Fast and Slow: Learning How the Brain Controls Movement
Posted on May 3rd, 2016
What if you couldn’t move faster even when you wanted to? Researchers thought that the part of the brain that determines how fast we perform voluntary movements, such as walking across a room or playing a melody on the piano, was a bit like a car. It has an accelerator to make movements faster and a brake to slow them down. Now, scientists at the Howard Hughes Medical Institute’s Janelia Research Campus have shown that, contrary to what was thought, the “brake” in this part of the brain can actually accelerate movements in mice, and the gas can rein them in. By determining how the brain controls movement, this discovery helps to explain the systematic slowing of movement in patients with Parkinson’s disease and could pave the way for interventions that allow patients to learn to perform everyday actions more fluidly.
Walking a little faster is no problem for most people, but patients with Parkinson’s disease struggle to accelerate voluntary movements. We have assumed for some time that “it’s almost as if only the brake works and the gas pedal doesn’t work,” says Janelia group leader Joshua Dudman. To better understand this effect, he and his colleague, research scientist Eric Yttri, wanted to find out more about the normal role of the basal ganglia, a brain region that is affected in Parkinson’s disease, in controlling voluntary movement. Within the basal ganglia, there are two main types of neurons known to promote (gas) or suppress (brake) movement.
In experiments described in an advance online publication May 2, 2016 in the journal Nature, Yttri and Dudman used a technique known as optogenetics to activate neurons in the basal ganglia during movements at specific speeds. By shining a laser through fine optical fibers that extend into the animals’ brains, the researchers could selectively stimulate either the gas or the brake neurons to ask how each group influenced future movement.
Yttri trained mice to move a small joystick with their front paws in order to get a sweet drink. The joystick was rigged such that a mouse has to make a choice to satisfy its thirst. The rodent has to push the joystick fast enough to obtain a drink of water, but if it pushes too rapidly it is wasting energy and ultimately limiting the total water it can consume. Every day, people make similar, albeit implicit, decisions about how rapidly they must act – deciding how fast to walk to the neighborhood restaurant on a lunch break. However, in Parkinsonian patients (and as Dudman and colleagues showed previously, Parkinsonian mice) all movements are slowed.
To gauge how forcefully a mouse was pushing, the researchers measured the speed of the joystick. On average, a mouse’s joystick movements take about half a second to complete. Dudman and Yttri first tested the effect of adding extra activity in either group of neurons during specific movements. If the push was predicted to be a swift one based upon its initial speed, the device rapidly activated one or the other group of neurons in the basal ganglia. With this procedure, the researchers could spur the mice to push the joystick systematically faster or slower on future movements, depending on which population of neurons the researchers activated.
Those results are consistent with the long-standing idea that separate populations of neurons in the basal ganglia serve as brake and gas pedal for movement. To determine whether these neurons always had the same effect on movement, the researchers asked what would happen if they activated the neurons when a mouse made a slow movement of the joystick. In this case, switching on the “gas pedal” neurons didn’t accelerate the animals’ movements. Now stimulation systematically slowed future movements. Dudman and Yttri saw a similarly reversed outcome when they triggered the “brake” neurons at the beginning of a slow push. The rodents surprisingly started to move the joystick systematically faster.
Dudman explains, “either one can speed you up or slow you down.” In other words, by showing that releasing the brake can speed movements and releasing the gas pedal can slow movements, the study suggests that we are using a combination of both pathways to regulate movement speed. To visualize how this system adjusts how we move, Dudman says, think of a racecar driver zipping around a track. Instead of either speeding up or slowing down, a driver uses both the gas and brake together to make controlled, but fast turns.
The researchers asked whether this control system could be what is disrupted in Parkinson’s disease. In patients with Parkinson’s, the cells that make a chemical messenger called dopamine die off. To simulate the loss of these cells in the mice, the researchers injected the animals with a compound that blocks dopamine receptors on neurons – mimicking an absence of dopamine. The stimulation that was previously sufficient to change the speed of movement now had no effect.
In addition to clarifying how the basal ganglia controls movements, these results have significant implications for treatment of Parkinson’s disease. Many patients already have implantable devices (deep brain stimulators) that provide electrical stimulation to the brain to improve movement. By selectively activating stimulation during specific movements, similar to what the mice received, such devices might allow patients to access to a normal range of movement speeds.
Learning How the Brain Recovers from Disruptions
Posted on April 14th, 2016
New research from scientists at the Howard Hughes Medical Institute’s Janelia Research Campus suggests that the brain is organized into modules that work together to maintain critical functions, even in the face of disturbances.
This structural organization may explain how neurons that store short-term memories can recover from significant disruptions—for example, enabling a quarterback to remember a planned play despite the distractions he encounters before he throws the football. According to the Janelia group leaders who led the study, experimental neuroscientist Karel Svoboda and theoretical neuroscientist Shaul Druckmann, this motif is likely to underlie other essential circuitry within the brain as well.
“This is how an engineer would build a mission-critical system,” says Svoboda. “You distribute the critical systems over multiple modules, and then the modules talk to each other so they can sense when one of them isn’t doing well and can correct for each other.”
The study, reported April 13, 2016 in an advance online publication in Nature, began with a surprising observation by postdoctoral fellow Nuo Li in Svoboda’s lab. The team had been studying neurons in mice that are involved in planning motor activity. After an animal is instructed by researchers to move in a particular way, groups of these neurons become active, signaling for several seconds until the animal completes the movement—a form of memory that outlasts the milliseconds that any single neuron can signal on its own. Scientists knew that an animal’s motor plan can persist even if this signaling is temporarily disrupted. Svoboda and his colleagues wanted to find out just how robust the underlying neural activity was.
To find out, they used a laser to switch off motor-planning neurons briefly before the animals in their experiments were allowed to complete a task. Monitoring the subsequent activity, they found that once the neurons were allowed to resume normal signaling, they quickly adjusted their activity to make up for the lost time. The mice carried on as if undisturbed, remembering which way they had been instructed to move and successfully completed their task. “We quenched [neural] activity to zero and saw that it came back to exactly the levels where it should have been,” Svoboda says. “It was a perfect—almost eerily perfect—example of robustness.”
Theoretical neuroscientists have modeled several ways in which neural circuits can establish robustness, so Svoboda shared his data with Druckmann and postdoctoral fellow Kayvon Daie, seeking an explanation for how the motor-planning neurons were able to recover so completely. “There is a rich history of models that have been suggested for such systems,” Druckmann says “But when we tried to compare experimental results to the models, we found that none of them show such strong robustness.”
“The fact that it recovers to exactly where it would have been had you not shut down the system is where all the models go wrong,” Druckmann says. Existing models showed neural activity picking up where it had left off after a disruption, so that the pause introduces a persistent delay in the normal activity pattern and activity remains slightly displaced from where it should be. “But what you see in the experiment is the exact opposite,” Druckmann says. “Somehow the activity catches up to where it needs to be.”
Something was missing. So, Svoboda says, “we went back to the biology to give us hints about how to construct the next level of models.”
Knowing that no group of neurons works in isolation, the researchers began to wonder if another brain area might have compensated for the disruption Svoboda’s team had introduced in their experiments. “The simplest thing that could be is that maybe this brain area is just looking at what another brain area is doing and copying it,” Druckmann says.
His modeling suggested the situation was slightly more complex, and that neuronal activity might return to where it should be after a disruption if the cells were in communication with another brain area carrying out the same function. “It’s like two roller coasters running in parallel and connected with a big rubber band,” Druckmann explains: If one of the coasters falls off the track, the other keeps going, and the rubber band eventually snaps the wayward coaster back where it should be.
Such an organization would explain the striking robustness observed in the experiments. “Once we had the right architectural principles, all of the preexisting models could be rescued,” Druckmann says. “We realized that we need two modules, they need to be redundant in the sense that each of them can independently generate the right dynamics, and they need to be connected. Once we rewrote them according to these principles, all of the models worked.”
The scientists devised a series of experiments in which they tested their model by disrupting motor-planning circuits on opposite sides of the brain, both separately and together. As expected, when they interrupted signaling in one region at a time, neurons recovered well. But when they disrupted both motor-planning regions at the same time, recovery was impaired and animals performed their task poorly. It looked as if maintaining the motor plan did indeed depend on at least one of the modules operating undisturbed.
In a final round of experiments, the researchers disconnected the two motor-planning regions from one another, then blocked signaling on one side. Although one module remained undisturbed, the disrupted module was unable to recover, supporting the idea that two modules must be in communication to correct for lapses in activity.
The scientists suspect their new model may explain the robustness of neural circuits beyond those they tested in the current study. “We think that this modularity probably happens in many incarnations,” Svoboda says. “From circuit analysis, we know that the right kind of circuit elements are there.”
Research Shows that Cortex Commands the Performance of Skilled Movement
Posted on February 1st, 2016
The essential function of the central nervous system is to coordinate movement. Skilled movement requires many elements including planning, initiation, execution, and refinement. Scientists at the Howard Hughes Medical Institute’s Janelia Research Campus have demonstrated in experiments with mice that the brain’s sensorimotor cortex is essential for initiating and executing a complex movement. Temporarily switching off this part of the brain causes mice to pause abruptly in the middle of a trained task, which they successfully resume as soon as motor cortex activity is restored.
It has been unclear whether the motor cortex is the brain’s primary director of voluntary movements, or whether its role is to fine-tune complex movements. The new study, led by Janelia group leader Adam Hantman and published February 1, 2016 in the journal eLife,demonstrates that activity in this part of the brain is critical for enacting a learned skill.
Neurons in the motor cortex fire as animals plan and carry out movements, but the precise effects of this activity have been unclear. Although artificially stimulating the cortex can evoke complex movements in animals, other researchers have found that mice with irreversible damage to the motor cortex retain the ability to complete a learned sequence of actions. Those animals do tend to lose dexterity, adapting by swinging the limb instead of carefully maneuvering the paw to the target, for example.
Hantman and colleagues thought they could learn more about the precise role of the motor cortex by manipulating it with optogenetics, a technique that allows precise temporal control over neural activity. The technique involves delivering light-sensitive proteins to specific nerve cells under study and then using a pulse of laser light of the right wavelength to turn on or shut off the activity of those nerve cells. Using this approach, Hantman and his team could inhibit the motor cortex at specific moments while the mice performed a task. Shutting down motor cortex activity for just a few seconds at a time would also make it unlikely that the mice would learn to compensate for any impairments in motor control, a potential confounding factor in prior attempts to probe the role of motor cortex.
Jian-Zhong (Jay) Guo, a senior scientist in Hantman’s lab, spent months patiently training mice to reach and grab for a food pellet, then bring the pellet to the mouth for consumption. The scientists then used optogenetics to manipulate activity in the animal’s motor cortex during the task. After the animals had begun reaching for the food, Guo used a laser beam to activate inhibitory neurons in the region of the motor cortex that controlled the action. The inhibitory neurons, in turn, switched off nearby excitatory neurons. “We were essentially using the cortex to turn itself off,” Hantman explains. Doing so, they thought, might make the mice struggle to succeed at the task.
In fact, the effects were much more dramatic. At any point after a mouse began to reach for its food pellet with its right paw, Hantman and his colleagues could abruptly halt the action by switching on the laser that silenced the left motor cortex. “This was a breathtaking thing when we saw it happen in the lab,” Hantman says. “As long as the laser was on, the animal was not able to progress its paw forward. It was as if we had achieved a remote control of the mouse.”
In some instances, the scientists inhibited the motor cortex after a mouse had grabbed its pellet, then watched as the animal used its other paw (controlled by the unmanipulated cortical hemisphere) to guide the stalled limb to its mouth. “He just couldn’t get that arm up to eat,” Hantman says. “He knew he had a pellet of food, and he so desperately wanted it, but he simply couldn’t get it into his mouth.”
In all of their experiments, Hantman’s team recorded the animals’ behavior and used new machine learning algorithms developed by Janelia group leader Kristin Branson and colleagues to track their movements efficiently. Automating this analysis allowed the team to test the effects of inhibiting the motor cortex in a variety of situations. But when they examined different behaviors, such as licking for food or grooming, mice continued their activities uninterrupted, despite motor cortex inhibition. The scientists concluded that the pausing effect was specific to a learned, complex, goal-motivated movement toward a target.
Remarkably, in the food-grabbing task, mice would pick up the action where they had left off as soon as the scientists switched off the laser, allowing normal activity in the motor cortex to resume. In fact, the scientists found that a hungry mouse familiar with the task would reach for a food pellet immediately after normal motor cortex activity was restored, even if no pellet was present and the action had not previously been initiated. “Just the mere inhibition and release of that inhibition seemed to evoke the entire sequence of the movement,” Hantman says. Despite the lack of an actual pellet, and regardless of the starting position of the reaching paw, the animals reliably reached toward the place they had learned the pellet should be. “That suggested to us that perhaps the cortex was specifying not the kinematics of the movement, but something about that learned end-point location,” Hantman says.
Hantman and his team are now eager to investigate how the motor cortex retains information about where a movement should be directed. “Some of the things that we noticed during these perturbations give us entry points to try and uncover the dynamics of the cortex that might be responsible for this kind of skilled action,” he says.
A New Platform for Brain-Wide Imaging and Reconstruction of Neurons
Posted on January 21st, 2016
The fine, branching tendrils that extend from a neuron often traverse great distances to form connections with their target cells. Tracing those connections is a major goal for scientists who want to understand exactly how the brain processes information, but technical obstacles have hindered the effort. Now, scientists at the Howard Hughes Medical Institute’s Janelia Research Campus have traced the complete paths of several neurons through the entire brain of a mouse, using technology they say can be scaled up to enable a larger mapping effort.
A multidisciplinary project team of Janelia scientists reports the development of a new imaging platform that combines a high-speed, high-resolution light microscope with new sample preparation methods and powerful image processing software to efficiently map neuronal projections cell by cell. They published their research online on January 20, 2016 in the journal eLife. “Being able to identify the targets of single neurons has been very difficult to do in the past,” says Michael Economo, a research scientist at Janelia who is the first author of the publication. “This microscope should really make that process more efficient.”
The publication includes reconstructions of five neurons that spread across the mouse brain in complex and diverse patterns, representing a milestone for Janelia’s MouseLight Project team, which aims to generate maps of neurons’ projections throughout the mouse brain.
“The goal of the project is to see where individual neurons in the brain send their messages, and to do this at scale,” says project leader Jayaram Chandrashekar. Ultimately, he says, the MouseLight team aims to trace the projections of thousands of neurons, a small percentage of the 70 million neurons in the mouse brain, but a huge increase over the handful of projections that have been reconstructed to date. “To do this,” he says, “we need to be able to follow individual neurons right from their cell body all the way to their termini, so we need very good resolution, and we need speed. The current technology lets us do both of these things.”
Although scientists are beginning to map out how different areas of the mouse brain connect to one another, they have little understanding of the cell-to-cell variation that exists within these broader connections. “There’s a major mystery about how many projection types there are, and this problem has been addressed in a very piecemeal manner,” says Karel Svoboda, a Janelia group leader who helped guide the development of the MouseLight project.
By tracing the projections of a significant percentage of the cells in the brain, the team expects to learn a lot about how the nervous system processes information. “A neuron’s projection field is complex but highly specific, and it essentially defines where that neuron sends information,” Svoboda says. “So knowing a neuron’s shape in great detail tells us about how signals are routed in the brain.”
The neuron-mapping effort began in the lab of former Janelia group leader Gene Myers, a computer scientist who is now at the Max Planck Institute of Molecular Cell Biology and Genetics. In 2010, Nathan Clack, a scientist in Myers’s lab (now at Vidrio Technologies), set out to apply imaging analysis methods being developed in Myers’s lab to the problem of tracing neurons in the mouse brain.
“It’s very difficult to take a picture of an entire neuron,” Clack says. “The axon — the business end of the neuron — is like a thin wire, as little as 80 nanometers in diameter, that travels for centimeters.” To capture a neuron’s slender extensions from beginning to end, he says “you really have to think about imaging the entire brain.”
Clack’s first priority, then, was to optimize a high-resolution microscope for the task. Working with Janelia’s Instrument Design & Fabrication team, he adapted a two-photon microscope—a technology that lets scientists image clearly beneath the surface of a tissue—to speed up its imaging rate dramatically. Even with the improved speed, the microscope would still take more than a week to image a complete mouse brain, so Clack developed control software to fully automate the process.
Clack then teamed with Chandrashekar and Economo to work out a method of readying a mouse brain for imaging. Brain tissue is clouded with fats that obscure the view beneath the surface, so unless these fats are removed, the tissue must be sliced into ultrathin sections to obtain high-resolution images. That process that can distort tissue, making it difficult to reconstruct images of structures that pass through many sections. Existing methods of removing fats were not compatible with a prolonged imaging period, so Janelia group leader Luke Lavis, a chemist, applied his expertise to the problem. Ultimately, the team developed a mild clearing strategy that allows the microscope to image clearly to about 250 microns beneath the surface of a sample. “This approach preserves the fluorescence [in labeled neurons] and gives us really high quality data,” Clack says.
To generate the images they would use to trace neuronal projections, Economo and his colleagues applied their clearing protocol to a mouse brain in which a small number of neurons had been brightly labeled with a fluorescent dye. They inserted this brain into the microscope and, after some initial set-up, automated imaging began.
To image a complete brain, the microscope collects about 20,000 separate but overlapping blocks of images, imaging as deep as it can, then slicing away a layer of tissue and repeating the process. The entire brain is imaged in about 100 slices, a process that takes about 7-10 days and generates tens of terabytes of data. Taking advantage of methods developed by Janelia’s scientific computing group to manage the large volume of data, the 20,000 image blocks are ultimately computationally stitched together to form a complete three-dimensional representation of the brain.
The last step, tracing the paths of the fluorescently labeled neurons to generate final reconstructions of each cell’s projections, was done manually, demanding 40 to 60 hours of work per cell. The team is now exploring ways to dramatically reduce the time that must be devoted to this task, working to improve cell labeling and computational analysis so that tracing can be done mostly automatically. “This step will never be 100 percent automated, but we believe it can be brought down to a few hours,” Chandrashekar says. That will be essential as the team works toward using their platform to complete about one neuronal reconstruction per day.