Amin Derouiche

Brain metabolism is a fascinating and exciting research topic. Many new metabolic features of the brain continue to be unveiled and much remains to be discovered. To truly enable scientific progress, it is imperative that discoveries and theories are continually challenged through scholarly discussion. Here we address an earlier editorial by Bolaños et al. 2025 to underline the importance of embracing and respectfully engaging in scientific debate. image

Synaptic activity imposes high demands of local energy production on astrocytes. However, the (an)aerobic pathways and fuel for generation of energy equivalents in astrocytes are still debated. Also, mechanisms to ensure rapid metabolic adaptation to bouts of neuronal activity have not been sufficiently explored. Here, we show a mechanism in astrocytes linking extracellular glutamate to upregulation of oxidative phosphorylation. We stimulated primary astrocytes with glutamate, and applied fluorescent immunocytochemistry with anti‐protein kinase Cδ (PKCδ), anti‐pyruvate dehydrogenase (PDH) and anti‐phospho‐PDH antibodies, and object oriented image analysis. Glutamate induces mitochondrial translocation of PKCδ and subsequent activation of the mitochondrial enzyme PDH—the point‐of‐no‐return in the utilization of carbohydrates. Using the specific mGlu5 antagonist 2‐Methyl‐6‐(phenylethynyl)pyridine hydrochloride (MPEP), the metabotropic glutamate receptor 5 (mGlu5) was identified as the key receptor inducing mitochondrial PKCδ translocation and PDH activation. We demonstrate by luminometric ATP assay and subtype‐specific inhibitors of PKC and mGlu5 that the distinct initial drop in intracellular ATP following glutamate application is counteracted by the mGlu5/PKCδ‐dependent mitochondrial activation. mGlu5 inhibition decreases ATP production also in astrocytes in the acute brain slice. Collectively, these findings reveal that astrocytes possess a potential for oxidative phosphorylation that can be stimulated by extracellular glutamate and the mGlu5/PKCδ/PDH axis, suggesting targets for pathologies involving excess glutamate. This also focuses the issue of activity‐induced glia‐neuronal metabolic interaction on perisynaptic energetics and the glia‐synaptic microenvironment. Up‐regulation of astrocytic metabolism via the mGlu5/PKCδ/PDH axis may affect only those perisynaptic astrocyte processes (PAPs) close to the active synapse(s), leaving other astrocyte domains and the whole cell unchanged. image

Astrocytes contribute to many higher brain functions. A key mechanism in glia-to-neuron signalling is vesicular exocytosis; however, the identity of exocytosis organelles remains a matter of debate. Since vesicles derived from the trans-Golgi network (TGN) are not considered in this context, we studied the astrocyte TGN by immunocytochemistry applying anti-Rab6A. In mouse brain, Rab6A immunostaining is found to be unexpectedly massive, diffuse in all regions, and is detected preferentially and abundantly in the peripheral astrocyte processes, which is hardly evident without glial fibrillary acid protein (GFAP) co-staining. All cells positive for the astrocytic markers glutamine synthetase (GS), GFAP, aldehyde dehydrogenase 1 family member L1 (Aldh1L1), or SRY (sex determining region Y)-box 9 (SOX9) were Rab6A⁺. Rab6A is excluded from microglia, oligodendrocytes, and NG2 cells using cell type-specific markers. In human cortex, Rab6A labelling is very similar and associated with GFAP⁺ astrocytes. The mouse data also confirm the specific astrocytic labelling by Aldh1L1 or SOX9; the astrocyte-specific labelling by GS sometimes debated is replicated again. In mouse and human brain, individual astrocytes display high variability in Rab6A⁺ structures, suggesting dynamic regulation of the glial TGN. In summary, Rab6A expression is an additional, global descriptor of astrocyte identity. Rab6A might constitute an organelle system with a potential role of Rab6A in neuropathological and physiological processes.

Astrocytes are increasingly perceived as active partners in physiological brain function and behaviour. The structural correlations of the glia–synaptic interaction are the peripheral astrocyte processes (PAPs), where ezrin and radixin, the two astrocytic members of the ezrin-radixin-moesin (ERM) family of proteins are preferentially localised. While the molecular mechanisms of ERM (in)activation appear universal, at least in mammalian cells, and have been studied in great detail, the actual ezrin and radixin kinases, phosphatases and binding partners appear cell type specific and may be multiplexed within a cell. In astrocytes, ezrin is involved in process motility, which can be stimulated by the neurotransmitter glutamate, through activation of the glial metabotropic glutamate receptors (mGluRs) 3 or 5. However, it has remained open how this mGluR stimulus is transduced to ezrin activation. Knowing upstream signals of ezrin activation, ezrin kinase(s), and membrane-bound binding partners of ezrin in astrocytes might open new approaches to the glial role in brain function. Ezrin has also been implicated in invasive behaviour of astrocytomas, and glial activation. Here, we review data pertaining to potential molecular interaction partners of ezrin in astrocytes, with a focus on PKC and GRK2, and in gliomas and other diseases, to stimulate further research on their potential roles in glia-synaptic physiology and pathology.

Assessing the amount and subcellular distribution of protein expression is a key component in modern cell biological and medical research. We studied protein kinase Cδ (PKCδ) as a potential regulator of mitochondrial metabolism in astrocytes, and sought to evaluate mitochondrial translocation of PKCδ since this is an important determinant of its function. Apart from visualizing compartment specific localization of mobile proteins such as PKCδ, we also wanted to determine what amount of a cell’s total content of a particular protein is located to a specific compartment, or translocated comparing control and experimental condition. We develop a semiquantitative parameter that indicates the relative protein distribution to two subcellular compartments, starting from standard two-channel fluorescence microscopy images. We studied the mitochondrial translocation of PKCδ in astrocytes using double immunofluorescence microscopy and object-oriented image analysis. In one channel, the protein of interest (PKCδ) is labeled, in the other the compartment or organelle of interest (mitochondria, using cytochrome oxidase IV). Both channels were binarized, turned into object populations, and the channel specific values for total area and integrated intensity extracted. From these values, the “intensity density ratio” (IDR) is calculated, a standardized parameter to easily compare distribution patterns in different cells or ROIs. IDR is highly sensitive to changes in localization pattern, and thus easily detects protein translocation in comparison between control and experimental condition. In our study, medium application of glutamate was found to result in partial PKCδ translocation to mitochondria, a statistically highly significant result based only on a limited number of acquired images.

In multiple sclerosis, coagulation factors have been shown to modulate inflammation. In this translational study, we investigated whether long-term anticoagulation with warfarin or rivaroxaban has beneficial effects on the course of autoimmune experimental encephalomyelitis (EAE). Female SJL/J mice treated with anticoagulants namely warfarin or rivaroxaban were immunized with PLP139–151. Stable anticoagulation was maintained throughout the entire experiment. Mice without anticoagulation treated with the vehicle only were used as controls. The neurological deficit was recorded during the course of EAE, and histopathological analyses of inflammatory lesions were performed. In preventive settings, both treatment with warfarin and rivaroxaban reduced the maximum EAE score as compared to the control group and led to a reduction of inflammatory lesions in the spinal cord. In contrast, therapeutic treatment with warfarin had no beneficial effects on the clinical course of EAE. Signs of intraparenchymal hemorrhage at the site of the inflammatory lesions were not observed. We developed long-term anticoagulation models that allowed exploring the course of EAE under warfarin and rivaroxaban treatment. We found a mild preventive effect of both warfarin and rivaroxaban on neurological deficits and local inflammation, indicating a modulation of the disease induction by anticoagulation.

Connexin 43 (Cx43) is the main astrocytic connexin and forms the basis of the glial syncytium. The morphology of connexin‐expressing cells can be best studied in transgenic mouse lines expressing cytoplasmic fluorescent reporters, since immunolabeling the plaques can obscure the shapes of the individual cells. The Cx43kiECFP mouse generated by Degen et al. (FASEBJ 26:4576, 2012) expresses cytosolic ECFP and has previously been used to establish that Cx43 may not be expressed by all astrocytes within a population, and this can vary in a region‐dependent way. To establish this mouse line as a tool for future astrocyte and connexin research, we sought to consolidate reporter authenticity, studying cell types and within‐region population heterogeneity. Applying anti‐GFP, all cell types related to astroglia were positive—namely, protoplasmic astrocytes in the hippocampus, cortex, thalamus, spinal cord, olfactory bulb, cerebellum with Bergmann glia and astrocytes also in the molecular layer, and retinal Müller cells and astrocytes. Labeled cell types further comprise white matter astrocytes, olfactory ensheathing cells, radial glia‐like stem cells, retinal pigment epithelium cells, ependymal cells, and meningeal cells. We furthermore describe a retinal Cx43‐expressing amacrine cell morphologically reminiscent of ON‐OFF wide‐field amacrine cells, representing the first example of a mammalian CNS neuron‐expressing Cx43 protein. In double staining with cell type–specific markers (GFAP, S100ß, glutamine synthetase), Cx43 reporter expression in the hippocampus and cortex was restricted to GFAP⁺ astrocytes. Altogether, this mouse line is a highly reliable tool for studies of Cx43‐expressing CNS cells and astroglial cell morphology.

Astrocytes in vivo extend thin processes termed peripheral astrocyte processes (PAPs), in particular around synapses where they can mediate glia-neuronal communication. The relation of PAPs to synapses is not based on coincidence, but it is not clear which stimuli and mechanisms lead to their formation and are active during process extension/ retraction in response to neuronal activity. Also, the molecular basis of the extremely fine PAP morphology (often 50 to 100 nm) is not understood. These open questions can be best investigated under in vitro conditions studying glial filopodia. We have previously analyzed filopodial mechanisms (Lavialle et al. PNAS 108:12915) applying an automated method for filopodia morphometry, which is now described in greater detail. The Filopodia Specific Shape Factor (FSSF) developed integrates number and length of filopodia. It quantifies filopodia independent of overall astrocytic shape or size, which can be intricate in itself. The algorithm supplied here permits automated image processing and measurements using ImageJ. Cells have to be sampled in higher numbers to obtain significant results. We validate the FSSF, and characterize the systematic influence of thresholding and camera pixel grid on measurements. We provide exemplary results of substance-induced filopodia dynamics (glutamate, mGluR agonists, EGF), and show that filopodia formation is highly sensitive to medium pH (CO₂) and duration of cell culture. Although the FSSF was developed to study astrocyte filopodia with focus on the perisynaptic glial sheath, we expect that this parameter can also be applied to neuronal growth cones, non-neural cell types, or cell lines.

Perivascular endfeet of astrocytes are highly polarized compartments that ensheath blood vessels and contribute to the blood–brain barrier. They experience calcium transients with neuronal activity, a phenomenon involved in neurovascular coupling. Endfeet also mediate the uptake of glucose from the blood, a process stimulated in active brain regions. Here, we demonstrate in mouse hippocampal tissue slices that endfeet undergo sodium signaling upon stimulation of glutamatergic synaptic activity. Glutamate-induced endfeet sodium transients were diminished by TFB-TBOA, suggesting that they were generated by sodium-dependent glutamate uptake. With local agonist application, they could be restricted to endfeet and immunohistochemical analysis revealed prominent expression of glutamate transporters GLAST and GLT-1 localized towards the neuropil vs. the vascular side of endfeet. Endfeet sodium signals spread at an apparent maximum velocity of ∼120 µm/s and directly propagated from stimulated into neighboring endfeet; this spread was omitted in Cx30/Cx43 double-deficient mice. Sodium transients resulted in elevation of intracellular magnesium, indicating a decrease in intracellular ATP. In summary, our results establish that excitatory synaptic activity and stimulation of glutamate uptake in astrocytes trigger transient sodium increases in perivascular endfeet which rapidly spread through gap junctions into neighboring endfeet and cause a reduction of intracellular ATP. The newly discovered endfeet sodium signaling thereby represents a fast, long-lived and inter-cellularly acting indicator of synaptic activity at the blood–brain barrier, which likely constitutes an important component of neuro-metabolic coupling in the brain. GLIA 2016