Learn with us online while the Exploratorium is temporarily closed. You can help us reopen—donate today.
View transcript- So it's a real pleasure to be here today. I'm looking forward to learning so many new things, and thank you so much, Jen, for this invitation. I'm gonna be telling you about some of my work. As Jen mentioned, I'm trained as a cell biologist, and my visualization is really focused on kind of trying to serve my community of cell biologists. And so, what I do is I create animations of molecular processes. And I'll show you a couple examples of that. But, I first wanted to give some idea of the motivation of why doing animation of these things. And so, the central problem, I think, facing molecular biologists, is the fact that molecules are tiny. So this is a great little interactive from the Genetics Science Learning Center, which is also based at the University of Utah. And what you can see are that molecules, while cells are maybe on the order of 10 microns, or something like that, molecules tend to be around maybe 10 nanometers. The wavelength of visible light tends to be between 500 to 800 nanometers, which means that most of the molecules, the proteins that molecular biologists are interested in studying, are smaller than the wavelength of light. And this means that we can never see them directly using visible light. And so that's problematic, because how do you actually think about things that are so small you can't see them? So, what biologists do is they do a lot of experiments that each can provide little bits of indirect data. And so, Jen, this is probably kind of data that's familiar to Jen. Most of the data that we collect, the raw data that we collect in cell biology and molecular biology, provides very indirect bits of information. And so a lot of the experiments, like biochemistry experiments, tell us something about where proteins might interact with each other, what parts of the protein are necessary, things like that, and those kind of data generally look something like this. So not particularly visual. We do do a lot of microscopy, as well. And so there's a lot of fluorescence microscopy happening in cell biology, and the kind of analogy that I like to think of is, if you have a protein that is glowing, it's kind of like being able to see someone from a mile away, who's holding a gigantic flashlight. You can see the light, but you can't make out the shape of that person. You can't make out who that person is. And that's what's kind of going on in these kinds of images. So, we can see where the proteins are in the cell, but we can't really make out their shape. There's also a lot of light microscopy that is looking at live cells. So that's the example here in the lower right. So this is an image I actually produced when I was in graduate school. And this is looking at a specific protein that moves in the cell. So looking at live cell microscopy, you can see how proteins are moving around, and there's also a lot of data these days looking at three-dimensional light microscopy, So being able to see cells in 3D and moving over time. And so this gives us a lot of information about what proteins are doing in the cell, more about their location and things like that. But if we're actually interested in what proteins look like, we can't rely on visible light. So what we do then, is we do other kinds of experiments where we essentially pull proteins out of the cell, purify them, and sometimes crystallize them, sometimes freeze them under cryo conditions. And so what this allows us to do is be able to use X-rays or electrons to be able to make out, essentially, the three-dimensional shape of a protein. Here's are some images of electron, some proteins that were frozen for cryo-electron microscopy, and that can give us results like what's shown in the top left, where we have a three-dimensional structure of a protein. So the thing to keep in mind, is that this is still quite indirect. You're pulling the protein outside of its native form, outside of the cell, and putting it in pretty unnatural conditions in order to figure out the three-dimensional shape. But, so this is all bits of indirect data. However, these give us a more visual idea of what might be going on. And from that, what scientists do is we come up with hypotheses. Typically, so these can be drawings. People are just basically taking all of these bits of data, the biochemistry, the cell biology and microscopy, whatever structures, and just basically thinking about all these bits of data, how do we combine this and synthesize this into an idea? And so these are examples of drawings that were made by my collaborators to talk about basically what they think is going on at a molecular scale, within the context of a cell. So these types of visualizations are considered to be, they're called the model figure. Basically, what do we think is happening? This is our hypothesis of what could be happening at a molecular scale. So I've worked with dozens of different researchers over the years to create visualizations of what they work on. And inevitably, every researcher grabs a pencil and paper and starts drawing. I think people have a really hard time explaining what they're doing without being able to show it in the form of some type of visualization. And it starts out as a hand drawing, just kind of putting an idea on paper, and then it gets refined into a figure that often is published in our publications. So this is a main way that scientists really communicate with each other is in the form of these publications. And, oftentimes, you can see these kinds of model figures as one of the last figures in the paper, that really summarizes our hypotheses of what we think is going on. So my introduction to animation came in graduate school. I was right nextdoor to a lab that studied a motor protein called Kinesin. We had joint group meetings, which meant that I watched dozens of presentations by graduate students and postdocs, talking about how Kinesin worked. So this is a figure that was published in a paper by Sarah Rice, who was a graduate student at the time, that talked about, basically, the hypothesis of how Kinesin walks along microtubules. And you can see, this is a pretty simple diagram that really kind of simplifies the structure of the protein, and is kind of giving a step-by-step idea of how this works. And so, from watching things like this, I really thought I understood how Kinesin worked. Then later in graduate school, about a year or two into it, the structure of Kinesin was solved, and Ron Vale hired an animator named Graham Johnson to create an animation of Kinesin. And this was also shown during group meeting. And the thing that really struck me about this was that it made me realize that I never really understood how Kinesin worked. My jaw just dropped and I was like, "This is the most intuitive, easy to understand thing. "Why aren't we all doing this? "Why are we just looking at triangles and circles, "when really, we have enough data, "dynamic data, structural data, "to put together visualizations "that really capture the information that we have?" And so, this was really the thing that drew me into thinking about animation. And so I'm gonna show you one of these animations that was made several years ago. This is a collaboration with Tom Kirchhausen, who studies this hard to pronounce process called Clathrin mediated endocytosis. I'll go ahead and play the animation, and I'll walk you through it. So this is outside of the cell. Now we're looking inside. So these green proteins are called adaptor proteins. They're called adaptors because they're able to bind to a protein named Clathrin, which can self-assemble into these regular, sort of soccer ball-looking shapes. Meanwhile, they can also capture proteins on the outside of the cell, which you'll be able to see in a moment here. So these are receptors on the outside of the cell. And so these adaptor proteins can basically link those two things together. Clathrin's main role is bending the membrane, which is a very, very energy-consumptive process. And so, it's now basically bending the membrane underneath, and forming a cup shape. The proteins coming in now, basically cause fission. The cutting the membrane between this newly formed vesicle, this kind of like a bubble of membrane, and the plasma membrane, which is at the surface of the cell. And so, this basically cuts the membrane, as shown. The music went away . I like the music. Can we turn up the music just a little bit? And so now we have proteins that are coming in that cause basically this Clathrin assembly to come apart. And so two major proteins are known to be involved in that. So this protein recognizes part of Clathrin, as well as part of the membrane of the vesicle. And the second protein recognizes this orange protein, and then another part of Clathrin, which is considered the Achilles heel. So it kind of binds it there across the entire cage, and is thought to kind of loosen everything up and allow the Clathrin to disassemble. So this is that process. So one of the things that I like to emphasize about this animation, and pretty much all the animations I create at the molecular scale, is that they're hypotheses. It's really one person, or maybe a small group's idea of how something like this occurs. And there's a lot of data that we have to support it, but there's a lot of, also, kind of guess work, and a lot of uncertainty involved at the molecular scale. And even with this particular animation, which was done with, like I said, Tom Kirchhausen, his idea of how this process changed over the years, and so we've gone through many different revisions. The part that's changed the most is this part on Dynamin. The fission process. So an earlier animation showed, so basically we didn't have a structure. We just knew it appeared at the neck, and it did something, and then fission happened. So that was shown in this animation. This animation that you just saw was a revision of that. We had a structure. We had sort of a start point and an end point, and so we showed that in this animation. It turned out that this mechanism was not popular, and I knew this because complete strangers would walk up to me at meetings, having seen this animation from a presentation Tom gave, and they would tell me, "You know that animation's pretty good, but you know, the Dynamin part, it's wrong." Complete strangers. Wouldn't even tell me their name. And so, it turned out a couple years later, I was invited to go to a conference on endocytosis. And my immediate reaction was, "You know, I don't actually work on endocytosis, right?" And the organizer said, "We know what you do. "We've all seen your animation, "and we'd like you to talk about the process." And so I thought that was great. So I decided to go to this meeting, and a couple months before this meeting, Tom called and said he was actually invited to give the keynote at the end of the conference. And he said, "You know, I think we need to change the Dynamin part of the animation, basically, before the conference." And so what we did was we worked on this animation, so during my talk I show the previous animation, and then I created new version of the animation that's shown here. So this is a little bit different. So the start point and the end points of the animations are actually the same as the previous, however the mechanism is different. Here, we have basically a Brownian ratchet-like motion that slowly closes the circle. And so this is the animation that Tom showed at the last day of the conference, and after that, there was kind of a social hour at the bar, and a couple people came up to me, people who specialize in thinking about this particular process, and one person said, "It's closer." So this gave me this impression that people who have been studying a process for years, maybe decades, have a movie in their heads. They're taking all of this abstract data and building a vision in their heads of how these molecular processes work. And when they see something that doesn't match that mental model, they get upset, which is great because, I think, in some ways it helps provide a dialogue. It allows people to understand where they disagree. And so these animations can really kind of bring that process, really make people think about where they disagree and what kind of experiments they could do to test which hypothesis is correct. So Tom has been a great collaborator in this, and so he says that, "Molecular 3D animations "inform both the scientist who creates them "and the audience that views them, "through an active process "leading to further inquiry and discovery." So one of the things I wanted to point out is that animation has a very steep learning curve. It took me months, years to basically get to the point where I could create sort of a semi-reasonable animation of these kinds of molecular processes. One of the projects that I've been really keen on thinking about is how do we democratize animation? How do we make it something that all researchers can really do without having to work through a specialist like me? Because I think a lot of the insights that are gained through animation are really come about through the interaction of playing around with different proteins, and moving them around on the screen. I'm involved in a couple different software projects that are aimed at really trying to think about how we can provide these tools to researchers. One of these projects is actually with the Allen Institute for Cell Science. Blair Lyons is here at the meeting. She's back there, and so please, if you're interested in these kinds of software development projects, you can talk to either me or to Blair about that. All right, so I wanted to show one more example This one also has sound so hopefully it'll play, because I'm not exactly sure I can come up with a good narration if it doesn't, but let's see. Oh, great. - Bacteria get around. - That's a little loud. - [Video Narrator] They move in sometimes mysterious ways. One of the most mysterious is via the pilus. To get ahead using pili, a pilus is thrown forward like a grappling hook. After touch down, the pilus is retracted back, and the cell is dragged along. But how does the cell make a grappling hook and pull it back in? Researchers used electron cryotomography to take many, many pictures of the pilus' machinery. Then, using genetic techniques, they took away some components and they labeled others, mapping the location of each protein in the structure. With this new understanding of the structure, they proposed a mechanism for how the pilus gets sent out and brought back. When activated, a motor-like protein adds units together into a chain, which pushes out through the membrane. When the extending chain touches a surface, the proteins change their shape, sending a message back to the machinery. In response to this message, a different motor reverses the process, retracting the pilus and pulling the cell along. Taking into account the size and forces at work here, One of the things that I was really interested in thinking about when I got into animation is, basically, how do we get animation to be used in the research process? And so I wanted to think about how we can, basically, replace the model figure with an animation. And so this was one of the examples where that was actually done. So this was in the publication that was published by my collaborators, Grant Jensen's group at Caltech. And you can see that the animation is embedded in the research article, which is exactly where the model figure would go. This also went under review, and the reviewers thought it was great for providing context, really allowing people to understand where this machine worked. It also, because I was involved pretty early in the process, right when they were getting these structures, the animation process also allowed them to gain insight into the structure. So one of the the things, Grant was interviewed talking about basically, the fact that when I first got the structure, I told them that I didn't understand how these little yellow proteins could get inside of this cage, because there didn't seem to be enough space. And that made them look back at the structure, and they realized there's this big conformational change, a big movement that happened that was really kind of, I think, initiated because of this question I had. So the video I just showed you was actually put on by Science Magazine, their Facebook page. You can see it right here. This is kind of my final point, but it gets 539 thousand views at the point I took a picture of this, which is more than Why are Lions Smarter than Tigers, and Why Fly When You Can Water Ski. And I point this out to molecular biologists, and cell biologists, because I think people really tend to think that the public isn't interested in molecules. They're too small, they're too abstract. And what I like to tell them, is that it's all about engaging them. It's all about telling them a story and being able to provide the context for why these things are interesting, that make it really accessible to the public. That's all I had for today, but my focus has really been on trying to get these kind of molecular animations to engage with both the scientists, as well as the public. Thank you.
University of Utah biochemist Janet Iwasa discusses the importance of visualization in forming scientific hypotheses. She shares her work visualizing protein function, a major driver of cellular processes. Because proteins are so small, their behavior isn’t directly observable and scientists must pull together many types of data in visualizations to understand their function. Translating their hypotheses into 3-D animations allow biochemists to communicate, critique and refine the research.
This talk was part of the Visualization for Informal Science Education conference held at the Exploratorium, which explored themes of interpretation, narration, broadening participation, applying research to practice, collaboration, and the affordances of technology.
VISUALISE was made possible thanks to generous support from the Gordon and Betty Moore Foundation and the National Science Foundation under Grant No. 1811163. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.
Pier 15
(Embarcadero at Green Street)
San Francisco, CA 94111
(415) 528-4444
The Exploratorium is a 501(c)(3) nonprofit organization. Our tax ID #: 94-1696494© 2021 Exploratorium | Terms of Service | Privacy Policy | Your California Privacy Rights |