Christopher Keenan
Introduction: My research team’s original question seemed simple enough: what sorts of neurons exist in a zebrafish tectum (the site of sensory integration in the zebrafish brain)? But, as with any worthwhile scientific endeavor, our project has not proven to be as easy or straightforward as the question may suggest. Over the past year, we have tested several theories of how tectal neurons develop in the zebrafish embryo. We were especially interested in multisensory neurons as these are the most important cells in the tectum. We were also interested in the development of multisensory neuron development in zebrafish as the development is completely unknown and issues with multisensory neuron development is associated with bipolar disorder and autism in humans. We employed several strategies to try and elucidate the truth.
Methodology: Our first strategy was focused primarily on development; we would inject zebrafish embryos with a plasmid (a circular piece of DNA) called kaede and this would cause certain cell populations in the zebrafish to glow green. Kaede proteins glow green unless exposed to UV light, then they undergo a conformational change and glow red. We controlled which cells glowed green using the UAS/GAL4 system of gene expression. This allowed us to activate only certain cell populations in specific transgenic fish lines. Our transgenic fish had labeling in the precursor cells to tectal neurons, neuroprogenitor cells.
This is a powerful tool, as we could photoconvert individual cells so our cell of interest would glow red and all other cells would glow green. This allowed us to image specific neurons. If we photoconverted a neuroprogenitor cell, all of its descendent cells would also glow red. Our strategy was therefore to photoconvert a single neuroprogenitor cell and record all of the cells it developed into, giving us information both on the morphology (shape) of cells developed in the tectum and the patterns on how these cells are developed. We would also take a video of the development using a confocal microscope, which would take a picture of the cells every 15 minutes for 20 hours. Once this was done, we could record videos of the neural development. All images and videos are taken using a confocal microscope with an anesthetized zebrafish embedded in agar.
After months spent on this strategy, while we were able to get some beautiful videos, it became apparent that this would not be an effective long term strategy due to two major problems. First, the cells developed off of the neuroprogenitor cell in dense clumps, in which it was impossible to distinguish one cell from another. While this taught us about the patterns of neural development, the neurons were too dense to distinguish individual neurons. Second, we would occasionally accidentally photoconvert two neuroprogenitor cells, ruining our data and a full week of work.
We changed tact and began photoconverting adult zebrafish neurons. We had several lines of fish expressing kaede in different patterns of neurons in the tectum. By imaging the adult neurons, we could determine both the neuron morphologies that show up in the tectum and the specific neurons that appear in our specific patterns. At this time, we took on another undergraduate researcher, Lindsey Woodward, who I trained in the skills for our project.
We would inject a number of embryos with kaede and, 5 days later, photoconvert several neurons in the morning and then image the neurons in the afternoon. We needed to wait several hours for the photoconverted protein to diffuse throughout the cell. We would repeat this process several times a week with multiple injections and photo conversion cycles. The one drawback to this process was that it was extremely slow going. We could only image two to six neurons per
afternoon, as each image was very time consuming. Additionally, occasionally our injections would not work or we would accidentally photoconvert several neurons at the same time, clouding our data. Photoconverting a single neuron is very difficult, because labeled neurons are often very close together in the tectum.
We began looking for other strategies to speed up the process and a lab worker, Nora Waltz, stumbled across a technology called Zebrabow. Without going into too much technical detail, Zebrabow is injectable, like kaede, but causes every new generation of neuron to glow a different color. For example, if the neuroprogenitor cell glowed red, the second generation would glow purple, the third would glow blue, and the fourth would glow green. (These colors are not precisely how the system works, but is an acceptable approximation for explanation purposes.)
We found that the Zebrabow plasmid we gained did not work as we had expected. We were unable to control the timing of the color changes and Zebrabow would not even appear to be present in our fish after injections. After working with Zebrabow for three months, in which I was away at an internship in Boston, Lindsey and Dr. Suli concluded that we needed to rebuild the plasmid as our copy were not effective.
At this point, I returned from my internship at the start of fall semester 2017. We decided to return to our previous strategy of individual photo conversion, which was effective even though it was slow. We trained a third person on the project, Jeffery Dunn, who helped us to speed up the process. After months of doing this, we accumulated many pictures but found that the pace is still very slow, as we are only able to dedicate 10-20 hours per week to our project due to our undergraduate course schedule. Additionally, we found that the variety of neuronal morphologies was so large that we were utterly unable to distinguish a pattern. So at this point in the project, we have many images of individual neurons with some preliminary patterning, but we are worried about the high volume of images needed and the time consuming nature of taking these images.
Summarized Results: Zebrafish tectums grow in dense wedge-like patterns from neuroprogenitor cell. Many different neuron cell morphologies are present in adult zebrafish, though similar morphologies can be found in similar sections of the tectum. For example, a unipolar neuron with a short axon and flat branching can frequently be observed on the extreme posterior end of the tectum.
Discussion/Future Direction: Our research is ongoing and we have numerous ideas on how to improve both the pace and the efficiently of the project. We intend to hone our focus to only multisensory neurons. We will first identify which neurons are multisensory, then photoconvert the cells to determine cell morphology. The protocol for identifying multisensory neurons has already been documented by Nora Waltz. At this point, we will then stain the samples with antibodies for various neurotransmitters to determine which the function of the neurons. Should this line of experimentation work, we will know the identity, morphology and function of multisensory neurons in the zebrafish tectum. All images taken have been recorded and archived and we are building a map of the placement of all neurons recorded.
Conclusion: Our project has been a long and wonderful one. We have taken many beautiful images and videos and are constantly readjusting our project to be more efficient and to gather better data. We intend to publish our findings when complete, and I am very grateful to the ORCA donors for giving me the opportunity to have spent a year working on a project on the forefronts of scientific knowledge. With a full class schedule, as well as a wife with a child on the way, I could not have done the research project without the grant money provided to me.