Epigenetic changes in burnout and depression
Burnout and depression present some of the main public health and socio-economic burdens in the modern society. Burnout also overlaps with depression, which hinders its clinical assessment. Despite an increasing amount of research on burnout, our knowledge on biological mechanisms in burnout remains scarce, as well as its comparison to depression on this level. Recently, epigenetics has emerged as the key mechanism by which environment and genetics interact, however no studies were conducted to test the relevance of epigenetics for burnout. In contrast, epigenetic mechanisms in depression have been somewhat studied, however often without simultaneous assessment of their functional effects. Therefore, the main objective of this PhD project was to investigate and compare DNA methylation changes, as one of the main epigenetic mechanisms, in burnout and depression.
To achieve this, first we conducted a systematic literature review (Chapter 2) in order to gain more insight into potentially significant genes and pathways on which we could further focus in our clinical studies. Based on the outcomes of Chapter 2, we made a selection of genes on which we focused in the clinical studies: the glucocorticoid receptor gene (NR3C1), the serotonin transporter gene (SLC6A4) and the brain-derived neurotrophic factor gene (BDNF).
In Chapter 3, we developed a sensitive and specific UPLC-MS/MS method for the simultaneous identification and quantification of salivary cortisol and cortisone, with improved performance compared to the previous methods. In addition, we assessed the impact of different covariates and the temporal variability of these hormones in healthy individuals over 1 week. We concluded that both cortisol and cortisone show high day-to-day variability and we identified the main covariates affecting their fluctuations. This method was applied to quantify salivary cortisol and cortisone in Chapter 4 and the covariates were included in the statistical models in this chapter.
Chapters 4-7 were based on two clinical studies: a cross-sectional study on subjects with burnout and healthy controls and a longitudinal study on depressed patients and healthy controls. In Chapter 4 and 5, we investigated DNA methylation patters of stress-related genes (NR3C1 and SLC6A4) in burnout (Chapter 4) and depression (Chapter 5), together with functional measurements of the (hypothalamic pituitary adrenal) HPA axis (cortisol and cortisone). In Chapter 4, we observed overall decreased NR3C1 methylation in burnout patients compared to healthy controls, indicating potentially increased expression and the overall increased sensitivity of the HPA axis in burnout. What could also support this hypothesis is the fact that we observed a negative association between salivary cortisone levels and the average NR3C1 methylation and a clear increase in cortisone levels in the burnout group.
In contrast, in Chapter 5, we observed uniformly increased methylation of NR3C1 in depressed patients, compared to the control group. In addition, we found blunted cortisol reactivity to a laboratory psychosocial stressor in depressed patients and the direct association between this outcome and the average NR3C1 methylation, which makes these finding convincing and sound. Another novel finding in this study was the predictive value of the baseline NR3C1 methylation for symptom improvement at the 8-week follow-up, which could also have potential translational implications, as a lot of effort is being made to identify biomarkers of treatment resistance in depression. Finally, an interesting observation is that we found DNA methylation at the exact same CpG position (CpG8, chr17:4835) of the SLC6A4 gene to be increased on both burnout and depression and negatively associated with cortisol levels in both outcomes, indicating potentially shared epigenetic mechanisms between these two phenotypes.
In Chapter 6 and 7, we focused on elucidating the role of BDNF methylation in burnout (Chapter 6) and depression (Chapter 7). In Chapter 6, we assessed a common polymorphism in the BDNF gene (Val66Met), DNA methylation of different BDNF regions and serum levels of the BDNF protein in subjects in burnout and healthy controls. We observed hypermethylation of the BDNF promoter of exon I and IV in subjects with burnout compared to healthy controls, which was also associated with downregulation of the serum BDNF protein levels and was associated with burnout symptoms dimensionally. We did not observe any role of the common BDNF polymorphism in these associations, indicating absence of genetic vulnerability at this specific locus, but rather emphasizing the environmental component in BDNF regulation in burnout.
In Chapter 7, we assessed the common polymorphism in the BDNF gene and DNA methylation of different BDNF regions in depression, focusing on two specific outcomes that play an important role in the BDNF pathway: 1) anhedonia and reward learning and 2) cognitive performance. Interestingly, we observed decreased methylation of BDNF promoter I in depressed patients compared to healthy controls, which is opposite to what we found in burnout. Moreover, we observed a moderation effect of BDNF methylation on the association between the BDNF polymorphism and impaired reward learning. This could be a distinguishing point between burnout and depression on a biological level, indicating that BDNF upregulation might be more specific for depression, particularly with regard to pathways involved in anhedonia and reward learning, which are not necessarily impaired in burnout. Finally, we observed the association between the BDNF polymorphism and cognitive performance (in the executive domain) in depressed patients, which was mediated by methylation of exon IX. This indicates that exon-specific epigenetic regulation BDNF might play a different role when it comes to different functional outcomes (anhedonia vs. cognitive performance) in depression.
To conclude, we observed DNA methylation changes in stress-related genes in burnout and depression, which had functional effects on the BDNF pathway and HPA axis functioning. Some identified patterns seem to be shared between these two phenotypes (SLC6A4 methylation), whereas other were changed in opposite direction (BDNF and NR3C1). Future replication studies on larger samples and longitudinal studies to better understand the evolution of these epigenetic patterns over time and in different stages of symptom progression and recovery are necessary to deepen the knowledge on this topic and understand its potential translational utility.