Piezos Thrive Under Pressure

REVIEWS
Piezos thrive under pressure:

mechanically activated ion channels
in health and disease
Swetha E. Murthy, Adrienne E. Dubin and Ardem Patapoutian
Abstract | Cellular mechanotransduction, the process of translating mechanical forces into
biological signals, is crucial for a wide range of physiological processes. A role for ion channels
in sensing mechanical forces has been proposed for decades, but their identity in mammals
remained largely elusive until the discovery of Piezos. Recent research on Piezos has underscored
their importance in somatosensation (touch perception, proprioception and pulmonary
respiration), red blood cell volume regulation, vascular physiology and various human
genetic disorders.
Membrane potential Cellular mechanotransduction is essential for detect- on the cell type (for example, for hair cells, the effect
The voltage across the plasma ing external mechanical stimuli relevant to touch and is neurotransmitter release onto auditory afferent end-
membrane that is dependent hearing, as well as for sensing forces from within the ings). The mechanotransduction channels also vary in
on the charge separated on the
body (for example, lung expansion and blood flow). gating kinetics (the rate at which ion channels transi-
outer and inner membrane
leaflets; usually, the resting
In vertebrates, dedicated cells such as inner ear hair tion between closed and open states)5. For example,
membrane potential is cells and somatosensory neurons are specialized to mechanical stimulation of cultured sensory neurons
negative (~–70 to –30 mV detect mechanical force with high sensitivity, which evoked non-selective cation currents with latencies in
depending on the cell type). is an integral part of their receptor function in the the sub-millisecond range and with inactivation (current
nervous system (BOX 1). However, it is now clear that relaxation in the presence of a stimulus) kinetics ranging
mechanotransduction is generally important for most from ~5 milliseconds to ~1 second5. Although vertebrate
cell types and regulates processes as diverse as cell ion channels selective for potassium cations have been
migration, proliferation and differentiation. At the shown to be activated by mechanical stimuli, the identity
core of these events is the conversion of force into bio- of any vertebrate mechanically activated non-selective
logical signals, which is accomplished mainly by the cation channel was elusive until the discovery of the
activation of ion channels, receptors and membrane Piezo1 and Piezo2 channels in 2010 (REF. 6).
proteins associated with extracellular, membrane and/or Here, we review the importance of the evolutionarily
intracellular components. conserved Piezo ion channel family. Global or tissue-
The first demonstration that mechanical forces specific knockout of genes encoding Piezo proteins
could directly activate ion channels (hence the general in mice or their knockdown in specific cell types has
term mechanically activated channels) was the seminal demonstrated that these channels have crucial roles in
work by Hudspeth and Corey in 1979. They argued several mechanosensory processes, such as vascular
that the activation of ionic currents in bullfrog auditory development and function, pulmonary respiration and
epithelial cells by mechanical stimulation was too fast sensory transduction. We discuss what is known about
(sub-millisecond range) to be compatible with a mech- the structure of Piezo ion channels and their mechano
anism involving second messengers1. Subsequently, the sensing and gating mechanisms. We then focus on
first eukaryotic stretch-activated cation current was anticipated and unexpected processes in which Piezo1
Department of Neuroscience, described in chick skeletal muscle2. The characteriza- and Piezo2 have important roles in vivo. For informa-
Howard Hughes Medical
tion of ionic currents recorded from various mechano tion on other mechanotransduction channels and more
Institute, The Scripps
Research Institute, La Jolla, sensory cells revealed that the channels mediating these general cellular mechanobiology, we recommend these
California 92037, USA. currents were fairly indiscriminate (non-s elective) reviews: (REFS 7–9) (see also BOX 2 for a brief overview of
Correspondence to A.P. with regard to the type of cation they passed3,4. The eukaryotic mechanosensitive proteins). We also direct
ardem@scripps.edu influx of cations and subsequent depolarizat ion of the reader to related reviews on Piezo ion channels (gen-
doi:10.1038/nrm.2017.92 the membrane potential invariably leads to an ‘excita- eral overviews10–12, molecular structure and function
Published online 4 Oct 2017 tory’ effect. However, the nature of this effect depends mechanisms13,14 and roles in sensory biology 15).
NATURE REVIEWS | MOLECULAR CELL BIOLOGY ADVANCE ONLINE PUBLICATION | 1

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS
Box 1 | Hair cell mechanotransduction and close in the continued presence of the stimulus
(a phenomenon called channel inactivation) in the milli-
The mammalian auditory system is composed of the cochlea (which is responsible for second range6,22. The underlying mechanism or molecu-
hearing) and a part of the vestibular organ (which controls balance and head movement). lar determinants of Piezo inactivation remain unknown.
Within these structures, specialized mechanosensory cells called hair cells detect Nevertheless, several (patho)physiological factors are
vibrations and transduce them into electrical signals that are relayed to sensory neurons.
known to affect the inactivation kinetics of Piezos
On the apical side of hair cells, cilia-like structures called stereocilia are arranged in order
of descending height in three rows that form the hair bundle. Stereocilia are connected with functional consequences, and these scenarios are
to one another via tip links formed by parallel homodimers of the cell–cell adhesion discussed in the following sections.
molecules cadherin 23 and protocadherin 15 (REF. 92). The mechanotransduction The use of selective blockers or activators of ion chan-
complex is present at the tip of stereocilia. Sound waves induce deflection of the nels helps to parse their role in physiological processes. To
stereocilia in each hair cell towards the tallest row, and this deflection has been proposed date, the compound Yoda1 (2-[5-[[(2,6‑dichlorophenyl)
to activate the mechanically activated ion channel. Deflection of stereocilia in the methyl]thio]-1,3,4‑thiadiazol‑2‑yl]pyrazine) is the
opposite direction (towards the shortest stereocilia) accordingly decreases channel only known activator of Piezo1; interestingly, Yoda1
opening. The unique arrangement and polarity of stereocilia is crucial for hair cell has no obvious effect on Piezo2 (REF. 23). Notably,
function and gating of their mechanosensitive channels. It is proposed that the channel chimaeras between PIEZO1 and PIEZO2 suggest that
involved in hearing is tethered to extracellular filaments (that is, tip links) and that
the region between PIEZO1 amino acid residues 1961
force-mediated movement of the filament tugs the channel open101. Although many
transmembrane proteins localize to the stereocilia, the exact nature of the pore-forming and 2063 (in the carboxy‑terminal region) is required for
subunit of the mechanotransduction channel in hair cells has so far been elusive. Yoda1‑dependent activation of the channel24. Whether
this region is indeed the Yoda1-binding site has not been
determined. Although its mechanism of action remains
Properties and structure of Piezos unknown, Yoda1 has helped to confirm Piezo1‑specific
A lack of homology between Piezos and other known downstream signalling events mediated by calcium in
ion channels and membrane proteins has impeded the endothelial cells and red blood cells (see the section on
use of homology-based predictions on Piezo topology or Physiological roles of Piezo1)25,26. Both Piezo1 and Piezo2
structure. Chimaera- and mutagenesis-based structure– are blocked by extracellularly applied GsMTx4 (spider
function studies, together with a medium-resolution venom toxin), which was shown decades ago to block
structure of the mouse Piezo1 channel determined by stretch-activated cation channels by altering the mem-
cryo-electron microscopy (cryo‑EM), have helped to brane curvature surrounding the channel27,28. Whether
shed light on certain details of the architecture of Piezos. the same mechanism applies to blocking Piezos is
not known.
General properties of Piezo channels. Piezos are the
largest known pore-forming multimeric (individual sub- Piezo channel architecture. Much of the structural
units arranged as trimers; see also below) ion channels, information about Piezos comes from a recent cryo‑EM
with each subunit composed of ~2,500 amino acids. map of the mouse Piezo1 channel. This map shows
For each Piezo1 subunit, 30–40 hydrophobic segments unequivocally that Piezo1 ion channels oligomerize as
are predicted by hydrophobicity analysis software16–19. homotrimers18. The best-resolved features are a central
FIGURE 1a illustrates one example depicting 38 predicted ‘core’ domain containing C-terminal segments of each
hydrophobic regions in PIEZO1. Whether all 38 seg- subunit and ‘blade’ structures formed by six pairs of
ments are transmembrane domains is currently not transmembrane segments (‘peripheral helices’) project-
settled (see below). Mechanical activation of the channel ing outward to the amino-terminal end of each subunit,
results in influx of sodium and calcium, which can lead suggesting a propeller-like architecture for the Piezo1
to the propagation of electrical signals and initiate intra- channel (FIG. 1b).
cellular secondary messenger pathways. Vertebrates have The core of the trimer comprises a large soluble extra-
two Piezo genes, PIEZO1 and PIEZO2, and the resulting cellular region (‘cap’ domain) linking the last two trans-
proteins share ~50% identity at the amino acid level. It is membrane helices from all three subunits. The crystal
likely that they evolved to take on specialized roles in the structure for the cap domain has been solved for mouse
particular tissues in which they are expressed: PIEZO1 is and Caenorhabditis elegans homologues16,18. Deleting the
expressed mainly in non-neuronal cells, whereas PIEZO2 entire cap domain of mouse PIEZO1 protein or mutating
is expressed in sensory neurons and in some specialized acidic residues in this domain (for example, mutating
mechanosensory structures (see also the more detailed Asp-Glu-Glu-Glu-Asp (residues 2393–2397) to Ala)
discussion of the physiological roles of Piezos below). alters the ability of the channel to selectively pass cat-
ions over anions. This suggests that the cap domain is
Biophysical properties of Piezo channels. Piezos are involved in sequestering cations near the pore18.
non-selective cation channels with very low perme The cryo‑EM structure predicts that the last trans-
ability to chloride17,20,21. The single-channel conduct- membrane domain of each subunit forms the pore-
ance for murine Piezo1 and Piezo2 is 29 pS and 24 pS lining helix of the channel18. Indeed, mutagenesis results
(picosiemens (pS) is a measure of conductance, which of substituted-cysteine accessibility method analysis
is the reciprocal of resistance), respectively, when meas- argue that residues in the last hydrophobic region face
ured in physiological saline17. Mechanically induced the ion-conducting pathway 21. Furthermore, molecu
macroscopic Piezo currents (ensemble currents from lar perturbation of specific negatively charged residues
many channels) activate in the microsecond timescale upstream and downstream of the last two putative
2 | ADVANCE ONLINE PUBLICATION www.nature.com/nrm

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS
transmembrane domains in Piezo proteins (Glu2133, transmembrane helices and the orientation and second
Glu2495 and Glu2496 in mouse PIEZO1; Glu2416, ary structure of the N‑terminal region of PIEZO1
Glu2769 and Glu2770 in mouse PIEZO2) alters their require further investigations, and a higher-resolution
single-channel conductance and cation-over-anion structure is now needed to resolve these issues.
selectivity and renders them insensitive to the broad- Interestingly, the cryo‑EM structure suggests that
spectrum cationic pore blocker ruthenium red17,21. These the peripheral helices have a considerable curvature
data indicate that acidic residues in the vicinity of the relative to the plane of the membrane, protruding into
putative pore-lining helix regulate ion permeation. the extracellular space as they radiate out from the
In the proposed model of Piezo1 channel archi- centre (FIG. 1b). Structural data from the mechanically
tecture emerging from the cryo‑EM studies, the large activated ion channels MscS channel from bacteria and
N‑terminal regions adjacent to the peripheral helices vertebrate potassium channel TRAAK (also known as
appear above the plane of the membrane, suggesting potassium channel subfamily K member 4) suggest that
an extracellular location18 (FIG. 1b). However, evidence the presence of these proteins in the lipid bilayer causes
from topology experiments (for example, the location local curvature in the membrane29,30. It is possible that a
of Myc-tagged residues) suggested that the N‑terminal similar model applies to Piezos, in which the peripheral
tails reside intracellularly 17. Furthermore, mass spectro helices follow the curve in the membrane caused by the
metry results identified phosphorylated residues within channel (FIG. 1c, left panel). Alternatively, the observed
the first 400 amino acids of the protein, indicating that bend in the protein could be an artefact of detergent-
these residues are probably intracellular 17. Finally, results based purification procedures, and, in cells, the channel
from topology experiments also confirm that at least 18 may in fact extend horizontally and align with a planar
out of the 38 predicted hydrophobic segments are transmembrane (FIG. 1c, right panel). Future structural studies
membrane domains, whereas the cryo‑EM structure on Piezos isolated in a more native environment, such
resolves only 14 transmembrane domains per subunit. as in nanodiscs or in lipid vesicles, could address this
However, because only 30% of the protein was modelled issue. Additional structural and functional analyses are
into the 3D density map, many more transmembrane required to elucidate the remaining structural motifs of
domains are likely to exist 17,18. Thus, the total number of these ion channels and shed light on Piezo geometry.
Box 2 | Eukaryotic proteins involved in mechanotransduction

Mechanically activated ion channels Tentonin3 (TMEM150C): global genetic ablation reduced mechanically
Ion channels that are directly activated by physical force are classified as activated currents in a subpopulation of mouse dorsal root ganglion
mechanically activated channels. In addition to Piezos, bona fide neurons106. Although initial studies suggested that expression of
eukaryotic mechanically activated ion channels include several others. TMEM150C in heterologous systems conferred mechanically activated
K2P channels (two-pore-domain potassium channels; KCNK family): currents, these currents were later shown to be dependent on
among the 15 members within this family, TREK1 (also known as endogenously expressed PIEZO1 (REF. 38).
potassium channel subfamily K member 2) was the first identified
Mechanosensitive ion channels
mechanically activated mammalian ion channel; subsequently,
When gating domains interact with lipids in the bilayer (channel–bilayer
TREK2 (also known as potassium channel subfamily K member 10) and
interface), those domains will be subject to the tension in the bilayer.
TRAAK (also known as potassium channel subfamily K member 4) were
Thus, it is not surprising that the activity of many ion channels with
discovered102. Asymmetric tension induced by membrane curvature
other known functions (for example, voltage- or ligand-gated channels)
is sufficient to activate these channels8. In the sensory system, K2P
can be directly or indirectly influenced by mechanical force. These
activation suppresses neuronal excitability by hyperpolarizing the
ion channels are primarily activated by a non-mechanical stimulus
membrane potential.
and are mainly modulated by mechanical force. Some examples of such
Transient receptor potential (TRP) channels: one of the members of this
mechanosensitive ion channels include the voltage-gated channels Kv1.1
group, TRPN1 (known as NompC in Drosophila melanogaster and TRP4 in
(REF. 4), Nav1.5 (REFS 107,108), HCN109 and Cav (REF. 110), the proton
Caenorhabditis elegans), is directly activated by mechanical forces when
channel Hv1 (REF. 111) and certain subtypes of N-methyl-d-aspartate
heterologously expressed, and mutations in its predicted pore alter its ion
(NMDA) receptors112. Modulation of the activity of these channels in
selectivity103. Although TRPN1 orthologues do not exist in mammals,
cells exposed to mechanical stimuli has been implicated in modifying
its close relative TRPA1 has been implicated in mechanical sensation
cellular function.
but has not been shown to open when challenged by mechanical forces.
Another TRP channel, TRPV4, is insensitive to mechanical indentation Mechanically activated proteins
and membrane stretch, but it has been reported to open in response to Ion channels are not the only source of force sensation in cells.
elastomeric pillar array-mediated membrane stretch97. Some studies Cytoskeleton-associated elements, which connect the cytoskeletal
also indicated that TRPC6 is mechanically active in kidney podocytes; elements to the lipid bilayer, and other membrane proteins engage in
however, this is currently debated104. transmitting mechanical signals from the membrane to the cell interior
Degenerin–epithelial sodium–acid-sensing ion channel (DEG–ENaC– and thus have essential roles in cellular mechanotransduction.
ASIC): sodium-selective channels that have two transmembrane domains This mechanoresponse is mostly associated with changes in protein
and were first described as mechanosensitive in C. elegans105. Ablation conformation. In addition, mechanical stimuli can elicit changes in
of DEG‑1 abolishes 80% of the mechanically activated current in ASH enzymatic activity leading to changes in signalling pathways. One such
neurons and mutations in its predicted pore region alter ion selectivity, conserved pathway involves mechanotransduction by the transcription
suggesting that DEG-1 contributes to, or is very close to, the pore of the factors Yes-associated protein (YAP) and transcriptional co-activator with
channel. Despite intense efforts, a clear role for homologous channels PDZ-binding motif (TAZ), which, in response to different mechanical
in mammals is lacking. stimuli, translocate to the nucleus to change gene expression12,113.

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS
Extracellular space
Plasma
membrane
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38
Intracellular space
b Channel (top view) Channel (side view) c Model 1: Piezo1 causes local Model 2: Piezo1 in its innate state is planar
deformation of the membrane to the membrane and extends laterally
Lipid bilayer
Extracellular
side
Intracellular
side
Carboxy-terminal end Proposed peripheral helices
Cap domain Proposed pore-lining region
Proposed amino-terminal region
Figure 1 | Topology and structural orientation of Piezo1. a | Predicted the periphery. Each predicted domain
NatureisReviews
highlighted in the density
| Molecular map.
Cell Biology
topology map of PIEZO1 protein depicting 38 putative transmembrane Piezo1 density map created in Chimaera; Electron Microscopy Data
segments. b | The left panel shows a top-view cryo-electron microscopy Bank (EMDB) ID: 6343. c | The two predicted models for the orientation
density map of the mouse Piezo1 channel. Although the channel is a of the Piezo1 channel in the membrane: the tilt in the peripheral helices,
homotrimer, each subunit is depicted in different colours to highlight the relative to the membrane, could outline the deformation in the
positions of the monomers relative to one another. The right panel shows surrounding membrane (left panel), or in its native environment,
a side view depicting the shape of the purified protein (only two out of the peripheral helices and amino-terminal region are in line with the
the three subunits are visible). The central cap domain is indicated in membrane (right panel). The density map of Piezo1 is used to represent
purple. The carboxy-terminal end of the protein is in the centre of the the channel. Part a is adapted with permission from REF. 17, Macmillan
channel complex, and the amino-terminal end of the protein is in Publishers Limited.
Mechanisms of action In the second hypothesis, the ‘force-from-filaments’

Several physiologically relevant forces activate Piezo ion model, the interaction (tethering) of the ion channel
channels, but little is known about how these forces are with the extracellular matrix (ECM) or intracellular cyto
transduced to channel opening or what features within skeletal proteins (FIG. 2b) that change their conformation
the protein respond to force. In vitro functional studies in response to force might cause the channel to open31
have begun to characterize and quantify the physical (FIG. 2b). Homotetrameric NOMPC (no mechanoreceptor
forces sensed by Piezos, and based on these advances, potential C) channels composed of six transmembrane-
activation models are emerging. domain subunits are gated by this tether mechanism.
In this case, the N-terminus of NOMPC interacts with
Activation mechanism. Two general mechanisms intracellular microtubules, an association required
have been hypothesized to underlie the activation of for mechanogating 33.
mechanically activated ion channels. In the ‘force- There is now evidence that Piezo1 is gated, at least
from-lipids’ model, membrane tension can alter lipid– in part, by direct membrane tension (‘force-from-lipid’).
protein interactions between the membrane and the Mechanical perturbation of reconstituted mouse PIEZO1
ion channel and result in channel activation31 (FIG. 2a). protein in droplet interface bilayers (containing no other
Bacterial mechanically activated ion channels and cellular components) opens Piezo1 with an estimated
some vertebrate channels belonging to the two-pore force of 3.4 mN/m (REF. 34). Similarly, lateral membrane
domain potassium (K2P) channel family behave accord- tension as low as 1.4 mN/m activates the channel in cellu-
ing to this model. When force is applied to the mem- lar membranes35,36. Interestingly, despite its large size and
brane, tension causes reorganization of lipids within unprecedented numbers of predicted transmembrane
and around the protein, causing the channel to open32. domains among membrane proteins, expression of

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS
PIEZO1 in human embryonic kidney cells does not by Apart from a mechanistic understanding of Piezo
itself change membrane stiffness as measured by atomic channel gating, molecular determinants of the chan-
force microscopy (AFM)37, suggesting that the presence nel’s mechanosensor also remain elusive. Based on the
of this large protein does not change the biophysical cryo‑EM structure, the N‑terminal region is suggested
properties of the membrane. Although it is possible that to play a role in mechanical force sensing 18,21; however,
Piezos can also be gated via ‘force-from-filament’ (FIG. 2b), the data supporting this claim have been questioned38.
direct evidence for this model is lacking (but see the Intriguingly, application of localized force using mag-
section on Modulation of Piezo activity below). netic nanoparticles to both the N‑terminal region and
the cap modulates the mechanosensitivity of Piezo1 but
is insufficient to directly activate the ion channel without
a Bilayers additional mechanical forces39. These functional studies
Proposed Piezo model 1 indicate that independent of the mechanosensor, specific
Channel closed regions of the protein are more susceptible to mechanical
perturbation than other regions.
Mode of activation. As Piezo1 is gated in part by direct

Channel open membrane tension, any physiological force that alters
membrane tension should, theoretically, activate the
channel. Indeed, a broad range of mechanical stimuli
robustly activate Piezos. Mechanical indentation of the
apical cell surface with a blunt glass probe or the appli-
Lateral cation of suction (negative pressure) to a membrane
Proposed Piezo model 2
membrane patch results in stretching of the membrane, and both
tension manipulations activate Piezos8. Such global or localized
stretching of the cell membrane is routinely experienced
by, for example, somatosensory nerve endings in the skin
and other organs, as well as by migrating cells. Piezo1 is
also activated by shear stress, that is, the frictional force
generated from fluid flow over cells, as exemplified by
sensing blood flow-mediated forces by endothelial cells
within arteries and veins40 (see also below).
In addition to forces applied to the apical surface
of lumen-facing cells, the basolateral surface of a cell
b Intact cells is also exposed to physical forces, and Piezos can be
Channel closed Channel open activated in these microenvironments. Accordingly,
Piezo-dependent force transduction at the cell–cell or
cell–matrix interface has been demonstrated by seeding
Force from and stretching cells on tensile substrates or on elasto-
lipids
meric pillar arrays41,42. AFM-driven compression of the
cell membrane also activates Piezo1, and this method
has helped to quantify the force required to activate
Piezo1 and the effect of membrane stiffness on Piezo1
activity (see next section)37,43. Finally, Piezos can also
Force from
filament? respond to repeated mechanical stimulation as observed
Actin during the rhythmic contraction and relaxation of blood
+ Actin vessels or during tactile vibration detection44.
Activity
– Actin
Modulation of Piezo activity
Owing to the widespread expression of Piezos and their
Force involvement in several physiological processes (see next
Figure 2 | Models of activation for Piezo channels.
Naturea | In either| of
Reviews the predicted
Molecular Cell Biology section for details), changes in the function of these chan-
conformations of Piezo1 (FIG. 1c), membrane tension could lead to channel opening. nels have profound consequences. For example, Piezo1
In the scenario where Piezo1 causes a deformation of the surrounding membrane, lateral gain‑of‑function mutations that slow channel inactiv
tension could lead to some ‘straightening’ of peripheral helices, ultimately causing ation kinetics by as little as 1.3‑fold have been associ-
channel opening. In addition, increased membrane tension could cause a conformational ated with dehydrated hereditary stomatocytosis (DHS),
change in the channel in the absence of any change in membrane curvature. b | The data
arguing that tight control of Piezo inactivation is impera
from experiments on synthetic lipid bilayers indicate that membrane tension is sufficient
to activate Piezo1, suggesting a force-from-lipid activation model. However, the actin tive45–47. Strikingly, at negative resting membrane poten-
cytoskeleton does have a role in fine-tuning the sensitivity of Piezo1 to mechanical tials, Piezo1 channels are largely in an inactivated state,
force in intact cells. Disruption of actin polymerization by cytochalasin D shifts Piezo1 suggesting that the percentage of channels that can be
channel mechanosensitivity to higher forces, as represented in the force–activity graph. readily activated at resting potentials is surprisingly low.
The Piezo1 density map (FIG. 1b) is used to illustrate the channel. Although the mechanism or physiological implications

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS
of this phenomenon are not completely understood, one states by affecting the conformational stability of any of
can speculate that conditions that alter this inactivated these states59. Additional studies are required to deter-
state could change the percentage of Piezo1 channels mine the mechanistic details of modulator-dependent
that can be activated48. Additionally, in some cell types, effects on Piezo channels. Altogether, these observations
variations in endogenous Piezo inactivation kinetics are suggest that Piezo activity can be fine-tuned in differ-
also observed, suggesting differential functions of Piezo- ent physiological environments, potentially to b etter
mediated mechanotransduction, but the cause of these suit the versatile roles of these channels in disparate
variations and their consequences on cellular function mechanotransduction pathways.
are not understood49,50.
Signalling pathways and proteins that change mem- Physiological roles of Piezo1
brane properties (such as membrane stiffness or curva- Piezo1 is expressed predominantly in non-neuronal
ture) have been shown to modulate Piezo channel activity cell types and is essential for transducing externally
by indirectly altering the channels’ mechanosensitivity. and internally applied forces at the plasma membrane.
For example, the threshold force to activate Piezo1 and In vivo characterization of phenotypes associated with
Piezo2 is reduced by co‑expression of stomatin-like Piezo1 deficiency or gain‑of‑function mutations have
protein 3 (STOML3; also known as SLP3). STOML3 is helped uncover many physiological functions mediated
related to the C. elegans mechanosensory protein MEC‑2 by Piezo1 channels.
Lipid rafts
and in mice is involved in normal cutaneous mechano
Membrane subdomains that
are ordered assemblies sensation51. It is proposed that by binding cholesterol and Cardiovascular mechanotransduction. The morpho
enriched in cholesterol, recruiting it to lipid rafts, STOML3 increases membrane logy and physiology of the heart and vasculature is
glycosphingolipids stiffness, making it easier for membrane tension to activ strongly influenced by mechanical forces. For example,
and proteins. ate Piezos37,42. In addition, several studies have suggested endothelial and smooth muscle cells that line mature
Bradykinin
that Piezo activity is modulated by cytoskeletal pro- blood vessels sense and respond to laminar and oscilla
An inflammatory peptide teins that have the ability to change membrane stiffness. tory shear stress and circumferential stretch60. During
that activates cognate Actin (in multiple cell types), dynamin (in chondrocytes) embryogenesis, shear stress initiates endothelial cell
G protein-coupled receptors. and filamin A (in smooth muscle) can alter Piezo1 activ- migration and alignment, which are essential events for
ity as a result of their effects on plasma membrane rigid- vessel formation and maturation61. In addition, shear
Dorsal root ganglion (DRG)
neurons ity 36,43,52. Other examples where Piezo activity is altered by stress also regulates vascular tone (the degree of vessel
Pseudo-unipolar sensory membrane properties include the suppression of Piezo1 constriction) by releasing endothelial-cell-derived relax-
neurons whose cell bodies are mechanically activated currents after treatment of cells ing factors, such as nitric oxide, which modulate arterial
located in the dorsal root of with cytochalasin D (an actin polymerization inhibitor)53 diameter 40,60. These mechanotransduction mechanisms
the spinal column and extend
a single bifurcating axon,
and the enhancement of Piezo1 and Piezo2 activity dur- have been well characterized in the cardiovascular sys-
one end of which innervates ing osmotic swelling, perhaps caused by increased resting tem during development and in adulthood; however, the
sensory organs and the other membrane tension53,54. identity and role of mechanically activated ion channels
which synapses onto neurons Piezo activity can also be regulated by other exogen in vascular physiology remained unknown until the
in the spinal cord. They are
ous and endogenous signals. For example, an acidic discovery of Piezos.
capable of detecting
mechanical, chemical and/or extracellular pH (~pH 6.3) attenuates Piezo1 currents PIEZO1 is expressed in embryonic and adult mouse
thermal stimuli and signal to (from exogenously expressed human PIEZO1) by stabil endothelial cells and is activated by shear stress62,63.
the central nervous system. izing the inactivated state of the channel55. As another Global and endothelial-cell-specific deficiencies
example, bradykinin potentiates Piezo2 activity in dorsal in Piezo1 in mice are embryonically lethal at mid-
Pericentrin
root ganglion (DRG) neurons56. In this case, bradykinin gestation, the developmental time point that coin-
A multifunctional scaffold for
anchoring proteins and protein receptor-mediated phosphorylation events enhance cides with the initiation of blood flow. In response to
complexes. Piezo2 channel activity up to sevenfold, in part by slow- shear stress, Piezo1‑mediated calcium influx activates
ing its inactivation kinetics. Interestingly, in DRG neu- c alpain, which proteolytically cleaves molecules associ-
Transient receptor potential
rons, Piezo2 is associated with pericentrin, knockdown of ated with focal adhesions, thereby disrupting the cell–
cation channel subfamily V
member 1 which enhances Piezo2‑mediated inactivating currents57. matrix attachment 62. Reorganization of focal adhesions
(TRPV1). An ion channel By contrast, activation of transient receptor potential cation within the cell facilitates endothelial cell migration and
activated by capsaicin and channel subfamily V member 1 (TRPV1; also known as alignment of cells in the direction of blood flow. Thus,
heat that is expressed in a capsaicin receptor and vanilloid receptor 1) attenuates Piezo1‑dependent signalling events enable endothelial
subpopulation of sensory
neurons, including nociceptors.
mechanically activated Piezo2 currents in DRG neurons cell reorganization, which is required for vasculogenesis
and in exogenously expressed HEK293 cells by depleting during development 62,63 (FIG. 3Aa).
Calpain membrane phosphoinositides (PI(4,5)P2 and PI(4)P)58. Notably, Piezo1 activity is also required for maintain-
A calcium-dependent cytosolic The underlying mechanisms by which pericentrin or ing basal blood pressure in adult mice, where it mediates
cysteine proteinase with roles
TRPV1 activation alter Piezo2 activity may include shear-stress-induced arterial dilation. Piezo1‑dependent
in cell migration, differentiation
and apoptosis. direct association with the channel or indirect effects calcium influx in response to shear stress leads to
on membrane tension. endothelial ATP release, which acts as a parahormone
Focal adhesions Biophysically, Piezo channel activity can be best to initiate P2Y2–Gq/G11 signalling. In endothelial cells,
Contact sites between cells described by a rudimentary four-state model comprising P2Y 2–G q/G 11-mediated signalling results in phos
and the extracellular matrix
that function as anchor points
a closed and an open state and two inactivated states44. phorylation of AKT and nitric oxide synthase to increase
for the cell and as biochemical Mutations in the channel or various modulators can nitric oxide production and release, which causes
signalling centres. potentially alter the rates of transition between these vasodilation26 (FIG. 3Aa).

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS
To maintain normal blood flow under physiological channel regulates the diameter and wall thickness of
conditions, blood vessels need to regulate both their small arteries during hypertension (FIG. 3Ab). Stretch-
lumen diameter and cross-sectional area. As a conse- induced Piezo1‑dependent calcium influx activates the
quence, arterial smooth muscle cells that surround ECM crosslinking enzyme transglutaminase, which
the endothelial layer experience variable amounts of has been implicated in vascular inward remodelling
stretch40. PIEZO1 is also expressed in smooth muscle (decreasing vessel diameter), placing Piezo1 upstream
cells of small-diameter arteries and mediates the of these events52. These studies indicate that as a shear
stretch-activated currents observed in these cells52. stress and stretch sensor, the Piezo1 channel mediates
Whereas PIEZO1 protein is not required for myogenic various mechanotransduction pathways essential for
tone in homeostatic conditions, the activity of this regulating cardiovascular physiology.
A Aa Endothelial cells Ab Smooth muscle cells Ac Red blood cells
Embryo
Piezo1 shear Piezo1 shear Piezo1 stretch Piezo1 stretch
Ca2+ Ca2+ Ca2+ Na+ and Ca2+
Calpain AKT
Proteolytic processing Transglutaminase Gardos channel
eNOS K+ efflux,
Components of focal adhesions ECM
• NO secretion remodelling water efflux and
Altered cell migration • Vasodilation dehydration
Endothelial cell organization Vascular tone Vascular remodelling Volume regulation
(increasing vascular
wall thickness)
No flow Flow H2O
H2O H2O
B Brain (side view)
Sensation of
light touch Ba
Nodose Merkel cell

and jugular LTMR
ganglia
Spinal cord
Dorsal root
Bc Lung ganglion
Bb Muscle
spindle
Sensing organ stretch

and regulating breathing Proprioception
(regulating limb position
and coordination)
Figure 3 | Physiological role of Piezo channels. A | Piezo1 is required for mechanotransduction

Nature Reviewsin the cardiovascular
| Molecular Cell Biology
system. PIEZO1 is expressed in different cell types in the cardiovascular system, including embryonic and adult
endothelial cells (part Aa) and smooth muscle cells (part Ab) in the blood vessels as well as in red blood cells (part Ac).
Below each cell type is a schematic outlining events resulting from the Piezo1‑mediated mechanotransduction pathway.
The cartoon below part A depicts the consequential cellular changes that occur upon Piezo1 activation in each cell type.
B | Piezo2 is required for sensory and respiratory mechanotransduction. PIEZO2 is expressed in low-threshold
mechanoreceptors (LTMRs, which are sensory neurons that innervate skin) and Merkel cells and is required for detecting
touch (part Ba), in sensory neuron endings that innervate skeletal muscle (muscle spindle) to detect changes in muscle
stretch (part Bb), and in vagal sensory neurons that innervate the lung and transduce mechanical stimuli (such as stretch)
exerted on the respective tissues to the central nervous system (part Bc). ECM, extracellular matrix; eNOS, endothelial
nitric oxide synthase.

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS
Mechanoregulation of red blood cell volume. Although by traction force in a substrate-stiffness-dependent

red blood cells (RBCs) experience considerable phys manner 71. Knockdown of PIEZO1 in these cells dis-
ical forces as they are transported within blood vessels rupts their ability to differentiate into neurons and pro-
and small capillaries, the possibility that mechanically motes their differentiation into neuron-supporting glial
activated ion channels could have a role in their physio cells (astrocytes), suggesting that the mechanically regu-
logy was unexplored until recently. This role came to lated lineage choice of neural stem cells is dependent on
light when the first reported human PIEZO1 mutation Piezo1 activity 71. In Xenopus laevis, tissue stiffness regu-
was described to cause DHS, a disease characterized by lates the rate of axon growth of retinal ganglion cells and
RBC dehydration and their increased permeability to the extent of their axonal branching in the optic tectum.
cations64. These abnormal haematological properties In this model, blocking Piezo1 activity via GsMTx4 or
suggested that Piezo1 activity is crucial for normal RBC piezo1 knockdown impedes in vitro and in vivo axonal
physiology. Subsequent studies in mice and zebrafish growth, suggesting that one of the pathways involved
demonstrated that Piezo1 is expressed on RBCs and is in substrate-driven neuronal mechanotransduction is
required for their volume regulation25,65,66. Functional Piezo1 dependent 72.
characterization of the human mutations demonstrated Several in vitro studies have also demonstrated that
that these amino acid changes caused slowing of Piezo1 knocking down Piezo1 in bladder and kidney epi-
inactivation kinetics (gain of function) and suggested thelial cells abolishes stretch-activated channel activ-
that the force-induced, Piezo1‑mediated, increased ity 50,73,74. In the kidney, deletion of Piezo1 specifically in
cationic influx into RBCs could be the cause of their urine-collecting ducts affects the ability of the kidney to
dehydration45–47. Indeed, stretch activation of Piezo1 concentrate urine after dehydration or fasting 74 via an
in RBCs causes calcium influx, which in turn activates unknown pathway. Overall, even though the functions
the calcium-activated potassium channel KCa3.1 (the of the Piezo1 channel in animal tissues are apparent,
Gardos channel), resulting in the efflux of K+ and water, the importance of Piezo1 in these various physiological
causing dehydration25 (FIG. 3Ac). contexts will have to be assessed in vivo in various organ-
isms to shed more light on how this channel regulates
Epithelial homeostasis. Epithelia serve as barriers animal physiology.
and thus need to maintain their integrity as a uniform
cellular sheet. During migration and division, epithe- Physiological roles of Piezo2
lial cells are stressed externally from neighbouring Piezo2 has important roles in sensory processes such as
cells or the ECM, as well as internally via their cyto the detection of (light) touch and proprioception and,
skeletal networks, and accordingly, both these processes rather unexpectedly, has also been shown to function in
are strongly regulated by mechanotransduction67,68. respiratory physiology. Other roles of Piezo2 in human
In zebrafish, piezo1 knockdown disrupted epithelial physiology have also surfaced through the study of
homeostasis by impeding cell extrusion induced by patients with inactivating or gain‑of‑function mutations
cell crowding, which resulted in epidermal cell mass in PIEZO2. This section discusses our understanding of
formation41. Interestingly, knocking down piezo1 in Piezo2 contributions to animal physiology, which were
zebrafish epithelial cells also attenuated cell division69. identified from mouse models and human diseases.
Piezo1 has been proposed to regulate cell extrusion
and division through its activity not just at the plasma Mechanotransduction in the sensory system. Somato
membrane but also at other intracellular compartments sensation provides important information about our
such as the nuclear envelope and endoplasmic reticu- environment and internal state and is dependent in
lum. These effects of Piezo1 on epithelial homeostasis part on mechanotransduction. External physical forces
can potentially affect processes such as wound healing applied to the skin, as well as mechanical forces within
or lead to tumorigenesis in humans. Indeed, mutations internal organs, are sensed by peripheral sensory neu-
in PIEZO1 were identified in some patients with colo rons that mostly reside in the DRG and various cranial
rectal adenomatous polyposis70. The consequence of ganglia (such as trigeminal and nodose). These neurons
Traction force these mutations on Piezo1 function has not been exam- relay information from the periphery to the brain8,75
A type of force generated ined yet, but it can be speculated that abnormal Piezo1 (FIG. 3B). Mechanical stimuli known to elicit neuronal
within the cell in response function in colorectal epithelial cells could contribute responses range from gentle brushing and noxious pin-
to change in its environment.
These forces are a result of
to tumorigenic polyp formation. Overall, additional prick of the skin to forces generated by internal organ
interactions between actin in vivo studies are needed to examine the importance distension. Disparate mechanically activated ion chan-
filaments, adhesion molecules of Piezo1‑mediated mechanotransduction for epithelial nels were postulated to sense these varied mechanical
and the extracellular matrix. homeostasis in mammals. stimuli; however, until recently, the identity of these
channels had remained a mystery.
Nociceptors
A subpopulation of Other biological processes. Substrate stiffness and tex- It was first shown in Drosophila melanogaster that
non-myelinated or lightly ture can regulate processes such as cell migration and Piezo is involved in detecting noxious mechanical p oking
myelinated sensory neurons neuronal axon growth, and mechanically activated applied to the dorsal side of the larva76. Although PIEZO2
that detect noxious stimuli ion channels have been implicated in these processes. is expressed in vertebrate nociceptors, a clear nocifensive
and are the initial receptors
that signal the presence of a
Recent studies have indicated a role for Piezo1 in the phenotype (that is, an inability to sense noxious mech
stimulus that can cause differentiation of neural stem cells. PIEZO1 is expressed anical stimuli) related to Piezo2 channel activity has
damage to the organism. in human neural progenitor cells and is activated not yet been observed6. Interestingly, PIEZO2 protein

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS
is critical for the sensation of innocuous touch and also be linked to chronic obstructive pulmonary disease,
mechanical forces in a number of model systems a condition where the Hering–Breuer reflex is attenu-
and in humans. In zebrafish, knocking down piezo2 ated91. The unexpected roles of Piezo2 in respiration
mRNA specifically in touch receptors during develop- highlight the potential requirement of this channel in
ment resulted in diminished tactile sensitivity of the sensory neurons innervating other internal sites exposed
embryo77. In mice, PIEZO2 is expressed in low-threshold to mechanical forces, including the aortic arch, stomach,
mechanoreceptor fibres as well as in specialized cuta- bladder and intestine (see also below).
neous mechanosensory epithelial cells known as Merkel
cells (FIG. 3Ba) . Deletion of Piezo2 in both sensory Other biological processes. One of the pressing
neurons and Merkel cells resulted in severe deficits in unknowns in the field of auditory sensation has been the
the s ensation of light touch78–81. identity of the mechanotransduction ion channel that is
Proprioception is based on sensing the degree of responsible for the conversion of forces elicited by sound
muscle stretch and provides us with the ability to sense waves into electrical signals in the inner ear (see BOX 1
limb position in space and to balance. An essential role for details). Although many candidate transmembrane
for Piezo2 in proprioception was recently described. channel-like proteins (such as TMC1, TMC2, THMS and
PIEZO2 is expressed in proprioceptive neurons in TMIE) localize at the site of mechanot ransduction
mice, and strikingly, selective deletion of Piezo2 in this and are required for auditory function, none of them
subpopulation of sensory neurons abolished muscle have been validated as the pore-forming ion channel
stretch-induced firing of these neuronal fibres and that underlies hearing 92. Additionally, PIEZO2 (but not
led to severe deficits in body coordination and limb PIEZO1) is expressed in cochlear and vestibular hair
position82,83 (FIG. 3Bb). cells at the apical side of the hair cell body, but func-
Remarkably, recent reports have demonstrated that tional studies have shown that it is not the long sought
PIEZO2 loss‑of‑function mutations in humans also leads after channel mediating auditory transduction. Mice
to loss of discriminative touch and proprioception84–87. lacking Piezo2 in the inner ear have only mild a uditory
To a large extent, the sensory deficits that these patientsdefects, which indicates that Piezo2 is not a part of the
experience phenocopies those observed in Piezo2 knock- main mechanosensory transduction complex 93. Piezo1
out mice. These results unequivocally demonstrate that and Piezo2 double knockout mice have the same pheno
Piezo2 is the primary mechanotransduction channel for type as Piezo2-knockout mice, ruling out the possibil
the sense of touch and proprioception. As pain pheno ity of a compensatory effect of PIEZO1 expression on
types have not yet been reported in mice or humans the putative Piezo2‑mediated mechanotransduction.
lacking functional Piezo2 channels, the future identifi However, Piezo2 was found to be responsible for the
cation of high-threshold mechanically activated ion ‘reverse p olarity current’ elicited when the hair cell
channels in sensory fibres underlying the transduction bundle is deflected in the direction opposite the normal
of noxious stimuli is of great importance. direction of stimulation. This reverse polarity current
is distinct from the current mediated by the p rimary
Mechanotransduction in the respiratory system. Unlike hair cell mechanotransduction channel, suggesting that
the case for touch and proprioception, the precise role two independent mechanically activated ion channels
of mechanotransduction in internal organs is less well are expressed in hair cells. In addition, Piezo2 function
Merkel cells defined. Constitutive and peripheral nervous system- was associated with the regulation of maturation of the
Specialized epithelial cells specific Piezo2‑deficient newborn mice die owing to primary mechanosensing machinery 93. As the reverse
present in the skin that are
in close contact with the
respiratory distress88. Piezo2 activity in s ensory neu- polarity current is activated only during hair cell matur
peripheral terminals of rons innervating the lung was also found to be crucial ation or when the sensory epithelium is disrupted,
low-threshold sensory neurons. for sensing organ stretch in adult mice; deletion of Piezo2‑mediated mechanotransduction could have
Piezo2 specifically in sensory neurons innervating the a regulatory role during hair cell development and in
Vagal neurons
lung markedly reduced lung stretch-induced firing of pathophysiological conditions.
Sensory neurons that have
their cell bodies within the vagal neurons and caused abnormal breathing p atterns In the gastrointestinal tract, enzyme secretion, bowel
nodose and jugular ganglia in adult mice. Piezo2 was also found to mediate the movement and the sensation of noxious stimuli are regu
situated in the vagus nerve Hering–Breuer reflex (apnoea to prevent excessive lated by enterochromaffin cells and other luminal epithe-
(tenth cranial nerve) and that stretching of the lung) in mice88 (FIG. 3Bc). Interestingly, lial cells via mechanotransduction pathways94. PIEZO2
innervate internal organs.
humans with inactive PIEZO2 variants survive birth is expressed in enterochromaffin cells and contributes
Arthrogryposis but, importantly, show shallow breathing at infancy 84,86. to mechanically activated currents in these cells95,96.
A condition characterized by Individuals with PIEZO2 gain‑of‑function mutations Future studies using in vivo models will probably fur-
joint contractures causing present with arthrogryposis and restrictive lung dis- ther elucidate the role of Piezo2 as a mechanosensor in
immobility of the joints,
ease89,90. Thus, Piezo2‑mediated respiratory function gastrointestinal physiology.
which is generally associated
with other disorders. seems to be fairly conserved. These findings under- The skeletal system is exposed to extensive mechani
line the role of mechanotransduction in respiration and cal loading. In some but not all species, Piezo2 contributes
Enterochromaffin cells raise the possibility that aberrant Piezo2 function might to signalling in chondrocytes. In porcine chondrocytes,
A subset of cells in the contribute to respiratory disorders, such as sudden Piezo1 and Piezo2 together mediate calcium signals
epithelium of the lumen of
the gastrointestinal tract that
infant death syndrome and asthma, which may involve when compressed at hypertrophic levels by AFM43.
regulates secretion of enzymes abnormal activity of sensory neurons innervating the In murine chondrocytes, detectable levels of Piezo1
and bowel movement. upper respiratory tract. Aberrant Piezo2 activity may (but not Piezo2) and TRPV4 protein were reported,

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS
Table 1 | List of PIEZO gene mutations with associated human disorders

Disease Description Protein alteration cDNA change
PIEZO1
Dehydrated Autosomal dominant G718S c.2152 G>A
hereditary haemolytic anaemia
G782S c.2344 G>A
stomatocytosis45,46,64,114 characterized by
erythrocyte dehydration. R808Q c.2423 G>A
Red blood cells show S1117L c.3350 C>T
abnormal ionic content
R1358P Not reported
K1878del c.5632–5634delAAG
R1943Q c.5828 G>A
A2003D c.6008 C>A
A2020T Not reported
A2020V c.6059 C>T
R2088G c.6262 C>G
T2127M c.6380 C>T
2166–2169 deletion of four Ks c.6495–6508delAGA
M2225R Not reported
R2302H c.6905 G>A
R2456H c.7367 G>A
R2488Q c.7463 G>A
E2496ELE (insertion of LE) Not reported
Colorectal Autosomal dominant H702Y c.2104 C>T
adenomatous disorder characterized by
I1007M c.3021 C>G
polyposis70 predisposition to cancer
V1712M c.5134 G>A
Y1763stop c.5289 C>G
R1955C c.5863 C>T
Generalized lymphatic Autosomal recessive E755stop c. 2263 G>T
dysplasia99,100 disorder characterized by
L939M c.2815 C>A
lymphoedema affecting
all body segments and S1153Yfs*21 c.3455 + 1G>A#
by pleural effusions E1630stop c.4888 G>T
G2029R c.6085 G>C
V2171F c.6511 G>T
Q2228stop c.6682 C>T
P2430L c.7289 C>T
R2456C c.7366 C>T
F2458L c.7374 C>G
PIEZO2
Distal Autosomal dominant M712V c.2134 A>G
arthrogryposis89,90,115 disorder characterized by
I802F c.2024 A>T
cleft palate and multiple
congenital contractures M998T c.2993 T>C
A1486P c.4456 G>C
T2221I c.6662 C>T
S2223L c.6668 C>T
T2356M c.7067 C>T
R2686H c.8057 G>A
R2718L c.7067 C>T
R2718P c.8153 G>C
E2727del c.8179–8181delGAA
Y2737Ifs*7 c.8208delA
S2739P c.8215 T>C

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS
Table 1 (cont.) | List of PIEZO gene mutations with associated human disorders
Disease Description Protein alteration cDNA change
PIEZO2 (cont.)
Marden–Walker Autosomal recessive, rare R2686C c.8056 C>T
syndrome115 connective tissue disorder
Gordon syndrome115 Autosomal dominant W2746stop c.8238_8245delGACTAGAG
and belongs to distal
R2686H c.8057 G>A
arthrogryposis disorders
Touch and Deficits in touch and R462stop c.1384 C>T
proprioceptive vibration sensation.
R1575stop c.4723 C>T
phenotype84–87 Proprioceptive defects;
loss of joint-position R1685stop c.5053 C>T
sense. Most patients also S517Tfs*48 c.1550_1552delGCTinsCGAA
exhibit scoliosis and joint
contractures S903stop c.2708 C>G
P1007Lfs*3 c.3019_3029delCCTGAGAACTT
R1658P c.5053 C>T
L1874fs*5 c.5621delT
fs*N refers to mutations leading to frameshift variants, with N denoting the length of the shifted reading frame. #3455 + 1G>A
denotes a G to A substitution at nucleotide +1 of an intron positioned between nucleotides 3455 and 3456 of the cDNA sequence.
and these channels together mediate the mechani- cause distal arthrogryposis 5 (DA5), a congenital dis-
cally activated current when the cell–matrix interface order characterized by severe joint contractures89. It is
is stimulated97. Obviously, we are only now starting thought that DA5 is a disease of muscles and joints,
to understand the roles of Piezos in internal organs and it is likely that Piezo2‑dependent hyperactivity in
and the skeletal system, and genetic models will be these tissues could contribute to the disease phenotype.
necessary to more precisely link Piezo-mediated However, at least in mice, PIEZO2 is not expressed in
mechanotransduction to their (patho)physiology. these tissues. Another possibility is that the disease
manifestation is due to hyperactive peripheral sen-
Human diseases linked to Piezos sory neuron activity. If so, then dampening of action
Whole-exome sequencing of patients with various dis- potential firing in PIEZO2‑expressing neurons may be
orders across different geographical and ethnic back- therapeutically beneficial in these patients.
grounds has identified novel gain‑of‑f unction and
loss‑of‑function mutations in PIEZO1 and PIEZO2 Conclusions and perspectives
genes. TABLE 1 summarizes these mutations and associ Various physiological processes dependent on and
ated disorders. Some of these disease phenot ypes influenced by mechanical inputs have traditionally been
confirm results from genetic studies in the mouse84–87 studied as separate entities. This approach was based on
(for example, touch and proprioception). Other pheno- the assumption that the key force-sensing components
types are unexpected and have triggered mechanistic of mechanotransduction machinery are distinct in dif-
studies in mice to understand how mutations in Piezos ferent systems (for example, gentle touch sensation by
cause the phenotypes observed in humans. For exam- peripheral sensory neurons and blood flow sensation
ple, human gain‑of‑function PIEZO1 mutations that by endothelial cells). The fact that the closely related
cause DHS have helped elucidate the role of Piezo1 in ion channels Piezo1 and Piezo2 have important roles in
RBC volume regulation45–47,64 (see above). Interestingly, diverse aspects of mechanotransduction has started to
there is evidence suggesting that mouse RBC dehydra unify disparate fields. Analysing the function of Piezo
tion is linked to reduced Plasmodium infection in vitro, channels in vivo has also helped to define the role of
and a DHS-mutation-carrying Piezo1 mouse model cellular mechanotransduction in biological processes
is less susceptible to Plasmodium-mediated cerebral where a precise role for mechanical forces was not
malaria. Strikingly, RBCs from individuals within clear. Examples include the role of Piezo1 in RBC vol-
the African population carrying a novel PIEZO1 ume regulation and Piezo2 in pulmonary respiration.
gain‑of‑function mutation are protected against in vitro Intracellular mechanobiology was largely thought to be
Plasmodium infection98. Generalized lymphatic dys- mediated by transmembrane proteins such as integrins,
plasia (GLD) co‑segregates with loss‑of‑function but the emerging role of the Piezo1 ion channel in trac-
mutations in PIEZO1 and is characterized by con- tion force sensing can potentially remodel this concept.
genital lymphoedema 99,100. Identification of these Furthermore, identification of phenotypes associated
GLD-associated mutations suggests an unexpected with human mutations in PIEZO1 and PIEZO2 has had
role for Piezo1 in the lymphatic system, and future a significant impact on our understanding of the in vivo
studies are needed to determine how Piezo1 contrib- roles of mechanotransduction-based processes and will
utes to the development and regulation of this system. undoubtedly shed light on the underappreciated causes
Finally, gain‑of‑function mutations in PIEZO2 can of the associated disorders.

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS
Despite the recent surge in Piezo-dependent Piezos on various mechanosensory events also raises
mechanotransduction research, several aspects of Piezo the possibility that these channels may associate with
channel architecture and physiology are still unknown. other known or unknown proteins to modify cellular
At the structural level, questions concerning the molecu signalling. Finally, in scenarios where Piezos fall short
lar determinants of the pore and the mechanosensor, of being the primary mechanotransduction channel,
and conformational changes leading to channel activ other novel ion channels remain to be identified. For
ation and inactivation remain to be answered. It will be instance, in sensory neurons, it is evident that apart from
important to decipher the features of Piezos that make Piezo2, there is at least one other (and probably more)
them such versatile mechanosensors. At the level of mechanically activated ion channel that detects high-
cell and systems biology, we have yet to understand the threshold forces. Current and future research on Piezos
mechanistic details of Piezo-dependent processes and to will continue to pave the way to a better understanding
determine the full extent of Piezo-mediated mechano of mechanotransduction processes and their importance
transduction in the body. The widespread impact of in regulating biological processes.
1. Corey, D. P. & Hudspeth, A. J. Response latency 20. Gnanasambandam, R., Bae, C., Gottlieb, P. A. 40. Davies, P. F. Flow-mediated endothelial
of vertebrate hair cells. Biophys. J. 26, 499–506 & Sachs, F. Ionic selectivity and permeation properties mechanotransduction. Physiol. Rev. 75, 519–560
(1979). of human PIEZO1 channels. PLoS ONE 10, e0125503 (1995).
2. Guharay, F. & Sachs, F. Stretch-activated single ion (2015). 41. Eisenhoffer, G. T. et al. Crowding induces live cell
channel currents in tissue-cultured embryonic chick 21. Zhao, Q. et al. Ion permeation and extrusion to maintain homeostatic cell numbers
skeletal muscle. J. Physiol. 352, 685–701 (1984). mechanotransduction mechanisms of in epithelia. Nature 484, 546–549 (2012).
3. Katta, S., Krieg, M. & Goodman, M. B. Feeling force: mechanosensitive Piezo channels. Neuron 89, 42. Poole, K., Herget, R., Lapatsina, L., Ngo, H.‑D.
physical and physiological principles enabling sensory 1248–1263 (2016). & Lewin, G. R. Tuning Piezo ion channels to detect
mechanotransduction. Annu. Rev. Cell Dev. Biol. 31, 22. Coste, B. et al. Piezo proteins are pore-forming molecular-scale movements relevant for fine touch.
347–371 (2015). subunits of mechanically activated channels. Nature Nat. Commun. 5, 3520 (2014).
4. Morris, C. E., Prikryl, E. A. & Joós, B. 483, 176–181 (2012). 43. Lee, W. et al. Synergy between Piezo1 and Piezo2
Mechanosensitive gating of Kv channels. PLoS ONE 23. Syeda, R. et al. Chemical activation of the channels confers high-strain mechanosensitivity to
10, e0118335 (2015). mechanotransduction channel Piezo1. eLife 4, articular cartilage. Proc. Natl Acad. Sci. USA 111,
5. Delmas, P., Hao, J. & Rodat-Despoix, L. Molecular e07369 (2015). E5114–E5122 (2014).
mechanisms of mechanotransduction in mammalian 24. Lacroix, J. J., Botello-Smith, W. M. & Luo, Y. 44. Lewis, A. H., Cui, A. F., McDonald, M. F. & Grandl, J.
sensory neurons. Nat. Rev. Neurosci. 12, 139–153 Chemical gating of the mechanosensitive Piezo1 Transduction of repetitive mechanical stimuli
(2011). channel by asymmetric binding of its agonist Yoda1. by Piezo1 and Piezo2 ion channels. Cell Rep. 19,
6. Coste, B. et al. Piezo1 and Piezo2 are essential Preprint at bioRxiv http://dx.doi. 2572–2585 (2017).
components of distinct mechanically activated cation org/10.1101/169516 (2017). 45. Albuisson, J. et al. Dehydrated hereditary
channels. Science 330, 55–60 (2010). 25. Cahalan, S. M. et al. Piezo1 links mechanical forces stomatocytosis linked to gain‑of‑function mutations
This study reports the identification of Piezos. to red blood cell volume. eLife 4, e07370 (2015). in mechanically activated PIEZO1 ion channels.
7. Holle, A. W. & Engler, A. J. More than a feeling: 26. Wang, S. et al. Endothelial cation channel PIEZO1 Nat. Commun. 4, 1884 (2013).
discovering, understanding, and influencing controls blood pressure by mediating flow-induced 46. Andolfo, I. et al. Multiple clinical forms of
mechanosensing pathways. Curr. Opin. Biotechnol. ATP release. J. Clin. Invest. 126, 4527–4536 (2016). dehydrated hereditary stomatocytosis arise from
22, 648–654 (2011). This study reports a role for PIEZO1 in regulating mutations in PIEZO1. Blood 121, 3925–3935
8. Ranade, S. S., Syeda, R. & Patapoutian, A. basal blood pressure in mice. (2013).
Mechanically activated ion channels. Neuron 87, 27. Bae, C., Sachs, F. & Gottlieb, P. A. The 47. Bae, C., Gnanasambandam, R., Nicolai, C., Sachs, F.
1162–1179 (2015). mechanosensitive ion channel Piezo1 is inhibited by & Gottlieb, P. A. Xerocytosis is caused by mutations
9. Sun, Z., Guo, S. S. & Fässler, R. Integrin-mediated the peptide GsMTx4. Biochemistry 50, 6295–6300 that alter the kinetics of the mechanosensitive
mechanotransduction. J. Cell Biol. 215, 445–456 (2011). channel PIEZO1. Proc. Natl Acad. Sci. USA 110,
(2016). 28. Alcaino, C., Knutson, K., Gottlieb, P. A., Farrugia, G. E1162–E1168 (2013).
10. Honoré, E., Martins, J. R., Penton, D., Patel, A. & Beyder, A. Mechanosensitive ion channel Piezo2 48. Moroni, M., Servin-Vences, M. R., Fleischer, R.
& Demolombe, S. in Reviews of Physiology, is inhibited by D-GsMTx4. Channels (Austin) 11, & Lewin, G. R. Voltage-gating of mechanosensitive
Biochemistry and Pharmacology Vol. 169 245–253 (2017). PIEZO channels. Preprint at bioRxiv http://dx.doi.org/
(eds Nilius, B. et al.) 25–41 (Springer International 29. Pliotas, C. et al. The role of lipids in 10.1101/156489 (2017).
Publishing, 2015). mechanosensation. Nat. Struct. Mol. Biol. 22, 49. Schneider, E. R. et al. Neuronal mechanism for acute
11. Parpaite, T. & Coste, B. Piezo channels. Curr. Biol. 27, 991–998 (2015). mechanosensitivity in tactile-foraging waterfowl.
R250–R252 (2017). 30. Brohawn, S. G., del Mármol, J. & MacKinnon, R. Proc. Natl Acad. Sci. USA 111, 14941–14946
12. Nourse, J. L. & Pathak, M. M. How cells channel their Crystal structure of the human K2P TRAAK, a lipid- (2014).
stress: interplay between Piezo1 and the cytoskeleton. and mechano-sensitive K+ ion channel. Science 335, 50. Peyronnet, R. et al. Piezo1‑dependent stretch-
Semin. Cell Dev. Biol. http://dx.doi.org/10.1016/ 436–441 (2012).  activated channels are inhibited by Polycystin‑2
j.semcdb.2017.06.018 (2017). 31. Árnadóttir, J. & Chalfie, M. Eukaryotic in renal tubular epithelial cells. EMBO Rep. 14,
This review summarizes the emerging role mechanosensitive channels. Annu. Rev. Biophys. 1143–1148 (2013).
of PIEZO1 in internal force sensing. 39, 111–137 (2010). 51. Lapatsina, L., Brand, J., Poole, K., Daumke, O.
13. Li, W., Gao, N. & Yang, M. Current Topics in 32. Brohawn, S. G., Campbell, E. B. & MacKinnon, R. & Lewin, G. R. Stomatin-domain proteins. Eur. J. Cell
Membranes (Academic Press, 2016). Physical mechanism for gating and mechanosensitivity Biol. 91, 240–245 (2012).
14. Wu, J., Lewis, A. H. & Grandl, J. Touch, tension, and of the human TRAAK K+ channel. Nature 516, 52. Retailleau, K. et al. Piezo1 in smooth muscle cells
transduction – the function and regulation of Piezo ion 126–130 (2014). is involved in hypertension-dependent arterial
channels. Trends Biochem. Sci. 42, 57–71 (2016). 33. Zhang, W. et al. Ankyrin repeats convey force to gate remodeling. Cell Rep. 13, 1161–1171 (2015).
This review highlights aspects of Piezo function. the NOMPC mechanotransduction channel. Cell 162, 53. Gottlieb, P. A., Bae, C. & Sachs, F. Gating the
15. Walsh, C. M., Bautista, D. M. & Lumpkin, E. A. 1391–1403 (2015). mechanical channel Piezo1: a comparison between
Mammalian touch catches up. Curr. Opin. Neurobiol. 34. Syeda, R. et al. Piezo1 channels are inherently whole-cell and patch recording. Channels (Austin) 6,
34, 133–139 (2015). mechanosensitive. Cell Rep. 17, 1739–1746 (2016). 282–289 (2012).
16. Kamajaya, A., Kaiser, J. T., Lee, J., Reid, M. 35. Lewis, A. H. & Grandl, J. Mechanical sensitivity 54. Jia, Z., Ikeda, R., Ling, J., Viatchenko-Karpinski, V.
& Rees, D. C. The structure of a conserved Piezo of Piezo1 ion channels can be tuned by cellular & Gu, J. G. Regulation of Piezo2 mechanotransduction
channel domain reveals a topologically distinct membrane tension. eLife 4, e12088 (2015). by static plasma membrane tension in primary
β sandwich fold. Structure 22, 1520–1527 (2014). 36. Cox, C. D. et al. Removal of the mechanoprotective afferent neurons. J. Biol. Chem. 291, 9087–9104
17. Coste, B. et al. Piezo1 ion channel pore properties are influence of the cytoskeleton reveals PIEZO1 is gated (2016).
dictated by C‑terminal region. Nat. Commun. 6, 7223 by bilayer tension. Nat. Commun. 7, 10366 (2016). 55. Bae, C., Sachs, F. & Gottlieb, P. A. Protonation of the
(2015). 37. Qi, Y. et al. Membrane stiffening by STOML3 human PIEZO1 ion channel stabilizes inactivation.
18. Ge, J. et al. Architecture of the mammalian facilitates mechanosensation in sensory neurons. J. Biol. Chem. 290, 5167–5173 (2015).
mechanosensitive Piezo1 channel. Nature 527, Nat. Commun. 6, 8512 (2015). 56. Dubin, A. E. et al. Inflammatory signals enhance
64–69 (2015). 38. Dubin, A. E. et al. Endogenous Piezo1 can confound Piezo2‑mediated mechanosensitive currents. Cell Rep.
This study reports the first medium-resolution mechanically activated channel identification and 2, 511–517 (2012).
cryo‑EM structure of PIEZO1. characterization. Neuron 94, 266–270.e3 (2017). 57. Narayanan, P. et al. Native Piezo2 interactomics
19. Prole, D. L. & Taylor, C. W. Identification and analysis 39. Wu, J., Goyal, R. & Grandl, J. Localized force identifies Pericentrin as a novel regulator of Piezo2
of putative homologues of mechanosensitive channels application reveals mechanically sensitive domains in somatosensory neurons. J. Proteome Res. 15,
in pathogenic protozoa. PLoS ONE 8, e66068 (2013). of Piezo1. Nat. Commun. 7, 12939 (2016). 2676–2687 (2016).

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS
58. Borbiro, I., Badheka, D. & Rohacs, T. Activation 80. Ranade, S. S. et al. Piezo2 is the major transducer of with non-immune hydrops fetalis. Nat. Commun. 6,
of TRPV1 channels inhibits mechanosensitive Piezo mechanical forces for touch sensation in mice. Nature 8085 (2015).
channel activity by depleting membrane 516, 121–125 (2014). 100. Lukacs, V. et al. Impaired PIEZO1 function in patients
phosphoinositides. Sci. Signal. 8, ra15 (2015). This study shows that in mice, PIEZO2 is involved with a novel autosomal recessive congenital lymphatic
59. Bae, C., Gottlieb, P. A. & Sachs, F. Human PIEZO1: in sensory mechanotransduction. dysplasia. Nat. Commun. 6, 83829 (2015).
removing inactivation. Biophys. J. 105, 880–886 81. Woo, S. H. et al. Piezo2 is required for Merkel-cell 101. Chalfie, M. Neurosensory mechanotransduction.
(2013). mechanotransduction. Nature 509, 622–626 (2014). Nat. Rev. Mol. Cell Biol. 10, 44–52 (2009).
60. Hahn, C. & Schwartz, M. A. Mechanotransduction in 82. Woo, S. H. et al. Piezo2 is the principal 102. Maingret, F., Patel, A. J., Lesage, F., Lazdunski, M.
vascular physiology and atherogenesis. Nat. Rev. Mol. mechanotransduction channel for proprioception. & Honoré, E. Mechano- or acid stimulation,
Cell Biol. 10, 53–62 (2009). Nat. Neurosci. 18, 1756–1762 (2015). two interactive modes of activation of the
61. Li, S., Huang, N. F. & Hsu, S. Mechanotransduction 83. Florez-Paz, D., Bali, K. K., Kuner, R. & Gomis, A. TREK‑1 potassium channel. J. Biol. Chem. 274,
in endothelial cell migration. J. Cell. Biochem. 96, A critical role for Piezo2 channels in the 26691–26696 (1999).
1110–1126 (2005). mechanotransduction of mouse proprioceptive 103. Delmas, P. & Coste, B. Mechano-gated ion channels
62. Li, J. et al. Piezo1 integration of vascular architecture neurons. Sci. Rep. 6, 25923 (2016). in sensory systems. Cell 155, 278–284 (2013).
with physiological force. Nature 515, 279–282 84. Chesler, A. T. et al. The role of PIEZO2 in human 104. Endlich, K., Kliewe, F. & Endlich, N. Stressed
(2014). mechanosensation. N. Engl. J. Med. 375, podocytes — mechanical forces, sensors, signaling
This is the first report of a role for PIEZO1 1355–1364 (2016). and response. Pflugers Arch. 469, 937–949 (2017).
in vascular development. This is the first paper to report that PIEZO2 105. Ben-Shahar, Y. Sensory functions for degenerin/
63. Ranade, S. S. et al. Piezo1, a mechanically activated is essential for mechanosensation in humans. epithelial sodium channels (DEG/ENaC). Adv. Genet.
ion channel, is required for vascular development in 85. Mahmud, A. A. et al. Loss of the proprioception 76, 1–26 (2011).
mice. Proc. Natl Acad. Sci. USA 111, 10347–10352 and touch sensation channel PIEZO2 in siblings 106. Hong, G.‑S. et al. Tentonin 3/TMEM150c confers
(2014). with a progressive form of contractures. Clin. Genet. distinct mechanosensitive currents in dorsal-root
64. Zarychanski, R. et al. Mutations in the 91, 470–475 (2016). ganglion neurons with proprioceptive function. Neuron
mechanotransduction protein PIEZO1 are associated 86. Delle Vedove, A. et al. Biallelic loss of proprioception- 91, 107–118 (2016).
with hereditary xerocytosis. Blood 120, 1908–1915 related PIEZO2 causes muscular atrophy with 107. Beyder, A. et al. Mechanosensitivity of Nav1.5,
(2012). perinatal respiratory distress, arthrogryposis, and a voltage-sensitive sodium channel. J. Physiol.
65. Faucherre, A., Kissa, K., Nargeot, J., Mangoni, M. E. scoliosis. Am. J. Hum. Genet. 99, 1406–1408 (2016). 588, 4969–4985 (2010).
& Jopling, C. Piezo1plays a role in erythrocyte volume 87. Haliloglu, G. et al. Recessive PIEZO2 stop mutation 108. Morris, C. E. & Juranka, P. F. Nav channel
homeostasis. Haematologica 99, 70–75 (2014). causes distal arthrogryposis with distal muscle mechanosensitivity: activation and inactivation
66. Shmukler, B. E. et al. Homozygous knockout of the weakness, scoliosis and proprioception defects. accelerate reversibly with stretch. Biophys. J. 93,
piezo1 gene in the zebrafish is not associated with J. Hum. Genet. 62, 497–501 (2016). 822–833 (2007).
anemia. Haematologica 100, e483–e485 (2015). 88. Nonomura, K. et al. Piezo2 senses airway stretch and 109. Lin, W., Laitko, U., Juranka, P. F. & Morris, C. E.
67. Macara, I. G., Guyer, R., Richardson, G., Huo, Y. mediates lung inflation-induced apnoea. Nature 541, Dual stretch responses of mHCN2 pacemaker
& Ahmed, S. M. Epithelial Homeostasis. Curr. Biol. 24, 176–181 (2016). channels: accelerated activation, accelerated
R815–R825 (2014). This paper demonstrates an unexpected role for deactivation. Biophys. J. 92, 1559–1572 (2007).
68. Eyckmans, J., Boudou, T., Yu, X. & Chen, C. S. PIEZO2 in respiratory mechanotransduction. 110. Calabrese, B., Tabarean, I. V., Juranka, P.
A hitchhiker’s guide to mechanobiology. Dev. Cell 21, 89. Coste, B. et al. Gain‑of‑function mutations in the & Morris, C. E. Mechanosensitivity of N‑type calcium
35–47 (2011). mechanically activated ion channel PIEZO2 cause channel currents. Biophys. J. 83, 2560–2574 (2002).
69. Gudipaty, S. A. et al. Mechanical stretch triggers rapid a subtype of distal arthrogryposis. Proc. Natl Acad. 111. Pathak, M. M. et al. The Hv1 proton channel responds
epithelial cell division through Piezo1. Nature 543, Sci. USA 110, 4667–4672 (2013). to mechanical stimuli. J. Gen. Physiol. 148, 405–418
118–121 (2017). 90. Okubo, M. et al. A family of distal arthrogryposis type (2016).
70. Spier, I. et al. Exome sequencing identifies potential 5 due to a novel PIEZO2 mutation. Am. J. Med. Genet. 112. Maneshi, M. M. et al. Mechanical stress activates
novel candidate genes in patients with unexplained A 167, 1100–1106 (2015). NMDA receptors in the absence of agonists. Sci. Rep.
colorectal adenomatous polyposis. Familial Cancer 91. Tryfon, S., Kontakiotis, T., Mavrofridis, E. & Patakas, D. 7, 39610 (2017).
15, 281–288 (2016). Hering-breuer reflex in normal adults and in patients 113. Ohashi, K., Fujiwara, S. & Mizuno, K. Roles of
71. Pathak, M. M. et al. Stretch-activated ion channel with chronic obstructive pulmonary disease and the cytoskeleton, cell adhesion and rho signalling
Piezo1 directs lineage choice in human neural stem interstitial fibrosis. Respiration 68, 140–144 (2001). in mechanosensing and mechanotransduction.
cells. Proc. Natl Acad. Sci. USA 111, 16148–16153 92. Zhao, B. & Müller, U. The elusive J. Biochem. 161, 245–254 (2017).
(2014). mechanotransduction machinery of hair cells. 114. Glogowska, E. et al. Novel mechanisms of PIEZO1
72. Koser, D. E. et al. Mechanosensing is critical for axon Curr. Opin. Neurobiol. 34, 172–179 (2015). dysfunction in hereditary xerocytosis. Blood http://
growth in the developing brain. Nat. Neurosci. 19, 93. Wu, Z. et al. Mechanosensory hair cells express two dx.doi.org/10.1182/blood-2017-05-786004 (2017).
1592–1598 (2016). molecularly distinct mechanotransduction channels. 115. McMillin, M. J. et al. Mutations in PIEZO2 cause
73. Miyamoto, T. et al. Functional role for Piezo1 in stretch- Nat. Neurosci. 20, 24–33 (2017). Gordon syndrome, Marden-Walker syndrome, and
evoked Ca2+ influx and ATP release in urothelial cell 94. Mawe, G. M. & Hoffman, J. M. Serotonin signalling distal arthrogryposis type 5. Am. J. Hum. Genet. 94,
cultures. J. Biol. Chem. 289, 16565–16575 (2014). in the gut — functions, dysfunctions and therapeutic 734–744 (2014).
74. Martins, J. R. et al. Piezo1‑dependent regulation of targets. Nat. Rev. Gastroenterol. Hepatol. 10,
urinary osmolarity. Pflugers Arch. 468, 1197–1206 473–486 (2013). Acknowledgements
(2016). 95. Wang, F. et al. Mechanosensitive ion channel Piezo2 The authors thank J. Kefauver, K. Marshall and K. Saotome
75. Carr, M. J. & Undem, B. J. Bronchopulmonary afferent is important for enterochromaffin cell response to for their helpful comments. A.P. is a Howard Hughes Medical
nerves. Respirology 8, 291–301 (2003). mechanical forces. J. Physiol. 595, 79–91 (2017). Institute investigator.
76. Kim, S. E., Coste, B., Chadha, A., Cook, B. 96. Tao, B. et al. Piezo2: a candidate biomarker for
& Patapoutian, A. The role of Drosophila Piezo visceral hypersensitivity in irritable bowel syndrome? Author contributions
in mechanical nociception. Nature 483, 209–212 J. Neurogastroenterol. Motil. 23, 453–463 (2017). A.P., S.E.M. and A.E.D. discussed the content of the article
(2012). 97. Rocio Servin-Vences, M., Moroni, M., Lewin, G. R. and contributed to the writing and editing of the manuscript.
77. Faucherre, A., Nargeot, J., Mangoni, M. E. & Poole, K. Direct measurement of TRPV4 and S.E.M. and A.E.D. also contributed by researching data for
& Jopling, C. piezo2b regulates vertebrate light touch PIEZO1 activity reveals multiple mechanotransduction the article.
response. J. Neurosci. 33, 17089–17094 (2013). pathways in chondrocytes. eLife 6, e21074 (2017).
78. Ikeda, R. et al. Merkel cells transduce and encode 98. Ma, S. et al. Common Piezo1 allele in African Competing interests statement
tactile stimuli to drive Aβ‑afferent impulses. Cell 157, populations causes xerocytosis and attenuates The authors declare no competing interests.
664–675 (2014). Plasmodium infection. Preprint at bioRxiv
79. Maksimovic, S. et al. Epidermal Merkel cells are http://dx.doi.org/10.1101/159830 (2017). Publisher’s note
mechanosensory cells that tune mammalian touch 99. Fotiou, E. et al. Novel mutations in PIEZO1 cause an Springer Nature remains neutral with regard to jurisdictional
receptors. Nature 509, 617–621 (2014). autosomal recessive generalized lymphatic dysplasia claims in published maps and institutional affiliations.

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.

Piezos Thrive Under Pressure

Uploaded by

Copyright:

Available Formats

Piezos Thrive Under Pressure

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Piezos Thrive Under Pressure

Uploaded by

Copyright:

Available Formats

REVIEWS

Piezos thrive under pressure:

NATURE REVIEWS | MOLECULAR CELL BIOLOGY ADVANCE ONLINE PUBLICATION | 1

2 | ADVANCE ONLINE PUBLICATION www.nature.com/nrm

Box 2 | Eukaryotic proteins involved in mechanotransduction

NATURE REVIEWS | MOLECULAR CELL BIOLOGY ADVANCE ONLINE PUBLICATION | 3

Mechanisms of action In the second hypothesis, the ‘force-from-filaments’

4 | ADVANCE ONLINE PUBLICATION www.nature.com/nrm

Mode of activation. As Piezo1 is gated in part by direct

NATURE REVIEWS | MOLECULAR CELL BIOLOGY ADVANCE ONLINE PUBLICATION | 5

6 | ADVANCE ONLINE PUBLICATION www.nature.com/nrm

A Aa Endothelial cells Ab Smooth muscle cells Ac Red blood cells

Piezo1 shear Piezo1 shear Piezo1 stretch Piezo1 stretch

Ca2+ Ca2+ Ca2+ Na+ and Ca2+

B Brain (side view)

Nodose Merkel cell

Sensing organ stretch

Figure 3 | Physiological role of Piezo channels. A | Piezo1 is required for mechanotransduction

NATURE REVIEWS | MOLECULAR CELL BIOLOGY ADVANCE ONLINE PUBLICATION | 7

Mechanoregulation of red blood cell volume. Although by traction force in a substrate-stiffness-dependent

8 | ADVANCE ONLINE PUBLICATION www.nature.com/nrm

NATURE REVIEWS | MOLECULAR CELL BIOLOGY ADVANCE ONLINE PUBLICATION | 9

Table 1 | List of PIEZO gene mutations with associated human disorders

10 | ADVANCE ONLINE PUBLICATION www.nature.com/nrm

NATURE REVIEWS | MOLECULAR CELL BIOLOGY ADVANCE ONLINE PUBLICATION | 11

12 | ADVANCE ONLINE PUBLICATION www.nature.com/nrm

NATURE REVIEWS | MOLECULAR CELL BIOLOGY ADVANCE ONLINE PUBLICATION | 13

You might also like