Mapping genomic loci implicates genes and synaptic biology in schizophrenia.

Schizophrenia has a heritability of 60-80%1, much of which is attributable to common risk alleles. Here, in a two-stage genome-wide association study of up to 76,755 individuals with schizophrenia and 243,649 control individuals, we report common variant associations at 287 distinct genomic loci. Associations were concentrated in genes that are expressed in excitatory and inhibitory neurons of the central nervous system, but not in other tissues or cell types. Using fine-mapping and functional genomic data, we identify 120 genes (106 protein-coding) that are likely to underpin associations at some of these loci, including 16 genes with credible causal non-synonymous or untranslated region variation. We also implicate fundamental processes related to neuronal function, including synaptic organization, differentiation and transmission. Fine-mapped candidates were enriched for genes associated with rare disruptive coding variants in people with schizophrenia, including the glutamate receptor subunit GRIN2A and transcription factor SP4, and were also enriched for genes implicated by such variants in neurodevelopmental disorders. We identify biological processes relevant to schizophrenia pathophysiology; show convergence of common and rare variant associations in schizophrenia and neurodevelopmental disorders; and provide a resource of prioritized genes and variants to advance mechanistic studies.

Splice-dependent trans-synaptic PTPδ-IL1RAPL1 interaction regulates synapse formation and non-REM sleep.

Alternative splicing regulates trans-synaptic adhesions and synapse development, but supporting in vivo evidence is limited. PTPδ, a receptor tyrosine phosphatase adhering to multiple synaptic adhesion molecules, is associated with various neuropsychiatric disorders; however, its in vivo functions remain unclear. Here, we show that PTPδ is mainly present at excitatory presynaptic sites by endogenous PTPδ tagging. Global PTPδ deletion in mice leads to input-specific decreases in excitatory synapse development and strength. This involves tyrosine dephosphorylation and synaptic loss of IL1RAPL1, a postsynaptic partner of PTPδ requiring the PTPδ-meA splice insert for binding. Importantly, PTPδ-mutant mice lacking the PTPδ-meA insert, and thus lacking the PTPδ interaction with IL1RAPL1 but not other postsynaptic partners, recapitulate biochemical and synaptic phenotypes of global PTPδ-mutant mice. Behaviorally, both global and meA-specific PTPδ-mutant mice display abnormal sleep behavior and non-REM rhythms. Therefore, alternative splicing in PTPδ regulates excitatory synapse development and sleep by modulating a specific trans-synaptic adhesion.

Power and optimal study design in iPSC-based brain disease modelling.

Studies using induced pluripotent stem cells (iPSCs) are gaining momentum in brain disorder modelling, but optimal study designs are poorly defined. Here, we compare commonly used designs and statistical analysis for different research aims. Furthermore, we generated immunocytochemical, electrophysiological, and proteomic data from iPSC-derived neurons of five healthy subjects, analysed data variation and conducted power simulations. These analyses show that published case-control iPSC studies are generally underpowered. Designs using isogenic iPSC lines typically have higher power than case-control designs, but generalization of conclusions is limited. We show that, for the realistic settings used in this study, a multiple isogenic pair design increases absolute power up to 60% or requires up to 5-fold fewer lines. A free web tool is presented to explore the power of different study designs, using any (pilot) data.

Early-life stress and amyloidosis in mice share pathogenic pathways involving synaptic mitochondria and lipid metabolism.

INTRODUCTION: Early-life stress (ES) increases the risk for Alzheimer's disease (AD). We and others have shown that ES aggravates amyloid-beta (Aß) pathology and promotes cognitive dysfunction in APP/PS1 mice, but underlying mechanisms remain unclear. METHODS: We studied how ES affects the hippocampal synaptic proteome in wild-type (WT) and APP/PS1 mice at early and late pathological stages, and validated hits using electron microscopy and immunofluorescence. RESULTS: The hippocampal synaptosomes of both ES-exposed WT and early-stage APP/PS1 mice showed a relative decrease in actin dynamics-related proteins and a relative increase in mitochondrial proteins. ES had minimal effects on older WT mice, while strongly affecting the synaptic proteome of advanced stage APP/PS1 mice, particularly the expression of astrocytic and mitochondrial proteins. DISCUSSION: Our data show that ES and amyloidosis share pathogenic pathways involving synaptic mitochondrial dysfunction and lipid metabolism, which may underlie the observed impact of ES on the trajectory of AD.

MS-DAP Platform for Downstream Data Analysis of Label-Free Proteomics Uncovers Optimal Workflows in Benchmark Data Sets and Increased Sensitivity in Analysis of Alzheimer's Biomarker Data.

In the rapidly moving proteomics field, a diverse patchwork of data analysis pipelines and algorithms for data normalization and differential expression analysis is used by the community. We generated a mass spectrometry downstream analysis pipeline (MS-DAP) that integrates both popular and recently developed algorithms for normalization and statistical analyses. Additional algorithms can be easily added in the future as plugins. MS-DAP is open-source and facilitates transparent and reproducible proteome science by generating extensive data visualizations and quality reporting, provided as standardized PDF reports. Second, we performed a systematic evaluation of methods for normalization and statistical analysis on a large variety of data sets, including additional data generated in this study, which revealed key differences. Commonly used approaches for differential testing based on moderated t-statistics were consistently outperformed by more recent statistical models, all integrated in MS-DAP. Third, we introduced a novel normalization algorithm that rescues deficiencies observed in commonly used normalization methods. Finally, we used the MS-DAP platform to reanalyze a recently published large-scale proteomics data set of CSF from AD patients. This revealed increased sensitivity, resulting in additional significant target proteins which improved overlap with results reported in related studies and includes a large set of new potential AD biomarkers in addition to previously reported.

SALM1 controls synapse development by promoting F-actin/PIP2-dependent Neurexin clustering.

Synapse development requires spatiotemporally regulated recruitment of synaptic proteins. In this study, we describe a novel presynaptic mechanism of cis-regulated oligomerization of adhesion molecules that controls synaptogenesis. We identified synaptic adhesion-like molecule 1 (SALM1) as a constituent of the proposed presynaptic Munc18/CASK/Mint1/Lin7b organizer complex. SALM1 preferentially localized to presynaptic compartments of excitatory hippocampal neurons. SALM1 depletion in excitatory hippocampal primary neurons impaired Neurexin1ß- and Neuroligin1-mediated excitatory synaptogenesis and reduced synaptic vesicle clustering, synaptic transmission, and synaptic vesicle release. SALM1 promoted Neurexin1ß clustering in an F-actin- and PIP2-dependent manner. Two basic residues in SALM1's juxtamembrane polybasic domain are essential for this clustering. Together, these data show that SALM1 is a presynaptic organizer of synapse development by promoting F-actin/PIP2-dependent clustering of Neurexin.

Blue Native PAGE-Antibody Shift in Conjunction with Mass Spectrometry to Reveal Protein Subcomplexes: Detection of a Cerebellar α1/α6-Subunits Containing γ-Aminobutyric Acid Type A Receptor Subtype.

The pentameric γ-Aminobutyric acid type A receptors (GABAARs) are ligand-gated ion channels that mediate the majority of inhibitory neurotransmission in the brain. In the cerebellum, the two main receptor subtypes are the 2α1/2ß/γ and 2α6/2ß/δ subunits. In the present study, an interaction proteomics workflow was used to reveal additional subtypes that contain both α1 and α6 subunits. Immunoprecipitation of the α6 subunit from mouse brain cerebellar extract co-purified the α1 subunit. In line with this, pre-incubation of the cerebellar extract with anti-α6 antibodies and analysis by blue native gel electrophoresis mass-shifted part of the α1 complexes, indicative of the existence of an α1α6-containing receptor. Subsequent mass spectrometry of the blue native gel showed the α1α6-containing receptor subtype to exist in two main forms, i.e., with or without Neuroligin-2. Immunocytochemistry on a cerebellar granule cell culture revealed co-localization of α6 and α1 in post-synaptic puncta that apposed the presynaptic marker protein Vesicular GABA transporter, indicative of the presence of this synaptic GABAAR subtype.

The proteome of granulovacuolar degeneration and neurofibrillary tangles in Alzheimer's disease.

Granulovacuolar degeneration (GVD) is a common feature in Alzheimer's disease (AD). The occurrence of GVD is closely associated with that of neurofibrillary tangles (NFTs) and GVD is even considered to be a pre-NFT stage in the disease process of AD. Currently, the composition of GVD bodies, the mechanisms associated with GVD and how GVD exactly relates to NFTs is not well understood. By combining immunohistochemistry (IHC) and laser microdissection (LMD) we isolated neurons with GVD and those bearing tangles separately from human post-mortem AD hippocampus (n = 12) using their typical markers casein kinase (CK)1δ and phosphorylated tau (AT8). Control neurons were isolated from cognitively healthy cases (n = 12). 3000 neurons per sample were used for proteome analysis by label free LC-MS/MS. In total 2596 proteins were quantified across samples and a significant change in abundance of 115 proteins in GVD and 197 in tangle bearing neurons was observed compared to control neurons. With IHC the presence of PPIA, TOMM34, HSP70, CHMP1A, TPPP and VXN was confirmed in GVD containing neurons. We found multiple proteins localizing specifically to the GVD bodies, with VXN and TOMM34 being the most prominent new protein markers for GVD bodies. In general, protein groups related to protein folding, proteasomal function, the endolysosomal pathway, microtubule and cytoskeletal related function, RNA processing and glycolysis were found to be changed in GVD neurons. In addition to these protein groups, tangle bearing neurons show a decrease in ribosomal proteins, as well as in various proteins related to protein folding. This study, for the first time, provides a comprehensive human based quantitative assessment of protein abundances in GVD and tangle bearing neurons. In line with previous functional data showing that tau pathology induces GVD, our data support the model that GVD is part of a pre-NFT stage representing a phase in which proteostasis and cellular homeostasis is disrupted. Elucidating the molecular mechanisms and cellular processes affected in GVD and its relation to the presence of tau pathology is highly relevant for the identification of new drug targets for therapy.

Glycine Receptor Complex Analysis Using Immunoprecipitation-Blue Native Gel Electrophoresis-Mass Spectrometry.

The pentameric glycine receptor (GlyR), comprising the α1 and ß subunits, is a major inhibitory ionotropic receptor in brainstem and spinal cord. GlyRs interact with gephyrin (GPHN), a scaffold protein that anchors the GlyR in the plasma membrane and enables it to form clusters in glycinergic postsynapses. Using an interaction proteomics approach, evidence of the ArfGEFs IQ motif and Sec7 domain 3 (IQSEC3) and IQ motif and Sec7 domain 2 (IQSEC2) as two novel synaptic proteins interacting with GlyR complexes is provided. When the affinity-isolated GlyR complexes are fractionated by blue native gel electrophoresis and characterized by mass spectrometry, GlyR α1ß-GPHN appears as the most abundant complex with a molecular weight of ≈1 MDa, and GlyR α1ß-GPHN-IQSEC3 as a minor protein complex of ≈1.2 MDa. A third GlyR α1ß-GPHN-IQSEC2 complex exists at the lowest amount with a mass similar to the IQSEC3 containing complex. Using yeast two-hybrid it is demonstrated that IQSEC3 interacts with the GlyR complex by binding to the GPHN G domain at the N-terminal of the IQSEC3 IQ-like domain. The data provide direct evidence of the interaction of IQSEC3 with GlyR-GPHN complexes, underscoring a potential role of these ArfGEFs in the function of glycinergic synapses.

MIR137 schizophrenia-associated locus controls synaptic function by regulating synaptogenesis, synapse maturation and synaptic transmission.

The MIR137 locus is a replicated genetic risk factor for schizophrenia. The risk-associated allele is reported to increase miR-137 expression and miR-137 overexpression alters synaptic transmission in mouse hippocampus. We investigated the cellular mechanisms underlying these observed effects in mouse hippocampal neurons in culture. First, we correlated the risk allele to expression of the genes in the MIR137 locus in human postmortem brain. Some evidence for increased MIR137HG expression was observed, especially in hippocampus of the disease-associated genotype. Second, in mouse hippocampal neurons, we confirmed previously observed changes in synaptic transmission upon miR-137 overexpression. Evoked synaptic transmission and spontaneous release were 50% reduced. We identified defects in release probability as the underlying cause. In contrast to previous observations, no evidence was obtained for selective synaptic vesicle docking defects. Instead, ultrastructural morphometry revealed multiple effects of miR-137 overexpression on docking, active zone length and total vesicle number. Moreover, proteomic analyses of neuronal protein showed that expression of Syt1 and Cplx1, previously reported as downregulated upon miR-137 overexpression, was unaltered. Immunocytochemistry of synapses overexpressing miR-137 showed normal Synaptotagmin1 and Complexin1 protein levels. Instead, our proteomic analyses revealed altered expression of genes involved in synaptogenesis. Concomitantly, synaptogenesis assays revealed 31% reduction in synapse formation. Taken together, these data show that miR-137 regulates synaptic function by regulating synaptogenesis, synaptic ultrastructure and synapse function. These effects are plausible contributors to the increased schizophrenia risk associated with miR-137 overexpression.

A Fast and Economical Sample Preparation Protocol for Interaction Proteomics Analysis.

A simple and fast immunoprecipitation (IP) protocol is designed with the sample preparation incorporated, applicable to both low and high throughput. This new protocol combines two procedures based on magnetic beads in 96-well plate format. Protein complexes are captured by antibodies and magnetic beads conjugated with protein A. Proteins are washed and on-bead digested by using Single-Pot solid-phase sample preparation (SP3). The whole IP-SP3 approach can be completed in one day, which is considerably faster compared to the classical approach. No major quantitative differences are found between SP3 and FASP (filter-aided sample preparation) or a longer incubation protocol. Taken together, the IP-SP3 protocol is a fast and economical approach easily applicable for large-scale protein interactome analysis.

Comparative Analyses of Data Independent Acquisition Mass Spectrometric Approaches: DIA, WiSIM-DIA, and Untargeted DIA.

Data-independent acquisition (DIA) is an emerging technology for quantitative proteomics. Current DIA focusses on the identification and quantitation of fragment ions that are generated from multiple peptides contained in the same selection window of several to tens of m/z. An alternative approach is WiSIM-DIA, which combines conventional DIA with wide-SIM (wide selected-ion monitoring) windows to partition the precursor m/z space to produce high-quality precursor ion chromatograms. However, WiSIM-DIA has been underexplored; it remains unclear if it is a viable alternative to DIA. We demonstrate that WiSIM-DIA quantified more than 24 000 unique peptides over five orders of magnitude in a single 2 h analysis of a neuronal synapse-enriched fraction, compared to 31 000 in DIA. There is a strong correlation between abundance values of peptides quantified in both the DIA and WiSIM-DIA datasets. Interestingly, the S/N ratio of these peptides is not correlated. We further show that peptide identification directly from DIA spectra identified >2000 proteins, which included unique peptides not found in spectral libraries generated by DDA.

Interaction proteomics of canonical Caspr2 (CNTNAP2) reveals the presence of two Caspr2 isoforms with overlapping interactomes.

Autism is a human developmental brain disorder characterized by impaired social interaction and communication. Contactin-associated protein-like 2 (Caspr2, CNTNAP2) is a known genetic risk factor of autism. However, how this protein might contribute to pathology is unclear. In this study, we demonstrate that Caspr2 is abundantly present in lipid raft and in the synaptic membrane but is highly depleted in the postsynaptic density. The Caspr2 protein level in hippocampus is present at a constant level during synapse formation and myelination from P0 to P84. Interaction proteomics revealed the interactors of Caspr2, including CNTN2, KCNAs, members of the ADAM family (ADAM22, ADAM23 and ADAM11), members of LGI family and MAGUKs (DLGs and MPPs). Interestingly, a short form of Caspr2 was detected, which lacks most of the extracellular domains, however, is still associated with ADAM22 and to a lesser extent LGI1 and Kv1 channels. The comprehensive Caspr2 interactome revealed here might aid in understanding the molecular mechanisms underlying autism. This article is part of a Special Issue titled Neuroproteomics: Applications in Neuroscience and Neurology.

Empirical Bayesian random censoring threshold model improves detection of differentially abundant proteins.

A challenge in proteomics is that many observations are missing with the probability of missingness increasing as abundance decreases. Adjusting for this informative missingness is required to assess accurately which proteins are differentially abundant. We propose an empirical Bayesian random censoring threshold (EBRCT) model that takes the pattern of missingness in account in the identification of differential abundance. We compare our model with four alternatives, one that considers the missing values as missing completely at random (MCAR model), one with a fixed censoring threshold for each protein species (fixed censoring model) and two imputation models, k-nearest neighbors (IKNN) and singular value thresholding (SVTI). We demonstrate that the EBRCT model bests all alternative models when applied to the CPTAC study 6 benchmark data set. The model is applicable to any label-free peptide or protein quantification pipeline and is provided as an R script.

Interaction proteomics reveals brain region-specific AMPA receptor complexes.

Fast excitatory synaptic transmission in the brain is mediated by glutamate acting on postsynaptic AMPA receptors. Recent studies have revealed a substantial number of AMPA receptor auxiliary proteins, which potentially contribute to the regulation of AMPA receptor trafficking, subcellular receptor localization, and receptor gating properties. Here we examined the AMPA receptor interactomes from cortex, hippocampus, and cerebellum by comprehensive interaction proteomics. The study reveals that AMPA receptor auxiliary proteins are engaged in distinct brain region-specific AMPA receptors subcomplexes, which might underlie brain region-specific differential regulation of AMPA receptor properties. Depending on the brain region, an interacting protein can be involved in an AMPA and a non-AMPA receptor complex.

GOAT: efficient and robust identification of gene set enrichment.

Gene set enrichment analysis is foundational to the interpretation of high throughput biology. Identifying enriched Gene Ontology (GO) terms or disease-associated gene sets within a list of gene effect sizes that represent experimental outcomes is an everyday task in life science that crucially depends on robust and sensitive statistical tools. We here present GOAT, a parameter-free algorithm for gene set enrichment analysis of preranked gene lists. The algorithm can precompute null distributions from standardized gene scores, enabling enrichment testing of the GO database in one second. Validations using synthetic data show that estimated gene set p-values are well calibrated under the null hypothesis and invariant to gene list length and gene set size. Application to various real-world proteomics and gene expression studies demonstrates that GOAT identifies more significant GO terms as compared to current methods. GOAT is freely available as an R package and user-friendly online tool for gene set enrichment analyses that includes interactive data visualizations: https://ftwkoopmans.github.io/goat .

Reduced MUNC18-1 Levels, Synaptic Proteome Changes, and Altered Network Activity in STXBP1-Related Disorder Patient Neurons.

Background: STXBP1-related disorder (STXBP1-RD) is a neurodevelopmental disorder caused by pathogenic variants in the STXBP1 gene. Its gene product MUNC18-1 organizes synaptic vesicle exocytosis and is essential for synaptic transmission. Patients present with developmental delay, intellectual disability, and/or epileptic seizures, with high clinical heterogeneity. To date, the cellular deficits of neurons of patients with STXBP1-RD are unknown. Methods: We combined live-cell imaging, electrophysiology, confocal microscopy, and mass spectrometry proteomics to characterize cellular phenotypes of induced pluripotent stem cell-derived neurons from 6 patients with STXBP1-RD, capturing shared features as well as phenotypic diversity among patients. Results: Neurons from all patients showed normal in vitro development, morphology, and synapse formation, but reduced MUNC18-1 RNA and protein levels. In addition, a proteome-wide screen identified dysregulation of proteins related to synapse function and RNA processes. Neuronal networks showed shared as well as patient-specific phenotypes in activity frequency, network irregularity, and synchronicity, especially when networks were challenged by increasing excitability. No shared effects were observed in synapse physiology of single neurons except for a few patient-specific phenotypes. Similarities between functional and proteome phenotypes suggested 2 patient clusters, not explained by gene variant type. Conclusions: Together, these data show that decreased MUNC18-1 levels, dysregulation of synaptic proteins, and altered network activity are shared cellular phenotypes of STXBP1-RD. The 2 patient clusters suggest distinctive pathobiology among subgroups of patients, providing a plausible explanation for the clinical heterogeneity. This phenotypic spectrum provides a framework for future validation studies and therapy design for STXBP1-RD.

Brain Region Differences in α1- and α5-Subunit-Containing GABAA Receptor Proteomes Revealed with Affinity Purification and Blue Native PAGE Proteomics.

GABAA receptors are the major inhibitory receptors in the brain. They are hetero-pentamers with a composition of predominantly two α, two ß, and one γ or δ subunit. Of the six α subunit genes, the α5 subunit displays a limited spatial expression pattern and is known to mediate both phasic and tonic inhibition. In this study, using immunoaffinity-based proteomics, we identified the α5 subunit containing receptor complexes in the hippocampus and olfactory bulb. The α1-α5 interaction was identified in both brain regions, albeit with significantly different stoichiometries. In line with this, reverse IPs using anti-α1 antibodies showed the α5-α1 co-occurrence and validated the quantitative difference. In addition, we showed that the association of Neuroligin 2 with α1-containing receptors was much higher in the olfactory bulb than in the hippocampus, which was confirmed using blue native gel electrophoresis and quantitative mass spectrometry. Finally, immunocytochemical staining revealed a co-localization of α1 and α5 subunits in the post-synaptic puncta in the hippocampus.

Suspension TRAPping Filter (sTRAP) Sample Preparation for Quantitative Proteomics in the Low µg Input Range Using a Plasmid DNA Micro-Spin Column: Analysis of the Hippocampus from the 5xFAD Alzheimer's Disease Mouse Model.

Suspension TRAPping filter (sTRAP) is an attractive sample preparation method for proteomics studies. The sTRAP protocol uses 5% SDS that maximizes protein solubilization. Proteins are trapped on a borosilicate glass membrane filter, where SDS is subsequently removed from the filter. After trypsin digestion, peptides are analyzed directly by LC-MS. Here, we demonstrated the use of a low-cost plasmid DNA micro-spin column for the sTRAP sample preparation of a dilution series of a synapse-enriched sample with a range of 10-0.3 µg. With 120 ng tryptic peptides loaded onto the Evosep LC system coupled to timsTOF Pro 2 mass spectrometer, we identified 5700 protein groups with 4% coefficient of variation (CoV). Comparing other sample preparation protocols, such as the in-gel digestion and the commercial Protifi S-TRAP with the plasmid DNA micro-spin column, the last is superior in both protein and peptide identification numbers and CoV. We applied sTRAP for the analysis of the hippocampal proteome from the 5xFAD mouse model of Alzheimer's disease and their wildtype littermates, and revealed 121 up- and 54 down-regulated proteins. Protein changes in the mutant mice point to the alteration of processes related to the immune system and Amyloid aggregation, which correlates well with the known major Alzheimer's-disease-related pathology. Data are available via ProteomeXchange with the identifier PXD041045.

Identifying true protein complex constituents in interaction proteomics: the example of the DMXL2 protein complex.

A typical high-sensitivity antibody affinity purification-mass spectrometry experiment easily identifies hundreds of protein interactors. However, most of these are non-valid resulting from multiple causes other than interaction with the bait protein. To discriminate true interactors from off-target recognition, we propose to differentially include an (peptide) antigen during the antibody incubation in the immuno-precipitation experiment. This contrasts the specific antibody-bait protein interactions, versus all other off-target protein interactions. To exemplify the power of the approach, we studied the DMXL2 interactome. From the initial six immuno-precipitations, we identified about 600 proteins. When filtering for interactors present in all anti-DMXL2 antibody immuno-precipitation experiments, absent in the bead controls, and competed off by the peptide antigen, this hit list is reduced to ten proteins, including known and novel interactors of DMXL2. Together, our approach enables the use of a wide range of available antibodies in large-scale protein interaction proteomics, while gaining specificity of the interactions.