PIEZO: How a high-school physics concept became relevant to osteoporosis, Alzheimer's and autism

From quartz crystals and electrical charges to bone formation, sensory processing and the brain's waste-disposal system.

 

From high school to the research lab

 

Key points

•        PIEZO1 and PIEZO2 are proteins that let cells sense physical force. Despite the name, they are not piezoelectric crystals: they are mechanically activated ion channels.

•        In bone, PIEZO1 helps osteocytes respond to mechanical loading. In aged mice, activating it restored the bone response, but only when combined with loading.

•        In mouse models of autism, PIEZO2 has been linked to tactile hypersensitivity and to oxytocin signalling. Evidence in autistic people is still missing.

•        In the brain, PIEZO1-related signalling has been linked both to improved fluid transport (ultrasound in Alzheimer's mice) and to synapse loss triggered by bacterial lipopolysaccharide (LPS) in mice. These results look contradictory, and they suggest a "Goldilocks" receptor: not too little, not too much.

•        A small human study of a vibrating mattress combined with nitric oxide raised blood levels of brain-derived Alzheimer's proteins. This may indicate increased waste clearance from the brain, which could be very beneficial.

 

A reason to pay attention in physics class

There is a connection between a principle taught in high-school physics and some of the more interesting developments in modern medicine.

The piezoelectric effect describes how certain materials generate an electrical charge when mechanical pressure is applied. More than a century later, scientists found that our own cells contain specialised proteins that detect mechanical forces and convert them into electrical and biochemical signals. They are called PIEZO channels, from the Greek piezein, to squeeze or press.

A clarification is needed at the start. PIEZO channels are not piezoelectric crystals, and they do not generate charge in the way quartz does. They are proteins in the cell membrane that open when the membrane is stretched or deformed, letting ions flow in. The name reflects a shared idea (force in, electrical signal out) rather than a precisely shared physical mechanism.

Research into these channels is opening up connections between exercise and bone strength, sensory processing in autism, oxytocin signalling and the movement of fluid through the brain. Sometimes the physics lesson you thought you would never use turns out to be relevant to some of the more interesting questions in medicine.

From piezoelectric crystals to biological sensors

The piezoelectric effect was discovered in 1880 by the French physicists Pierre and Jacques Curie. They showed that pressure applied to certain crystals, including quartz, could generate an electrical charge. The reverse effect also occurs: applying an electrical field can make certain materials change shape.

In 2010, a team in Ardem Patapoutian's laboratory identified two previously unknown mechanically activated ion channels, named PIEZO1 and PIEZO2. Patapoutian shared the 2021 Nobel Prize in Physiology or Medicine with David Julius for discoveries concerning receptors for temperature and touch.

PIEZO proteins sit within cell membranes. When a cell is stretched, compressed or otherwise mechanically deformed, these channels can open, allowing positively charged ions, particularly calcium, to enter the cell. This converts a physical force into an electrical and biochemical signal. The process is known as mechanotransduction.

PIEZO1 and PIEZO2 have different distributions and functions. PIEZO1 is particularly important in blood vessels, lymphatic vessels and bone. PIEZO2 is especially important in sensory neurons, where it helps us detect touch, pressure, body position and movement.

The distinction matters because the same general mechanism can have very different consequences depending on which cells express the channel, and, as we will see, on the context.

1. PIEZO1 and osteoporosis: why mechanical loading matters

Bone is not an inert structure. It is living tissue that constantly remodels itself in response to mechanical demands.

One of the most important cells involved is the osteocyte. These cells are embedded within the bone matrix and act as sensors, detecting mechanical loading and the movement of fluid through the tiny channels surrounding them.

When we walk, run, jump or lift weights, forces are transmitted through bone. Osteocytes detect these forces and communicate with other bone cells, helping regulate the balance between bone formation and bone resorption.

PIEZO1 appears to be an important part of this sensing mechanism. When mechanical forces activate PIEZO1 in osteocytes, calcium enters the cells and triggers downstream signalling, including pathways involving Wnt signalling, and other regulators of bone remodelling.

One consequence is the regulation of sclerostin, a protein that normally inhibits bone formation. Mechanical loading can reduce sclerostin signalling, allowing bone-building activity to increase. This helps explain why appropriate mechanical loading is so important for maintaining bone density.

Does the response to PIEZO1 weaken with age?

A study published in August 2026 examined why older bone becomes less responsive to mechanical stimulation.

The researchers used 19-month-old mice and found that their bones had a blunted anabolic response to mechanical loading. Giving Yoda1, a chemical compound that activates PIEZO1, was also insufficient on its own. However, combining Yoda1 with mechanical loading restored the anabolic response in aged cortical bone.

The research findings suggest that age-related changes in how osteocytes respond to mechanical forces may be partly overcome by enhancing PIEZO1 signalling while providing an appropriate mechanical stimulus.

The distinction matters: the study does not show that PIEZO1 simply disappears with age. It shows that the bone's response to mechanical stimulation becomes less effective, and that PIEZO1 activation can help restore that response.

Yoda1 is a research compound, not an approved osteoporosis treatment, and the findings are from mice, so we cannot yet assume the same approach will work in people. The result does offer a molecular explanation for why mechanical loading matters, and it raises the possibility that future osteoporosis treatments might combine mechanical stimulation with drugs that help bone cells sense that stimulation.

2. PIEZO2 and autism: could sensory hypersensitivity involve mechanical sensing?

Many autistic people have unusual sensory responses. Touch, pressure, clothing, hair brushing, light contact or unexpected physical contact can be uncomfortable or even painful.

The mechanisms are complex. They may involve sensory receptors, peripheral nerves, spinal processing and the way the brain interprets incoming signals. PIEZO2 is a natural candidate for investigation because it is one of the principal proteins responsible for detecting mechanical stimuli.

Human genetic studies have already shown how important PIEZO2 is. People with rare loss-of-function mutations can have profound difficulties with proprioception, the sense of where their body is in space, as well as abnormalities in discriminative touch. This establishes that PIEZO2 is essential for normal human mechanosensation. It does not establish that PIEZO2 abnormalities are a general cause of autism.

PIEZO2 and tactile hypersensitivity in an autism mouse model

A 2025 investigated PIEZO2 in BTBR mice, a commonly used animal model of autism.

The researchers found increased IGF-1 signalling in dorsal root ganglia, clusters of sensory nerve cells outside the spinal cord. IGF-1 appeared to regulate PIEZO2 expression, and they identified an IGF-1-PIEZO2 pathway associated with tactile hypersensitivity. Interfering with this pathway reduced tactile hypersensitivity and also improved some social behaviours in the mice. A gentle-touch intervention reduced the abnormal expression of IGF-1 and PIEZO2 in sensory ganglia and improved the measured behavioural outcomes.

There are three implications. First, sensory hypersensitivity may not be only a problem of how the brain processes sensory information; some of the difference may originate further down the sensory pathway, in the peripheral nervous system. Second, excessive activation or expression of a mechanosensory pathway could contribute to an exaggerated response to ordinary physical stimuli. Third, it raises the question of whether different forms of tactile stimulation could influence sensory signalling.

PIEZO2, oxytocin and social behaviour

A study in July 2025 asked whether mechanical stimulation could influence oxytocin-producing neurons in the hypothalamus, a brain region involved in social behaviour, stress responses and hormonal regulation.

The researchers injected superparamagnetic iron oxide nanoparticles directly into the paraventricular nucleus of the hypothalamus in mice. A rotating magnetic field then produced mechanical stimulation, and the results depended on frequency. At 10 Hz, the system increased PIEZO2 expression in oxytocin-producing neurons, leading to calcium influx and increased oxytocin signalling. In mice exposed to valproic acid during development, an experimental autism model, the intervention improved several autism-like behavioural measures. When the researchers reduced PIEZO2 expression, the beneficial effects were lost, which suggests PIEZO2 was part of the mechanism linking mechanical stimulation to oxytocin release.

Is PIEZO1 itself an autism-related gene?

There are also some emerging genetic and epigenetic observations involving PIEZO1.

A 2026 study analysed de novo genetic variants in 41,367 parent-child trios from three family-based cohorts. PIEZO1 appeared among a group of genes showing statistically significant male-specific enrichment in one of the gene-based analyses. This is potentially relevant given autism's well-known sex bias. But it is a preprint that has not yet been peer reviewed, the study included both affected and unaffected individuals, and the findings do not establish PIEZO1 as a recognised high-confidence autism gene. It is a candidate for further investigation.

A separate 2025 study examined DNA methylation of PIEZO1 in infants with neurodevelopmental disorders. The researchers studied 24 infants with neurodevelopmental disorders and 22 typically developing infants, and found lower methylation at a group of PIEZO1 sites in the clinical group. Methylation can influence gene regulation, so this raises the possibility that PIEZO1 expression is altered during early neurodevelopment. But the clinical group was a mixture of conditions rather than autism alone, methylation was measured in buccal cells rather than brain tissue, and the researchers did not measure PIEZO1 gene or protein expression. It is an early observation, not evidence of a specific autism mechanism.

What does this mean for autism?

We now have several separate lines of research:

•        PIEZO2 may contribute to tactile hypersensitivity through sensory neurons outside the brain (mice).

•        PIEZO2 may influence oxytocin release through hypothalamic neurons (mice).

•        PIEZO1 has appeared in a large genetic analysis as a candidate requiring further investigation.

•        PIEZO1 methylation differences have been reported in a small group of infants with neurodevelopmental disorders.

These findings suggest that mechanosensitive ion channels may be relevant to several biological processes already of interest in autism research, and different individuals or subgroups could have quite different alterations.

Where this leaves us. The most informative next steps would be human studies that measure tactile detection and discomfort thresholds in autistic participants alongside PIEZO2 genotyping or expression data; peer-reviewed replication of the PIEZO1 de novo variant signal; and replication of the methylation finding in larger, autism-specific groups. Until then, PIEZO2 is a plausible mechanism for some sensory differences, not an established one.

3. PIEZO1, brain pulsations and Alzheimer's disease

The brain has its own system for moving fluid and clearing waste products, commonly referred to as the glymphatic system. It involves the movement of cerebrospinal fluid (CSF) along pathways surrounding blood vessels, with exchange between CSF and fluid in the brain tissue. Waste products, including soluble amyloid-beta, can then be carried towards drainage pathways. The system appears to be particularly active during sleep.

The glymphatic concept is still debated. Some researchers question how much fluid is moved by bulk flow rather than diffusion, how reliably clearance can be measured in humans, and whether sleep and anaesthesia change it in the way early mouse studies suggested. 

Why does PIEZO1 matter here? Blood vessels constantly expand and contract in response to changes in blood pressure and flow, and these movements create mechanical forces that can influence the surrounding fluid. PIEZO1 is expressed in vascular endothelial cells, which line blood vessels, and lets them detect mechanical forces and change their signalling. This provides a possible link between vascular pulsations, mechanical stimulation and the movement of fluid through the brain. If these processes become less effective with age, the clearance of soluble proteins could be affected, which would be relevant to Alzheimer's disease, where accumulation of amyloid-beta and other abnormal proteins is a major feature.

Ultrasound activates a PIEZO1-related pathway in Alzheimer's mice

A 2026 study investigated whether low-intensity focused ultrasound could improve glymphatic-associated transport in 5xFAD mice, a widely used Alzheimer's disease model. The researchers applied ultrasound to the hippocampal region over four weeks.

They reported increased CSF tracer movement into the brain and improved transport of solutes through brain tissue. They also found increased soluble amyloid-beta in deep cervical lymph nodes, reduced amyloid deposits in the hippocampus and prefrontal cortex, and improvements in learning and memory tests. The effects were accompanied by increased PIEZO1 signalling in brain vascular tissue, and when the researchers inhibited PIEZO1-related signalling, many of the benefits were reduced.

This is experimental evidence that PIEZO1-related mechanotransduction may contribute to the effects of ultrasound on brain fluid transport. It remains an animal study and does not show that ultrasound can prevent or treat Alzheimer's disease in humans.

PIEZO1, gut bacteria and microglia: a double-edged sword?

So far PIEZO1 has appeared as a helpful sensor: it helps bone respond to loading, and its activation was linked to better fluid transport in the ultrasound study. A 2025 paper suggests the picture is less simple.

The researchers were interested in lipopolysaccharide (LPS), a component of the outer wall of Gram-negative bacteria and a powerful inflammatory trigger. LPS has been found at higher levels in Alzheimer's brains, but it is not clear how it gets past the blood-brain barrier. One suggestion is that it travels inside bacterial extracellular vesicles (bEVs), tiny membrane packages released by gut bacteria.

The team isolated bEVs from the faeces and blood of people and mice. LPS levels in these vesicles, or in plasma vesicles overall, were higher in people with Alzheimer's disease or mild cognitive impairment than in healthy controls. In mice, vesicles from Gram-negative bacteria reached the brain, while those from Gram-positive bacteria did not. Neutralising the LPS, or using bacteria engineered to lack it, prevented the vesicles from getting in.

Once in the brain, the vesicles activated microglia, raised levels of the complement proteins C1q and C3, and led to loss of synapses within 24 hours, with microglia engulfing more synaptic material. PIEZO1 was involved. The vesicles increased Piezo1 and calcium levels in microglial cells, and blocking Piezo1 with the inhibitor GsMTx4 reduced the effect. Mice lacking Piezo1 in myeloid cells (the cell family that includes microglia) were largely protected from the synapse loss. Samples of Alzheimer's cortex also showed more Piezo1 in microglia.

This looks very different from earlier work. A 2023 study in cultured microglia found that activating PIEZO1 with Yoda1 reduced LPS-induced inflammation, and other work suggests PIEZO1 may help microglia engulf amyloid plaques. The authors of the new paper themselves describe PIEZO1 as a "double-edged sword".

Several cautions apply. Most experiments were short-term injections of vesicles into young mice, with synapse counts at 24 hours. There were no memory tests and no chronic Alzheimer's model, and several comparisons used only three animals per group. GsMTx4 is a non-specific inhibitor and only partly blocked the effects. Here PIEZO1 seems to be switched on by vesicle uptake and inflammation rather than by physical force, so this is not mechanotransduction in the usual sense. It is a mouse and cell-culture mechanism, not a treatment.

It does carry a lesson for the rest of this post: boosting PIEZO1 in one tissue says little about what it will do in another. I return to this in Section 6.

Thanks to our reader Stephen for suggesting a post on LPS and PIEZO1.

4. The vibrating mattress: a new human study

The story took another turn with a study published in November 2026. Researchers investigated a combination of nitric oxide-mediated vasodilation and a specially designed whole-body vibration intervention intended to stimulate PIEZO1. The study involved 29 healthy adults.

The vibration was delivered with a commercially available Neurosonic Mobile Mattress, using a specially modulated vibration pattern rather than simple continuous vibration. The researchers also gave a sublingual nitric oxide intervention. Nitric oxide causes blood vessels to relax and dilate, while mechanical stimulation of PIEZO1 can produce complementary changes in vascular signalling. The idea was to influence the movement of blood and fluid through the brain by combining the two mechanisms. Nitric oxide is a recurring theme on this blog, including the vascular effects of agmatine and eNOS.

Following the combined intervention, the researchers measured increases in blood levels of several brain-derived proteins:

•        Amyloid-beta 40 (Aβ40)

•        Amyloid-beta 42 (Aβ42)

•        Glial fibrillary acidic protein (GFAP)

They also observed changes in MRI measurements of brain fluid dynamics and brain pulsations. A similar rise in blood levels of brain-derived proteins was reported after intermittent hypercapnia, which I discussed in July so the mattress result fits a developing pattern in which vascular oscillations are associated with proteins appearing in the blood.

What the study does and does not show

The increases in circulating proteins are consistent with a temporary movement of soluble brain-derived proteins towards the bloodstream. That is a hypothesis, not the only possible reading. Release from blood cells or the vessel wall, the haemodynamic effects of vasodilation and a falling blood pressure, or, in the case of GFAP, an astroglial stress response could all contribute. Telling these apart would need studies that measure clearance directly, follow the proteins over a longer period and include control interventions.

The effect was also acute. A rise in blood proteins after a single session does not show net clearance from the brain over time. An increase in amyloid-beta in the blood does not prove that amyloid plaques have been removed, and it does not demonstrate better memory, reduced Alzheimer's risk or any clinical benefit. The study was small and involved healthy volunteers, not people with Alzheimer's disease.

It is nonetheless a first human indication that a PIEZO1-targeted mechanical intervention may influence brain fluid dynamics and the movement of brain-derived proteins, and a bridge between laboratory work on PIEZO1 and the possibility of non-pharmacological ways to influence brain fluid transport.

5. Could glymphatic dysfunction also be relevant to autism?

This brings us back to a question I have raised before on this blog: could impaired glymphatic drainage be relevant to autism?

A study in August 2026 adds a piece of evidence. Researchers analysed MRI data from five cohorts in the Autism Brain Imaging Data Exchange, involving 250 participants: 128 autistic individuals and 122 neurotypical controls. They used diffusion tensor imaging analysis along the perivascular space (DTI-ALPS), an indirect MRI measure that estimates water diffusion along pathways associated with blood vessels. It is sometimes used as a proxy for glymphatic function, although it does not directly measure clearance of waste products from the brain.

They found lower ALPS index values in the autistic group overall. The difference was particularly apparent in adults: there was no significant group difference among participants younger than 18, whereas autistic adults had lower values than neurotypical adults. Among autistic adults, the ALPS index was also associated with depressive symptoms, but not with IQ or the measured severity of autistic traits.

How this fits with earlier findings

In July I discussed a study of 78 autistic children and 48 typically developing controls that reported reduced DTI-ALPS values in children. The August analysis found no significant difference under 18, so the picture in children is not yet consistent. Differences in cohorts, scanners, age ranges and analysis methods could all contribute, and taken together the studies suggest a possible difference rather than an established one.

The interpretation needs care. The August study was cross-sectional, so it cannot tell us whether reduced perivascular fluid movement contributes to autism, develops later in life, or reflects other factors. Sleep, vascular function, mood and other health factors could all be relevant, and DTI-ALPS is an indirect marker rather than a direct measurement of glymphatic clearance.

If future studies confirm that some autistic people have altered glymphatic function, it would be worth asking whether this is associated with sleep disturbances, vascular differences or other biological features, and whether interventions that influence vascular pulsations or mechanical signalling have any relevance. At present, there is no evidence that vibrating mattresses or other PIEZO-targeted interventions improve glymphatic clearance in autistic people.

6. One principle, many applications: a Goldilocks receptor

What makes PIEZO research so interesting is that one biological mechanism lets physical forces influence very different systems. In bone, mechanical loading helps osteocytes regulate bone formation. In the peripheral nervous system, PIEZO2 helps detect touch, pressure and body position. In the hypothalamus, PIEZO2 can influence oxytocin-producing neurons. In blood vessels and lymphatic vessels, PIEZO1 helps cells respond to mechanical forces. And in the brain, these processes may influence the movement of fluid and the clearance of soluble proteins.

It is tempting to think of PIEZO channels as a universal switch that could be turned up to improve health. The evidence points the other way. PIEZO looks more like a Goldilocks receptor: not too little, not too much.

•        Bone: Yoda1 alone did not restore the response in aged mouse bone. It worked when combined with mechanical loading, so the stimulus and the channel both mattered.

•        Microglia: In cultured microglia, activating PIEZO1 with Yoda1 reduced LPS-induced inflammation. In mice, LPS-carrying bacterial vesicles activated microglial Piezo1 and drove synapse loss. Same channel, opposite-looking outcomes.

•        Stimulus: In the oxytocin study the effect depended on frequency, and only 10 Hz produced the described benefit.

The effects depend on which PIEZO channel is involved, which cells express it, the intensity, timing and frequency of stimulation, and the biological context. Too little activity may be a problem in one tissue, while excessive or prolonged activation could be harmful in another. I would treat "Goldilocks" as a working hypothesis rather than a finding: the studies show opposite effects in different settings, but none has mapped a dose-response curve for PIEZO activity.

The research also shows why mechanical stimulation is attracting renewed interest. Exercise, vibration, ultrasound and other physical inputs may have biological effects beyond moving muscles or loading bones, because they can influence how cells sense and respond to their physical environment. Regular exercise remains the simplest and best-supported way to apply mechanical stimulus to bone, and I have discussed how exercise may relate to glymphatic function [link: 22 July 2026 post].

Conclusion

The discovery of PIEZO channels created an unexpected bridge between physics and biology. A principle first demonstrated in crystals more than 140 years ago has a biological counterpart in proteins that let cells convert mechanical forces into electrical and chemical signals.

For osteoporosis, PIEZO1 helps explain how bones detect mechanical loading and why that response can become less effective with age. For autism, PIEZO2 research offers possible mechanisms for tactile hypersensitivity and oxytocin signalling, PIEZO1 deserves further investigation, and the MRI findings raise the question of whether altered brain fluid transport matters for some people. For Alzheimer's disease, animal studies and a small human study suggest that mechanical stimulation of vascular PIEZO1 pathways may influence brain fluid dynamics, while the LPS work is a reminder that the same channel can contribute to harm in another context.

Perhaps there really was a good reason to pay attention in physics class.

 

Selected research papers

PIEZO channels and mechanotransduction

•        Coste B. et al. (2010). Piezo1 and Piezo2 are essential components of distinct mechanically activated cation channels. Science. DOI: 10.1126/science.1193270

•        Chesler A.T. et al. (2016). The Role of PIEZO2 in Human Mechanosensation. New England Journal of Medicine. DOI: 10.1056/NEJMoa1602812

Bone and osteoporosis

•        Zhao D. et al. (2026). Piezo1 Activation Rescues Anabolic Response to Mechanical Loading in Aged Bone via Connexin 43 Hemichannels. Aging Cell. DOI: 10.1111/acel.70687

Autism and sensory processing

•        Zhai J. et al. (2025). The effect of modulation Piezo2 by IGF-1 on tactile hypersensitivity in BTBR model mice. Life Sciences. DOI: 10.1016/j.lfs.2025.123449

•        Liu S. et al. (2025). PIEZO2 is the underlying mediator for precise magnetic stimulation of PVN to improve autism-like behavior in mice. Journal of Nanobiotechnology. DOI: 10.1186/s12951-025-03557-x

•        Turner T.N. (2026). De Novo Variation in Autism by Sex and Diagnostic Status in 41,367 Parent-Child Trios. medRxiv preprint. DOI: 10.64898/2026.01.26.26344889

•        Mascheroni E. et al. (2025). Exploring PIEZO1 DNA methylation in infants with neurodevelopmental disorders. Frontiers in Psychology. DOI: 10.3389/fpsyg.2025.1593609

Alzheimer's disease, microglia and glymphatic transport

•        Zhang L. et al. (2026). Ultrasound enhances glymphatic-associated solute transport via Piezo1-related mechanotransduction in 5xFAD mice. Journal of Cerebral Blood Flow & Metabolism. DOI: 10.1177/0271678X261470064

•        Tuunanen J. et al. (2026). Vasomechanic treatment is associated with increased efflux of human brain proteins. Nature Communications. DOI: 10.1038/s41467-026-78894-y

•        Zhao X., Yu J. et al. (2025). Gut-derived bacterial vesicles carrying lipopolysaccharide promote microglia-mediated synaptic pruning. Alzheimer's & Dementia. DOI: 10.1002/alz.70331

•        Piezo1 channel-mediated Ca2+ signalling inhibits LPS-induced activation of the NF-kB inflammatory signalling pathway and generation of TNF-alpha and IL-6 in microglial cells. Glia (2023), 71(4):848-865. [author list to be added]

•        The mechanosensitive Piezo1 channel mechanism of Alzheimer's disease and implications for the development of therapeutic or early detection strategies. Frontiers in Aging Neuroscience (2025). DOI: 10.3389/fnagi.2025.1707659 [author list to be added]

Glymphatic function in autism

Knudsen L.V. et al. (2026). Reduced glymphatic function in autism revealed by the diffusion tensor analysis along the perivascular space index.  Journal of Neural Transmission. DOI: 10.1007/s00702-026-03249-2