Gap in the knowledge

I have argued that rumination can be regarded as an unsubstitutable endophenotype for major depression, I have detailed candidate genes proven in its background so far, and I have delineated two new directions of candidate gene studies: the folate metabolism on the one hand, and new candidates with regard to the serotonergic system: childhood maltreatment and 5-HT2A receptor, on the other hand.

According to this scope, for candidate genes to test I chose two SNPs from two folate genes: rs1801133 from the MTHFR gene, because of its contradictory associations with both depression and cognitive performance, and rs11754661 from the MTHFD1L gene, because it has been investigated neither with depression nor with cognitive vulnerability yet. My hypotheses were to test their associations with ruminative response style measured by the most widely used RRS.

As an additional candidate, I chose rs3125 from HTR2A gene encoding the 5-HT2A

receptor protein. Rs3125 resides in a microRNA (miRNA) binding site (https://snpinfo.niehs.nih.gov/snpinfo/snpfunc.html), and binds five different miRNAs (https://snpinfo.niehs.nih.gov/cgi-bin/snpinfo/mirna.cgi?2_rs3125) (170). From these miRNAs, miR-539 is bound by the G allele of rs3125, and shows decreased expression in ACC in an animal model of chronic neuropathic pain (171). This is interesting because ACC volume and resting state activity had an inverse association with rumination (see chapter 2.1.4.3 for details), and ACC had a reduced 5-HT2A receptor binding in treatment-resistant depressed patients (see chapter 2.5.2 for details). Moreover, the C allele of rs3125 has been proven to be a risk for depressive symptoms in cardiac patients (172),

41

and one possible way through which cardiovascular disorders may be associated with depression is a form of perseverative cognition: rumination (35, 36).

Besides miRNA binding, another epigenetic modification, methylation, has also been revealed with respect to the functioning of the HTR2A gene, linking environmental impacts to psychiatric or cognitive outcomes. Methylation level of the promoter region of HTR2A gene within placenta was positively associated with the newborn infants’

attention score and negatively with quality of movement score (173), which scores predict medical and behavioural problems later in childhood, making the precise timing of epigenetic regulation processes inevitable to consider in the background of adult psychiatric disorders (174). In the promoter region of HTR2A, a cytosine at position -1439 is methylable only if the allele is G at the adjacent polymorphic -1438 A/G (rs6311) site (173, 175, 176). Investigating the frontal lobe, schizophrenic and bipolar patients showed a reduced HTR2A expression compared to controls, and, correspondingly, this -1439 position was found to be more methylated in them than in controls (175). Moreover, antipsychotic use, regardless of whether typical or atypical antipsychotics, was related to a lower methylation level than that of drug-free patients (175). In addition and being consistent also with gene expression levels, methylation level at -1439 decreased with age in individuals with C/C genotype, but this was true only in controls, but not in schizophrenic and bipolar patients (175). (Note that this nomenclature stands for the complementary strand to the one in the -1438 A/G name.) Regarding rs6311 genotype in itself, Fiocco et al, 2007 (177) found that rs6311 G/G (C/C on the other strand) genotype was related to higher levels of depression, neuroticism, emotion-based coping strategies, and cortisol response following a psychosocial stressor, all of which can be associated to rumination (see chapters 2.1.1.1., 2.1.2. and 2.1.4.2 for details). Consequently, I also chose rs6311 from the promoter region of HTR2A to test in association with rumination, expecting the G allele (C allele on the other strand) to be the risk. Thus, I could tag two distinct epigenetic regulatory mechanisms within HTR2A: miRNA binding and methylation.

Based on the literature reviewed in chapter 2.5.1, when testing the association of these two HTR2A SNPs with RRS rumination, I also considered childhood adversity and the two rumination subscales: brooding and reflection.

42

With regard to all SNP-rumination associations, in the perspective of the endophenotype concept, I also tested whether or not these possible findings can be related to depression.

To test the possibility of robustness, I also tested replicability of the findings within the two subsamples of my study sample: the one recruited in Budapest, Hungary and the one in Manchester, United Kingdom.

Regarding rs6311 and depression, moderating effects of different types of stress have been demonstrated to be contradictory. Dressler et al, 2016 (178) found that rs6311 A/A genotype (T/T on the other strand) denoted a risk for higher CES-D depression only in case of a high level of childhood adversity, but this interaction effect on depression was entirely mediated by cultural consonance in family life. It means the degree to which the individual incorporates salient cultural models into own beliefs and behaviours (178). A allele (T allele on the other strand) of rs6311 has been positively related to either seasonal affective disorder or specifically to its winter-type, but not to seasonality itself (179, 180), but negative results have emerged with regard to seasonal affective disorder, too (181).

These contradictory results with rs6311 may be explained by gene x gene interactions, specifically, G allele (C allele on the other strand) decreases in vitro and ex vivo promoter activity in function of genotypes on other HTR2A promoter SNPs (182). Thus, in case of finding no significant association between rs6311 and depression in any of our models significant for a rumination phenotype, we test its association with a complex depression-anxiety phenotype in a complex Bayesian model, to further seek its possible role in depression, if not through rumination. Since a third type of stress, recent stress has been found to have a huge impact on depression (183), this complex model will include recent stress, besides six other polymorphisms previously related to depression: HTR1A rs6295 (184), SLC6A4 5-HTTLPR (185), BDNF rs6265 (108), GALR2 rs8836 (186), CNR1 rs7766029 (187) and P2RX7 rs7958311 (188).

43