Consolidation of metabolomic, proteomic, and GWAS data in connective model of schizophrenia

Abstract

Despite of multiple systematic studies of schizophrenia based on proteomics, metabolomics, and genome-wide significant loci, reconstruction of underlying mechanism is still a challenging task. Combination of the advanced data for quantitative proteomics, metabolomics, and genome-wide association study (GWAS) can enhance the current fundamental knowledge about molecular pathogenesis of schizophrenia. In this study, we utilized quantitative proteomic and metabolomic assay, and high throughput genotyping for the GWAS study. We identified 20 differently expressed proteins that were validated on an independent cohort of patients with schizophrenia, including ALS, A1AG1, PEDF, VTDB, CERU, APOB, APOH, FASN, GPX3, etc. and almost half of them are new for schizophrenia. The metabolomic survey revealed 18 group-specific compounds, most of which were the part of transformation of tyrosine and steroids with the prevalence to androgens (androsterone sulfate, thyroliberin, thyroxine, dihydrotestosterone, androstenedione, cholesterol sulfate, metanephrine, dopaquinone, etc.). The GWAS assay mostly failed to reveal significantly associated loci therefore 52 loci with the smoothened p < 10^-5 were fractionally integrated into proteome-metabolome data. We integrated three omics layers and powered them by the quantitative analysis to propose a map of molecular events associated with schizophrenia psychopathology. The resulting interplay between different molecular layers emphasizes a strict implication of lipids transport, oxidative stress, imbalance in steroidogenesis and associated impartments of thyroid hormones as key interconnected nodes essential for understanding of how the regulation of distinct metabolic axis is achieved and what happens in the conditioned proteome and metabolome to produce a schizophrenia-specific pattern.

Figure 1

Main anthropometric and psychometric data accrued for subjects with schizophrenia (n = 49). Complete records can be appreciated in the Supplementary Appendix A (Demography) and include the corresponding characteristics for the control group of healthy donors (n = 50) and for the validating group of patients with schizophrenia (n = 28). All groups under consideration were aligned by age and genders ratio items. Both, study population and validating group were leveled in the duration from prodrome and the duration from manifestation. The hereditary loading was slightly inclined in the study cohort (61%) cmpare to the validating group (46%; see Supplementary Appendix A) however this did not affect the final results and conclusions. Subjects of study cohorts were tested with several scales to assess the severity and progression of symptoms (Mann–Whitney test at a raw p < 0.05): PANSS Positive and Negative Syndrome Scale, BFCR scale Bush–Francis Catatonia Rating Scale, NCS4 the 4-Item Negative Symptom Assessment, SAS Simpson-Angus Scale, DSM-5 Diagnostic and Statistical Manual of mental disorders, fifth edition, FAB Frontal Assessment Battery.

Figure 2

Discriminant analysis of studied cohorts and scatter volcano plot of the most significantly altered proteins and metabolites. Sparce partial least-squares discriminant analysis (sPLS-DA) for proteome (A) and metabolome (B) data type with 0.95 ellipse confidence level. The designed score scattering plots show relationship between the control group and patients with schizophrenia, and the degree of variations that were explained by each component consisted of PC1 = 30% and PC2 = 6% for proteomic data (A) and of PC1 = 5% and PC2 = 2% for metabolomic data (B). A volcano plot for the most significant variables determined in proteome (C) and in metabolome (D) that enabled to discriminate the control group from patients with schizophrenia (Supplementary Appendix C, Appendix D). Only proteins and metabolites with scores exceeding FC = 2 (in linear scale) and p-value above 0.05 (Mann–Whitney U-test) were considered as significant and were engaged for the reconstruction of multilayer molecular events chain. All other analyses were performed with the in-house scripts written in R (version 3.2.0; R Foundation for Statistical Computing, Vienna, Austria; https://www.r-project.org/).

Figure 3

The proposed integrative scheme of interplay between proteome and metabolome layers in patients with schizophrenia. Observed elements (metabolites and proteins) are highlighted by yellow frame and FC (fold-changes) value; the FC observed but insignificant (after running the validation cohort of patients with schizophrenia) are designated by asterisk (*). Although it is almost impossible to figure out the very initial point, still there are several biological processes explicitly outlined on the proposed plane. It seems that the immune response and lipids metabolism might be one of the initial points affected in schizophrenia. The declination of APOA2, APOA4 and APOC1 and lipocalins APOE, APOM, APOH, CLU in combination with the increased oxidative-induced factors (GPX3, APOD, SELENOP) might be a clue. It is worth noticed that the consolidation of apolipoproteins with action of FASN, GPX3 and PEDF in steroidogenesis produces significant influence on the neuroimmune HPA (hypothalamic–pituitary–adrenal) axis. By this reason, some steroids, including DHEA and precursors of 17-ketosteroids, and the balance between androgen and estrogen (sex hormones) pathways bears a special significance. The concept of DHEA role was considered as important in pathogenesis of schizophrenia even before due to pro-oxidative and defending properties initiated through affecting on PPARα/γ1/γ2 receptors and, thus, regulation of apolipoproteins family genes expression. The connecting point is enhanced by the expression of ROS-sensitive APOD that triggers FASN and, consequently, enhance ROS generation. It entails to initiation of complement cascade as a response on the local inflammatory, where some complement factors (C5, C3, and C2 complement factors) enhance cAMP production through the Ca²⁺-dependent signaling. In turn, it boosts excessive NO production which readily interacts with ROS species and aggravates immune response closing the loop with cAMP. This, in sense, may found a response in the action of PTGIS prostaglandin. The progenitor guiding gene PTHG was consistently found in the studied subjects as associated loci and it is perfectly match with the hypothesized escalation of inflammation/immune response and steroidogenesis stressed through cAMP and Ca²⁺-dependent signaling. Another sort of neuroimmune regulation undergoes through the axis of prolactin and its surrounding. Here, the comprehensive and complementary interplay between prolactin and 17β-estradiol, and opposed impact on adiponectin and dopamine are crucial. Apart direct regulation of steroidogenesis, it impacts on the transformations of tyrosine including its transformation products dopamine, dopaquinone, and catecholamines, which has been observed imbalanced. As an obvious consequence, TRH and thyroid hormones tend to decrease which is agreed with the obtained data and previously reported observations about hypothyroidism in the affected patients.

Figure 4

Circular plot for the Multi Omics connections between the serum-based proteome, metabolome and SNPs recognized in patients with schizophrenia. The external environment of the graph shows the most affected biological pathways where the identified meaningful molecular factors (Table 3) are involved in. The graph size was equal to 554 with assortative value of 0.346. Connections (colored lines, or edges) entering and outputting in/from different pathways (nodes, or vertices) illustrate a track caused by common elements between molecular layers (proteome—red line; metabolome—green line; gene with SNP—blue color) that has been defined in the study (Table 3) and consolidated with the current knowledge about pathophysiological mechanism of schizophrenia. If entity (node) has no connection (edge) with other node, there is no element (protein, metabolite, or gene with SNT), which is shared between two or more distinct pathways, so that element is specifically attributed to certain pathway and makes a loop on itself. The more elements (proteins, metabolites, genes with SNP) are congregated and focused in certain pathways (node), the larger size of centrality can be expected. The number of nodes with the centrality more than 10 (i.e., the number of edges attached to the node) was equal to 30, while the centrality more than 20 refers to 5 hubs. The centrality reflects the level of occurrence of a biological process. In turn, the occurrence has been was estimated through the frequency of identification of the corresponding molecular factors in the studied samples. The larger centrality indicates the higher frequency of the molecular factor identification.