Novel candidate genes for ECT response prediction—a pilot study analyzing the DNA methylome of depressed patients receiving electroconvulsive therapy

Nicole Moschny, Tristan Zindler, Kirsten Jahn, Marie Dorda, Colin Davenport, Lutz Wiehlmann, Hannah Benedictine Maier, Franziska Eberle, Stefan Bleich, Alexandra Neyazi, Helge FrielingClinical 2020 · DOI

Abstract

We suggest ten novel candidate genes to be implicated in either ECT response or its possible mechanism. Because of the small sample size of our pilot study, our findings must be regarded as preliminary.

Introduction

Major depressive disorder (MDD) is a leading cause of global disability. While standard antidepressants are widely used, approximately 30–50% of patients fail to achieve remission. For these treatment-resistant cases, electroconvulsive therapy (ECT) remains one of the most potent clinical interventions. However, ECT's success rate is roughly 50% in naturalistic settings, and clinicians currently lack reliable biomarkers to predict which patients will benefit. This often leads to a "trial and error" approach, delaying effective care. This pilot study moves beyond traditional "usual suspect" genes (like BDNF) to explore the broader DNA methylome—the landscape of chemical modifications that regulate gene expression—to identify novel predictive biomarkers for ECT response.

Methods: A Discovery-Based Approach

The researchers conducted a longitudinal study involving 17 patients with treatment-resistant MDD, with 12 patients providing complete data for the final epigenetic analysis. Peripheral blood mononuclear cells (PBMCs) were collected at four critical time points: before and after the first ECT session, and before and after the final session of the treatment course.

Unlike previous studies that utilized lower-resolution arrays, this team employed the TruSeq Methyl Capture EPIC Kit combined with next-generation sequencing (NGS). This high-resolution approach allowed for the analysis of over 3.3 million CpG sites, including often-overlooked regions such as introns and long non-coding RNA (lncRNA) transcripts. Clinical response was defined as a ≥ 50% reduction in the Montgomery-Åsberg Depression Rating Scale (MADRS) score.

Key Results: Identifying the "Top Ten" Candidates

The analysis identified ten candidate genes associated with ECT. Eight of these genes (RNF175, RNF213, TBC1D14, TMC5, and WSCD1) showed baseline methylation differences that predicted treatment response. Two additional genes (AQP10 and TRERF1) exhibited methylation changes during the treatment course, suggesting they may be involved in the biological mechanism of ECT itself.

Heatmap of DNA methylation at the RNF213 locus, illustrating distinct patterns between ECT responders and non-responders.

The differences were particularly striking in the RNF213 gene. Despite the small cohort, specific probes reached high levels of statistical significance (p < 0.001), even after adjusting for multiple testing. Furthermore, responders and non-responders showed significant clinical divergence at baseline, including differences in leukocyte counts and the duration of their current depressive episode.

Key Insights: Autophagy and Vascular Health

A major finding of this study is the clustering of candidate genes within specific biological pathways. Several genes, including RNF213, RNF175, and TBC1D14, are integral to the ubiquitin-proteasome system and autophagy. Autophagy is the cellular "recycling" process responsible for clearing damaged proteins. This suggests that ECT may exert its therapeutic effects by modulating protein turnover or "cleaning up" cellular debris in the brain.

Longitudinal stability of mean DNA methylation levels across four treatment time points for candidate genes.

Additionally, the identification of RNF213—a gene known for its role in vascular health and Moyamoya disease—points toward a potential link between vascular remodeling, angiogenesis, and recovery from depression. Crucially, the methylation patterns of these predictive genes remained remarkably stable throughout the treatment course, reinforcing their potential as baseline biomarkers for clinical decision-making.

Limitations and Future Directions

As a pilot study, the primary limitation is the small sample size (n=12). While the results are statistically significant, they require validation in larger, independent cohorts. Additionally, the use of blood-based markers (PBMCs) serves as a proxy for brain-level changes; while often correlated, the relationship is not absolute. Future research should determine if these epigenetic signatures are specific to ECT or represent a broader biological marker of recovery applicable to other rapid-acting treatments like ketamine or rTMS.

Conclusion

This study identifies ten novel candidate genes that may predict or mediate the response to ECT in patients with treatment-resistant depression. By highlighting pathways related to autophagy and vascular health, these findings provide a foundation for developing personalized treatment strategies and deepening our understanding of ECT’s biological mechanisms.

Summary updated

Abstract

We suggest ten novel candidate genes to be implicated in either ECT response or its possible mechanism. Because of the small sample size of our pilot study, our findings must be regarded as preliminary.
BibTeX
@inproceedings{moschny2020novel,
  author = {Nicole Moschny and Tristan Zindler and Kirsten Jahn and Marie Dorda and Colin Davenport and Lutz Wiehlmann and Hannah Benedictine Maier and Franziska Eberle and Stefan Bleich and Alexandra Neyazi and Helge Frieling},
  title = {{Novel candidate genes for ECT response prediction—a pilot study analyzing the DNA methylome of depressed patients receiving electroconvulsive therapy}},
  booktitle = {Clinical},
  year = {2020},
  doi = {10.1186/s13148-020-00891-9},
}
Figure