Harvard University - Department of Molecular & Cellular Biology


by Charlie Schmidt

November 1st, 2006

Naoshige Uchida
Of all the systems in biology, the mind could be the hardest to conceptualize. How can we understand the way neurons produce the mind, with its powers of perception and decision-making? While most of us ponder that question from time to time, stumped by the splendid mystery of it all, Naoshige Uchida, is trying to answer it. A neuroscientist and new Assistant Professor in MCB, Uchida studies the neural circuits that link sensory organs with decision centers in the brain. Surrounded by packing boxes at his new office in the Harvard Biolabs, he explains that his objective is to find out how the brain codes and uses sensory information to trigger thought, action, and learning. "And by looking at those processes, we may at some point get to where the mind starts to emerge,"

Uchida approaches his goal from the bottom up--by tracing the circuits that drive simple decisions in rodents. Sporting cranial electrodes to record neural activity, the animals generate data that ideally could pinpoint where decisions occur in the brain. The decisions in this case are imposed on the rodents. In his experiments, Uchida uses a device that forces the animals to think about how to get a drink of water. That device, of his own invention, contains three ports: a center port that delivers selected odors to an animal's probing nose, and a pair of adjacent ports that provide water drops. Specific odors direct animals to either the left or right ports for sustenance; to drink, they need to match odors with their associated water-ports correctly. If they choose incorrectly, they come up dry. Uchida admits the technique measures simple processes. Scientists know relatively little about the circuitry of thought, he says, so these types of studies pave the way for more advanced research.   "We're really doing the basic work," he explains. "It could have been done before but it wasn't. This is a very good system for measuring behavior and brain activity. It gives us a simple, binary output that we can easily study."

From Kyoto to Cold Spring Laboratory

Uchida's fascination with the mind stems from his days as a young student in Japan, where he was born and raised. Curious about physiology of thought, he studied biology at Kyoto University, assuming that comparative studies with different species, and with animals in different stages of development, might yield clues to the mind. Uchida stayed at Kyoto University for nearly a decade, completing a BA in science in 1992, an MS in developmental biology in 1994, and a PhD in biology in 1997. For his graduate research, he studied with Masatoshi Takeichi, a specialist in tissue patterning. Dr. Takeichi was famous for revealing how proteins called cadherins regulate cell-to-cell adhesion during the formation of functional cell groups. Uchida's work built on those findings by showing that cadherins also play key roles in neuronal recognition and organization.

That work ultimately launched a second postdoc, this time with Kensaku Mori, at the RIKEN Brain Science Institute. With Mori, Uchida turned his attention to brain physiology, investigating how odorant molecules effect neuronal activity in a region, called the olfactory bulb, that receives inputs from receptors in the nose. That work showed that specific odors evoke corresponding patterns of spatial activity in anesthetized animals. But while valuable, the findings were limited by their experimental approach. Because he was studying drugged animals, Uchida was missing a behavioral component that also connects scent and neuronal activity patterns. In other words, the "thoughts" linking scents to actions were absent from his analysis.

A rat performing an odor discrimination task (Uchida and Mainen, 2003).
Hoping to fill that gap, Uchida came to the United States in 2000 for a fellowship with Zachary F. Mainen, at Cold Spring Harbor Laboratory. There, he developed the "psychophysical assay," with its three ports, that he still uses today. The assay, inspired by a similar method used in primates, has since been adopted by other researchers in the field; Uchida uses it to study how scent drives decision processes in rodents, while others use sounds or visual representations as sources of sensory information. Uchida says rodents make optimal subjects for his research: they're cheap, naturally inquisitive, and they learn the assay quickly, with just 5-6 days of training. What's more, unlike most primate studies, which record the activity of just one neuron at a time, Uchida uses electrode arrays that allow him to measure the activity of many neurons simultaneously. "[The arrays make it possible to] record interactions among different regions of the brain," he explains. "We can see how single neurons, or groups of neurons, talk to each other during decision-making."  

Having just left Cold Spring Laboratory, Uchida says he now plans to focus his MCB research on the dynamic learning processes that drive decision-making. Learning, he explains, requires animals to update action strategies according to the outcomes of prior experiences. Uchida modifies those outcomes in the laboratory by changing the amounts, probability, or timing of rewards (i.e., drinks of water). These simple manipulations give a degree of experiential texture to the learning process and allow Uchida to study how and where outcome recollections are stored in the brain. But getting to the larger goal--namely, the wiring of the mind itself--requires more resolved experimental methods, he acknowledges. Perception, which guides the mind's view of the world, is incredibly complex and multi-dimensional. Uchida's work helps lay the groundwork for studies that will likely go on for generations.

In the meantime, his research could yield timely clinical advances. Patients with schizophrenia and addiction disorders typically make impulsive choices, possibly because they suffer from defects in decision-making mechanisms, Uchida explains.   "We may be able to model these impulsive choices in rodent systems to study the causes of these diseases," he says. "We think some of these [conditions] are regulated by information flows in certain parts of the brain. By looking at the dynamics of many neurons 'talking' with each other we might be able to provide insights into the disease's clinical aspects--that's the long-term goal."

Uchida says his upcoming research will focus on three key questions: How populations of neurons code sensory information in the brain; how sensory information triggers action for decision-making; and how animals learn to update decision processes from prior experience. These questions give form to a broader agenda that seeks to establish causal links between the activity of specific neuronal circuits and behavior and learning dynamics. The behavioral techniques and electrophysiological tools behind this research are ready to go, he says. Upcoming challenges will be to somehow integrate those methods with emerging molecular tools that up- or down-regulate specific neurons. That capacity, he says, will enable more controlled hypothesis testing. Assay observations can't confirm hypotheses in isolation, Uchida stresses. Molecular tools will ideally add a new dimension to his research that reinforces experimental findings.

Looking forward, Uchida says he's pleased and excited to be at Harvard, and anticipates numerous collaborations with other MCB scientists. Among them are Catherine Dulac, who develops genetic and molecular technologies in mice, and Markus Meister, who contributes critical experience in the analysis of large data sets. Harvard's new Center for Brain Science, where he will also be Assistant Professor, figures prominently in Uchida's future plans. These combined resources could accelerate his contributions to the field, and help create a foundation from which ever greater understandings of the mind emerge.

View Nao Uchida's Faculty Profile