The brains behind the Brain Science Institute
One day two years ago, a Johns Hopkins neuro-oncologist named Jaishri O'Neill Blakeley was conversing with a colleague about her research on tumors of the nervous system. The tumors are called plexiform neurofibromas, and because they grow on nerves, they wreak havoc—deformation, physical impairment, pain. They begin to appear early in childhood, and as if they were not bad enough in their benign form, they can develop into a particularly nasty form of cancer known as malignant peripheral nerve sheath tumors. Whether the neurofibromas are benign or malignant, there is no good cure—no surgery to excise them without substantial collateral damage, no drugs to kill the cancer.
The other party in Blakeley's conversation was Barbara Slusher. Slusher is director of the NeuroTranslational Drug Discovery Program, which is nestled in the Johns Hopkins Brain Science Institute. BSi began life in 2007 when a large anonymous gift enabled the vision of the late Jack Griffin, a longtime Johns Hopkins neurologist. From the start, the institute has not been short on ambition, as attested by its mission statement: "to solve fundamental questions about brain development and function and to use these insights to understand the mechanisms of brain disease." The institute has made more than $20 million in grants across more than 20 academic departments throughout Johns Hopkins in Medicine, the Krieger School of Arts and Sciences, the Whiting School of Engineering, the School of Education, the Bloomberg School of Public Health, Peabody Conservatory, Kennedy Krieger Institute, and the Applied Physics Laboratory. It has helped secure state-of-the-art research equipment, technology that goes by names like a 7 Tesla MRI magnet and a transcranial magnetic stimulator, and established Barbara Slusher's drug translation program to aid in the development of novel pharmaceuticals.
During their conversation, Blakeley mentioned to Slusher that neurofibromas are thought to develop from cells known as nonmyelinating Schwann cells. Slusher had spent 18 years working in the pharmaceutical industry before coming to Johns Hopkins and was familiar with a lot of drug research. Now she had a question for Blakeley: Did she know of an enzyme, GCP2? Slusher had performed more than a decade of drug discovery research on this enzyme, which is also expressed in the prostate. She had just reviewed an imaging study where scientists evaluating the enzyme in a prostate cancer patient with neurofibromatosis had by accident discovered that the enzyme was also highly abundant in neurofibromas. Could that information be useful?
Slusher and Blakeley began to collaborate on a closer look at GCP2 and found that in benign neurofibromas, presence of the enzyme indeed is strikingly high. But in the malignant forms of the peripheral nerve tumors, GCP2 drops to nearly nothing. The researchers realized two implications. First, find a way to block the receptors for the enzyme, which may be important to the functioning of the tumor cells, and you might be able to inhibit growth of benign neurofibromas. Second, GCP2 might serve as a biomarker on which to base more precise diagnosis between the benign and malignant forms of the tumors, which are hard to tell apart even in biopsy. Because neurofibromas develop on nerves—"[The nerve] looks a lot like a python that has just swallowed a rabbit," Blakeley says— they are nearly impossible to cut out without nerve damage, often with devastating consequences like paralysis. That makes surgery a last resort. But symptoms that indicate the tumors have become cancerous are hard to detect until it is too late to do anything else.
The prostate researchers had developed an isotope that binds to GCP2. Under a form of imaging called single-photon emission computed tomography, the isotope lights up, marking wherever it has bound to GCP2. Blakeley and Slusher believe this means the enzyme could be used to monitor benign tumors and warn when they are about to turn malignant. They envision imaging a child with neurofibromas at regular intervals, using the isotope to look for GCP2 expression. If the tumors are benign, the enzyme will light up. If tumors stop lighting up, that could alert doctors that the neurofibromas are about to become cancerous and it is time for last-resort surgery. "That would take an otherwise asymptomatic person, whom we would not have operated on, and instead of waiting for the cancer, we could get it out before it's cancer," Blakeley says. "That would change the way we practice medicine in neurofibromas."
Changing the practice of medicine is one goal of the institute. "I regard BSi as one of the most important events in the history of neuroscience at Hopkins," says Solomon Snyder, the eminent neuroscientist for whom the School of Medicine's neuroscience department is named. "Its funding has facilitated the recruitment of several outstanding new faculty as well as enabled the retention of some of our stars who had been offered hard-to-resist packages from other institutions. More important, BSi has initiated myriad activities bringing together faculty and students from all Hopkins campuses in common enterprise."
Johns Hopkins President Ronald J. Daniels is a big supporter of BSi. "The Brain Science Institute exemplifies the power, potential, and necessity for cross-cutting collaboration to advance the neurosciences. Since its inception, BSi has championed the kind of creative thinking that increasingly occurs at the nexus of disciplines. With visionary support, BSi has laid the foundation not only for cutting-edge research in the brain sciences, but also for research into the confluences of science, education, and art."
At BSi, inquiry and application walk hand in hand. If the institute had a mantra, it might be "collaborate, discover, translate." One good translation of "trans- late" in this context is "drugs"—finding new drugs or effective new applications for existing drugs. The average time required to discover and develop a new drug to treat the brain in the United States is 15 to 20 years. The average cost is $1 billion to $2 billion. And 95 percent of these drugs fail in clinical trials and thus never make it to market. So it should not come as a surprise that many of the major pharmaceutical companies are exiting or significantly decreasing their efforts in discovering drugs for the central nervous system—too risky, too expensive. This exodus comes just as a substantial portion of the U.S. population is aging, and as lifespan extension means more and more people will confront neurological dysfunction and diseases like Alzheimer's, dementia, and stroke. In response, many research universities have begun stepping up to fill the void.
Early in 2008, the pharmaceutical giant Eisai acquired a Baltimore company, MGI Pharma. When Eisai announced it was moving the company to Boston, BSi saw an opportunity and hired away more than a dozen experienced pharmaceutical scientists who preferred to continue their work in Baltimore. This one move created the NeuroTranslational Drug Program directed by Slusher. In its first four years, it has landed more than 20 translational grants and more than $3.5 million in contracts, filed for five patents on new drugs and technologies, and founded a Baltimore-based pharma company, Cerecor, to aid in commercializing Hopkins' translational discoveries.
There are many good reasons for developing drugs that can treat brain injury. To name a few: cerebral palsy, Alzheimer's disease, ALS, traumatic brain injury, multiple sclerosis, brain tumors. Getting drugs into the body is easy— needles and pills—but getting them to the right parts of the brain is hard. First they must pass through the blood-brain barrier, which has evolved to prevent that very thing—foreign substances entering the brain. Succeed in getting something therapeutic past the barrier and into brain tissue and you then must move it through the brain to wherever it is needed, then somehow get it to act on (preferably only on) the cells causing the disease or injury. The task is not impossible, but it is awfully damned hard.
Kannan Rangaramanujam is an Indian-born professor of ophthalmology at the Wilmer Eye Institute who has received funding from and collaborates closely with BSi. For obvious reasons, he goes by Dr. Kannan or just plain Kannan. He is a chemical engineer who used to work in polymer physics and nanomedicine. At Wilmer, he works on how to deliver therapeutic nanoparticles to precise areas of inflammation in the human body, and he has branched out from studying age-related macular degeneration to also researching new methods that deliver drugs to parts of the brain that have gone wrong.
Common to brain diseases and injuries is inflammation. Inflammation impairs the blood-brain barrier just enough for drug molecules to slip past. Scientists have explored delivering drugs through the barrier by attaching them to ligands, which are molecules that can be selected to bond to the specific brain cells that require the drug. Ligands are like bicycle messengers with one fundamental shortcoming—they can be directed to a neighborhood, but not to a specific address. If a ligand happens to come close to the drug's target, it will make its delivery to the right cells. But it cannot be sent straight to where the drug is needed. Kannan uses an analogy from his native country: "To recognize me and shake my hand, you have to be in my vicinity. If you are in the train station in Bombay, there is no way you will find me." To anyone trying to deliver drugs via ligands, the brain is the Bombay train station.
So Kannan has been working with a smart messenger called a dendrimer. Dendrimers are tree-like molecules with a spectacular ability to go straight to inflammation. If you bind a drug molecule to a dendrimer and inject it into an animal with a healthy brain, it goes nowhere near the brain and the animal's body eventually clears it. But if you inject it into an animal with a brain injury, it will move through the blood-brain barrier and straight to the inflamed region. Better than that, it will go straight to the inflammatory cells that require the drug.
In his faculty office, Kannan finds a video clip from a recent study. In the video is a pair of hairless, newborn rabbit kits placed on white mats. The kits are littermates with induced leg impairments characteristic of cerebral palsy; when they try to walk, they fall over and can barely manage a directionless crawl. For the study, researchers injected one kit with saline. The other was injected with an amino acid derivative called NAC (shorthand for N-acetylcysteine, already used for treating Tylenol overdoses) attached to dendrimers. Video from five days later shows no change in the control animal that received the saline. But the kit that received the single dendrimer-NAC injection walks so well it has to be gently nudged back onto the mat. Kannan has to contain his excitement. "'Cure' is a strong word," he says. "I don't want to use the word 'cure.' But this is a dramatic motor-function recovery." He does not yet know if the effect lasts or wears off. He does not yet know if what works on a rabbit kit will work in people with cerebral palsy. But he knows that dendrimers took the NAC straight to where the rabbit's brain needed it, and that could have huge implications for treating brain disease.
If a promising treatment that capitalizes on the capabilities of dendrimers emerges from Kannan's work, BSi has expertise—Slusher and all those former MGI Pharma researchers—in how to move from lab to pharmaceutical to clinic. The neuroscience translation unit essentially acts as a small biopharmaceutical company within Johns Hopkins. "They can test new drugs for us. That's simply not available in a typical academic laboratory," Blakeley says. "We have the insight of a pharmaceutical company applied to the expertise of a laboratory. If I didn't have BSi, I'd be hoofing it to AstraZeneca and Pfizer and Sanofi, and I'd be looking at what they have and negotiating to see if I could have it. Here, with BSi, we can say, 'Here's the science. Now, can you make us a molecule that matches the science?'" Researchers advised or funded by BSi also note the speed with which they can work through parts of the research process. "We review grant applications inside of a month, compared to NIH, which is nine to 12 months," says Blakeley. "We're very focused. We did an assessment of what the field needs, and we identified gaps and said we were going to fund to fill those gaps."
In Amy Bastian's lab there is a treadmill unlike any you may have encountered in a gym. It is called a split-belt treadmill, and it allows researchers to manipulate each leg's stride separately to study what they call incoordination—limps and other uneven, unbalanced gaits that result from strokes and brain injuries.
Bastian, a Johns Hopkins professor of neuroscience and director of the motor analysis laboratory at the Kennedy Krieger Institute, uses BSi support to study movement disorders that arise from abnormalities to the central nervous system, and the special treadmill is useful for trying to figure out what is happening in the brain during walking. For example, she has found that when she creates a limp on the treadmill, the walker's brain adjusts to do what it can to even out the impairment. After time spent walking with the machine-induced limp, put the person on a normal treadmill and she will still limp, but now with the other leg, and will have to unlearn that unbalanced gait to resume walking normally. "I can tell you the new limp is going to happen, you can think about it all you want, but you can't do anything about it," Bastian says. "This is something that takes place in lower motor centers in the brain and probably in the spine."
She has been zeroing in on the cerebellum, a smaller part of the brain that lies under the massive cerebrum. If someone suffers brain damage from a stroke, the damage usually takes place in the cerebrum. This is fortunate, in a way, because Bastian has found that people with damaged cerebrums can be trained on the special treadmill to walk with more symmetrical strides, but people with damage to the cerebellum cannot. This would suggest the cerebellum as the seat, at least for locomotion, of what Bastian calls error-based learning. She believes that in someone with a limp, the cerebellum perceives the asymmetry of the stride as an error, and tries to correct it. "Step by step, you have this error and you update what the motor commands should be to get rid of the error." If the cerebellum is damaged by stroke, an individual can no longer learn a new walking pattern.
This led Bastian to wonder if by stimulating an intact cerebellum, she could help people with lesions in the cerebrum to learn faster. She has been experimenting with placing wet pads, something like the electro pads used for an EEG or EKG, on the cheeks and on the head nearest the cerebellum, and running weak electrical current between them. She has found that this stimulation induces faster motor learning on the treadmill. Why is not yet clear. Perhaps the current excites more neurons, giving the person more brain cells to work with. "We think we're tapping into a learning mechanism for stroke patients that might be advantageous because they don't have to think about how to walk. The learning takes place in a sort of subconscious way."
One difficulty, though, has been getting the improved walking pattern to stick. The new learning decays quickly, as if the brain for some reason comes to prefer the post-stroke, asymmetrical walk. "We don't understand why people go back to their crummy pattern when they're perfectly capable of producing a normal walking pattern," she says. "So there's something else, maybe in the nervous system, that's stuck on the bad pattern." She has been measuring the metabolic cost to the body of walking. Perhaps the process of changing to the newer, more symmetric pattern costs significantly more metabolically, so the brain tries to conserve energy by rejecting it. Perhaps they are simply stuck in a habit that is hard to unlearn. Whatever the explanation, there is a premium on finding means of learning that not only do not fade, but do not take weeks of repetition to create lasting change, because it is hard to get people to complete long rehabilitation programs. Says Bastian, "It's tedious, it's boring, and it can be painful or uncomfortable. It works and has tons of benefits, but people just don't do it."
They might do it were it more fun, and if it better engaged the senses. Tucked away in the Carnegie Building on the medical campus—you feel like a lab rat running a maze when you try to find it—is the Kata Project: a group of engineers, programmers, and artists developing animal simulations and games with potential applications in both therapy and brain science. Kata—a Japanese word for "form" and a reference to a practiced, choreographed pattern of movement—is funded in part by BSi and President Daniels' Science of Learning Initiative, and based in a lab run by John Krakauer, a systems neuroscientist and stroke neurologist with keen interests in motor skill learning and brain plasticity after stroke.
Though his medical specialty is stroke, Krakauer became interested in another form of impairment, traumatic brain injury, when he and a collaborator, Johns Hopkins Associate Professor of Anesthesiology Robert Stevens, noticed that no one had collected longitudinal data on people recovering from TBI. From a research standpoint, people with TBI are a tough group to work with. They are typically young, male, and subject to high-risk behavior—a preponderance of traumatic brain injuries result from vehicular accidents, violent sports, and fights. TBI often leaves the injured person forgetful and irritable and impatient with being considered sick. Those are not the sort of people eager to submit to long therapy and years of monitoring and testing in the name of science.
Krakauer believes one way in might be the iPad. Omar Ahmad, Promit Roy, and Kat McNally of the Kata Project have developed something for the Apple tablet that currently goes by the informal name of "the ant game." It features a sort of cartoon ant-car hybrid moving across the iPad's screen. To play the game, you control the path of the ant-car by how you tilt the iPad, and score points by how well you keep the car in the center of the road during a 90-second run. After about 10 rounds of play, the game begins to introduce distractions. Targets in different colors appear and move across the screen. The player may be instructed to ignore blue targets but make red targets disappear by tapping on them, all the while keeping the ant-car on course down the middle of the road. The game measures both motor skill and the player's ability to handle the cognitive load imposed by the distraction of the targets. The game records movement trajectories and a score for each round and thus forms a data set that is a record of recovery for someone with TBI. Young men usually like playing video games, so Krakauer hopes they will be agreeable to playing the game once a week for a year as part of a longitudinal study.
Kata also has been working on an extraordinary physical simulation of a swimming dolphin. The dolphin swims in a lovely, soothing seascape on a large monitor, controlled by a hand-held motion capture device shaped something like a flashlight. The task is to learn how to move your hand and arm to make the dolphin swim, dive, roll, and leap out of the water. As a game, it is fun—one day in the lab, a visiting graduate student named Lorenzo demonstrates a striking ability to get the dolphin to leap and spin in the air, and makes the Kata gang laugh by saying, "I'm thinking of taking this back to my lab to decrease productivity." But the bigger effort is to create an immersive, aesthetically pleasing environment for stroke patients who need to rehabilitate motor function. Someone engaged in the tedious business of regaining precise movement of the hand and arm might find learning to make a dolphin leap from a virtual ocean far more entertaining than the repetitive exercises of conventional rehab.
Funding programmers and visual artists to create iPad games and virtual worlds is one example of BSi's willingness to pay for the sort of creative research that can have a hard time securing money from standard sources like NIH. Another is the work of Charles Limb, an otolaryngologist, neuroscientist, and musician fascinated by how human beings perceive complex sound, especially music. "Music is an entity that reveals so much about human nature, throughout history, throughout all cultures, all epochs," he says. "I have this idea that music cannot be understood as just an artifact of the need for entertainment. I think our brains are hardwired to perceive, seek, and produce these kinds of entities for very basic biological reasons. I think it's deeply linked to how we generate new ideas well beyond music."
Limb has received a lot of media attention for placing jazz musicians in a functional magnetic resonance imaging machine, then asking them to improvise on a miniature keyboard while the fMRI records their brain activity. By studying what happens in the brain when a musician spontaneously invents something, he hopes to gain a better understanding of how the brain creates. This ties in with his broader interest in neuroaesthetics, which can be thought of as the neuroscience of art. Limb notes that while Johns Hopkins has had many people studying basic sensory perception, that research was not being linked to artistic or aesthetic practices, behaviors, or goals. "There was a gap between people who were studying the basic biology of sensor y perception and those who were interested in art," he says. Limb has received research support arising from BSi's interest in the intersection of science and the arts. "It's a real uphill battle to get funding. Everyone finds the work interesting, but no one wants to pay for it."
BSi believes that fundamental to translating basic research to clinical applications is investment in the next generation of neuroscientists. Daniel O'Connor is a 36-year-old neuroscientist working in the institute's Synapses, Circuits, and Cognitive Disorders Program. He studies how the brain forms our perception of the world. Decades of work in psychology and neuroscience have demonstrated that perception is malleable, something that we construct from what our senses take in. Through sight, hearing, smell, and touch we sample our environment, taking a continuous sequence of sensory snapshots. Then, our brain goes to work. For example, we assume certain regularities based on personal history, and thus have an expectation of the external world. In our brains, that expectation interacts with our senses to form our idea of what's out there moment by moment. O'Connor wants to understand how the brain does this using a simple model.
In his lab, he works with mice and their sense of touch. (Think whiskers.) "The beautiful thing about the mammalian brain is that a lot of the core architecture is conserved all the way from mice to humans," O'Connor says. One experiment in his lab involves placing a mouse in a black box, to negate its vision, and stimulating its sense of touch by slightly moving one of its whiskers. The mouse has been trained to report—"as best we can do with a critter that can't tell us what it perceives"—by licking. If it feels something touch its whisker, it licks a sensor. The experimenter then dials down the stimulus. Now sometimes the mouse detects the touch, and other times it fails. The reduced stimulus is the same each time, so what happens in the brain when the mouse detects it and when it does not? Did the animal's mind wander? Can the scientists change whether the mouse expects the stimulus at a certain moment, then see how that changes the activity in the sensory cortex? What does it mean, neurologically, to expect something, or to pay attention? (Lick the sensor if you know the answer.)
BSi's director, Jeffrey Rothstein, says, "BSi has enabled Johns Hopkins to continue to expand and improve the productivity of its clinical and basic research. BSi funding has led to more than $100 million in additional new NIH grants, and its efforts have led to increases in new patents, new early candidate drugs, the spinoff of new biotech companies, and licensing agreements between Hopkins neuroscientists and pharmaceutical and biotech companies."
Looking ahead, the institute has outlined where it intends to focus additional effort, including autism, brain and spinal injuries, post-traumatic stress disorder, and personalized therapies for neurodegeneration. A number of scientists who work on BSi-funded research speak of the institute as a model for the sort of collaborative, creative, more nimble research centers that, in their view, universities must foster if they are to remain at the forefront of biomedical research. Rothstein says, "That's the unique aspect of BSi. At its core, it's about coordinated collaboration, whether that's for advancing basic research or, perhaps more importantly, translational research. Investigators want more flexibility to work across the disease or organ boundaries that traditionally define departments. They seek new academic organizations centered on common methods, products, or processes that apply to many nervous system diseases or health-related specialties."
You would have to look long and hard to find anyone in academia who does not endorse cross-disciplinary collaboration. But two linked questions arise. At an institution like Johns Hopkins that encompasses such a vast array of scientific endeavors with leading researchers in every discipline, what is so hard about finding other creative minds from other domains to collaborate with? Why does it take research institutes dedicated to encouraging and facilitating the sort of creative encounters everyone agrees are the future of science?
Sometimes it just comes down to the practicalities of time, distance, and workload. The people who work in all the various scientific domains and sub-domains and sub-sub-domains are scattered across Johns Hopkins' campuses, vast in number, perhaps separated only by floors of a building or even a corridor, but often separated by city blocks or miles. Finding someone working on a specific sort of science can be a time-consuming hunt, and fruitful chance encounters at the mailboxes or in a cafeteria are unlikely.
"BSi has identified the various hurdles to more productive collaborations in translational research," Rothstein says. "These barriers have been addressed by creating an efficient mechanism for connecting individual investigators or teams to experts knowledgeable in ways of overcoming hurdles that arise during a translational research project." Or, as Blakeley puts it in plainer language, "I work about 100 hours a week, which doesn't leave time to have coffee. It doesn't leave time to socialize in the cafeteria. Most clinician-scientists are not in public spaces to connect with one another." BSi serves to bring the right people and the right resources together. "What BSi says is, 'Come here; we will have the people who matter to you in the room. And we'll provide coffee.'"
Correction: An earlier version of this story included an inaccurate description of the synthetic dolphin developed by the Kata Project. The dolphin is a physical simulation.