Behavioral genetics: meet molecular biology

During its first 30 years, from roughly 1960 to 1990, the modern discipline of behavioral genetics was based almost entirely on twin and family studies. Those studies made a strong case for the importance of genes in behavior, but the connection always remained loose and statistical. Only in rare cases could a direct connection between a particular gene or set of genes and a particular behavior be made.

In the past decade and a half, all that has changed with the introduction of bioinformatics, genetic engineering and other techniques that allow researchers to measure, analyze and manipulate genetic material rapidly and easily. These techniques have changed the composition of the field of behavioral genetics, engaging the interest of new groups of researchers beyond psychology--molecular biologists, medical doctors and others--who had previously seen behavior as too slippery for biological research.

This shift took place during a time when interest in genetics was exploding. The announcement in 2000 of a completed draft of the human genome--the total complement of genes found in the nucleus of each human cell--and the 50th anniversary of Francis Crick and James Watson's discovery of the structure of DNA in 2003 marked the high points.

Today, expectations of quick rewards from the use of these new techniques are lower than they were during the first flush of excitement. It is now clear that a single gene for complex disorders such as depression is unlikely to exist, let alone be found, even with the most sophisticated methods. Complex behavioral traits, researchers are finding, are influenced by tens if not hundreds of genes, each interacting with the environment and each other in unpredictable ways.

Nonetheless, behavioral genetics continues to hold out the promise of better understanding the biological basis of behavior--hence the field receives strong support from the National Institutes of Health and other grant-making institutions concerned with the intersection of behavior and health.

"There's more and more a proper recognition that you have to understand behavior and genetics and how they work together if you want to understand how people stay healthy or become unhealthy," says John Hewitt, PhD, director of the Institute for Behavioral Genetics at the University of Colorado at Boulder, and chair of the APA Science Directorate's task force on genetics.

New tools and collaborations

The new techniques have not replaced the classic methods in behavioral genetics: twin and family studies that used genetic relatedness to search for genes associated with behavior (see page 46). In fact, twin studies remain one of the best ways of identifying genetic markers linked to complex behavioral traits, according to researchers such as John DeFries, PhD, founder of the journal Behavioral Genetics and former director of the Institute for Behavioral Genetics.

Increasingly, however, such studies are being used not as end-points in themselves, but as stepping stones toward molecular genetics studies that can identify particular genes and their functions, says DeFries.

Ten years ago, before the Human Genome Project and the proliferation of inexpensive genetic tests, a researcher studying a particular behavioral disorder might have had access to tests for three or four genes, says Jonathan Flint, MD, a behavioral geneticist at Oxford University--and the information available about those genes would have been minimal. "Now you click on the Internet and you can find information for the whole genome," he says. Such information is now available not just for the human genome, but also for common laboratory animals such as mice.

This flood of data means that the ability to gather, organize and analyze biological information is becoming increasingly critical. Flint's lab, like many others, has recently hired a bioinformatics specialist to stay up-to-date on methods for mining the gigabytes of data now available.

New techniques are also providing scientists with ways of directly manipulating genes in animals and observing the altered genes' effects on behavior. Mice have proved to be especially amenable to such manipulation. There are now thousands of different strains of single-gene mutants and "knockout" mice--animals in which a single gene has been altered or disabled.

APA genetics task force member Jeanne Wehner, PhD, of the University of Colorado at Boulder, is among those who have studied such knockouts. Although her training is in biochemistry, she and her laboratory do work that is primarily psychological. Using tests of learning and cognition, they look for behavioral differences in strains of genetically manipulated mice.

One knockout-mouse strain they have studied is missing the gene for protein kinase C gamma, a cellular "second-messenger" that communicates between surface receptors and the internal machinery of neurons in the brain and spinal cord.

Like standard mice, these knockouts can be trained to respond to a stimulus in exchange for a reward. However, in experiments where rewards are given for withholding a response--rather than for responding immediately after each stimulus--the mice tend to perform poorly. This, together with their tendency to drink more alcohol than standard mice, is taken as indication of their impulsivity.

Wehner and her colleagues at the University of Colorado are now trying to flesh out the links between protein kinase C gamma and its possible effects on human behaviors such as drug abuse and alcoholism.

One set of studies, led by a member of Wehner's lab, neuroscientist Barbara Bowers, PhD, is examining the effects of protein kinase C within the cell. She is testing the hypothesis that the missing gene affects serotonin receptors, which are known to be involved in emotion and motivation.

Another set of studies, led by Marissa Ehringer, PhD, a human genetics researcher also at the Institute for Behavioral Genetics, is trying to bridge the gap between animals and humans. As part of a larger project on adolescent anti-social behavior, Ehringer is looking for evidence that humans show variation in the gene for protein kinase C gamma and whether that variation has consequences for behavior.

As with much of today's behavioral genetics research, the protein kinase C studies would be impossible without the collaboration of people from a variety of disciplines: the biologists who created the knockout animals, the neuroscientists and psychologists who designed and implemented the animal behavior studies, and the psychologists and medical geneticists who are looking for genetic variation in humans.

Promises and challenges

The proliferation of new techniques has raised expectations of what behavioral genetics can do. But, as many researchers are quick to note, those expectations can sometimes be seriously out of touch with the real promises and challenges of the field.

"The most common misunderstanding--and it's almost willful misunderstanding right now--is that there's going to be a simple answer to a complex question," says Hewitt.

Typically, this takes the form of claims that "the gene" for some complex trait--sexual orientation, for instance, or alcoholism--has been discovered. The media deserve some blame for exaggerating the significance of new research findings, but as Hewitt notes, researchers are not guilt-free: The temptation to play along with the hype in order to increase support for the field is strong.

Recent research is making such a stance increasingly untenable, however. The deeper scientists delve into the genetics of complex behaviors, the more they find that such behaviors are influenced by tens or hundreds of interacting genes, each accounting for only a small portion of the overall variance.

That it is not genes alone, but rather genes in interaction with the environment that produce complex behaviors, is also receiving increasing support, says psychologist Terrie Moffitt, PhD, of the Institute for Psychiatry at King's College London.

Moffitt and her colleagues, for instance, have studied two genes that affect the breakdown and uptake of neurotransmitters in the brain. They have found that the genes have significant effects on depression and antisocial behavior--but only in people who are exposed to particular environmental stressors (see Further Reading below).

Other research is showing that the idea that the heritability of a given trait can be determined once and for all is mistaken. In reality, heritability for complex behavioral traits--the amount of variance in a population accounted for by genetic factors--can vary dramatically within populations (see sidebar).

Even those conducting animal research, which in many ways is easier to interpret than research on humans, have faced challenges. With knockout mice, for instance, developmental psychologists have been quick to point out that removing a gene from an embryonic stem cell and allowing that cell to grow into a genetically modified mouse is not the same as turning the gene off in an otherwise normal adult. The missing gene could have widespread effects on how the organism develops.

In response, says Wehner, geneticists are now producing mice with conditional or inducible knockouts--genes that are inactive only during certain developmental stages, or that can be turned on or off using drugs or changes in environmental conditions. Even so, progress has been slow. Such knockouts are extremely difficult to make, she notes, and they have limitations of their own.

New techniques may help researchers overcome at least some of those challenges. One particularly promising area, says Flint, is the combination of behavioral genetics with visualization tools in biology.

In living animals, including humans, functional MRI and other brain-imaging techniques are providing increasingly high-resolution maps of large-scale neural activity. Meanwhile, in cells, molecular techniques such as tagging enzymes with green fluorescent protein are allowing researchers to watch changes in gene expression as they occur. Researchers are also hoping to make increasingly direct connections between animal models and clinical research, says Hewitt. Right now, a number of interesting candidate genes have been identified in animals, but links to human behavior are sparse. Visualization tools such as those described by Flint may help bridge the gap.

These techniques may bring behavioral geneticists one step closer to their ultimate goal: discovering how neurons--shaped by interactions between genes and the environment--give rise to behavior. "The marrying of those different technologies will enable some of the most exciting science in the next 10 years," says Flint.