Microglia are the brain’s immunocompetent macrophages with a unique feature that allows surveillance of the surrounding microenvironment and subsequent reactions to tissue damage, infection, or homeostatic perturbations. Thereby, microglia’s striking morphological plasticity is one of their prominent characteristics and the categorization of microglial cell function based on morphology is well established. Frequently, automated classification of microglial morphological phenotypes is performed by using quantitative parameters. As this process is typically limited to a few and especially manually chosen criteria, a relevant selection bias may compromise the resulting classifications. In our study, we describe a novel microglial classification method by morphological evaluation using a convolutional neuronal network on the basis of manually selected cells in addition to classical morphological parameters. We focused on four microglial morphologies, ramified, rod-like, activated and amoeboid microglia within the murine hippocampus and cortex. The developed method for the classification was confirmed in a mouse model of ischemic stroke which is already known to result in microglial activation within affected brain regions. In conclusion, our classification of microglial morphological phenotypes using machine learning can serve as a time-saving and objective method for post-mortem characterization of microglial changes in healthy and disease mouse models, and might also represent a useful tool for human brain autopsy samples.

Sporadic Alzheimer's disease (AD) is marked by a lengthy preclinical phase during which patients are nonsymptomatic but show pathology in variable manifestations. Whether or not neuroinflammation occurs in such nondemented individuals is unknown. We evaluated the medial temporal lobe of 66 nondemented subjects, aged 42–93, in terms of tau pathology, Aβ deposition, and microglial activation. We show that 100% of subjects had neurofibrillary degeneration (NFD), 35% had Aβ deposits, and 8% revealed microglial activation in individuals where early amyloid formation was apparent by Congo Red staining. Amyloid‐induced neuroinflammatory clusters of Iba1, CD68, and ferritin‐positive microglia were evident in the immediate vicinity of aggregated Aβ. Microglia in the adjacent neuropil were nonactivated. Thus, neuroinflammation in AD represents a highly localized phagocyte reaction, essentially a foreign body response, geared toward removal of insoluble Aβ. Because clustered microglia in some amyloid plaques were dystrophic and ferritin‐positive, we hypothesize that these cells were exhausted by their attempts to remove the aggregated, insoluble Aβ. Our findings show that the sequence of pathologic events in AD begins with tau pathology, followed by Aβ deposition, and then by microglial activation. Because only 8% of our subjects revealed all three hallmark pathologic features, we propose that these nondemented individuals were near the threshold of transitioning from nonsymptomatic to symptomatic disease. The onset of neuroinflammation in AD may thus represent a tipping point in AD pathogenesis. Our study suggests that the role of microglia in AD pathogenesis entails primarily the attempted removal of potentially toxic, extracellular material.

Polyunsaturated fatty acids (PUFAs) and eicosanoids are important mediators of inflammation. The functional role of eicosanoids in metabolic-syndrome-related diseases has been extensively studied. However, their role in neuroinflammation and the development of neurodegenerative diseases is still unclear. The aim of this study was the development of a sample pretreatment protocol for the simultaneous analysis of PUFAs and eicosanoids in mouse liver and brain. Liver and brain samples of male wild-type C57BL/6J mice (11-122 mg) were used to investigate conditions for tissue rinsing, homogenization, extraction, and storage. A targeted liquid chromatography-negative electrospray ionization tandem mass spectrometry method was applied to quantify 7 PUFAs and 94 eicosanoids. The final pretreatment protocol consisted of a 5-min homogenization step by sonication in 650 μL n-hexane/2-propanol (60:40 v/v) containing 2,6-di-tert-butyl-4-methylphenol at 50 μg/mL. Homogenates representing 1 mg tissue were extracted in a single step with n-hexane/2-propanol (60:40 v/v) containing 0.1% formic acid. Autoxidation was prevented by addition of 2,6-di-tert-butyl-4-methylphenol at 50 μg/mL and keeping the samples at 4°C during sample preparation. Extracts were dried under nitrogen and reconstituted in liquid chromatography eluent before analysis. Recovery was determined to range from 45% to 149% for both liver and brain tissue. Within-run and between-run variability ranged between 7% and 18% for PUFAs and between 1% and 24% for eicosanoids. In liver, 7 PUFAs and 15 eicosanoids were quantified; in brain, 6 PUFAs and 21 eicosanoids had significant differences within the brain substructures. In conclusion, a robust and reproducible sample preparation protocol for the multiplexed analysis of PUFAs and eicosanoids by liquid chromatography-tandem mass spectrometry in liver and discrete brain substructures was developed. Keywords Extraction. Eicosanoids. Brain lipids. Mass spectrometry. Polyunsaturated fatty acids. Tissue sample preparation Abbreviations ACN Acetonitrile ARA Arachidonic acid BHT 2,6-Di-tert-butyl-4-methylphenol COX Cyclooxygenase DHET Dihydroxyeicosatrienoic acid EET Epoxyeicosatrienoic acid ETE Eicosatetraenoic acid

Obesity arising from excessive dietary fat intake is a risk factor for cognitive decline, dementia and neurodegenerative diseases, including Alzheimer’s disease. Here, we studied the effect of long-term high-fat diet (HFD) (24 weeks) and return to normal diet (ND) on behavioral features, microglia and neurons in adult male C57BL/6J mice. Consequences of HFD-induced obesity and dietary changes on general health (coat appearance, presence of vibrissae), sensory and motor reflexes, learning and memory were assessed by applying a phenotypic assessment protocol, the Y maze and Morris Water Maze test. Neurons and microglia were histologically analyzed within the mediobasal hypothalamus, hippocampus and frontal motor cortex after long-term HFD and change of diet. Long periods of HFD caused general health issues (coat alterations, loss of vibrissae), but did not affect sensory and motor reflexes, emotional state, memory and learning. Long-term HFD increased the microglial response (increased Iba1 fluorescence intensity, percentage of Iba1-stained area and Iba1 gene expression) within the hypothalamus, but not in the cortex and hippocampus. In neither of these regions, neurodegeneration or intracellular lipid droplet accumulation was observed. The former alterations were reversible in mice whose diet was changed from HFD to ND. Taken together, long periods of excessive dietary fat alone do not cause learning deficits or spatial memory impairment, though HFD-induced obesity may have detrimental consequences for cognitive flexibility. Our data confirm the selective responsiveness of hypothalamic microglia to HFD.

Plant sterols (PSs) cannot be synthesized in mammals and are exclusively diet-derived. PSs cross the blood-brain barrier and may have anti-neuroinflammatory effects. Obesity is linked to lower intestinal uptake and blood levels of PSs, but its effects in terms of neuroinflammation—if any—remain unknown. We investigated the effect of high-fat diet-induced obesity on PSs in the brain and the effects of the PSs campesterol and β-sitosterol on in vitro microglia activation. Sterols (cholesterol, precursors, PSs) and polyunsaturated fatty acid-derived lipid mediators were measured in the food, blood, liver and brain of C57BL/6J mice. Under a PSs-poor high-fat diet, PSs levels decreased in the blood, liver and brain (>50%). This effect was reversible after 2 weeks upon changing back to a chow diet. Inflammatory thromboxane B2 and prostaglandin D2 were inversely correlated to campesterol and β-sitosterol levels in all brain regions. PSs content was determined post mortem in human cortex samples as well. In vitro, PSs accumulate in lipid rafts isolated from SIM-A9 microglia cell membranes. In summary, PSs levels in the blood, liver and brain were associated directly with PSs food content and inversely with BMI. PSs dampen pro-inflammatory lipid mediators in the brain. The identification of PSs in the human cortex in comparable concentration ranges implies the relevance of our findings for humans.