Abstract
Perturbations of cellular and systemic osmolarity severely challenge the function of all organisms and are consequently regulated very tightly. Here we outline current evidence on how cells sense volume perturbations, with particular focus on mechanisms relevant to the kidneys and to extracellular osmolarity and whole body volume homeostasis. There are a variety of molecular signals that respond to perturbations in cell volume and osmosensors or volume sensors responding to these signals. The early signals of volume perturbation include integrins, the cytoskeleton, receptor tyrosine kinases, and transient receptor potential channels. We also present current evidence on the localization and function of central and peripheral systemic osmosensors and conclude with a brief look at the still limited evidence on pathophysiological conditions associated with deranged sensing of cell volume.
Osmotic water flux across the plasma membrane, resulting in altered cellular volume and ionic strength, severely affects cell function. Most vertebrates counteract such perturbations by maintaining a remarkably stable osmolarity in the extracellular fluid (ECF; in mammals, close to 300 mOsm) and by possessing a variety of generally indirect mechanisms of volume regulated ion transport that allow individual cells to monitor and recover their volume following osmotic swelling or shrinkage. The purpose of this review is to outline the current evidence on how cells sense volume perturbations, with particular focus on mechanisms relevant to the kidneys and to ECF osmolarity/whole body volume homeostasis. The signaling events downstream from the osmosensor and the volume regulatory ion transport proteins involved in the process of regulatory volume decrease (RVD) and regulatory volume increase (RVI) after osmotic swelling and shrinkage, respectively, have been reviewed in detail elsewhere.1–4 We also discuss a variety of pathophysiological consequences to disturbances in cell-volume sensing.
Two terms are used to describe the osmotic relation between the cell and its surroundings: osmolarity and tonicity. Osmolarity is an absolute term that can be measured typically by freezing-point depression, whereas tonicity is defined by the osmotic gradient across a membrane; by definition, hypertonic exposure causes cell shrinkage, and hypotonic exposure causes cell swelling. In contrast, exposure of a cell to a fully permeable osmolyte has an osmotic effect, but no tonicity effect, and does not alter cell volume. In this review, we will use the terms osmosensor and osmosensing unless there is direct experimental evidence that the modality sensed is tonicity that perturbs cell volume. Tonicity will be used for describing the experimental perfusion of a cell with an anisotonic solution or for processes in which a direct dependence on cell volume, rather than osmolarity, is established.
BASAL MECHANISMS IN CELLULAR VOLUME SENSING
Arguably the most fundamental issue in osmosensing is precisely what is sensed when a cell is exposed to osmotic stress. Surprisingly, this issue is still not resolved completely. At least three types of osmotic stress signals can be identified in eukaryotic cells: changes in macromolecular crowding, ionic strength, and mechanical or chemical changes in the lipid bilayer or the extracellular matrix (ECM) and the cytoskeleton to which it is tethered.1,5–7 Osmosensory systems respond to these signals, resulting in a series of signal transduction events that, in turn, activate volume regulatory, protective, and adaptive events.1 All three types of osmotic stress signals can be part of the osmosensory machinery in various model systems, and, in some cases, several signals may contribute to a given response, as examples that follow will show.
A poorly understood issue in volume sensing is where, subcellularly, the sensors are located and how that impacts on the signals that can be perceived. As discussed in detail elsewhere,1 altered curvature and/or composition of the plasma membrane can initiate a number of signaling events central to the cellular sensing of osmotic stress. However, while altered membrane stretch or curvature is seemingly an attractive mechanism for volume sensing, many cells have such a large membrane reserve that this mechanism is not likely to be globally relevant.1,8 On the other hand, proteins located in specific subcellular domains, such as caveolae or lipid rafts, may well experience altered stretch or curvature relevant to osmosensing.9–12 Moreover, several volume-sensitive transporters and channels are highly sensitive to membrane curvature and lipid composition.7,13–16
Another subcellular domain of interest in the context of osmosensing is the primary cilium, a microtubule-based structure emanating from the mother centriole of most cells in the body controlling essential cellular functions.17,18 The primary cilium contains several complexes of transient receptor potential (TRP) channels, with proposed roles in volume sensing and signaling (see Transient Receptor Potential Channels (TRPs) and Mechanosensitive Channels). The primary cilium has also been assigned an essential role in renal flow sensing,19,20 although this has been partially challenged recently.21,22
Sensors, Transducers, and Effectors?
In fungi, two-component histidine kinase systems initiate the osmosensory response.23 In higher eukaryotes, this system is lacking and a number of others are employed. These are often parsed in terms of the primary volume sensor, the signaling events transmitting the signal (the transducers), and the mechanisms actually mediating the volume regulatory, protective, and adaptive events in response to osmotic stress (the effectors). In many cases, however, it is hard to separate the sensor from the transducer, an example being the cytoskeleton (see below), and some effectors likely directly sense the osmotic or mechanical perturbation. Examples include stretch-activated channels such as two-pore K+ channels,15,16 as well as, possibly, the transcription factor TonEBP (see below). Figure 1 illustrates some mechanisms through which a membrane protein may be activated by osmotic or mechanical perturbations. Further details are provided below.
Mechanisms through which a membrane protein may sense cell volume perturbations. (A) Direct stretch-sensitivity. (B) Direct or indirect tethering to the cytoskeleton. (C) Changes in membrane curvature, either mechanical or through changes in membrane composition. (D) Via interaction with integrins activated by the cell volume perturbation. See text for examples and details.
The Cytoskeleton in Osmosensing
Rapid, extensive reorganization of the actin cytoskeleton in response to osmotic stress has been demonstrated in a wide variety of cell types.24–30 The specific reorganization pattern varies between cell types but typically involves a net increase in F-actin content in the cortical region, loss of stress fibers, and increased length of F-actin protrusions in hypertonically shrunken cells, and the reverse pattern in swollen cells.1 The mechanisms involved in this reorganization are complex; however, several important moieties have been identified, including Rho family G proteins,31,32 cofilin,33 the ezrin/radixin/moesin (ERM) proteins ezrin34 and moesin,35 nonmuscle myosin II,36 cortactin, and the WASP/Arp2/3 system.28,37 Recent data also suggests the microtubule-based (MT) cytoskeleton is regulated by osmotic shrinkage.1,38,39 In fibroblasts and retinal pigment epithelial cells, we find that, in marked contrast to the actin cytoskeleton, MTs collapse upon hyper-osmotic stress and the MT plus the TIP protein end-binding protein (EB1) are lost from the tips, whereas hypotonic stress is associated with increased MT polymerization and EB1 tip-tracking.39 Is the cytoskeleton osmosensory, then? The sequence of events from osmotic volume perturbations to cytoskeletal reorganization is incompletely elucidated. However, available evidence suggests that, after osmotic shrinkage, very early events include changes in cellular PtdIns(4,5)P2 content, which can lead to changes in ERM protein activity and, in turn, to altered Rho activity,34,40 and which could also, tentatively, be upstream of the shrinkage-induced change in cofilin and WASP/Arp2/3 activity.1 Volume-dependent integrin activation (see below) could also be upstream of osmotic changes in Rho activity through p190RhoGAP, which is known to couple integrin activation to Rho.41
The cytoskeleton clearly contributes to the osmosensitivity of some ion transport proteins. An important example is TRPV4-mediated Ca2+ influx, which is part of the signaling activated by hypotonic stress in many epithelial cells and is F-actin dependent.1 Another is the Na+/K+/2Cl− co-transporter, NKCC1, an important player in RVI, which is dependent on F-actin both for quiescence in the unstimulated state and for activation by cell shrinkage.32,42,43
Integrins as Cell Volume Sensors
Integrin receptor heterodimers (α, β subunits) associate with a large array of cytoskeletal and signaling proteins to form Focal Adhesions (FAs). FAs are major cellular signaling hubs that link the actin cytoskeleton to integrins, and thus the ECM, playing an essential role in the regulation of cell survival, migration, and proliferation.46 In our view, it has yet to be proven whether integrins can act as primary osmosensors.1 However, integrins are clearly activated rapidly after cell volume perturbations in many cell types and have been proposed to serve as volume sensors both after swelling47–51 and after shrinkage.52–55
For instance, in the rat liver, cell swelling increases the plasma membrane level of activated β1-integrin.56 Interestingly, integrins form macromolecular complexes with certain ion channels, constituting a platform for downstream signals depending on both integrin signaling and channel activity.57,58 Several volume-sensitive K+ channels, including Kv1.359,60 and Kv11.1,57,61 are regulated by β1 integrins and, in turn, couple to other signaling molecules, and, for Kv11.1,62 also to growth factor and chemokine receptors.57 In myocytes, integrin stretch activates the volume-regulated anion channel (VRAC) through activation of focal adhesion kinase (FAK), Src, and the receptor tyrosine kinase (RTK) EGF receptor (EGFR, also known as ErbB1R).47–49 Finally, integrins are involved in the hypertonicity-induced increase in expression of tonicity-responsive enhancer-binding protein (TonEBP).55
Upon integrin activation, the nonreceptor tyrosine kinase, FAK, which is a central component of the FA, is recruited and undergoes phosphorylation on Tyr397 in the N-terminal domain.63 This recruits Src,64 phosphatidyl-inositol 3-kinase (PI-3K), and phospholipase Cγ (PLC-γ), and subsequent phosphorylation of Tyr925 promotes binding to the RTK-bound adaptor, Grb2.65 Thus, FAK is an important coordinator of integrin and RTK-mediated cellular signaling events. Other important phosphorylation sites on FAK are Tyr576 and Tyr577 in the kinase domain,66 and Tyr861 between the kinase domain and the Focal adhesion Targeting (FAT) domain.67 Hypotonic stress stimulates phosphorylation of tyrosine residues on FAK in several cell types,68–71 in a manner reported to depend on Rho family G proteins,68 or on the RTK, ErbB4.71 As noted above, in myocytes, FAK modulates VRAC after integrin stretch, yet this mechanism is distinct from that of swelling-activation of VRAC.47,72 Moreover, in endothelial cells, there seems to be no involvement of FAK in activation of VRAC.73 Hyper-osmotic stress likewise stimulates FAK phosphorylation in some cell types.74–77 FAK Tyr861 phosphorylation is stimulated by hyper-osmotic stress in many cell types, including epithelial cell lines, in which the phosphorylation is inhibited by Src inhibitors but insensitive to F-actin disrupting agents.76 Finally, we find that in NIH 3T3 fibroblasts, hyper-osmotic stress induces a rapid, transient increase in Src-dependent FAK phosphorylation at Tyr576 and Tyr,861 and independent phosphorylation at Tyr397 (B. Jørgensen, SFP, EKH, unpublished).
Receptor Tyrosine Kinases (RTKs)
Hypotonic swelling activates EGFR in several cell types,78–81 and interplay between integrins, FAK, Src and EGFR in swelling-activation of VRAC has been proposed.49 Hypertonic shrinkage phosphorylates both EGFR and the insulin receptor in some cell types, and shrinkage-induced receptor clustering has been proposed as a mechanism of osmosensing.82 On the other hand, cell shrinkage inhibits some RTKs, including PDGF receptor β (PDGFRβ), resulting in reduced Akt and extracellular signal regulated kinase (ERK) activity.83,84 Moreover, in hypertonically stressed rat hepatocytes, EGFR activation is downstream of a Src family kinase,85 and in Vero kidney cells, shrinkage-inhibition of EGFR signaling is downstream of Ras.83 Thus, EGFR is clearly not a primary osmosensor in all cell types. However, the role of RTK regulation as an early signal in the osmotransduction process is well established. Interaction between RTKs and integrins probably contributes to the complexity, both because of the many converging signaling pathways, as noted above, and because integrin activation is known to transactivate RTKs.49,86,87
Transient Receptor Potential Channels (TRPs) and Mechanosensitive Channels
The mammalian TRP channels are divided into six subfamilies: the canonical (TRPC), vanilloid (TRPV), melastatin (TRPM), mucolipin (TRPML), ankyrin (TRPA1), and polycystin (TRPP) families.88–91 The selectivity of TRPs range from cation nonselective to highly Ca2+-selective.88,91 Several TRPs are sensitive to cell volume and/or to membrane expansion or stretch.1,7,92,93 Some of these, such as TRPC1, TRPC6, TRPV4, and TRPM7, are ubiquitously expressed and might be volume sensors.88,89,91 While a detailed discussion of this is outside the scope of this review, it is important to note that experimental separation of osmosensitivity and mechanosensitivity is challenging, as is determination of the precise stimulus that triggers activation of a channel in response to cell swelling or membrane stretch.7,94 Hence, whether swelling or stretch regulates TRP activity is not always clear. The TRPV family is, by far, the best studied in the context of volume sensing. TRPV1 is unequivocally shown to be involved in osmosensing, as indicated from studies of TRPV1−/− mice95 (see below). When expressed in yeast, however, TRPV1 responds to heat but not to hypotonicity, leading to the suggestion that TRPV1 may not be directly osmosensitive or mechanosensitive.96 It is, however, shown that an N-terminal variant of TRPV1 is essential for osmosensory transduction in mouse SON neurons97 (see also below). TRPV4, the mammalian homologue of C. elegans OSM-9, is clearly a swelling-activated channel.95,98–100 Thus, TRPV4 mediates swelling-activated Ca2+ influx,101,102 cells from TRPV4−/− mice have reduced RVD rates,95 and mammalian TRPV4 rescues mechanosensitive and osmosensitive defects in OSM-9 mutants.103 In mammalian cells, hypotonic swelling activates phospholipase A2,104 releasing arachidonic acid (AA), and swelling-activation of TRPV4 is dependent on the AA metabolite 5′, 6′-epoxyeicosatrienoic acid (EET).105–108 Interestingly, although yeast cannot synthesize EET, TRPV4 also responds to hypotonicity when expressed in yeast.96 This may suggest that direct mechanosensitivity of TRPV4 is possible under some conditions. The mechanism is unclear, as it was previously shown that swelling-activation of TRPV4 was independent of direct membrane stretch.98,109,110 TRPV4 activity is central to actin cytoskeleton-dependent RVD,111 and forms a supramolecular complex containing regulatory kinases and cytoskeletal proteins.112 Furthermore, TRPV4 is activated after integrin activation by stretch.113 While a detailed discussion of this will not be given here, a number of other protein-protein interactions appear to be involved in the osmosensitivity of TRPV4, including With No Lysine kinase (WNK4)114 and aquaporin 5.115,116
TRPM7 is also proposed to be activated both by mechanical stress117 and by osmotic swelling, and knockdown of TRPM7118 attenuates RVD. Finally, TRPC1 and TRPC6 are reported to be mechanosensitive119,120; however, this has been disputed recently.121,122 It is possible that stretch sensitivity of these TRPs is indirect and requires components that are differentially expressed in different cell types.123 In accordance with this suggestion, TRPs interacts with many molecules of relevance to osmosensing, such as PLCγ-1,124 PtdIns(4,5)P2, aquaporin 5, various protein kinases, and cytoskeletal proteins.88,91,124,125 For instance, for TRPC6/3, the angiotensin II receptor (AT1R) acts as the primary mechanosensor upstream of the channel.125 Nonetheless, it is a contentious issue whether mammalian TRPs are directly mechanosensitive.7,126 Genetic screens in C. elegans127,128 and D. melanogaster129 have identified mechanosensitive TRPs.130 In bacteria, stretch-activated, nonselective cation channels (SACs) have been cloned and analyzed in detail.94,131 Nonselective cation currents activated by cell swelling and stretch has also been demonstrated in cells from vertebrate organisms.132–134 Recently, two proteins, Piezo1 and Piezo2, were identified, overexpression of which results in mechanically activated currents very similar to SACs.136 Whether Piezos contain the pore region of SAC or, rather, are components necessary for channel function is unclear.
CENTRAL OSMOSENSORS
Even very minor changes in systemic osmolarity are associated with severe symptoms,137 and extracellular fluid (ECF) osmolarity is, accordingly, very tightly regulated. ECF hyper-osmolarity (generally caused by increased salt intake relative to water) stimulates central osmoreceptor cells, which transmit this signal to orchestrate a series of signaling and ion transport events in other central and peripheral cells. Collectively, this results in increased thirst and in vasopressin (AVP) release, the latter, in turn, leading to increased free water retention through antidiuresis and natriuresis. Conversely, when ECM hypo-osmolarity is detected by osmoreceptor neurons, the net result is reduced AVP release and diuresis, and thus reduced water retention.138 Figure 2 outlines the overall mechanisms of systemic osmoregulation.
Overview of the processes of extracellular fluid (ECF) osmolarity disturbance and regulation. ECF hyper-osmolarity causes osmotic shrinkage of osmosensory cells (likely both central and peripheral). This stimulates AVP release, increased water retention and Na+ excretion, increased thirst and increased Na+ appetite, collectively leading to normalization of ECF osmolarity and osmosensory cell volume. Conversely, ECF hypo-osmolarity results in osmosensory cell swelling, leading to reduced AVP release and, in turn, water excretion, salt uptake, and reduced thirst, and, hence, normalization of ECF osmolarity and osmosensory cell volume. See text for details.
The precise location of the central osmosensory mechanisms responding to changes in systemic osmolarity has only recently been elucidated, and much is still obscure regarding these essential homeostatic mechanisms.138,139 A major part of the systemic response to perturbations in ECF osmolarity is mediated by osmoreceptor neurons found in two locations in the central nervous system (CNS). Available evidence suggests that the most influential intrinsically osmosensitive neurons in the brain reside within the forebrain lamina terminalis, specifically in the organum vasculosum laminaee terminalis (OVLT) and the subfornical organ (SFO).139 Intrinsically osmosensitive neurons are also found in magnocellular neurosecretory cells (MNCs) in the supraoptic nucleus (SON) and paraventricular nucleus (PVN) in the hypothalamus.140
As first described by Verney,141 these osmoreceptor neurons perceive ECM volume perturbations and translate them into electrical activity that is transmitted to other neurons as changes in action potential firing, eventually resulting in altered AVP release from the pituitary gland. Which, if any, of the volume sensory mechanisms outlined above are employed by the central osmosensory neurons? In general, it seems that osmoreceptor neurons respond to hyper-osmotic stimuli by activation of a nonselective cation conductance, resulting in plasma membrane depolarization and thus increased action potential firing rate, whereas the reverse is true for hypo-osmotic stimuli.142,143 The available evidence indicates that of the three main types of volume signals outlined above, the central osmosensors react to the mechanical stimulus elicited by cell volume change.138,144
Biophysical characteristics and inhibitor profiles indicated, a number of years ago, that at least some of these nonselective cation channels responsible for central osmosensing belong to the TRP family. Specifically, an N-terminal variant of TRPV1 appears to be activated by cell shrinkage and is involved in central osmosensing of ECF hyper-osmolarity, at least in the acute phase97 but possibly not in the chronic phase.145,146 In contrast, TRPV4 seems to be essential in systemic sensing of hypo-osmolarity. Thus, TRPV4−/− mice have aberrant systemic osmosensitivity,147,148 and TRPV4 staining in the brain reveals high expression in the OVLT and SON.148 The general mechanisms through which TRPV4 is activated by osmotic cell swelling are outlined above. In the context of central osmosensing, it is interesting to note that the central response to perturbations of ECF osmolarity is strongly dependent on the actin cytoskeleton.144 Moreover, in a rat model of liver cirrhosis, TRPV4 expression is increased, and its lipid raft association decreased, in the SON, correlating with elevated AVP release.149 This suggests that lipid rafts may regulate TRPV4 function and that altered TRPV4 expression and localization may contribute to deranged AVP release in cirrhosis.
In addition to neuronal TRPV1 and TRPV4, a number of other mechanisms are likely to contribute to systemic osmosensing. Thus, several stretch-activated two-pore K+ channels were identified in MNG neurons,150,151 and, recently, hyper-osmotic dehydration-induced activation of a voltage-dependent K+ current152 and L-type Ca2+ currents153 was reported in these cells. Finally, swelling-activated glial taurine release seems to contribute to AVP release through activation of neuronal glycine receptors.154
PERIPHERAL OSMOSENSORS
While surprisingly little is known about peripheral sensing of systemic osmotic perturbations, the existence of peripheral osmoreceptors has been suggested by two sets of intriguing observations. The first is the presence of anticipatory mechanisms: osmotically induced hormonal or behavioral responses occur much faster than any other change detected in the ECF, that is, before the information can reach the central osmosensors. Thus, water drinking quenches thirst,155–158 and changes in gastric osmotic load due to water or food intake modulate AVP secretion159–161 before any alteration in ECF osmolarity. Second, local changes in osmotic concentration (for example, due to oral or intragastric sodium or water load) elicit larger regulatory responses than what the same load provokes when administered intravenously162 Based on these findings, peripheral osmoreceptors have been proposed to localize to the oropharyngeal area,163 along the gastrointestinal tract (stomach, duodenum),159,164,165 in the liver,166 portal vein,167,168 and splanchnic mesentery.169 These receptors might sense changes in osmolarity, cell volume, or salt concentration (sodium monitor).138 So far, neither the sensed parameters, nor the identity of the receptors, nor the sensing mechanisms have been established with absolute certainty. Nonetheless, osmotic activation of TRPV4 channels (see above) emerges as an important mechanism.162,170 Accordingly, TRPV4−/− mice exhibit impaired adaptive responses to changes in gastroduodenal osmolarity.171 The osmosensors may be local cells or sensory neurons such as vagal afferents or—as increasing evidence suggests—neurons of the dorsal root ganglia.172–177 Once stimulated, they send signals to the CNS, where they modulate thirst and satiety and AVP secretion and/or to peripheral effector organs, particularly to the kidney. Regarding the latter, increased gastric osmolarity reduces sympathetic renal nerve activity, which, in turn, suppresses the intrarenal renin-angiotesin system, resulting in enhanced natriuresis.178 Clearly, the characterization of peripheral systemic osmosensors is a topic of primary importance, which lags behind our more in-depth understanding of the central mechanisms and warrants more attention in the future.
KIDNEY CELLS AS POTENTIAL PERIPHERAL OSMOSENSORS?
While the mechanisms through which renal epithelial cells sense and regulate their own volume have been well characterized,1,2,29 much less is known regarding the possible role of kidney cells as peripheral sensors of systemic osmolarity. Interestingly, TRPV4 is abundantly expressed along the nephron,179,180 particularly enriched along in the collecting duct. This raises the possibility that tubular cells might themselves act as systemic osmosensors. Accordingly, hypotonicity was shown to provoke TRPV4-triggered ATP release from the thick ascending limb.181 This phenomenon is of particular interest, because swelling-induced ATP release in the macula densa may be a key mediator of tubuloglomerular feedback.182 Besides its effects on renal hemodynamics, extracellular ATP also controls salt reabsorption in several nephron segments.183 Moreover, TRPV4 and polycystin-2 (TRPP2) were shown to form a polymodal sensory complex in tubular cells, which acts as a flow-activated, Ca2+-permeable channel in the primary cilium.22 Finally, increased flow alters tubular K+ secretion and Na+ reabsorption in a TRPV4-dependent manner.184 This scenario then raises a multitude of interesting questions. What are the effectors through which renal epithelial cells relay information about systemic osmolarity to other cell groups—does this involve ATP and/or other paracrine factors? How do different modalities (changes in osmolarity versus flow) acting through the same sensor get integrated into the final response?185 What is the contribution of TRPV4 versus other renal TRP channels, such as TRPM3,179 to the overall responses of renal cells to aniso-osmolarity? How are the sensory inputs through TRP channels integrated with signals emanating from other cellular osmosensors such as integrins or growth factors? Future research should address these intriguing questions, using genetically manipulated tubular cells and transgenic animals with tubule-specific deletion of TRPV4 or other osmosensors or volume sensors.
ARE MACROPHAGES MOBILE PERIPHERAL OSMOSENSORS?
It has been repeatedly shown that, during salt loading, Na+ can accumulate in the skin without commensurate water,186,187 presumably due to negatively charged glycosaminoglycans. This Na+ is in dynamic equilibrium with the viscous connective tissue and, in this sense, it is not osmotically inactive. Indeed, the skin and the lymphatic tissue have higher osmolarity than the blood plasma.188,189 According to a new paradigm, this local hypertonicity activates TonEPB signaling in mononuclear phagocytes, which, in turn, secrete vascular endothelial growth factor-C (VEGF-C).190–192 VEGF-C induces lymphatic vessel formation and endothelial nitric oxide synthase expression, thereby modulating BP, volume homeostasis, and, presumably, kidney function. Notably, deletion of macrophages augments the Na+ load-induced rise in BP.190 These findings are especially intriguing, since, originally, TonEPB was shown to be essential for the induction of osmoprotective genes in highly hyper-osmotic environments, such as the kidney medulla.193,194 However, as the above-mentioned data show, in addition to these renoprotective functions exerted in a robustly hyper-osmotic milieu, TonEBP may play an important regulatory role in nonrenal cells under (nearly) iso-osmotic conditions.192 Future studies should address whether TonEBP, which may be considered as a self-contained tonicity- or ionic-strength sensor/effector system,195–197 plays additional roles in central or peripheral osmosensing during health and disease.
DISTURBANCES IN OSMOSENSING: A RENAL VIEWPOINT
The kidney in is involved in osmosensory pathophysiology in two conceptually different aspects. First, it is the main effector organ of systemic osmoregulation responsible for water and salt homeostasis, and, as such, it is controlled by central osmosensing machinery through both hormonal and neuronal pathways. Accordingly, insufficiency of AVP production or secretion causes a plethora of diseases, including eneuresis,198 nocturia,199 and central diabetes insipidus. Conversely, incapability of the kidney to respond to AVP results in nephrogenic diabetes insipidus, either acquired or congenital, the latter most frequently caused by various mutations affecting AVP Receptor-2 (90%) or aquaporin-2 (10%). This field has been extensively covered by excellent reviews.200–203 Second, since the kidney itself, as discussed above, emerges as a systemic osmosensing and mechanosensing organ, pathologies can be associated with anomalies of the molecular apparatus responsible for these functions in the renal cells. As noted above, such molecular sensors and the related signaling systems are highly enriched in the primary cilium. Recently, numerous so-called ciliopathies have been identified in which impaired ciliary function underlies severe mechanosensory, developmental, or degenerative disorders.204–206 Of these, the prototypic example is autosomal dominant polycystic kidney disease, which is caused by mutations in polycystin-1 (PKD1, TRPP1) or polycystin-2 (PKD2, TRPP2). Under normal conditions, these proteins form a mechanosensitive cation channel in the primary cilium, which opens upon the flow-induced ciliary bending, generating a Ca2+ signal.20,207 PKD mutations impair mechanically-evoked Ca2+ transients and lead to the formation of fluid-filled cysts and enhanced cell proliferation, implying that intact mechanotransduction is essential for normal tubular function and structure.208,209 As mentioned above, PKD2 can form a Ca2+-conducting complex with the osmo-/volume-sensitive TRPV4,22 raising the possibility that osmotic stimuli might also contribute to the preservation of normal kidney architecture and prevent cyst formation. On the other hand, it should be stressed that TRPV4−/− mice and zebrafish lack renal cysts, arguing against a role for at least TRPV4 in preventing cyst formation.22 Finally, although, so far, no TRPV4 or TRPM3 mutations have been directly associated with known renal disorders (see, however,210) it is noteworthy that an array of other TRP channel malfunctions has been shown to play pathogenic roles in various kidney diseases, including diabetic nephropathy (TRPC1), focal segmental glomerulosclerosis (TRPC6), and hypomagnesemia (TRPM6).179
CONCLUSIONS
An interesting point, often forgotten in the discussion of cellular volume regulation, is that a cell that regulates its own volume in response to an ECM perturbation is, in effect, compromising the common good of the organism it is part of, by negatively impacting on systemic volume regulation. Thus the evolutionary development of systemic osmoregulation must have modified the elementary volume-regulating responses of specialized cells. What determines whether cellular or systemic volume regulation is prioritized? Notably and logically, osmoreceptor neurons seem to lack functional volume-regulatory machinery and thus act as osmometers, swelling and shrinking in a manner correlating with the difference in tonicity between the cell and the ECF to which they are exposed.8 In our view, essential areas of research needed to be explored in future studies include the extended molecular identification of osmosensors and volume sensors and their subcellular locations, the clarification of the nature of the upstream signals of volume change, and the mechanisms whereby they interact with the primary sensors in health and disease.
Disclosures
Work in the author's laboratories is supported by the Danish Council for Independent Research: Natural Sciences (SFP, EKH) and Medical Sciences (SFP), the Novo Nordisk Foundation (SFP, EKH), the Lundbeck Foundation (SFP, EKH), the Danish Cancer Society (SFP), the Natural Sciences & Engineering Council of Canada (AK, contract grant number 227908), and Canadian Institutes of Health Research (AK, contract grant number MOP 86535).
Footnotes
Published online ahead of print. Publication date available at www.jasn.org.
- Copyright © 2011 by the American Society of Nephrology
REFERENCES
- 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.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.
- 45.
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
- 84.↵
- 85.↵
- 86.↵
- 87.↵
- 88.↵
- 89.↵
- 90.↵
- 91.↵
- 92.↵
- 93.↵
- 94.↵
- 95.↵
- 96.↵
- 97.↵
- 98.↵
- 99.↵
- 100.↵
- 101.↵
- 102.↵
- 103.↵
- 104.↵
- 105.↵
- 106.↵
- 107.↵
- 108.↵
- 109.↵
- 110.↵
- 111.↵
- 112.↵
- 113.↵
- 114.↵
- 115.↵
- 116.↵
- 117.↵
- 118.↵
- 119.↵
- 120.↵
- 121.↵
- 122.↵
- 123.↵
- 124.↵
- 125.↵
- 126.↵
- 127.↵
- 128.↵
- 129.↵
- 130.↵
- 131.↵
- 132.↵
- 133.↵
- 134.↵
- 135.
- 136.↵
- 137.↵
- 138.↵
- 139.↵
- 140.↵
- 141.↵
- 142.↵
- 143.↵
- 144.↵
- 145.↵
- 146.↵
- 147.↵
- 148.↵
- 149.↵
- 150.↵
- 151.↵
- 152.↵
- 153.↵
- 154.↵
- 155.↵
- 156.↵
- 157.↵
- 158.↵
- 159.↵
- 160.↵
- 161.↵
- 162.↵
- 163.↵
- 164.↵
- 165.↵
- 166.↵
- 167.↵
- 168.↵
- 169.↵
- 170.↵
- 171.↵
- 172.↵
- 173.↵
- 174.↵
- 175.↵
- 176.↵
- 177.↵
- 178.↵
- 179.↵
- 180.↵
- 181.↵
- 182.↵
- 183.↵
- 184.↵
- 185.↵
- 186.↵
- 187.↵
- 188.↵
- 189.↵
- 190.↵
- 191.↵
- 192.↵
- 193.↵
- 194.↵
- 195.↵
- 196.↵
- 197.↵
- 198.↵
- 199.↵
- 200.↵
- 201.↵
- 202.↵
- 203.↵
- 204.↵
- 205.↵
- 206.↵
- 207.↵
- 208.↵
- 209.↵
- 210.↵