The brain is in a constant state of change. Neurons continuously rewire themselves in response to inputs from within, and from the external world. Despite this, the brain retains abilities and memories throughout our lives. However, there are still huge gaps in neuroscientists’ knowledge of how such plasticity affects behavior.
Understanding how the brain changes over time requires a new generation of tools that can work at infinitesimal scales that were once inconceivable. And it requires a new level of collaboration between brain researchers, engineers, and computational scientists that, until recently, didn’t exist.
The Kavli Institute for Fundamental Neuroscience (Kavli IFN) at the University of California, San Francisco (UCSF) will create a framework for connecting neuroscientists with tool builders at UCSF and nearby institutions. The Kavli IFN’s co-directors, Roger Nicoll and Loren Frank, recently discussed the importance of plasticity and stability in the brain, and the need for engineering and computation-based tools for advancing neuroscience.
- Loren Frank, PhD – Co-director of Kavli IFN. Frank is a professor of physiology at UCSF, and a Howard Hughes Medical Institute investigator. His research explores how neural activity underlies learning and memory and guides decision-making. He has also worked extensively to develop new tools for monitoring and manipulating neural activity in awake, active animals.
- Roger Nicoll, MD – Co-director of Kavli IFN. Nicoll is a professor in the departments of cellular and molecular pharmacology and of physiology at UCSF. His work has been foundational in establishing how chemical signals strengthen and weaken connections between neurons – a key mechanism in learning and memory. Nicoll is also the director of the Neuroscience Graduate Program at UCSF.
The following is an edited transcript of the discussion. The participants have had the opportunity to amend or edit their remarks.
The Kavli Foundation (TKF): The Kavli Institute for Fundamental Neuroscience at least initially aims to home in on a single area of inquiry: neuroplasticity. Why did you choose this?
Roger Nicoll: The brain is anatomically very stable and yet it changes all the time. This plasticity underlies functions like learning and memory and is critical in making us who we are as individuals. We felt that by probing brain function through this lens, we would be able to understand the brain on multiple levels.
Loren Frank: Also, brain plasticity is an area where UCSF has particular historical strengths. Roger is too modest to say it, but his work has been crucial for understanding how individual synapses can change. Similarly Michael Merzenich, who is now professor emeritus at UCSF, has pioneered our understanding of how the brain retains plasticity throughout the lifespan. We want to build on our unique strengths in this area as a platform to move the field forward.
TKF: And how will the Kavli IFN’s focus evolve and unfold?
Frank: I think brain plasticity will always be an element of the Kavli IFN’s work, but the wonderful thing about an institute that carries on forever is that it can adapt as the times change and as the most relevant questions change. We could imagine, for example, that in 50 years the technology might exist to do all the studies we’re doing now, but on the human brain rather than in animal models.
Nicoll: Part of the Institute’s ability to adapt will involve technological innovation. We can’t predict where brain research will eventually go. But what we have in mind for this Institute is having a framework for researchers to partner with engineers at the Lawrence Berkeley National Laboratory and Lawrence Livermore National Laboratory -- two nearby federal research facilities with world-class expertise in technology development – and other key institutes here at UCSF. Through these partnerships we hope to develop state-of-the-art recording probes and a lot of other technologies. We’re certain that combining these various technologies will be essential for moving forward.
TKF: The Kavli IFN’s founding members will span three institutions – UCSF and the two national laboratories. How will these alliances bolster the technological capabilities of the institute?
Nicoll: Loren and others here have already established collaborations with folks at the national laboratories, but we realized that we needed to formalize the process.
Frank: Right. These connections exist, but only among a few of us. Right now, most researchers at UCSF don’t even know who to talk to at the national labs. We want to extend those connections to the whole neuroscience community by creating a “phone book" and "dating service,” if you will, to hook up researchers with shared interests.
Establishing close communication between engineers with technological expertise and researchers trying to solve a neuroscience problem is a real challenge—especially within the confines of academia. Academic engineers need to innovate. They aren’t necessarily accustomed to making reliable, functional tools that are easily transferable to scientists who want to make measurements in the lab. In contrast, what the national labs do is build instruments that work reliably – whether it’s a particle accelerator at CERN or a retinal implant to help a blind person see.
The Kavli IFN will provide people with the infrastructure to start understanding each other’s problems. If someone wants to build a new sensor, say, to understand how cells change their activity, they’ll know to go talk to the people who do nanoscale engineering at Lawrence Berkeley’s Molecular Foundry, who know a lot about that.
TKF: What are some of the collaborations between UCSF neuroscientists and instrument builders at the national labs that are already underway?
Frank: One is a project that Eddie Chang is doing. He is the head of a massive DARPA [Defense Advanced Research Projects Agency] grant to develop treatments for depression and post-traumatic stress disorder in humans. They’re trying to design new devices for stimulating the human brain that could relieve some of these psychological problems. Creating those devices requires a fabrication facility, as well as people who understand how to build something that can sit in a human brain for years and not come apart. That’s actually really hard – brains were not evolved for having physical objects stuck in them, as it turns out. So he’s working with a group at Lawrence Livermore to do that.
Similarly, we have a BRAIN Initiative grant with the same group at Lawrence Livermore to develop long-term recording electrodes for animal models. Right now we can record from about 100 different electrodes at a time, which is enough to get a sense of what one part of one brain area is doing, but not to study a circuit as a whole. So we’re working with them and with a chip designer in Southern California to improve this by a factor of 10. The goal is to be able to record from a thousand electrodes in a freely behaving rat. That will allow us to look at how signals are propagating from one brain area to another at a level of detail that we simply couldn’t achieve before.
TKF: What about projects with the Lawrence Berkeley Lab?
Frank: The Berkeley Lab has a Molecular Foundry – a facility for building nanoscale tools – and there’s an amazing physicist there named Peter Denes. He’s interested in turning technologies developed from particle accelerators into really high-density recording systems for the brain. He and Eddie are working together and I’ve been talking with him also.
TKF: How will the Kavli IFN facilitate such connections?
Frank: That’s the dating service component, which we call the Technology Core. The plan is to have a couple people whose job will be to understand what technologies are available and what technologies are needed by the neuroscience community and to make the connections.
TKF: Dr. Nicoll, can you talk a little about how tools and technology development have played a role in your lab?
Nicoll: My lab has focused more on advances in mouse genetics and imaging than on engineering state-of-the-art electrodes and probes. We use what are called conditional knockout mice, in which a gene is deleted only in a specific organ or group of cells. That technology has been around for some 20 years, but we’ve found a clever way of using it to modify one single cell in an otherwise normal sea of cells. We can then take proteins of interest, change them in some specific way, and then put them back into a cell to create a sort of modified synapse. With this technique we can study synaptic plasticity at its most fundamental level. We can ask questions about whether or not particular elements of the protein are integral to the phenomenon we’re looking at. So it’s provided a level of genetic rigor that was unheard of in the past.
TKF: Your recent work examines not just how the nervous system stays adaptable throughout life, but also how it remains stable and maintains functions even in the face of such plasticity. How do these two opposing processes work together?
Nicoll: At a cellular level, at least, these two different phenomena take place hand-in-hand, but on a somewhat different time scale. You have this very rapid plasticity when you learn something – the synapses are modified very quickly and very specifically. But the problem is that these synaptic modifications introduce an instability into the network. That means you need to readjust the overall input to maintain that overall level of neural activity. This is called homeostasis, and I’ve been studying that along with the synaptic plasticity associated with learning and memory. Graeme Davis, who’s also at UCSF, has devoted much of his time to understanding this process as well. Interestingly, though the end result is a modification of a certain type of receptor, they do so by very different underlying molecular mechanisms. Both of these processes are very important for storing information, but also for maintaining a stable network and system.
TKF: Dr. Frank, you mentioned some of the tool development projects your lab is collaborating on with the national lab facilities. Can you also describe the big questions that these technologies are being used to address?
Frank: We’re interested in what are called replay events in the hippocampus, which is a brain region critical for forming new memories. The hippocampus gets activated when an animal runs around and explores its space. Then those same patterns of activity re-occur in the hippocampus when the animal isn’t moving, but sped up by a factor of about 20. These replay events recapitulate sequences from previous experience, and they seem to be critical for memory.
We’re trying to explore how these patterns of activity in the hippocampus engage the rest of the brain. To do that, we need to be able to record from a lot more cells than we can now. We have the technological capability to capture activity in the hippocampus, but we want to understand how that activity might travel to downstream brain regions. Our project with the Lawrence Livermore fabrication lab to develop electrodes that can be implanted long-term in the brains of rats aims to help us do that.
TKF: What happens when you interfere with these patterns of activity, rather than just record them?
Frank: We showed a few years ago that if we built a system that detects these replay events and then just turned off the hippocampus for a tenth of a second each time one occurred, we would seriously impair the animal’s ability to learn and make memory-guided decisions. But to understand what these different patterns of brain activity do, we need to be able to understand the information that they convey – ideally by manipulating them individually.
It looks like replay events might be ways that the animal represents possible future options. It’s as though the animal is playing through possibilities in its head to figure out what choice to make. So if that’s true – and we’re not sure yet that it is – then we should be able to interfere with its ability to think about certain possibilities. That in turn should change the way the animal behaves and the rest of the brain processes information. These replay events last only about a fifth of a second, so if we want to interrupt the signal before the animal finishes its thought, we need a way of decoding it quickly. We’re collaborating with statisticians and a engineers to develop systems that can do that.
TKF: Dr. Frank, you’ve talked before about the importance of using patterns of activity to understand circuits. Another key alliance for the Kavli IFN is the California Institute for Quantitative Biosciences (QB3), a systems biology institute housed at UCSF that spans three UC campuses. How will the Kavli IFN use that connection?
Frank: This is another case where we have these amazing, brilliant quantitative people in such close proximity but have not built bridges between our work and theirs that would allow us to collaborate effectively. Scientists at the QB3 are looking at cellular networks, protein networks, or gene networks in many of the same ways that we’re looking at neural networks. So the question is, how do we learn from each other and inspire each other? Are there tools that they’ve developed that we could translate to the brain?
Nicoll: Right. And the idea behind establishing all of these close ties is to inspire a new type neuroscientist who is able to bring all these strands together.
TKF: How will the Kavli IFN promote a new way of thinking about neuroscience?
Frank: Well, first there’s what we call the “infectious meme” possibility. As an example of that, two new scientists who were just hired in the department of pharmaceutical chemistry and who work on cellular networks have mentioned an interest in neuroscience. So a couple of us are meeting with them to encourage their budding interest and welcome them to the fold. We’re making a real effort to get people outside of neuroscience hooked so that they will contribute their ideas to the field. Then there’s the training element. Roger is head of UCSF’s Neuroscience Graduate Program and is perhaps better suited to address this, but I’ll just briefly say that neuroscientists now also need to be versed in engineering, math, and statistics. They don’t necessarily need to be able to do the engineering themselves, but they need to know enough to effectively speak to the engineers.
Nicoll: I agree. Moving forward, neuroscientists will have to have a much broader knowledge base that includes engineering and computation. Through the Kavli IFN we’ll be making sure to put young researchers in an atmosphere where they can become facile with all these ways of thinking. For one thing, our students will be able to do rotations in the labs of our partner institutions, which will broaden their experiences significantly. We’re also working to set up a scheme for funding small pilot projects specifically for students and postdocs that would allow them to take advantage of the Institute’s infrastructure and set up their own collaborations with researchers at those facilities.
Frank: Many of us are excited to build up capacity here in the quantitative and computational side of neuroscience by bringing in faculty with that expertise. My hope and expectation is that adding this glue between all of these different institutions and having more faculty at UCSF who are interested in quantitative perspectives on neuroscience problems will raise the bar from a training perspective as well.
TKF: California has been especially progressive in supporting neuroscience, most prominently through the state neurotechnology grant program, the Cal-BRAIN Initiative. How will the Kavli IFN leverage this state-wide support?
Frank: The support we will receive from The Kavli Foundation to launch the Kavli IFN will be incredibly valuable, but the Institute’s goal isn’t necessarily to fund lots of individual research in many different labs. Instead, the Kavli IFN will bring researchers together in a dynamic way, so that they can successfully get grants through Cal-BRAIN, the BRAIN Initiative, the National Science Foundation and National Institutes of Health, DARPA, or other funding sources to advance their research. The federal government is still by far the largest funder of research, and to be successful as a researcher you almost have to have federal support. But to compete for that effectively, we need interdisciplinary teams that can combine insight from several fields. Building that kind of collaboration requires a framework. So the Kavli IFN will essentially provide the structure on which highly innovative neuroscience can grow.