The Role of Animal Models in Migraine Research: A Conversation with Andy Russo

“I agree that there needs to be more crosstalk between researchers and clinicians. As a basic science researcher, I feel it’s important that clinicians keep us grounded; it’s really easy to get sidetracked by interesting nuggets of data, but clinicians need to ground our knowledge in what actually happens to patients. One important clinical aspect of migraine we need to look at more carefully in basic research is variability among patients – why do symptoms differ among different people, and how can we use mouse models to understand this?” – Andy Russo

Andrew Russo, PhD, is a professor of molecular physiology and biophysics, and a professor of neurology, at the University of Iowa, Iowa City, US. He studies how neurons respond to changes in their environment, with a particular focus on regulation of calcitonin gene-related peptide (CGRP) and its actions in migraine.

In this Migraine Science Collaborative interview, Russo chats with freelance writer Fred Schwaller about his career, the role of animal models in migraine research, and much more. The interview has been edited for clarity and length.

Let’s start at the beginning: How did your career in migraine research get started?

I did not start out studying migraine. I started out as a biochemist interested in how gene expression responds to changes in the environment. One gene I started studying back in the 1980s was CGRP, but at that time I had no idea if it had a known role in migraine. It all started after a medical student joined my lab and said, “Andy, did you know that CGRP is elevated in migraine?”

Initially, I didn’t think much of it, as a lot of things change with migraine, but then he told me that triptans bring CGRP levels back down. That’s when I started getting interested in CGRP and migraine.

I had a postdoc in my lab at the time, Paul Durham [see MSC interview with Durham]. He was studying the transcription factors that control the CGRP gene, but we were hitting our heads against the wall on this project. So Paul did a complete 180 on his project and switched over to migraine, looking at how sumatriptan is decreasing CGRP release in migraine. We kept plugging away at gene regulation in migraine, which worked out really well.

Andy Russo

How did your previous training in biochemistry help you think that you could advance the migraine field?

I trained as a biochemist in my 20s. It helped me come into the migraine field through the back door. In the mid-2000s, I went to a neuropeptide meeting in the winter at a ski resort. When I was on a ski lift, I started thinking that we need to study CGRP because it’s important in migraine, but we have no idea why it’s important or what it’s doing. I thought I could make an advance here. From my biochemistry background, I figured that migraine patients must be more susceptible to CGRP, possibly because they have more receptors for CGRP.

So how did you study this?

I wanted to model this in animals, and because I had experience with transgenic mice, I decided to genetically engineer a mouse line that is more sensitive to CGRP by increasing expression of the RAMP1 subunit of the CGRP receptor [RAMP1, or receptor activity-modifying protein 1, is a protein necessary for the binding of CGRP to its receptor]. We purposely planned the mouse so that we’d have flexibility to target expression of CGRP receptor subunits to different parts of the body using different gene promoters [DNA sequences where gene transcription is initiated]. We got darn lucky it worked so well.

But then we thought, How do we tell if a mouse has a migraine? I was always interested in sensory perception since my graduate studies on bacterial chemotaxis, and it seemed to me that migraine is a hypersensitivity disorder, especially with sensitivity to light [photophobia]. From a naive perspective, measuring light/dark preferences seemed straightforward. We did a light/dark assay and found that mice with overexpression of RAMP1 had photophobia, and this was magnified when the mice were injected with CGRP [see here and here]. Later we found the mice also had pain hypersensitivity in response to CGRP [see here and here].

If I had known more about mouse behavior, I may never have made this big leap in my lab’s direction. I was fortunate to have a great lab team willing to take risks and faculty colleagues willing to help educate me on the nuances of mice.

I was presenting the behavior work at conferences and apologizing for the variance in the data. As a biochemist, I was used to the 5% variance standard, but people told me they believed my behavior data because it was so variable. While at present this variance is annoying, I believe that in the future it will be a key to unlocking epigenetic mechanisms that determine susceptibility and resilience to sensory stimuli.

What do you think are your most important contributions to the migraine field?

The first thing that comes to mind is that I feel so very fortunate that my lab’s work has helped lay the foundation for the CGRP-based drugs that are helping so many people. But for just my lab, I think that my most important contribution was showing that CGRP can cause photophobia to dim light in transgenic CGRP receptor mice but not in wild-type mice, while in bright light, CGRP made both transgenic and wild-type mice sensitive to light.

Extrapolating from mice to people, this suggests that we all have the potential to have a migraine symptom, photophobia, from just a small tip of the scale – that is, twofold more RAMP1 in transgenic mice. I think that is a good reminder that we are all on a spectrum of a disease state, or as I tell my first-year medical student class when discussing physiology and genetics, we are all mutants.

Why are animal models so crucial for understanding migraine?

That’s an important question. Translatability of mouse studies has a poor track record, especially in the pain field. To avoid this, we build our studies based on human clinical data, and then dive into mechanisms in animals. Modeling migraine in animals has advanced the field significantly by telling us how CGRP is causing migraine. We have sophisticated tools to engineer mice in order to answer questions that are in no way possible [to address] in humans, especially in fields like epigenetics.

How well do the animal models of migraine recapitulate migraine in humans?

Will we ever know if a mouse has a migraine? I don’t know, but what we can know is if mice exhibit migraine-like symptoms. There are now many different mouse models of migraine, all of which have strengths and weaknesses.

For example, there are animal models of migraine that come from injecting neuropeptides like CGRP or inflammatory substances. These models are good for replicating CGRP sensitivity in humans, but a weakness here is that people don’t normally go around injecting themselves with CGRP to trigger migraine. There is also the cortical spreading depression model of migraine aura. However, one must question whether this is relevant to migraine without aura.

There’s no perfect model. We see differences among models, but that’s not a bad thing – it might even tell us something about the heterogeneity of migraine we see in patients. As long as we remain grounded in the fact that migraine is heterogeneous and use animal models to understand the mechanisms of different aspects of migraine, the models all have great potential. Clinicians and basic scientists need to start pooling resources on this topic.

How can we facilitate crosstalk between basic science researchers and clinicians to better model migraine and improve translatability?

I agree that there needs to be more crosstalk between researchers and clinicians. As a basic science researcher, I feel it’s important that clinicians keep us grounded; it’s really easy to get sidetracked by interesting nuggets of data, but clinicians need to ground our knowledge in what actually happens to patients. One important clinical aspect of migraine we need to look at more carefully in basic research is variability among patients – why do symptoms differ among different people, and how can we use mouse models to understand this?

How do we best match animal models to clinical research problems?

The American Headache Society [AHS] has a committee working on that exact question right now and will report on this at the annual meeting in Austin this summer. Stay tuned….

You’re on another AHS committee focused on bringing more people into the migraine field. What are you discussing?

Aside from a paucity of dedicated funding for migraine research, one of the challenges we have in the field right now is getting more people into it. We’ve got very low numbers of people doing research in migraine, especially basic research.

One way to change this is to find ways to give research time to clinicians. At the American Headache Society, we’re starting a mentoring program where early clinical investigators and even established clinicians can get advice about research topics in migraine. The idea is that clinicians can come in and say, “Here is a clinical problem, and this is what I’m thinking of doing.” Then more experienced researchers can say, “That’s a great idea. Have you thought about X and Y? And here’s what you need to do next to make this a strong research proposal.” At the moment we’re focusing more on neurologists, but maybe we should be more broad-minded about other disciplines like endocrinology, gastroenterology, or obstetrics and gynecology.

As far as bringing in more basic scientists, this is tied up with removing the stigma of migraine for the general population. And I really think that the CGRP story is going to help a lot with that. Because then, basic scientists can realize that there’s something tangible here – something biologically meaningful to focus on in a complex disease state that is more than just a headache.

What are the biggest questions for the migraine field going forward?

Epigenetics is a really exciting avenue we can use our migraine mouse models for, looking in particular at epigenetic changes that underlie sensitivity to sensory input in migraine. This just isn’t possible at the moment in clinical studies, but animal models will hopefully tell us more about this. And the sky is the limit in terms of looking at how migraine changes the brain in all the different pathways associated with pain and hypersensitivity to all sensory input, both external input such as light and sound, but possibly even internal input [interoception].

Second, I really think that big advances in the field are going to be made with induced pluripotent stem cells [iPSCs]. The technology is not quite there yet, but it will come. I’m excited about the potential of working with clinicians to study receptors in iPSCs from different patients. For example, we can study the molecular differences between patients who do or don’t respond to particular migraine treatments. The great thing is that this has tremendous translational potential.

Fred Schwaller, PhD, is a freelance science writer based in Germany. Follow him on Twitter @SchwallerFred