In both vertebrates and invertebrates, including leeches, microglia are rapidly activated by central nervous system (CNS) damage and migrate to the lesions. Adenosine triphosphate (ATP), nitric oxide (NO) and endocannabinoids have been implicated in controlling activation and migration, but details of the mechanisms are uncertain. This dissertation tests the hypothesis that endocannabinoids coordinate and influence the microglial response to nerve cord crushing. Chapter 1 reports that application of endocannabinoids to nerve cords at concentrations as low as 100 nM for arachidonylethanolamide (anandamide, or AEA) reduced the number of migrating microglia, but not when cords were pretreated with 10 µM of the CB2 cannabinoid receptor (CB2R) antagonist SR144528. In addition, immunoblots confirmed the expression of CB1-like and CB2-like receptors and immunohistochemistry showed that they were concentrated at lesions, where microglia accumulated. Benzoyl ATP (BzATP) also reduced microglia accumulation, an effect blocked by pretreatment of nerve cords with SR144528, whereas the G-protein coupled P2YR agonists uridine triphosphate (UTP) and methylthio-ATP (MeSATP) at 100 µM did not reduce accumulation. This result suggested that P2X7R activation elicited production and release of a CB2R agonist that influenced microglia movement. Chapter 2 reports that extracellular ATP levels were highest in the CNS within the first 30 min of injury and remained above unharmed controls for at least 2 hours. Application of 10 units (U) of the ATPase apyrase to nerve cords reduced accumulation of microglia at lesions, another indication that microglia require extracellular ATP to accumulate. Chapter 3 reports that AEA stopped ATP-induced movement of microglia, and that this effect was blocked by pretreatment of nerve cords with the CB1R antagonist SR141716A (10µM), the nitric oxide synthase inhibitor NG-nitro-L-arginine methyl ester (L NAME, 2 mM) or the NO scavenger 2-(4-carboxyphenyl)-4,4,5,5-teramethylimidazoline-oxyl-3-oxide (carboxy-PTIO or cPTIO, 1 mM). Thus the migration of microglia to lesions is regulated not only by ATP acting on P2YR and by NO acting on soluble guanylate cyclase (sGC), but also potentially by ATP binding to P2X7-like receptors to increase the production of cannabinoids. Cannabinoids, binding to the CB1R and CB2R cause production of NO, which suppresses microglia movement.

The p38 mitogen-activated protein kinase [p38(MAPK)] mediates responses to extracellular stressors. An increase in the phosphorylated form of p38(MAPK) [p-p38(MAPK)] has been associated with early events in Alzheimer disease (AD). Although most often associated with processes including apoptosis, inflammation and oxidative stress, p-p38(MAPK) also mediates beneficial physiological functions, such as cell growth, survival and phagocytosis of cellular pathogens. Amyloid plaques [β-amyloid aggregates] are a hallmark of AD-related pathology. As p38(MAPK) has been detected in the vicinity of senile plaques, we combined immunohistochemistry and stereological sampling to quantify the distribution of plaques and p-p38(MAPK)-immunoreactive (IR) cells in the sensorimotor cortex of 3-, 6- and 10-month-old TgCRND8 mice. This animal model expresses an aggressive nature of the AD-related human amyloid-β protein precursor (APP). It was confirmed by the appearance of both dense-core (thioflavin-S-positive) and diffuse plaques, even in the youngest mice. p-p38(MAPK)-IR cells were associated with both dense-core and diffuse plaques, but the expected age-dependent increase in the density of plaque-associated p-p38(MAPK)-IR cells was restricted to dense-core plaques. Furthermore, the density of dense-core plaque-associated p-p38(MAPK)-IR cells was inversely correlated with the size of the core within the given plaque, which supports a role for these microglia in restricting core growth. p-p38(MAPK)-IR cells were also observed throughout wildtype and TgCRND8 mouse cortical parenchyma, but the density of these non-plaque-associated cells remained constant, regardless of age or genotype. We conclude that the constitutive presence of p-p38(MAPK)-IR microglia in aging mouse brain is indicative of a longitudinal role for this kinase in normal brain physiology. Additionally, the majority of p-p38(MAPK)-IR cells were predominantly co-immunoreactive for the Macrophage-1 (CD11b/CD18) microglial marker, regardless of whether they were associated with plaques or localized to the parenchyma. We suggest that the facts that a pool of p-p38(MAPK)-IR microglia appears to restrict b-amyloid plaque core development and the non-pathological role of p-p38(MAPK) in parenchyma, needs to be considered when anticipating targeted p38(MAPK) therapeutics in the context of clinical AD.

The p38 mitogen-activated protein kinase [p38(MAPK)] mediates responses to extracellular stressors. An increase in the phosphorylated form of p38(MAPK) [p-p38(MAPK)] has been associated with early events in Alzheimer disease (AD). Although most often associated with processes including apoptosis, inflammation and oxidative stress, p-p38(MAPK) also mediates beneficial physiological functions, such as cell growth, survival and phagocytosis of cellular pathogens. Amyloid plaques [β-amyloid aggregates] are a hallmark of AD-related pathology. As p38(MAPK) has been detected in the vicinity of senile plaques, we combined immunohistochemistry and stereological sampling to quantify the distribution of plaques and p-p38(MAPK)-immunoreactive (IR) cells in the sensorimotor cortex of 3-, 6- and 10-month-old TgCRND8 mice. This animal model expresses an aggressive nature of the AD-related human amyloid-β protein precursor (APP). It was confirmed by the appearance of both dense-core (thioflavin-S-positive) and diffuse plaques, even in the youngest mice. p-p38(MAPK)-IR cells were associated with both dense-core and diffuse plaques, but the expected age-dependent increase in the density of plaque-associated p-p38(MAPK)-IR cells was restricted to dense-core plaques. Furthermore, the density of dense-core plaque-associated p-p38(MAPK)-IR cells was inversely correlated with the size of the core within the given plaque, which supports a role for these microglia in restricting core growth. p-p38(MAPK)-IR cells were also observed throughout wildtype and TgCRND8 mouse cortical parenchyma, but the density of these non-plaque-associated cells remained constant, regardless of age or genotype. We conclude that the constitutive presence of p-p38(MAPK)-IR microglia in aging mouse brain is indicative of a longitudinal role for this kinase in normal brain physiology. Additionally, the majority of p-p38(MAPK)-IR cells were predominantly co-immunoreactive for the Macrophage-1 (CD11b/CD18) microglial marker, regardless of whether they were associated with plaques or localized to the parenchyma. We suggest that the facts that a pool of p-p38(MAPK)-IR microglia appears to restrict b-amyloid plaque core development and the non-pathological role of p-p38(MAPK) in parenchyma, needs to be considered when anticipating targeted p38(MAPK) therapeutics in the context of clinical AD.

L’activació microglial és un fenomen fisiològic que permet resoldre alteracions a l’homeostasi. Quan l’estímul desencadenant de l’activació microglial és proinflamatori, es desencadena una resposta inflamatòria en la que la micròglia produeix citocines i altres factors proinflamatoris que tenen potencial neurotòxic. Si aquesta activació es dóna de forma exacerbada o es cronifica, pot acabar tenint un efecte perniciós. De fet, diversos autors postulen que l’activació microglial pot agreujar el curs de patologies com l’Alzheimer o el Parkinson. És per això que en aquesta tesi ens vam plantejar la inhibició del fenotip proinflamatori de l’activació microglial com una diana terapèutica en les malalties neurodegeneratives. El primer objectiu va ser establir i caracteritzar un model in vitro de neuroinflamació, en el que estudiar el patró d’activació microglial a cultius tractats amb un estímul proinflamatori (LPS + interferó gamma). Així, vam determinar que la micròglia estimulada amb LPS/IFN-gamma s’activa amb un patró de resposta proinflamatori que es caracteritza per un increment en l’expressió dels enzims iNOS i COX-2 i en la producció de NO, TNF-alfa i IL-6. En una segona fase es va establir i caracteritzar un model de cocultiu de neurones i micròglia on estudiar la neurotoxicitat induïda per activació microglial en el que vam determinar que l’activació de la micròglia en resposta al LPS/IFN-gamma té un efecte neurotòxic principalment mediat per la producció de NO, mentre que el TNF-alfa la IL-6 i la COX-2 tenen un paper secundari. Una vegada establert el model experimental vam abordar la modulació de l’activació microglial inhibint els factors implicats en l’activació proinflamatòria, en concret els C/EBPs, ja que aquests factors regulen l’expressió de molts dels gens implicats en la resposta proinflamatòria microglial. La nostra hipòtesi era que la inhibició dels C/EBPs disminuiria aquesta resposta i tindria un efecte neuroprotector. La inhibició es va abordar des d’una aproximació farmacològica i una aproximació genètica. La inhibició farmacològica es va fer usant el flavonoide crisina que té efectes antiinflamatoris mediats per la inhibició dels C/EBPs en macròfags. Així, vam determinar que la crisina té un efecte antiinflamatori i neuroprotector i que aquest està mediat, si més no en part, per la inhibició de C/EBP-delta. Posteriorment, mitjançant la utilització de micròglia de ratolins deficients en C/EBP-beta o C/EBP-delta, vam determinar que l’absència de C/EBP-beta o C/EBP-delta inhibeix l’activació microglial i la neurotoxicitat associada. Els C/EBPs per tant són una possible diana terapèutica addicional per al tractament de patologies que cursen amb neuroinflamació. Per últim es va plantejar la inhibició de la sobreactivació microglial proinflamatòria potenciant els senyals inhibidors que mantenen a la micròglia en un estat quiescent, en concret la parella lligand-receptor CD200-CD200R. La nostra hipòtesi era que modulant l’activació de CD200R es podria inhibir la sobreactivació microglial i la neurotoxicitat associada. Així el tercer objectiu va ser estudiar, d’una banda quins estímuls poden modular l’expressió de CD200 a les neurones, i de l’altra quin és el paper de CD200 i CD200R en la modulació de l’activació microglial per part de les neurones. Per abordar aquest objectiu es va establir un model de mort neuronal apoptòtica i un model de mort neuronal necròtica per excitotoxicitat. En aquest model vam determinar que l’expressió de CD200 augmenta en neurones que estan morint, ja sigui per apoptosi com per necrosi, el que podria constituir un mecanisme de control de l’activació microglial en presència de mort neuronal (no es pot descartar que l’efecte observat sigui degut a una major estabilitat CD200 en front a altres proteïnes neuronals). Per últim ens vam plantejar estudiar l’efecte de la modulació de la senyalització CD200-CD200R sobre l’activació microglial i la neurotoxicitat associada, demostrant quel’expressió proteica de CD200 a neurones no es modula per estímuls antiinflamatoris com la IL-4 i la IL-10 ni proinflamatoris com el LPS i l’IFN-gamma. En resum, els resultats d’aquesta tesi mostren la utilitat dels models experimentals in vitro per estudiar la resposta inflamatòria de la micròglia i la neurotoxicitat resultant. Amb aquests models hem pogut establir la importància de la família de factors de transcripció C/EBPs en la regulació de la resposta inflamatòria microglial i suggerir la rellevància de CD200 en la comunicació entre la micròglia i les neurones que han entrat en un procés de mort cel•lular. / Microglial activation is a physiologic phenomenon that occurs in response to alterations of homeostasis. When microglia are activated by an pro-inflammatory stimulus there is an inflammatory response where cytokines and other pro-inflammatory factors with neurotoxic potential are released. A chronic or exacerbated response can have harmful effects; in fact many authors connect microglial pro-inflammatory overactivation with degenerative diseases such as Alzheimer’s or Parkinson’s disease. Given that, the main goal of this thesis was develope strategies to inhibit pro-inflammatory microglial activation that could serve as therapeutic target in neurodegenerative diseases. To do so we set up and characterized an in-vitro model of neuroinflammation using LPS + interferon-gamma as a proinflammatory stimulus. Treated microglial cells undergo a proinflammatory response pattern characterized by the release of NO, TNF-alpha and IL-6 and the increase of iNOS and COX-2 expression. When microglial cells are co-cultured with cortical neurons, the pro-inflammatory response results in neurotoxicity, an effect mediated mainly by NO release, whereas TNF-alpha, la IL-6 and COX-2 may have a secondary role in the process. Using this model we tested whether we could modulate microglial activation by inhibiting C/EBPs, a family of transcription factors involved in the regulation of pro-inflammatory genes. We first used a pharmacologic approach treating microglial cells with an anti-inflammatory compound named crisin. We determinated that crisin has anti-inflammatory and neuroprotective effects and that these effects are mediated,at least in part, by C/EBP-delta inhibition. We then used a genetic approach to inhibit C/EBPs, by using microglial cells from KO mice. We showed that the lack of C/EBP-beta or C/EBP-delta inhibits microglial activation and the associated neurotoxicity. Therefore, we suggest that C/EBPs are a potential therapeutic target in the treatment of diseases where neuroinflammation is present. We also modulated microglial activation by potentiating the inhibitory signals that keep microglia in a resting state, such as the neuronal ligand CD200 and its microglial receptor CD200R. We showed that CD200 expression appears to increase in response to neuronal death but not in response to pro-inflammatory stimuli such as IL-4, IL-10 or anti-inflammatory stimuli such as LPS and IFN-gamma. In summary we demonstrated that in-vitro models can be used to study the pro-inflammatory and neurotoxic effect of the microglial response. Using this approach we determinated the relevance of C/EBPs in regulating microglial activation and our studies support an important role of CD200 in the crosstalk between microglia and neurons that undergoing cell death.

Neurodegeneration describes the progressive loss of structure and function of neurons, leading ultimately to cell and organism death. Although the initiating factors of neurodegenerative diseases such as Alzheimer’s, Parkinson’s, Huntington’s, and Amyotrophic Lateral Sclerosis may be different, they share common pathophysiologies. Proteinopathies, as these diseases are now termed, are characterized by atypical deposits of proteins, often due to misfolding. Associated with these deposits are dysfunctional mitochondria, oxidative stress, disrupted axonal transport, inflammation, and apoptotic cell death. If this occurs in motor neurons, as in ALS, ataxia precedes death with little or no change in cognition. On the other hand, if the deposits are found in cortical neurons, as in Alzheimer’s disease, the outcome is dementia and motor function remains largely intact. Each disease is selective for particular types of neurons and brain regions. Although research has elucidated much of the molecular biology involved in these diseases, their initiating causes remain largely unknown. Most of our current understanding originated with the identification of gene mutations that cause rare familial forms of these diseases. As a result, numerous strains of transgenic animals have been developed to study neurodegenerative disease phenomena and were central to the studies presented in this body of work. Novel routes of drug and gene delivery are described here as well as characterization of the mouse models studied. In particular, this work demonstrates that the blood brain barrier is disrupted in ALS followed by the formation of autorosettes in ALS mice. In various Alzheimer’s disease mouse models, it was demonstrated that the acute phase reactant alpha-1-antichymotrypsin (ACT) not only interacts with amyloid plaques, but also induces tau phosphorylation in vivo; tying together these disease hallmarks. It was also shown that small fragments of Aβ (1-11) could disrupt the formation of mature amyloid plaques in these mice. Lastly, it was demonstrated that mature plaques could also be decreased by intracranial delivery of granulocyte-macrophage stimulating factor (GM-CSF). My dissertation research goal was to understand and develop these treatment strategies based on protein disaggregation, neuroprotection, and inflammation, meanwhile developing novel methods for targeted delivery of molecules into the CNS of mice.

Amyotrophic lateral sclerosis (ALS) is a progressive adult onset neurodegenerative disease characterized by selective death of the upper and lower motor neurons of the brain and spinal cord. Neuromuscular synapses are lost leading to paralysis and ultimately death. Non-neuronal cells, such as astrocytes, oligodendrocytes, and microglia, have been shown to contribute to ALS disease progression in mouse models. Microglia, the innate immune cells of the central nervous system, have been shown to be activated in ALS and contribute to disease progression. Hundreds of mRNAs have shown to be dysregulated in a variety of ALS cell types and tissues, including total spinal cord, acutely isolated microglia, and in vitro differentiated motor neurons. These mRNAs can be regulated post-transcriptionally by microRNAs (miRNAs), which are small endogenous non-coding RNAs with important regulatory roles in a wide range of cellular processes. This dissertation examines the contribution of miRNAs to ALS disease progression in microglia. I acutely isolated primary microglia from the spinal cords of transgenic mice overexpressing human wild type (WT) SOD1 and human G93A SOD1. I used small RNA sequencing to profile the miRNAs that are expressed during disease progression, and identified miRNAs that are differentially expressed. I confirmed these results by quantitative PCR and examined the expression changes of predicted targets in a microglia RNA-seq dataset. Here I show that miRNAs are dysregulated in acutely isolated microglia from SOD1 G93A transgenic mice, and that miR-155, a pro-inflammatory miRNA, and miR-210, a hypoxia-inducible miRNA, are significantly upregulated during disease progression. In addition, miR-1198-5p, miR-182, miR-503, and miR-668 are also dysregulated, and predicted mRNA targets of all six of these miRNAs are differentially expressed during disease progression. To my knowledge, this is the first analysis of miRNA expression in microglia during ALS disease progression. This work contributes to the understanding of the contribution of a non-neuronal cell type to ALS disease progression and serves as a paradigm for studies in other non-neuronal cell types, such as astrocytes and oligodendrocytes, and other ALS mouse models.

A novel cell culture system was designed to serve as a model of neuroinflammation. Neurons, astrocytes, and microglia derived from embryonic and perinatal rat cortical tissue were combined in a three-dimensional hydrogel utilizing a method that facilitated cell maturation and viability. Chemical challenge of the cultures with a broad pro-inflammatory stimulus resulted in the production of inflammatory cytokines and other associated molecules commensurate with the response observed in vivo and in other in vitro systems. It was hypothesized that mechanical deformation of the multitypic neural cell cultures would produce a similar response and thus validate the system as an in vitro model of traumatic brain injury-induced neuroinflammation. Mechanical injury delivered using custom-manufactured culture chambers and injury devices successfully imparted a moderate level of cell death to the cultures. It was determined that a mechanically-induced inflammatory response required chemical stimulation prior to the injury. The research presented here describes the generation and characterization of a novel in vitro culture system and its implementation in experiments designed to model secondary injury mechanisms associated with injury-induced neuroinflammation. The findings of these studies, applications of the culture system, and future research avenues are discussed.

3,4-(±)-Methylenedioxymethamphetamine (MDMA, ecstasy) is a widely abused amphetamine derivative. The metabolism of MDMA is thought to be a necessary component of MDMA-induced neurotoxicity, as direct administration of MDMA into the central nervous system of rats failed to reproduce the hallmark serotonin deficits seen following systemic administration of MDMA. Mechanistic questions remain regarding how MDMA elicits this neurotoxicity. Work of this thesis was undertaken to examine how MDMA-induced neurotoxicity is affected by the activity of two polymorphic enzymes involved in the metabolism of MDMA, namely cytochrome P450 family member 2D6 (CYP2D6) and catechol-O-methyltransferase (COMT), as well as the potential role microglia play in the facilitation of this neurotoxicity. Inhibition of CYP2D1, the homolog of human CYP2D6 in the rat, resulted in an attenuation of serotonergic neurotoxicity following MDMA-administration. In both a pharmacological model and a genetic model of CYP2D1 inhibition, serotonin deficits were alleviated when compared to normal-activity CYP2D1 counterparts. Inhibition of COMT, the primary detoxication enzyme in the MDMA pathway, resulted in potentiation of MDMA-induced neurotoxicity. In a pharmacological model of COMT inhibition, rats displayed greater long-term serotonin deficits after COMT inhibition. Mice devoid of COMT proved sensitive to the lethal hyperthermic effects of MDMA, illustrating the importance of this enzyme in preventing the acute toxicity of MDMA. Brain lesions often elicit a microglial response. Microglia have the potential of both beneficial and deleterious actions in the brain. Whether microglia are activated by nerve terminal degeneration produced by MDMA is an area of ongoing debate. Systemically delivered MDMA produces a modest increase in the amount of microglial cells present in the parietal cortex of rats over a one-week period. MDMA also increased the phagocytic activity of microglia in the cortex. The studies described herein support the hypothesis that metabolism is critical in MDMA-induced neurotoxicity. Furthermore, as both CYP2D6 and COMT are polymorphic in the human population, certain individuals are more at risk for severe serotonergic toxicity following MDMA administration. Finally, while microglia are likely not the cause of MDMA-induced neurotoxicity, contributions of these cells cannot be dismissed.

Intracerebral hemorrhage (ICH) is a devastating form of stroke that results from rupture of a blood vessel in the brain. Tissue inside the hematoma is irreversibly damaged soon after ICH onset and when this thesis research began, there was a dearth of information regarding pathological changes outside the hematoma. Inflammation is often proposed as a mechanism of injury, but very little information was available to show that inflammatory cells were in the right place at the right time to cause secondary brain injury. Using the collagenase-induced model of ICH, this work sought to better define spatial and temporal relationships between secondary brain injury and the inflammatory response after ICH. To test the hypothesis that reducing inflammation can protect the brain from secondary injury, minocycline, an antibiotic with established anti-inflammatory effects, was administered 6 hours after ICH onset. A small number of neurons die in the parenchyma bordering the hematoma between 6 hours and 3 days after ICH onset. This area was not associated with neutrophil infiltration, and most activated microglia/macrophages did not accumulate until after most neuron death had occurred. Despite a pronounced microglial response and prolonged increase in expression of many inflammatory genes, including complement receptor-3, interleukin-1 beta, interleukin-6, and interleukin-1 converting enzyme, no dying neurons were observed further outside the hematoma at any time. Interestingly, less early neuron death was observed in aged than in young animals, without a concomitant difference in the amount of tissue lost at 28 days. However, aged animals had less early microglial activation and a larger residual lesion, which might have resulted from impaired phagocytosis by activated microglia/macrophages. Minocycline was less effective in reducing microglial activation in aged animals, and did not reduce neuron death in either young or aged animals. Edema and BBB disruption was associated with degradation of the basal lamina protein, collagen type IV, and that damaged vessels are associated with tumor necrosis factor-alpha (TNFα)-positive neutrophils and active matrix metalloprotease-12 (MMP-12), all of which were reduced by delayed minocycline treatment. In contrast to ischemic stroke, there is a limited ‘penumbra’ outside the hematoma. Nevertheless, BBB damage in this region appears to be a potential target for protection. Furthermore, the prominent inflammatory response that continues for days after ICH does not appear to be associated with damage to other areas of the brain. Minocycline appears to protect the BBB by reducing neutrophil infiltration and the MMP-12 mediated basal lamina degradation. Future studies should investigate other targets for protection (i.e., white matter injury), and seek drugs that modulate the inflammatory response in aged animals and promote lesion resolution.

Microglia, the resident immune cells of the central nervous system, can become activated following injury, disease, or infection. In vitro, they can be activated by stimuli, which determine the inflammatory phenotype they will develop. In this thesis, stimulating microglia with tumor necrosis factor- and interferon- resulted in classical activation, characterized by proliferation, increased transcription of complement receptor 3 and major histocompatibility class II molecules, and elevated production and transcription of interleukin-1 and nitric oxide. Stimulation with TNF and IFN also changed the intensity of phosphorylated (activated) cyclic adenosine monophosphate response element binding protein immunoreactivity in microglia. Specifically, cells differentiated into populations with high or low pCREB intensity. This was the first example of such a response in microglia and was representative of what occurred in vivo, after ICH. Thus, the characterization of this model will be useful for future studies of this and other intracellular pathways of classically activated microglia.