Logo: to the web site of Uppsala University

Depression is a multifactorial disorder representing a significant public health burden. Previous studies have linked multiple single nucleotide polymorphisms with depressive phenotypes and suicidal behavior. MAD1L1 is a mitosis metaphase checkpoint protein that has been linked to depression in GWAS. Using a longitudinal EWAS approach in an adolescent cohort at two time points (n = 216 and n = 154), we identified differentially methylated sites that were associated with depression-related genetic variants in MAD1L1. Three methylation loci (cg02825527, cg18302629, and cg19624444) were consistently hypomethylated in the minor allele carriers, being cross-dependent on several SNPs. We further investigated whether DNA methylation at these CpGs is associated with depressive psychiatric phenotypes in independent cohorts. The first site (cg02825527) was hypomethylated in blood (exp(beta) = 84.521, p value similar to 0.003) in participants with severe suicide attempts (n = 88). The same locus showed increased methylation in glial cells (exp(beta) = 0.041, p value similar to 0.004) in the validation cohort, involving 29 depressed patients and 29 controls, and showed a trend for association with suicide (n = 40, p value similar to 0.089) and trend for association with depression treatment (n = 377, p value similar to 0.075). The second CpG (cg18302629) was significantly hypomethylated in depressed participants (exp(beta) = 56.374, p value similar to 0.023) in glial cells, but did not show associations in the discovery cohorts. The last methylation site (cg19624444) was hypomethylated in the whole blood of severe suicide attempters; however, this association was at the borderline for statistical significance (p value similar to 0.061). This locus, however, showed a strong association with depression treatment in the validation cohort (exp(beta) = 2.237, p value similar to 0.003) with 377 participants. The direction of associations between psychiatric phenotypes appeared to be different in the whole blood in comparison with brain samples for cg02825527 and cg19624444. The association analysis between methylation at cg18302629 and cg19624444 and MAD1L1 transcript levels in CD14+ cells shows a potential link between methylation at these CpGs and MAD1L1 expression. This study suggests evidence that methylation at MAD1L1 is important for psychiatric health as supported by several independent cohorts.

Introduction Depression is a major global burden with unclear pathophysiology and poor treatment outcomes. Diagnosis of depression continues to rely primarily on behavioral rather than biological methods. Investigating tools that might aid in diagnosing and treating early-onset depression is essential for improving the prognosis of the disease course. While there is increasing evidence of possible biomarkers in adult depression, studies investigating this subject in adolescents are lacking.Methods In the current study, we analyzed protein levels in 461 adolescents assessed for depression using the Development and Well-Being Assessment (DAWBA) questionnaire as part of the domestic Psychiatric Health in Adolescent Study conducted in Uppsala, Sweden. We used the Proseek Multiplex Neuro Exploratory panel with Proximity Extension Assay technology provided by Olink Bioscience, followed by transcriptome analyses for the genes corresponding to the significant proteins, using four publicly available cohorts.Results We identified a total of seven proteins showing different levels between DAWBA risk groups at nominal significance, including RBKS, CRADD, ASGR1, HMOX2, PPP3R1, CD63, and PMVK. Transcriptomic analyses for these genes showed nominally significant replication of PPP3R1 in two of four cohorts including whole blood and prefrontal cortex, while ASGR1 and CD63 were replicated in only one cohort.Discussion Our study on adolescent depression revealed protein-level and transcriptomic differences, particularly in PPP3R1, pointing to the involvement of the calcineurin pathway in depression. Our findings regarding PPP3R1 also support the role of the prefrontal cortex in depression and reinforce the significance of investigating prefrontal cortex-related mechanisms in depression.

The causes of depression are complex, and the current diagnosis methods rely solely on psychiatric evaluations with no incorporation of laboratory biomarkers in clinical practices. We investigated the stability of blood DNA methylation depression signatures in six different populations using six public and two domestic cohorts (n = 1942) conducting mega-analysis and meta-analysis of the individual studies. We evaluated 12 machine learning and deep learning strategies for depression classification both in cross-validation (CV) and in hold-out tests using merged data from 8 separate batches, constructing models with both biased and unbiased feature selection. We found 1987 CpG sites related to depression in both mega- and meta-analysis at the nominal level, and the associated genes were nominally related to axon guidance and immune pathways based on enrichment analysis and eQTM data. Random forest classifiers achieved the highest performance (AUC 0.73 and 0.76) in CV and hold-out tests respectively on the batch-level processed data. In contrast, the methylation showed low predictive power (all AUCs < 0.57) for all classifiers in CV and no predictive power in hold-out tests when used with harmonized data. All models achieved significantly better performance (>14% gain in AUCs) with pre-selected features (selection bias), with some of the models (joint autoencoder-classifier) reaching AUCs of up to 0.91 in the final testing regardless of data preparation. Different algorithmic feature selection approaches may outperform limma, however, random forest models perform well regardless of the strategy. The results provide an overview over potential future biomarkers for depression and highlight many important methodological aspects for DNA methylation-based depression profiling including the use of machine learning strategies.