Peripheral MicroRNA Signatures in Adolescent Depression

Abstract

Background: Adolescent depression is linked to enduring maladaptive outcomes, chronic severity of symptoms, and poor treatment response. Identifying epigenetic signatures of adolescent depression is urgently needed to improve early prevention and intervention strategies. MicroRNAs (miRNAs) are epigenetic regulators of adolescent neurodevelopmental processes, but their role as markers and mediators of adolescent depression is unknown.

Methods: Here, we examined miRNA profiles from dried blood spot samples of male and female adolescents with clinical depression and psychiatrically healthy male and female adolescents (N = 62). We processed and sequenced these samples using a small RNA protocol tailored for miRNA identification.

Results: We identified 9 differentially expressed (DE) miRNAs (adjusted p value < .05), all of which were upregulated in adolescents with depression. At future follow-ups post blood collection, expression of miR-3613-5p, mir-30c-2, and miR-942-5p were positively associated with depression severity but not anxiety, suggesting a stronger link to persistent depression symptoms. Expression of miR-32-5p inversely correlated with hippocampal volume, highlighting a potential neurobiological basis. Common predicted gene targets of the DE miRNAs are involved in neurodevelopment, cognitive processing, and depressive disorders.

Conclusions: These findings lay the groundwork for identifying adolescent peripheral miRNA markers that reflect neurodevelopmental pathways that shape lifelong psychopathology risk.

Keywords: Adolescence; Depression; Development; Dried blood spots; Mental health; MicroRNA.

Figure 1

Overview of dried blood spot–derived miRNAs DE between healthy control participants and adolescents diagnosed with MDD. (A) The vast majority of mapped reads are dominated by miRNAs. Showing representational image of mapping summary by exceRpt for one-third of the samples combined. (B) Volcano plot showing differential expression analysis between control and MDD groups of 1332 miRNAs. In gray are miRNAs that did not pass any significance threshold. In blue are miRNAs that passed the nominal p value but were below an FC of 1. In red are miRNAs that passed the nominal p value and were FC > 1, with top DE miRNAs that passed false discovery rate correction shown with enlarged circles. (C) Summary table of the top DE miRNAs by p_adjusted < .05, with upregulated FC in the MDD group compared with the control group. (D) Linear models of miRNAs as predictive values for depression severity, measured with RADS T scores, adjusted for the length of the follow-up interval. (E) Linear models of miRNAs as predictive values for anxiety measured with MASC T scores, adjusted for age at follow-up, revealed a lack of prediction. No outliers were detected as per Cook’s distance cutoff >1. [Figures were created in BioRender.com under McGill University Department of Psychiatry plan.] DE, differentially expressed; FC, fold change; lincRNA, long intergenic noncoding RNA; MASC, Multidimensional Anxiety Scale for Children; MDD, major depressive disorder; miRNA, microRNA; misc_RNA, miscellaneous RNA; piRNA, PIWI-interacting RNA; RADS, Reynolds Adolescent Depression Scale; rRNA, ribosomal RNA; scaRNA, small Cajal body-specific RNA; snoRNA, small nucleolar RNA; snRNA, small nuclear RNA; TEC, transcription elongation complex; tRNA, transfer RNA.

Figure 2

Top DE miRNAs were expressed in the human brain, and miR-32 was negatively associated with hippocampal volume. (A) The miTED revealed human brain expression levels of each of the top DE miRNAs. (B) TissueAtlas dataset allowed evaluation of top DE miRNA expression in specific human brain regions. (C) Hippocampal volume comparison by group did not show statistical differences. (D) Partial correlation of miR-32 by hippocampal volume adjusted by ICV and MDD medication use showed a negative association. [Figures were created in BioRender.com under McGill University Department of Psychiatry plan.] DE, differentially expressed; ICV, intracranial brain volume; MDD, major depressive disorder; miRNA, microRNA; miTED, MicroRNA Tissue Expression Database; TIGER, Teen Inflammation Glutamate Emotion Research.

Figure 3

Bioinformatics-based analysis identified highly targeted genes and their association with gene ontologies, pathways, and diseases. (A) Illustration of the top DE miRNAs and their shared predicted gene targets (n = 127). (B) Visualization of the interacting GO terms for the multi-miRNA target genes. (C) Localizing the enrichment of the genes in the body yielded peripheral organ hits as well as some specific brain regions. (D) Performing GO analysis for multi-miRNA target genes, we identified many associations, here showing an excerpt of only the notable biological processes (the pathways/processes to which that gene product’s activity contributes) and cellular components (where the gene products are active). Notable enriched (E) process networks and (F) disease associations for multi-miRNA target genes using MetaCore. [Figures were created in BioRender.com under McGill University Department of Psychiatry plan.] BA, Brodmann area; BP, biological process; CC, cellular component; DEG, differentially expressed gene; ERK, extracellular signal-regulated kinase; FDR, false discovery rate; GO, gene ontology; miRNA, microRNA; TGF, transforming growth factor.

Figure 4

Comparison of the most expressed miRNAs in DBSs in comparison to miRNA expression in other peripheral fluids suggests resemblance to the serum profile. (A) The heatmap of the top 20 expressed miRNAs in the current DBS samples. (B) Based on the additional MicroRNA Tissue Expression Database data, we extracted expression profiles for miRNAs in other whole blood–related samples to compare their similarity with DBS proportions. [Figures were created in BioRender.com under McGill University Department of Psychiatry plan.] DBS, dried blood spot; miRNA, microRNA; TIGER, Teen Inflammation Glutamate Emotion Research.