Of Mice and Models

New research shows that neurons across species are not created equal. What does this mean for animal research?

by Jon Bardin • Posted September 23, 2008 01:16 PM


If the key to our cognitive success is functional specificity, then synaptic complexity is the underlying cause.

The history of pharmaceutical trials is littered with cases of drugs that show promise in mice but ultimately fail when tested in humans, often for reasons poorly understood by the scientists who study them. Such failures generally are explained in one of two ways: Either there is a problem with the underlying biological model itself, or there is a problem with the way the success of the model is being measured.

Now a recent breakthrough by the Cambridge neuroscientist and geneticist Seth Grant may provide a third possibility. In a report published in the June 2008 issue of Nature Neuroscience, Grant and his colleagues analyzed synapses in organisms of increasing evolutionary complexity, from single-celled organisms to vertebrates. They found that more advanced organisms also had more complex synapses, allowing neurons to communicate in more complicated ways.

This finding upends the classical model of intelligence, in which the number of neurons, not their complexity, predicts the capacity for greater intelligence and higher-order behaviors. Compounding this, many mouse models are based on incomplete behavioral measures because the true biology of the disease is unknown. Depression in mice, for example, is often measured by how fast a mouse swims.

As more information comes to light about the cross-species molecular complexity of synapses, some of our fundamental assumptions about translational neuroscience may require a new perspective, which Grant believes we can only fully develop once more cross-species research is conducted. "We need to know more about different species, about how these molecular differences manifest themselves in humans versus primates, in different types of primates, in humans versus rodents, before we can truly know how good a model these species are for understanding what underlies human behavior and disease."

Some of our fundamental assumptions about translational neuroscience may require a new perspective.

Perhaps the most fascinating component of Grant's research is that it shows not only that there are differences between invertebrates and vertebrates on a molecular level but also that, as Grant puts it, "the expanded collection of proteins that vertebrates have at their disposal have been specifically utilized by big brains to create regional specification, and a larger behavioral toolbox" than those of less-evolved animals. Grant's team discovered this connection by comparing the variability of different genes across the brain with the evolutionary chronology of those genes. Genes that were more ancient were less variable across the brain, while more recently evolved genes could be attributed to a certain part of the brain. If the key to our cognitive success is functional specificity, then synaptic complexity is the underlying cause.

If human synapses are unlike those of lesser mammals on some essential level, then complex human brain diseases — especially those that involve complicated behaviors — may reflect such species-specific complexity. Though Grant stresses there is currently no direct evidence that demonstrates such differences among vertebrates, he suggests that his research provides a model, driven by evolution, that explains how such differences might exist: "If certain new and evolving genes undergo a mutation and stop working properly, mental illness develops."

For now there is no reason for translational researchers to put down their mice. As Grant points out, we have gained significant knowledge of how the brain works by studying organisms even less advanced than rodents. Yet we may find that it's harder for researchers to model behavioral disorders such as autism, schizophrenia, and Alzheimer's in animals than many had previously thought.