Analysis of blood-based gene expression signature in first-episode psychosis

An Investigation of How Normalisation and Local Modelling Techniques Confound Machine Learning Performance In a Mental Health Study

Machine learning (ML) is increasingly deployed on biomedical studies for biomarker development (feature selection) and diagnostic/prognostic technologies (classification). While different ML techniques produce different feature sets and classification performances, less understood is how upstream data processing methods (e.g., normalisation) impact downstream analyses. Using a clinical mental health dataset, we investigated the impact of different normalisation techniques on classification model performance. Gene Fuzzy Scoring (GFS), an in-house developed normalisation technique, is compared against widely used normalisation methods such as global quantile normalisation, class-specific quantile normalisation and surrogate variable analysis. We report that choice of normalisation technique has strong influence on feature selection. with GFS outperforming other techniques. Although GFS parameters are tuneable, good classification model performance (ROC-AUC > 0.90) is observed regardless of the GFS parameter settings. We also contrasted our results against local modelling, which is meant to improve the resolution and meaningfulness of classification models built on heterogeneous data. Local models, when derived from non-biologically meaningful subpopulations, perform worse than global models. A deep dive however, revealed that the factors driving cluster formation has little to do with the phenotype-of-interest. This finding is critical, as local models are often seen as a superior means of clinical data modelling. We advise against such naivete. Additionally, we have developed a combinatorial reasoning approach using both global and local paradigms: This helped reveal potential data quality issues or underlying factors causing data heterogeneity that are often overlooked. It also assists to explain the model as well as provides directions for further improvement.

What genes are differentially expressed in individuals with schizophrenia? A systematic review

Schizophrenia is a severe, complex mental disorder characterized by a combination of positive symptoms, negative symptoms, and impaired cognitive function. Schizophrenia is highly heritable (~80%) with multifactorial etiology and complex polygenic genetic architecture. Despite the large number of genetic variants associated with schizophrenia, few causal variants have been established. Gaining insight into the mechanistic influences of these genetic variants may facilitate our ability to apply these findings to prevention and treatment. Though there have been more than 300 studies of gene expression in schizophrenia over the past 15 years, none of the studies have yielded consistent evidence for specific genes that contribute to schizophrenia risk. The aim of this work is to conduct a systematic review and synthesis of case-control studies of genome-wide gene expression in schizophrenia. Comprehensive literature searches were completed in PubMed, EmBase, and Web of Science, and after a systematic review of the studies, data were extracted from those that met the following inclusion criteria: human case-control studies comparing the genome-wide transcriptome of individuals diagnosed with schizophrenia to healthy controls published between January 1, 2000 and June 30, 2020 in the English language. Genes differentially expressed in cases were extracted from these studies, and overlapping genes were compared to previous research findings from the genome-wide association, structural variation, and tissue-expression studies. The transcriptome-wide analysis identified different genes than those previously reported in genome-wide association, exome sequencing, and structural variation studies of schizophrenia. Only one gene, GBP2, was replicated in five studies. Previous work has shown that this gene may play a role in immune function in the etiology of schizophrenia, which in turn could have implications for risk profiling, prevention, and treatment. This review highlights the methodological inconsistencies that impede valid meta-analyses and synthesis across studies. Standardization of the use of covariates, gene nomenclature, and methods for reporting results could enhance our understanding of the potential mechanisms through which genes exert their influence on the etiology of schizophrenia. Although these results are promising, collaborative efforts with harmonization of methodology will facilitate the identification of the role of genes underlying schizophrenia.

Profile of the Nicotinic Cholinergic Receptor Alpha 7 Subunit Gene Expression is Associated with Response to Varenicline Treatment

INTRODUCTION

Smoking is considered the leading cause of preventable morbidity and mortality worldwide. Studies have sought to identify predictors of response to smoking cessation treatments. The aim of this study was to analyze a possible association of target gene expression for smoking cessation with varenicline.

METHODS

We included 74 smokers starting treatment with varenicline. Gene expression analysis was performed through the custom RT² Profiler qPCR array assay, including 17 genes. Times for sample collection were before the start of therapy (T0) and two weeks (T2) and four weeks (T4) after the start of treatment.

RESULTS

For gene expression analysis, we selected 14 patients who had success and 13 patients resistant to varenicline treatment. Success was considered to be when a patient achieved tobacco abstinence until the fourth week of treatment and resistant was when a patient had not stopped smoking as of the fourth week of treatment. We observed a significant difference for CHRNA7 gene expression: in the resistant group, samples from T2 and T4 had lower expression compared with T0 (fold change: 0.38, P = 0.007; fold change: 0.67, P = 0.004; respectively).

CONCLUSION

This exploratory clinical study, searching for a possible predictor of effectiveness for varenicline, reaffirmed the association of the α7 nAChR subunit for nicotine dependence and smoking therapy effectiveness with varenicline.

A multimodal attempt to follow-up linkage regions using RNA expression, SNPs and CpG methylation in schizophrenia and bipolar disorder kindreds

The complexity of schizophrenia (SZ) and bipolar disorder (BD) has slowed down progress in understanding their genetic roots. Alternative genomic approaches are needed to bypass these difficulties. We attempted a multimodal approach to follow-up on reported linkage findings in SZ and BD from the Eastern Quebec kindreds in chromosomes 3q21, 4p34, 6p22, 8p21, 8p11, 13q11-q14, 15q13, 16p12, and 18q21. First, in 498 subjects, we measured RNA expression (47 K Illumina chips) in SZ and BD patients that we compared with their non-affected relatives (NARs) to identify, for each chromosomal region, genes showing the most significant differences in expression. Second, we performed SNP genotyping (700 K Illumina chips) and cis-eQTN analysis. Third, we measured DNA methylation on genes with RNA expression differences or eQTNs. We found a significant overexpression of the gene ITGB5 at 3q25 in SZ and BD after multiple testing p value adjustment. SPCS3 gene at 4q34, and FZD3 gene at 8p21, contained significant eQTNs after multiple testing corrections, while ITGB5 provided suggestive results. Methylation in associated genes did not explain the expression differences between patients and NARs. Our multimodal approach involving RNA expression, dense SNP genotyping and eQTN analyses, restricted to chromosomal regions having shown linkage, lowered the multiple testing burden and allowed for a deeper examination of candidate genes in SZ or BD.

Differential gene expression analysis in blood of first episode psychosis patients

BACKGROUND

Psychosis is a condition influenced by an interaction of environmental and genetic factors. Gene expression studies can capture these interactions; however, studies are usually performed in patients who are in remission. This study uses blood of first episode psychosis patients, in order to characterise deregulated pathways associated with psychosis symptom dimensions.

METHODS

Peripheral blood from 149 healthy controls and 131 first episode psychosis patients was profiled using Illumina HT-12 microarrays. A case/control differential expression analysis was performed, followed by correlation of gene expression with positive and negative syndrome scale (PANSS) scores. Enrichment analyses were performed on the associated gene lists. We test for pathway differences between first episode psychosis patients who qualify for a Schizophrenia diagnosis against those who do not.

RESULTS

A total of 978 genes were differentially expressed and enriched for pathways associated to immune function and the mitochondria. Using PANSS scores we found that positive symptom severity was correlated with immune function, while negative symptoms correlated with mitochondrial pathways.

CONCLUSIONS

Our results identified gene expression changes correlated with symptom severity and showed that key pathways are modulated by positive and negative symptom dimensions.

The Effects of APOE4 on Mitochondrial Dynamics and Proteins in vivo

This study examined the effects of apolipoprotein E4 (APOE4), the most prevalent genetic risk factor for Alzheimer's disease (AD), on proteins involved in mitochondrial dynamics and autophagy, in the hippocampus of targeted replacement mice. Immunohistochemical measurements revealed that the levels of the mitochondrial fusion-mediating protein, MFN1, were higher, whereas those of corresponding fission-regulating protein, DRP-1, were lower in the hippocampus of ApoE4 mice than in the corresponding ApoE3 mice, indicating that APOE4 is associated with increased mitochondrial fusion and decreased fission. A similar ApoE4-driven decrease in DRP-1 was also observed in AD brains. The levels of the mitochondrial proteins COX1 and Tom40, were higher in the ApoE4 mice, which is consistent with the increased fusion. Measurements of the levels of cleaved PINK1 and parkin, which mark and target mitochondria for mitophagic degradation, revealed lower levels of cleaved PINK1, suggesting reduced mitochondrial membrane potential, and higher levels of parkin in the hippocampus of ApoE4 compared with the ApoE3 mice, indicating altered mitophagy. The levels of the ubiquitin-binding scaffold protein, p62/SQSTM1, which directs selected cargo to the autophagosomes, were also higher in the ApoE4 mice. These findings suggest that APOE4 is associated with enhanced mitochondrial fusion and decreased fission. Additionally, the results indicate that mitophagy/autophagy is reduced in ApoE4 mice, resulting in higher levels of proteins such as parkin and p62, which are normally degraded during this process. Taken together, these results suggest a novel mechanism that may underlie the pathological effects of APOE4 and indicate that use of APOE4 genotyping could pave the way for identification of novel APOE4-related therapeutic targets.

Longitudinal Analyses of Blood Transcriptome During Conversion to Psychosis

The biological processes associated with the onset of schizophrenia remain largely unknown. Current hypotheses favor gene × environment interactions as supported by our recent report about DNA methylation changes during the onset of psychosis. Here, we conducted the first longitudinal transcriptomic analysis of blood samples from 31 at-risk individuals who later converted to psychosis and 63 at-risk individuals who did not. Individuals were followed for a maximum of 1 year. Blood samples were collected at baseline and at the end of follow-up and individuals served as their own controls. Differentially expressed genes between the 2 groups were identified using the RNA sequencing of an initial discovery subgroup (n = 15 individuals). The most promising results were replicated using high-throughput real-time qPCR in the whole cohort (n = 94 individuals). We identified longitudinal changes in 4 brain-expressed genes based on RNAseq analysis. One of these genes (CPT1A) was replicated in the whole cohort. The previously observed hypermethylation in NRP1 and GSTM5 during the onset of psychosis correlated with a decrease in corresponding gene expression. RNA sequencing also identified 2 co-expression networks that were impaired after conversion compared with baseline-the Wnt pathway including AKT1, CPT1A and semaphorins, and the Toll-like receptor pathway, related to innate immunity. This longitudinal study of transcriptomic changes in individuals with at-risk mental state revealed alterations during conversion to psychosis in pathways and genes relevant to schizophrenia. These results may be a first step toward better understanding psychosis onset. They may also help to identify new biomarkers and targets for disease-modifying therapeutic strategies.

Can Peripheral Blood-Derived Gene Expressions Characterize Individuals at Ultra-high Risk for Psychosis?

The ultra-high risk (UHR) state was originally conceived to identify individuals at imminent risk of developing psychosis. Although recent studies have suggested that most individuals designated UHR do not, they constitute a distinctive group, exhibiting cognitive and functional impairments alongside multiple psychiatric morbidities. UHR characterization using molecular markers may improve understanding, provide novel insight into pathophysiology, and perhaps improve psychosis prediction reliability. Whole-blood gene expressions from 56 UHR subjects and 28 healthy controls are checked for existence of a consistent gene expression profile (signature) underlying UHR, across a variety of normalization and heterogeneity-removal techniques, including simple log-conversion, quantile normalization, gene fuzzy scoring (GFS), and surrogate variable analysis. During functional analysis, consistent and reproducible identification of important genes depends largely on how data are normalized. Normalization techniques that address sample heterogeneity are superior. The best performer, the unsupervised GFS, produced a strong and concise 12-gene signature, enriched for psychosis-associated genes. Importantly, when applied on random subsets of data, classifiers built with GFS are "meaningful" in the sense that the classifier models built using genes selected after other forms of normalization do not outperform random ones, but GFS-derived classifiers do. Data normalization can present highly disparate interpretations on biological data. Comparative analysis has shown that GFS is efficient at preserving signals while eliminating noise. Using this, we demonstrate confidently that the UHR designation is well correlated with a distinct blood-based gene signature.

Epigenetics and gene expression profile in first-episode psychosis: The role of childhood trauma

Childhood Trauma (CT) mediation of the epigenome and its impact on gene expression profile could provide a mechanism for the gene-environment interaction underling psychosis. We reviewed the evidence concerning epigenetic and gene expression modifications associated with CT in both First-Episode Psychosis (FEP) and healthy subjects. In order to explore the relative role of psychosis itself in determining these modifications, evidence about FEP and epigenetics/gene expression was also summarized. We performed a systematic search on PubMed, last updated in December 2016. Out of 2966 potentially relevant records, only 41 studies were included. CT resulted associated: in FEP subjects, with global DNA hypo-methylation and reduced BDNF gene-expression; in healthy subjects, with hyper-methylation of SLC6A4, NR3C1, KITLG, and OXTR; hypo-methylation of FKBP5, IL-6, and BDNF; increased IL1B, IL8, and PTGS gene expression; and decreased SLC6A4 gene expression. FEP showed global DNA hypo-methylation; increased methylation and reduced gene expression of GCH1; hyper-expression of MPB, NDEL1, AKT1, and DICER1; and hypo-expression of DROSHA, COMT, and DISC1 in comparison with healthy controls.

Peripheral blood gene expression of acute phase proteins in people with first episode psychosis

BACKGROUND

There is a growing interest in the association between schizophrenia and the activation of inflammatory system with signs of acute phase (AP) response. Majority of such studies had focused on C-reactive protein (CRP). The aims of the present study were (i) to examine the gene expression profiles of other acute phase proteins (APP), namely haptoglobin (HP), alpha-1 antitrypsin (A1T), and alpha-2 macroglobulin (A2M) in patients with first episode psychosis (FEP) over a period of three months and (ii) to explore the association between APP levels and severity of symptoms.

METHODS

In this study, HP, A1T and A2M gene expression levels from whole blood were measured at recruitment, 1- and 3-month follow-up visits using quantitative PCR (qPCR) in 43 patients with FEP and in 57 healthy controls. Diagnoses was ascertained on the Structured Clinical Interview for DSM-IV-TR. Severity of symptoms in patients was assessed on the Positive and Negative Syndrome Scale (PANSS) and a previously validated 5-factor PANSS structure was applied in the subsequent analyses.

RESULTS

The FEP sample comprised of 28 (65.1%) individuals with schizophrenia, 12 (27.9%) with schizophreniform disorder and 3 (7%) with schizoaffective disorder. HP gene expression level was noted to be significantly higher in patients than controls at all three time points: recruitment (P=0.049), 1-month follow up (P=0.002) and 3-month follow up (P=0.005). PANSS positive, depression, and excitement symptom factors showed significant associations with HP (P=0.002), A1T (P=0.016) and A2M (P=0.034), respectively. These findings remained significant after controlling for age, gender, smoking status and accumulated chlorpromazine dosage.

CONCLUSION

The current study provides information on HP, A1T and A2M gene expression profiles in FEP patients and their associations with psychopathology. This provides support for the hypothesis that inflammation is related to schizophrenia and further encourages studies on immune-inflammatory markers to understand the relationship between inflammation and schizophrenia.

Predicting Autism Spectrum Disorder Using Blood-based Gene Expression Signatures and Machine Learning

Objective

The aim of this study was to identify a transcriptomic signature that could be used to classify subjects with autism spectrum disorder (ASD) compared to controls on the basis of blood gene expression profiles. The gene expression profiles could ultimately be used as diagnostic biomarkers for ASD.

Methods

We used the published microarray data (GSE26415) from the Gene Expression Omnibus database, which included 21 young adults with ASD and 21 age- and sex-matched unaffected controls. Nineteen differentially expressed probes were identified from a training dataset (n=26, 13 ASD cases and 13 controls) using the limma package in R language (adjusted p value <0.05) and were further analyzed in a test dataset (n=16, 8 ASD cases and 8 controls) using machine learning algorithms.

Results

Hierarchical cluster analysis showed that subjects with ASD were relatively well-discriminated from controls. Based on the support vector machine and K-nearest neighbors analysis, validation of 19-DE probes with a test dataset resulted in an overall class prediction accuracy of 93.8% as well as a sensitivity and specificity of 100% and 87.5%, respectively.

Conclusion

The results of our exploratory study suggest that the gene expression profiles identified from the peripheral blood samples of young adults with ASD can be used to identify a biological signature for ASD. Further study using a larger cohort and more homogeneous datasets is required to improve the diagnostic accuracy.

GAD1 Gene Expression in Blood of Patients with First-Episode Psychosis

γ-Aminobutyric acid (GABA), the primary inhibitory neurotransmitter, has often been studied in relation to its role in the pathophysiology of schizophrenia. GABA is synthesized from glutamate by glutamic acid decarboxylase (GAD), derived from two genes, GAD1 and GAD2. GAD1 is expressed as both GAD67 and GAD25 mRNA transcripts with the former reported to have a lower expression level in schizophrenia compared to healthy controls and latter was reported to be predominantly expressed fetally, suggesting a role in developmental process. In this study, GAD67 and GAD25 mRNA levels were measured by quantitative PCR (qPCR) in peripheral blood of subjects with first-episode psychosis (FEP) and from healthy controls. We observed low GAD25 and GAD67 gene expression levels in human peripheral blood. There was no difference in GAD25 and GAD67 gene expression level, and GAD25/GAD67 ratio between patients with FEP and healthy controls. PANSS negative symptoms were associated with levels of GAD25 mRNA transcripts in patients with FEP. While the current study provides information on GAD25 and GAD67 mRNA transcript levels in whole blood of FEP patients, further correlation and validation work between brain regions, cerebrospinal fluid and peripheral blood expression profiling are required to provide a better understanding of GAD25 and GAD67.

Candidate gene networks and blood biomarkers of methamphetamine-associated psychosis: an integrative RNA-sequencing report

The clinical presentation, course and treatment of methamphetamine (METH)-associated psychosis (MAP) are similar to that observed in schizophrenia (SCZ) and subsequently MAP has been hypothesized as a pharmacological and environmental model of SCZ. However, several challenges currently exist in diagnosing MAP accurately at the molecular and neurocognitive level before the MAP model can contribute to the discovery of SCZ biomarkers. We directly assessed subcortical brain structural volumes and clinical parameters of MAP within the framework of an integrative genome-wide RNA-Seq blood transcriptome analysis of subjects diagnosed with MAP (N=10), METH dependency without psychosis (MA; N=10) and healthy controls (N=10). First, we identified discrete groups of co-expressed genes (that is, modules) and tested them for functional annotation and phenotypic relationships to brain structure volumes, life events and psychometric measurements. We discovered one MAP-associated module involved in ubiquitin-mediated proteolysis downregulation, enriched with 61 genes previously found implicated in psychosis and SCZ across independent blood and post-mortem brain studies using convergent functional genomic (CFG) evidence. This module demonstrated significant relationships with brain structure volumes including the anterior corpus callosum (CC) and the nucleus accumbens. Furthermore, a second MAP and psychoticism-associated module involved in circadian clock upregulation was also enriched with 39 CFG genes, further associated with the CC. Subsequently, a machine-learning analysis of differentially expressed genes identified single blood-based biomarkers able to differentiate controls from methamphetamine dependents with 87% accuracy and MAP from MA subjects with 95% accuracy. CFG evidence validated a significant proportion of these putative MAP biomarkers in independent studies including CLN3, FBP1, TBC1D2 and ZNF821 (RNA degradation), ELK3 and SINA3 (circadian clock) and PIGF and UHMK1 (ubiquitin-mediated proteolysis). Finally, focusing analysis on brain structure volumes revealed significantly lower bilateral hippocampal volumes in MAP subjects. Overall, these results suggest similar molecular and neurocognitive mechanisms underlying the pathophysiology of psychosis and SCZ regardless of substance abuse and provide preliminary evidence supporting the MAP paradigm as an exemplar for SCZ biomarker discovery.

Changes in gene expression and methylation in the blood of patients with first-episode psychosis

Schizophrenia is a severe mental health disorder with high heritability. The investigation of individuals during their first-episode psychosis (FEP), before the progression of psychotic disorders and especially before treatment with antipsychotic medications, is particularly helpful for understanding this complex disease and for the identification of potential biomarkers. In this study, we compared the expression of genes that are involved in neurotransmission and neurodevelopment of antipsychotic-naive FEP in the peripheral blood of patients (n=51) and healthy controls (n=51). In addition, we investigated the differentially expressed genes with respect to a) DNA methylation, b) the correlation between gene expression and clinical variables (PANSS), and c) gene expression changes after risperidone treatment. Expression levels of 11 genes were quantified with SYBR Green. For methylation analysis, bisulfite sequencing was performed. A significant decrease in GCH1 mRNA levels was observed in FEP patients relative to controls. Also, when we compare the FEP patients after risperidone treatment with controls, this difference remains significant, and no significant differences were observed in GCH1 mRNA levels when comparing patients before and after risperidone treatment. Additionally, although the differences were non-significant after Bonferroni correction, the expression of GCH1 seemed to be correlated with PANSS scores, and the GCH1 promoter region was more methylated in FEP than in controls, thus corroborating the results obtained at the mRNA level. Few studies have been conducted on GCH1, and future studies are needed to clarify its potential role in the progression of schizophrenia.

Neurotransmitter receptor and regulatory gene expression in peripheral blood of Brazilian drug-naïve first-episode psychosis patients before and after antipsychotic treatment

Little is known about how genes expressed in blood relate to schizophrenia or antipsychotic use. We analyzed gene expression in 10 first-episode psychosis patients and nine controls using PCR Arrays. GABRR2 and CHRNA3 were found to be differentially expressed after risperidone treatment. These genes may be regulated by antipsychotic use.

Blood-based gene-expression predictors of PTSD risk and resilience among deployed marines: a pilot study

Susceptibility to PTSD is determined by both genes and environment. Similarly, gene-expression levels in peripheral blood are influenced by both genes and environment, and expression levels of many genes show good correspondence between peripheral blood and brain. Therefore, our objectives were to test the following hypotheses: (1) pre-trauma expression levels of a gene subset (particularly immune-system genes) in peripheral blood would differ between trauma-exposed Marines who later developed PTSD and those who did not; (2) a predictive biomarker panel of the eventual emergence of PTSD among high-risk individuals could be developed based on gene expression in readily assessable peripheral blood cells; and (3) a predictive panel based on expression of individual exons would surpass the accuracy of a model based on expression of full-length gene transcripts. Gene-expression levels were assayed in peripheral blood samples from 50 U.S. Marines (25 eventual PTSD cases and 25 non-PTSD comparison subjects) prior to their deployment overseas to war-zones in Iraq or Afghanistan. The panel of biomarkers dysregulated in peripheral blood cells of eventual PTSD cases prior to deployment was significantly enriched for immune genes, achieved 70% prediction accuracy in an independent sample based on the expression of 23 full-length transcripts, and attained 80% accuracy in an independent sample based on the expression of one exon from each of five genes. If the observed profiles of pre-deployment mRNA-expression in eventual PTSD cases can be further refined and replicated, they could suggest avenues for early intervention and prevention among individuals at high risk for trauma exposure.