Proteomic profile in postmortem brain in chronic schizopfrenia

Abstract

Schizophrenia is a complex psychiatric disorder involving dysregulation of multiple pathways. In the last decades, the cortico-cerebellar-thalamo-cortical circuit has been proposed to play a key role in cognitive impairments in schizophrenia. The cerebellum is a brain area that forms part of this circuit that modulates synaptic responses of cortical regions, it has been proposed to play an important role in schizophrenia pathophysiology. Furthermore, the cerebellum, in connection with the dorsolateral prefrontal cortex, is involved in executive and working memory function. Thus, we hypothesized that the altered proteomic profile in the brain participates in the molecular network dysfunction in chronic schizophrenia. In this doctoral thesis we aimed to compare the proteomic profile in the postmortem cerebellar cortex and the prefrontal cortex of individuals with chronic schizophrenia by using (i) mass spectrometry, (ii) bioinformatic analyses to identify altered molecular networks, (iii) two double-hit schizophrenia murine models induced by maternal deprivation combined with an additional stressor for the hit proteins from the altered network in the cerebellum, (iv) immunohistochemistry techniques and (v) 3D projection analysis for unexplored candidate hits in human cerebellum. First, we performed a pilot proteomic analysis on postmortem human cerebellar tissue from patients with schizophrenia (n=4) and control (n=4) subjects in a pool design using differential isotope peptide labelling followed by liquid chromatography tandem mass spectrometry (LC-MS/MS). Our results showed 1412 quantified proteins and 99 significantly altered proteins. 11 candidate proteins were selected as the most robust candidates from the enriched biological functions comprising cell communication/signal transduction in schizophrenia. In this study we report 68 new proteins altered in schizophrenia. In our individual proteomic study of the cerebellum, we analysed 12 cerebellar samples of subjects with schizophrenia and 14 healthy control individuals using one-shot liquid- tandem mass spectrometry. We identified a proteomic signature composed of 2578 identified proteins and 1474 quantified proteins with 250 significantly dysregulated proteins. This study reported 167 new proteins that are altered in the cerebellum in schizophrenia. Moreover, our study allowed us to identify the proteome profile and molecular networks altered in the cerebellum in chronic schizophrenia. Hierarchical clustering allowed schizophrenia subjects to be segregated from controls. Furthermore, our analyses showed that the 250 altered proteins could be under the transcriptional control of only11 transcription factors. The pathways regulated by these 11 transcription factors were related to transport, signalling, inflammation, and apoptosis.

Our network generation analysis in the cerebellum showed two well-defined networks. The network generated from the up-regulated proteins showed a mixed module with enriched pathway interactions. This module consisted of proteins from vesicle-mediated transport and axon guidance pathways. In the axon guidance pathway, CLASP1 was a hit protein that we further studied in two independent double-hit murine models for schizophrenia. Our results showed CLASP1 to be down-regulated in the murine model of maternal deprivation combined with social isolation, while our proteomic study showed it to be up- regulated. The second network was formed by two modules generated from proteins that were found to be down-regulated in the cerebellum: an energy module and a neutrophil degranulation module. The energy module was well defined in our analysis, with NDUFB9 found to be a down-regulated hit protein in our proteomic study and in the double-hit murine model of maternal deprivation combined with social isolation. The second module was neutrophil degranulation. METTL7A was found to be a hit protein in this pathway. In our study, we found . and METTL7A was found to be consistently down-regulated both in schizophrenia subjects and murine models for schizophrenia. We demonstrated for the first time the expression of METTL7A in Bergmann glia cells in the human cerebellum. We also detected METTL7A in one of its canonical localizations, in lipid droplets in white matter cells. The prefrontal cortex analysis revealed 4407 identified proteins, 1989 quantified proteins and 43 significantly altered proteins. The enriched pathways were mainly related to the immune system. The network generated from the enriched pathway showed a mixed module with interactions between MHC class II antigen presentation, membrane trafficking, Golgi-to-ER retrograde transport, Nef-mediated CD8 down-regulation and the immune system.

Together, the results presented in this Thesis suggest an imbalance in the immune system in two brain regions, the cerebellum and the prefrontal cortex. This suggests that the nervous system could be susceptible to an imbalance in the immune system in schizophrenia subjects. Moreover, the vesicle-mediated transport pathway was found to be altered in both the cerebellum and the prefrontal cortex, two brain areas that participate in the cortico-cerebellar thalamo-cortical circuit. This result suggests a possible alteration in synaptic efficacy and communication between these areas in schizophrenia.