Taking the Next Step in Understanding the Genetic Roots of Mental Illness

Taking the Next Step in Understanding the Genetic Roots of Mental Illness

Posted: March 31, 2020
Taking the Next Step in Understanding the Genetic Roots of Mental Illness

Story highlights

What has study of the human genome taught us about the relationship between genetic variation and individual risk for illnesses like schizophrenia, bipolar disorder, and autism? Among other things, that much of the risk in most people comes from the part of the genome that doesn’t contain genes, but instead, the 98% of our genetic material that regulates our genes—tells them when to switch “on” and “off,” for how long, and in which types of cells.


The human genome’s three billion pairs of DNA “letters” are a code of instructions packed tightly in the center of every cell, bearing our genetic inheritance. The sequence of those letters, which holds so much potential to help us understand health and illness, has been known to science for less than 20 years.

Since the full human sequence was first assembled, in the early 2000s, much of the news about how our genes are involved in psychiatric illnesses like schizophrenia, bipolar disorder, and autism has centered on the discovery of variations in the DNA sequence— variations that scientists have been able to correlate with increased illness risk.

Now, a new phase of genome research has begun, powered by major advances in analysis pioneered by dozens of experts involved in an National Institute of Mental Health (NIMH)-funded research consortium called PsychENCODE. Among its founding members are 17 investigators who are members of BBRF’s Scientific Council or have received BBRF grant awards and prizes.

Their project, launched in 2015, has moved an important step beyond the identification of DNA variations associated with elevated risk for specific disorders. In PsychENCODE’s first set of results—a set of 11 papers published in the journals Science, Science Translational Medicine, and Science Advances—the focus is on figuring out how DNA variations perturb the brain’s biology, impairing its normal function.

Obtaining a multi-dimensional picture of how genetic variation affects mechanisms in the brain, say members of PsychENCODE, is essential if genome discoveries are to be translated into a basis for new treatments.

What is specifically new in PsychENCODE’s mission is its focus on understanding how genetic variations affect the way the human genome is regulated—the biological processes that determine how, when, and where in the brain genes are activated and silenced.


Figuring out how our genes are controlled has always been a part of genome science. But its special importance for understanding the mechanisms involved in psychiatric illness has taken a while to become a focal point of research. To understand why this new phase in research is important, we review in this article the deepest roots of the question, which can be traced to the years just after publication of the human genome. That was when researchers began to realize that they weren’t going to discover a single “gene for schizophrenia” or any other psychiatric illness, as some may have hoped.

Rather, researchers discovered that risk for psychiatric illnesses tends to be “highly polygenic.” This means that many combinations of DNA variations— cumulatively occurring in as many as 1,000 of our 21,000 genes—contribute to risk, when viewed at the level of the entire human population.

In light of this discovery, the question for an individual becomes: Which of these many variations, if any, do I carry in my own genome? And how, if at all, might the variations that I have in my genome affect my mental health and that of my children and grandchildren? Answers to these questions involve understanding what “risk” means in the genomic context.


Every one of us, on the basis of our unique gene sequence alone, carries some measurable risk of psychiatric illness, just as we do for cancer and other illnesses. And as with cancer and other illnesses, risk from our genes is only part of the equation. Other factors impact an individual’s risk, such as the way the activity of their genes is affected by environmental factors, ranging from conditions in the womb to those of early childhood and beyond. These interactions affect, in varying degrees, the impact (if any) that genome variations will have on an individual’s mental health.

In most people, the genetic portion of risk for psychiatric illness—schizophrenia or bipolar disorder, for example—is extremely low. But in a small yet significant minority, it is very high. One person in 100 develops schizophrenia, and about two in 100 is diagnosed with bipolar disorder. These are common, not rare diseases— and yet most people are unlikely to be affected by them as a consequence of variations in their genetic material.

Schizophrenia and bipolar disorder are complex disorders, meaning that they are typically caused by multiple factors which interact, both genetic and environmental. In this respect they are unlike disorders caused by problems in a single gene, like cystic fibrosis or sickle-cell disease.

Importantly, people who have inherited DNA variations that confer risk do not necessarily develop an illness. In addition to environmental factors that interact with gene activity, other factors, biologically protective and conferring resilience, are thought to be involved in determining whether any individual remains healthy or develops an illness. These moderating factors are still poorly understood.

As for the genetic portion of total risk: Each illness has its own genome-based risk profile, which can now be “mapped” onto the full human genome sequence. So far, investigators have validated 147 genome locations where commonly occurring variations in the DNA sequence slightly raise an individual’s risk for schizophrenia. The search is still in progress; many more risk locations, or “loci,” in the genome are likely to be discovered as more genomes of both affected and unaffected people are sequenced and the sample size of studies grows. A significant but smaller number of commonly occurring genome risk variants have also been validated for bipolar disorder and autism spectrum disorder, as well as several other psychiatric disorders, and the search goes on in those diagnostic categories, too.

Some of the DNA variations that are associated with increased illness risk overlap across diagnostic boundaries—about 50% of those for schizophrenia have also been found to be risk factors in bipolar disorder, although not all of the shared variations have the same significance in the two disorders. Risk factors also overlap for schizophrenia and autism, leading to the hypothesis that some of the same underlying biological processes are disturbed in the two illnesses. This is a hopeful notion, since the discovery of a therapy in one disorder might therefore also help people with a different, but genetically related, diagnosis.


Most of the common DNA variations raising risk for psychiatric illness affect short sequences of DNA, and sometimes only a single DNA letter. It wasn’t until 2007 that researchers discovered that every one of us carries small variations—dozens or hundreds—in our genome, relative to the human “reference” sequence (a representative sample based on multiple individuals of different races and ethnicities). This fact, which surprised scientists, provides a critical clue about how genome variations relate to health. Since all of us have DNA variations, and most of us are not ill, we can safely conclude that most small variations have no major impact on our wellbeing. The key question is: which variations matter? Part of the answer has to do with their location. Some places in the human genome are much more sensitive than others. Among the variations that have healthconsequences are those that prevent genes from doing the job they have evolved to do. Sometimes, a single-letter DNA change can seriously impair the way a gene functions.

More typically, “gene-disrupting variations” involve larger structural flaws in the genome that are random and usually are present from birth, ranging from the deletion of a lengthy DNA sequence that contains one or more essential genes, to large-scale events like the breaking of a chromosome and the reattachment of the fragments to other chromosomes. While these large-scale events are rare, they account for a significant share of people diagnosed with psychiatric illness. Some researchers believe the figure may be 30% or more in autism, for instance.

What are the “key genes” whose disruption might have a causal role in a neurodevelopmental disorder like autism? They might include genes whose function is essential in order to build the brain during the fetal period; or genes whose activation is critical while newly born brain cells are wiring up to form circuits at the dawn of life. It’s relatively easy to imagine how rare, large-scale variations in DNA could impact one or more key genes. What’s not yet clear is how common, small variations can combine to have major impact.


This helps explain why PsychENCODE has set out to comprehensively understand gene regulation in cells of the brain. The project proceeds from the observation that many of the common variants discovered—such as the 147 found so far in schizophrenia—tend to cluster in parts of the genome that are not occupied by genes, but rather, by areas of DNA that regulate genes.

“Regulatory areas,” it turns out, occupy most of the genome. In spatial terms, genes themselves only account for about 2% of the full human sequence. The “other 98%” is composed heavily of regulatory DNA, genetic code that regulates the activity of our genes. What does regulatory DNA do? It influences the timing of when specific genes are active and inactive; it can control processes that block or provide access to the DNA that encodes proteins, and thus it can govern how much of various proteins….are made in which kinds of cells…at different moments of time…in different parts of the body.

To get a picture of gene activity in the human brain, PsychENCODE researchers have assembled a high-quality collection of postmortem brain specimens, representing “neurotypical” individuals as well as people who had been diagnosed with three psychiatric illnesses—schizophrenia, bipolar disorder, and autism.

Over 2,000 brains, harvested and frozen within hours of death, have been assembled from various meticulously curated collections and divided into samples that are shared in research labs spanning the nation. These brains provide snapshots of what is happening in actual tissue from 5 weeks of embryonic life, through the fetal period, and into the time after birth, with a heavy emphasis on infancy, childhood and adolescence, and ending in fully mature brains as old as 64 years post-birth.

As members of PsychENCODE point out in one of their 2018 papers, “Understanding the causes of neuropsychiatric disorders requires knowledge not just of endpoint differences between healthy and diseased brains but also the developmental and cellular contexts in which these differences arise.” Each postmortem brain is an endpoint for a single brain, but when layered genomic assessments have been made of the entire collection, a large set of snapshots can be pieced together to form a kind of movie, showing gene activity and gene regulation in the brain over the lifespan. What the initial studies have revealed is explained in the companion story on the next page.

Written By Peter Tarr

Click here to read the Brain & Behavior Magazine's March 2020 issue