Can we predict accurately what people perceive by studying their neural activity? Treue and colleagues take a step in this direction by using the response profiles of neurons in a motion-sensitive area of the monkey brain, area MT, to predict how humans will perceive moving stimuli.
The authors' stimuli consisted of patterns of moving dots, in which some dots move in one direction while others move in a different direction. Normally, we (and almost certainly monkeys) perceive these patterns as 'transparent motion', such that the two sets of dots appear to be sliding past each other, as if on two moving panes of glass.
Each pattern causes large numbers of neurons to fire in area MT, and the firing rates of the cells can be plotted on a graph, giving a shape that looks like the famous bell curve, with a single peak in the middle. The puzzle is how we perceive two directions of movement simultaneously; do two patterns of dots cause a graph with two peaks?
By making careful recordings from monkey MT, Treue et al concluded that the 'twin peaks' explanation was unlikely; our perceptual abilities are much better than this explanation would predict. Instead, the authors reasoned, there must be something more subtle about the exact shape of the graph which our brains can recognize. If so, different stimuli that cause the same shape of graph should be perceptually indistinguishable. The authors tested this idea in human volunteers, comparing patterns containing two sets of dots with other patterns containing three sets. Remarkably, their subjects were unable to detect the fact that the direction of movement was different in the two types of stimuli—just as predicted based on the 'exact shape' hypothesis.
The next question (a challenge for mathematical theorists) will be how the brain can decode the shape of the graph to infer the two directions of motion. In the meantime, the fact that we can predict a novel and unexpected visual illusion from studies of brain activity is a sign of how far we have come in understanding the physiological basis of perception.
Jennifer Groh of Dartmouth College discusses this work in an accompanying News & Views. Also, in this month's editorial, Charles Jennings discusses how Treue's work argues against the claim by science writer John Horgan that neuroscientists are incapable of answering fundamental questions about how the brain works.
Over the past few years, an entire industry has arisen around the belief that an enriched environment promotes the development of babies' brains. Although some of the claims are certainly exaggerated, it is well known from animal studies that an enriched environment can promote changes in both behavior and brain structure. How this happens is still very unclear, but a new study by Joe Tsien and colleagues at Princeton University uses a combination of genetics, behavior and electron microscopy to look at the possible basis for this effect.
Tsien, whose recent study on 'intelligent' mice made international headlines, had previously developed a sophisticated genetic trick to knock out genes in a specific part of the brain, area CA1 of the hippocampus, which has been implicated in learning and memory. He has used this method to study the function of a neurotransmitter receptor called the NMDA receptor, which plays a central role in the synaptic changes thought to underlie many types of memory. Mice whose hippocampus lacks NMDA receptors are known to have difficulty finding their way through a maze, and the authors now show that they also have poor memory for nonspatial stimuli such as smells and unfamiliar objects.
Is this defect genetically hardwired, or can it be overcome given the right environment? Tsien and colleagues tested this by allowing their mutant mice to live in what amounted to a rodent playpen, filled with toys that were changed every few days to stimulate their exploratory tendencies. After two months of exposure to this environment, the mutant mice had overcome their deficits and performed as well as their normal littermates on almost every memory task in which they were tested.
The results suggest that something happens in the brain that allows it to learn even without NMDA receptors. To find out what, the authors used electron microscopy to look at the fine structure of the hippocampus. They found that the enriched environment led to the formation of more dendritic spines, the tiny structures on which synapses are formed. This presumably signifies a denser network of neural connections, and it is tempting to think that this increased density somehow allows the mice to learn even in the face of a genetic defect that would otherwise prevent learning. This may be hard to prove, as it is difficult to be sure that changes in the hippocampus (rather than elsewhere in the brain) are responsible for the improved performance.
Whatever the answer, the results make it clear that memory deficits need not be fixed but can be overcome given the right environment. As with any such study, one must be cautious about extrapolation to humans. Nevertheless, the 'zero-to-three' lobby might take note that the effects of environmental enrichment are not confined to early development; the mice were 1.5 to 2 months old before they were transferred to the enriched cages, which in human terms would correspond to early adulthood.
Howard Eichenbaum and Kristen Harris of Boston University discuss these findings in an accompanying News & Views article.
In 1997, flashing lights in an episode of the Pokemon cartoon triggered seizures in 685 Japanese children, drawing public attention to photosensitive epilepsy, a common disorder with a prevalence of 0.5-0.8% between ages 4 and 14. The origin of the condition is unclear (although inheritance plays some role), and the underlying brain dysfunction has been poorly understood. Now Vittorio Porciatti and colleagues (University of Pisa) suggest that the problem arises from a defect in the brain's ability to control its responses to visual stimulation.
Our brains must cope with scenes that vary widely in contrast—think of the difference between bright sunshine and fog—and so the visual system has a contrast control mechanism, roughly analogous to the contrast adjustment button on a television. A person's sensitivity to contrast can be measured by recording electrical activity at the scalp in response to flickering visual stimuli. Porciatti and colleagues now report that patients with photosensitive epilepsy showed abnormal cortical activity in such tests. In normal subjects, activity recorded from scalp electrodes increases up to a luminance (black-white) contrast of 20%, and then the response levels off. In patients, however, the response continued to increase up to 90% contrast, becoming twice as high as the maximum in normal people. This abnormal response was found at frequencies of 4-10 Hz, the range known to trigger seizures, but not at higher frequencies. The authors conclude that photosensitive epilepsy may result from severely impaired contrast gain control in visual cortex. The prevalence of this condition is increasing, apparently as a result of increased use of TV and video games, and the authors suggest that a better understanding of the condition may allow the design of safer video graphics.
How does an animal know its own location? The key seems to be the so-called 'place cells' of the hippocampus, which fire strongly when an animal is in a particular place, and less so when it moves to a new location.
How is the activity of place cells controlled? Not surprisingly, visual landmarks are important, but the story must be more complex than this, because place cells can fire even in darkness. Apparently, animals can use information about self-motion, which comes from the vestibular sensors of the inner ear (which detect head movement). As the animal wanders around the darkened cage, the place cells must take each new turn into account to update their representation of the animal's location. Current thinking is that visual and vestibular signals can each drive place cells, but that the two systems must be constantly recalibrated to make sure that they remain in agreement.
Now, in one of the more unusual experiments in the history of neuroscience, James Knierim, Bruce McNaughton and Gina Poe at the University of Arizona and the University of Texas-Houston Medical School have asked what happens to this system in microgravity, when the vestibular signals are disrupted. Their experiment was performed aboard the space shuttle Columbia during the 1998 Neurolab mission. Rats were trained to walk around a three-dimensional track (an 'Escher staircase') in which they could make three 90-degree turns to complete the track, as opposed to four that would be required on a flat surface. (Imagine flying from the north pole to the equator, then turning 90 degrees left and flying a quarter of the way around the earth before turning left again and flying north back to the pole.) In the weightless environment of space, the vestibular system is confused; leftward and rightward turns are detected normally, but movement signals for the other axes are disrupted by the absence of gravity, and therefore cues about self-motion and landmark cues should be conflicting. Can place cells still respond under such conditions?
The researchers demonstrated that the hippocampus was still able to generate a stable place map even when vestibular signals and landmark cues were decoupled. Using a battery of electrodes implanted into the hippocampus, they recorded place cell activity as the rats explored a flat surface on the ground, before the flight. The rats were then launched into orbit, and further recordings were made as they explored their new microgravity environment. By the ninth day in space, place cells were signaling the rat's position as accurately as they did on earth.
The technical obstacles to doing such experiments during space flight are enormous, and to obtain even a simple result with a few animals is a remarkable achievement. Of course, much more work would be required to understand in any detail how place fields can emerge in microgravity. The answers would be of considerable scientific interest, not least because of their relevance to the problem of disorientation in astronauts. Whether they justify the massive expense involved of course is debatable, and the prospect for follow-up studies will depend in large part on the fate of the international space station.
