Episode 2: Making The Jump (Transcript)

Dr. Wickersham [0:00] 

I’m trusting you Peter.

Peter [0:03] 

If you could use one word to describe it.

Dr. Wickersham [0:06] 

Bread-y, I’m feeling like that’s a hot dog. I only got the bun, I think

Peter [0:15] 

That is fine. You can open up your eyes now. It was a hot dog on gluten free bread (we made sure). I think the reason we ended up choosing a hot dog was we think that dogs are associated with rabies, so what’s the closest food we could go for and we thought of hot dog.

Dr. Wickersham [0:31] 

Yeah. Nice.

Peter [0:41] 

Hi, my name is Peter and I will be your host for the Gastronauts podcast. Here at Gastronauts, we are committed to understanding communication in the body, and in particular how our gut talks to our brain. In this podcast, we take a deeper dive into the mind and motivations of leading science and their work. We hope that by getting to know the scientists behind the science, we will be able to bridge the gap between different scientific spheres, and between the scientific community and mainstream culture. So let’s dive right in, The Gastronauts Podcast: Between Two Spheres.

We’re really thrilled to have Dr. Ian Wickersham here today. Ian studied physics as an undergrad here at Duke University and then went on to do his PhD at UCSD in neurobiology. He later did his postdoc at MIT, and now serves as head of the genetic neuro-engineering group at MIT. He focuses on developing powerful and precise techniques to study the structure of the brain. He’s taken advantage of really unique features of viruses and modified them so that they can infect particular cells and light up so we can visualize brain networks. One of these viruses that he’s worked with is rabies virus. And I’ll be honest, when I first heard of using rabies virus as a tool, I got a picture of […] a rabid salivating dog with somewhat of a crazed look. And I was wondering, […] how did you get drawn to rabies? Did it take some convincing to get you to work with a virus that has somewhat of a scary reputation?

Dr. Wickersham [2:40] 

Well, no. And of course, I actively wanted to work with it, so I had to convince others more than the other way around. But the thing about rabies virus is that while it is terrible pathogen that does kill many people worldwide every year, it is a tremendously useful, naturally occurring tool for neuroscientists, because the way that it spreads is between synaptically-connected neurons. It spreads between neurons in a way that doesn’t kill the neurons outright for quite some time. So when I started my PhD, we were looking for ways of identifying connected neurons in the brain. Because neurons in the brain are of very many different cell types and they’re all mixed in together and they have these long processes, axons and dendrites that overlap. And it’s impossible to tell just by looking at them, even if you are lighting up different types of them or staining them. It’s impossible to tell which are connected to each other. And the exquisitely precise connectivity between neurons is basically maybe the most unique aspect of the brain as opposed to other organs. So it’s very important for understanding how the brain works, or how one little aspect of the brain works, to understand how the neurons that are involved in that behavior that you’re interested in, are connected to each other. So we needed a tool that would allow us to identify cells that are connected to a targeted group of other cells and rabies virus was beckoning as the most promising way of doing that.

Peter [4:29] 

That’s really interesting. You mentioned rabies, more so as a tool, than kind of a pathogen. And your lab really focuses on developing tools and prototypes and modifying these viruses so that we can use them to understand different circuits and different connections. And I was wondering, when we go through developing these tools, we go through an iterative process where you go through prototype one or generation one, generation two, generation three. How do you know when generation one is ready? How do you know when you’ve made a construct that you think this is kind of a tool that we can sell to other people or let other people know about?

Dr. Wickersham [5:06] 

Well, as far as the first-generation version went, there was nothing else out there. And so as soon as we had a basic sort of demonstration that we could do this, and I should say that the system that we were trying to invent was a way of specifically labeling neurons that are directly connected to a targeted group of neurons. So basically, we would be able to selectively infect whatever type of neuron one wanted in the brain with a modified form of rabies virus and allow the rabies virus to spread not throughout the entire brain, like wild type rabies virus does, but only to the cells, which are directly synaptically connected to the starting neuron population.

Peter [5:55]

So you control kind of the specificity of how its spread.

Dr. Wickersham [5:58]

We control two things, actually, one was the specificity of which cells it would infect to start with, and we also controlled the number of synaptic steps that it would travel, basically. So naturally occurring rabies virus will infect essentially any neurons it encounters if you were inject in the brain. And it would, once it infected those cells, it would spread in a retrograde direction, that is from the starting cells to the cells, which are pre-synaptic to them.

Peter [6:29] 

And pre-synaptic means that it forms the connection upstream of it.

Dr. Wickersham [6:33] 

That’s right, to cells, which release neural transmitter on to the cells that you’re starting with. And wild-type rabies virus will simply replicate within those cells, and keep spreading to the cells, which are pre-synaptic to them, and so on, and spread throughout the entire brain. And we wanted a system that would allow us to label only the directly pre-synaptic cells, so that we could very precisely identify sort of a connection matrix between cell types in the brain.

Peter [7:01] 

So kind of like a controlled rabies infection.

Dr. Wickersham [7:05] 

Yeah, that is the first-generation system being able to do that. And of course, we publish it as soon as we had any kind [of data], like one does basically in neuroscience. Once you have something which is novel, no one has ever shown before, even if it’s not perfect, you want to get it out there. So without waiting for perfection, one simply publishes each new, major advance.

Peter [7:30] 

But at the same time, you want to make sure that the toxicity or the lethality of the rabies is controlled as well.

Dr. Wickersham [7:36] 

Exactly, and the toxicity of rabies viruses is maybe its biggest downside. It’s certainly less toxic than many other viruses. And the reason, if you will, that it’s less toxic is the rabies virus wants to keep the cells of the of the nervous system intact, so that the host, the infected animal, will be able to go on and transmit the virus. Basically, the viral life cycle depends on not trashing the nervous system as it spreads through it so that the animal is in a position to implement that behavior, which causes the spread the virus. So basically, the virus is already fairly non-toxic. But within neuroscience, there are many, many experiments that one would like to do that involve manipulations or study over a long period of time, not just a few days or a couple of weeks. But in mice, for example, that are learning a task, that are developing over time, and one would want to see how the synaptic connected networks of neurons evolve their responses over time as the mouse learns a task, for example, something like that. And that basically has not really been possible with the rabies virus systems that have been out there, or mainly the system that has been out there, because it does kill the neurons. And so a huge effort in my lab has been in recent years to develop non-toxic versions of the so called model synaptic tracing system, the rabies virus tracing system.

Peter [9:13] 

And the mono-synaptic tracing system is kind of just having a jump one synapse, right and not continuing to spread, and that’s how you control a little bit of the toxicity?

Dr. Wickersham [9:17] 

Well, that’s how we control the spread of the virus, but that doesn’t directly do anything about the toxicity per se, of the virus to the infected cells. So in the second, and now, third generation versions that we have underway, the rabies virus appears to be completely non-toxic. And so that means that we can label pre-synaptic cells, and then leave them alive indefinitely so that they can be studied and manipulate for long-term behavioral experiments.

Peter [9:52] 

Okay, so what I’m getting is kind of your generation one is […] you have this vision set forth, you want to be able to label a pathway and […] you found a way to do it. It is modestly toxic, relatively low toxicity. And then when you went to generation two, you’re really trying to focus on how can we reduce this toxicity  almost zero? And then how can we get this labeling to persist?

Dr. Wickersham [10:16] 

Yeah, exactly. I mean, the very first pass was, if I may say so is a huge step forward, because basically, it is, and remains the only way of identifying cells which are directly connected to some population of the brain that you’re interested in, without already having a hypothesis that you can then test. So for example, let’s say you’re interested in dopaminergic cells of the membrane. These are cells which project to the cortex and depending on the cells, the striatum, and other places in the brain, and they’re tremendously important for motivation, and reward and movement control. And being able to identify what the inputs are to the those tremendously important, dopaminergic neurons in the membrane will allow neuroscientists to map out the entire circuit manipulates various inputs and see something basic about how the organization of this key system of the brain or the systems probably more properly, is put together. Now, the first order question that one has is what are those cells and where are they? And so the first generation rabies virus system was just to answer that […] anatomical question: what and where are the cells which are present? Wherever they are in the brain, it will label them. And so just to be able to do that anatomical mapping before going on, and then maybe manipulating pre-synaptic cells with other means.

Peter [11:48] 

So what really makes this rabies still powerful is kind of the resolution that you get, is that correct? Because my previous understanding of neurobiology is we have these brain regions, and we know they’re, to anthropomorphize, talking to one another. But we don’t know specifically, what neuron is talking within this brain region to this other neuron, right? And the rabies virus gives you the ability to look at this direct connection between two cells.

Dr. Wickersham [12:12] 

That’s correct. That’s exactly right. So what one can do without rabies virus or with a variety of other tracers, and what people had done before this came along was see where in the brain there are cells that project to a brain region. But with any of those techniques, there’s no way of seeing which cells in the targeted region, those upstream neurons project to. So you basically can get, with high resolution, mapping of cells everywhere in the brain, that project to a place. But before the rabies virus system came along, there was no way of identifying which cells in that target location, were post-synaptic to, all these pre-synaptic neurons. So what we can do with the rabies virus is select the cells that you’re interested in, mapping the inputs to, and then map the inputs specifically to them. And they’re different, essentially, everywhere in the brain, there are many different cell types and they differ in their conductivity by and large. And so essentially, you may have differences just in degree of the numbers of neurons in a particular brain region that project to one type of cell in, let’s say the cortex, or another. Or you may have completely non overlapping types of cells, which are pre-synaptic to the two, for example. And this allows you a much higher resolution ability to map conductivity.

Peter [14:00] 

Yeah, so that really shows like even neurons that are close in space that are localized may not be part of this network that we’re looking at. To pick your brain a little bit. Where do you see the future of this going? I think structure belies function. And I think understanding this network is important for us to really know how the brain works. And I was wondering […] where do we see this going forward? How do you think, our understanding of these networks? Or how do you think we can better understand these networks with new techniques within the next 10 years or so?

Dr. Wickersham [14:32] 

Yeah, well, for one thing, be able to see more about those specifically pre-synaptic cells than just where they are and what they look like, and what they express and so on, would be good. So if you can, with a non-toxic version of all of this, do patterned stimulation of pre-synaptic neurons to be able to perturb the activity of these pre-synaptic cells, and see how that affects the activity of these targeted post -synaptic cells. And one can with the rabies virus, do mapping of inputs to not just a group of neurons, but a single neuron. And this allows beautiful anatomical mapping for a start of inputs to individual cortical neurons, but so you can express calcium indicators in those pre-synaptic cells, and be able to image, for example, visual response properties of all of those pre-synaptic cells, as well as of the single targeted synaptic cell and then see, for example, whether a neuron in the visual cortex, for example, that responds in a certain way to visual stimuli, gets input from primarily other neurons that respond in that same way, or whether it’s doing something maybe a bit more interesting. So in other words, answering the question, which I think is one of the most fundamental questions of neuroscience, how does a single neuron take the input that it’s getting and process that information to produce its own output?

Peter [16:02]

So what I’m hearing is moving forward, you think kind of understanding these on a single cell level, how we integrate information, more how these neurons integrate information in a single cell level and how they can take that information and integrate and send that message over that synapse to another cell, is where you see the future of these techniques moving forward.

Dr. Wickersham [16:23]

Yeah, well, so I think that’s, that’s a really great goal and a motivating goal. In terms of the futures of the techniques themselves, we should soon have a highly efficient, completely non-toxic mono-synaptic tracing system that would allow those sorts of experiments. I mean, at that point, neuroscientists of all stripes can use the technique and do whatever exciting science they want with it. We as the tool developers are going on and developing other tools that will allow other exciting science.

Peter [16:55] 

Now that’s really exciting moving forward, I can’t wait to see what your group does in the future. I’m going to go transition a little bit more now towards you as a scientist in your own career development. Moving from physics to neuroscience is kind of a big jump. And I think that I’d be interested in hearing how you made the jump from being a student in one field into a researcher in another?

Dr. Wickersham [17:23] 

So I mean, in my case, I was always interested in in the brain, not necessarily neuroscience, per se. But from the perspective of wanting to understand how it worked. I was always interested in physics, too. I enjoyed it. So that was my major in college. But I also became interested in the brain, and minored in neuroscience here at Duke. And was thinking about, though, from the perspective of neural networks [and] that was my, that was my motivation. So I was thinking, Okay, how can how can we build a brain? So now, I think the majority of people probably go into neuroscience, the goal of understanding the brain in its entirety, and so on, you know, figuring out how consciousness works, or building robots that think and that sort of thing. And then you sort of get to neuroscience, and it’s like, “okay, well, what you can actually do is this…” [What] might be a more direct path to take for actually building architectures that are more like thinking in a sense than, than we’ve had before. But going into neuroscience, my motivation was just to understand as much as I could about the organization of the brain.

Peter [18:28] 

What about the brain in particular? Is it just the fact that it is such a poorly understood network? Is it because we don’t know much about it? What drew you towards the brain?

Dr. Wickersham [18:37] 

Oh, I just think intelligence is cool. And, you know, it seems like okay, well, we should be able to build a machine that does that. So that really was the motivation. And I better start by understanding what we do know about the brain. So I ended up joining a PhD program- but, it just turns out that we don’t understand all that much is the truth of it and that’s sort of the glass half empty, or 99% empty, kind of a view of it. But in fact, in fact, we know an enormous amount, it must be said, but that basically, that was my trajectory, I was coming at it from a from a neural network sort of point of view, and interest and in construction, doing programs and architectures that […] were intelligent, and so on. But [when I] got to actual neuroscience, and realized that what I wanted to know- that information was not out there. From my perspective, what I wanted to know, for starters, was okay, what are the connections between all these different types of […] cells? What are the connections between them? What are they doing? And the tools for knowing all that didn’t exist.

Peter [19:45] 

And you decided to make them yourself!

Dr. Wickersham [19:46] 

Yeah. So […] it just seemed like the bigger bang for the buck, as it were, in terms of the impact that I could have, rather than using existing tools to study laboriously some circuit in the brain and get, you know, sort of incomplete answers about that circuit would be to develop tools that would allow people working in any aspect of the brain to just have much more powerful and precise experiments.

Peter [20:19] 

Yeah, something that just came to my mind was, your research is really focused on the synaptic labeling, or, first off, this mono-synaptic labeling how you jump from one cell to another. I was thinking, […] this kind of reminds me about how people make the jump from being a trainee to being a mentor. So how did you make that transition from kind of being a scientist, learning about neurobiology, learning about these neural circuits to being someone who was a pioneer in the field. Were there some tools or mindsets that you could […] share with people who are trying to make that transition from trainee to mentor.

Dr. Wickersham [21:00] 

So I guess my focus has always throughout been pretty much the same in just trying to develop tools that would just basically allow a huge increase in the abilities of experimental neuroscientists to study circuitry. So I mean, that was kind of like the driving characteristic in the PhD and throughout, it’s still the same thing. It’s just that now I can do a lot more, because I don’t have to do it all myself. And so that is, it is a different role, for sure. But analogously, if you can just make tools that will allow other people to find out things about the brain, then more will be found out than if you were trying to just do it yourself with the existing tool. Similarly, if you have ideas, and can delegate them to your co-workers, a lot can get done.

Peter [21:55] 

So it certainly sounds to me that when you were a trainee, kind of, the fascination with the tools to understand the network is what kind of drew you in, when you were first getting interested in research. But then, as you were trying to study these networks, you realized that you needed to build your own network of, you know, collaborators that would help develop this.

Dr. Wickersham [22:15] 

Throughout I was learning from mentors, postdocs in the same lab-

Peter [22:20] 

And how do you go about looking for this help, do kind of just see who’s in the area reach out to them,?

Dr. Wickersham [22:25] 

That’s a really good way to do it. The truth is, they’re there are great people all over the place, wherever you are. And so you want to find them. It’s just certainly at a place like the Salk, or MIT or Duke, it’s easy to find world class people who are happy to teach you. The degree of collegiality, I think, differs between places, but I’ve had certainly good luck with people being happy to help.

Peter [22:51] 

That’s really nice to hear, kind of, how collegial it is for science, because I remembering […] going back [to our earlier conversation], you’re thinking, Oh, you know, when I discover something, I really need to push it out. There’s a sense of competition, and having people that are available to you having, you know, a group of people within your institute or between institutes that are really driven and motivated to kind of help you succeed is important to taking that next step.

Dr. Wickersham [23:18] 

That’s right. That’s right. And it’s almost always win-win. And that’s the only really effective way of collaborating. Generally, it’s an academic collaboration and if there’s a significant amount of help, they’ll get some of the credit for the work and in many cases, be able to use the technology or whatever that’s being developed in their own work. People have a stake in the outcome like that, then they’ll be more motivated to help.

Peter [23:42] 

Right? So when forming collaborations, you want to make sure that it is a win-win for both parties, right? So we’re both people can advance their own work to advance the field and develop.

Dr. Wickersham [23:51] 

And generally people share authorship on the paper. And if the project is exciting enough that it’s going to be a high impact publication, just to put it in that sort of crass form, then […] that’s in their interests to participate in the development of that.

Peter [24:10] 

That’s really great. Well, I really want to thank you for taking the time to be on this podcast. Dr. Wickersham. Oh, I think we learned a lot about forming connections in the brain in our science, then yeah, thank you so much for your time.

Dr. Wickersham [24:23] 

My pleasure. Thank you very much.

Peter [24:34] 

Well, there you have it. Forming a collaborative network is easy when you take the time to get to know your colleagues. We challenge you to reach out. Share your ideas with the people down the hall, or the person sitting next to you at your next lunch or seminar. Thank you all so much for listening, and we’ll see you on the next episode. For more of our content, you can follow us on Twitter @gutbrains, or visit our website at thinkgastronauts.com. The Gastronauts Podcast would be impossible without the incredible team that we have here. Meredith Schmehl is our producer and the music composer. Dr. Laura Rupprecht is our social media manager. And special thanks to the founders of Gastronauts: Dr. Diego Bohórquez and the Bohórquez laboratory.