A Serendipitous Path to the Study of Microglia: A Conversation with Michael Salter

Editor’s Note: The third North American Pain School (NAPS) took place June 24-28, 2018, in Montebello, Quebec, Canada. This educational initiative brought together leading experts in pain research and management to provide 30 trainees—part of the up-and-coming generation of pain researchers—with scientific education, professional development and networking experiences.

Six of the trainees were also selected to provide first-hand reporting from the event, including summaries of talks presented at the meeting. Here, Donald Iain MacDonald, summarizes a talk delivered by Michael Salter, a pain researcher at The Hospital for Sick Children in Toronto.

What got you interested in studying pain?
I can date my interest in pain back to precisely one lecture in medical school during my second year, in our neurophysiology section, at the University of Western Ontario. The talk was on the gate control theory of pain. Immediately, I thought that this was extremely strange, and I wanted to learn more about it.

A couple of weeks later, Ron Melzack, co-creator of the gate control theory, happened to be in town, and I went to his evening lecture. There might have been 15 people there, and Ron gave one of his amazing talks about what pain is like. He showed incredible photos of people undergoing trephination or hypnosis, and I thought that I had to learn more about this. From then on, I was really hooked!

Did you practice medicine before pursuing basic research?
No, I decided partway through medical school that I wasn’t really cut out to be a clinician. I had already decided that what I was really interested in was the cellular and molecular aspects of pain, and had already agreed to go do a PhD with Jim Henry at McGill University immediately after medical school. So I never got a license to practice.

What did you work on for your PhD at McGill?
At the time, we were in the heyday of trying to understand synaptic transmission and synaptic mediators. Jim Henry had started to popularize the idea that substance P was an important transmitter for pain, so much so that many people even to this day think the “P” in “substance P” stands for pain—which, of course, it does not!

I was interested in going to McGill because Jim had been doing amazing recordings from single neurons in the spinal cord of the anesthetized cat. It was fascinating that you could record the activity of individual nerve cells, and then squirt agents onto the cells to change their activity and try to understand how they were functioning.

You work on microglia and sex differences now. Why did it take so long for people to recognize how important microglia are?
It’s not just microglia that were relegated to second-class citizens—it was the whole area of glia. For the history of neuroscience, the major focus, and maybe rightly so, has been on neurons— on the electrically excitable cells, or at least cells that you could envisage being electrically excitable. This goes all the way back to Galvani and his experiments with the frog leg; you could stimulate the leg electrically and it moves. Obviously it’s the electrical transmission of information that’s important.

So one can understand why there was this focus on neurons. Then, Virchow had the idea that the non-neuronal elements—the glia—were like glue. “Glia” is actually Greek for “glue.” It’s like thinking about the bricks and mortar in your house. Some people are interested in that, but most people are interested in what goes on in the house. That was part of the reason that glia had been largely ignored. But we now know that glia are much more than just glue!

What got us thinking about microglia was really a serendipitous observation. A number of us, including Kazuhide Inoue and Makoto Tsuda in Japan, were interested in the potential role in pain of purinergic signaling in the spinal dorsal horn. Kazuhide and Makoto observed that a specific subtype of purinergic receptor, called P2X4, was necessary for pain hypersensitivity after peripheral nerve injury. Remarkably, in the dorsal horn these receptors were expressed exclusively on microglia. So microglia cells had to be involved in pain hypersensitivity!

Then we discovered that depleting microglia, or inhibiting their signaling to dorsal horn neurons, could abolish mechanical hypersensitivity due to nerve injury in male but not female mice. Thus, pain hypersensitivity is dependent on microglia in males, but independent of microglia in females. So there is a profound sex difference at the cellular and molecular level, even though the degree of pain hypersensitivity in females is not different from that in males.

It was this one singular observation that really changed my view of the role of these cells in the context of pain. And now the evidence for microglia involvement in pain has increased quite dramatically over the last few years. The number of publications is sort of hockey sticking, to use a Canadian metaphor.

Where is the microglia field headed? Is there a lot of hype?
The subtleties of microglial signaling are really starting to become apparent, so it’s not hype. But undoubtedly we’re in this rising phase where there will be enhanced expectations and perhaps some misunderstanding. For instance, there’s still this persistent misconception that microglial cells are either on or off. But they’re on all the time in their normal surveillance mode.

The way I think about these cells is that they can adapt in a variety of different ways, some of which are quite easily seen by changes in the number of the cells or their morphology. But there are other, probably more subtle, ways the cells can be involved in physiological or even pathological changes that you might not see if you just look at morphology.

We’re now really trying to understand how signals the microglia receive are interpreted, processed and how those cells communicate with neurons, astrocytes, and other cells.
I think we haven’t hit the peak yet, but undoubtedly it will go up, and then come back down to some steady state level.

Across your career, what was your favorite experiment? Was it the “eureka” moment you had about microglia?
Actually, one of my favorite experiments was one that I did as a grad student, when we were studying the spinal gating of pain. I had been doing electrophysiological recordings in the spinal cord of anaesthetized cats, and had come up with this idea that inhibition of dorsal horn firing by inputs from large fibers came about through adenosine-mediated inhibition of the dorsal horn cells.

One night I was there alone in the McIntyre building at McGill University. I was using the only available adenosine receptor blockers at the time, theophylline and caffeine. I injected caffeine intravenously into a cat to see if I could block or affect the inhibitory response I’d been working on. The response went completely away in a couple of minutes. And I waited. And then the inhibition very slow returned. I was literally running down the hallway jumping, yelling ‘yay!’

What are your thoughts on one of the NAPS debate motions this year—are we going to discover a silver bullet for pain in the next 30 years?
My perspective on that has really changed. Starting out as a biological scientist, but medically trained, I thought, yes, maybe there will be one treatment. In the pain field, we were led to that view by the apparent success of opioids and nonsteroidal anti-inflammatories, where large portions of the population had some pain relief from those drugs.

For me, chronic pain is a disease of the nervous system. And as a disease it’s likely to be like cancer; it’s not a single disease. Even if you’re talking about one type of cancer, the molecular architecture of the changes that occur is different across different individuals.

I really don’t think there’s any reason to expect that pain is any different, and we’ve got evidence for that now. The microglial sex difference is a clear example where degenerative mechanisms do occur and at least two different broad categories of mechanisms—the immune-microglial side of things and the neuronal side of things—can drive the same behavior.

But we shouldn’t necessarily throw out the idea that there may be other ways to inhibit pain in broad segments of the population. In our model, at least from the sex difference perspective, the molecular differences are on the immune-microglial side, and the similarities are all on the neuronal side. So there is at least a possibility that one could look at the neuronal side for a silver bullet. But these things are not mutually exclusive. In the end, I don’t think we should put all our eggs in one basket.