Microglia respond to CNS injuries and diseases with complex reactions, often called “activation.” A pro-inflammatory phenotype (also called classical or M1 activation) lies at one extreme of the reactivity spectrum. There were several motivations for this study. First, bacterial endotoxin (lipopolysaccharide, LPS) is the most commonly used pro-inflammatory stimulus for microglia, both in vitro and in vivo; however, pro-inflammatory cytokines (e.g., IFNγ, TNFα) rather than LPS will be encountered with sterile CNS damage and disease. We lack direct comparisons of responses between LPS and such cytokines. Second, while transcriptional profiling is providing substantial data on microglial responses to LPS, these studies mainly use mouse cells and models, and there is increasing evidence that responses of rat microglia can differ. Third, the cytokine milieu is dynamic after acute CNS damage, and an important question in microglial biology is: How malleable are their responses? There are very few studies of effects of resolving cytokines, particularly for rat microglia, and much of the work has focused on pro-inflammatory outcomes. Here, we first exposed primary rat microglia to LPS or to IFNγ+TNFα (I+T) and compared hallmark functional (nitric oxide production, migration) and molecular responses (almost 100 genes), including surface receptors that can be considered part of the sensome. Protein changes for exemplary molecules were also quantified: ARG1, CD206/MRC1, COX-2, iNOS, and PYK2. Despite some similarities, there were notable differences in responses to LPS and I+T. For instance, LPS often evoked higher pro-inflammatory gene expression and also increased several anti-inflammatory genes. Second, we compared the ability of two anti-inflammatory, resolving cytokines (IL-4, IL-10), to counteract responses to LPS and I+T. IL-4 was more effective after I+T than after LPS, and IL-10 was surprisingly ineffective after either stimulus. These results should prove useful in modeling microglial reactivity in vitro; and comparing transcriptional responses to sterile CNS inflammation in vivo.

BackgroundMicroglial cells are highly mobile under many circumstances and, after central nervous system (CNS) damage, they must contend with the dense extracellular matrix (ECM) in order to reach their target sites. In response to damage or disease, microglia undergo complex activation processes that can be modulated by environmental cues and culminate in either detrimental or beneficial outcomes. Thus, there is considerable interest in comparing their pro-inflammatory (‘classical’ activation) and resolving ‘alternative’ activation states. Almost nothing is known about how these activation states affect the ability of microglia to migrate and degrade ECM, or the enzymes used for substrate degradation. This is the subject of the present study.MethodsPrimary cultured rat microglial cells were exposed to lipopolysaccharide (LPS) to evoke classical activation or IL4 to evoke alternative activation. High-resolution microscopy was used to monitor changes in cell morphology and aspects of the cytoskeleton. We quantified migration in a scratch-wound assay and through open filter holes, and invasion through Matrigel™. A panel of inhibitors was used to analyze contributions of different matrix-degrading enzymes to migration and invasion, and quantitative real-time reverse transcriptase PCR (qRT-PCR) was used to assess changes in their expression.ResultsVinculin- and F-actin-rich lamellae were prominent in untreated and IL4-treated microglia (but not after LPS). IL4 increased the migratory capacity of microglia but eliminated the preferential anterior nuclear-centrosomal axis polarity and location of the microtubule organizing center (MTOC). Microglia degraded fibronectin, regardless of treatment, but LPS-treated cells were relatively immobile and IL4-treated cells invaded much more effectively through Matrigel™. For invasion, untreated microglia primarily used cysteine proteases, but IL4-treated cells used a wider range of enzymes (cysteine proteases, cathepsin S and K, heparanase, and matrix metalloproteases). Untreated microglia expressed MMP2, MMP12, heparanase, and four cathepsins (B, K, L1, and S). Each activation stimulus upregulated a different subset of enzymes. IL4 increased MMP2 and cathepsins S and K; whereas LPS increased MMP9, MMP12, MMP14 (MT1-MMP), heparanase, and cathepsin L1.ConclusionsMicroglial cells migrate during CNS development and after CNS damage or disease. Thus, there are broad implications of the finding that classically and alternatively activated microglia differ in morphology, cytoskeleton, migratory and invasive capacity, and in the usage of ECM-degrading enzymes.

BackgroundAcute CNS damage is commonly studied using rat and mouse models, but increasingly, molecular analysis is finding species differences that might affect the ability to translate findings to humans. Microglia can undergo complex molecular and functional changes, often studied by in vitro responses to discrete activating stimuli. There is considerable evidence that pro-inflammatory (M1) activation can exacerbate tissue damage, while anti-inflammatory (M2) states help resolve inflammation and promote tissue repair. However, in assessing potential therapeutic targets for controlling inflammation, it is crucial to determine whether rat and mouse microglia respond the same.MethodsPrimary microglia from Sprague-Dawley rats and C57BL/6 mice were cultured, then stimulated with interferon-γ + tumor necrosis factor-α (I + T; M1 activation), interleukin (IL)-4 (M2a, alternative activation), or IL-10 (M2c, acquired deactivation). To profile their activation responses, NanoString was used to monitor messenger RNA (mRNA) expression of numerous pro- and anti-inflammatory mediators, microglial markers, immunomodulators, and other molecules. Western analysis was used to measure selected proteins. Two potential targets for controlling inflammation—inward- and outward-rectifier K+ channels (Kir2.1, Kv1.3)—were examined (mRNA, currents) and specific channel blockers were applied to determine their contributions to microglial migration in the different activation states.ResultsPro-inflammatory molecules increased after I + T treatment but there were several qualitative and quantitative differences between the species (e.g., iNOS and nitric oxide, COX-2). Several molecules commonly associated with an M2a state differed between species or they were induced in additional activation states (e.g., CD206, ARG1). Resting levels and/or responses of several microglial markers (Iba1, CD11b, CD68) differed with the activation state, species, or both. Transcripts for several Kir2 and Kv1 family members were detected in both species. However, the current amplitudes (mainly Kir2.1 and Kv1.3) depended on activation state and species. Treatment-induced changes in morphology and migratory capacity were similar between the species (migration reduced by I + T, increased by IL-4 or IL-10). In both species, Kir2.1 block reduced migration and Kv1.3 block increased it, regardless of activation state; thus, these channels might affect microglial migration to damage sites.ConclusionsCaution is recommended in generalizing molecular and functional responses of microglia to activating stimuli between species.Electronic supplementary materialThe online version of this article (doi:10.1186/s12974-017-0941-3) contains supplementary material, which is available to authorized users.

BackgroundMicroglia migrate during brain development and after CNS injury, but it is not known how they degrade the extracellular matrix (ECM) to accomplish this. Podosomes are tiny structures with the unique ability to adhere to and dissolve ECM. Podosomes have a two-part architecture: a core that is rich in F-actin and actin-regulatory molecules (for example, Arp2/3), surrounded by a ring with adhesion and structural proteins (for example, talin, vinculin). We recently discovered that the lamellum at the leading edge of migrating microglia contains a large F-actin-rich superstructure (‘podonut’) composed of many podosomes. Microglia that expressed podosomes could degrade ECM molecules. Finely tuned Ca2+ signaling is important for cell migration, cell-substrate adhesion and contraction of the actomyosin network. Here, we hypothesized that podosomes contain Ca2+-signaling machinery, and that podosome expression and function depend on Ca2+ influx and specific ion channels.MethodsHigh-resolution immunocytochemistry was used on rat microglia to identify podosomes and novel molecular components. A pharmacological toolbox was applied to functional assays. We analyzed roles of Ca2+-entry pathways and ion channels in podosome expression, microglial migration into a scratch-wound, transmigration through pores in a filter, and invasion through Matrigel™-coated filters.ResultsMicroglial podosomes were identified using well-known components of the core (F-actin, Arp2) and ring (talin, vinculin). We discovered four novel podosome components related to Ca2+ signaling. The core contained calcium release activated calcium (CRAC; Orai1) channels, calmodulin, small-conductance Ca2+-activated SK3 channels, and ionized Ca2+ binding adapter molecule 1 (Iba1), which is used to identify microglia in the CNS. The Orai1 accessory molecule, STIM1, was also present in and around podosomes. Podosome formation was inhibited by removing external Ca2+ or blocking CRAC channels. Blockers of CRAC channels inhibited migration and invasion, and SK3 inhibition reduced invasion.ConclusionsMicroglia podosome formation, migration and/or invasion require Ca2+ influx, CRAC, and SK3 channels. Both channels were present in microglial podosomes along with the Ca2+-regulated molecules, calmodulin, Iba1 and STIM1. These results suggest that the podosome is a hub for sub-cellular Ca2+-signaling to regulate ECM degradation and cell migration. The findings have broad implications for understanding migration mechanisms of cells that adhere to, and dissolve ECM.

The Ca2+-activated K+ channel, KCa3.1 (KCNN4/IK1/SK4), contributes to “classical,” pro-inflammatory activation of microglia, and KCa3.1 blockers have improved the outcome in several rodent models of CNS damage. For instance, blocking KCa3.1 with TRAM-34 rescued retinal ganglion neurons after optic nerve damage in vivo and, reduced p38 MAP kinase activation, production of reactive oxygen and nitrogen species, and neurotoxicity by microglia in vitro. In pursuing the therapeutic potential of KCa3.1 blockers, it is crucial to assess KCa3.1 contributions to other microglial functions and activation states, especially the IL-4-induced “alternative” activation state that can counteract pro-inflammatory states. We recently found that IL-4 increases microglia migration – a crucial function in the healthy and damaged CNS – and that KCa3.1 contributes to P2Y2 receptor-stimulated migration. Here, we discovered that KCa3.1 is greatly increased in alternative-activated rat microglia and then contributes to an enhanced migratory capacity. IL-4 up-regulated KCNN4 mRNA (by 6 h) and greatly increased the KCa3.1 current by 1 day, and this required de novo protein synthesis. The increase in current was sustained for at least 6 days. IL-4 increased microglial migration and this was reversed by blocking KCa3.1 with TRAM-34. A panel of inhibitors of signal-transduction mediators was used to analyze contributions of IL-4-related signaling pathways. Induction of KCNN4 mRNA and KCa3.1 current was mediated specifically through IL-4 binding to the type I receptor and, surprisingly, it required JAK3, Ras/MEK/ERK signaling and the transcription factor, activator protein-1, rather than JAK2, STAT6, or phosphatidylinositol 3-kinase.The same receptor subtype and pathway were required for the enhanced KCa3.1-dependent migration. In providing the first direct signaling link between an IL-4 receptor, expression and roles of an ion channel, this study also highlights the potential importance of KCa3.1 in alternative-activated microglia.