Bilal Ersen Kerman

Anti‐amyloid therapies for Alzheimer's disease (AD) modestly slow cognitive decline but carry significant risk of amyloid‐related imaging abnormalities (ARIAs), brain swelling, and hemorrhage, particularly in apolipoprotein E ε4 carriers. Cerebral amyloid angiopathy (CAA) and vascular inflammation drive this vulnerability, highlighting the need for complementary strategies targeting upstream mechanisms of vascular injury. Cytosolic phospholipase A2 (cPLA2) regulates arachidonic acid and lysophosphatidylcholine‐derived lipid signaling at the intersection of amyloid burden, oxylipin dysregulation, blood–brain barrier disruption, and neurovascular inflammation. By depleting protective membrane plasmalogens while amplifying inflammatory lipid mediators, cPLA2 creates a state of vascular vulnerability predisposing to ARIAs. This Perspective article synthesizes evidence from human, preclinical, and translational studies positioning cPLA2 as an upstream driver of CAA‐related inflammation and vascular vulnerability in AD. We discuss biomarker and imaging approaches to assess cPLA2 activity in vivo and outline how targeting this pathway may enhance anti‐amyloid therapy safety by mitigating ARIA risk.

Cellular senescence is a major contributor to aging‐related degenerative diseases, including Alzheimer's disease (AD), but much less is known about the key cell types and pathways driving senescence mechanisms in the brain. We hypothesized that dysregulated cholesterol metabolism is central to cellular senescence in AD. We analyzed single‐cell RNA‐seq data from the ROSMAP and SEA‐AD cohorts to uncover cell type‐specific senescence pathologies. In ROSMAP snRNA‐seq data (982,384 nuclei from postmortem prefrontal cortex), microglia emerged as central contributors to AD‐associated senescence phenotypes among non‐neuronal cells. Homeostatic, inflammatory, phagocytic, lipid‐processing, and neuronal‐surveillance microglial states were associated with AD‐related senescence in both ROSMAP (152,459 microglia nuclei from six brain regions) and SEA‐AD (82,486 microglia nuclei) via integrative analysis. We assessed top senescence‐associated bioprocesses and demonstrated that senescent microglia exhibit altered cholesterol‐related processes and dysregulated cholesterol metabolism. We identified three gene co‐expression modules representing cholesterol‐related senescence signatures in postmortem brains. To validate these findings, we applied these signatures to snRNA‐seq data from iPSC‐derived microglia（iMGs) exposed to myelin, Aβ, apoptotic neurons, and synaptosomes. Treatment with AD‐related substrates altered cholesterol‐associated senescence signatures in iMGs. This study provides the first human evidence that dysregulated cholesterol metabolism in microglia drives cellular senescence in AD. Targeting cholesterol pathways in senescent microglia is an attractive strategy to attenuate AD progression.

Background: Apolipoprotein E4 (ApoE4) dysregulates cholesterol metabolism in the brain promoting the accumulation of lipid droplets, synaptic dysfunction and associates with a greater Alzheimer’s disease risk. We previously demonstrated that ApoE4 reduces the function of ATP‐binding cassette transporters A1 (ABCA1) which has important roles in cholesterol transport. The mechanisms of how ApoE4 induces less ABCA1 activity are poorly understood.Method: ABCA1 binding proteins were analyzed with mass spectrum after immunoprecipitation. Immortalized astrocytes, primary astrocytes, mouse brain and human postmortem brain samples were used to test the association of ABCA1 and caveolin‐1 levels. Small interfere RNA was used to knockdown caveolin‐1 expression and 3H labelled cholesterol was used to test the cholesterol efflux of ABCA1.Result: Here, we enriched ABCA1 binding protein complexes from ABCA1‐overexpressing cells treated with recombinant ApoE3 or ApoE4 proteins for a proteomic screen implicating endocytosis as a major pathway differentiating ApoE3 from ApoE4. Among the protein targets discovered, caveolin‐1 expression increased after ApoE4 treatment. Caveolin‐1 and ABCA1 interactions were confirmed using immunoprecipitation experiments. Cholesterol accumulation in astrocytes induced caveolin‐1 expression and promoted ABCA1 degradation. Greater caveolin‐1 and lower ABCA1 expression was observed in brains of 8, 18 and 22 months‐old ApoE4‐TR mice compared with the corresponding ApoE3‐TR mice. Isolation of single cell types from adult mouse ApoE‐TR brains confirmed lower ABCA1 and greater caveolin‐1 expression was largely driven by astrocytes and not microglia or neurons enriched fractions.Conclusion: In conclusion, cholesterol accumulation induced by ApoE4 promotes ABCA1 degradation in astrocytes partly through a caveolin‐1 dependent endocytosis pathway.

Docosahexaenoic acid [22:6(n-3), DHA], a polyunsaturated fatty acid, has an important role in regulating neuronal functions and in normal brain development. Dysregulated brain DHA uptake and metabolism are found in individuals carrying the APOE4 allele, which increases the genetic risk for Alzheimer's disease (AD), and are implicated in the progression of several neurodegenerative disorders. However, there are limited tools to assess brain DHA kinetics in vivo that can be translated to humans. Here, we report the synthesis of an ω-radiofluorinated PET probe of DHA, 22-[¹⁸F]fluorodocosahexaenoic acid (22-[¹⁸F]FDHA), for imaging the uptake of DHA into the brain. Using the nonradiolabeled 22-FDHA, we confirmed that fluorination of DHA at the ω-position does not significantly alter the anti-inflammatory effect of DHA in microglial cells. Through dynamic PET-MR studies using mice, we observed the accumulation of 22-[¹⁸F]FDHA in the brain over time and estimated DHA's incorporation coefficient (K*) using an image-derived input function. Finally, DHA brain K* was validated using intravenous administration of 15 mg/kg arecoline, a natural product known to increase the DHA K* in rodents. 22-[¹⁸F]FDHA is a promising PET probe that can reveal altered lipid metabolism in APOE4 carriers, AD, and other neurologic disorders. This new probe, once translated into humans, would enable noninvasive and longitudinal studies of brain DHA dynamics by guiding both pharmacological and nonpharmacological interventions for neurodegenerative diseases.

Myelin is an essential component of the nervous system and myelin damage causes demyelination diseases. Myelin is a sheet of oligodendrocyte membrane wrapped around the neuronal axon. In the fluorescent images, experts manually identify myelin by co-localization of oligodendrocyte and axonal membranes that fit certain shape and size criteria. Because myelin wriggles along x-y-z axes, machine learning is ideal for its segmentation. However, machine-learning methods, especially convolutional neural networks (CNNs), require a high number of annotated images, which necessitate expert labor. To facilitate myelin annotation, we developed a workflow and software for myelin ground truth extraction from multi-spectral fluorescent images. Additionally, to the best of our knowledge, for the first time, a set of annotated myelin ground truths for machine learning applications were shared with the community.

Motivation: Identifying and prioritizing disease-related proteins is an important scientific problem to develop proper treatments. Network science has become an important discipline to prioritize such proteins. Multiple sclerosis, an autoimmune disease for which there is still no cure, is characterized by a damaging process called demyelination. Demyelination is the destruction of myelin, a structure facilitating fast transmission of neuron impulses, and oligodendrocytes, the cells producing myelin, by immune cells. Identifying the proteins that have special features on the network formed by the proteins of oligodendrocyte and immune cells can reveal useful information about the disease. Results: We investigated the most significant protein pairs that we define as bridges among the proteins providing the interaction between the two cells in demyelination, in the networks formed by the oligodendrocyte and each type of two immune cells (i.e. macrophage and T-cell) using network analysis techniques and integer programming. The reason, we investigated these specialized hubs was that a problem related to these proteins might impose a bigger damage in the system. We showed that 61%–100% of the proteins our model detected, depending on parameterization, have already been associated with multiple sclerosis. We further observed the mRNA expression levels of several proteins we prioritized significantly decreased in human peripheral blood mononuclear cells of multiple sclerosis patients. We therefore present a model, BriFin, which can be used for analyzing processes where interactions of two cell types play an important role. Availability and implementation: BriFin is available at https://github.com/BilkentCompGen/brifin.

Automated screening systems in conjunction with machine learning-based methods are becoming an essential part of the healthcare systems for assisting in disease diagnosis. Moreover, manually annotating data and hand-crafting features for training purposes are impractical and time-consuming. We propose a segmentation and classification-based approach for assembling an automated screening system for the analysis of calcium imaging. The method was developed and verified using the effects of disease IgGs (from Amyotrophic Lateral Sclerosis patients) on calcium (Ca²⁺) homeostasis. From 33 imaging videos we analyzed, 21 belonged to the disease and 12 to the control experimental groups. The method consists of three main steps: projection, segmentation, and classification. The entire Ca²⁺ time-lapse image recordings (videos) were projected into a single image using different projection methods. Segmentation was performed by using a multi-level thresholding (MLT) step and the Regions of Interest (ROIs) that encompassed cell somas were detected. A mean value of the pixels within these boundaries was collected at each time point to obtain the Ca²⁺ traces (time-series). Finally, a new matrix called feature image was generated from those traces and used for assessing the classification accuracy of various classifiers (control vs. disease). The mean value of the segmentation F-score for all the data was above 0.80 throughout the tested threshold levels for all projection methods, namely maximum intensity, standard deviation, and standard deviation with linear scaling projection. Although the classification accuracy reached up to 90.14%, interestingly, we observed that achieving better scores in segmentation results did not necessarily correspond to an increase in classification performance. Our method takes the advantage of the multi-level thresholding and of a classification procedure based on the feature images, thus it does not have to rely on hand-crafted training parameters of each event. It thus provides a semi-autonomous tool for assessing segmentation parameters which allows for the best classification accuracy.

Myelin is an essential component of the nervous system and myelin damage causes demyelination diseases. Myelin is a sheet of oligodendrocyte membrane wrapped around the neuronal axon. In the fluorescent images, experts manually identify myelin by co-localization of oligodendrocyte and axonal membranes that fit certain shape and size criteria. Because myelin wriggles along x-y-z axes, machine learning is ideal for its segmentation. However, machine-learning methods, especially convolutional neural networks (CNNs), require a high number of annotated images, which necessitate expert labor. To facilitate myelin annotation, we developed a workflow and software for myelin ground truth extraction from multi-spectral fluorescent images. Additionally, to the best of our knowledge, for the first time, a set of annotated myelin ground truths for machine learning applications were shared with the community.