Mutations in two kinases of the WNK family, PRKWNK1 and PRKWNK4, cause a Mendelian disease featuring hypertension with hyperkalemia (elevated serum K⁺) (79). Subsequent studies revealed that WNK1 and WNK4 are components of a novel signaling pathway that determines the balance between NaCl reabsorption and K⁺ secretion in the kidney’s distal nephron via regulation of different ion transporters and channels (Refs. 32, 80, 87; for review, see Ref. 20). Recent work suggests that WNK kinases might also play a role in the regulation of ion transport outside the kidney. Na-K-Cl and K-Cl cotransporters, two different branches of the SLC12 gene family, are molecular mediators of cellular Cl⁻ entry and Cl⁻ exit, respectively, in cells throughout the body; their coordinated activities are necessary for the regulation of neuronal excitability, the homeostasis of blood pressure, and the maintenance of cell volume (21, 28). K-Cl cotransporters transport Cl⁻ out of cells, are inhibited by phosphorylation, and are activated by dephosphorylation; Na-K-Cl cotransporters mediate Cl⁻ influx and show the opposite regulation (27, 39). Because of their reciprocal regulation by the same stimuli, researchers have postulated that a common “Cl⁻-/volume-sensitive regulatory kinase” regulates the Na-K-Cl and K-Cl cotransporters (45–47). However, the identity of this kinase, the associated components of its signaling pathway, and the mechanism of its action have remained elusive. Evidence now suggests WNK kinases are components of a signaling pathway that regulates the Cl⁻influx and efflux branches of the SLC12 family.

Roles of SLC12 Cation/ Cl⁻ Cotransporters

The cation/Cl⁻ cotransporter gene family (solute carrier family 12,SLC12) comprises intrinsic membrane proteins that transport Cl⁻ ions, together with Na⁺ and/or K⁺, across plasma membranes (FIGURE 1A; Ref. 20). Driven by electrochemical gradients established by the Na⁺-K⁺-ATPase, they mediate secondary active transport. A balanced stoichiometry exists between transported cations and Cl⁻ so that transport is electroneutral. Members of the SLC12 gene family are divided into two branches defined by their stoichiometry of transported ions, sensitivity to inhibitors, and phylogeny. SLC12 cotransporters are targets of therapeutic drugs and are mutated in several human diseases (11).

The Na-K-Cl cotransporters facilitate the coupled movement of Na⁺ and Cl⁻, with or without K⁺, into cells. Two genes encode Na-K-Cl cotransporters. NKCC1 (Na, K, Cl cotransporter; SLC12A2) is ubiquitously expressed and plays a “housekeeping” role in volume regulation during hypertonicity, in addition to mediating basolateral Cl⁻ entry in Cl⁻-secreting epithelia (27). NKCC2 (SLC12A1) is expressed exclusively on the apical membrane of the thick ascending limb in the kidney and is the target of loop diuretics (e.g., furosemide) (37). The third gene encodes the apical Na-Cl cotransporter (NCC; SLC12A3) that is expressed almost exclusively in the renal distal convoluted tubule (19); NCC is the target of thiazide diuretics. Loss-of-function mutations in NKCC2 and NCC, respectively, cause Bartter’s syndrome type I [Online Mendelian Inheritance in Man (OMIM) no. 601678] and Gitelman’s syndrome (OMIM no. 263800), inherited diseases featuring low blood pressure due to renal loss of NaCl and hypokalemic alkalosis due to increased activity of the reninangiotensin system and consequent increased electro-genic reabsorption of Na⁺ via ENaC accompanied by increased secretion of K⁺ and H⁺ (68, 69). As expected from its wide expression, mice with targeted deletion of NKCC1 suffer impairments in multiple systems, including hearing, pain perception, salivation, spermatogenesis, and the control of extracellular fluid volume (9, 11). NKCC1 null mice are hypotensive, in part, from the absence of NKCC1-mediated vasoconstriction (14, 51, 78). Recent data reveal that NKCC1 facilitates seizures in the developing brain and that bumetanide (a pharmacological inhibitor of NKCC1) might be a novel, efficacious drug for the treatment of neonatal seizures (13).

The K-Cl cotransporters (KCC1–KCC4; encoded by SLC12A4–7) transport Cl⁻ coupled with K⁺ and move Clout of cells. KCC1 is ubiquitously expressed with a housekeeping role in cellular volume regulation during hypo-tonicity (22). KCC2 expression is neuronal specific (56); KCC2 knockout mice have a profound seizure disorder and exhibit anxiety-like behavior due to deranged GABA signaling (31, 73, 87). KCC3 is also highly expressed in neurons, in addition to all segments of the nephron (49). In humans, mutations in KCC3 cause Andermann’s syndrome (OMIM no. 218000), a severe inherited peripheral neuropathy associated with agenesis of the corpus callo-sum (30). Mice with targeted deletion of KCC3 display neurodegeneration of the peripheral and central nervous system (CNS), in addition to arterial hypertension (3, 65). KCC4 is expressed ubiquitously but has been localized in the nephron, inner ear, and various regions of the CNS (2, 53); mice null for KCC4 develop distal renal tubular acidosis and deafness (2).

Na-K-Cl and K-Cl Cotransport is Regulated by an Unknown Volume/Cl⁻-Sensitive Kinase

The coordinated function of Na-K-Cl and K-Cl cotransport is necessary in numerous physiological processes. One example is the process by which cells defend themselves against changes in cell volume (38). The abilities to sense and respond appropriately to changes in cell volume are key homeostatic mechanisms for cells that do not have cell walls. Cell volume is maintained by a system that can rapidly raise or lower the intracellular Cl⁻ concentration ([Cl⁻]_i); this can be achieved by altering the balance between Cl⁻ influx and efflux across cell membranes. An important mediator of Cl⁻ entry is NKCC1; Cl⁻ exit is largely mediated by K-Cl cotransporters. In hypertonic extracellular fluid, water leaves cells, and cells shrink. During the process of regulatory volume increase, cell volume is rapidly reestablished, in part, by the coordinated activation of NKCC1 and inhibition of K-Cl cotransporters; this increases net K⁺ and Cl⁻ entry into cells, thereby raising cell osmolarity and driving re-uptake of water. Conversely, in hypotonic extracellular fluid, water enters cells, and cells swell. During the process of regulatory volume decrease, the coordinated inhibition of NKCC1 and activation of K-Cl cotransporters helps re-establish normal cell volume by increasing net K⁺ and Cl⁻ exit out of cells, thereby lowering cell osmolarity and driving water loss.

Coordination of Na-K-Cl and K-Cl cotransport is also important in the control of neuronal excitability (10, 57). Excitable cells like neurons experience constant changes in their membrane potential due to ion flux through plasma membrane channels. Transmembrane concentrations of cations are maintained via Na⁺-K⁺-ATPase activity. During the spread of action potentials, the concentration of Cl⁻ is also under constant challenge from membrane potential changes. Neurotransmitters such as GABA and glycine act via opening of Cl⁻ channels; as a consequence, the resting level of intracellular Cl⁻ is a key determinant of the response to these neurotransmitters. If [Cl⁻]_i is below its equilibrium potential, Cl⁻ enters the cell, resulting in hyperpolarization and inhibition. If [Cl⁻]_i is above its equilibrium potential, GABA signaling results in Cl⁻efflux, depolarization, and neuronal excitation. [Cl⁻]_i, and hence the neuronal response to GABA, is largely determined by a balance between Cl⁻ efflux and Cl⁻influx through KCC2 and NKCC1, respectively (10, 57, 63). Early in development, NKCC1 activity predominates over KCC2, raising [Cl⁻]_i and rendering GABA an excitatory neurotransmitter. With postnatal development, KCC2 activity increases, and NKCC1 activity decreases, lowering [Cl⁻]_i and rendering GABA an inhibitory neurotransmitter (10, 57). Excitatory GABA signaling occurs widely in the brain in the neonatal period, with circadian rhythm in many different brain centers, and in the peripheral nervous system. GABA-generated inhibitory inputs are essential in maintaining appropriate electrical activity in the brain, balancing signals generated by excitatory neurotransmitters (e.g., glutamate), and thus preventing the spread of excitatory activity. Derangement in the balance of the activities of NKCC1 and KCC2 results in hypo- or hyperexcitability of neurons due to altered homeostasis of intracellular Cl⁻ (10, 48, 57). This alteration results phenotypically in epilepsy, altered pain perception, peripheral neuropathies, or mood disturbances (10, 57).

These examples show the importance of Na-K-Cl and K-Cl cotransport in setting intracellular Cl⁻ activity and its consequent physiological effect. Recent studies have begun to reveal the molecular components of the mechanism responsible for the synchronized regulation of Cl⁻movement into and out of cells. SLC12 cotransporter activities are known to be regulated by phosphorylation and dephosphorylation. Dephosphorylation activates K-Cl cotransporters and inhibits Na-K-Cl cotrans-porters, whereas phosphorylation has the opposite effects (1, 15, 21, 39). Specifically, cell swelling, high intracellular Cl⁻, and protein phosphatases, by promoting cotransporter dephosphorylation, stimulate K-Cl cotransporters but inhibit Na-K-Cl cotransporters, resulting in decreased intracellular Cl⁻. In contrast, cell shrinkage, low intracellular Cl⁻, and protein phosphatase inhibitors, by promoting cotransporter phosphorylation, activate Na-K-Cl cotransporters but inhibit K-Cl cotransporters, thus resulting in increased intracellular Cl⁻. Researchers have long postulated that a common Cl⁻/volume-sensitive regulatory kinase mediates the regulation of both the Na-K-Cl and K-Cl cotransporters by promoting the reciprocal phosphorylation and dephosphorylation of Na-K-Cl and K-Cl cotransporters in response to changes in cell volume/intracellular Cl⁻ (FIGURE 1A; Refs. 45–47). This kinase is likely part of a larger regulatory pathway, comprised of sensors (sensitive to changes in intracellular Cl⁻, cell volume, and/or extracellular tonicity) and transducers (kinases and linked phosphatases) that ultimately result in cotransporter phosphorylation/dephosphorylation.

Although NKCC1 contains known consensus phosphorylation sites for protein kinase C, casein kinase II, and protein kinase A, none of these kinases has been shown to directly phosphorylate and regulate NKCC1 (12). Myosin light chain kinase, c-jun NH₂ kinase, and Rho kinase have modest effects on NKCC1 activation in hyperosmotic (stimulating) conditions, but none of these kinases bypass the extracellular tonicity requirement for NKCC1 activation or reciprocally regulate the two different branches of the SLC12 family (12). Thus the identity of the Cl⁻/volume-sensitive regulatory kinases is still unclear.

WNK Kinases are a Family of Novel Kinases Necessary for Electrolyte Homeostasis

The novel WNK family of serine-threonine kinases is named for having the unique substitution of cysteine for lysine at an almost invariant position in subdomain II of their catalytic core (hence WNK: with no K = lysine; Ref. 81; FIGURE 1B). In humans, the four genes encoding WNK1, WNK2, WNK3, and WNK4 are located on chromosomes 12, 9, X, and 17, respectively (79).

Recent studies have demonstrated that WNK kinases play important roles in the regulation of cation-coupled chloride transport in the kidney. Wilson et al. (79) demonstrated that mutations in either WNK1 or WNK4 cause pseudohypoaldosteronism type II (PHAII; OMIM no. 145260), a rare autosomal-dominant syndrome of hypertension with hyperkalemia and metabolic acidosis. PHAII phenotypes are the mirror image of those in Gitelman’s syndrome (an autosomal recessive disease featuring hypotension with hypokalemia and metabolic alkalosis due to loss-of-function mutations in the SLC12 cotransporter NCC), and PHAII phenotypes are corrected by the administration of thiazide diuretics (drugs that inhibit NCC) (26, 66). Thus, at the time of their discovery, it was proposed that WNK1 and WNK4 were involved in the regulation of NCC in the distal nephron. Indeed, immunolocalization experiments revealed that WNK1 and WNK4 are both most highly expressed in this location (26, 66). Subsequent in vitro experiments have demonstrated that WNK4, via a mechanism dependent on its catalytic activity, is a potent inhibitor of NCC activity by altering the expression of functional cotransporters at the plasma membrane (25, 80, 87, 88). WNK1 appears to lie upstream of the WNK4 signaling pathway, abolishing the inhibitory effect of WNK4 at NCC (87, 88). Different PHAII-causing missense mutations in WNK4 were also shown to abrogate the inhibitory effect of WNK4 on NCC (25, 80, 87). Thus the PHAII phenotype, in part, could be explained by increased NCC activity achieved through overexpression of WNK1 or dysregulation by mutant WNK4. Subsequent studies suggested that, through the coordinated regulation of NCC, a K⁺ channel (ROMK1), and paracellular Cl⁻ flux at the tight junction, WNK4 is an important determinant of the balance between NaCl reabsorption and K⁺ secretion in the kidney’s distal nephron (34, 80, 86, 87).

In addition to the role of WNK kinases in the kidney, a role for their involvement in the regulation of ion transport outside the kidney has been suggested from immunolocalization studies that demonstrate their expression in a wide variety of Cl⁻-transporting epithelia (5, 33); moreover, initial expression studies in Xenopus laevis oocytes revealed that WNK4 regulates the activity of NKCC1 (33), the ubiquitous mediator of Cl⁻ entry.

WNK3 Reciprocally Regulates the Na-K-Cl and K-Cl Cotransporters

The above studies established the key roles of WNK1 and WNK4 in the regulation of NaCl and K⁺ homeostasis and raised the question of whether the related proteins WNK2 and WNK3 might play related or unique roles. WNK3 has recently been shown to coordinately regulate the Na-K-Cl and K-Cl cotransporters by reciprocally altering transporter phosphorylation and dephosphorylation (FIGURE 1A; Refs. 36, 62). Heterologous coexpression of WNK3 with of any of the three Na-K-Cl cotransporters (NCC, NKCC1, and NKCC2) resulted in a potent increase in metolazone-sensitive ²²Na⁺ (NCC) or bumetanide-sensitive ⁸⁶Rb⁺ (NKCC1/2) uptake in oocytes. WNK3 activated NKCC1/2 by increasing the phosphorylation of two amino terminal threonine residues (36, 62); these same residues have been shown previously to be necessary for cotransporter activation (7, 16, 23, 24). These effects of WNK3 were seen even when oocytes were exposed to extracellular hypotonicity, a condition in which NKCC1/2 is normally dephosphorylated and therefore inactivated. In contrast, coexpression of WNK3 with the K-Cl cotransporters KCC1, KCC2, KCC3, or KCC4 resulted in complete inhibition of these cotransporters, even when oocytes were exposed to extracellular hypotonicity, a condition in which K-Cl cotransporters are otherwise dephosphorylated and maximally active. Strikingly, coexpression of these cotransporters with a kinaseinactive WNK3 mutant (KI-WNK3) resulted in the opposite effects: Na-K-Cl cotransporters were dephosphorylated and inhibited, whereas K-Cl cotransporters were dephosphorylated and activated. The stimulation of K-Cl cotransport by KI-WNK3 in hypertonic (inhibiting) conditions was reversed by caly-culin A and cyclosporine A, inhibitors of protein phosphatase 1 and 2B, respectively, revealing that this effect of WNK3 is dependent on its functional interaction with specific protein phosphatases shown previously to be required for K-Cl cotransporter activation (8).

The localization of WNK3 to cells in tissues that express members of the SLC12 family that are known to modulate the balance between Cl⁻ entry and exit suggests the in vivo relevance of the regulatory activities of WNK3 (36, 62). WNK3 is coexpressed with NKCC1/KCC2 in hippocampal, cerebellar, and cortical GABAergic neurons, with NKCC1/KCC1 in Cl⁻-transporting extrarenal epithelia and with KCC3/4, NCC, and NKCC2 in the nephron. In hippocampal and cerebellar neurons, expression of WNK3 mRNA is developmentally regulated in a pattern nearly identical to that of KCC2; before postnatal day 10, WNK3 expression is sparse but becomes intense by postnatal day 21 (36).

The knowledge that WNK1 autophosphorylation and kinase activity is increased substantially by elevated NaCl and other hypertonic challenges (including glucose, sucrose, mannitol, sorbitol, KCl, and urea) in a wide variety of cell lines (including fibroblasts, distal convoluted tubule kidney cells, and colon epithelial cells) (43, 83), coupled with the fact that WNK3 reciprocally regulates the Na-K-Cl and K-Cl cotransporters by affecting the net phosphorylation state of transporters, suggests that WNK3 could be a Cl⁻/volume-sensitive kinase that coordinates the activity of SLC12 family members. Studies examining WNK3 kinase activity in response to changes in cell volume and intracellular Cl⁻will be an important topic of future investigation.

Since Na-K-Cl cotransporters are activated by phosphorylation, and K-Cl cotransporters are inhibited by phosphorylation, the following model could account for the effects of WNK3: wild-type WNK3 increases cotransporter phosphorylation, whereas the expression of KI-WNK3 promotes cotransporter dephosphorylation (FIGURE 2). It is unclear whether wild-type WNK3 increases transporter phosphorylation through direct phosphorylation, activation of another downstream kinase, or inhibition of protein phosphatases. It is also unclear whether KI-WNK3 decreases transporter phosphorylation by acting in a dominant-negative manner, or whether it truly has a unique function compared with wild-type WNK3. What is evident is that, in the presence of KI-WNK3, the phosphorylation of cotransporters decreases due to the activity of phosphatases, because (at least for K-Cl cotransporters) specific protein phosphatase inhibitors reverse KI-WNK3 stimulatory effect. One explanation of these data is that wild-type WNK3 phosphorylates and inactivates phosphatases (increasing the net phosphorylation state of transporters), whereas KI-WNK3 is unable to phosphorylate and inhibit phosphatases (allowing net transporter dephosphorylation to occur) (FIGURE 3). Perhaps the effects of KI-WNK3 mimic a normal biochemical event in vivo, established by switching WNK3 from an active to inactive kinase in response to upstream signals. Because WNK1 is strongly activated by hypertonic stress (43), WNK3 could be a sensor (or part of a sensor) that monitors changes in intracellular Cl⁻ and/or cell volume, as well as an effector that regulates transporter phosphorylation/dephosphorylation.

Thus, in its kinase active and inactive state, WNK3 modulates the dynamic equilibrium between the kinase and phosphatase activities that regulate Cl⁻ entry and exit mechanisms in oocytes. The coexpression of WNK3 with its cation/Cl⁻ cotransporter targets in vivo suggest the effects of WNK3 are physiologically relevant.

WNK1 and WNK4 Regulate Cation/ Cl⁻ Cotransporters Via a Signal Transduction Pathway Involving Ste20-Type Protein Kinases

Other WNK kinases have recently been shown to regulate Na-K-Cl and K-Cl cotransport through activation of another serine-threonine kinase, SPAK (Ste20/SPS1-related, proline alanine-rich kinase; also known as PASK), a Ste20-type kinase. Ste20-type serine-threonine kinases, named after the founding member of the kinase superfamily discovered by genetic analysis of mating in the Saccharomyces cerevisiae (42), are divided into two groups: the p21-activated kinases (PAKs) and the germinal center kinases (GCKs) (6). PAKs are distinguished by the presence of a kinase domain located in the COOH terminus and an NH₂-terminal GTPase-binding domain. GCKs lack GTPase-binding domains, and the kinase domain is located in the NH₂ terminus. Ste20-type kinases play essential roles in signal transduction pathways that regulate apoptosis, development, cell-cycle control, cell growth, and cell stress responses (6).

PASK is a member of the GCK-VI subfamily, which includes the Drosophila gene Fray, the Caenorhabditis elegans gene Gck-3, and the mammalian oxidative stress-response protein 1 (OSR1) (4, 75). Piechotta et al. (59, 60) demonstrated that PASK and OSR1 interact with the NH₂ terminus of NKCC1 in the yeast two hybrid system via a specific binding motif ([R/K]FX[V/I]). PASK and NKCC1 co-localize in salivary glands and the choroid plexus (59, 60); earlier studies revealed PASK expression in neurons and several other Cl⁻-transporting epithelia, including the distal nephron (75).

The binding of PASK to NKCC1, and its coexpression with NKCC1 in Cl⁻-secreting epithelia, made it an attractive candidate for a kinase involved in NKCC1 regulation. Dowd et al. (12) demonstrated that overexpression of a dominant-negative form of PASK (DN-PASK) reduced NKCC1 activation in human embryonic kidney cells by ~70%. Overexpression of wild-type PASK caused only a small (~10%) increase in NKCC1 activity. DN-PASK inhibited the phosphorylation of the same two critical NH₂-terminal threonines in NKCC1 that are phosphorylated in the presence of WNK3 (12, 36). Protein phosphatase type 1 inhibitor calyculin A restored NKCC1 activation in the presence of DN-PASK, demonstrating that the effect of DN-PASK resulted from an alteration in kinase/phosphatase dynamics. Moreover, the phosphorylation state of PASK coincided with that of NKCC1, both increasing approximately fivefold in low intracellular Cl⁻. These data provided compelling evidence that PASK, a Cl⁻-sensitive kinase, is involved in the regulation of NKCC1. However, the small effect of wild-type PASK on NKCC1 activity suggested that certain co-factor(s) might be necessary for PASK’s full activation of NKCC1.

Yeast two hybrid analysis identified several proteins that interact with PASK, including the MAPK p38, heat shock protein 105, the apoptosis-associated tyrosine kinase (AATYK), and WNK4 (59, 60). Gagnon et al. (17) hypothesized that PASK might interact with other kinases to regulate NKCC1 and KCC2. Experiments revealed that NKCC1 function, measured in both isotonic (basal) and hypertonic (stimulated) conditions, was not significantly affected by coexpression of PASK or WNK4 alone; however, coexpression of PASK and WNK4 together resulted in a significant increase in NKCC1 activity that was insensitive to external osmolarity or cell volume. NKCC1 activation was dependent on the catalytic activities of both PASK and WNK4, because catalytically inactive versions of either kinase did not stimulate NKCC1. The physical interaction of WNK4 with PASK was also required for NKCC1 activation, because a mutant WNK4 deficient in PASK interaction did not affect NKCC1. Coexpression of PASK mutants that harbor mutations in the activation loop (T243A or T247A) with WNK4 not only prevented but robustly inhibited NKCC1 activity in cotransporter-injected oocytes (18). In addition, WNK4 and PASK, when expressed together but not alone, diminished the activity of the KCC2 under both isotonic (basal) and hypo-tonic (stimulated) conditions. DN-PASK, when expressed with WNK4, activated KCC2 in isotonic conditions. OSR1, a kinase closely related to SPAK, exhibited similar functional activation of NKCC1 when coex-pressed with WNK4 (18). These studies revealed a novel interaction between the WNK and Ste20-type kinases in the regulation of Na-K-Cl and K-Cl cotransport (17, 18).

Recent biochemical studies have shed insight into the mechanism behind WNK4/PASK’s regulation of NKCC1. Vitari et al. (76) and Moriguchi et al. (52) demonstrated that WNK4 bound and phosphorylated PASK at Thr-233 and Ser-373 in mammalian cells. PASK, when coex-pressed with WNK4, induced the phosphorylation of NKCC1 at its NH₂ terminus; expression of PASK or WNK4 alone had no effect on NKCC1 phosphorylation. Furthermore, PASK, when coexpressed with WNK4, could phosphorylate the NH₂ terminus of NKCC2 and NCC. These studies demonstrate that WNK4 binds and phosphorylates PASK; this phosphorylation event activates PASK, which in turn phosphorylates and activates NKCC1 (FIGURE 3). This same sequence of events has been demonstrated with WNK1 and OSR1; WNK1 can phosphorylate OSR1 (at sites homologous to the residues on PASK phosphorylated by WNK4), which in turn leads to NKCC1 activation via the OSR1-mediated phosphorylation of the NKCC1 amino terminus. Moreover, in vitro, any combination of WNK1 or WNK4 with OSR1 or PASK is sufficient for NKCC1 activation. It is unclear whether a similar phosphorylation cascade involving the WNK and Ste-20-type kinases regulates the K-Cl cotransporters.

Thus WNK and Ste20-type kinases are likely components of a complex signaling pathway that modulates cellular Cl⁻ entry and exit by altering the net phosphorylation state of the Na-K-Cl and K-Cl cotransporters (FIGURES 2 AND 3). How, or if, WNK3 interacts with the WNK4(1)/PASK(OSR1) pathway is unknown. Indeed, since all four WNK kinases contain potential binding sites for SPAK/OSR1, and both WNK1 and WNK4 can phosphorylate SPAK/OSR1, the separation of distinct WNK3 vs. WNK4(1)/SPAK(OSR1) pathways may be premature. However, WNK3’s unique properties [e.g., its ability to reciprocally regulate Na-K-Cl and K-Cl cotransporters when expressed alone in oocytes vs. the requirement of PASK(OSR1) coexpression for WNK4(1) regulation of these transporters in the same experimental system] suggest that it is not merely redundant to the function of WNK4(1)/PASK(OSR1), at least in oocytes. Since the activation of K-Cl cotransporters by KI-WNK3 is dependent on functioning protein phosphatases, wild-type WNK3’s regulation of its targets might occur indirectly through the inhibition of phosphatases. Conversely, data suggest that wild-type WNK4(1) phosphorylates PASK(OSR1), which in turn directly phosphorylates transporters. Perhaps WNK3 functions as a regulator of transporter dephosphorylation via interactions with phosphatases, whereas WNK4(1) regulates transporter phosphorylation via interaction with other Ste20-type kinases.

Physiological Implications and Future Directions

The recent data linking the WNK kinases to the reciprocal regulation of Na-K-Cl and K-Cl cotransport in oocytes and mammalian cell culture systems provide compelling reasons to speculate what physiological roles these kinases play in vivo.

Cell Volume Regulation: Coordination of Cl⁻ Influx and Efflux

The WNK and Ste20-type kinases fit the profile for the volume/Cl⁻-sensitive kinases that regulate cell volume. WNK3 [and WNK4(1)/PASK(OSR1)] increase NKCC1 activity while inhibiting K-Cl cotransport by promoting net transporter phosphorylation; in contrast, KI-WNK3 [and inactive forms of WNK4(1)/PASK(OSR1)] inhibits NKCC1 and activates K-Cl cotransporters by promoting net transporter dephosphorylation. Because the activity of WNK and Ste20-type kinases can be modulated by changes in cell volume, the opposite effects seen with the catalytically active and inactive forms of these kinases are likely physiologically relevant. In response to hypotonicity and cell swelling, WNK3 and WNK4(1)/PASK(OSR1) could be inactivated, promoting K-Cl cotransport over NKCC1 cotransport, thus promoting loss of solute and water. In response to hypertonicity and cell shrinkage, WNK3 and WNK4(1)/PASK(OSR1) could be activated, promoting NKCC1 cotransport over K-Cl cotransport, thus mediating gain of solute and water. One way to test this hypothesis in vivo would be to expose RBCs or primary cultures of epithelial cells from different WNK or Ste20-type kinase knockout mice to challenges in extra-cellular tonicity, then monitoring changes in cell volume. How WNK and Ste20-type kinases might enable cells to sense volume perturbations and transduce those changes into regulatory responses will be an interesting topic for future investigation. The bimodular domain structure of both kinase families, with NH₂-terminal kinase domains and COOH-terminal regulatory domains that contain binding sites for protein phosphatases, SH2/SH3 binding domains, coil-coil motifs, and cytoskeletal binding regions, suggest that these kinases are not simply relay stations along a linear phosphorylation cascade. Instead, they may serve as scaffolds, bringing other regulatory proteins into proximity with transporters. PASK has already been shown to bind actin and tubulin, translocate from the cytoplasm to the cytoskeleton in response to hyperosmotic stress in vivo, and serve as a physical linker between NKCC1 and WNK4 (17, 18, 59, 60, 74). Stimulation by hypertonicity or cell shrinkage might cause conformational changes in cytoskeleton proteins that could activate WNK3 and/or facilitate the interaction of WNK4(1) with PASK(OSR1) at downstream targets. This interaction could, in turn, trigger a phosphorylation cascade that activates Na-K-Cl cotransporters but inhibits K-Cl cotransporters. Conversely, cell swelling could have the opposite inhibitory effect on WNK3 and WNK4(1)/PASK(OSR1) activity, allowing net dephosphorylation of targets to take place by protein phosphatases.

Neuronal Excitability: Setting [Cl⁻]_i

The presence of WNK3 in GABAergic neurons has important implications for neuronal function. For example, the change from GABA excitation in prenatal life to GABA inhibition during adulthood, which is due to decreased NKCC1 and increased KCC2 activities with postnatal development, could in part be due to changes in WNK3 activity, or increases in WNK3 expression during postnatal developmental could counterbalance the rise in KCC2 activity seen during this time period. Derangements in the normal balance of NKCC1 and KCC2 activities have been shown to be responsible for certain types of seizures. These derangements could be due to altered regulation by WNK3. Measuring the GABA reversal potentials, neuronal [Cl⁻]_i and seizure thresholds of genetically altered WNK3 mice could test these hypotheses.

Dynamic regulation of intracellular Cl⁻ by WNK3 might also help explain the circadian variation of the response to GABA from excitatory to inhibitory that occurs in the supraoptic nucleus (SON), suprachiasmatic nucleus (SCN), and other areas of the reticular activating system that control the sleep/wake cycle (44, 58, 67, 72, 77). Neurons in these hypothalamic nuclei exhibit GABA-depolarizing currents during the day, but GABA-hyperpolarizing responses at night. WNK3 could regulate the cyclic firing activity of SON and SCN neurons by activating NKCC1 and inhibiting KCC2 during daytime and, doing the opposite, at night. This, in turn, could be achieved by altering the expression or activity of WNK3 with circadian rhythmicity. Consistent with this hypothesis is the high expression of WNK3 in these anatomical centers (36) and the facts that WNK kinases are phosphorylated by clock-associated proteins and show circadian rhythmicity of their expression in Arabidopsis thaliana (54, 55). Do WNK kinases, and WNK3 specifically, have circadian variation in their expression levels in the mammalian brain? Perhaps mice with targeted deletion of WNK3 have altered circadian variation of blood pressure, hormone secretion (cortisol or prolactin), or deranged sleep-wake cycles. Plants with deletion of the WNK3 Arabidposis homolog might have altered light-dark cycles.

In the brain, PASK is expressed in the gray and white matter of the spinal cord, the nodes of Ranvier in the sciatic nerve, and in the brain stem; more rostral areas of the CNS are devoid of PASK, except for the choroid plexus (60). In the choroid plexus, PASK discretely immunolabels the apical membrane in a pattern that closely matches NKCC1 (60). In contrast, in the choroid plexus from NKCC1 knockout mice, PASK immunostaining is found solely in the cytoplasm (59). It will be important to see whether PASK or WNK3 (which is expressed at intercellular junctions in the choroid plexus) plays a role in CSF secretion. A detailed characterization of WNK1, WNK4, and OS.R1 expression in the brain is needed to understand whether these kinases might regulate neuronal cation/Cl⁻ cotransport.

Notably, KCC2 exhibits considerable isotonic transport, in contrast to the other three K-Cl cotransporters that are active only in hypotonic conditions (21). A unique COOH-terminal domain has recently been implicated in this constitutive transport by KCC2 (50), which is furthermore resistant to the phosphatase inhibitor calyculin A (50, 70). Therefore, although WNK4(1)/PASK(OSR1) and WNK3 appear to affect constitutive activity of KCC2 in oocytes, in isotonic environments in vivo (i.e., the brain), other pathways may contribute to KCC2 regulation that are different from those involved in swelling/Cl⁻-induced K-Cl cotransport (50). Other molecules, such as CLC-2 Cl⁻ channels and carbonic anhydrase, might also be involved.

Transepithelial Ion Transport Coordinating Apical with Basolateral and Paracellular with Transcellular Ion Fluxes

WNK and Ste20-type kinases localize to Cl⁻-transporting epithelia such as the choroid plexus in the brain, ducts of salivary glands, parietal cells of the stomach, pancreatic ducts, bile ducts, and the epididymis (5, 33, 59, 60, 75). These cells actively transport electrolytes and water to regulate the osmolarity and ionic composition of body fluids such as cerebrospinal fluid, saliva, gastric juice, pancreatic secretions, bile, and seminal fluid. Activation of basolateral NKCC1 and apical Cl⁻ channels such as CFTR mediates fluid secretion in many of these epithelia (27). Coordinated simultaneous activation of NKCC1 and apical Cl⁻ channels, and inhibition of basolateral Cl⁻“leak” channels, would increase net secretory Cl⁻ and water transport. Because Ste20-type kinases have been shown to regulate both Cl⁻ channels (e.g., CLH-3b) and cation-Cl⁻ cotransporters (71), it will be important to investigate whether WNKs and/or the Ste20-type kinases coordinate apical and basolateral Cl⁻ transport.

Furthermore, the expression of WNK kinases at inter-cellular junctions and their ability to regulate both trans-cellular cation/Cl⁻ cotransporters and paracellular Cl⁻flux via phosphorylation of claudins (pore-forming proteins of the tight junction) suggest that they might coordinate transcellular and paracellular Cl⁻ transport (35, 86), a necessity for efficient transepithelial Cl⁻ transport. In this light, it is interesting that aldosterone has recently been shown to induce the phosphorylation of Claudin-4 and increase paracellular Cl⁻ transport between RCCD2 collecting duct cells (41). This hormone also stimulates K⁺ secretion through ROMK1 and Na⁺ reabsorption via NCC and ENaC (29). Each of these pathways is regulated by WNK4 and/or WNK1 (34, 40, 84, 85). It is plausible that WNK kinases are involved in the coordinated response of the transcellular and paracellular pathways to this hormone. In a single-cell system, such as a tight epithelium that is aldosterone sensitive, could induction of WNK4 kinase expression simultaneously regulate amiloride-sensitive Na⁺ currents and facilitate paracellular Cl⁻ flux? Moreover, could selective knockdown of WNK4 by siRNA in these cells abrogate the increase in ENaC activity or paracellular Cl⁻ conductance induced by aldosterone?

WNK1 and WNK4 are known to play important roles in the regulation of blood pressure, in part through regulation of NCC in the DCT (20). It will be important to see whether these kinases interact with the Ste20 kinases to execute their functions in the distal nephron; the expression of PASK in the DCT makes this a plausible hypothesis (75). Moreover, investigating the function of WNK3 in the kidney, with a focus of possible interactions between WNK3 and KCC3 in the PCT, NKCC2 in the TAL, NCC in the DCT, and KCC4 in the CCD, might shed insight into the regulation of electrolyte handling in these nephron segments. The fact that WNK3 stimulates the phosphorylation of the exact same threonine residues on NKCC2 that are phosphorylated, and necessary for NKCC2’s activation, in response to vasopressin suggests that WNK3 signaling lies downstream of this hormone (62). It will be of interest to examine the blood pressure, serum and urine electrolyte levels, and urine osmolarities of WNK3 (and other WNK and Ste20-type kinase) knockout mice.

Conclusion

It is becoming increasingly clear that the coordinated regulation of the cellular Cl⁻ influx and efflux branches of the SLC12 gene family (the Na-K-Cl and K-Cl cotransporters, respectively) does not take place via a single Cl⁻/ volume-sensitive kinase but rather occurs through a complex signaling pathway involving the WNK and Ste20-type serine-threonine kinases and associated proteins like phosphatases. Dissection of this molecular cascade is in its most nascent stage. However, compelling data derived from studies in Xenopus oocytes and mammalian cell culture systems has provided strong impetus to investigate the role of WNK and Ste20-type kinases in in vivo models, like knockout or transgenic mice. These future studies will likely provide novel insight into the normal physiological mechanisms governing epithelial Cl⁻ transport, regulation of neuronal excitability, and homeostasis of cell volume. Since mutations of several members of the WNK kinase family and SLC12 cation/Cl⁻ cotransporters cause rare inherited syndromes that include hypertension, epilepsy, or electrolyte disturbances as phenotypes, there is hope that, by examining the interaction between members of these gene families, insight into the pathophysiological mechanisms of certain human diseases will be achieved.