MSACL 2016 EU Abstract

A Multi-Platform Mass Spectrometric Approach for Investigating Schizophrenia

Elizabeth Want (Presenter)
Imperial College, London

Bio: Elizabeth Want is a Senior Lecturer in Molecular Spectroscopy in the Department of Surgery and Cancer at Imperial College, London and the Director of the Imperial International Phenome Training Centre. She joined Imperial College in 2006 after working as a postdoctoral researcher at the Scripps Research Institute in La Jolla, CA. Her research at Imperial College involves the development, optimisation and application of LC-MS methodologies for the analysis of biological samples, largely in the context of metabolic phenotyping. She applies these methods to biomedical research areas including toxicology, cardiovascular disease, neonatal disease and development, and neurological diseases.

Authorship: Elizabeth J. Want,(1) Hendrik Wesseling,(2) Elaine Holmes,(1) and Sabine Bahn (2,3)
1)Section of Biomolecular Medicine, CSM, Department of Surgery and Cancer, Faculty of Medicine, Imperial College, London. 2)Department of Chemical Engineering and Biotechnology, University of Cambridge, Cambridge. 3)Department of Neuroscience, Erasmus Medical Center

Short Abstract

Schizophrenia affects 1% of the world’s population. Heterogeneous symptoms hinder current therapies; new models are needed to elucidate mechanisms and identify biomarkers. Metabolic profiling (UPLC-MS) and proteomics on serum/brain from a chronic phencyclidine (PCP) rat model showed a larger effect on the hippocampus than the frontal cortex. Alterations in cytokines, fibroblast growth factor-2, lipid metabolism and superoxide dismutase levels were observed. Abnormalities in NMDA-receptor associated pathway, kainate, AMPA and GABAergic signalling correlated with behavioural functions. Findings could lead to increased understanding of perturbed glutamate receptor signalling in schizophrenia, potential novel drug targets, and translation of surrogate blood markers to the clinic.

Long Abstract

Background

The neuropsychiatric disorder schizophrenia affects 1% of the world’s population. It manifests with a broad, heterogeneous range of symptoms, hindering the effectiveness of current therapies. Currently there is limited understanding of the molecular pathology underlying schizophrenia, and few well-characterized animal models. Thus, new models capable of reproducing core pathological features of schizophrenia are needed to elucidate pathological disease mechanisms, identify biomarkers for improved diagnosis and discover potential novel drug targets.

Metabolic profiling using ultra performance liquid chromatography-mass spectrometry (UPLC-MS) was combined with comprehensive LC-MS based global label-free and targeted proteomics (selected reaction monitoring and multiplex-immunoassay) on serum and brain tissues from a chronic phencyclidine (PCP) rat model. Here, NMDA-receptor hypofunction is induced through non-competitive NMDAR-receptor antagonism. This multi-platform approach can increase confidence in model validity at the molecular level and aid drug discovery studies. To date, this is the largest study of the PCP rat model using a combination of ‘omics technologies to analyse distinct brain regions implicated in schizophrenia.

Methods

Metabolic profiling. Brain tissue samples were extracted using optimised in-house protocols. A 10 μL aliquot from each study sample was combined to produce a quality control (QC) sample. UPLC−MS analysis was performed using a Waters XEVO G2 Q-TOF mass spectrometer coupled to an Acquity UPLC−MS system (Waters Corporation, Milford, MA, USA). Separation was performed on a 2.1 × 100 mm (1.7 μm) HSS T3 Acquity column with a linear gradient of water:methanol employed over 28 min. QC samples were injected ten times at the start of the run to condition the column and after every ten samples to assess instrument stability. Data were processed using the freeware XCMS with standard parameters. Metabolite feature tables (m/z, retention time, intensity values) were imported into SIMCA-P (Umetrics) for multivariate analysis.

Label-Free LC−MSE Analysis. Digested brain tissue samples were analyzed using a splitless nanoACQUITY for reversed-phase chromatographic peptide separation, coupled to a Waters Q-TOF Premier mass spectrometer (Waters Corporation; Milford, MA, USA). Fragment ions were matched to corresponding precursor peptide ions using retention time, mass accuracy, and other physicochemical properties. Data were processed with ProteinLynx Global Server v.2.4 (Waters Corporation; Milford, MA, USA) and Rosetta Elucidator v.3.3 (Rosetta Biosoftware; Seattle, WA, USA). Aligned peaks were extracted and abundance measurements obtained by integrating retention time, m/z, and intensity values, with normalization to total ion current. PLGS2.4 using the Swiss-Prot rodent reference proteome (Uniprot, March 2013 release) was used for protein identification searches.

Label-Based Selected Reaction Monitoring Mass Spectrometry. Digested frontal cortex and hippocampus proteomes were analyzed using targeted label-based selected reaction monitoring (SRM) mass spectrometry on a Xevo TQ-S mass spectrometer coupled to a nanoAcquity UPLC system (Waters Corporation; Milford, MA, USA). Multiplex SRM assays were developed using a high throughput strategy. Resulting SRM data were analysed using the freeware Skyline. Differential abundance of analytes between cPCP-treated rats and control animals was calculated using the freeware MSstats.

Results

To elucidate cPCP treatment effects in the peripheral circulation and identify potential surrogate biomarkers for schizophrenia, serum levels of 64 analytes were measured. Five analytes were significantly altered in cPCP rat serum (p < 0.05); (IL-5, IL-2, IL-1β), fibroblast growth factor-2 (FGF-2) and macrophage inflammatory protein 1a (MIP-1α).

Using label-free LC−MSE, 555 proteins were identified in the frontal cortex and 937 proteins in the hippocampus. Of these, 79 frontal cortex proteins (14%) and 501 hippocampus proteins (53%) were significantly changed due to cPCP treatment. Protein level alterations of 22 enzymes were detected in the frontal cortex, of which 10 (45%) catalyse a metabolic reaction, and 139 enzymes in the hippocampus, of which 94 (68%) catalyze a metabolic reaction.

Consistent with proteomics, global metabolic profiling of brain tissue showed a greater effect of PCP in the hippocampus. 1057 metabolite features were identified after filtering using relative standard deviation across both models and brain regions. No significantly changed features were detected in the frontal cortex of the cPCP rat model, while 426 features were significantly changed in the hippocampus. The top 10 significant hits were selected and metabolite identification performed using the HMDB and Pubchem databases.

Combined with quantitative mass spectrometry, pathway analysis can help identify functional links or the causality of complex physiological crosstalk in an in vivo context. Ingenuity Pathways Analysis (IPA) using the total of all changed proteins in the frontal cortex and hippocampus (regardless of the magnitude of change), identified a decrease in neurodevelopment associated biological functions in the frontal cortex. The hippocampus was associated with a decreased activation of the biological processes, plasticity of synapse, exocytosis of vesicles, behaviour, spatial memory, and increased activation of movement of rodents, paralysis and conditioning. This matches the reported behavioural readouts associated with the cPCP animal model in the literature. GO-enrichment analysis of the proteomic changes revealed the most robust enriched biological functions across both brain regions were associated with small GTPases and Rho signalling proteins.

Conclusions

Brain tissue profiling identified changes in a wide range of proteins induced by cPCP treatment. Individually, protein changes were subtle, indicating homeostatic disequilibrium. Chronic PCP treatment had a larger effect on the hippocampal proteome and metabonome compared to the frontal cortex. A trend toward an anti-inflammatory state was observed, with alterations in cytokine levels (IL-5, IL-2, IL-1β) and fibroblast growth factor-2. Metabolic profiling revealed changes in lipid metabolism, particularly glycerophospholipids, further supported through altered superoxide dismutase levels, indicative of oxidative stress and apoptotic pathway alterations. Bioinformatic pathway analysis confirmed abnormalities in NMDA-receptor associated pathways in both brain regions, as well as alterations in kainate, AMPA and GABAergic signalling in the hippocampus. These findings were also correlated with hippocampal behavioural functions.

These molecular changes parallel findings observed in humans, where changes in hippocampal function have been linked to schizophrenia, opening up new avenues of research, as previous studies focused on potential frontal cortex abnormalities. This study could lead to increased understanding of how perturbed glutamate receptor signalling affects other relevant biological pathways in schizophrenia. This may lead to the discovery of potential novel drug targets for improved treatment, while surrogate markers in blood can be translated to the clinic.


References & Acknowledgements:


Financial Disclosure

DescriptionY/NSource
GrantsyesWaters Corporation
Salaryno
Board MemberyesMSACL EU Scientific Committee
Stockno
Expensesno

IP Royalty: no

Planning to mention or discuss specific products or technology of the company(ies) listed above:

no