Visual Overview
Abstract
Volume changes deviating from original cell volume represent a major challenge for cellular homeostasis. Cell volume may be altered either by variations in the external osmolarity or by disturbances in the transmembrane ion gradients that generate an osmotic imbalance. Cells respond to anisotonicity-induced volume changes by active regulatory mechanisms that modify the intracellular/extracellular concentrations of K+, Cl–, Na+, and organic osmolytes in the direction necessary to reestablish the osmotic equilibrium. Corrective osmolyte fluxes permeate across channels that have a relevant role in cell volume regulation. Channels also participate as causal actors in necrotic swelling and apoptotic volume decrease. This is an overview of the types of channels involved in either corrective or pathologic changes in cell volume. The review also underlines the contribution of transient receptor potential (TRP) channels, notably TRPV4, in volume regulation after swelling and describes the role of other TRPs in volume changes linked to apoptosis and necrosis. Lastly we discuss findings showing that multimers derived from LRRC8A (leucine-rich repeat containing 8A) gene are structural components of the volume-regulated Cl– channel (VRAC), and we underline the intriguing possibility that different heteromer combinations comprise channels with different intrinsic properties that allow permeation of the heterogenous group of molecules acting as organic osmolytes.
Introduction
Cell volume is characteristic of each cellular lineage and is maintained with few variations through the cell life. Conditions inducing changes in cell volume represent a major challenge for cell homeostasis and may even culminate in cell death. To keep cell volume constant, water fluxes should be in equilibrium between the intracellular and extracellular compartments. This occurs under conditions of isotonicity, but any fluctuations in the osmolarity of the extracellular milieu or in the distribution of osmotically active solutes drives water fluxes in the necessary direction to reach a new equilibrium, which disturbs cell volume. Except for cells in the kidney and intestine, which normally face large variations in external osmolarity, the extracellular milieu maintains a well-controlled osmolarity. This is modified only under pathologic conditions leading to hyponatremia or in other pathologies that alter the distribution of osmolytes in the extra-intracellular compartments (Song and Yu, 2014; Pedersen et al., 2016). The intracellular osmolarity also endures continuous variations owing to transient and local osmotic microgradients generated during physiologic processes, such as uptake and release of metabolites, cytoskeletal remodeling, synthesis and degradation of macromolecules, exocytosis and secretion (Pedersen et al., 2001; Strbák, 2011; Platonova et al., 2015). Even these localized changes in cell volume should be regulated to prevent changes in the concentration of molecules free in the cytosol, protein misfolding, or alterations in cell and organelle architecture. Cell volume is modified during migration, proliferation, and cell adhesion; therefore, volume change has been proposed as a signaling event for these processes (Lang et al., 2005; Stutzin and Hoffmann, 2006; Dubois and Rouzaire-Dubois, 2012).
When volume is perturbed as a result of changes in the extracellular osmolarity, cells respond by activation of a plethora of signaling pathways and mechanisms intended to protect them from the challenge of volume change; remodeling of the cytoskeleton, changes in adhesion molecules, activation of stress and survival signals are among the adaptive mechanisms triggered by the altered volume (Pasantes-Morales et al., 2006). At the same time, cells set in motion an active regulatory process directed to partially or fully restore the original volume (Cala, 1977; Olson et al., 1986; Hoffmann, 1992; Okada, 2004). This adaptive mechanism, which has been highly preserved during evolution, essentially consists of the redistribution of osmotically active solutes in the necessary direction to equilibrate water fluxes facing the new osmotic condition (Lang, 2007; see Hoffmann et al., 2009 for a comprehensive review of cell volume regulation). Osmolytes that accomplish cell volume regulation are the ions present in high concentrations in the intracellular or extracellular compartments—such as Na+, K+ and Cl–—and a group of heterogeneous organic molecules, such as amino acids, polyamines, and polyalcohols (Verbalis and Gullans, 1991; Pasantes-Morales et al., 1998; Burg and Ferraris, 2008). Volume-regulated channels and cotransporters translocate ion osmolytes, and cotransporters and pores, as yet fully identified, mobilize organic osmolytes; channels participate predominantly in the volume-regulatory process activated by cell swelling, whereas cotransporters play a more important role in the cell response to shrinkage (Kahle et al., 2015). Besides their role in volume regulation, channels are involved in the pathophysiologic cell volume changes generated in isotonicity that lead to necrotic death, as well as in those associated with apoptosis (Bortner and Cidlowski, 1998, 2007; Okada et al., 2006). This review focuses on: 1) channels through which ions and organic osmolytes permeate during regulatory volume decrease, 2) ion channels participating in volume regulatory increase, and 3) channels implicated in normotonic cell volume changes leading to apoptotic and necrotic cell death.
The Role of Channels in Regulatory Volume Decrease
A decrease in external osmolarity modifies the water concentration in the extracellular space, which activates water inward-directed fluxes and increases the cell volume with a magnitude proportional to the extent of the osmolarity change. As mentioned in the Introduction, most animal cells respond to this situation with an active process of volume recovery known as regulatory volume decrease (RVD) (Cala, 1977), which is accomplished by extrusion of the major intracellular ions, K+ and Cl–, and organic osmolytes. Most of the corrective fluxes occur via K+ and Cl– channels and diffusion pathways for the organic osmolytes (Hoffmann and Lambert, 1983; Sánchez-Olea et al., 1991; Okada and Maeno, 2001; Hoffmann et al., 2009) (Fig. 1).
K+ Channels.
K+ is a major intracellular osmolyte and contributes substantially to RVD. The osmosensitive K+ fluxes permeate across a large variety of K+ channels, that are involved in multiple physiologic processes in the cell other than volume regulation, but are triggered via signals evoked by swelling. Four main classes of K+ channels are activated during RVD: voltage-gated K+ (Kv) channels, Ca2+-activated K+ (KCa) channels, inwardly rectifying K+ (KIR) channels, and two-pore–domain K+ channels (K2P) (Fig. 1). The kind of channel involved differs in the various cell types and responds to one or several of the signals elicited by the volume change, such as depolarization, membrane stretch, and elevation of intracellular Ca2+ [Ca2+]i. In epithelial and other cell types, swelling evokes a large increase in [Ca2+]i, and RVD in these cells is consistently Ca2+-dependent owing to the predominant role of KCa channels in the volume-corrective K+ fluxes. The high- and intermediate-conductance K+ channels (BKCa, IKCa) are the channels mainly implicated in RVD. High conductance or big potassium (BK) channels are activated by swelling in the intestine, kidney, and bronchial epithelia, chondrocytes, pituitary tumor cells, and osteoblast-like cells (reviewed in Hoffmann et al., 2009). Activation of BK channels occurs by Ca2+ influx mediated by the TRPV4 subtype, establishing a link between BKCa channels and some transient receptor potential (TRP) channels (discussed later) (Jin et al., 2012). IKCa channels participate in RVD in T lymphocytes, erythrocytes, hepatocytes, osteosarcoma cells, intestine 407 cells, proximal tubule cells, and lens epithelia (rev. in Hoffmann et al., 2009). In other cells—such as brain cells—following hypotonic swelling, depolarization generated by the Cl– efflux occurs across the volume-regulated Cl– channel, and in these cells RVD is essentially Ca2+-independent (Pasantes-Morales et al., 1994; Morán et al., 1997) and the K+ regulatory fluxes permeate across voltage-gated K+ channels. Subtypes of Kv involved in the different cell types include Kv1.3 in lymphocytes and murine spermatozoa, Kv1.4.2 and Kv1.4.3 in myocytes, and Kv1.5 in lymphocytes and spermatozoa. KCNQ1 (Kv7.1), in various subunit arrangements, participates in RVD in human mammary epithelial cells, rat hepatocytes, cells from the inner ear, and in mouse trachea epithelium (reviewed in Hoffmann et al., 2009). The inwardly rectifying K+ channels Kir4.1 and Kir4.5, which sense small changes in osmolarity (Grunnet et al., 2003; Soe et al., 2009), colocalize with aquaporin 4 (AQP4) in the glial endfeet of the blood-brain barrier, suggesting that they play a role in swelling and/or volume control in these cells. Kir channels regulate extracellular K+ homeostasis during the K+ spatial buffering through the glial syncytium, and this transcellular K+ transport is accompanied by transmembrane water fluxes, likely via a Kir4.1/AQP4 complex (Nagelhus et al., 1999; Dibaj et al., 2007). The two-pore–domain K+ channels participate in RVD in lymphocytes, Ehrlich ascites cells, astrocytes, and kidney cells (Niemeyer et al., 2001; Kirkegaard et al., 2010; Andronic et al., 2013).
The variety of K+ channels in different cell types that are involved in RVD and triggered by stimuli concurrent with cell swelling raises the question about the existence of a purely swelling-gated K+ channel. A recent study, in which AQP1 and a number of K+ channel types were coexpressed in Xenopus oocytes (Tejada et al., 2014) identified the Kca 4.1 channel (Slick; Slo2.1) as the most sensitive to volume changes. Slick currents increased markedly in hypotonicity and decreased in hypertonicity, and these changes were not observed in the absence of AQP1, indicating that channel activity follows the changes in cell volume and not in osmolarity. Thus, Slick may be a purely volume-sensitive K+ channel. Although Kca 4.1 channel was originally included in the Kca channel nomenclature, it is now known that it is activated by internal Na+ and Cl– and not by Ca2+.
The Volume-Regulated Cl– Channel.
Volume-sensitive Cl– fluxes play a key role in RVD; in contrast to the diversity of K+ channels implicated in RVD in various cell types, Cl– fluxes permeate across only one type of channel, the volume-regulated anion channel (VRAC), also named volume-sensitive outwardly rectifying anion channel. The channel is also referred to as volume-sensitive organic osmolyte and anion channel to denote the possibility that it permeates organic osmolytes. VRAC will be the name used throughout this review. VRAC, widely expressed in animal cells, is essentially inactive under isotonic conditions and opens slowly by cell swelling or by a decrease in ionic strength under isosmotic conditions (Emma et al., 1997; Voets et al., 1999; Pedersen et al., 2016). In biophysical terms, the Cl– current (ICl–swell) carried by VRAC exhibits mild outward rectification and voltage inactivation at positive membrane potentials, and this differs among cell types (Nilius et al., 1994; Nilius and Droogmans, 2001; Okada, 2004; Akita and Okada, 2014). This outward rectification defines VRAC and establishes clear differences from other volume-sensitive Cl– channels, such as channels with no rectification (maxi-anion channel), inward rectification (CLC2), steep outward rectification (Cl–C3 and anoctamine), or no rectification (bestrophin). The single-channel VRAC conductance is 50–80 pS at positive and 10–20 pS at negative membrane potentials. VRAC exhibits a low field-strength anion permeability (I>Br>Cl>F). In this respect, VRAC differs from the ClC family members, which are typically selective for chloride over iodide. VRAC permeates large molecules such as gluconate, glutamate, and benzoate, and the pore size of this channel has been estimated to be about 11–17 Å (Nilius et al., 1997).
The search for VRAC-specific blockers has been largely unsuccessful, but the long list of nonspecific VRAC blockers (reviewed in Pedersen et al., 2016) include: 1) conventional anion channel inhibitors, such as DCPIB, DIDS, SITS, NPPB, and the acidic di-aryl-urea NS3728; 2) transporter blockers, such as fluoxetin, phloretin, and mefloquine; 3) the antiestrogens tamoxifen, clomiphene, and nafoxidine; 4) tyrosine kinase inhibitors, such as genistein, tyrphostin A23, and PD98059; 5) purinergic receptor blockers suramin, reactive blue 2, and PPADS; and 6) carbenoxolone and riluzol. However, in addition to inhibiting VRAC, these inhibitors block a variety of other Cl– channels, transporters, or receptors to some extent (Pedersen et al., 2016). The failure to find a selective VRAC blocker has complicated efforts to establish the molecular identity of this channel.
The VRAC gating mechanisms are not fully understood, but allosteric nonhydrolyzed-bound ATP and permissive cytosolic Ca2+ are necessary for channel activation. It has been proposed that G protein-mediated signal transduction pathways and membrane structures, such as caveolin, cholesterol, and actin cytoskeleton influence the channel activation mechanism. It has been suggested that the Ras-Raf-MEK-ERK pathway and the Rho/Rho kinase are modulators of VRAC activity (Nilius et al., 1999). A volume-transduction pathway mediated by the epidermal growth factor receptor and NAD(P)H oxidase-derived H2O2 has been proposed as a signaling pathway for VRAC activation (Varela et al., 2004). Epidermal growth factor receptor is activated by hyposmolarity and is an early signal that modulates osmolyte efflux pathways (Pasantes-Morales et al., 2006). Additionally, reduction in ionic strength is a key element for VRAC activation; early studies by the Nilius group (Nilius et al., 1998; Sabirov et al., 2000) demonstrated that reducing ionic strength activates a Cl– current, with biophysical and pharmacological features identical to those of VRAC, even without cell swelling. This mechanism of VRAC activation is of high physiologic relevance. Under physiologic conditions, the osmolality changes are small and gradual and, in most cases, insufficient to generate significant changes in cell volume or in membrane distension, but the osmotic disequilibrium still should be corrected. Corrective Cl– fluxes permeating across VRAC, activated by local changes in the ionic environment, appear as the best option for an isovolumetric regulation (see following). More direct evidence on the role of ionic strength in VRAC gating is further discussed in the context of the identification of LRRC8 proteins as components of the VRAC structure.
In 2014, two independent groups (Qiu et al., 2014; Voss et al., 2014), using a similar fluorescence assay for a screening of human genome small-interfering RNA libraries, identified a multispan transmembrane protein derived from a gene of unknown function named LRRC8A (leucine-rich repeat containing 8A). The LRRC8A subunit protein was postulated to be an integral component of the VRAC structure. LRRC8A belongs to the LRRC8 family, which consists of five members, LRRC8A–LRRC8E. All the LRRC8 proteins have four putative transmembrane domains and contain up to 17 leucine-rich repeats. Studies by the groups mentioned earlier suggest that LRRC8 isoforms, similar to the pannexins to which they are related, organize in heteromeric complexes to form a functional channel with VRAC properties. It was also shown that the presence of LRRC8A in the multimeric complex is essential, but not sufficient alone for VRAC activity, since at least one of the other LRRC8 isoforms should be present together with LRRC8A. Genetic ablation of LRRC8B–LRRC8E did not separately affect VRAC currents, but disruption of the five genes abolished VRAC activity. The functional channel can be restored by expression of LRRC8A together with one of the other four LRRC8 subunits (Qiu et al., 2014; Voss et al., 2014). The largest reductions in ICl–swell amplitudes are observed in LRRC8E−/− and in LRRC8(C/E)−/− cells and the lowest reductions in LRRC8B−/−, LRRC8D−/− cells (Voss et al., 2014). Different combinations of LRRCA and other heteromers define intrinsic properties of the channel. Coexpression of LRRC8A and LRRC8E induces faster inactivation at less positive potentials, whereas coexpression of LCRRC8A and LRRC8C has the opposite effect; this coexpression markedly reduces the inactivation rate (Voss et al., 2014). Different combinations of LCRR8 heteromers also strikingly modify VRAC permeability. As discussed in detail later, VRAC permeability to taurine has a strict requirement of LCRR8A and LCRR8D, whereas LCRR8BCE subunits appear unnecessary (Planells-Cases et al., 2015) (Fig. 2).
The possibility that LRRC8A constitutes the anion-conducting pore domain of VRAC is questioned (Akita and Okada, 2014) on the basis of the lack of influence on currents across the LCRR8 complexes of charged amino acid mutations in predicted transmembrane domains. This suggests that the postulated VRAC structure may require a regulator molecule. Decrease, and not an increase, in IClswell following overexpression of LCRR8A also is suggestive of a regulator molecule. Recently, Syeda et al. (2016) assembled models of different combinations of LCRR8A and LCRR8BD subunits in which the pore-forming channels were incorporated into bilayers. Using this approach it was possible to demonstrate that, as previously shown (Qiu et al., 2014; Voss et al., 2014), the different combinations of LCRR8A with LCRR8B–D subunits determine intrinsic channel properties, such as rectification, inactivation kinetics, and relative anion permeability. The study also showed that reducing intracellular ionic strength directly gates the channel, even in the absence of mechanical changes in the membrane or the cellular components, such as cytoskeletal structures or elements of signaling chains. The relevance of these findings is that differently composed heteromers formed by combinations of LRRC8 subunits may result in a large variety of channels with different intrinsic properties and different regulatory mechanisms, which helps explain the differences in the inactivation rate of VRAC in various cell types and the possibility of permeating molecules as heterogeneous as the group of organic osmolytes. It also points to the possibility that one or some of the LCRR8 subunits forming VRAC contain the sensor for changes in ionic strength, the proposed channel gating (Cannon et al., 1998; Nilius et al., 1998; Sabirov et al., 2000). The relevance of these findings to define the molecular identity of VRAC is extensively discussed in several excellent recent reviews (Pedersen et al., 2015, 2016; Stauber, 2015; Jentsch et al., 2016).
Organic Osmolytes.
Organic osmolytes, which contribute to cell volume regulation in most animal cells (Kinne, 1993), are small molecules of heterogeneous structure that include amino acids and derivatives (taurine, glutamate, glycine, GABA, and N-acetylaspartate), polyalcohols (myo-inositol and sorbitol), and amines (glycerophosphorylcholine, betaine, creatine/P-creatine, and phosphoethanolamine). The concentration of organic osmolytes varies in the different tissues; glutamate, myo-inositol, creatine, taurine, and N-acetylaspartate are present in the highest concentration in the brain (Verbalis and Gullans, 1991), whereas glycerophosphorylcholine, betaine, myo-inositol, and sorbitol are the major organic osmolytes in renal cells (Beck et al., 1992).
Organic osmolytes are important in the process of volume regulation; in contrast to ions, which generate adverse effects that disturb the structure of macromolecules or affect neuronal excitability, many organic osmolytes are compatible molecules. Thus, in the long term, organic osmolytes tend to replace ionic osmolytes (Verbalis and Gullans, 1991). In the nervous system, glutamate, γ-aminobutyric acid, and glycine are an exception since they play a dual role as osmolytes and neurotransmitters, and, therefore, changes in their concentration at the extracellular space disturb nervous excitability or lead to neuronal death by excitotoxicity (discussed later). Taurine is particularly suitable for an osmolyte role since it is largely free in the cytosol, is not a protein amino acid, and participates in few reactions in the cell (Pasantes-Morales et al., 1998). Compared with other osmolytes, taurine shows the largest efflux (9- to 22-fold) and the lowest osmolarity threshold (Tuz et al., 2001), which could reflect a higher efficiency of the taurine efflux pathway or more availability of taurine pools for release compared with osmolytes involved in other cell functions.
The decrease in intracellular concentration of organic osmolytes as part of RVD is accomplished largely by solute extrusion rather than by molecular degradation. Efflux of organic osmolytes occurs via energy-independent, bidirectional leak pathways with net solute movements driven by the concentration gradient (Hoffmann and Lambert, 1983; Sánchez-Olea et al., 1991). It has been reported consistently that organic osmolyte fluxes are greatly reduced by VRAC blockers, but it is still not known whether Cl– and organic osmolytes permeate across a common pathway, most likely an anion channel like VRAC. The pore size of VRAC is sufficiently large to allow such osmolytes as taurine and glutamate to permeate, and indeed currents carried by glutamate and anionic taurine across VRAC have been demonstrated (Banderali and Roy, 1992). However, taurine is an electroneutral zwitterion and at physiologic pH, has no net charge; therefore, it is unable to permeate across an anion channel. The same is true for polyols and most organic osmolytes. A number of differences between VRAC and the osmosensitive taurine efflux pathway have been documented (Lambert and Hoffmann, 1994) and discussed in detail in reviews by Shennan (2008) and Hoffmann et al. (2009). A most noticeable difference is the time course of Cl– (traced by 125I–) and taurine efflux in various cell types (Pasantes-Morales et al., 1994; Stutzin et al., 1999) (Fig. 2); Studies showed that I– efflux is fast and rapidly inactivating, whereas taurine efflux is slower and sustained (Fig. 2). As clearly underlined in Shennan (2008) and Stutzin et al. (1999), if taurine and Cl– hypotonic fluxes were occurring via an identical pathway, such large differences in the time course should not be observed. Though similar to VRAC-mediated currents, isotonic taurine efflux is evoked by a decrease in ionic strength (Cardin et al., 1999; Guizouarn and Motais, 1999). The recent discovery of the LRRC8 family of proteins forming volume-sensitive heteromeric channels with different permeability offers a real possibility to establish the molecular identity of the organic osmolyte efflux pathway(s). Studies by Qiu et al. (2014), Voss et al. (2014), and Planells-Cases et al. (2015) show that in LRRC8A−/− HEK, HCT116, and HeLa cells, the osmosensitive taurine efflux is essentially abolished, stressing the importance of this LCRR8 family isoform for the taurine efflux pathway, which is similar to VRAC (Qiu et al., 2014; Voss et al., 2014). Interestingly, also in 2014, a study by Hyzinski-García et al. (2014) showed similar results in cultured astrocytes from the mouse brain cortex. The relevance of this study is that astrocytes from primary cultures are not cell lines but originate from unmodified tissue. In this respect it is also noteworthy that LCRR8A is present in a variety of mouse tissues (Qiu et al., 2014). The effects of genetic ablation of the different LCRR8 family isoforms on IClswell and on taurine efflux shown in studies by Voss et al. (2014) and Planells-Cases et al. (2015) revealed that deletion of LCRR8A abolished both IClswell and taurine, deletion of LCRR8E and C reduced IClswell (taurine efflux was not examined), and deletion of LCRR8D did not affect IClswell but markedly decreased the hypotonic taurine efflux (Planells-Cases et al., 2015). Comparison of results in LRRC8(B,D,E)−/− and LRRC8(B,C,D)−/− cells confirmed that LRRC8D is dispensable for IClswell but indispensable for taurine fluxes. As predicted by Stutzin et al. (1999) and Shennan (2008), these results indicate that the combination of LCRR8 heteromers, particularly the presence or absence of the LCRRL8D subunit, confers two different molecular entities. In fact, the LCRRL8D subunit, or lack of a subunit, confers two different channels, one for permeation of anions and the other with high permeability to a neutral molecule such as taurine. Although both channels have LCRR8A as an integral element and are activated by swelling/ion strength reduction, these channels exhibit different intrinsic properties, such as inactivation and permeability, suggesting that the pore structure must be significantly different in these two molecular entities; a larger pore and differences in the amino acid residues between different combinations of LCRR8 subunits are necessary to permit the passage of molecules like taurine and other neutral organic osmolytes. The predictable existence of a spectrum of channels constructed by multiple combinations of LCRR8 subunits is relevant to understanding the puzzling existence of permeability pathways for organic osmolytes that are not anions and have dissimilar molecular structure but are still gated by reduced ionic strength and inhibited by VRAC blockers.
The variety of LCRR8-based channels specific to permeate different osmolyte groups, including Cl–, represent a useful tool to estimate the contribution of these different osmolyte pathways to cell-volume regulation. Presently, it has been shown that in LCRR8A−/− HEK293, HeLa, and HCT116 cells, RVD is attenuated, and in the cancer cell line BHY, RVD is abolished (Qiu et al., 2014; Voss et al., 2014; Sirianant et al., 2016). These likely reflect differences in the contribution to the regulatory process of other LCRR8A-independent mechanisms, such as the K+/Cl− cotransporter KCC. Parallel analysis of the effect of LCRR8A genetic ablation, together with blockers of KCCs as in HeLa cells in the study by Sirianant et al. (2016), would help to clarify this question.
Transient Receptor Potential Channels.
Transient receptor potential (TRP) channels are widely expressed in vertebrate tissues and are cellular sensors for a large variety of stimuli. Most of the channels are nonselective cation channels, with the exception of a few members that are Ca2+-selective. TRPs of the mammalian superfamily are grouped into several subfamilies by sequence similarity (reviewed in Pedersen et al., 2005; Owsianik et al., 2006; Nilius and Owsianik, 2011; Nilius and Szallasi, 2014). A member of the vanilloid subfamily, the TRPV4 channel is an osmosensitive channel that contributes to RVD in numerous cell types (Vriens et al., 2004; Plant, 2014), including chondrocytes (Lewis et al., 2011.), corneal epithelial cells (Pan et al., 2008), keratinocyte cell lines (Becker et al., 2005), salivary gland cells (Aure et al., 2010), and airway epithelial cells (Arniges et al., 2004). In these cell types, RVD is Ca2+-dependent and TRPV4 is proposed as the pathway for Ca2+ influx evoked by swelling and the ensuing activation of KCa channels involved in RVD (Fernández-Fernández et al., 2002, 2008; Jin et al., 2012). In astrocytes and retinal Muller cells, TRPV4 is functionally linked to AQP4 (Benfenati et al., 2011; Jo et al., 2015), and this association is a requirement for RVD. TRPV4 associates with AQP5 in acinar and salivary gland cells (Liu et al., 2006; Hosoi, 2016) and with AQP2 in renal cells (Galizia et al., 2008).
The activation of TRPV4 by cell swelling is still unclear (Pedersen and Nilius, 2007), but the following possibilities have been proposed: 1) direct membrane stretch with integrins as intermediate molecule (Matthews et al., 2010); 2) interaction with supramolecular complexes containing regulatory kinases and cytoskeletal proteins (Becker et al., 2009); 3) interaction with the arachidonic acid metabolite 5,6-EET (Watanabe et al., 2003); and 4) phosphorylation by WNKs, particularly WNK4 (Fu et al., 2006). Results of TRPV4 genetic manipulation suggest that its role is not restricted to cellular processes of volume regulation, but that it is also implicated in systemic osmosensing; TRPV4 genetic ablation alters functions, such as drinking behavior, serum osmolarity, and antidiuretic hormone plasma levels (Liedtke and Friedman, 2003). Osmosensing neurons and cells in kidney epithelium express TRPV4 and may represent its site of influence on systemic osmoregulation (Ciura and Bourque, 2006; Cohen, 2007).
Other members of the transient receptor potential channels family, such as TRPV1, TRPV2, TRPC1, TRPC5 and TRPC6, TRPM3, TRPM7 are possibly involved in RVD; these channels are osmotically activated channels, but detailed studies on their role in volume regulation are largely missing (Plant, 2014).
Isovolumetric Regulation
In most animal species, extracellular osmolarity is tightly regulated and abrupt, and large changes in osmolarity rarely occur. In contrast, small and gradual changes in osmolarity associate with a number of metabolic reactions and cell functions. Exposing cells to small and gradual osmolarity reductions have shown that cells are able to maintain a constant volume over a wide range of tonicities via an active mechanism of continuous volume adjustment. This adaptive response, known as isovolumetric regulation (IVR), is accomplished by the efflux of intracellular osmolytes (Lohr and Grantham 1986, Mountian and Van Driessche, 1997; Souza et al., 2000). The osmolytes involved in IVR are the same as in RVD, i.e., Cl–, K+ and organic molecules. Fig. 3 illustrates as an example the Cl– currents in glioma cells evoked by gradual or sudden reductions in osmolarity (Fig. 3B). This Cl– current activates early, at osmolarity reductions of about 3% and is blocked by niflumic acid and NPPB (Fig. 3B), which suggests that it’s similar to the current across VRAC. This point should be clarified in light of new findings regarding the LCRR8 family members as integral elements of VRAC. K+ conductance is also activated in these cells during IVR with different osmolarity thresholds in the presence or absence of Ca2+ (Ordaz et al., 2004). In a variety of cell types, gradual changes in osmolarity activate taurine efflux with different efflux threshold (Pasantes-Morales et al., 2000; Souza et al., 2000; Ordaz et al., 2004). The lowest efflux is observed in cerebellar granule neurons where taurine efflux increases significantly at osmolarity reductions of only 2 mOsm (Fig. 3A) (Tuz et al., 2001). This compensatory mechanism is then likely to be preferred by neurons facing physiologic changes in cell volume.
Channel-Transporter Interactions in Regulatory Volume Increase
Increase in external osmolarity activates intracellular water exit driven by the osmotic disequilibrium, and, accordingly, this leads to a decrease in cell volume. If this is not corrected, cells undergo apoptotic death. In healthy cells, hypertonicity activates a regulatory mechanism that aims to restore normal cell volume even if the external anisotonic condition persists. This adaptive mechanism, known as regulatory volume increase (RVI) (Burg et al., 2007), occurs when mechanisms that increase the concentration of intracellular osmolytes are set in motion. This tends to re-establish the osmotic equilibrium between extracellular and intracellular compartments. Electroneutral cotransporters and ion exchangers both participate in this process to increase the cytosolic levels of Na+ and Cl– (Fig. 4), and NCC and Na+/K+/2Cl cotransporter (NKCC) family cotransporters (Arroyo et al., 2013) and Na+/H exchangers are the main contributors to RVI (Cala, 1980; Alexander and Grinstein 2006). Organic osmolytes are also significantly involved in RVI (Fig. 4); cell shrinkage increases the expression of Na+-dependent cotransporters known to operate for various molecules acting as organic osmolytes (Sánchez-Olea et al., 1992; Burg and Ferraris, 2008). In contrast to RVD, in which the corrective fluxes of osmolytes permeate essentially across ion channels and diffuse pathways for organic osmolytes, in RVI, channels play only a minor role, and the regulatory process is essentially accomplished by cotransporters and exchangers. It has been proposed that unidentified hypertonicity-activated cation channels participate in RVI (Wehner et al., 2006). A study in HeLa cells (Wehner et al., 2006) showing RVI reduction by SKF-96365, a blocker of TRP channels, raised the question of whether the hypertonicity-induced cation channels are actually TRP channels, and recent evidence pointed to TRPM2 as the channel involved. A recent report, also in HeLa cells (Numata et al., 2012), showed that nucleotides adenosine diphosphate ribose (ADPr) and cyclic ADPr, reported as typical TRPM2 activators (Pedersen et al., 2005; Pedersen and Nilius, 2007), generate cation currents similar to those elicited by hypertonicity, previously known as hypertonicity-activated cation channels. In the same line, small-interfering RNA silencing of TRPM2 expression or the precursor enzyme of the TRPM2 nucleotide activators abolish the hypertonicity-elicited cation current and mildly reduce RVI. Cloning of TRPM2 identified the ΔC–splice variant as the molecular entity corresponding to the nucleotide-activated current (Numata et al., 2012), which suggests that TRPM2 may represent the molecular identity of the hypertonicity-activated cation channels. Another member of the TRP channel family, TRPV2, seems to participate in RVI via an interesting connection with the electroneutral cotransporter NKCC, which as previously mentioned is a main effector in RVI. The suggested mechanistic chain of events relating TRPV2 channels and RVI involves a TRPV2-dependent accelerated depolarization followed by Ca2+ release from intracellular sources and the subsequent phosphorylation of STE20/SPS1-related proline/alanine-rich kinases (SPAKs), which is essential for activation of NKCC. This proposed pathway is supported by studies in skeletal muscle showing that expression of TRPV2 negative dominant reduces the RVI efficiency at the time that depolarization is impaired and the Ca2+ response is diminished (Zanou et al., 2015). Interestingly, the upregulation of all transporters involved in RVI, ions and organic osmolyte transporters, occurs by the concerted action of transcription factor tonicity-responsive enhancer binding protein/osmotic response element–binding protein. As part of the regulatory response to cell volume decrease, tonicity-responsive enhancer binding protein/ osmotic response element–binding protein increases the expression of osmoprotective genes, including Na+-dependent transporters, responsible for the accumulation of organic osmolytes (Ferraris and Burg, 2006).
Channels Involved in Volume Changes in Necrotic and Apoptotic Death
Changes in cell volume are characteristic features of cell death and represent one of the main differences between necrosis and apoptosis. In necrotic death, cell swelling and depletion of intracellular ATP are distinctive traits. During the necrotic process, water accumulates in the cytosol and organelles, and the membrane surface architecture is deformed by prominent, stationary blebs that eventually lead to membrane rupture and cell death (Jurkowitz-Alexander et al., 1992; Barros et al., 2001). In contrast, in apoptosis, reduction in cell volume is a central event and a hallmark in this program of cell death (Bortner and Cidlowski, 1998; Yu and Choi, 2000). Coined phrases for these cell volume changes include “apoptotic volume decrease” (Hughes et al., 1997; Maeno et al., 2000) and “necrotic volume increase,” also known as oncotic swelling or cytotoxic swelling. A variety of channels are involved in these changes in cell volume.
Necrotic Volume Increase.
Necrosis associates with the extensive cell loss that accompanies pathologic conditions, such as ischemia and ischemia/reperfusion, cardiovascular diseases, and several neurodegenerative conditions. Cytotoxic swelling in ischemia results from the arrest of oxygen-dependent ATP synthesis, which reduces the activity of the Na+/K+ ATPase, impedes the transmembrane Na+/K+ exchanges, and results in dissipation of ion gradients. In the brain, swelling during ischemia is the first step in a cascade of ion gradient disturbances between the intravascular and extracellular compartments that culminates in blood brain barrier injury and vasogenic edema (Annunziato et al., 2013, Kahle et al., 2015). Besides ischemic swelling, brain cells also swell during trauma, seizures, and spreading depression. Swelling is essentially owing to Na+, K+ and Cl– overload, which drives inwardly-directed water fluxes, and both channels and cotransporters participate in this process (Fig. 5). Characteristic of brain ischemia is the marked increase in extracellular K+ (K+o), which is often elevated from a physiologic level of around 3 mM up to 60 mM. This high K+o level reverses the KCC direction, and—together with the ischemia-activated expression of NKCC—results in large inward ion and water fluxes (Kahle et al., 2015). Na+ influx also permeates across a nonspecific cation channel, which, on the basis of recent evidence, is identified as transient receptor potential melastatin-4 channel (TRPM4). The TRPM4 pore is selective for monovalent cations with similar permeability for K+ and Na+ and is impermeable to divalent cations. Evidence in support of TRPM4 as a major mechanism for Na+ influx in ischemia includes the following observations: 1) Two main regulators of TRPM4 (i.e., ATP and intracellular Ca2+) are deeply disturbed by conditions promoting necrosis. This deregulation favors the opening of TRPM4, Na+ overload and NVI. 2) Redox imbalance by excessive generation of free radicals, a condition related to pathologies leading to necrosis, induces a sustained activity of TRPM4 (Simon et al., 2010). 3) Pharmacological or genetic manipulation of TRPM4 has a marked influence on the swelling phase of the necrotic process. 4) activation of TRPM4 channel by ATP depletion in COS-7 cells leads to depolarization and progressive bleb formation (Chen and Simard, 2001; Chen et al., 2003), but this is not observed in cells lacking the channel (Simard et al., 2012). 5) Induced upregulation of TRPM4 protein in endothelial cells from human umbilical vein causes Na+ overload, cell volume increase, and necrotic death, effects which are prevented by pharmacologic inhibition of the channel (Gerzanich et al., 2009; Becerra et al., 2011). It should be noted that ADP and AMP are reported to block TRPM4b currents (Nilius et al., 2004); however, during the first minutes of ischemia, ADP levels decrease in parallel to the increase in ATP and AMP levels (Guarnieri et al., 1993; Phillis et al.,1995). Regardless, the concentration of AMP is so low (3% and 5% of ATP levels in heart and brain, respectively) (Guarnieri et al., 1993; Phillis et al.,1995) that the increase may not affect TRPM4 activation. TRPM4 is not constitutively present in the central nervous system but is transcriptionally upregulated in neurons, astrocytes, oligodendrocytes, and microvascular endothelial cells after the onset of ischemia (Gerzanich et al., 2009). Necrotic death of endothelial cells caused by TRPM4 activation is also responsible for the capillary fragmentation, known as progressive hemorrhagic necrosis, observed in traumatic injury of the spinal cord. The capillary damage is prevented in TRPM4−/− or by antisense oligodeoxynucleotides directed against TRPM4 (Gerzanich et al., 2009). TRPM4 channel also contributes to cytotoxic swelling, cell death, blood-brain barrier breakdown, and vasogenic edema in ischemia and trauma (Simard et al., 2006; Zweckberger et al., 2014; Martinez-Valverde et al., 2015). Together, these findings are consistent with the idea that the TRPM4 channel plays a role in cell swelling and necrotic death. Other channels, including members of the TRP family (TRPV4, TRPM2, and TRPM7) and the acid-activated cation channels, also increase in expression and functional activity after an ischemic episode, and their blockade reduces the size of the infarct. However, their role in inducing ischemic injury is more related to Ca2+ overload and protease activation than to cytotoxic swelling.
In brain cells, VRAC participates in cytotoxic swelling and ulterior necrotic neuronal death, particularly in the penumbra where intracellular ATP that is necessary for VRAC activation still remains. Neuronal death results from the increase in extracellular glutamate driven by the reversal of Na+-energy-dependent transporters and the swelling-activated, VRAC-mediated glutamate efflux from swollen astrocytes (Mongin, 2016). The consequent overactivation of AMPA and NMDA ionotropic receptors causes Na+ and Ca2+ overload at the same time that the ischemic depolarizing condition activates VRAC in neurons, moving Cl– into the cell driven by its electrochemical potential. All this culminates in NVI and a necrotic cascade. The effect of VRAC blockers on reducing both swelling and neuronal death (Inoue et al., 2007; Mongin, 2016) suggests that VRAC has a role in this deleterious effect. Necrosis is a delayed pattern of excitotoxicity, but NMDA receptor overfunction also activates the apoptotic machinery (Linnik et al., 1993).
Hemichannels, the constituent elements of the Gap junctions, allow transmembrane movements of large groups of heterogeneous molecules of different sizes, such as water, Na+, K+, Ca2+, ATP, or adenosine, among others. Hemichannels are expressed in the membrane of astrocytes, neurons, oligodendrocytes, and vascular endothelium. Unregulated hemichannels have been implicated in hypoxia, ischemia, and brain trauma, where they generate disruption of intracellular ionic homeostasis and cytotoxic swelling. Consistent with this action, downregulation of hemichannels reduce cell swelling in astrocytes and reduce the ischemic injury (Davidson et al., 2015).
Apoptotic Volume Decrease.
Apoptosis, or programmed cell death, is a physiologic mechanism that commits cells to individual death fate (Kerr et al., 1972). Apoptosis occurs normally during development and aging as a homeostatic mechanism to maintain optimal cell populations and eliminate excessive or potentially harmful cells. Apoptosis is characterized by changes in the cell structure, notably cell nuclear condensation and DNA fragmentation, membrane blebbing, and the formation of apoptotic bodies. Programmed cell death is activated by either intrinsic or extrinsic stimuli. Intrinsic stimuli are molecules or situations of stress that result in mitochondrial dysfunction, like UV radiation, nitric oxide, staurosporine, thapsigargin, and dexamethasone, whereas extrinsic stimuli are ligands of death receptors such as Fas or other members of the tumor necrosis factor receptor superfamily, including CD95. As mentioned earlier, cell volume decrease is a characteristic trait of apoptotic death, and depending on the cell type, the decrease in cell volume may be in the range of 40–80%. The term “apoptotic volume decrease” was coined for this unique condition (Hughes et al., 1997; Maeno et al., 2000). In terms of time-course, apoptotic volume decrease (AVD) occurs in two phases, an initial phase that starts 0.5–2 hours after exposure to the inductor and a late phase detected about 3 hours later. The two phases of AVD are accomplished by the same mechanism that includes outward fluxes of K+ and Cl– moving across specific channels and a decrease in osmolyte intracellular concentration and water outflow resulting in AVD (Kondratskyi et al., 2015). It has yet to be determined whether this decrease in cell volume is a passive factor or an active signal in the induction of apoptosis, but evidence does exist that points to an explicit role for volume decrease in the apoptotic process; blockade of the main osmolyte pathways that reduce AVD impairs the progression of apoptosis (Yu and Choi, 2000; Grishin et al., 2005).
The increase in K+ currents that were detected as an early event in the cell death program (Yu et al., 1997) first suggested that K+ channels contribute to AVD. Regardless of the cell type or the inductor stimulus, the role of K+ channels as the K+ efflux pathway leading to AVD and loss of intracellular potassium is recognized as a distinctive feature in the apoptotic program, (reviewed in Burg et al., 2006; Bortner and Cidlowski, 2007; Orlov et al., 2013). A diversity of K+ channels mediate the K+ efflux during AVD in different cell types; voltage-gated, Ca2+-activated of intermediate and large-conductance K+ channels, inward rectifier, and two-pore–domain K+ channels participate in AVD. Different subtypes of Kv channels contribute to AVD, including Kv 1.2 and Kv 2.1 in neurons, Kv 1.3 in T-lymphocytes, and Kv1.5 in COS-7 and in vascular smooth muscle. Kv channels of undefined subtypes account for AVD and K+ loss in hippocampal neurons, sympathetic neurons, cerebellar granule neurons, corneal epithelial cells, and cardiomyocytes (reviewed in Remillard and Yuan, 2004; Bortner and Cidlowski, 2007; Lang, 2007). The intermediate-conductance IKCa channels mediate the apoptotic K+ current in lymphocytes, thymocytes, and glioblastoma cells, and the large-conductance BK channels play this role in artery smooth muscle cells, glioma, and superficial colonocytes, whereas inward rectifier IRK channels are involved in apoptosis in liver cells, HeLa cells, and some neuronal cell lines (reviewed in Bortner and Cidlowski, 2007). Members of two-pore–domain potassium channels TASK1 and TASK3 participate in apoptosis in cerebellar granule neurons and in exocrine pancreatic cells (Patel and Lazdunski, 2004). Hydrogen peroxide–mediated apoptosis is associated with activation of TWIK-related K+ (TREK) channels and of hERG K+ channels. The significance of intracellular K+ decrease in the apoptotic process is more complicated than simply an involvement as an osmolyte and a main contributor to AVD. The decrease in intracellular K+ levels has per se an effect on apoptotic mechanisms, independent of AVD, and is more related to the release of cytochrome C (Yu, 2003; Remillard and Yuan, 2004). Various K+ channel subtypes localized at the mitochondrial membrane regulate mitochondrial volume and contribute to the organelle homeostasis.
Together with K+, Cl– efflux is an essential component of AVD, during which K+ efflux hyperpolarizes the cell and is followed by a Cl– outward movement directed by its electrochemical gradient, which maintains the ionic balance and electroneutrality of the cell. Activation of anion currents facing intrinsic and extrinsic apoptotic stimuli is observed in a variety of cells (Okada et al., 2006), and these anion currents are similar to those carried by VRAC regarding outward rectification, ATP dependence, and pharmacological profile. However, VRAC is activated by a swelling and ionic strength reduction, whereas volume decrease and ionic strength increase characterizes apoptosis. Therefore, if VRAC participates in AVD, the channel volume set point must shift to a lower level or the channel is gated by other mechanisms. Overproduction of reactive oxygen species, or a change in cytosolic ATP, both concurrent with AVD, may be signals to either reduce the set point or to modify a hypothetic modulator (Shimizu et al., 2008). Although this critical point is not as yet clarified, there is evidence showing that AVD induction is reduced by VRAC blockers in various cell types. Although it should be considered that those blockers are not specific for VRAC or for Cl– channels. The recent findings identifying LCCR8A as a structural element of VRAC provides additional tools to define the VRAC contribution to AVD. The study by Planells-Cases et al. (2015) in KBM7 and HAP1 cells showed that a channel formed by the combination of LCRR8A and LCRR8D subunits participates in AVD and apoptosis. This channel composition is similar to that of the channel proposed to allow taurine to permeate, and dissimilar to the typical VRAC. In contrast, in HeLa cells, Sirianant et al. (2016) exhibited no effect of LCRR8A knockdown in staurosporin-induced cell shrinkage.
The significant role of organic osmolytes in cell volume regulation either in RVD and in RVI is well documented, as discussed earlier. In contrast, evidence regarding organic osmolyte contribution to volume decrease during apoptosis is rather scarce (Wehner et al., 2003). However, the importance of organic osmolytes was emphasized in the quantitative analysis made by Model (2014), which demonstrated that the loss of organic molecules is necessary to account for the osmotically obligated water exit in AVD. Taurine cell loss is documented in apoptosis in Jurkat lymphocytes (Lang et al., 1998) and in cerebellar granule neurons (Morán et al., 2000). Similar to VRAC, taurine efflux is evoked by cell swelling or ionic strength reduction, two conditions absent in apoptosis. A number of studies report antiapoptotic effects of taurine (Lambert et al., 2015), but this seems unrelated to AVD. Recent findings that demonstrate resistance to cancer chemotherapy by a platinum-based drug, attenuated taurine efflux and reduced expression of LCRR8D, suggest a role for the LRRC8D-containing VRAC entity in drug extrusion (Planells-Cases et al., 2015; Voets et al., 2015). As for apoptosis, the normotonic activation of VRAC under these conditions remains to be defined.
Comparing RVD and AVD mechanisms demonstrates that essentially the same type of K+ channels participate in both processes in the same cell type. For instance, voltage-dependent Kv channels are those involved in both RVD and AVD in neurons, whereas Ca2+-activated K+ channels are implicated in epithelial cells. This selectivity is explained in the thermodynamic analysis of ion fluxes described in Orlov et al. (2013). The analysis in Orlov’s review refers only to AVD, but appears valid also for K+ and Cl– flux kinetics in RVD. Interestingly, activation of channels in AVD occurs under isotonic conditions, suggesting marked differences in the intracellular signals responsible for the efflux of osmolytes during the two processes. Interesting as well is the fact that the sustained cell shrinkage during AVD does not result in activation of RVI (Maeno et al., 2006) in cisplatin‐induced cell shrinkage and subsequent apoptotic cell death
Concluding Remarks
The key role of ion channels together with cotransporters in the active processes of cell volume regulation is now well established. The intricate and complex signaling pathways directed to activate/inactivate these channels represents a vast field for present and future research. The firm establishment of the molecular identity of the volume-regulated anion channel and the pore(s) for permeation of organic osmolytes are also intriguing aspects in the physiology of volume-related channels. Identification of the mechanisms responsible for cytotoxic swelling may ideally contribute to the formulation of drugs with the potential to prevent the chain of events disturbing the ionic homeostasis that ultimately result in brain edema. Although extensive literature has reported drugs and maneuvers directed to reduce swelling and injury resulting from ischemic episodes, all have failed in clinical trials. There is also strong evidence, however, of their relevance in the ion-gradient disturbances generating cytotoxic and vasogenic brain edema. The ischemic phenomenon is complex and involves a plethora of factors and signals, complicated further by those resulting from reperfusion. A comprehensive understanding of the events concurrent with ischemia/superfusion, including those related to the channels involved, is part of the integral approach necessary to formulate effective therapeutic targets for this pathology.
Acknowledgments
The author thanks Dr. Gerardo Ramos-Mandujano for invaluable help in the preparation of the manuscript and figures. ASPET thanks Dr. Katie Strong for copyediting of this article.
Authorship Contributions
Wrote or contributed to the writing of the manuscript: Pasantes-Morales.
Footnotes
- Received March 3, 2016.
- Accepted June 22, 2016.
Work in the author’s laboratory is supported by Dirección General de Asuntos del Personal Académico (DGAPA) at the Universidad Nacional Autónoma de México (UNAM) [Grant No. PAPIIT IN205916 ].
Abbreviations
- AMPA
- α-Amino-3-hydroxy-5-methylisoxazole-4-propionic acid hydrate
- AQP4
- aquaporin 4
- AVD
- apoptotic volume decrease
- BK
- big potassium
- DCPIB
- 4-[(2-butyl-6,7-dichloro-2-cyclopentyl-2,3-dihydro-1-oxo-1H-inden-5-yl)oxy]butanoic acid
- DIDS
- 4,4′-diisothiocyanostilbene-2,2′-disulfonic acid
- IKCa
- intermediate-conductance (channels)
- IVR
- isovolumetric regulation
- KCa
- Ca2+-activated K+ (channels)
- KCC
- K+/Cl− cotransporter
- KIR
- inwardly rectifying K+ channels
- K2P
- two-pore–domain K+ channels
- Kv
- voltage-gated K+ (channels)
- NKCC
- Na+/K+/2Cl cotransporter
- NMDA
- N-methyl-d-aspartic acid
- NPPB
- 5-nitro-2-(3-phenylpropylamino) benzoic acid
- RVD
- regulatory volume decrease
- RVI
- regulatory volume increase
- SITS
- 4-acetamido-4′-isothiocyanatostilbene-2, 2′-disulfonic acid
- TBOA
- dl-threo-β-benzyloxyaspartic-acid
- TRP
- transient receptor potential
- TRPM4
- transient receptor potential melastatin-4 channel
- VRAC
- volume-regulated anion channel
- Copyright © 2016 by The American Society for Pharmacology and Experimental Therapeutics