Jason Shepherd on Scientific Discovery

Taking on the major challenge of understanding how experience shapes neural networks and how circuits are modified by proteins/genes,  Jason Shepherd has garnered worldwide recognition through the research in his lab, Shepherd Lab at the University of Utah School of Medicine. At TEDMED 2018, Jason shared why we might have viruses to thank for the biology behind memory storage and encoding. Watch his Talk “How an ancient virus spread the ability to remember” and read his about his journey through Scientific Discovery below. 


What goes into scientific discoveries? Movies will have you think that discoveries are made by lone geniuses in moments of inspiration. The reality is that this is rarely the case, scientific discovery is a long and often tedious process that requires a team of people. In my own research lab, we recently made a surprising connection between two seemingly unrelated topics; viruses and memory. 

This connection was made through observation, rather than through inspiration. We study a gene called Arc, which is essential for making long-lasting memories in the brain. A focus of my lab is to understand how and why this gene is so important for information storage. A technician in the lab, Nate Yoder, wanted to study the biochemistry of Arc protein. To do this, we engineered bacteria to produce a ton of Arc protein that we could purify. Nate found, however, that Arc protein behaved strangely. It seemed like it was much bigger than predicted and this was probably because single Arc proteins were clumping or aggregating. Perhaps we were just unable to purify Arc properly. Still, Nate was curious to know what the protein looked like so with the help of Adam Frost he took some images of Arc protein using an electron microscope. This allowed him to resolve Arc protein at very high magnification. Strikingly, instead of clumps of protein, we saw these beautiful “soccer ball” structures (image 1). 

This observation led us down a rabbit hole of unique biology. Turns out these soccer ball structures look just like the protein shells or capsids that viruses make. Why would a neuronal protein form something that looks like a virus capsid?! We are still trying to understand this surprising discovery, but Arc seems to have retained many properties of viruses. Oh yes, and we think that this gene evolved from an ancient ancestor of the retroviruses, like HIV, called retrotransposons. These rogue elements, along with ancient viral infections have left us with bloated genomes comprising of up to 50% of our own human DNA. In some cases, it seems, evolution has used these sequences and repurposed them to create new genes. Wild! 

Back to the process of doing science. Another common fallacy is that science results in black and white answers. In biology, this is rare. Scientists can be wrong in their interpretation of the data. They can be wrong in how they designed an experiment and the results can be messy. The key is replication and figuring out many different ways to get at the answer. For example, another lab headed by Vivian Budnik and work led by her postdoctoral fellow Travis Thomson independently found that a gene that looked like Arc in the fly, also seemed to behave like a virus. When we purified the fly Arc protein, we also saw that it could form capsids. So here we have two examples of different genes having retained the ability to form virus-like capsids! Even more surprising, we think that the fly Arc gene is actually unrelated to the mammalian gene; evolution repurposed a similar kind of retrotransposon in the fly lineage 100s of millions of years after the mammalian Arc gene.

If it happened twice, it probably happened many times more. We and others are on the hunt to find other genes that may have similar properties. Most important for my own research, we want to understand why you need a virus-like protein to make long-term memories. All of this reinforces, to me, the intricate and complicated path evolution has taken that has led to the amazing structure of the human brain. We hope that this science will not only lead to understanding how the brain works, but potentially to applications like gene therapy. Currently, we rely on modified viruses to get gene therapy into human cells but these still elicit an immune response and are often not very efficient. What if we could use proteins that look like viruses but are already made in our bodies? Nature often has the best solutions to hard problems; we just have to figure out how. It takes a team of dedicated people and a little bit of luck to reveal Nature’s secrets. 


TEDMED & Massive Science: TEDMED boasts a proud partnership with Massive Science, a digital science media publication that brings together scientists and the science-curious public. The team at Massive joined us onsite at TEDMED 2018, and by various speakers including Jason Shepherd. Check out their coverage of Jason’s TEDMED 2018 talk: “A protein in your brain behaves like a virus, infecting your cells with memories.