Pflügers Archiv - European Journal of Physiology

, Volume 455, Issue 5, pp 775–785

Mechano-sensitivity of ENaC: may the (shear) force be with you

Authors

    • Institute of Animal PhysiologyJustus-Liebig-University Giessen
  • Wolfgang G. Clauss
    • Institute of Animal PhysiologyJustus-Liebig-University Giessen
Invited Review

DOI: 10.1007/s00424-007-0332-1

Cite this article as:
Fronius, M. & Clauss, W.G. Pflugers Arch - Eur J Physiol (2008) 455: 775. doi:10.1007/s00424-007-0332-1

Abstract

The epithelial Na+ channel (ENaC) is the rate-limiting step for Na+ absorption in various vertebrate epithelia and deeply enmeshed in the control of salt and water homeostasis. The phylogenetic relationship of ENaC molecules to mechano-sensitive Degenerins from Caenorhabditis elegans indicates that ENaC might be mechano-sensitive as well. Primarily, it was suggested that ENaC might be activated by membrane stretch. However, this issue still remains to be clarified because controversial results were published. Recent publications indicate that shear stress represents an adequate stimulus, activating ENaC via increasing the single-channel open probability. Basing on the experimental evidence published within the past years and integrating this knowledge into a model related to the mechano-sensitive receptor complex known from C. elegans, we introduce a putative mechanism concerning the mechano-sensitivity of ENaC. We suggest that mechano-sensitive ENaC activation represents a nonhormonal regulatory mechanism. This feature could be of considerable physiological significance because many Na+-absorbing epithelia are exposed to shear forces. Furthermore, it may explain the wide distribution of ENaC proteins in nonepithelial tissues. Nevertheless, it remains a challenge for future studies to explore the mechanism how ENaC is controlled by mechanical forces and shear stress in particular.

Keywords

Epithelial Na+ channelENaCMechano-sensitivityMechano-sensitive channelDegenerinsMembrane stretchShear stressDegenerin/ENaC family

Introduction

Living organisms are exposed to a variety of mechanical forces and have adapted organs, cells, and molecules to convert these forces into cellular signals bearing the possibility to respond to those stimuli. Although the transduction of mechanical forces seems to be a ubiquitous feature within evolution [48], the particular mechanisms are poorly understood. In this review, we focus on the mechano-sensitive properties of the epithelial Na+ channel (ENaC). ENaC is originally known from Na+-absorbing vertebrate epithelia as, for example, the epithelia of colon, kidney, bladder, lung, and skin. There is growing evidence that ENaC responds to mechanical forces, which is further supported by the identification of ENaC proteins in nonepithelial tissues contributing to mechano-sensing.

The main known function of ENaC is linked to the control and maintenance of salt and water homeostasis by representing the rate-limiting step for the uptake of Na+ [8, 30, 61]. Malfunction of ENaC activity is associated with severe diseases, e.g., it was impressively demonstrated that α-ENaC knockout mice die shortly after birth because they are not able to reabsorb the fluid from the lungs [35]. In this context, it is also known that a decreased ENaC activity is associated with the formation of pulmonary edema [49]. Other examples underlining the physiological function of ENaC are a hereditary form of hypertension known as Liddle’s syndrome, which is associated with hyperactivity of ENaC due to defective degradation processes [61, 68], as well as mutations leading to a loss-of-function, known as pseudohypoaldosteronism [61].

Primary experimental hints concerning the function of ENaC base on observations published by du Bois-Reymond in 1848 [24], demonstrating that isolated frog skins maintain a potential difference between the internal and external side, with the internal side becoming more positive with respect to the external side. This observation already implicated the suggestion that this might represent a net transport of positively charged ions from the external to the internal side of the skin—corresponding to the absorption of Na+ ions. This phenomenon was further investigated by Ussing and coworkers and resulted in a model for Na+-absorbing epithelia known as the “two-membrane hypothesis” (Fig. 1) [43, 75]. Ussing and colleagues recognized that the apical side of the amphibian skin behaves like a Na+ (permeable) electrode, whereas the basolateral side behaves like a K+-selective electrode. In this model, Na+ transport is accomplished via the following steps: (1) Na+ enters the epithelial cell passively, by passing the outward-facing membrane following an electrochemical gradient. This gradient is provided and maintained by an active, oxygen-consumptive process due to an “ion pump” later on identified as the Na/K-ATPase. The energy-consumptive process allows the (2) extrusion of Na+ against its electrochemical gradient across the inward-facing membrane. During this step, Na+ is transported in exchange with K+, which is then (3) recycled via K+-conductive pathways (Fig. 1). The importance of this model should be considered, as it is still the basis of our understanding how active Na+ reabsorption is accomplished by Na+-absorbing epithelia.
https://static-content.springer.com/image/art%3A10.1007%2Fs00424-007-0332-1/MediaObjects/424_2007_332_Fig1_HTML.gif
Fig. 1

Basic function of Na+-absorbing epithelia as developed by Ussing and colleagues. This principle of Na+ absorption is known as the two-membrane hypothesis developed from studies using isolated frog skin. Up to now, this is still the basic mechanism for our understanding of Na+ absorption across epithelia. Following this model, Na+ passes the outward-facing membrane (om) downward its electrochemical gradient and is then extruded in exchange with K+ by the Na/K-ATPase (P) due to an energy-consumptive process across the inward-facing membrane (im). Furthermore, the basolateral membrane is characterized by a high potassium conductance to “recycle” K+ ions and to maintain intracellular K+ concentrations at constant levels. Subsequent studies identified that ENaC is the major candidate for the Na+-conductive pathway across the apical membrane of Na+-absorbing epithelia (tj tight junctions; the scheme was modified from [75])

Although, in the following years, Na+-selective and amiloride-sensitive permeable pathways across the apical membrane of different epithelia were well described—including identification of the ENaC by single-channel recordings from rat cortical colleting tubule [54]—the molecular identity of the channel remained unknown for several years. Finally, in 1993, two different groups succeeded to isolate DNA from epithelial cells, which was suggested to encode for ENaC [11, 45]. This was proven by the functional expression of the proteins in Xenopus oocytes, which reconstituted amiloride-sensitive currents [11, 45]. Furthermore, it was found that the identified proteins belong to a new gene family (Degenerin/ENaC family) whose members were identified to be involved in the mechano-transduction pathway in Caenorhabditis elegans [11, 45]. One year later, the basic structure of ENaC was revealed, demonstrating that the functional channel comprises three homologue subunits (α, β, and γ [12]). However, the molecular architecture of ENaC as well as the stoichiometry of the subunits forming the ion channel remain to be clarified. Different models were proposed including a channel consisting of nine subunits [67], eight subunits [27], as well as a heterotetrameric structure of two α, one β and γ subunits [5, 28, 61].

Other important findings within the last decade were the identification of new ENaC proteins such as the pore-forming delta (δ) subunit [76] as well as the expression of ENaC proteins in nonepithelial tissues, e.g., testis [17, 36], brain [4, 25, 32, 76], nerve endings [20, 23], vascular cells [21], and skeletal cells [42, 73]. Interestingly, it was suggested that the δ subunit might be limited to nonepithelial tissues, but Ji et al. [40] recently detected δ-subunit expression in lung epithelial cells, in pancreatic cells, and in colon epithelial cells.

Background concerning the mechano-sensitivity of ENaC

The concept of ENaC mechano-sensitivity is deduced from the phylogenic relationship of the ENaC proteins to the mechano-sensitive Degenerins identified in C. elegans. These proteins were discovered by screenings for mutations in animals with abnormal behavior in response to touch compared with wild-type animals [15, 19, 72]. It was recognized that mec genes encode mechano-sensitive ion channels and that mutations of these genes are causative for touch insensitivity and/or degeneration of mechano-sensory neurons in C. elegans.

But other experimental findings may further support the suggestion that ENaC is a mechano-sensitive ion channel—the finding that ENaC subunits are expressed in nonepithelial tissues, which are obviously exposed to mechanical stimuli such as the nerve endings of rat foot pad [20], the nerve endings innervating the aortic arch and carotid sinus [22, 23], and vascular cells [21]. Thus, it is suggested that ENaC might be directly involved in mechano-transduction processes such as touch sensation as well as in the control of blood pressure [21, 23, 38].

Experimental evidence that ENaC is a “mechano-gated” channel

In the following section, we focus on the experimental evidence concerning the mechano-sensitivity of ENaC (summarized in Table 1). The section is subdivided in two parts with respect to the used mechanical stimuli: (1) membrane stretch, which was applied via increased hydrostatic pressure, negative pressure via the patch pipette (suction), osmotic cell swelling, or intracellular injection into Xenopus oocytes expressing ENaC proteins, and (2) shear forces as administrated via a fluid stream passing the outer surface of the investigated cells, and thus the expressed ENaC proteins.
Table 1

ENaCs from different species and tissue origins were reported to be affected by various mechanical stimuli

Species

ENaC subunit(s)

Model

Mechanical stimulus

Effect

Reference(s)

Bovine

α

Lipid bilayers

Hydrostatic pressure

+

[6]

α

Xenopus oocytes

Stretch (injection)

+

[6]

Rat

α

Lipid bilayers

Hydrostatic pressure

+

[37]

α, β, γ

Lipid bilayers

Hydrostatic pressure

+

[37]

Rat

Native ENaC

Isolated CCD cells

Stretch (negative pressure)

?

[55]

Rat

α, β, γ

Xenopus oocytes

Osmotic swelling/shrinkage

−/+

[39]

Rat

α, β, γ

Xenopus oocytes

Osmotic swelling

No effect

[7]

Stretch (injection)

No effect

Human

Not known

B lymphocytes

Hydrostatic pressure

+

[1]

Human

α

Transfected fibroblasts

Stretch (negative pressure)

+

[42]

Mouse

α, β, γ

Xenopus oocytes

Shear stress

+

[13, 62]

Rabbit

Native ENaC

Isolated CCD tubules

Shear stress

+

[51, 62]

Xenopus

Native ENaC

A6 cells

Stretch (negative pressure)

+

[46]

Rabbit

Native ENaC

Bladder epithelium

Hydrostatic pressure

+

[78]

Xenopus

α, β, γ

Xenopus oocytes

Shear stress

+

[3]

Rat

α, β, γ

Xenopus oocytes

Shear stress

+

[3]

+ Stimulation, − inhibition, ? unclear, CCD cortical colleting duct, A6 cells from Xenopus distal kidney tubules

Membrane stretch

As mentioned above, the suggestion that ENaC could be a mechano-sensitive ion channel originated soon after ENaC was recognized as a relative of the C. elegans MEC and DEG proteins. The group surrounding Benos provided in 1995 the first evidence that cloned ENaC proteins might respond to mechanical forces. Bovine αENaC subunits were reconstituted in planar lipid bilayers and exposed to hydrostatic pressure. This procedure was found to increase ENaC activity, and these results were additionally confirmed in Xenopus oocyte expression experiments, where the bovine clone was further activated by cell swelling [6]. These results were confirmed by similar experiments using the rat ENaC clone in lipid bilayers as well as oocyte experiments [37]. Contradictory results were published from a study expressing αβγENaC in Xenopus oocytes demonstrating that cell swelling decreases ENaC activity, whereas cell shrinkage increases ENaC activity [39]. In another study using human B lymphocytes, an amiloride-sensitive Na+ current was reported, which was activated by hydrostatic pressure [1]. Kizer et al. [42] reconstituted mechano-sensitive currents by transfection of the α-ENaC subunit into fibroblasts. However, it appears to be difficult to interpret this data, as the transfected fibroblasts were found to express endogenous mechano-sensitive ion channels as well [77].

Other investigations so far have shown contradictory findings concerning the activation by membrane stretch. For example, Awayda and Subramanyam [7] found no evidence for mechanical activation by membrane stretch. This publication was contradictory to prior published data [6, 37], indicating that the mechano-sensitivity of bilayer-reconstituted ENaC could be due to the contamination of non-ENaC proteins during the carryover process. This issue was further critically discussed by Rossier [58], who addresses the suspected biophysical properties of the channels in bilayers in contrast to native ENaC biophysical properties—especially comparing the single-channel conductance and the selectivity (Li+ > Na+ >> K+) [30]. Doubts concerning a mechano-sensitive ENaC activation via membrane stretch were also published in a study from Palmer and Frindt [55]. They failed to provide evidence concerning this issue in native cortical collecting duct cells and pointed out the limitation of the cell-attached configuration with respect to the activation of ENaC by membrane stretch. They argued that local membrane deformations might be responsible for the observed inconsistent effect(s) and concluded that the used protocol might not be appropriate to stimulate ENaC. Another suggestion to explain the controversy was introduced by Ma et al. [46]. They demonstrated that ENaCs in A6 kidney cells were mechanically activated by membrane distention, but this effect was compensated by the release of ATP, which subsequently inhibits ENaC via the activation of purinergic receptors. But this principle remains to be confirmed in other cell systems, e.g., the oocyte expression system.

An important fact to explain the discrepancy between the results obtained from reconstituted ENaC in bilayers vs native cell membranes may be further provided by the indication that reconstituted membranes exhibit other physical attributes regarding the expansion/tension properties of the membranes due to mechanical deformation compared with native cell membranes [33]. Furthermore, it has to be considered that it is difficult to use negative/positive hydrostatic pressures in combination with patch clamp experiments, as the properties do not represent the physiological situation regarding endogenous membrane tensions. Thus, excessive pressures applied via the patch pipette may lead to an “overstimulation” of the investigated channels [34].

It is difficult to asses whether or not ENaC might be activated by membrane stretch as applied via hydrostatic pressure, osmotic cell swelling, and/or cell swelling via injection. Because we were not able to prove a mechanical activation either of rat or of the Xenopus ENaC clone in Xenopus oocytes exposed to hyposmotic challenge, as well as intracellular injection (unpublished data), we propose that membrane stretch does not represent an adequate stimulus to activate native ENaC localized in native cell membranes. However, this issue needs further investigations—especially to reveal the putative role of ATP and purinergic receptors concerning the stretch-induced ENaC activation.

Shear stress (laminar flow)

The physiological relevance of shear stress as an adequate stimulus for ion channel activation was originally identified on isolated endothelial cells by the whole cell patch clamp technique [53]. It is suggested that, in the vasculature, the blood flow represents a fast direct stimulus to regulate the vascular tone by mechanical-induced activation of K+ channels leading to an increase in intracellular Ca2+ concentrations [2, 53]. A new attempt concerning the mechano-sensitivity of ENaC was undertaken by Kleyman and colleagues also using shear stress as mechanical stimulus [63]. In this context, it should be considered that ENaC-expressing epithelia are also exposed to shear forces under physiological conditions, and the most obvious example concerning this issue is probably represented by the flow of urinary fluid in the kidney tubules [31]. Experimentally, shear stress was generated via a fluid stream directed to the surface of the ENaC-expressing membranes. The principle was used for two different approaches: (1) ENaC-expressing oocytes were exposed to a fluid stream, which was generated via a tubule placed in close proximity to the oocytes; (2) isolated cortical collecting duct tubules from rabbits were perfused by different flow rates. In both systems, an increased Na+ absorption, and thus ENaC activity, was observed in response to increased flow rates [63]. These observations demonstrated that ENaC was activated in a well-known in vitro model employing the two-electrode voltage-clamp technique on ENaC-expressing oocytes, as well as in a more physiological system as provided by freshly isolated cortical collecting duct tubules. In both systems, the investigated ENaCs are expected to represent the physiological native form of the ENaC.

Concerning the activation of Na+ absorption, two mechanisms are basically indicated under physiological conditions. This might be either accomplished via an increase in the activity of ENaC by an increased single-channel open probability (PO) or via an increased number of active ion channels (N) [59]. New insights concerning the shear-activated mechanism were provided by Carattino et al. [13, 14]. These studies indicated that the shear-force-induced effect is due to an increased PO. For the studies, modified ENaCs were used, which were either sensitive to chemical modifications by sulfhydryl-reactive agents or constitutive active. The main observations were that the modified channels (defined by a high PO) were not activated by laminar shear stress in comparison to wild-type ENaC [13] and that the region spanning the amino acids between 580 and 589 within the α subunit is involved [14]. The indication that shear stress activates ENaC via altered gating properties and thus an increased open probability was further supported by a recent published study showing that the shear-induced activation was not affected either by maneuvers affecting membrane trafficking or by changes in Ca2+ levels [51].

However, further evidence concerning the underlying mechanism was recently published by our group [3]. We were able to analyze the effect of laminar shear stress on ENaC in patch clamp recordings employing the outside-out configuration. Therefore, Xenopus [57] and rat ENaCs [12] were heterologously expressed in Xenopus oocytes. In the outside-out configuration, we were able to expose the excised patches to laminar shear stress and to monitor ion channel activity in parallel (Fig. 2a). By this procedure, we found that the activation of a laminar fluid stream within the perfusion tubule resulted directly in an activation of the expressed ENaC orthologs. The obtained activation was primarily due to an increased open probability of the channels (Fig. 2b,c). The number of channels, which were activated by shear stress, was not significantly altered [3]. This finding was further confirmed in a series of two-electrode voltage clamp recordings employing strategies that are known to activate ENaC via an increased open probability. In these experiments, ENaC molecules were first pharmacologically activated by the gadolinium cation [29], zinc cation [65], or glibenclamide [16, 64] and were then exposed to laminar shear forces. In accordance to the patch clamp data and to the prior published indications by Carattino et al. [13, 14], the shear-force-induced effect was decreased [3]. Employing another strategy provided by proteolytic cleavage, which is known to increase the numbers of active ENaC molecules [60], we found that the response to shear force was stronger compared to control conditions and to the experiments with the maneuvers which increase the open probability of the channels [3].
https://static-content.springer.com/image/art%3A10.1007%2Fs00424-007-0332-1/MediaObjects/424_2007_332_Fig2_HTML.gif
Fig. 2

ENaC orthologs originated from rat colon, and Xenopus kidney cell line (A6) were activated by laminar shear stress (LSS). a Excised patches in the outside-out configuration were moved in front of a tubule, which was perfused to generate shear forces. b Rat ENaC activity was first recorded for approximately 1 min under control conditions. Then the fluid stream was activated (LSS, 1 min), and finally, amiloride (10 μM, ami.) was added to the perfusate to determine baseline levels. These experiments showed that the open probability of ENaC was increased. c Although, in some experiments with the Xenopus ortholog, the number of activated channels was also increased (n = 2 of 7 recordings), these changes were not statistically significant. For further details, please consider reference [3] from where the figure was modified

From these studies, we conclude that shear forces provide a suitable stimulus for the activation of ENaC, as shear-activated inward currents were solely observed in ENaC-expressing oocytes and were completely sensitive to amiloride and benzamil, respectively [3, 13, 63]. This suggestion is further underlined by the fact that the observations were obtained from different experimental models including the oocyte expression system using various ENaC orthologs, as well as native isolated cortical collecting duct tubules. Additionally, these results provided by different groups and by various techniques including single-channel recordings are in accordance to each other.

Putative mechanism of ENaC activation

The mechanism of how ENaC might sense shear forces is yet unknown. Our data give evidence that the extracellular loops play a major role in this mechanical activation. This is underlined by the finding that proteolytic cleavage of the loops produced contradictory results comparing the two used orthologs [3]. The rat ortholog was stimulated by shear stress after proteolytic cleavage, whereas the Xenopus ortholog was inhibited. This contrary observation could be reasoned by different cleavage patterns in the extracellular loops. These different cleavage patterns affect channel gating directly (activation vs inhibition) indicating a connection between the loops and the gating machinery. A promising region participating in shear stress activation was identified within the pre-M2 region of the α subunit between amino acid 580 and 589 [14]. Amino acid substitutions in this region were characterized by altered responses to shear force with respect to delay times (time between application of shear and ion channel activation), time constants of activation (τ), and the magnitude of activation. Although, in our study, we have not monitored these parameters, this would be of interest for future studies because ENaC-activity-modifying drugs (e.g., Gd3+, Zn2+, glibenclamide, and trypsin) may change these parameters in a similar manner like amino acid substitutions.

Nevertheless, we suggest that the large extracellular loops behave like antenna extending into the luminal compartment (Fig. 3a). Taking into consideration that these structures could detect the movement of particles as well as fluid and are coupled with the gating machinery of ENaC (e.g., the pre-M2 domain of the α subunit as identified by Carattino et al. [14]), it is likely that deflection of this “sensory complex” induces conformational changes within the pre-M2 domain, and thus influences the rate of Na+ ions passing the pore (Fig. 3b,c). Two possibilities are evident with this respect: (1) the ENaC loops are solely deflected by shear stress without being tethered to extracellular cell matrix components (Fig. 3b); (2) ENaC loops are tethered to extracellular matrix proteins (Fig. 3c). But this hypothesis concerning the putative role of the extracellular loops has to be confirmed by future studies.
https://static-content.springer.com/image/art%3A10.1007%2Fs00424-007-0332-1/MediaObjects/424_2007_332_Fig3_HTML.gif
Fig. 3

Putative model concerning the mechano-sensitivity of ENaC, considering an assembly of two α, one β and γ subunits. a It is feasible that the large extracellular loops play a major role in the detection as well as the transduction of shear forces. This may occur by coupling of the mechano-sensitive structure with the gating machinery of the channel localized in close proximity to the ion channel pore. Basing on this idea, two possibilities exist: b either the loops are free-standing and are deflected by the movement of surrounding particles (e.g., fluid) producing shear forces (indicated by arrow), c or the loops are associated with components of the extracellular matrix as it is known for the C. elegans touch receptor (Fig. 4). Up to now, nothing is known about such an interaction of ENaC proteins with extracellular components and extracellular matrix proteins, respectively

Although some functional domains within the ENaC channel were identified, little is known about the distinct role of the extracellular loops. Each of the identified subunits has two cystein-rich domains (CRDs) located in the extracellular loops, which are highly conserved within the members of the DEG/ENaC protein family [9, 41]. The CRDs found in the Degenerin subunits are suggested to be involved in modulating ion channel gating [41, 71]. On the other hand, it has been shown that the gating process of ENaC plays an important role concerning the shear-stress-induced activation [14]. Taking these two findings together, this may support our idea about a link between the extracellular loops and the gating machinery localized in the pre-M2 region.

Considering the phylogenetic relationship of ENaC to Degenerins, it might be appropriate to have a closer look on the touch-sensitive complex of C. elegans, as this structure is much better characterized. The touch-sensitive complex in C. elegans neurons comprises different proteins (Fig. 4a): pore-forming subunits (MEC-4, MEC-10), proteins which are important for proper channel localization and formation (MEC-6), subunits which are suggested to organize a proper lipid environment (MEC-2, UNC-24), as well as components forming an extracellular matrix surrounding the complex (MEC-1, MEC-5, MEC-9), and tubulins (MEC-7, MEC-12) [10, 44]. Thus, the mechano-receptor complex of C. elegans shares various proteins with distinct functions. This complex is activated by membrane stretch, inducing conformational changes in the pore-forming subunits (Fig. 4b) [10].
https://static-content.springer.com/image/art%3A10.1007%2Fs00424-007-0332-1/MediaObjects/424_2007_332_Fig4_HTML.gif
Fig. 4

A model of the mechano-sensitive touch receptor complex from C. elegans comprising the ENaC homologue Degenerin proteins (e.g., MEC-4, MEC-10). a This touch receptor complex consists of different proteins, including those forming the channel, tubulins, as well as extracellular matrix proteins forming the mantle. In this model, the pore-forming subunits are associated with the mantle, which is crucial for mechano-sensitivity. b The touch receptor is suggested to be activated by membrane distention (stretch) caused by displacement (e.g., during touch) of the channel relative to the membrane. The figure was modified from [10]

Currently, it is unknown whether or not mechano-sensitivity of ENaC is due to a similar mechanism. The main reason for this is that there is now evidence concerning the interaction of the extracellular loops with extracellular components such as matrix proteins (e.g., MEC-1, MEC-5, MEC-9 as known from C. elegans). In C. elegans, the interaction of the pore-forming subunits with the extracellular matrix, called mantle, is crucial for the maintenance of mechano-sensitivity [26]. On the other hand, there is evidence that ENaC interacts with intracellular parts of the cytoskeleton [50, 66]. However, integrating this finding into the C. elegans model does not explain mechano-sensitivity, as it is recently suggested that the intracellular components of the pore-forming receptor complex are not directly tethered to the corresponding tubulins (MEC-7, MEC-12; [10]). This is indicated by observations that animals with null mutations in the mec-7 and mec-12 genes produce touch-insensitive animals, but the expressed receptors still produce mechano-sensitive currents [52]. Intracellular tethering of ENaC might be adequate to detect membrane stretch (with yet unsatisfactory published results, see above) but not the shear forces appearing at the extracellular cell surface. This is more obviously a role for the extracellular loops and/or the extracellular matrix.

The lack of experimental data concerning the assembly and stoichiometry of ENaC constitutes a major limitation to develop a model to explain the mechano-sensitivity of ENaC. In line with this point, there is little understanding concerning the interaction of the ENaC subunits with its environment, e.g., the extracellular matrix surrounding the channel. Especially the composition of the extracellular matrix needs to be further elucidated because it is known that this component plays a major role as a “shear-force transducer” in endothelial cells [18, 79] as well as in mechano-sensitivity of the touch receptor complex in C. elegans.

Putative physiological significance of ENaC mechano-sensitivity

Concerning the physiological significance of a mechanical ENaC activation by shear force, it has to be considered that ENaC function is mainly characterized in Na+-absorbing epithelia (e.g., kidney, colon, and lung). Although ENaC proteins have been detected in various nonepithelial tissues (see above), functional data concerning the role of ENaC in these tissues/cells are scarce. Therefore, we would like to focus on the physiological role of mechanical ENaC activation in Na+-absorbing epithelia.

The mechanical ENaC activation by shear forces in polarized epithelia might represent a novel, nonhormonal regulatory mechanism. It can be compared with established nonhormonal mechanisms known as “self-inhibition” and “feedback-inhibition.” These two mechanisms are defined by the ability that ENaC is directly controlled by extracellular (self-inhibition) and/or intracellular (feedback-inhibition) Na+ ions. These phenomena are suggested to limit the Na+ absorption during periods of high salt delivery [56] to avoid excessive changes of the intracellular Na+ concentration as well as cell volume of the epithelial cells [74]. This means that, under physiological conditions, ENaC molecules can directly sense Na+ concentrations and thereby control their activity in dependence to local Na+ concentrations.

With respect to the idea that ENaC is regulated by such a nonhormonal mechanism depending on local microenvironments (e.g., Na+ concentrations), it seems reasonable that shear forces might represent a stimulus for a similar nonhormonal mechanism, by coupling the rate of Na+ absorption to luminal transport of extracellular substrates (e.g., fluid or feces). The appearance of shear forces is evident for different Na+-absorbing epithelia. For example, the flow of the urinary fluid in the kidney tubules is known to represent a suitable stimulus providing shear forces at the luminal surface of the kidney epithelial cells [31, 62, 63]. In this way, the rate of Na+ absorption could be directly coupled to the flow rate passing the tubules. This means that a high flow rate through the tubules will provide high shear forces, and this may result in an increased Na+ absorption [63]. This principle is also evident in the oocyte expression system demonstrating that shear-induced ENaC activation correlates with the magnitude of the applied shear forces [13].

Another example for the physiological relevance of ENaC mechano-sensitivity is represented by the lung epithelium. It was recently indicated that the lung epithelium—beside distention—is also exposed to shear forces mainly due to inspiratory and expiratory airflows [69, 70]. Furthermore, the phasic movement of the lung causes shear forces at the surface of the epithelial cells because the fluid covering the pulmonary tract behaves relatively static. It is also obvious that the viscosity of this fluid layer will affect the magnitude of shear forces, where a high viscosity of the airway surface layer (thickened fluid layer) might provide higher shear forces compared to a layer exhibiting low viscosity. From this point of view, an increased viscosity of the liquid layer may lead to higher shear forces at the surface of the epithelial cells and could therefore further increase ENaC activity. This speculation might be of physiological relevance in patients with cystic fibrosis where the viscosity of the airway surface layer is elevated and an increased Na+ absorption across ENaC is observed [47].

Additionally, it seems reasonable that a similar mechanism can appear in the colon. In this example, the luminal passage of particles within the intestine will directly provide a local stimulus to activate ENaC in the apical membrane of the epithelial cells. This feature might ensure increased ENaC activation during time periods with high luminal transport rates.

In nonepithelial tissues, the mechano-sensitivity of ENaC proteins is more evident. This is, for example, clearly underlined by the presence of ENaC subunits in the vasculature. Endothelial cells are permanently exposed to shear forces due to the blood flow, and ENaC proteins are suggested to be involved in the control of blood pressure [21, 23, 38]. Discovery of ENaC subunits in touch-sensitive structures, e.g., rat foot pad [20], also indicates a significant role for ENaC proteins in touch-sensation and thus mechano-transduction. The localization in these touch-sensitive neurons further indicates a mechanism more related to the function of the C. elegans Degenerins.

Summary and concluding remarks

Recently, investigations established that ENaCs are mechano-sensitive ion channels. More specifically, it seems that shear forces represent an adequate stimulus. In contrast, it is not doubtlessly clarified whether or not native ENaCs are activated by membrane stretch. The mechanism how ENaC is activated by shear stress remains unknown, and this is mainly due to the lack of experimental data concerning the stoichiometry and the assembly of the ENaC proteins forming the channel, as well as its microenvironment with special focus on the extracellular matrix surrounding the channel. Basing on experimental data, it seems likely that the extracellular loops play a major role concerning the detection and transduction of shear forces. We hypothesize that the mechanical activation via shear stress is of considerable physiological relevance because many ENaC-expressing tissues are exposed to shear forces. This ENaC feature could therefore represent a novel nonhormonal regulatory mechanism. Last but not least, ENaC mechano-sensitivity will underline the evolutionary conservation of this property as a basic functional feature within the Degenerin/ENaC superfamily.

Acknowledgments

The authors want to thank M. Althaus, R. Bogdan, and J. Strauss for helpful comments and discussions to improve the manuscript. This work is supported by DFG (Deutsche Forschungsgemeinschaft).

Copyright information

© Springer-Verlag 2007