Mechanosensation—the ability to sense force—is a fundamental sensory process in people and other animals. However, while researchers have made progress in identifying molecular mechanisms of mechanosensation in animal studies, an understanding of these processes in people has remained largely elusive. Now, new research begins to illuminate the transducers of force in humans.
Using whole-exon sequencing, two research groups at the National Institutes of Health (NIH), Bethesda, US, one led by Carsten Bonnemann of the National Institute of Neurological Disorders and Stroke and the other led by Alexander Chesler of the National Center for Complementary and Integrative Health, identify mutations in PIEZO2, an ion channel previously identified in Drosophila and mouse studies of mechanosensation, in two patients with previously undiagnosed neuromuscular disorders. Using extensive sensory testing, the researchers further find that the mutations result in defects in response to vibration, touch discrimination, and joint proprioception, but had no effect on measures of pain, leaving the identity of the mechanical pain transducer unknown.
“It is relatively rare to see such nice agreement between animal models and human studies,” wrote Ardem Patapoutian, Scripps Research Institute, La Jolla, US, in an email to PRF. “Further analysis of these patients will in turn extend our knowledge of what Piezo2 ion channels do,” added Patapoutian, who discovered along with colleagues Piezo1 and Piezo2 in cell studies and later elucidated their mechanosensory role in fruit flies and mice, but was not involved in the current work.
The study was published online September 21 in the New England Journal of Medicine.
When humans are like mice
The discovery and characterization of two stretch-gated ion channels, Piezo1 and Piezo2, were important first steps in identifying transducers for touch and possibly pain. Studies in Drosophila showed a mechanosensory role for the channels, with knockout of Piezo2 in the flies drastically reducing the ability to sense noxious force (see related PRF news story). Later studies in mice revealed that conditional knockout of Piezo2 had no effect on the ability to sense inflammatory, neuropathic, or mechanical pain, but inhibited the ability to sense light touch (see related PRF news story). Subsequent research showed that Piezo2 knockout mice also had uncoordinated body movements and abnormal limb positioning, suggesting Piezo2 also plays a role in proprioception (Woo et al., 2015). However, whether PIEZO proteins serve similar mechanosensory functions in humans was largely unknown.
In the new study, the researchers examined two unrelated patients with undiagnosed neuromuscular disorders who presented with highly unusual yet similar phenotypes. Both patients had impairments of ambulation, proprioception, vibratory sense, and fine motor skills, as well as difficulty performing reaching tasks. Electrodiagnostic testing revealed reduced sensory nerve responses. There were no deficits in tests of cognition, nor were there any central nervous system abnormalities according to functional magnetic resonance imaging (fMRI).
Bonnemann and colleagues performed whole-exome sequencing on the two patients to search for a genetic basis underlying these symptoms. Surprisingly, both patients carried compound-inactivating variants of the same gene, PIEZO2. The first patient was compound heterozygous for two alleles with null variants, while the second patient carried a null variant on one allele and a missense variant on the other—ultimately resulting in a loss of function of PIEZO2 in both patients.
According to Chesler, the discovery of the PIEZO2 mutations by Bonnemann was quite surprising. Indeed, at the time of the exon sequencing, Bonnemann’s group did not yet know that the two patients might have altered perception of touch. However, when Chesler described to Bonnemann the previous studies of Piezo2 knockout mice in their first meeting, “his jaw dropped,” Chesler said. Together, Chesler and Bonnemann were able to use the exon sequencing to find the gene and then use previous information about Piezo2 to guide mechanosensory testing in the two patients.
Is it working?
To examine the functioning of the PIEZO2 variants, the scientists transfected human embryonic kidney (HEK293) cells, which do not endogenously express PIEZO2 or respond to mechanical stimuli, with mutated variants of mouse Piezo2 similar to those found in the two patients, or with wild-type Piezo2. Electrophysiological experiments showed that inward currents in response to mechanical stimulation were only observed in cells expressing wild-type Piezo2 and not in those with mutated variants. This suggested that the human variants did not create functional PIEZO2 protein.
Some genetic differences between the two patients were apparent. Reverse transcriptase polymerase chain reaction (RT-PCR) of RNA from skin biopsy samples showed that PIEZO2 messenger RNA (mRNA) was nearly undetectable in one of the patients, suggesting a deficit of gene transcription. The other patient had detectable levels of PIEZO2 mRNA, but further DNA sequencing showed the PIEZO2 transcripts were derived partly from normally spliced PIEZO2 mRNA but primarily from a missense PIEZO2 variant, likely resulting in non-functional PIEZO2 protein.
Whether the mutations lead to a complete loss of PIEZO2 function remains unclear. “The authors assume these mutations are killing the function of the protein. There is some evidence for that, but it’s hard to prove,” said Gary Lewin, Max-Delbrück Center for Molecular Medicine, Germany, who was not involved in the study. “There is still a little bit of a question mark regarding [the degree of] loss of function of these alleles, and I think it’s often the case that more severe mutations [would] lead to lethality.” Mice with complete depletion of Piezo2 die soon after birth, suggesting a role for the protein in normal development. However, the authors suggest that, unlike in mice, it is possible humans can survive without PIEZO2.
What does PIEZO2 do in people?
Given previous research showing that Piezo2 primarily affects the sensation of touch, the investigators used a variety of mechanosensory tests in the two patients to understand the role of the protein. Glabrous (hairless) skin of the palm and fingertips of both patients had markedly decreased sensitivity to punctate touch (by von Frey filaments), gentle stroking with a brush, and vibration. In a two-point touch discrimination task, whereas control patients were able to discriminate between a one-point stimulus and a two-point stimulus with 100 percent accuracy, the PIEZO2-deficient patients only showed approximately 40 percent accuracy, scoring no better than chance.
On hairy skin, sensitivity to vibration was reduced in both patients, but unlike glabrous skin, punctate and gentle touch were not affected. These results suggest that hairy skin is innervated by mechanoreceptors that do not require PIEZO2, similar to what is observed in mice. fMRI scans supported these findings, showing brain activation in response to gentle brushing on hairy but not glabrous skin.
Proprioception was also tested by asking the patients to make large and small movement of the arms and legs. With smaller movements of the joints, control participants were able to detect the direction of movement with 100 percent accuracy, but the patients could only detect the direction of movement with 40 to 60 percent accuracy―no better than chance.
Finally, during a learned motor task (reaching a finger from the nose to a target kept at arm’s length), PIEZO2-deficient patients performed similarly to controls. However, when the task was performed blindfolded, both patients had intense dysmetria—an inability to control the distance and speed of movement—when nearing the target, while the control subjects did not. In addition, the velocity of the arm movements of the two patients while blindfolded was approximately 10 times more variable compared to when their eyes were open, or compared with blindfolded controls.
What about pain?
The patients’ responses to mechanical pain (both pinprick and pressure) and thermal pain (both heat and cold) stimuli did not differ from responses in control subjects, supporting data from animal knockout studies that Piezo2 does not transduce mechanical pain. This leaves uncertain how noxious mechanosensation operates in people.
“This is really asking the million-dollar question,” said Chesler. “We have previously observed no change in mechanical pain [in mice lacking Piezo2], but moving forward we need to ask, Is all mechanical pain created equal? There are a lot more questions we haven’t even delved into. If there is a single mechanosensor for pain, there would be incredible significance in that discovery.”
As for the current work, it provides an intriguing example of how basic science can inform clinical research. PIEZO2 deficiency in humans “essentially reproduces all of the features you might expect in the mouse, which is important to know for drug development,” Lewin said. In turn, human studies give researchers the opportunity to address questions that are difficult to answer in animals, and also raise new questions that can then be taken back to animals for further study. The new research “gives us a rich appreciation for the information that the mechanosensor is relaying to the brain and allows us to get at questions of perception, which is difficult to do in a mouse. It is a reiterative cycle from the bench to the bedside and then back to the bench,” Chesler said.
Hillary Doyle is a PhD candidate and science writer studying pain and analgesia at Georgia State University in Atlanta.
Image credit: akinshin/123RF Stock Photo.