Despite decades of research, scientists still don’t know all the types of neurons that transmit the sensation of mechanical pain. Now, researchers add one more piece to the puzzle, by discovering a new population of specialized mechanosensory nociceptors.
Using an intersectional chemical-genetic approach in mice, a team led by Alexander Chesler, National Institutes of Health (NIH), Bethesda, US, identifies a new class of cutaneous sensory neurons with an anatomy commonly associated with touch neurons, but with response properties resembling A-delta pain-sensing neurons. With a combination of in vivo calcium imaging, electrophysiology, and optogenetics, they find that these neurons are high-threshold mechanoreceptors (HTMRs) whose circumferential endings wrap around the base of individual hair follicles and respond to hair pull. The discovery of these neurons lays the foundation for exploring their function in chronic pain.
“It’s a really interesting study because the field has long assumed that nociceptors make free nerve endings in the skin and don’t innervate specialized structures, but these fast-conducting pain neurons actually do by wrapping around guard hair follicles in the mouse’s skin,” says Cheryl Stucky, Medical College of Wisconsin, Milwaukee, US, who was not involved in the study.
The research was published online August 16 in Neuron.
A new technique for an old problem
Somatosensation is the sensory process that allows us to feel the outside world, whether a cool breeze, a painful bee sting, or the warmth of a fire. Accordingly, a vast range of potential stimuli must be encoded.
“How do you depict the chaotic gentle swirl of a summer breeze?” asks Chesler. “It’s an incredibly rich signal that we’re extracting information from and that’s being used for discrimination of the input as well as for protection and social cues."
Given the sheer number of stimuli, researchers have struggled to identify and categorize all the different types of sensory neurons. Nociceptors, for example, have been categorized by how quickly they conduct their signals, which is determined by the diameter of the fibers and whether they are myelinated or unmyelinated. They’re also categorized according to the neurochemical markers they contain. However, these classifications have been imperfect and do not always accurately reflect their physiology (Le Pichon and Chesler, 2014).
With the recent development of in vivo calcium imaging techniques to directly visualize the firing of groups of neurons under a microscope, Chesler and his team thought they could more directly connect a sensory stimulus to responses by specific populations of neurons.
Going green
The investigators began by turning to calcium imaging with GCaMP6f. This protein, which can be genetically expressed in mice, glows green when it comes in contact with calcium. When a neuron fires an action potential, calcium rushes into the cytosol, where GCaMP6f is waiting. As GCaMP6f binds calcium, the entire cell flashes green, indicating neuronal activation.
Unlike electrophysiology, which detects one or a few neurons firing at a time, calcium imaging makes it possible to watch a symphony of hundreds to thousands of neurons firing together. Further, researchers can express GCaMP6f in specific subtypes of neurons by expressing it in tandem with proteins that exist only in the subtype of interest.
For their initial proof-of-concept studies, Chesler and his team expressed GCaMP6f in neurons expressing the heat-sensing ion channel TRPV1. Although expressed primarily in C-fibers that detect heat in adult mice, TRPV1 is broadly expressed in multiple types of neurons during development. Thus, they could place GCaMP6f into a large population of diverse sensory neurons, perfect for confirming the technique was working.
The researchers first tested this procedure in trigeminal ganglion (TG) neurons, using a surgical procedure that exposed the entire TG. This enabled them to watch for hours as neurons fired in response to sensory stimulation of the face, operating a video to record what they saw. As the video began recording, the team gently stroked the cheek with a cotton swab. Remarkably, a series of neurons could be seen firing across the TG.
In addition to the neurons responding to this non-painful, low-threshold mechanical stimulation, some neurons responded to painful high-threshold mechanical stimulation induced by pulling hairs on the face, while others responded to changes in temperature.
Digging down
Next exploring a more specific population of neurons, here in adult mice, the group expressed GCaMP6f in neurons expressing CGRP, a neuropeptide found in peptidergic nociceptors. This time, they found TG neurons that responded to hair pull, heat, or both, but not ones that responded to gentle touch, modestly warm temperatures, or cold. The heat-sensitive neurons had small cell bodies and, unsurprisingly, also expressed TRPV1. The neurons sensitive to hair pull, however, had larger cell bodies and lacked TRPV1.
Further, immunohistochemical analysis of skin samples revealed two cell populations with distinct anatomies: neurons that had free nerve terminals and those that wrapped around hair follicles in a lasso-like manner. This latter morphology had been seen in previous studies but only with low-threshold mechanical touch neurons, and never with nociceptors.
The team wondered if these lasso-like neurons were the same ones that fired when they pulled on the mouse’s hair. To examine this, they turned to resiniferatoxin (RTX), a chemical that tightly binds TRPV1, overactivating and killing the neurons that express the ion channel. RTX killed the heat-sensitive neurons, leaving behind only neurons sensitive to hair pull, effectively creating a chemical-genetic approach for isolating this population. Calcium imaging confirmed that the remaining neurons were sensitive to the pulling of hair, but not to heat, after RTX treatment.
“Since RTX has a storied past [of not always killing all TRPV1-positive neurons], we also verified the ablation of TRPV1-positive neurons by immunostaining and confirmed that the mice had reduced sensitivity to heat with a hot plate assay,” explained Chesler.
Anatomical analysis revealed that the free nerve terminals had disappeared, but the lasso-like neurons remained. Given their circumferential anatomy and that they only responded to high-threshold mechanical stimulation, the investigators named them Circ-HTMRs.
What do Circ-HTMRs do?
The investigators knew the Circ-HTMRs could be activated by hair pull, but they also wanted to show that these cells were sufficient to induce a nociceptive response. So they turned to optogenetics, which was also recently used to help identify a different population of A-fiber mechano-nociceptors innervating the skin that mediates the response to pinprick pain (Arcourt et al., 2017; also see PRF related news story).
The team expressed the light-sensitive protein Channelrhodopsin-2 (ChR2), instead of GCaMP6f, in Circ-HTMRs. To specifically activate Circ-HTMRs, they flashed blue light onto the hairy portion of the hindpaw. Immediately, the mice flinched and jumped, confirming the cells as a nociceptive population.
As another measure of pain, Chesler’s team performed a two-chamber place aversion assay in which they shined blue light onto the shaved back or head of ChR2-expressing mice in one chamber but not in the other chamber. The mice avoided the chamber where they were exposed to blue light, confirming that Circ-HTMR activation was inducing pain.
Electrophysiological properties
Chesler and his colleagues found fascinating the rapid speed with which the animals responded to blue light. Much more is known about slow-conducting C-fiber nociceptors than the fast-conducting A-delta nociceptors. The rapid response to Circ-HTMR activation suggested the newly identified cells were A-delta fibers.
To measure the conduction velocity of Circ-HTMRs, the researchers moved their experiments to the dorsal root ganglia (DRG) and combined calcium imaging with electrophysiology. After electrical stimulation of the hindpaw to trigger action potentials, calcium imaging again revealed the presence of the Circ-HTMRs. Electrophysiological recordings of the speed with which an action potential traveled from the paw to the cell body revealed that Circ-HTMRs had fast conduction velocities, similar to that of A-delta nociceptors.
The group also found that Circ-HTMRs responded to other high-threshold mechanical stimulation, such as von Frey hairs or pinching of the skin. Interestingly, these stimuli resulted in sustained cell activation, suggesting that Circ-HTMRs were slowly adapting, another characteristic of nociceptors.
Receptive fields
Moving back into the TG with calcium imaging, the researchers found that pulling a single guard hair on the cheek reliably activated small but reproducible groups of neurons. This presented an interesting opportunity.
In these experiments, “Since we’re doing in vivo imaging, as opposed to electrophysiology, allowing us to record from many neurons at one time, we can map the receptive fields of multiple neurons in the same mouse,” explained Chesler.
They did so by using a von Frey hair to stimulate specific areas of the cheek while watching, in real-time, the TG neurons that fired.
“What we found is a tiling pattern where each neuron’s receptive field partially overlaps. It’s like a GPS [global positioning system] where, by having multiple points, it allows you to have very high spatial acuity with large receptive fields, as opposed to having lots of neurons with very small receptive fields,” explained Chesler.
Mouse versus human hair
The anatomy of the Circ-HTMRs may have functional relevance to chronic pain in people.
“We know that some patients develop allodynia, but it doesn’t really occur in the glabrous skin, such as your palms. Instead, it happens in the hairy skin,” explained Stucky. “Hair movement could be where Circ-HTMRs are involved in allodynia.”
Future work could explore changes in Circ-HTMR physiology following injury. However, it’s unclear how mouse guard hairs relate to human hairs.
“Human adults have vellus hair, which is very fine and thin, and terminal hair, which is the type that makes up areas like the eyebrows. We don’t really know how the afferents that innervate mouse guard follicles correlate with afferents that innervate these follicles,” said Stucky.
Nevertheless, the new research takes a big step toward a better understanding of mechanical pain. “This is an incredibly basic science study,” Chesler said, “but the goal is to help with something that is very important medically. To do that, a deep understanding of these neural mechanisms is essential.”
Nathan Fried is a postdoctoral fellow at the University of Pennsylvania, Philadelphia, US.
Image credit: Ghitani et al., 2017.
