Dr. Gordus, thank you for taking part in our OneNeuro interviews. Could you summarize the research of you and your lab?
We’re a behavioral neuroscience lab in the Department of Biology, and we’re interested in how behaviors are encoded in brains. In our research, we use animals with very small brains, such as invertebrates, specifically Caenorhabditis. elegans, a small nematode, and Uloborus diverus, which is a hackled orb-weaving spider.
Are you creating new knowledge or working to solve an existing problem?
Our work with invertebrates is basic biology, we’re trying to answer fundamental questions. The advantage of using small invertebrates is that they have a smaller number of neurons, yet many of the genes they use to encode information are present in all animals. The worms, for example, have 302 neurons in their brains, and they have a very stereotypical development. That means every worm is close to having the same brain, but they are all individuals and make individual decisions. We are curious to understand what influences notions of individuality and how internal states drive our interpretations of sensory information and our responses to the world. We can think of these internal states as arousal states or emotions, colloquially.
We typically don’t assign terms like emotions to invertebrates, but they can have different arousal states, such as fear or satiety or hunger. We know that these states are influenced by compounds in their brains called neuromodulators, and many are very conserved amongst all animals. We are curious to understand how these states are encoded. With the worms, they have a defined number of neurons, which makes it a tractable system to know, explicitly, how these modulators affect specific neurons, and how these neurons in turn encode behavior. Arousal states can be difficult to infer because we cannot ask the animals how they feel. Instead, we have to infer them based on behavior, but behaviors can be very probabilistic. For example, a cat will lick its paw when it’s content, but cats lick their paws all the time, so that behavior is not an absolute definition of being content. Instead, the probability of observing this behavior is higher if the cat is content.
However, when it comes to animal architecture, many animal architectures are built in phases. Structural integrity is a reflection of behavioral integrity. Animals organize their behaviors over a long spatial, temporal scale to construct something effectively. So, the advantage there is that they are leaving behind a physical record of their intent. I don’t have to ask the spider, “Why are you doing this?” I know because I can see that they’re building a specific structure. So, we leverage that information to understand how these different structures are encoded in the brain. That’s the end goal: To understand how the structure is a physical manifestation of these changing internal states.
Does any of your research involve 3-D structures?
That is the nice thing about modeling an orb web – it’s a two-dimensional, highly regular structure. A lot of animal architecture is very three-dimensional with complex geometries. An orb has defined stages, such as one that we identify as the so-called proto web.
View an orb web (VIDEO):
The hackled orb-weaving spider has long anterior legs and a distinctive way of sitting on their web with their legs together. We use this species because they are easy to work with. These tiny spiders lack venom glands and can be housed together since they are not cannibalistic.
What we see here is the proto web stage, sped up 100-fold. These spiders build in the dark, so we use infrared light. The proto web stage is dynamic; spiders put down, remove, and adjust lines randomly, spending hours in this stage. They also pause for long periods, likely assessing structural integrity. If disturbed, they abandon the proto web, indicating a bad place to build.
If the spider likes the spot, she adjusts the proto web, making it more radial, adds a frame, an auxiliary spiral, and then a capture spiral. The capture spiral is adhesive to trap prey, and she removes the temporary spiral as she builds it.
In our research, we use videos and machine vision tools to identify behaviors and associate them with web-building phases. Spiders organize behaviors differently in each stage, and we aim to understand how this is encoded in their brains. The web is an elegant, two-dimensional structure. Spiders build blindly but seem to know where they are. They can exit a web phase, explore, make adjustments, and return to where they left off, showing path integration. This ability, found in many animals, seems to have evolved independently in arthropods and vertebrates.
What led you to go into this field of research?
When I was in grad school, I worked in a chemistry lab on biotechnology development and got interested in systems biology. I read a fascinating paper by Michael Elowitz at Cal Tech and became intrigued by information and biology. Chemistry is inherently entropic, but our cells and brains have evolved structured responses to the world, largely due to network topology. In grad school, I focused on cell signaling but wanted to apply these ideas to the brain.
I pursued a postdoc with C. elegans because it had a full neuronal map. I thought we could apply graph theory to a brain, but it wasn’t simple. I treated the neural network as a passive computer, but the brain is dynamic. The worm is constantly thinking, regardless of sensory input. My postdoc work ended up being about how network states influence behavioral responses to sensory input. The worm is always thinking, and so is every brain.
I saw an orb web in Central Park and was struck by its beauty. I wondered how it was encoded in the spider’s brain. The web is elegant, created by an animal with a brain no bigger than a fruit fly’s. I read behavioral and neuroscience research on spiders and contacted researchers. Spiders are difficult to work with, being cannibals and only active for a few months a year. I read about hackled orb-weavers by William Eberhard, who mentioned they were easier to work with. They are small, can be raised as a group, build indoors, don’t have venom glands, and are active year-round. They belong to the Uloborids, a small, atypical family of orb-weavers. I care about the behavior, regardless of their relation to other spiders.
The species I use is native to California, and both my dad and sister are wildlife biologists there, so I thought they would be able to find them quite easily. Ironically, it was my brother, who works in water treatment, who found several in a greenhouse in California, and he mailed them to me in little cheese containers from a pizza place.
That was the beginning of my foray into research with spiders. In fact, I had just started this work when I came to Johns Hopkins in 2016. When I interviewed here, it was all about my worm research, but during the last 10 minutes of my talk, I changed directions and talked about my possible research on spiders. I originally thought of the spider work as being a side project, but we were able to make great progress through the efforts of several outstanding undergraduate and graduate students who worked on the assays and the genome assembly. Our success is really a testament to the creativity and persistence of these students since most of our experiments have been built from scratch.
Who are your collaborators in this research?
We have a variety of collaborations at Johns Hopkins. We collaborate with Chen Li in the Department of Mechanical Engineering. We’re interested in understanding how spiders dynamically sense vibrations in their web, and Chen has built a spider robot that he’s using to test the observations that we have. He is helping us to gain an understanding of how the spider dynamically senses different frequencies on the web; how it locates prey. I have a collaboration with Conor McMenamin at the School of Public Health that is a mosquito project. We are interested in the volatile compounds from humans that attract mosquitoes, and we work together on two projects. One is doing field studies in Zambia, testing mosquitoes in their native habitat. And the second is building a pipeline to make it easier to image brain activity from invertebrates. We’ve developed a laser microsurgery setup that makes it easier to dissect invertebrate cuticle, which is typically very difficult to do because of their extremely small heads.
What do you recommend to students interested in research like yours?
Learn how to code. Much of our work involves programming in some sense. Behavioral and genomic work involves a lot of coding. Even the worm where we record from neurons and quantify the responses involves coding. Knowing how to code greatly expands the projects an undergrad can undertake, with the least amount of training upfront. And, depending upon their abilities and interests, they may work on more independent and research-focused problems through their time in the lab. Currently I have a few seniors in the lab. One is doing experiments with spiders to see how they handle errors or damage to the web. Another one is assisting a graduate student in looking at how or if there is intergenerational transfer of behavior in C. elegans.
What do you enjoy doing in your spare time?
I enjoy the outdoors. I’m from Oakhurst, California, just south of Yosemite National Park, and I grew up hunting, fishing, hiking, camping. I still do all of that. Also, my son is in the Boy Scouts, and I enjoy being outdoors with him. My dad is a wildlife biologist, and really instilled a love of nature in us.
Anything else you would like to add?
Spiders are our friends. There is a paper that came out a couple of years ago in Current Biology about the spread of misinformation, and the topic they investigated was spider myths because those are some of the most common myths out there. There are a lot of sensationalist anecdotes about spiders that are almost all false. Most “spider bites” are insect bites, or bacterial infections. I don’t mind spiders in my house because they eat the insects I don’t want in my house, like ants, termites, and flies.
Please contact Dr. Gordus if you have questions or comments. He may be reached by email: [email protected]
Above Image: Andrew and son fishing / Andrew Gordus
Above Image: The web of Uloborus diversus – This is a photo of a female Uloborus diversus (hackled orb-weaver) hanging from her web. The image shows three of an orb web’s structures: radii, spirals, and the fuzzy stabilimentum. The other two web structures that are typically built (the proto-web and the auxiliary spiral) are transient features that are built during the process of construction, but are removed and not part of the final structure. / Andrew Gordus