CD86, intercellular adhesion molecule 1 (ICAM-1), and CD40) are required for full T-cell activation. Without this costimulatory signal, T cells will arrest in a nonactivated state, referred to as anergy. Considering that the immune system can be detrimental or beneficial to brain function, either by causing tissue damage or contributing to repair (Shaked et al. 2004; Wraith and Nicholson 2012), microglial ability to present antigens to T cells through MHC class II could allow these normally quiescent cells to play a yet undescribed role in shaping the outcome of certain neurological diseases where antigen presentation is critically involved such as MS and EAE (O’Keefe et al. 2002). Some studies have demonstrated that treatment of microglia with granulocyte macrophage colony-stimulating factor (GM-CSF) can transform these cells into a DC-like phenotype in vitro (Ponomarev et al. 2005). Usually identified as CD11c+ cells, brain DCs are found in perivascular regions of the CNS and in peripheral blood but might be derived from a microglial precursor as well (Prodinger et al. 2011). However, it remains unresolved whether APCs have the capacity to leave the CNS and transport CNS-derived antigens to the draining lymph nodes. Given the observations that the cerebrospinal fluid (CSF) and CNS parenchyma lack naïve T cells under physiological conditions, it is commonly accepted that adaptive immune responses against CNS antigens are initiated in the periphery, and subsequently propagated to the CNS by circulating memory T cells, which are restimulated by antigens within the CNS (Ransohoff and Engelhardt 2012). The involvement of T cell–APC interactions in EAE has been evaluated using adoptive transfer experiments with activated T cells, either myelin-specific or ovalbumin-specific. In these experiments, the ovalbumin-specific T cells remained in the subarachnoid space (SAS), whereas myelin-specific T cells readily invaded the parenchyma. Ovalbumin-specific T cells were only detected in the brain parenchyma when ovalbumin-pulsed APCs were placed in the SAS (Bartholomaus et al. 2009). Therefore, these data suggest that antigen-specific interactions between T cells and APCs occurring within the SAS are essential for the development of EAE (Ransohoff and Engelhardt 2012). The fact that most studies exploring microglial interactions with T cell activation were performed in vitro prompts careful interpretation, as it was shown that microglia that develop in mixed glial cultures display a more activated phenotype than microglia in pure cultures (Carson et al. 1998). In vivo microglia express low levels of accessory molecules required for efficient antigen presentation and have weak antigen-presenting activity (Ransohoff and Engelhardt 2012). Although adult microglia can express costimulatory molecules under certain conditions (Carson et al. 1998; Garcia et al. 2013; Zhang et al. 2002), they failed to present a peptide antigen to naïve T cells in vitro (Carson et al. 1998). Furthermore, adult microglia were less efficient, when compared to CD11c+ DCs, to restimulate primed T cells and to induce T cell proliferation in vitro (Ford et al. 1995; Garcia et al. 2013). Similarly, during Theiler’s murine encephalomyelitis virus-induced demyelinating disease, CNS-infiltrating macrophages were more highly activated based on their MHC class II expression than resident microglia. However, microglia isolated at the onset of disease were as efficient as infiltrating macrophages at inducing Th1 proliferation and IFNγ production (Mack et al. 2003). Although it was reported that EAE disease severity is ameliorated when microglia are selectively depleted before disease induction via systemic administration of ganciclovir (an antiviral nucleotide analogue) to mice engineered to express the thymidine kinase of herpes simplex virus (HSVTK) under the CD11b promoter (Heppner et al. 2005) (see Chap. 16 for further reading about this model), the exact mechanism linking microglia to APC functions in vivo remains unsolved.

Although microglia display enhanced clearance of myelin debris in CNS autoimmunity models (Nielsen et al. 2009), compared with naïve/physiological conditions, it is still unclear how the scavenging mechanism relating to antigen processing and presentation differs between microglia and professional CD11c-positive DCs. Recently it was shown that oligodendrocytes secrete small membrane vesicles containing myelin antigens, called exosomes, which are efficiently taken up by microglia both in vitro and in vivo. However, these exosomes are preferentially internalized by microglia devoid of antigen presenting capacity (Fitzner et al. 2011). Therefore, it was proposed that a decreased exosome internalization by MHC class II-positive microglia might allow myelin antigens to be transported with the interstitial fluid to the meningeal macrophages, which could then present myelin-derived antigen to myelin-specific T cells present in the cerebrospinal fluid (Fitzner et al. 2011; Ransohoff and Engelhardt 2012). It has been demonstrated, however, that IFNγ enables brain CD11c-expressing cells to become effective APCs and to induce antigen-specific T cells expressing the CD4 glycoprotein at their surface (CD4+) to proliferate and secrete cytokines specific to the differentiated Th1/Th17 effector T cell subtypes, which are considered as encephalitogenic (Gottfried-Blackmore et al. 2009). In an entorhinal cortex lesion model, the fact that microglia do not upregulate MHC class II also indicates that they could represent a DC population capable of producing immunological tolerance, i.e., a condition where the immune system does not attack an antigen, under particular circumstances. However, evidence is still lacking to validate a role for microglia expressing the CD11c marker in transporting antigens from the brain to lymphoid organs (Prodinger et al. 2011). Not all microglia express CD11c, only a subpopulation of about 1.1 ± 0.1 % of microglia in naïve controls, and 4 ± 0.8 % at 3 weeks after cuprizone treatment expressed CD11c (Remington et al. 2007). It was shown that CD11c-positive microglial cells in culture are approximately three times better at stimulating T cell activation than CD11c-negative cells, indicative of enhanced antigen presentation capacity associated with this phenotype. Also, resident brain microglia appeared to be key functional players during adaptive responses in the sub-chronic 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP, dopaminergic neurotoxin) mouse model of PD that is associated with T-cell recruitment into the CNS (Depboylu et al. 2012). Using bone marrow chimeric mice to distinguish the resident microglial population from infiltrating leukocytes, in combination with retrograde neuronal tracing, it was shown that resident microglia upregulate MHC class II after MPTP administration. Importantly, microglia also contained neuronal tracer and appeared in close contact with CD4+ T cells in the lesion area. Therefore, resident microglia represent likely candidates for the presentation of antigens to infiltrating activated T lymphocytes during dopaminergic cell death.

More than a decade ago, when the concept of attenuating Aβ accumulation in the brain by eliciting adaptive immune responses emerged, it was the first time that a self-peptide was introduced as a vaccine. Preclinical studies were successful, but the initial clinical trials were halted because of the development of severe inflammatory reactions in the vaccinated AD patients (Nicoll et al. 2003; Schenk 2008). While a number of reports suggested that T cells were activated in the patients and could be found both in the periphery and brain tissues, the question on how those T cells gained access to the brain has been investigated by some groups. In this regard, microglia were proposed to be indirectly involved in the transendothelial migration and activation of T cells. Their involvement with the regulation of adaptive T cells effector responses was evidenced by their high levels of TNFα release in response to Aβ application in vitro, thus resulting in over-expression of the chemokine receptor CXCR2 (mouse equivalent of the human IL-8 receptor) on peripheral T cells derived from AD patients (Liu et al. 2010a). During EAE, disease severity was significantly exacerbated in hypermyelinated PLP-Akt-DD transgenic mice compared to wild-type animals with “normal” myelin content (Jaini et al. 2013). PLP-Akt-DD transgenic animals have been engineered to drive, under the proteolipid PLP promoter, expression of the constitutively active Akt-1 which harbors aspartic acid (D) at positions Thr308 and Ser473, two sites where phosphorylation leads to full activation of Akt. In these mice, expression of Akt-DD dramatically increased myelin production, correlating with increased neurological score and exacerbated neuronal damage. There was also an increased number of resident microglia in the hypermyelinated mice, but their individual capacity to present antigens was not increased, as revealed by the levels of IL-17 (a pro-inflammatory cytokine inducing and mediating pro-inflammatory responses) secreted by T cells isolated from these mice in the presence of microglia (Jaini et al. 2013). This work suggests that the load of self-antigens imposed on the APCs could have an important influence on the disease onset and progression (Fig. 5.1). Therefore, the drastic differences between microglia versus professional DCs, as APCs, support the notion that microglial activity, if well controlled, is a crucial step in determining neuronal survival.

2.2 Chemokine Receptors

Microglia express most chemokine receptors in vivo, including CCR3, CCR5, CCR8, CXCR3, CXCR4, and CX3CR1, depending on the CNS microenvironment that shapes their effector phenotype (Biber et al. 2002; Cardona and Ransohoff 2007; Glabinsk and Ransohoff 2001; Prinz and Priller 2010; Ransohoff 2002; Trebst et al. 2001; Trebst et al. 2003), with actions that promote either tissue damage and neurotoxicity or neuroprotection and tissue repair (Cardona et al. 2006; Garcia et al. 2013; Prinz et al. 2011) (also see Chap. 6). Several studies support the concept that microglia are a heterogeneous population based on the selective expression of chemokine receptors. For example, transcriptome analyses in combination with protein validation revealed that nearly all microglial cells express high levels of complement C1q, while only subsets of microglia express the CCR1 ligand, CXCL14, and TREM2 (triggering receptor expressed on myeloid cells 2; which regulates microglial activation and phagocytosis) (Schmid et al. 2009). Adding another layer of complexity, the regulation of microglial function is also influenced by the other CNS resident cells. For example, several studies in PD models suggest a scenario in which astrocyte-microglial communication impacts T cell responses. Astrocytes exposed in vitro to neuron-derived α-synuclein induce expression of transcripts for chemokines such as CCL3-7, CCL12, 19 and 20, and CXCL1,2,4,5,9,10, 11, 12, and 16, and CX3CL1 (Harms et al. 2013). Microglial expression of the chemokine receptors CCR3, CCR5, CXCR3 and CX3CR1 makes them highly responsive to their respective ligands produced by astrocytes (Cowell et al. 2006; Flynn et al. 2003; Gautier et al. 2012). Chemokine receptor-ligand interactions also correlated with morphological changes in microglia and increased expression of MHC class II in vivo (Garcia et al. 2013). Most importantly, microglial responses to aggregated α-synuclein induced antigen processing and presentation, driving T cell proliferation and cytokine release, as well as neurodegeneration in a mouse model of PD in vivo (Harms et al. 2013), suggesting that antigen presentation could promote an inflammatory environment detrimental to neuronal survival. A similar role for microglia/MHC class II was also demonstrated in a facial nerve axotomy model, in which the physical distance between the injury to the facial nerve and the neuronal localization in the brainstem provides a good model to avoid direct CNS trauma and/or disruption of the BBB (Byram et al. 2004). Therefore, these results implicate a central role for microglial MHC class II in the activation of adaptive immune responses, indicating that this process could be viewed as a neuroprotective target.

Microglia represent important components of the CNS immune response in MS and EAE, alongside peripherally derived macrophages and DCs. The mechanism by which these cell types repopulate during acute and chronic brain inflammation have been explored. The chemokine receptor CCR2 that signals in response to several monocyte chemoattractant molecules (known as CCL2, CCL8, and CCL16) is crucial for the accumulation of macrophages and DCs to the sites of inflammation, although its role in mobilizing microglia is still controversial. There is evidence suggesting that CD11b+ Ly6ChiCCR2+CD62L+ (or adhesion molecule L-selectin+) monocytes are mobilized from the bone marrow into the bloodstream, via a GM-CSF-dependent pathway before EAE relapses (King et al. 2009). However, microglia are recognized by a CD11b+ Ly6C−CCR2−CD62L+ phenotype (Carson et al. 1998), and the lack of microglial CCR2 upregulation in CCR2-RFP reporter mice supports a selective role for CCR2 in the trafficking of macrophages (King et al. 2009; Saederup et al. 2010). However, it was reported that over-expression of CCL2 also leads to morphological activation of microglia and pro-inflammatory cytokine release in the CNS (Selenica et al. 2013), although a direct interaction between exogenous CCL2 and CCR2 on microglia is uncertain. CCR2 also targets the recruitment of CCR2+CD11c+ DCs. As expected, the expression of multiple chemokine receptors by myeloid cells may account for the synergistic effects observed when analyzing DC trafficking or microglial mobilization within the CNS.

The pathways that influence chemokine receptor expression are complex and cytokines are the most studied candidates in this process. However, the highly conserved cell–cell communication pathway Notch, initially identified as a pleiotropic mediator of cell fate in invertebrates, has emerged as an important regulator of immune cell development and function. It was recently shown that microglia express the Notch ligand DLL4 and that DLL4 blockade reduces the neurological symptoms of EAE, correlating with a downregulation of CCR2 and CCR6 expression on T cells. Although microglia are not the only cell type expressing DLL4 in the brain (Benedito and Duarte 2005; Shutter et al. 2000), microglia–T cells interactions mediated by DLL4 and Notch receptors could contribute to the regulation of T cells chemokine receptors expression, and their responses to organ-specific chemokine production (Reynolds et al. 2011).

In addition to microglia–T cell interactions, microglia-neuron communication is now recognized as a key mechanism in the regulation of microglial function. Several studies have confirmed the predominant microglial expression of CX3CR1 (Cardona et al. 2006; Gautier et al. 2012; Jung et al. 2000), whose ligand CX3CL1 is expressed in neuronal membranes (Bazan et al. 1997; Mizoue et al. 1999; Rossi et al. 1998). The chemokine module is released by action of proteases including A Disintegrin and metalloproteinase domain-containing protein 10 and 17 (ADAM10 and ADAM10) and cathepsin S (Clark et al. 2009; Garton et al. 2001; Hundhausen et al. 2003). Binding of CX3CL1 to CX3CR1 on microglia has been viewed as neuroprotective because this interaction inhibits microglial activation in selected models of neurodegeneration (Bachstetter et al. 2011; Cardona et al. 2006; Rogers et al. 2011). Although some studies suggest that CX3CR1 deficiency could be neuroprotective and/or anti-inflammatory in various contexts of disease such as AD (Fuhrmann et al. 2010; Lee et al. 2010; Mattison et al. 2013), in EAE, the absence of CX3CR1 correlated with an enhanced neurological disease and increased microglial activation, reflected by their increased MCH class II expression and proliferation. These features of microglia also correlated with an increased T cell proliferation. Furthermore, higher IFNγ and IL-17 levels were detected in cerebellar and spinal cord tissues of CX3CR1-deficient mice (Garcia et al. 2013).

Overall chemokine receptors regulate microglial mobilization within the CNS, but the observation of higher frequencies of IFNγ- and IL-17-producing T cells in the lymphoid tissues of CX3CR1-deficient mice, and the enhanced T cell proliferation induced by CX3CR1-deficient DCs, demonstrate that besides their role in chemoattraction, some chemokine receptors such as CX3CR1 may contribute to the establishment of adaptive immune responses via regulation of antigen presentation function.

3 Microglia and the BBB

The BBB is the interface between the CNS and the periphery, fulfilling two main functions, acting as a physical barrier and a selective exchange barrier aimed to maintain the proper CNS microenvironment for optimal neuronal function. This mechanical separation of the CNS is accomplished by the presence of specialized endothelial cells tightly attached one to another via tight junctions and adherent junctions (Hawkins and Davis 2005; Hermann and ElAli 2012). These junctions are formed by the tight junction proteins occludins and claudins, as well as the junctional adhesion molecules E-cadherin, P-cadherin, and N-cadherin, whose ultimate role is to restrict blood-borne molecules and peripheral cells from entering the CNS (Pardridge 2003; Wilson et al. 2010). Extensive work in the last decade has unraveled the presence of a specialized intrinsic innate immune system in the CNS, which also involves the BBB as an active contributor, as opposed to being an immunological neutral and passive barrier (Muldoon et al. 2013). It has been observed not only that peripheral immune cells can cross an intact BBB (Carson et al. 2006), but that the BBB modulates the differentiation of infiltrating CD14+ monocytes into DCs through the influence of BBB-secreted TGFβ and GM-CSF. These data confer a more active role of the BBB in the intrinsic innate immunity of the CNS. Despite the limited infiltration of peripheral immune cells into the CNS under physiological conditions, it is well known that neutrophils, eosinophils, T lymphocytes, monocytes, and other immune cells can infiltrate the CNS parenchyma after injuries, infections, and chronic diseases such as MS (Wilson et al. 2010).

Although it is unclear whether BBB dysfunction during pathological conditions is due to a loss of signals provided by the CNS, or to breakdown signals produced during pathology, recent work has shown that communication between the BBB and microglia is one of the early events involved in EAE. In particular, disruption of the barrier and subsequent leakage of the plasma protein fibrinogen has been identified as a signal triggering microglial motility towards perivascular regions, microglial activation and clustering, as well as axonal damage, using two-photon in vivo imaging. Importantly, these changes in microglia were shown to occur prior to the onset of neurological changes, suggesting their relevance to the disease pathogenesis (Davalos et al. 2012) (also see Chap. 4 for more information on these findings).

Activated microglia and infiltrating macrophages secrete a wide range of pro-inflammatory signals, including NO and ROS, which have been categorized as a major mechanism leading to axonal damage in MS (Lassmann 2010), through the initiation of mitochondrial dysfunction which, in turn, leads to the production of mitochondrial-derived ROS (Haider et al. 2011). Indeed, increased ROS generation has been observed in close proximity to fibrin-mediated microglial clusters at the peak of EAE, compared to distal areas with normal microglial density, or healthy controls (Davalos et al. 2012). Since fibrin is formed from fibrinogen through the action of the protease thrombin, it represents a molecular marker for BBB disruption and fibrinogen leakage into the CNS parenchyma. Interestingly, in vitro treatment of microglia with fibrinogen induced a 5.1-fold increase of H2O2 and a 7.4-fold increase in inducible nitric oxide synthase (iNOS) gene expression (Davalos et al. 2012), strongly suggesting that fibrinogen might contribute to ROS generation during inflammatory demyelinating lesions. These data highlight the importance of fibrin deposition for the development of axonal degeneration within microglial clusters. Postmortem analysis of MS brains has reinforced this finding of microglial clustering within normal-appearing white matter (van Horssen et al. 2012), although microglial association with the disrupted BBB remains a challenge to assess, due to a limited effectiveness of molecular probes to identify microglia at their different states of “activation” in heavily fixed human specimens, which also complicates investigation of their association with BBB leakiness.

This aggregation of microglial cells at perivascular areas before disease onset may constitute an early indicator of new lesion formation, as demyelination and axonal damage are exacerbated at the peak of EAE, coincident with rapid microglial process extension and retraction at the sites of BBB leakage showing fibrin deposition (Fig. 5.2). Despite the strong correlation between perivascular fibrinogen deposition and microglial clusters formation at the early stages of neuroinflammation in EAE, the role of fibrinogen or other blood products within the CNS and their interaction with microglial cells in the context of other CNS pathologies remains to be investigated.

Fig. 5.2
figure 2

Microglia and the interface to peripheral communication. Fibrinogen appears to be a key modulator of microglial activity as a result of early blood–brain barrier (BBB) damage. Once extravasated, it is lysed into fibrin by thrombin and interacts with parenchymal microglia, leading to an increased expression of inducible nitric oxide synthase (iNOS) and reactive oxygen species (ROS) that accounts for neuronal damage associated with certain neuroinflammatory conditions such as EAE

Full size image

Both in vivo and in vitro studies support the idea that microglia release inflammatory mediators that regulate BBB permeability. Coculture models of microvascular endothelial cells showed that LPS-activated microglia induce BBB leakiness by producing ROS through nicotinamide adenine dinucleotide phosphate (NADPH) oxidase (Matsumoto et al. 2012; Sumi et al. 2010) and TNFα release (Nishioku et al. 2010). Microglial activation is evident in the aged CNS and breakdown of the BBB and blood–retinal barrier in the aging brain suggests that microglial actions at the vasculature interface may contribute not only to neuroinflammatory conditions, but also to the pathogenesis of age-related neurological disorders (see Chap. 13 for additional information on aging). Further understanding of the mechanisms involved in the coupling between the BBB and microglia in innate immunity will require to investigate the role of microglia in the regulation of BBB dynamics during normal physiological conditions, considering that primary microglial cells in culture express a recently discovered tight junction protein, tricellulin (Mariano et al. 2011), whose role in the CNS and immune function remains to be elucidated.

4 Summary

It is now well accepted that microglia do not only participate in the genesis of immune responses during CNS infection/injury, but also play key roles in maintaining homeostasis. During CNS development and inflammatory conditions, microglia display the ability to phagocytose debris and apoptotic cells, release cytokines, and upregulate MHC class II to modulate effector T cell responses. These functions of microglia do not occur in isolation but are influenced by neighboring cells, including astrocytes, neurons, and the BBB. Microglia are capable of producing a vast array of cytokines and chemokines and are therefore able to influence the CNS environment in a very effective manner. Microglia can also respond to soluble mediators produced by neurons and astrocytes, and a current challenge is to dissect the signals that enable such an orchestrated network to sustain optimal CNS function. The most recognized function of microglia as phagocytes and scavengers is temporally regulated to allow for appropriate antigen recognition and presentation during adaptive immune responses. As our knowledge of microglial functions during homeostasis and pathology continues to expand, it may be possible to consider harnessing microglial activity to benefit CNS integrity and improve outcomes following neuroinflammatory injury. For many years researchers have been studying the mechanisms that lead to microglial activation and enhance their phagocytic properties. Another interesting venue involves the use of microglia as shuttles of anti-inflammatory molecules. Although the issue of identifying microglial origins is still a moving target, progress has been made to show that microglia are a distinct myeloid subset with unique properties, making them distinct from DCs and macrophages. Balancing neuroprotective versus neurotoxic properties of microglia has been and will continue to be a topic of arduous investigation, with the goal of modulating CNS inflammation due to autoimmunity, infection, or neurodegeneration.