A conversation with Dr. Emma Burrows (OHBM 2024 keynote interview series pt. 6)
Dr. Emma Burrows leads a team aiming to understand how genetic mutations give rise to changes in brain connectivity and behavior across the lifespan, as observed in complex brain conditions such as Autism Spectrum Disorder (ASD) and dementia. Emma and her team use a touchscreen behavioral analysis that allows them to get an impressive readout of how mice solve problems, remember and pay attention. When mice perform the task correctly, they receive a milkshake reward. These tasks are similar to the ones used by neuropsychologists to assess people. Her team’s work aims to bridge the translational divide between how we measure brain conditions in animals and in the clinic.
Her work has attracted significant funding and several awards, including a Victoria Fellowship to train with pioneers of touchscreen rodent cognitive testing in Cambridge, United Kingdom. She is an internationally recognized expert in touchscreen testing who has published work on behavioral phenotyping of mice modeling a range of brain disorders (including Autism Spectrum Disorder, Schizophrenia and Dementia), been invited to speak about her work to a wide range of audiences, supervised over 20 graduate students, and mentored many early career researchers. If you’d like to read a summary of one of Dr. Burrows’s recent works, see our Brain Bites summary here.
Rahul Gaurav (RG): Dr. Burrows, it’s wonderful to have you with us today. Could you tell us about your work, and could you share what specifically drew you to study brains using animal models?
Emma Burrows (EB): I work with mice to better understand complex human brains. The mice I work with get to play computer games that allow them to earn strawberry milkshake rewards. What's important about these games is that they mimic what psychologists use in the clinic to understand how people learn, remember, and pay attention.
I also have a strong interest in understanding complex brain conditions like autism. This was sparked by my early work with an autistic boy during his school transitions. One thing that I found fascinating was how differently he saw the world. His brain was wired differently from mine, and I saw the challenges he faced. But I also saw the many strengths that his differently wired brain provided him, which ignited my interest in brains and my desire to choose neuroscience as I progressed through my science degree.
RG: That’s fascinating! Given your focus on autism, could you elaborate on the specific behavioral or functional aspects that interest you most? Also, why are animal models crucial for understanding these behaviors and functions?
EB: Our ability to orient attention to new stimuli or to maintain our focus on them is just essential for everyday activities like learning. So, an attention difference can completely change the way that we process information. For me, this is the most interesting behavior. Profound changes in the attention system are some of the earliest identifiable features of autism and they could potentially underlie core traits that are commonly used to describe autism, like atypical social communication and the presence of restrictive and repetitive interests and behaviors.
However, attention is a complex behavior—what we put our spotlight on in our world. It's not something you can study if you have neurons in a dish. If you have an atypical attention, it impacts the way you move through life and the way you learn. So, it affects your education and also how you engage in work. Autistic people do not have the same access to education and work. Therefore, we need more research in this area to reduce the burden on them. Mice provide us with a wonderful way to understand this complex attention system. We can either introduce a specific genetic change or change their environment. We could also raise mice with running wheels to look at the effect of physical exercise on the attention system, as an example. These experiments cannot be done in humans. We use mice because they are incredibly similar to us in many regards, and they have roughly 90% of the genetic material we have.
RG: In your paper on Nature Neuropsychopharmacology, you introduced a touchscreen-based Posner-style cueing task. Could you explain how this innovative approach has helped uncover new insights into the attention mechanisms in autism?
EB: My team, including Dr. Shuting Li and Professor Katherine Johnson, have been using a novel touchscreen platform to explore attention orienting in mice with a genetic mutation, the R451C mutation in neuroligin-3, that relates to autism. We have designed and validated a mouse attention-orienting task which is based on the classic human cognitive paradigm called the Posner cueing task. This task uses touchscreen technology to capture nose poke responses on a touch-sensitive computer screen, and mice are required to voluntarily wait at the center of this touchscreen until the appearance of a target either on the left or the right-hand side of the screen. While the mice are waiting, we present a cue, which indicates the location of the target. This enables the mice to either facilitate their response if it's a valid cue or slow down their response to the target if it's an invalid cue. Through changing the feature of this cue, we can test whether mice are using their externally directed attention, which we term exogenous orienting, or their goal-directed attention, which involves internal cues or endogenous orienting. We've shown that mice respond quicker to validly cued targets, similar to humans when assessed with the equivalent task.
This is the first task to assess attention orienting in mice, which enables us to ask mechanistic questions about how this attention system works in both typical and neurotypical brains. My team has used this task to ask questions about autism, in particular, how a specific mutation that has been linked to autism might impact this attention system. Mice carrying the autism-associated R451C mutation in the neuroligin-3 gene have been characterized to show autism-relevant phenotypes in other behavioral tasks. But we are the first to look at attention orienting in this model.
RG: Could you tell us what were the main challenges in the field, and how the touchscreen-based cognitive tests addressed these challenges in your research?
EB: I think the real problem we're trying to solve is that translation from animal studies to humans has been a longstanding challenge in neuroscience. It's clearly multifaceted, but one of the issues is that we're not using the same assessment methods to translate between animals and humans. Touchscreen-based cognitive testing methods were developed for use in rodents to solve this problem. They are purported to substantially improve the translation between preclinical studies and clinical trials.
Clearly, there are some big differences between rodent and human cognitive testing, and I think this is important to note. One of the things that is very different between the two is that humans will come into a room, be given verbal instruction about what their task involves, sit down in a chair, do that task, and be out within an hour. Whereas our animals, who receive no verbal instruction, show up to their task, but we actually need to train them on the very basics of what the touchscreen is. The earliest stages are basic Pavlovian conditioning; we need them to understand that interacting with the screen gets them a shot of strawberry milkshake, which requires a significant amount of training. Once we've got the animal to understand that touching the screen equals a reward, we need to increase the complexity of the stimuli daily. This again can take almost a month with training, which is different from the one-shot human cognitive testing. But other than that, I think touchscreens do offer rodent cognitive behavioral neuroscientists many advantages in that we can show up, put these animals into these boxes, and they drive their own learning. They are not as stressed, and they have a very similar experience; it does reduce a lot of the unknowns.
One of the things that is different about our touchscreen attention task compared to other attention tasks that focus on different subtypes of attention is that our animals have to hold their nose in front of the screen. Knowing that our mice have their nose pointed at the front of the screen means that we know they see the cue, and we also have an accurate readout of their response time, which is a very clinically relevant output. In humans, response times are what researchers care about—how quickly can a human respond to a target? In animal studies, we care about whether an animal gets a target correct or not. This is a reassuring output measure for us because we need to know that the animal can actually do the task. Response times have not been historically very reliable because an animal might be facing away from the screen or might be poking its nose in the area where it gets its milkshake reward when the stimuli appear on the screen. The only other option would be to ask the animal to sit in a chair or to head-fix it. However we believe that our task avoids any additional stress caused by fixating an animal in a location. If these animals don't choose to do the task, they don't get to do it. In fact, we don't actually show them stimuli unless they've shown up and have stopped at the front of the screen.
RG: So in developing animal models, particularly for complex conditions like autism, what criteria do you consider essential to ensure that the model is valid?
EB: An animal is never going to be a human. However, we do use some criteria for understanding if we're getting close to modeling something that is relevant. The first one is to construct validity—can you create something in an animal that looks like the human condition? In the context of an autism mouse model, we look at genetics. To be able to find a slight genetic variation in a human genetic sample and introduce this to a mouse is one of the ways that we model construct validity.
Face validity is about whether it looks the same. In the context of how we measure different autism-relevant behaviors in mice, the touchscreen paradigm is a really great way of ensuring you have face validity because you can conduct the same task in both humans and mice. Do they show the same sorts of behaviors? Can we observe the same thing across species?
The last one is predictive validity—does the animal behave in the same way as humans if given therapeutics? This is something that can be tested in many different human conditions but not so well in autism as there are no therapeutics that reliably work in autism.
RG: Then how can we optimize the use of animal models?
EB: We need to be thinking about the Three Rs principles when we design animal experiments: replacement, reduction, and refinement. We design our experiments to ensure that the small number of animals we use gives us the appropriate power to ask the questions we want to ask. One of the things I am very proud of in my research is that the animals we use in our touchscreen experiments are used for a long time; they voluntarily show up to a certain degree. They become really accustomed to the concept that at the exact time of the day, they will expect to be taken into the touchscreen testing room, they will be able to play their computer game, and get shots of strawberry milkshakes for an hour. Then they get to go back into their cage and rest. Each animal produces lots of different touches on the screen. We are also getting a wealth of data from each individual animal as we can measure response times, learning rates, and more.
The work is also designed to ensure that our animals are treated with the utmost care. We know every single animal's body weight, and we get to know their nuanced behavior. We need to make sure that they are healthy, happy, and willing to learn. An animal that is not healthy, happy, and not willing to learn will sit in the box and go to sleep. This is a consideration of how we optimize animal models through clever experimental design and ensure that we're appropriately powered.
As for animal husbandry, earlier in my neuroscience career I studied environmental enrichment. This involves raising animals in boxes where they have access to running wheels, tunnels, and many complex cognitive challenges to boost their brains and give them opportunities to live a life that is closer to that of an animal that would not live in a cage. One of the things I found in this early work was that this boosts the connections that exist in a key area of the brain involved in learning and memory, the hippocampus. At the bare minimum in our laboratory, we include housing that gives them security and enables them to build really complicated and beautiful nests. In other groups, I have seen more standardized processes that include tunnels and various types of nesting to encourage different behaviors in their animals.
RG: Shifting focus slightly from the specifics of your current research, I'm curious about the broader challenges you've faced. What would you say has been the most significant obstacle in your research career?
EB: I think the key to the team that I have is that we include psychologists, who have no experience working with animals, working alongside myself, who only have experience working with animals. One of the biggest early challenges that we encountered was the completely different language that we used. Our backgrounds are very different, as are our thinking styles and the information we possess. In order to design an animal experiment that is relevant to humans, you need to ensure that you're measuring the same thing, not just through the task design, but also through the history of how that thing has been measured in the human field. We've just undertaken a systematic review to compare rodent tasks to human tasks. We noticed that people weren't even using the same terminology when they were talking about the type of attention they were looking at. Additionally, the task designs weren't even close to the relevant task designs that human researchers were using. I think careful task design comes from breaking down those barriers that exist between different disciplines within neuroscience. And that, I think, itself requires us to be willing to be uncomfortable because miscommunications cause a degree of human discomfort. The key to my collaboration with Dr. Shuting Li and Professor Katherine Johnson has been a willingness to sit down and break down those communication barriers. That's how we've achieved the best animal research.
Nowadays, there are a lot of people undertaking really sophisticated cognitive behavior research. It requires a lot of work to get animals to perform such wonderfully complex behaviors. However I'm hoping that the research I'll be talking about in the keynote will inspire others to break down barriers and collaborate across disciplines. It's only when you actually step outside of your discipline, share with others and work together that you see the benefits. It's slower, but I think it produces better results. So, I'm hoping that this style of work will become more commonplace.
RG: Reflecting on the people who have shaped your journey, could you share who has been particularly influential in your career?
EB: I'm going to start with every graduate student I have worked with. I have learned so much from people who are less experienced in terms of years than me. So, a big thank you to every student I have supervised and also mentored along my journey. I think that the bidirectional exchange of ideas and growing as individuals as you go through your research career has been one of the most uplifting experiences I've had as a researcher.
If I had to name one person I've come across, other than all the giants in neuroscience and the multiple people whose shoulders we stand upon, it would be a close personal friend and one of my first mentors, Professor Julie Bernhardt. Julie is a world-leading stroke rehabilitation expert. I have learned a lot from her in the way that she leads. Julie has not only inspired through setting new standards in her own field and being excellent and accountable, but she's also brought a lot of people along with her. Her ability to collaborate across different countries and disciplines has been utterly exciting for me to watch. I cannot believe the impact she's had not just on the field, but also on other people within the field. Her generosity in giving and also stepping aside when it's appropriate has been remarkable. I've seen her be involved in the genesis of an idea or a project and then step aside to let someone else lead. That's a really wonderful way of leading in research, because when we can remove our egos from the work, then we can start focusing on the problem, which is actually to help people and to make people's lives easier and less burdened.
RG: As we wrap up, what advice would you share with aspiring researchers to help them navigate their careers?
EB: I met Fred, also known as Rusty Gage, early on in my career. Fred Gage is best known for challenging the longstanding dogma that the creation of new neurons, or neurogenesis, does actually occur in the human brain. I met him when I was first becoming aware of the difficulties of sustainably managing my place in a highly competitive field. His advice was to “follow your nose—when you are interested, it's easier to work hard.” I understand that appropriately resourcing the work you do and thinking ahead is clearly important for planning impactful research. However, ensuring that my focus has always been on the fun of doing the work and the people I'm working with, rather than solely on achieving the goal, has really continued to give me energy. I think it's also energized others I've worked with as well. So, that concept of focusing on the process, as well as the reason why you're there—that interest and the meaning behind the work—is crucial.