Mental illnesses such as bipolar disorder and schizophrenia are very heritable, but the search for culprit genes can seem hopelessly complex. Perhaps dozens contribute to mental illness, and they interact with brain chemistry, brain structures and the environment in a dizzying array of ways. Even more confounding, nearly identical symptoms can result from very different behind-the-scenes malfunctions, researchers say.

"[Mental illnesses] as we currently define them are far too complex and far too multifactorial to understand," says Todd Gould, MD, a research fellow at the National Institute of Mental Health's Laboratory of Molecular Pathophysiology. "We need to break down and decompose these disorders into parts that can be more tractable."

In the past, scientists tried to link genes to Diagnostic and Statistical Manual-defined categories, but psychologists and geneticists are now collaborating to make finer-grained distinctions, says Irving Gottesman, PhD, a University of Minnesota behavioral geneticist and psychology professor. Instead of searching for depression genes, for example, researchers are looking for genes that contribute to the emotional regulation problems that may underlie the disorder.

Shining a light on these hidden mechanisms--known as endophenotypes--is a natural fit for psychologists says Gottesman, who introduced the term in the 1970s, along with his colleague James Shields.

"Many of the features that are now being focused on are exactly the things psychologists have been studying, but in a different context," he says. "We studied them for their own value without worrying whether they would be predictors or precursors for mental illness."

From normal variation to mental illness

This approach, though young, has already spotlighted a number of backstage players in schizophrenia. One team of researchers, headed by Tyrone Cannon, PhD, a psychology professor at the University of California, Los Angeles, recently linked a gene to working memory deficits, which in turn may contribute to schizophrenia.

The researchers performed blood and memory tests on 118 adult monozygotic and dizygotic twins and found that a sequence of base pairs, "TCG," occurs at the gene known as DISC1 in about 75 percent of the participants with schizophrenia and 64 percent of healthy participants. Even healthy individuals with the gene, however, tended to score lower on tests of working memory than those with a less common variant, according to a study published in a 2005 issue of the Archives of General Psychiatry (Vol. 62, No. 11, pages 1,205-1,213).

What's more, brain imaging showed that participants with the TCG sequence tended to have relatively low gray matter density in their prefrontal cortex--an area related to working memory. The results dovetail with past research that linked the gene to decreased connections between neurons, Cannon notes.

Although these differences seem dire, it's worth noting that most people carry the variant but only a few develop schizophrenia. Therefore it makes sense to first link the gene to commonly occurring mental phenomena--like slight working memory deficits--he says. Keeping an eye out for hidden pathology is what separates the endophenotype approach from past attempts to identify genes for mental illness, says Cannon.

"The key is, you are assessing not just the risk for the illness phenotype, but you are assessing for the endophenotype as well--whether that be a personality trait, a score on a cognitive test or a brain scan," says Cannon.

Studies like Cannon's are building up a vocabulary of genes and their biological effects, which may interact to lead to schizophrenia, says Gould. By doing this, scientists may someday be able to diagnose which factors are at play in any given individual with the disorder and tailor treatment accordingly.

The flexible brain

Though researchers have established a handful of potential culprit genes and their intermediate effects for schizophrenia--and more are on the way--the search for endophenotypes in mood disorders is only just beginning, says Cannon.

One potential candidate is the serotonin-transporter gene--known as 5-HTT--which gives instructions to produce a molecule that moves serotonin out of the synapse. A variant of the gene, which is shorter than the more common version, seems to increase the rate of production of the molecule, moving serotonin too quickly and potentially leading to symptoms of depression. However, like the DISC1 variant, the 5-HTT mutation is much more common than full-blown mental illness--many people have the variant of the gene without developing depression, says Stephen Manuck, PhD, a psychology professor at the University of Pittsburgh. So he devised a study to look at what environmental factors might also come into play.

Manuck and his collaborators recruited 139 people, ages 24 to 60, with varying socioeconomic status, as measured by income and education level--and used blood samples or cheek swabs to determine what variant of 5-HTT they carried. They then administered a dose of fenfluramine--a serotonin-releasing agent--to increase the serotonin in the participants' brains, and gauged their sensitivity to the neurotransmitter by measuring the amount of prolactin in the participants' blood. (The hormone is released in response to increased serotonin levels.)

The researchers found that people with the short version of the gene showed lower prolactin levels, suggesting decreased sensitivity to serotonin--but the effect was only found among people with low socioeconomic status. Those who did not face the challenges of poverty showed a normal response regardless of what variation of the gene they had, according to the study, published in Psychoneuroendocrinology (Vol. 29, No. 5, pages 651-668.)

The finding suggests that the short version of 5-HTT may lead to the endophenotype of increased sensitivity to one's environment, says Manuck. If that environment is a challenging one, the pathway can be in the direction of depression--though the gene may not code specifically for any feature of the disorder, he notes.

Studies such as this show how teasing out the middle steps from genes to disorder can help researchers track the complex ways that the environment affects gene expression, Gottesman says. A better understanding of how mental illness unfolds may eventually help practitioners develop early interventions, he says.

"We are sneaking up on these genes, and we are doing it by entering a program of research I call optimal reductionism," he says. "You can get so much more mileage out of observing people not at the obvious level of their diagnosis...but by asking what are the precursors of those symptoms."