Skip to main content

Verified by Psychology Today

Reviewing the Evidence for Mental Illness Being Epigenetic

DNA modification has emerged as the critical factor in mental illness.

When people ask me to explain the imprinted brain theory, I usually start by pointing out that the paradox of mental illnesses such as schizophrenia or autism is that, although they are known to run in families and show high concordances between twins, they cannot be explained by classical, Mendelian genetics. But the paradox can be resolved if you consider the possibility that the expression of genes, and not just their inheritance, is critical. Specifically, the theory invokes variations in gene expression such as imprinting and X-inactivation, which along with other similar effects, come about by way of various types of DNA modification (DNAM) (diagram below).

Epigenomics (2015) 7(3), 427–449
The interplay between DNA and the histone proteins that define the functional state of chromatin. While histone and DNA molecules are tied together, any modification of the amino acids of the histone tails, such as acetylation (Ac) or methylation (Me) would change the electric charge of histone proteins and alter its binding affinity to the attached DNA and modify the accessibility of transcription factors to the DNA, and thus the corresponding gene expression. As shown, there are several enzymes that add or remove methyl or acetyl groups based on the cellular context making the epigenetic modifications quite a complex process.
Source: Epigenomics (2015) 7(3), 427–449

Today this is often called epigenetics, and as a recent review article points out, interest in it has come about in part because of the failure to discover classical Mendelian genes that could account for mental illnesses. Despite tens of thousands of cases and controls being examined in genome-wide association scans, no single gene or group of genes with a significant effect has been identified. As a result, examination of the potential role of epigenetic modifications has become an alternative to classical genetics for mental diseases, just as the imprinted brain theory proposes. And as the theory would predict, many genes associated with autism, schizophrenia, and bipolar disorder have been found to show epigenetic modification, and indeed are listed and discussed in this review in too much detail to be repeated here.

Because brain cells do not replicate themselves much, epigenetic effects such as DNA modification via factors such as differential mythelation illustrated above come into play at different times so that age is one of the major determinants of brain-specific DNAM alterations. Indeed, based on a whole-genome DNAM profiling of the frontal cortex in more than 100 different individuals, the overall levels of methylation were found to increase with age. Brain development, in other words, is not set in stone once you are an adult. On the contrary, these findings suggest that brain cells remain capable of epigenetic modification throughout life.

In addition to nutritional factors, maternal distress also affects the epigenetics of the brain in the fetal period in ways which are retained in adulthood. In rats, stress during pregnancy decreases brain-derived neurotrophic factor expression associated with changes in the amygdala and hippocampus of offspring both at weaning and adulthood. Prenatal stress also decreases expression of a corticotrophin-releasing hormone gene associated with an increase in the responsiveness of the hypothalamic-pituitary-adrenal axis to the point of producing acute stress and anxiety-like behaviors in adolescent rats. Prenatal stress in pregnant mice has also been found to affect epigenetic factors in the offspring, and produces schizophrenia-like effects in adults that are in turn attenuated by antipsychotic drugs. In addition, depression, nutritional elements and vitamins, body mass index, the infant’s gestational age, birth order and birth weight have been associated with epigenetic changes of many genes. Indeed, the latter factor has also been shown to confirm the highly counter-intuitive predictions of the imprinted brain theory in a huge cohort, as I pointed out in a recent post.

In human studies, childhood physical aggression has been linked to lower serotonin synthesis in the orbito-frontal cortex during adulthood (measured by positron emission tomography) associated with altered DNAM of the serotonin transporter gene (5-HTT) promoter and regulatory regions of many cytokines and transcription factors.

Additionally, a comparative whole-genome DNAM analysis uncovered methylation changes in … saliva samples from 96 maltreated children who were separated from their parents (because of child abuse/neglect) compared with matched control subjects. Such long-lasting DNAM alterations were also observed by whole-genome DNAM analysis of the hippocampal tissues of individuals with traumatic life events (25 cases vs 16 controls) in childhood and adolescence.

Electro-convulsive treatment (ECT) has long been known to be one of the most effective treatments for psychotic illnesses, and this review goes on to note that “In addition to drugs, clues from animal studies uncovered that even neuronal activity induced by electroconvulsive stimulation (mimicking ECT in humans) in adult mice renders DNA demethylation of specific neuronal genes … and promotes neurogenesis in hippocampus.” Indeed, as Randy Jirtle has suggested to me, this may also explain why ECT, although effective, seems to wear off with time and repetition: perhaps because the epigenetic changes it induces are transitory rather than permanent. But however that may be, the authors of this review go on to speculate that “in addition to electroconvulsive therapy (ECT), new therapeutic approaches such as transcranial magnetic stimulation (TMS) that can provoke targeted neuronal activities may attain applications to induce epigenetic modifications in specific neuronal pathways affected in psychotic diseases.”

Whatever else you may say about it, the basic claim of the imprinted brain theory that gene expression is critical in neuro-development has now been thoroughly vindicated. And as I have repeatedly pointed out in countless previous posts, detailed studies of specific genes such as XIST (to quote just the most recent example) are progressively endorsing the theory’s provocatively precise predictions about exactly what these patterns of gene expression are.

(With thanks to Tamara Shengelia and Vikas Chandra for their help.)