Search

INTRODUCTION: Cerebrovascular dysfunction is a pathological hallmark of Alzheimer's disease (AD). Nevertheless, detecting cerebrovascular changes within bulk tissues has limited our ability to characterize proteomic alterations from less abundant cell types. METHODS: We conducted quantitative proteomics on bulk brain tissues and isolated cerebrovasculature from the same individuals, encompassing control (N = 28), progressive supranuclear palsy (PSP) (N = 18), and AD (N = 21) cases. RESULTS: Protein co‐expression network analysis identified unique cerebrovascular modules significantly correlated with amyloid plaques, cerebrovascular amyloid angiopathy (CAA), and/or tau pathology. The protein products within AD genetic risk loci were concentrated within cerebrovascular modules. The overlap between differentially abundant proteins in AD cerebrospinal fluid (CSF) and plasma with cerebrovascular network highlighted a significant increase of matrisome proteins, SMOC1 and SMOC2, in CSF, plasma, and brain. DISCUSSION: These findings enhance our understanding of cerebrovascular deficits in AD, shedding light on potential biomarkers associated with CAA and vascular dysfunction in neurodegenerative diseases. [ABSTRACT FROM AUTHOR]

Background: In Alzheimer's Disease (AD), a sub‐population of disease associated microglia present increased Kv1.3 channel abundance. Pharmacological blockade of Kv1.3 reduces Aβ pathology and dampens pro‐inflammatory function of microglia. To identify Kv1.3‐regulated neuroinflammatory mechanisms in microglia, we examined the interactome of Kv1.3 channels in vitro, and Kv1.3 deletion and inhibition in vivo. Methods: TurboID, a proximity‐dependent biotin ligase, was fused to the N‐term or C‐term of Kv1.3 and transduced in BV2 microglia. Kv1.3 and TurboID presence and activity were confirmed via qPCR, electrophysiology, flow cytometry, and western blot. Mass spectrometry(MS) of biotinylated proteins, under resting and lipopolysaccharide(LPS)‐treated conditions, identified Kv1.3 interactomes. Kcna3(Kv1.3)‐floxed mice were generated and crossed to CMV‐Cre mice for global Kv1.3 deletion. Three and six month old 5xFAD mice were treated with PAP1, a small molecule intracellular blockade of Kv1.3, for three months. At nine months of age behavioral tests were performed. Aβ pathology was evaluated using ELISA and Immunofluorescence. Brain nuclei were isolated for snRNAseq. CSF, plasma, and brain were processed for MS. Results: Proteomics of BV2‐Kv1.3‐TurboID microglia identified distinct N‐term (n = 991) and C‐term Kv1.3 (n = 849) interactors. Kv1.3 N‐term interacts with plasma membrane proteins (e.g. Cars1, Psma2) and mitochondrial trafficking proteins (Timm23). C‐term interactors were modified by LPS‐stimulation (C3, STAT1, Oasl1) (n = 36), and partly dependent on a PDZ‐binding motif (Snx3, ND3, n = 70). Predicted by MS, western blot confirmed mitochondrial Kv1.3 presence. In vivo global Kv1.3 deletion was verified by qPCR. 9month 5xFAD mice with the Kv1.3 blockade show a reduction in Aβ pathology and a rescue of fear conditioning. Brain cell type‐specific and tissue‐level effects of Kv1.3 blockade will be evaluated by snRNAseq and MS. Conclusions: We identified N and C‐term domain‐specific and context‐dependent interactors of Kv1.3 channels in microglia, including several pro‐inflammatory C‐term interactors. We developed and validated a novel Kv1.3‐floxed line for global, and conditional, Kv1.3 deletion studies in mouse models of AD pathology. Kv1.3 blockers reduce Aβ pathology and correct behavioral deficits in 5xFAD mice. The transcriptomic and proteomic studies of brain and biofluids from this cohort will provide mechanistic insights and identify novel biomarkers of therapeutic efficacy of Kv1.3 blockers. [ABSTRACT FROM AUTHOR]

Background: Neurofibrillary tangles (NFTs) comprised of the phosphorylated Tau protein are a core pathological feature of Alzheimer's disease (AD) and several other neurodegenerative diseases collectively termed tauopathies include progressive supranuclear palsy, corticobasal degeneration, and subtypes of frontotemporal dementia. Physiological Tau is involved in many processes essential for cell function via protein‐protein interactions. However, under pathological conditions, the gain or loss of Tau protein interactions is an important consequence of disease pathophysiology and may result in the 'rewiring' of Tau protein‐protein interaction (PPI) networks. Therefore, modeling and identifying Tau interacting partners could reveal signaling mechanisms and therapeutic targets in tauopathies. Method: To gain an unbiased understanding of the aggregated Tau interactors, we generated a split Turbo Tau (sTurbo Tau) proximity labeling system. We chose to express the isolated Tau repeat domain (RD) containing a pro‐aggregation P301L familial tauopathy substitution. Immunocytochemistry (ICC) and Western blot (WB) were performed in sTurbo Tau expressing cells following biotin labeling. Biotinylated proteins were affinity purified and analyzed in triplicate via mass spectrometry (MS) to identify these Tau interacting partners. Differential expression and Gene Ontology (GO) analyses were performed. We overlapped the sTurbo Tau interactors in cell culture and Tau interacting partners in AD brain lysates to further validate our model. Result: ICC displayed speckle‐like Tau‐positive aggregates in the cytoplasm and nucleus that colocalized with biotin. WB with streptavidin dye also revealed high and low molecular weight biotinylated proteins in sTurbo Tau cells. MS identified nearly 1,900 proteins enriched in sTurbo Tau lysates compared to controls. GO analysis of these putative Tau interacting proteins confirmed various pathways, including cellular structural components, RNA binding, and translation. This overlap revealed 176 shared partners with roles in translation, synaptic processing, RNA‐binding, and proteasome function. Candidate spliceosome proteins, U1‐70K, U1A, and SNRPD2, have previously been shown to aggregate and colocalize with Tau in AD. Additionally, HSP90 family members were enriched in the overlapped datasets and are the target of mechanistic studies. Conclusion: Collectively, we established a cell‐based Tau proximity labeling approach to identify Tau co‐aggregation partners relevant to human AD. This provides insight into novel pathways and therapeutic targets in disease. [ABSTRACT FROM AUTHOR]

Background: Despite decades of research into the cellular and molecular contributions of brain cell types to Alzheimer's disease (AD) pathogenesis, there are no glial‐specific disease‐modifying therapies. To address this, our group has developed the innovative cell‐specific in vivo biotinylation of proteins (CIBOP) approach to quantify astrocyte‐specific in vivo proteomic and transcriptomic contributions to AD under homeostatic and neuroinflammatory conditions. Method: In the CIBOP approach, TurboID, a biotin ligase, is selectively expressed in the cell type of interest using a conditional Cre/lox genetic strategy to label the cellular proteome. Using mass spectrometry (MS)‐based proteomics, we have found that TurboID biotinylates many RNA‐binding and ribosomal proteins. Leveraging this, we extended the CIBOP approach to obtain representative cell type‐specific transcriptomes, along with the proteome. We crossed astrocyte specific Cre (Aldh1l1‐Cre‐ert2) and Rosa26TurboID/wt floxed mice for astrocyte‐specific proteomic labeling (astrocyte‐CIBOP). Astrocyte‐CIBOP and control (Cre‐only) mice received tamoxifen, followed by biotin‐containing water and systemic lipopolysaccharide (via intraperitoneal injections) to induce a neuroinflammatory state. Biofluids (CSF and plasma) and cortical tissue were collected. While maintaining RNA‐protein interactions, cortical tissue was lysed, biotinylated proteins were enriched via streptavidin beads, and then RNA was eluted. Immunofluorescent microscopy (IF) and biochemical assays were performed to confirm astrocyte‐specific proteomic biotinylation. Result: Western blot analysis of the cortex and CSF confirmed biotinylation of the cellular proteome of astrocyte‐CIBOP mice under homeostatic and neuroinflammatory conditions when compared to controls. IF images show that the biotinylation signal is astrocyte specific, proteomic labeling does not impact astrocyte morphology, and high levels of signal was detected at astrocytic end‐feet. RNA gel electrophoresis shows low RNA yield from control mouse brain streptavidin pulldowns, and high levels of RNA from astrocyte‐CIBOP brain pulldowns. MS‐based proteomics of astrocyte‐derived biotinylated proteins and RNA‐sequencing from biotinylated protein‐RNA interactions are underway. Conclusion: We validate the CIBOP approach to capture the cortical Aldh1l1‐positive astrocyte proteome and transcriptome in vivo under homeostatic and neuroinflammatory conditions and demonstrate labeling of astrocyte‐derived CSF proteins. Our novel in vivo cell type‐specific and native‐state dual ‐omics approach provides complementary transcriptomic and proteomic information to investigate disease mechanisms, discover new biomarkers, and identify novel therapeutic targets. [ABSTRACT FROM AUTHOR]

Background: Perineuronal nets (PNNs) are extracellular matrix structures that surround neurons and proximal dendrites and regulate synaptic transmission, plasticity, and several facets of memory. PNNs are commonly associated with parvalbumin‐positive (PV+) fast‐spiking interneurons which are hypothesized to play physiological roles in memory and cognition. Despite a growing literature related to PNNs and PV+ neurons' roles in synaptic maintenance and memory, much remains unknown regarding how these structures are affected in neurodegenerative diseases such as Alzheimer's Disease (AD). In this study, we used the 5xFAD amyloidosis mouse model to investigate how AD‐like pathology affects PNNs and PV+ neurons Method: Using immunofluorescence microscopy, we stained 5xFAD and Wild Type (WT) mice at 1.8, 3, 6, and 10 months of age. We stained for PV+ neurons (using anti‐PV antibody), PNNs (using a WFA antibody), and for Amyloid‐Beta (Aβ) plaques (using a 4G8 antibody). Images were then processed using ImageJ to quantify the presence of PV+ neurons, PNNs, and the association between them in several regions of interest across genotypes and age groups. Result: We found that PNNs, but not PV+ neurons, are significantly depleted in the anterior cortex, posterior cortex, and subiculum of the 5XFAD mice in early amyloidosis (3‐6 month old 5XFAD mice). Furthermore, analyses investigating the co‐localization of PNNs and PV+ neurons in these regions found that PNNs were depleted surrounding PV+ neurons in an age‐dependent manner in the 5XFAD mice compared to WT. Specifically, in early amyloidosis (3‐6 months) there was a general depletion of PNNs, but in later amyloidosis (6 ‐10 months) the depletion of PNNs was preferentially surrounding PV+ neurons. Finally, assessing the relationship between Aβ plaque proximity and the co‐localization between PNNs and PV+ neurons revealed that PNNs closer to Aβ plaques were more likely to be associated with PV+ neurons Conclusion: Collectively, these findings may indicate the time‐dependent impact of Aβ pathology on PNNs, both in general and specifically surrounding PV+ neurons. Our ongoing studies will further characterize PNN and PV changes across a wider age spectrum. [ABSTRACT FROM AUTHOR]

Background: Immune and vascular mechanisms of aging and Alzheimer's disease (AD) are mediated by microglia and endothelial cells, respectively. Transcriptomic studies of endothelial cells and microglia have revealed distinct and overlapping molecular features between these cell types while brain endothelial proteomic signatures remain unknown. Using concurrent isolation of microglia and endothelial cells from adult mouse brain along with multiplex quantitative proteomics, we contrast the proteomic profiles of these cell‐types and define their unique age‐related proteomic characteristics, for the first time. Method: The study cohort consisted of 3 and 7‐month‐old C57BL/6J mice (N = 6/age, 3M and 3F). Endothelial cells and microglia were concurrently isolated from fresh, whole mouse brain using fluorescence‐activated cell sorting (FACS). On average, 14,000 endothelial cells and 20,000 microglia were sorted directly into 8M urea and digested. Peptides from each sample were randomized into four batches and labeled using 16‐plex tandem‐mass‐tag (TMT) kits. After high pH fractionation, peptides were analyzed by liquid chromatography‐tandem mass spectrometry (LC‐MS/MS) with synchronous precursor selection MS3 based quantification (SPS‐MS3). Raw files were processed using Proteome Discoverer. The spectra were searched against UniProt mouse proteome database. Differential expression analysis and Gene Ontology analysis (GO) was performed. Result: Quantitative proteomics of endothelial cells and microglia from adult mice identified 512 proteins. Of these, 199 proteins differentiated microglia from endothelial cells (Fig.1A) of which 20 proteins had >4‐fold higher expression in microglia (e.g., Rgs10) and 60 proteins had >4‐fold higher expression in endothelial cells (e.g., Hspa12b). 51 proteins demonstrated age‐dependent changes in microglia (Fig. 1B; 18 increased and 33 decreased). 35 proteins demonstrated age‐dependent changes in endothelial cells (Fig.1B; 15 increased and 20 decreased). GO analysis showed age‐dependent increase in structural, morphogenesis, and intermediate filament proteins and decrease in cell motility and macromolecular biosynthetic proteins in microglia. Age‐dependent changes in endothelial cells included increased gene transcription, splicing, and subcellular transport and decrease in chromatin organization and cytokine production proteins. Conclusion: In this direct comparison of proteomic profiles of adult mouse microglia and endothelial cells, we report proteomic signatures that differentiate microglia and endothelial cells and identify unique age‐dependent changes in microglia and endothelial cells. [ABSTRACT FROM AUTHOR]

Background: Despite advancements in systems‐level analyses to Alzheimer's disease, global proteomic profiling of brain homogenate cannot resolve molecular changes occurring in distinct cell types. Proximity labeling of proteins by the biotin‐ligase TurboID, coupled with mass spectrometry (MS) analyses of biotinylated proteins, can resolve cell‐type specific proteomes from brain without cell type enrichment. Before in‐vivo applications, it is important to identify proteins preferentially labeled by TurboID in mammalian cells, and directly compare biotinylated proteomes with whole cell proteomes. We determined whether TurboID‐labeled proteomes in microglial (BV2) and neuronal (N2A) cell lines are representative of total cellular proteomes under homeostatic and activated conditions. Method: We generated N2A and BV2 cell lines that stably express TurboID fused to a nuclear export sequence. Cells incubated in 200μM biotin and 1μg/mL lipopolysaccharide (LPS) for 48hrs to mimic inflammatory stress. Western blotting and label‐free quantitation MS identified biotinylated proteins in whole‐cell‐lysates and streptavidin‐enriched preparations. Result: Western‐blot confirmed robust biotinylation in TurboID N2A and BV2 cell‐lines. LFQ‐MS quantified >3,000 proteins in inputs and >2,400 proteins in biotinylated proteomes from both cell types. In BV2 and N2A inputs, TurboID expression and biotinylation didn't impact the proteome, evidenced by few differentially‐expressed proteins and absence of cellular respiration measures in mitochondrial stress‐tests, comparing TurboID and non‐TurboID cells. Of all the proteins quantified in BV2 and N2A TurboID inputs, ∼50% were also quantified in the biotinylated proteomes in both cell types. Compared to input proteomes of both cell types, the biotinylated proteome was enriched in vesicular, ribosomal, mitochondrial, cytoplasmic and cytoskeletal proteins. The biotinylated proteome successfully differentiated BV2 and N2A cells. LPS stimulation altered levels of 759 proteins in BV2 cells in the input proteome, of which the top increased (e.g. Irg1, Oasl1, Cyb5a) and decreased (e.g. Mrc1, Mgl2, Arg1) proteins showed concordant changes in the biotinylated BV2 proteome. Conclusion: Our analyses show that proteomic labeling by TurboID in microglial and neuronal cell lines doesn't disrupt cellular function. The resulting biotinylated proteomes are reasonably representative of the whole‐cell proteome and captures the most robust alterations induced by immune stimuli in microglia. These foundational data will guide future applications of proximity‐labeling for cell‐type‐specific proteomics in‐vivo. [ABSTRACT FROM AUTHOR]

Background: Alzheimer's disease (AD) is the most common neurodegenerative disorder defined by progressive pathological protein aggregation (amyloid‐beta and tau) and deterioration of cognitive function. Microglia‐mediated neuroinflammation is a key pathological component of AD; however, there are critical gaps in our understanding of how microglia perpetuate AD pathology. One proposed mechanism of microglia‐mediated neuroinflammation and neurodegeneration is exosome release because of their critical role in the transport of macromolecules between cells to facilitate intercellular communication. Thus, it is possible that microglia‐derived exosomes transfer pathogenic cargo which could perpetuate AD pathology. The proteomic profiles and influence of different microglia‐derived exosomal populations on AD pathology remain unknown. We hypothesize that different microglia states determine the molecular composition of exosomes. Method: We treated a murine microglia cell line, BV2 cells, with various cytokines to polarize them and collect their exosomes for downstream mass spectrometry (MS) analyses. Three groups of BV2 cells (n=4/group) were treated with either lipopolysaccharide (LPS) to polarize to a pro‐inflammatory state, interleukin 10 (IL‐10) to polarize to an anti‐inflammatory state, or transforming growth factor beta (TGF‐β) to polarize to a homeostatic state. Untreated BV2 cells served as a control group. Following 72 hours of treatment, BV2 cells were lysed and cell culture media was collected for exosome isolation. Result: Transmission electron microscopy images and western blotting for exosomal marker, CD9, confirmed exosome purification by our isolation method. In MS studies, we identified 533 proteins in exosome fractions and 1,866 proteins in BV2 cell proteomes. We found that known exosome related proteins, Sdcbp and Igsf8, were significantly increased in the exosomal proteome compared to the whole cell proteome. We identified proteins that are differentially expressed across polarization with unique proteins increased in exosomes derived from LPS, IL‐10, and TGF‐β treated BV2 cells. Validation studies of these top hits are ongoing, and studies are underway to replicate these findings in primary microglia. Conclusion: Our results indicate that exosomes derived from microglia adopt distinct state‐associated protein profiles which may have differential effects on other cell types. This work will guide future studies concerning the role of exosomal cargo in perpetuating AD pathology. [ABSTRACT FROM AUTHOR]

Background: Women are disproportionately affected by Alzheimer's disease (AD), even after adjusting for age and socioeconomic status. Sex‐based biological differences in immune response, including hormonal and X‐linked immune genes, may contribute to differences in disease progression. However, we have limited understanding of the role of sex in modulating immune response and how these differences are affected by AD pathology. Here, we hypothesized that female and male mice would exhibit distinctive neural immune signaling in response to challenge with the pro‐inflammatory stimulus lipopolysaccharides (LPS), and that these differences would be exacerbated in model of AD pathology compared to wild‐type controls. Methods: Female and male six‐month‐old 5xFAD amyloid beta (Aβ) and wild type (WT) littermate mice were interperitoneally injected with LPS or saline vehicle once daily for 0,1, or 4 days (N=4). Luminex multiplexed ELISAs was used to quantify inflammatory cytokines, MAPK intracellular signaling phospho‐proteins in cortical tissues. Results: In the saline group, we observed no differences between females and males in either 5xFAD or WT mice. Additionally, in wild type mice, we observed no significant differences between females and males after either one or four daily injections of LPS. In contrast, in 5xFAD mice injected with LPS, we found significant differences between females and males after either one or four once daily injections. In particular, female 5xFAD mice after a single injection exhibited significantly increased profile (p=0.014, two tailed t‐test) of phospho‐proteins and cytokines, including MEK, cJUN, IL‐13, IL‐17, IL‐10, and IL‐1α compared to males. After four once‐daily LPS injections, female 5xFAD mice exhibited a significantly different profile (p=0.016, two tailed t‐test) consisting of increased phosphorylation of ATF2, Mek, cJUN, and HSP‐27, together with decreased expression of multiple proinflammatory cytokines including IL‐9, VEGF, MIP‐1β compared to males. Conclusions: Our data reveal that although sex is not a significant determinant of neural immune response in WT mice, it significantly affects neural immune response to LPS challenge in 5xFAD mice. These data suggest distinctive immune processes in female vs male 5xFAD mice that are not present in healthy controls. Future work will consider the effects of Aβ pathology on these results. [ABSTRACT FROM AUTHOR]

Background: Biologically relevant insights into cellular disease mechanisms of neurons and glia can be obtained by complimentary molecular profiling of the transcriptome and proteome of these cells. While mutually exclusive pipelines are available, the information surveyed is often from different samples leading to poor correlation between the transcriptome and proteome. Our goal was to develop a method for concomitant cell type‐specific analyses of RNA and protein. Method: We utilized a proximity‐labeling strategy that uses the biotin ligase, TurboID, to efficiently label the proteome, in a mouse microglial BV2 cell line. We created a stable BV2‐TurboID cell line that expresses TurboID fused to a nuclear export sequence. Biotin treatment of BV2‐TurboID cells resulted in robust biotinylation of the cellular proteome as confirmed by Western blot and by label‐free quantitation mass spectrometry (LFQ‐MS) of biotinylated proteins enriched with streptavidin beads. LFQ‐MS revealed that TurboID biotinylates several RNA‐binding proteins (RBPs) including ribosomal units, suggesting that transcripts associated with RBPs may also be pulled down simultaneously with proteins. To test this, we homogenized BV2‐Turbo and control BV2 cells, enriched biotinylated proteins with streptavidin beads while maintaining RNA‐protein interactions, and then eluted RNA. Result: Quality control studies showed negligible mRNA from streptavidin pulldowns from control BV2 cells, while BV2‐TurboID pulldowns had larger mRNA yields with high quality. NanoString neuroinflammatory profiling (800 genes) of whole cell RNA from control BV2 and BV2‐TurboID cells and streptavidin pulldowns from both cell types were performed. We observed a 23‐fold higher mRNA yield in the BV2‐TurboID pulldowns compared to control BV2 cells for 550 genes included in the analyses. Transcript abundances from total RNA and BV2‐TurboID pulldowns were highly comparable (R2=0.97; no differentially expressed genes) with equivalent abundance of microglial genes (e.g., Spp1 and Apoe) suggesting a faithful transcriptome capture. Conclusion: Our novel TurboID proximity labeling approach can simultaneously capture cell type‐specific transcriptomes and proteomes. We are now validating this method in other cell types using RNA‐sequencing and MS approaches. Once validated, this concurrent RNA and protein profiling approach can be applied to in vivo and ex vivo model systems to investigate the distinct roles brain cell types play in development, aging, and disease. [ABSTRACT FROM AUTHOR]

Background: Alzheimer's disease (AD) is a devastating neurodegenerative disease characterized by progressive neuronal loss, accumulation of amyloid‐beta(Aβ) plaques, and neuroinflammation. Microglia, the immune cell of the brain, transition to disease associated microglia (DAM) in AD. A subset of DAM, known as proinflammatory DAM, promote neuronal injury and release proinflammatory cytokines. Our lab established that proinflammatory DAM highly express Kv1.3 potassium channel, which consist of a homotetramer of Kv1.3 that interacts with a homotetramer of Kvβ2. Blockade of the Kv1.3 channel reduces microglial immune responses in mouse models. Both Kv1.3 and Kvβ2 contain motifs that interact with immune signaling proteins, however more evidence is needed to support this hypothesis. TurboID is a biotin ligase that biotinylates proteins within a 10nm radius. Split‐TurboID consists of two functionally inactive fragments of TurboID that within proximity of each other, fuse and actively biotinylate proteins within proximity. We will identify molecular interactors of Kv1.3 channels in mammalian cells using proximity labeling. Method: We fused TurboID or Split‐TurboID to the N‐ or C‐terminal Kv1.3 and Kvβ2 using a 15 amino‐acid inert linker and transfected the constructs into Hek293 cells. TurboID not fused to Kv1.3 and sham transfection acted as controls. Flow cytometry and patch‐clamp electrophysiology confirmed the presence of functional Kv1.3 channels. Biotinylated proteins were streptavidin bead affinity‐enriched from cell lysates and confirmed using western blots. Result: Label free quantitation mass spectrometry (LFQ‐MS) identified 2,195 biotinylated proteins of which very few were enriched in negative controls. There were 313 proteins as N‐term interactors, 385 as C‐term interactors, and 226 proteins common to both N and C term fusions, representing proteins indirectly associated with Kv1.3 channels. we identified 14 unique N‐term interactors (including metabolic proteins, eg. Naca, Txnl1, Stub1) and 38 unique C‐term interactors (including membrane trafficking proteins, eg. Ank3, Plekha7, Erlec1). Split‐TurboID interactions between Kv1.3 and Kvβ2 indicate that these interacting proteins are largely dependent on the fully functional Kv1.3 with the Kvβ2 subunit present. We packaged these plasmids into lentiviruses to transduce into mouse microglial cell lines. Conclusion: Our findings indicate Kv1.3 channels interacting with Kvβ2 are a novel molecular mechanism of immune regulation in mammalian cells. [ABSTRACT FROM AUTHOR]