Taking a Direct Route: A Novel Explanation for Migraine with Aura

Cerebrospinal fluid can directly activate neurons in the trigeminal ganglia, according to an animal study.

What is the origin of migraine headache pain? It’s a question that the field has long grappled with, and in doing so placed the focus on trigeminal neurons in the meninges. For migraine with aura, the idea is that cortical spreading depression (CSD), a wave of electrical activity across the brain that is considered the physiological correlate of aura, sensitizes and activates those neurons in the dura mater, one of the membranes that surrounds the brain.

A new study shows, however, that researchers should also consider an additional mechanism of pain: Direct communication between the cortex and the trigeminal ganglion, a peripheral structure underneath the brain, at the base of the skull, that contains the cell bodies of trigeminal neurons.

Using a CSD model of migraine in mice, researchers led by Maiken Nedergaard, University of Copenhagen, Denmark, and University of Rochester Medical Center, US, and Martin Kaag Rasmussen, University of Copenhagen, show that cerebrospinal fluid (CSF) flows directly to the trigeminal ganglion from the cortex. Then, certain proteins at elevated levels in the CSF – not only the familiar calcitonin gene-related peptide (CGRP) but other signaling molecules as well – directly activate nociceptors in the ganglion.

The identification of this novel pathway not only contributes to the understanding of migraine pathophysiology, but has clinical implications, too.

The researchers “show in a mouse model of migraine that after CSD, subarachnoid CSF carries signals from the cortex directly to cell bodies in the trigeminal ganglia, where they activate nociceptors through a pathway that bypasses meningeal trigeminal afferents,” wrote Andrew Russo, University of Iowa, Iowa City, US, and Jeffrey Iliff, University of Washington, Seattle, US, in an accompanying commentary. “The demonstration that the trigeminal ganglia lies within the blood-brain barrier and the identification of the signals that connect aura and headache may provide a new path for preventing and treating migraine.”

The research and commentary appeared in the July 5, 2024, issue of Science.

A change of plans
Rasmussen, the first author of the study, told Migraine Science Collaborative in an interview that, at the start, his focus was in fact on migraine’s tried-and-true pathophysiological mechanism. But that approach to the research didn’t pan out.

“Our initial plan was to study trigeminal afferents in the dura mater,” he said. “We were also interested in whether meningeal lymphatic vessels and the drainage pathways around them would present CSF-bound molecules to the dural afferents, so I spent a long time looking at these afferents, and we just didn’t see a lot of stuff happening up there. We were also having a hard time recording any activity in the dural afferents.”

A key result from experiments using a tracer to track the flow of CSF suggested that it was time to move on to another anatomical location.

“We then compared CSF concentrations in different areas of trigeminal innervation, and, to our surprise, we found a much higher concentration of CSF tracer in the basal skull compared to the dorsal skull,” referring to the bottom and top of the skull, respectively. “And then we just continued with this new, unknown signaling pathway towards the trigeminal ganglion.”

Does it get in – and how?
That story would move forward with experiments that first determined whether CSF could directly access the trigeminal ganglion. Several techniques allowed the team not only to follow CSF flow into the ganglion, but also to measure associated neuronal activity, in live mice.

They confirmed that CSF could enter the extracellular space around the trigeminal ganglion, and this was accompanied by an increase in the activity of neurons in the ganglion. That suggested that substances in CSF (known as CSF solutes) could directly activate trigeminal neuron receptors.

From where could CSF travel to get to the ganglion? It turned out that CSF containing protein made in the cortex of the brain reached the ganglion, revealing a route from the central to the peripheral nervous system.

CSF mainly entered the ganglion at the root of the trigeminal nerve – but how was that even possible in the first place? That was a relevant question since peripheral nerves outside of the cranium are usually enveloped by a connective tissue sheath with a tight barrier that prevents external solutes from getting into the extracellular space around the nerves. But the portion of the trigeminal nerve nearest to the nerve root lacked this barrier, unlike regions of the nerve farther away, which did have the barrier.

Other experiments showed that CSF could enter the extracellular space of the ganglion through transport from the inner subarachnoid space (a compartment between the arachnoid mater and pia mater in which CSF flows). It could also enter via brainstem tissue.

From the visual cortex to the trigeminal ganglion
To better understand how CSF solutes from the central nervous system enter the trigeminal ganglion and play a role in migraine with aura, the investigators turned their attention to the visual cortex, since that area is the most common site of migraine aura.

The CSF surrounding the visual cortex flowed mainly to the ganglion rather than to the dura mater – and particularly to the ganglion on the same side of the cortex (the ipsilateral ganglion) where the researchers had injected the CSF tracer allowing them to follow the CSF flow pathway. That might explain a commonly recognized feature of migraine, according to Rasmussen.

“To our surprise, the CSF flow pathway goes to the same side. So when a protein is released in the right cortex in response to CSD, there is a much higher concentration in the right trigeminal ganglion compared to the left trigeminal ganglion. That’s really interesting because migraine is most commonly a one-sided headache, and this distribution pattern of cortically released proteins could explain why,” Rasmussen said.

Further, it took about a half-hour for the tracer injected in the visual cortex to appear in the CSF in the ganglion. That was in the ballpark of the typical delay between aura and the subsequent headache in people, which ranges from about 5 to 60 minutes.

Changes in proteins
With knowledge of the anatomic pathway from the visual cortex to the trigeminal ganglion, the researchers next wanted to assess whether that pathway actually played a role in migraine with aura. Mass spectrometry – a technique enabling the identification and quantification of molecules in samples, in the current case samples of CSF – was put to use for this purpose.

The team discovered that CSF from adult mice exposed to CSD showed alterations in about 155 proteins – that was a change of about 11% of all the proteins (1,425) present in CSF from control animals not exposed to CSD.

Matching their data against a ligand-receptor database, the researchers saw that, of the entire complement of proteins in CSF, 65 were proteins that could bind to 143 distinct receptors in the trigeminal ganglion. Most of these interactions were stimulatory rather than inhibitory.

Looking more specifically at proteins at higher levels in animals subjected to CSD, the investigators identified 12 proteins capable of interacting with 28 distinct receptors in the trigeminal ganglion. One of those proteins was CGRP, but the other 11 are not as well known and are, Rasmussen said, ripe for further investigation.

“What we think is most interesting is to continue with these 11 proteins and describe them in more detail, and determine whether they can provoke a migraine phenotype in animals. And if we find that one or several of them, or perhaps a cocktail of them, can induce a strong migraine phenotype, then I think we would like to see if we can do human provocation studies,” he said.

“These 11 proteins could also become a drug target one day,” he added. “Even in a clinical setting, if you had CSF samples from migraine patients during an attack and before an attack, maybe you would be able to identify, hopefully, the same molecules that we see or perhaps even other molecules,” Rasmussen said.

Finally, the receptors for the dozen proteins are present not only in sensory neurons but in other cell types, including vascular cells, fibroblasts, and Schwann cells (a type of glial cell).

Let’s get excited
Could the changes in CSF proteins after CSD affect the activity of trigeminal sensory neurons? Final experiments addressed that important question, with the group collecting CSF from mice exposed to CSD for an hour and then applying it to the exposed trigeminal ganglia of animals that had not been exposed to CSD.

There was an increase in the activity of trigeminal neurons exposed to CSF from the CSD animals. This was not the case, however, with CSF collected from animals 2.5 hours after the first CSD.

The researchers say this suggests that the protein changes after CSD are short lived, and so those changes may be responsible for the initiation of migraine, but not its maintenance.

An additional experiment showed that CGRP was one of the proteins in CSF that could increase trigeminal neuron activity.

A new song, but the classic tune still applies
The mass spectrometry method and the results it provided are what stand out the most to Rasmussen.

“I’m excited both about the molecules that we identified but also this whole toolbox that we have developed,” he told MSC.

Rasmussen also said that the current results do not imply that meningeal mechanisms of migraine are not important.

“We’re not trying to refute what is happening in the meninges. We do also see that this CSF and all the pain-provoking molecules also go to the dorsal dura mater, but just in much lower concentrations. But, in the meninges, you can also still think that these same proteins can start processes that could be responsible for the headache, longer term, for instance, through mast cell degranulation, which has been studied in the past.”

The commentators expressed a similar sentiment.

“It seems likely that communication via both the direct CSF-borne and the classic trigeminal afferent pathways co-occur and contribute to peripheral sensitization in migraine with aura,” wrote Russo and Iliff. “A key implication of these observations is that the trigeminal ganglia as well as the meninges should be considered as a target for peripherally restricted drugs that are used to attenuate pathologies initiated in the cortex.”

Notably, the new mechanism for migraine with aura could play a role in the non-aura subtype, too.

“One thing that makes us believe that this could also be relevant in migraine without aura is that there are a few studies out there showing that CGRP is also increased in cerebrospinal fluid of patients with migraine without aura,” Rasmussen said. “And we have some MRI images of humans in our paper that show that the fluid pathway likely is also similar, and that this lack of a barrier around the root of the trigeminal ganglion is likely also present in humans.”

In terms of future research, Rasmussen said that mass spectrometry could be used to look at other types of molecules in CSF, including very small peptides, and lipids, which the current study did not pursue.

In the meantime, while another recent animal study also reported changes in CSF flow as a contributing factor to migraine, in that case using the nitroglycerin model of chronic migraine (see MSC related news), the role of the glymphatic system – the system for waste clearance in the central nervous system – in migraine is still a nascent area of research.

“Although this [current] work provides some of the strongest data to date for a role of the glymphatic system in migraine, there is much to discover about fluid and solute transport in the brain. Perhaps the journey toward understanding the role that glymphatic function and dysfunction play in a wide range of neurological and psychiatric conditions has only just begun,” wrote Russo and Iliff.

Neil Andrews is a science journalist and executive editor of the Migraine Science Collaborative. Follow him on X (formerly Twitter) @NeilAndrews

Image credit: 123RF Stock Photo.

Reference
Trigeminal ganglion neurons are directly activated by influx of CSF solutes in a migraine model.
Rasmussen et al.
Science. 2024 Jul 5;385(6704):80-86.