In the silent film Modern Times, Charlie Chaplin is a waiter delivering dinner to a rather impatient customer. To do so, he must navigate across a crowded dance floor. While his body gets lost among the crowd, we can still follow him because he holds the wine bottle high above his head and the crowd. This is how biophysicist Melike Lakadamyali explained the basics of virus tracking using fluorescence microscopy at the 2011 TEDxBarcelona conference.
In her lab at the Institute of Photonic Sciences (IPS) in Barcelona, Spain, she applies fluorescence microscopy to follow single viruses as they infect living cells to understand the mechanisms of viral infection. But during her postdoc at Harvard University, she found herself in the middle of a perfect storm: she was at the right time and the right place to be a pioneer in the application of the super-resolution microscopy technique called STORM to the field of connectomics.
In 2006, Lakadamyali earned her Ph.D. from Harvard’s Department of Physics after five years of graduate research in the lab of Harvard professor of chemistry and chemical biology Xiaowei Zhuang. Her work focused mostly on developing methods for single-virus tracking in live cells, and, along with her colleagues from Zhuang's lab, she published a paper in the PLoS Biology in 2007 that demonstrated the imaging of poliovirus entry into living cells (1).
But this wasn’t the only research that was going on in Zhuang’s lab. At the same time, other members of the lab, namely researchers Michael J. Rust and Mark Bates, were developing a super-resolution microscopy technique called stochastic optical reconstruction microscopy (STORM). In 2007, they published a paper in Nature Methods describing the technique, which uses photoswitchable fluorophores to obtain nanometer accuracy in fluorescence microscopy, effectively breaking the diffraction limit of traditional fluorescence microscopy (2).
This caught Lakadamyali’s interest. “It’s such an promising technique for biology. So, when I was looking for a postdoc position, I was looking for an opportunity that would allow me to apply this technique to an interesting problem in collaboration with Xiaowei Zhuang,” says Lakadamyali. And she didn’t have to look very far.
Elsewhere at Harvard, Jeff Lichtman was working to solve just that problem in neurobiology. To trace neural pathways, connectomics researchers have traditionally relied upon electron microscopy because of its high resolution. But because it uses immunogold as a label, the technique’s molecular specificity isn’t very good. In addition, the images produced by electron microscopes are grayscale, so different neural populations cannot be differentiated by different colors.
In contrast, Lichtman has taken a fluorescent microscopy approach for tracing these neural pathways, developing transgenic techniques to label neural cells. In a 2007 paper published in Nature, Lichtman and colleagues described a technique to genetically label neurons in mouse brain tissue with approximately 90 different colors (3). This allowed them to study the circuitry of the neurons in an unprecedented way. They called the technique “brainbow”. Since then, the technique has been applied to a variety of other animal models used in neurobiology, including zebrafish and Drosophila.
“Neuroscience is one of those fields that has benefited from technology, and in particular fluorescence imaging techniques. For example, two-photon confocal imaging was really a revolutionary technique for this field. And optogenetics is another example of how light can be used to propel the field forward,” says Lakadamyali.
But since Lichtman’s brainbow technique relied on fluorescence microscopy, its resolution was restricted by the diffraction limit of visible light. So, in the end, the neuronal processes could not be traced in as great detail using the brainbow technique as was originally hoped.
Eye of the STORM
In 2007, a group of five researchers at Harvard, Stanford University, and the Massachusetts Institute of Technology (MIT) received funding to apply 3D super-resolution microscopy to the brainbow technique. This group was dubbed the BrainSTORM Consortium and consisted of Zhuang, Lichtman, Sebastian Seung, Stephen Smith, and Joshua Sanes.
Lakadamyali was at the right place at the right time. As a result of this funding, she was offered a postdoc position in Lichtman’s lab. Her goal was to bring brainbow below the diffraction limit, a little bit closer to the resolution of electron microscopy. And almost immediately, challenges began to present themselves.
The goal of connectomics is to trace the connectivity of neurons, which means imaging large fields of view. Previously, super-resolution microscopy had typically been used to image one cell in great detail. Now she was trying to apply this to map the interactions of several neurons across a large volume of tissue. To accomplish this, Lakadamyali would have to take an image, move the sample, then repeat that imaging-repositioning cycle over and over again, which would not have been a particularly enjoyable task. “No one wants to sit through an overnight imaging session, clicking buttons and moving the sample,” she laughs. “So, automation and high throughput was one challenge that we had to tackle.”
With the help of Harvard research associate Hazen Babcock from Zhuang’s lab, Lakadamyali automated the different components of a STORM microscope to get the necessary throughput. Specifically, the two automated the microscope stage so that it could be moved along the x, y, and z axis with high precision. Each image taken overlapped the previous field of view, so that in the end, the images could be stitched together.
“Once you set up how you want to change your laser powers and which region you want to image, the microscope does the rest of the imaging. So, in the end, it was quite a highly automated procedure,” says Lakadamyali.
Another challenge involved label density. Super-resolution techniques such as STORM depend on a high label density. Lakadamyali had to improve the amount of green fluorescent protein per unit of volume so that the neurons looked continuously labelled. The trick was to label the membrane rather than the cytoplasm, which improved the technique's resolution.
In a paper published in PLoS ONE in 2012, Lakadamyali and colleagues reported on the 3D super-resolution imaging technique for tracing neural connectivity in cultured neurons (4). After only a couple of months, it’s already been cited by two other papers.
Somewhere Over the Brainbow
In the end, 3D BrainSTORM will be another tool to improve the speed and efficiency of connectomics, and researchers in the field will gladly accept the help. “Each technique has its own advantages. They are complementary to one another,” says Lakadamyali. “Especially when used in combination, they could be even more powerful than using each technique separately.”
For example, the high molecular specificity and multicolor capability provided by STORM will allow scientists to look at molecular content, such as in the synaptic connections of neurons. On the other hand, if you wanted to primarily look at the morphology of a neuron, electron microscopy can provide a better image of the ultrastructure.
But electron microscopy requires rather demanding sample preparation and, despite recent advances in automation, is still relatively slow. Meanwhile, super-resolution imaging is getting faster. So, an effective strategy for future connectomics studies might be to first image a sample using 3D super-resolution fluorescence techniques and then use electron microscopy to look at a particular region of interest in greater detail.
“You can correlate your molecular map to the ultrastructure,” says Lakadamyali. “We are starting to see examples of this already, and it will be a very empowering technique.”
The next step for the 3D super-resolution brainbow technique is to apply it to actual brain tissue, as opposed to cultured neurons as reported in the PLoS ONE paper. The major challenge for this will be to increase the label density, which remains the limiting factor. Traditionally, researchers have embedded brain tissue samples in plastic prior to imagei scanning, but this process compromises the antibody’s ability to label its epitope. To overcome this, a new sample preparation method will be necessary.
Of course, there are other avenues of development that will follow, such as live-cell imaging. “When super-resolution techniques were first developed, the speed was only compatible with imaging fixed cells, but over the past 10 years, the speed has improved dramatically,” says Lakadamyali. “There’s still a lot of push in this direction, and room for improvement, so we’re going to start to see faster and faster imaging from super-resolution.”
At the Max Planck Institute for Biophysical Chemistry, super-resolution microscopy pioneer Stefan W. Hell and colleagues have already begun making headway in this area. During the 1990s, Hell had developed the super-resolution microscopy technique called stimulated emission depletion (STED) (5). In a paper published in Science in February 2012, Hell and colleagues reported on a super-resolution microscopy technique that provided them with a view of live neurons within a mouse brain (6).
But Lakadamyali has left further development of the 3D super-resolution brainbow technique in the capable hands of her Harvard mentors. After her postdoc in Lichtman’s lab ended, Lakadamyali was offered a position at the IPS in Barcelona, Spain where she is continuing to use super-resolution imaging to answer other biological questions that interest her.
In some respect, she has come full circle back to her Ph.D. dissertation on single-particle tracking of influenza. But now, she is using super-resolution to understand the virus’ transport and entry mechanisms. In addition, she is collaborating with other researchers in Spain to study psychiatric diseases and chromatin’s role in transcription. “My training is in physics. I’m a physicist, but I’m interested in biology and biological problems. The biology motivates me and drives me.”
1. Brandenburg, B., L.Y. Lee, M. Lakadamyali, M.J. Rust, X. Zhuang, and J.M. Hogle. 2007. Imaging poliovirus entry in live cells. PLoS Biol. 5(7):e183+.
2. Rust, M.J., M. Bates, and X. Zhuang. 2006. Sub-diffraction-limit imaging by stochastic optical reconstruction microscopy (STORM). Nat. Methods 3(10):793-796.
3. Livet, J., T.A. Weissman, H. Kang, R.W. Draft, J. Lu, R.A. Bennis, J.R. Sanes, and J.W. Lichtman. 2007. Transgenic strategies for combinatorial expression of fluorescent proteins in the nervous system. Nature 450(7166):56-62.
4. Lakadamyali, M., H. Babcock, M. Bates, X. Zhuang, and J. Lichtman. 2012. 3D multicolor Super-Resolution imaging offers improved accuracy in neuron tracing. PLoS ONE 7(1):e30826+.
5. Klar, T.A., S. Jakobs, M. Dyba, A. Egner, and S.W. Hell. 2000. Fluorescence microscopy with diffraction resolution barrier broken by stimulated emission. Proc. Natl. Acad. Sci. 97(15):8206-8210.
6. Berning, S., K.I. Willig, H. Steffens, P. Dibaj, and S.W. Hell. 2012. Nanoscopy in a living mouse brain. Science 335(6068):551.
7. Lakadamyali, M. 2012. High resolution imaging of neuronal connectivity. Journal of Microscopy:no.