Stress, Heritability, and Genetic Factors Influencing Depression, PTSD, and Suicidal Behavior

To better understand how heritability relates to the marked heterogeneity in the presentation of major depression, Nguyen and colleagues (1) used a sibling-based design to analyze data from 1.5 million individuals in the Swedish registry. In contrast to many other studies, an important feature of this study was that diagnosis of MDD had to be made by a “specialist.” Data from individuals with MDD were between 4 and 36 years of age were used in the analyses. The use of different sibling relationships in this sample, full sibling pairs and half sibling pairs, provided varying and predictable degrees of shared genetics, and it was assumed that the sib-pairs in general shared the same environment. Ultimately, 419,495 sibling pairs were used in the analysis and 46,255 of these individuals had a diagnosis of major depression. Eighteen different depression subgroups were identified as defined by numerous characteristics (e.g., severity, recurrence number, comorbidities, suicidality, impairment, and age at first diagnosis). The heritability across the subgroups ranged from 30.5% to 58.3%. Significantly higher heritabilities were found for the MDD subgroups characterized by increased disability (58.3%) and that with youth onset (55.1%), while the lowest heritabilities were found for the subgroups characterized by a single episode (34.4%) and no comorbidity with anxiety disorders (32.2%) or other psychiatric disorders (30.5%). Furthermore, when genetic correlations were performed between the subgroups, a wide range of shared genetics was found, from 0.33 to 0.90, suggesting some shared as well as differing underlying genetics associated with the different subgroups. In their editorial (2), Dr. Mary-Ellen Lynall from the University of Cambridge and Dr. Andrew McIntosh from the University of Edinburgh further comment on the results and point out the value and potential limitations of including half-siblings in the design.

Dr. Anna Docherty, along with over 200 other authors, uses GWAS data from two large datasets to determine the most reliable information related to the genetics of suicidal behavior (3). Findings from both datasets, the International Genetics Suicide Consortium and the Million Veteran’s Program, have been previously reported, and now the researchers perform a meta-analysis by combining these datasets: 29,782 patients with suicide attempts or suicide deaths and 519,961 comparators from the International Genetics Suicide Consortium, and 14,089 veterans with a history of suicide attempt and 395,064 comparators from the Million Veteran Program. The combined total sample for the overall analysis was 43,871 individuals with suicide attempts and 915,025 comparators. Importantly, this sample size allowed for multiple ancestry analyses. In the combined sample, 81% of the individuals were deemed to be of European ancestry, 11% had significant African ancestry admixtures located in the U.S., 5% had East Asian ancestry admixtures, and 3% had Hispanic/Latino ancestry admixtures located in the U.S. Thirteen percent of the sample had died by suicide. Eight significant loci were identified when all the samples were combined. The overall heritability of suicide attempts based on this GWAS analysis was 5.7%. The various single nucleotide polymorphisms (SNPs) significantly associated with suicide attempts were generally intergenic, intronic, or in untranslated regions of the genes. Interesting positive findings involved the sodium- and chloride-dependent glycine transporter 1 gene (SLC6A9), which has been associated with schizophrenia, the dopamine receptor 2 gene (DRD2), and the neuroligin 1 gene (NLGN1). NLGN1 is of interest as it is a gene that is involved in the formation of synapses and neuroplasticity and has also been associated with neurodevelopmental disorders including autism. The separate European ancestry analyses yielded four more significant loci including one found in the major histocompatibility complex (MHC), which contains 224 genes and members of this complex have been associated with schizophrenia (the SNP identified is in close proximity to neurogenic locus notch homolog 4 [NOTCH4]). Separate analyses of the African, East Asian, and Hispanic/Latino ancestries did not reveal significant loci, but their overall predicted heritability scores were around 10%. When examining how the significant loci associated with suicide attempts related to patterns of gene expression, the investigators found that affected genes had enriched expression in the brain and pituitary as well as for genes associated with antipsychotic and antidepressant drug targets. Additionally, genetic correlations revealed some shared genetics with other psychiatric and physical health related factors. In her editorial (4), Deputy Editor Dr. Elisabeth Binder from the Max Planck Institute-Munich elaborates on the findings and highlights the potential importance of the new discoveries facilitated by increasing the diversity of the sample and its size.

Chatzinakos and colleagues (5) use single nuclear RNA sequencing to characterize gene expression alterations in brain cells from the dorsolateral prefrontal cortex of individuals with PTSD and MDD. Tissue from the dorsolateral prefrontal cortex was selected because of the involvement of this region in these illnesses, as well as its important role functioning as a cognitive-emotion interface. The approach used here is very powerful as it allows for an examination of the specific cell types and the molecules within these cells that may be altered in relation to PTSD and MDD. In addition to single nuclear sequencing, the researchers used other cutting-edge methods to enhance an understanding of the detected alterations. These included bulk RNA sequencing, neuroimaging transcriptome wide association studies, and in vitro methods using neurons derived from induced pluripotent stem cells exposed to the synthetic glucocorticoid, dexamethasone. While PTSD and MDD are highly comorbid and both triggered by stressors and trauma, they have some distinctly different features including differential alterations of the hypothalamic-pituitary-adrenal axis (HPA), the product of which is the stress-related release of the glucocorticoid cortisol. For example, in MDD the HPA tends to be hyperresponsive related to decreased glucocorticoid negative feedback mechanisms whereas in PTSD, the opposite appears to be the case in which sensitized glucocorticoid receptor signaling may lead to blunted cortisol responses. The initial brain samples used in this study came from PTSD (N=11), MDD (N=10), and comparison (N=11) subjects. To assure reliability of the findings, analyses were performed in another set of samples from separate individuals, PTSD (N=5), MDD (N=5), and comparison (N=5). While numerous interesting findings emerged, some of the key findings included uncovering differential gene expression between PTSD and MDD in excitatory and inhibitory neurons as well as in astrocytes. Interestingly, the expression of genes in the glucocorticoid signaling pathway was significantly enriched for PTSD compared with MDD, and the in vitro dexamethasone-treated neuron studies were consistent with patterns of gene expression observed in the brains from PTSD patients. In relation to PTSD, several interesting, involved genes appeared to be located on the 17q21.3.1 locus. One of particular interest involved a region linked to the corticotrophin releasing hormone 1 receptor gene (CRHR1), the product of which is a receptor known to be widely involved with the stress response and hypothesized to play an important role in mediating stress-related psychopathology. Taken together, these findings shed new light on molecular mechanisms within dorsolateral prefrontal cortical cells that may contribute to PTSD and MDD. The findings provide an understanding of the molecular mechanisms that may underlie some of the differences in the pathophysiology of these stress-related and highly comorbid illnesses. Finally, it is important to emphasize that this type of study has the potential to direct new treatment development based on an understanding of alterations in brain function at the cellular and molecular level.

While numerous childhood risk factors have been associated with the development of depression, most studies have only examined a limited number of potential factors. To better understand the relations among multiple factors related to the expression of depressive symptoms, Farhat et al. (6) assessed the impact of neurodevelopmental traits, stressors, and emotion dysregulation on depression across development using statistical methods that allowed for an understanding of the complex interactions among these variables. To accomplish this, the investigators performed parallel analyses with data from 4,407 individuals, independent twins from the Twins Early Development Study and 10,351 individuals from the Avon Longitudinal Study of Parents and Children. Initial assessments of traits associated with neurodevelopmental disorders (i.e., measures reflective of ADHD, autism spectrum disorder, intellectual disability and communication and learning disorders), two social-environmental stressors, and measures of emotional dysregulation were collected between 7 and 11 years of age. These measures were then used to predict depressive symptoms at 12, 16, and 21 years of age. Of interest, many significant correlations were found between any two of the neurodevelopmental trait measures and significant correlations were also found between individual differences in the neurodevelopmental traits with the magnitude of depressive symptoms. However, after considering the effects of the stressors and emotional dysregulation, the associations between neurodevelopmental traits and depressive symptoms were no longer significant. Interestingly, individual differences in some of the stress response and emotional dysregulation ratings were correlated with both neurodevelopmental traits and depressive symptoms, further suggesting the possibility that the association between neurodevelopmental traits and the development of depressive symptoms is mediated by other factors such as stressors and the capacity to regulate emotion.

Marr and colleagues (7) used two cohorts of mother-infant pairs to characterize longitudinal trajectories of stress, anxiety, and depression in mothers during pregnancy and 1 month after pregnancy with the aim of understanding how these measures relate to their infants’ trajectory of negative affect development and amygdala resting functional connectivity. The first sample included 115 mother-infant pairs, and for 60 of these infants resting state MRIs were obtained at approximately 1 year of age. The behavioral data from the second sample (2,156 mother-infant pairs) was used to assess the replicability of the findings determined from the first sample. Using a machine learning approach with data from the first cohort, composite stress scores (combining measures of stress, anxiety, and depression) were analyzed to define different subgroups based on the magnitude of mothers’ stress ratings across the entire pregnancy and first postpartum month, as well as the pattern of change of mothers’ stress levels over this time. Across this perinatal period, high and low stress magnitude clusters were identified. In addition, four clusters were found based on different patterns of change in the stress over time scores. In general, the trajectory clusters were replicated in the larger sample. Next, it was determined from the first cohort, and replicated in the second cohort, that one of the trajectory clusters (trajectory 3), characterized by a mid-gestational trough and a late gestational increase in stress, was predictive of a pattern of infants’ negative affect development in a way that differed from the average developmental trajectory of negative affect. Specifically, from 12 to 24 months of age, infants in this group demonstrated a lower magnitude and flattening of their negative affect growth curve. Of note, this finding remained significant when controlling for maternal levels of postnatal depression. While the long-term significance of this is not known, the authors suggest that the normative trajectory of negative affect development is important for acquiring emotion regulation/coping strategies. Finally, in the 60 infants that received MRIs, some of the maternal trajectory clusters were associated with increased infant amygdala resting state functional connectivity with such regions as the ventromedial prefrontal cortex and anterior insula. Increased amygdala connectivity was also observed in infants from the mothers in trajectory 3, which to reiterate was characterized by a mid-gestational trough followed by a late gestational increase in stress. The authors conclude that it is the timing of stress and its pattern over pregnancy that may be most relevant in determining neonatal outcomes and point to the last trimester as being a hypothetically critical period. The findings from this paper are complicated, yet potentially important, and are discussed more fully in an editorial authored by Dr. Alex Shackman from the University of Maryland and Dr. Dylan Gee from Yale (8).

In relation to MDD and PTSD, and across all psychiatric disorders, the interactions between one’s genetic endowment and environmental factors, especially those related to stress, are major determinants of the likelihood of developing illness. In this issue of the Journal, we are presented with new data that further our understanding of these influences in relation to depression, PTSD, and suicidal behavior. The major findings highlighted from these studies include: 1) in characterizing the different presentations of depression, the highest heritabilities were found in individuals with the most disability and those in which the first episode occurred during childhood; 2) in the largest analysis to date, eight SNPs were identified that are significantly related to suicide attempts and these include loci associated with the dopamine receptor 2 gene (DRD2) and the neuroligin 1 gene (NLGN1); 3) an understanding of the gene expression-based molecular alterations within the dorsolateral prefrontal cortex that may underlie some of the differences in the pathophysiology of PTSD and MDD; 4) while neurodevelopmental traits have been associated with the development of depressive symptoms, it appears that this relation may be mediated by other factors during childhood such as stress exposure and individual differences in the ability to regulate emotion; and 5) understanding trajectories of stress exposure during pregnancy provides insights into the development of infants’ negative affect and amygdala function.