PIEZO ion channels: force sensors of the interoceptive nervous system

GRAPHICAL ABSTRACT

Many organs are designed to move: the heart pumps each second, the gastrointestinal tract squeezes and churns to digest food, and we contract and relax skeletal muscles to move our bodies. Sensory neurons of the peripheral nervous system detect signals from bodily tissues, including the forces generated by these movements, to control physiology. The processing of these internal signals is called interoception, but this is a broad term that includes a wide variety of both chemical and mechanical sensory processes. Mechanical senses are understudied, but rapid progress has been made in the last decade, thanks in part to the discovery of the mechanosensory PIEZO ion channels (Coste et al., 2010). The role of these mechanosensors within the interoceptive nervous system is the focus of this review. In defining the transduction molecules that govern mechanical interoception, we will have a better grasp on how these signals drive physiology.

Sensory cues arise from all organ systems, but the boundary between interoception and exteroception is not well defined. We are using the definition of skin as the border, with interoception referring to the processing of signals generated from within the body below the skin, a definition provided by the NIH (Chen et al., 2021). As the field continues to grow, consensus will be reached on which processes are included in interoception.

Mechanosensory molecules in the nervous system

PIEZO1 and PIEZO2 are sensitive, mechanically gated ion channels (Coste et al., 2010). In cultured dorsal root ganglia (DRG) sensory neurons, PIEZO ion channels are required for a large portion of mechanosensory responses (Murthy, 2023). The discovery of PIEZOs unraveled a long-standing mystery about the sensors responsible for fundamental mechanical senses, including touch (Ranade et al., 2014; Chesler et al., 2016). Evidence for endogenous chemical agonists of PIEZO ion channels is scant, which points to their prominent and specific role in mechanosensation.

We focus on PIEZO ion channels in this review because, in our view, these offer the best road to gain insights into mechanical interoception for the current state of the field. Their abundant expression throughout the sensory nervous system, in 7 out of 11 defined cell types of the nodose ganglion (Kupari et al., 2019), and almost all somatosensory neuron subtypes (von Buchholtz et al., 2021; Qi et al., 2023), as well as a rapidly growing body of research, indicates that they are important for numerous interoceptive processes (Figure 1). Other bona fide mechanosensory ion channels are expressed in the sensory nervous system, such as the TMEM63 cation channel family (Murthy et al., 2018a) and TREK and TRAAK potassium channels (Brohawn et al., 2014), and the molecular origin of some mechanical responses remains unknown (Murthy, 2023). Exciting future work will reveal the complete identities and physiological functions of these ion channel families, but PIEZO ion channels remain the mechanosensory channel family that has been implicated most widely as key transducers of mechanical force in the nervous system. Here we will outline the many roles PIEZO ion channels play in mediating mechanosensation directly within interoceptive sensory neurons.

Figure 1:

Summary of the known roles for PIEZO ion channels in neurons that contribute to mechanical interoception. Neuron types that mediate these functions are noted in parentheses. Abbreviations: (DRG) dorsal root ganglia; (N) nodose ganglion; (P) petrosal ganglion. Illustration is based on Leonardo Da Vinci’s Vitruvian Man drawing, with the blindfold signifying our ability to sense where our limbs are in space in the absence of visual cues (proprioception).

Interoception can occur by direct activation of sensory neurons as well as through non-neuronal cells that influence the nervous system. For example, specialized PIEZO2-positive mechanosensitive enterochromaffin cells release serotonin to influence gastrointestinal motility (Alcaino et al., 2018), and PIEZO-positive urothelial cells shape urinary reflexes (Marshall et al., 2020; Dalghi et al., 2021). The variety of mechanosensitive epithelial cell types and their physiological functions exceeds the scope of this review, but they undoubtedly play important roles in modulating or stimulating sensory neuron responses and physiology. Nonetheless, the direct activation of sensory neurons alone is sufficient to elicit rapid sensations and sensory reflex arcs to drive physiology and behavior.

Pushing, pulling, dancing, and prancing: PIEZO2 in proprioception

Proprioception defines our ability to sense the position of our limbs in space, allowing us to generate coordinated movements such as walking, running, swimming, or even performing intricately choreographed dance routines. This process is orchestrated by sensory feedback from the musculoskeletal system to shape motor activity. The main types of sensory end organs responsible for proprioception are Golgi tendon organs and muscle spindles, which are DRG-residing sensory neurons. These highly sensitive structures are strategically positioned within tendons and muscles to detect changes in muscle tension and muscle length (respectively). This informs the central nervous system of limb positions and initiates reflex arcs that protect against overstretching. An example of this is the classic patellar reflex, which is tested by a hammer tap under the knee.

Previous work has identified PIEZO2 as the main mechanotransduction ion channel for proprioception in mice (Woo et al., 2015; Florez-Paz et al., 2016) and humans (Chesler et al., 2016). Conditional PIEZO2-deficient mice display an abnormal gait and lack muscle stretch responses (Woo et al., 2015). Similarly, humans lacking PIEZO2 show severe sensory ataxia and dysmetria resulting from an inability to sense limb positions (Chesler et al., 2016). Despite the absence of proprioception, PIEZO2-deficient individuals are still capable of carrying out a variety of daily activities through learned compensatory mechanisms, such as using visual cues to keep track of where their limbs are in space (Chesler et al., 2016). While these individuals have taught us about the various functions of PIEZO2, of equal importance is learning what PIEZO2 is not responsible for: they can write, type, speak, chew, and carry out directed ballistic movements, all of which likely require some PIEZO2-independent mechanosensory feedback for coordination (Nagel & Chesler, 2022). Remaining sensations from skin stretch or deep pressure could serve this function. For more in-depth reading on the role of PIEZO2 in proprioception, a comprehensive review is available (Nagel & Chesler, 2022).

Discovery of PIEZO2 as the mammalian proprioceptive mechanosensor has been instrumental to understanding the pathophysiology behind debilitating diseases and identifying relevant therapeutic targets. For example, distal arthrogryposis is an inherited developmental syndrome that is characterized by the presence of multiple joint contractures in the absence of primary neurologic or muscular deficits, with one subtype (DA5) being linked to a PIEZO2 gain-of-function mutation in humans (Serra et al., 2022). Interestingly, Ma et al. (2023) discovered that muscular Botox injections within a critical developmental window can counteract PIEZO2 overactivity and alleviate some of the musculoskeletal deficits in DA5 mice, which has promising therapeutic potential (Ma et al., 2023). Angelman syndrome (AS) also affects the nervous system, resulting in movement and balance problems, among other deficits. Building on the observation that AS mice display a sensory ataxia phenotype that is reminiscent of Piezo2-deficiency, Romero et al. (2023) discovered that Ube3a-deficient AS mice display reduced PIEZO2 activity in sensory neurons (Romero et al., 2023). They suggest that this decrease in PIEZO2 activity could explain AS-associated locomotor deficits which are unaccounted for by deficits of cerebellar origin (Bruinsma et al., 2015; Romero et al., 2023). Remarkably, dietary supplementation of AS mice with linoleic acid can restore PIEZO2 activity back to normal levels. This was also observed with linoleic acid supplementation to culture conditions of AS patient-derived sensory neurons. The ability of dietary fatty acids to modulate cellular membrane properties, and thus mechanical gating, brings hope for non-invasive therapeutic intervention to patients suffering from debilitating diseases. This includes musculoskeletal pain, as PIEZO2 signaling contributes to functional deficits and pain behaviors in conditions like osteoarthritis (Obeidat et al., 2023). These studies demonstrate how understanding the basic physiology behind interoception, and characterizing the cells/molecules governing it, can directly contribute to advancing healthcare for a broader patient population.

How do you know when to go? PIEZOs in urination

Unlike most organs systems which rely on a combination of chemosensory and mechanosensory signals, the lower urinary tract (LUT) relies primarily on mechanical cues for normal function, making it an excellent model for studying mechanosensation. The urinary bladder operates in two modes: storage and voiding. During the filling phase, the bladder can maintain low internal pressure through an active sensory process which initiates detrusor muscle relaxation to accommodate the incoming urinary volume (de Groat, 2006). Upon crossing a certain stretch threshold, the nervous system prompts micturition reflexes to promote urination. These reflexes are governed by different LUT muscle groups which must work together to ensure continence and efficient bladder voiding (de Groat, 2006). For example, in storage mode, a guarding reflex is activated during early filling to contract the urethral sphincter and maintain continence. On the other hand, in voiding mode, urine flow through the urethra further engages micturition reflexes by enhancing detrusor muscle contraction to ensure complete emptying of the bladder (de Groat, 2006). These important reflexes rely on mechanical force sensing by different cell types to assess and convey the bladder’s filling status to the central nervous system and ensure normal urinary function. Both urothelial cells and sensory neurons innervating the LUT have been shown to be mechanosensitive (Apodaca et al., 2007; Merrill et al., 2016; Janssen et al., 2017; Marshall et al., 2020). In response to bladder stretch, the urothelium releases signaling molecules (e.g., ATP), which are thought to mediate downstream signaling for urinary reflexes (Ferguson et al., 1997; Mochizuki et al., 2009). Bladder stretch also directly induces increased excitation of innervating sensory neuronal afferents (Zagorodnyuk et al., 2009; Marshall et al., 2020). Ultimately, the sympathetic and parasympathetic systems control urinary tract motor functions, but the sensory neurons that initiate and coordinate these reflexes are thought to entirely arise from DRGs (de Groat, 2006).

Although many ion channels have been proposed to contribute to bladder mechanosensitivity, none had definitively been pinpointed as necessary for bladder fullness sensing (Cockayne et al., 2000; Mochizuki et al., 2009; Umans & Liberles, 2018). This changed with the discovery that PIEZO2 expression in both urothelial cells and bladder sensory neurons is required for proper bladder-filling sensation and micturition reflexes in mice and humans. Specifically, mice lacking PIEZO2 showed signs of poor urinary control and impaired urinary reflexes (Marshall et al., 2020). Moreover, humans with PIEZO2-deficiency reported an inability to sense when their bladders are filling, decreased voiding frequency, irregular urinary stream, and trouble fully emptying their bladders (Marshall et al., 2020). These data implicate PIEZO2 as a key mechanosensor that is important for low-threshold bladder fullness sensation and efficient urination in mammals. Detecting the difference between low and high volumes in the bladder allows us to sense different degrees of bladder fullness and know when to visit the restroom accordingly. Interestingly, in a mouse model lacking PIEZO2 in caudal tissues, DRG firing was absent during low, but not high, bladder pressure (Marshall et al., 2020). This is consistent with PIEZO2-deficient individuals reporting bladder fullness sensation only when urgently full (Marshall et al., 2020), and suggests an unknown ion channel is responsible for high bladder pressure sensation. Intriguingly, the ion channel that detects high threshold mechanical stimuli in the skin, like a pinprick, is also unknown. These could be shared mechanisms, much like innocuous mechanosensation in the skin and bladder both rely on PIEZO2. Thus, the high threshold mechanosensory ion channel will be an important molecule to identify in future work.

Lower urinary tract problems, such as overactive bladder, underactive bladder, incontinence, and interstitial cystitis, are highly prevalent among people aged 40+, making studying LUT pain and dysfunction a pressing issue. Important questions surrounding the distinct contributions of LUT cell types to bladder function remain unanswered, for example, what are the precise downstream pathways through which urothelial cells affect underlying muscle and activate sensory neurons? PIEZOs are generally not required for high-threshold mechanical pain sensation, but PIEZO2 is required for mechanical allodynia (pain from gentle touch) in the skin (Murthy et al., 2018b; Szczot et al., 2018). More recently, PIEZO2 was suggested to be involved in bladder pain, but thorough genetic studies are needed to validate these findings (Liu et al., 2023). PIEZO1 in non-neuronal cells was implicated in pathological signaling during cystitis (Liu et al., 2018; Beca et al., 2021), a mechanism that may involve urothelial ATP release inducing bladder hyperactivity (Cook & McCleskey, 2000). Characterizing the potential role of PIEZO ion channels in physiological bladder mechanosensation, and how this is altered in LUT pathologies, will expand our understanding of mechanosensation in the LUT and lay the groundwork for the discovery of therapeutic targets.

Pump it up and air it out: PIEZOs in cardiopulmonary function

If you were to close your eyes and revisit a stressful moment in your life, think: how did you feel inside? For many, the answer may include a racing heart and rapid breathing. Our ability to identify our heart rates when they go high and our breathing patterns when they go awry are forms of interoception which alert us to our bodies’ internal states. There are also many processes beneath our awareness that are required for the homeostatic regulation of these organs. The cardiovascular and respiratory systems are intimately linked via hemodynamic, mechanical, chemosensory, and neurohormonal pathways. This allows them to work together towards one main goal: delivering oxygen and nutrients to cells while eliminating carbon dioxide and metabolic waste (Cross et al., 2020). For example, during strenuous exercise, central and local control mechanisms increase blood flow and breathing rate to meet the body’s demand for oxygenated blood supply and metabolic waste elimination. Many mechanical processes involved are critically important for this coordination. One example is the baroreflex: stretch sensing within the cardiovascular system to allow the body to adjust heart rate and vascular tone. This reflex is important for keeping blood pressure under control and is modulated by breathing rate (Eckberg & Orshan, 1977). Thus, breathing and heart rate are always intertwined.

In addition to sensing chemical cues to monitor oxygen levels within the blood, the lungs need to sense mechanical cues, such as stretch, to initiate appropriate reflexes that protect airway integrity. These include inhalation/exhalation, apnea, swallowing, and cough (Prescott et al., 2020). For example, PIEZO2 in airway-innervating sensory neurons was identified as a key mechanosensor where it plays a role in the Hering-Breuer reflex. This inflation-induced response is initiated by vagal lung stretch sensors that induce apnea (halt breathing) to avoid lung overinflation (Nonomura et al., 2017). Interestingly, despite being expressed in approximately 50% of the P2ry1⁺ nodose neurons that mediate a coordinated airway protection response, characterized by apnea, vocal fold adduction, swallowing, and expiratory reflexes, P2ry1/PIEZO2-double positive neurons do not drive this response (Prescott et al., 2020). Thus, distinct lung mechanosensory neuron populations must drive apnea for different purposes.

Global Piezo2 knockout mouse pups experience respiratory distress and die shortly after birth, potentially due to failed lung inflation (Nonomura et al., 2017; Xiong et al., 2022). The biological phenomenon linking sensory neuron mechanosensation to initial lung inflation, however, remains a mystery. Interestingly, human patients with PIEZO2 loss-of-function variants are also reported to have shallow breathing during infancy (Chesler et al., 2016). Nonetheless, Piezo2^−/− mice do not show signs of embryonic developmental impairment within the pulmonary system. Adult mice with conditional Piezo2 knock-out maintain normal breathing frequencies, which suggests that other mechanosensory inputs besides Piezo2 are responsible for regulating breathing rate (Nonomura et al., 2017). These mysteries call for further studies to identify the key molecules and processes involved.

Throughout the cardiovascular system, baroreceptors detect stretch, which in turn is used to modulate heart rate and vascular tone to prevent extreme episodes of hypo- or hyper-tension. Multiple different mechanosensory receptors perform this function. Sensors in arterial walls detect carotid and aorta stretch to maintain steady arterial pressure (Kirchheim, 1976). An increase in venous and atrial pressure increases heart rate (tachycardia) to compensate by drawing more blood from the venous system (Bainbridge reflex) (Kuhtz-Buschbeck et al., 2017). Stretch sensing by vagal neurons within the heart’s left ventricular wall elicits the opposite inhibitory effect by slowing heart rate (bradycardia), and stimulating vasodilation and hypotension (Bezold-Jarisch reflex) (Aviado & Guevara Aviado, 2001; Kuhtz-Buschbeck et al., 2017; Lovelace et al., 2023). There are also intrinsic mechanisms of stretch sensing within the heart. These many sensory stretch reflexes reveal this system is under complex mechanical control.

In the nodose-petrosal-jugular (NPJ) ganglion complex in mice, where baroreceptor cell bodies reside, both Piezo1 and Piezo2 are expressed and required for the normal homeostatic baroreflex (Zeng et al., 2018). This was determined by using a classical pharmacological stimulus to constrict peripheral blood vessels (phenylephrine), which increases blood pressure and induces a homeostatic decrease in heart rate. This baroreflex, presumably dominated by the arterial stretch sensors, was absent in Piezo1/2 double knockout mice. Similar to humans suffering from baroreflex failure, Piezo1/2 double knockout mice also show labile hypertension and increased blood pressure variability (Zeng et al., 2018). A deeper dive into the anatomy of arterial baroreceptors showed that PIEZO2-positive NPJ ganglia neurons form macroscopic “claws” with mechanosensory end-nets around the aortic arch (Min et al., 2019). Interestingly, either PIEZO ion channel is sufficient for an intact baroreflex. It is not clear if this redundancy arises from their function in the same or different neuronal populations. Importantly, impaired baroreflex has serious real-world consequences, including arrhythmias, hypertension, and syncope, which can greatly impact patients’ quality of life and, in some cases, become life-threatening (Ketch et al., 2002). The discovery of PIEZO1 and PIEZO2 as the critical mediators of this fundamental reflex solved a long-standing mystery and opened more questions about the interplay of these sensors within the many cell types that modulate cardiac function. Recent work has identified neuronal subtypes that control mechanical reflexes in the heart (Lovelace et al., 2023), but the mechanotransduction mechanisms of the atrial and ventricular baroreceptors are not fully explored.

PIEZO-mediated gut feelings

From the moment food is ingested, it undergoes a series of transformations as it is transported through the gastrointestinal (GI) tract. Mechanosensation plays a crucial role in this process from the first step of mastication and feeding behaviors (Moayedi et al., 2018), all the way to ensuring proper digestion. GI contents exert mechanical forces on the walls of the gut, enabling mechanosensory neurons to detect valuable information regarding the location, volume, and physical properties of the ingested food. These mechanical signals play a pivotal role in the coordination of physiological processes such as mastication and swallowing, sphincter coordination, digestion, peristalsis and mixing within the intestine, and ultimately the process of defecation. By detecting and responding to these mechanical cues, the nervous system ensures the smooth progression of food through the GI tract, and the regulation of meal size and frequency.

Electrophysiological studies have delineated mechanoreceptor subsets based on their sensitivity to mechanical stimuli and activation threshold (Berthoud et al., 2004; Kentish et al., 2013). Three anatomically distinct vagal terminal endings in the GI are proposed to respond to mechanical stimuli. First, the intraganglionic laminar endings (IGLEs) form a web-like structure around the myenteric neurons and respond to distension and tension. Second, the intramuscular arrays (IMAs) are found parallel to smooth muscle fibers and are abundant near the sphincters. They are proposed to be stretch receptors, in a role analogous to muscle spindles (Phillips & Powley, 2000). Lastly, endings in the mucosal layer are proposed to detect the physical properties of ingested food and mediate responses to mucosal stroking (Powley et al., 2011; Kentish et al., 2013). Similar ending structures to IGLEs, which also transduce force but arise from DRGs, are called intraganglionic varicose endings (IGVEs) (Spencer et al., 2016; Servin-Vences et al., 2023; Wolfson et al., 2023). Some enteric neurons are also mechanosensitive (Kunze et al., 1998; Mao et al., 2006), which likely allows them to coordinate peristalsis and ensure mixing with the digestive enzymes. Given the many cell types involved, it is challenging to experimentally distinguish the precise stimuli to which each GI sensory neuron responds. For example, the stomach wall experiences both tension and stretch when distended, thus engaging most mechanosensory neurons. Moreover, artificial filling of the stomach does not account for accommodation reflexes that occur during feeding (Phillips & Powley, 2000). Therefore, it is important to develop genetic tools and techniques to parse the individual contributions of neuronal subtypes.

The motility required for digestion relies on mechanosensing at multiple levels in the GI tract. Mouse genetics and single-cell RNA sequencing techniques have allowed us to begin to elucidate some of these mechanosensory populations (Bai et al., 2019; Hockley et al., 2019; Zhao et al., 2022; Qi et al., 2023). For example, developmental studies revealed that Prox2⁺ and Runx3⁺ mark the majority of vagal neurons in the esophagus and play a role in esophageal motility (Lowenstein et al., 2023). These are PIEZO2⁺ neurons, although the explicit role of PIEZO2 was not defined (Lowenstein et al., 2023). Typically, homeostatic reflexes that mediate functions like motility are either mediated by the intrinsic nervous system or extrinsic vagal innervation, whereas spinal innervation from the DRG is thought to mediate pain sensation. This is also true in the colon, where PIEZO2 contributes to high-threshold colon distension and to visceral mechanical hypersensitivity (Wolfson et al., 2023; Xie et al., 2023). These important findings shed light on the mechanisms of GI pain, which are prominent in conditions like Crohn’s disease and irritable bowel syndrome. Surprisingly, it was also recently discovered that PIEZO2 is found in DRG IGVE afferents and regulates healthy gut motility. Genetic studies in mice indicate that PIEZO2 activity slows whole-gut transit time in the presence of GI contents, presumably to facilitate proper absorption (Servin-Vences et al., 2023; Wolfson et al., 2023). In contrast to whole-gut transit, in the colon, PIEZO2 seems important to speed up the expulsion of large fecal pellets. PIEZO2 was required for most DRG responses to mechanical stimuli delivered to the colon (Servin-Vences et al., 2023), although some high-threshold responses remain when investigating specific neuronal subtypes (Wolfson et al., 2023). This might implicate PIEZO2 dysfunction in constipation. Interestingly, these roles for PIEZO2 function in two separate circuits that control motility: the whole-gut transit circuit pathway recruits sympathetic output, but the colon expulsion pathway does not (Servin-Vences et al., 2023). In line with these functions, humans without PIEZO2 also exhibit impaired bowel sensation and motility (Chesler et al., 2016; Servin-Vences et al., 2023). The discovery of these novel reflexive roles for DRG innervation in the gut underscores how delineating mechanosensory molecules can also uncover previously unknown aspects of fundamental physiology.

Anecdotally, we know that a large meal can make us feel full and prevent us from eating more. Defining the identities of mechanoreceptors in the GI tract, and the potential contribution of PIEZO ion channels, could help us understand complex behaviors like feeding. Certain stretch receptors, like vagal Glp1r⁺ IGLEs in the stomach, are activated by gastric distension (Williams et al., 2016). Vagal Oxtr⁺ neurons form IGLEs in the intestine, and both of these populations contribute to fullness sensing (Bai et al., 2019). Within these defined mechanosensory neuron populations, the key transduction molecules are not entirely known. There are indications that PIEZO2 could be involved. In Drosophila melanogaster, piezo plays a pivotal role in GI mechanosensation, with knock outs showing extreme overconsumption, even leading to bursting of the crop (Wang et al., 2020; Min et al., 2021). It is possible that this role for PIEZO ion channels is at least partially evolutionarily conserved, which makes them promising candidate molecules for controlling food intake. Delineating transduction molecules and cells that modulate feeding will further our understanding of the many ways mechanosensation drives GI function, and potentially open up important avenues for therapeutic intervention for the prevalent problems of eating disorders and obesity.

Mechanosensory mysteries

It is clear that PIEZO2 is employed widely by the nervous system for a variety of internal mechanosensory functions. Although Piezo1 is expressed in both vagal and DRG neurons, it is found in fewer sensory neurons than Piezo2, and thus far has only been implicated as a key interoceptive mechanosensor for the baroreflex. In DRGs, PIEZO1 contributes to mechanical itch signaling in the skin, but its other neuronal roles are not yet defined (Hill et al., 2022). What do Piezo1-expressing neurons detect elsewhere in the body? It will be important to define the neuronal roles for PIEZO1 in the future.

PIEZO1 plays varied roles in mediating non-neuronal mechanosensory responses that might communicate to the nervous system. Non-neuronal cells certainly have important roles in shaping interoceptive processes. For example, Piezo1 is highly expressed across the urothelium, which lines the bladder lumen, (Dalghi et al., 2019) and might coordinate with PIEZO2 to mediate mechanosensation in these cells (Dalghi et al., 2021). Although urothelial cells are known to detect and respond to mechanical force by releasing signaling molecules, like ATP (Ferguson et al., 1997; Knight et al., 2002), it is not fully understood how the activity of these cells might affect mechanosensory neurons within the bladder, which will be an important line of inquiry. Piezo1 is also highly expressed in other endothelial cells (including pulmonary microvasculature), smooth muscle cells, endocardial cells, and alveolar epithelial cells (Xiong et al., 2022). How the responses from these specialized mechanosensory cell populations influence surrounding sensory neurons and physiology is an area that is still ripe for exploration.

PIEZO ion channels at the cell membrane confers mechanosensitivity resulting in cation influx, which has many downstream consequences. For this reason, it is likely that cells exert tight control over Piezo expression and function. For example, nociceptors that mediate internal pain sensing are typically not firing in normal or healthy states, which is why they are referred to as “silent nociceptors”. A proposed Chrna3+ population of silent nociceptors expresses Piezo2 transcript at all times, but only becomes mechanosensitive when exposed to inflammatory mediators (Prato et al., 2017). The basis for this regulation is not entirely understood, although proteins that could contribute are being identified (Nees et al., 2023). It is likely that many protein regulators of PIEZO activity will be important for controlling mechanosensation, including MyoD-like family inhibitor proteins, which appear to directly bind to PIEZOs (Zhou et al., 2023). Few such regulatory proteins are known, which makes this an exciting area of active research.

PIEZO ion channels are expressed in a variety of tissues where the activity of other sensory ion channels could be coordinated. For instance, PIEZO1 and TRPV4 are co-expressed and their activity is linked in multiple tissue types, like chondrocytes, osteoblasts, and urothelium, and their physiological effects could oppose each other in endothelial tissues (Servin-Vences et al., 2017; Ihara et al., 2018; Yoneda et al., 2019; Gao et al., 2022; Endesh et al., 2023; Nagai et al., 2023; Steinecker-Frohnwieser et al., 2023; Rong et al., 2024). At the cellular level, mechanical force transmission depends on membrane fluidity and cellular attachments, so the context of target tissues and their mechanical properties will change how the nervous system encodes mechanical force (Romero et al., 2019). Regulation in different tissues can also occur at the genetic level: splice variants of PIEZO ion channels with different kinetic properties could drive the sensitivity and response of different cell types (Szczot et al., 2017). Mechanosensation in some tissues can also change the mechanical properties of an organ by leading to hyperplasia, hypertrophy, or even dysplasias. Changes in tissue, cell membrane or ECM stiffness could lead to different levels of activation for innervating sensory neurons and could alter sensations or homeostatic functions in the body. Thus, understanding the regulation of PIEZO ion channel expression and activity in multiple cell types is an important field that will undoubtedly yield interesting insights into the processes of internal mechanosensation.

Aside from PIEZOs, other proteins have been proposed to be mechanosensory ion channels, but for many, their true functions are controversial or unknown. For example, TACAN (TMEM120A) was proposed to be a mechanically gated ion channel that plays a role in sensing mechanical pain (Beaulieu-Laroche et al., 2020). However, the recorded single-channel mechanically evoked currents are unusually small and these electrophysiological experiments could not be replicated (Gabrielle & Rohacs, 2023). Others have shown that TACAN inhibits PIEZO2, further casting doubt on its true role as a mechanosensor (Del Rosario et al., 2022). TMEM150c was also proposed to be mechanosensitive, but follow-up studies demonstrated that the currents recorded depend on endogenous PIEZO1 (Dubin et al., 2017), indicating these are unlikely to be novel mechanically gated ion channels. Although the mechanosensitivity of some epithelial sodium channels (ENaCs) is clear in invertebrates, the evidence for instrinsic mechanosensitivity is disputed for mammalian ENaC channels (Arnadottir & Chalfie, 2010). TMEM87a/ELKIN1 was found to mediate mechanosensitive currents in cancer cell lines, yet the relevance of this ion channel as a mechanosensor in normal cell physiology remains unclear (Patkunarajah et al., 2020). In 2018, OSCA/TMEM63 was identified as an evolutionarily conserved family of mechanically activated ion channels, but the in vivo roles of this family also remain elusive (Murthy et al., 2018a). Additional work is needed to uncover the remaining mechanosensors that could contribute to interoceptive functions.

The initial transduction mechanisms for force detection are just the very first steps in internal mechanosensation: after these signals are generated, many more complex processes must occur for the nervous system to integrate sensory features to mediate sensation and physiological control. Although all systems are informed by mechanical inputs and can stimulate reflex arcs, in some systems we can consciously control nearly all mechanical movements (skeletal muscle proprioception), while in others we only control parts of the motor system (the sphincters necessary for defecation and urination). Mechanical sensing can occur without our conscious awareness or control (peristalsis in the healthy gut), or only be detected during intense activity (heartbeat). What are the fundamental features that drive the different levels of conscious sensation, and how might this change in different states? Unlocking the key mechanosensory molecules of the interoceptive nervous system is a critical first step towards grasping these bigger questions about how the brain senses the body.

Biographies

Yasmeen M. F. Hamed earned her B.Sc. in Pharmacy and Biotechnology from the German University in Cairo and completed an M.Sc. at The University of Goettingen studying cardiac regeneration in the Mexican cavefish. She is now a Ph.D. Candidate in the Marshall lab where she uses PIEZO ion channels as molecular tools to understand bladder mechanosensation in physiology and disease.

Dr. Kara L. Marshall is a Freeman Hrabowski Scholar with the Howard Hughes Medical Institute and an Assistant Professor at Baylor College of Medicine and the Neurological Research Institute at Texas Children’s Hospital in Houston, TX. Dr. Marshall is supported as a McNair Scholar by the Robert and Janice McNair Foundation. After earning her Ph.D. from Columbia University, she did postdoctoral work at Scripps Research studying mechanosensation in the urinary tract and other systems. Her lab continues this line of research to understand peripheral and central mechanisms of internal mechanosensation.