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

Abstract

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.

Figure 1

A multi-step translational work-flow for identifying methamphetamine-associated psychosis (MAP) biomarkers. First, weighted gene co-expression network analysis (WGCNA) analysis built a global co-expression network and identified 24 co-expression modules. On the hierarchical cluster tree, each line represents a gene (leaf) and each group of lines represents a discrete group of co-regulated genes or gene modules (branch) on the clustering gene tree. Each gene module is indicated by the colour bar below the dendrogram, and subsequently functionally annotated then integrated with recorded clinical and biological data to identify candidate gene modules representing functional biomarkers of MAP. Second, differential gene expression and class prediction methods identified 20 candidate MAP biomarkers (14 were recycled from the second split on the tree). A Bayesian-like convergent functional genomic (CFG) approach prioritized our panel of biomarkers specific to MAP and biomarkers were placed within an empirically derived biological framework. For each step, the corresponding figure and/or table is listed providing a quick reference. LOOCV, leave-one-out cross-validation; RFE, recursive feature elimination.

Figure 2

Significant methamphetamine-associated psychosis (MAP) findings from differential analysis of module eigengene (ME) values across controls (white), MA subjects (light grey) and MAP subjects (dark grey). Modules specific to MAP include (a) ubiquitin (UB)-mediated proteolysis, (b) RNA degradation and (c) circadian clock. Indicated for each module are number of overlapping genes from the module ∩ out of total genes in the term. Enrichment P-values are Bonferroni corrected for multiple comparisons. A Bayes analysis of variance (parameters: conf=12, bayes=1, winSize=5) was used on the ME values to test for significance between the groups and P-values were corrected for multiple comparisons where (*) implies post hoc-corrected P-value significance <0.05 and (⁺) indicates P-value significance <0.05 without post hoc correction.

Figure 3

Top candidate blood biomarkers for methamphetamine-associated psychosis (MAP). (a) Convergent functional genomic (CFG) evidence and scoring are depicted on the right side of the pyramid. Genes in bold have been found in external publications. Genes found in methamphetamine (METH)-free studies investigating schizophrenia (SCZ, ^†) and psychosis (*) are as indicated. (b) Overlapping gene–disease relationships including CFG-validated genes within gene modules (ubiquitin-mediated proteolysis and circadian rhythm) and single-gene biomarkers. Nodes represent genes and edges indicate gene–disease relationships. Node shape denotes empirically derived functions from our network analysis. Green shading indicates biomarkers from our machine-learning analysis including 14 unique genes separating controls from METH dependants. Grey nodes represent CFG-validated biomarkers of delusion (psychosis) or SCZ.^, Node border colour in turquoise indicates gene signatures across MAP, general psychosis and SCZ studies. Venn diagram depicts lack of overlap from curated haloperidol gene signatures onto the 128 candidate MAP genes (61 UPS+39 clock+25+20=128 genes (while accounting for overlap across lists)).