Recently, an eclectic group of neuroscientists, engineers, computer scientists and even an astronomer gathered for the first time to share their science, brainstorm and hopefully begin fruitful collaborations.
As director of the Kavli NDI, Huganir opens the meeting with gratitude for the support of the Kavli Foundation and the university, and with excitement about the progress he envisions around three pillars of innovation: neuroexperiments, neuroengineering and neurodata. “We are starting with 45 members from 14 different departments who epitomize the Kavli ideal, but if you get Kavli NDI funding, you automatically become a member, and all of our events will be open to the whole university,” says Huganir, who is also director of the Department of Neuroscience.
He then requested applications for the first funding opportunities: a postdoctoral fellow who would bridge two interdisciplinary labs and a visiting scientist fund to bring relevant scientists to Johns Hopkins for a week to promote Kavli-inspired collaborations around the country and the world.
Though the institute is primarily an interpersonal reality, it will have some dedicated space in the bioengineering department on the Homewood campus. A large video monitor there and at the school of medicine will create a portal uniting the two campuses 24/7, further facilitating the exchange of ideas.
Neuroexperiments That Ask Intriguing Questions
To start that exchange, 13 Kavli NDI members gave brief talks about their work. Without any PowerPoint slides, neuroscientist David Linden
steps up to the podium. He begins by telling, or reminding, the audience that one of the dogmas of modern neuroscience is that memories are saved in the brain through synapses. That is, the answer to “2+2=?” is somehow encoded in a series of connections between specific neurons that are formed or strengthened during the learning process. But then, he mentions three situations in which mammals lose at least one-third of their synapses without any noticeable memory loss: during hibernation, after castration and during menstruation. And in each case, synapses are regained upon waking, with testosterone replacement therapy or with the continuation of the female’s cycle, respectively.
“Do women lose and regain the same synapses each month, or is it a different group that is lost and gained?” Linden queries. “Is there a way to predict which synapses will be lost? What is the purpose of this inefficient process?”
To answer all of those questions, “I need to be able to monitor individual synapses as animals go through these changes,” he tells his audience. “Right now, that’s only possible in the brains of laboratory animals, so I’m hoping someone can work with me to develop the right device to see this process in humans.”
Huganir says that that kind of problem-solving is exactly what he expects to see more of as the institute matures. “Our colleagues at the Whiting School of Engineering and the Applied Physics Laboratory have so much expertise,” he says. “The funds from the Kavli Foundation allow us to tap into that in unprecedented ways.”
Neuroengineering to Build Better Tools
One tool that might help Linden comes from the lab of biomedical engineer Xingde Li
. With a background in physics, most of his research at Johns Hopkins has been focused on developing imaging technologies for clinical applications. Then, four years ago, his research caught the eyes of Huganir, Dwight Bergles
and Marshall Shuler
as they began to prepare their Kavli application. Would Li be interested in developing tools for imaging rodent brains? He gave an enthusiastic yes and has been an integral part of the Johns Hopkins Kavli team ever since.
At the symposium, he spoke about how far his research has come. Bergles and others in the neuroscience department have been developing ways to peer deep within the brain to spy on the activity of both neurons and glial cells, the two main types of cells in the brain. Their technology has come a long way but always with a catch: The rodents’ heads have to be kept still during imaging, which limits the questions they can ask to simple behaviors, like eating and running.
Building on his knowledge of patient imaging tools, Li began with an endomicroscope, a flexible fiber-optic cable that physicians use to view the deepest recesses of the human body. With a little trial and error, he modified it to be mountable on a mouse’s head. Because the cord weighs less than 1 gram, the rat or mouse can move freely without feeling the tug of a leash. Now, the team plans on studying how the brain changes during exercise, and maybe Li will team up with Linden to study synapse loss too.
Neurodata: Making Sense of It All
But the solution of these technical problems creates another problem, which you can relate to if you’ve ever reached your smartphone’s maximum storage capacity. Bergles notes that a single imaging experiment from the rodent brain can generate many terabytes of data, creating challenges for analyzing and archiving such large data sets. Other labs are also generating large quantities of data — mostly images, but also the detailed molecular differences between specialized cells in the brain.
To be transformed into scientific discoveries, all of that data have to be stored and analyzed. And that’s where astronomer Alex Szalay
comes in. At the Kavli symposium, Szalay reported on a project he led in the 1990s called the Sloan Digital Sky Survey
— a digital hub for astronomical data and analysis that is now the world’s most used astronomy “facility.”
“It was the cosmic version of the Human Genome Project,” he says, “but this one was open to the public, and some of them made very interesting discoveries.”
Szalay is now working with computational engineers like Miller, Joshua Vogelstein
and Randal Burns
to build a cloud-based “SciServer” so that other fields of science can benefit in the same way. They have already developed several “brain clouds,” cyberspaces dedicated to collecting millions of brain images from people of different ages, ethnicities and health statuses, all at the service of the national scientific community.
“Every brain is different,” says Miller, co-director of the Kavli NDI and founder of a field called computational anatomy, “but they should all contain about 300 unique structures with typical sizes, shapes and connections to other structures. They should also grow and age in similar ways. We use computers to describe those structures and processes as equations so we can learn what ‘healthy’ and what ‘unhealthy’ look like. Because most of the roughly 1,000 neuropsychiatric illnesses result in changes of volume, shape or connection in one structure or another, we expect that these databases will allow more diseases to be diagnosed early on with a simple image comparison.”
Vogelstein, too, has built several brain clouds, but his are geared toward understanding the brains of research animals. One is entirely devoted to data from “clarity brains,” brains treated with special chemicals that make the tissue clear, allowing researchers to see cells and their connections in all of their three-dimensional complexity.
Speaking the Same Language
Huganir and Miller believe that the connections between scientists of diverse fields, like Bergles and Szalay, is one reason the Kavli Foundation chose to fund an institute here. But that diversity also creates a challenge: Every field has its own jargon, so scientists can have a hard time understanding each other’s work.
agrees. She and Loyal Goff
collaborate on a project to map the diversity of neurons in the cerebral cortex. Though they are both in the Department of Neuroscience, Goff’s expertise is molecular and computational biology, while Brown’s is neural circuits. Brown says they spend two to six hours each week discussing their data. “That’s a lot of time and energy, but it’s super productive,” she says. “Working together allows us to address questions neither of us could tackle alone.”
, another neuroscientist attending the symposium, adds that collaborators have to understand the biological problem at hand in all of its complexity, and “I have to put in the effort to understand how they generated a particular answer or insight to my question,” she says. “With time, we will each learn to trust alternative approaches.”
At the symposium, Huganir offers encouragement after the first coffee break. “Different words mean different things in different fields, so it will be a process to learn to speak the same language,” he says. “It will take time to learn the power of each lab, but it will transform our research.”