The crossroads of psychiatric epigenomics

Traditionally, the origin of mental disorders has been attributed to the combination of genetic and environmental risk factors. Since environmental effects are difficult to pinpoint, a strong emphasis has been placed on genetics. However, despite the increasing scale and scope of contemporary genetic studies on mental disorders, it is evident that structural DNA differences cannot explain all facets of these disorders. The beginning of 21st century in psychiatric research was marked by the arrival of epigenetics, followed by its large scale version – epigenomics1. The rapidly increasing popularity of epigenetic approaches in psychiatric diseases, and human morbid biology in general, can be explained by several factors. A series of cell biology studies showed that histone proteins are not just passive scaffolds for packing the two sets of 2 meter-long DNA strands into the micrometer-sized cell nucleus. Acetylation, methylation, phosphorylation and other types of chemical modifications of histones determine accessibility to local chromatin, which, in turn, regulates transcription factor binding and gene activation. The discovered regulatory functions of histone modifications resonated well with those of DNA modifications, which included methylation and other covalent chemical attachments to cytosines. By this time, DNA modifications had already been shown to account for monoallelic expression of imprinted genes, suppression of genomic retroelements, and X chromosome inactivation2. With some exceptions, the density of modified cytosines in gene regulatory elements correlates with gene transcriptional activity. The realization that epigenetic factors play a pivotal role in the regulation of genes and genomes put them on par with DNA sequence variation. It has become evident that the failure of epigenetic “software” can be as detrimental as changes in the DNA sequence “hardware” . Furthermore, the responsiveness of the epigenetic code to the environment allows it to be an “interface” between genes and the environment, thereby positioning epigenetic studies to offer new insights to DNA-environment interactions in determining complex phenotypes3. For example, one of the main mysteries in human morbid biology is why identical twins frequently do not share phenotypes. It has been known from isogenic plants and inbred animal studies that genetically identical organisms can exhibit numerous epigenetic differences, some of which may translate to different phenotypic outcomes4. Notably, not all epigenetic variations need to be caused by environmental factors; stochastic errors of maintenance of epigenetic profiles may accrue over time, resulting in significant molecular and, by corollary, phenotypic differences. Epigenetics has identified some “blind spots” in psychiatric research, with the most notable being age- and time-dependent phenotypic changes in organisms. There is increasing evidence that epigenomes age in a partially deterministic fashion. The best illustration of programmed aging is the epigenetic “clock”: age-dependent methylation changes at several hundred cytosines that allow for the precise prediction of an individual's chronological age5. Slow deviation from healthy epigenetic aging may not immediately cause health problems, nor even for an extended period of time. Such a mechanism may explain why individuals carrying inherited disease risk factors remain disease-free for several decades after birth. Following disease onset, epigenomes may also fluctuate, which has the potential to translate into disease remissions and relapses. Finally, the accumulation of epigenetic aging changes in elderly adults may surpass the effects of disease epigenomes, resulting in partial recovery of psychiatric symptoms in late life. The rapidly increasing interest in psychiatric epigenetics and epigenomics is best illustrated by PubMed statistics. Over the last 18 years, the annual number of publications with the key words “epigenetic AND psychiatric” has increased nearly 80-fold (6 and 478 papers in 2000 and 2018, respectively). The majority of studies focus on DNA methylation analysis. The typical study investigates peripheral blood samples collected from hundreds, and in rare cases thousands, of psychiatric patients and controls, with the use of Illumina microarrays that interrogate ~450,000 cytosine sites in gene regulatory and coding regions. It is worth noting that Illumina microarrays profile only a small fraction of cytosines with the potential to carry brain disease-relevant epigenetic changes. Analysis of the entire methylome of multiple individuals in populational studies has thus far been prohibitively expensive, leaving large parts of the epigenome unexplored. Despite significant efforts, psychiatric epigenetic and epigenomic findings are modest thus far, and their interpretation is difficult6. The majority of studies have been performed using white blood cells. To date, it is still not clear whether such “surrogate” tissues can be useful for mental disorder studies, given that epigenetic status and dynamics in neurons and glial cells are distinct from those of white blood cells. The studies performed in post-mortem brain are of particular interest, but they also face the issue of separating genuine disease-causing epigenetic signals from those that result from living with the disease or unrelated processes taking place over an individual's lifetime. Another concern is brain cell type heterogeneity – the mixture of many different subtypes of neurons and glial cells. If proportions of such cell subtypes were to vary, the detected epigenetic differences may reflect cellular differences, rather than the sought-after disease-related epigenetic changes. Finally, even in large and well-designed studies, the mean DNA methylation differences between psychiatric patients and controls rarely exceeded 1%7, which has made biological interpretation difficult, especially considering the large variation exhibited by epigenetic marks across individuals. At the same time, basic epigenetics is progressing rapidly. Recent large epigenomic studies such as PsychENCODE documented numerous new layers in epigenetic and chromatin organization of the brain8. The list of epigenetic marks is increasing, and it has been recently detected that major monoaminergic neurotransmitters, such as serotonin, can be attached to histones and facilitate gene expression in neurons9. In parallel, experimental approaches have become more sophisticated and informative. Several laboratory innovations are of particular interest for psychiatric epigenomics. First, single-cell approaches are redefining the meaning of epigenetic stochasticity and directly address the issues of cell type differences in the brain. Second, easily available somatic cells, such as fibroblasts, can be reprogrammed into neurons, partially addressing the need for brain tissue. Third, CRISPR-Cas9 technology can be used not only for editing genomes, but also epigenomes, which is of considerable interest for modeling disease components in tissue culture and animals. Fourth, progress in computational strategies has enabled the integration of epigenomic data with genomics, transcriptomics, and metabolomics. The comprehensive trans-omic approaches enable the identification of hub elements and cellular pathways centrally involved in disease. Given the rapid developments in molecular biology and brain imaging technologies, an ideal experiment – a prospective epigenomic study in the living brain of psychosis-predisposed individuals – may not be science fiction in the near future. Despite the challenges thus far, epigenetics and epigenomics remain an important part of the psychiatric research agenda. There are still no better ways to explain the numerous dynamic features of complex diseases, which by definition do not conform with the stability of DNA sequence. Uncovering the mechanisms of discordance in monozygotic twins or the delayed age of psychosis onset would be of major importance for precision psychiatry. The success and progress of psychiatric epigenetics relies on the ever improving experimental and computational tools and, more importantly, on the diligence and creativity of scientists working on this very interesting, but also challenging, part of human biology.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.