WNK-SPAK/OSR1 signaling: lessons learned from an insect renal epithelium

Abstract

WNK [with no lysine (K)] kinases regulate renal epithelial ion transport to maintain homeostasis of electrolyte concentrations, extracellular volume, and blood pressure. The SLC12 cation-chloride cotransporters, including the sodium-potassium-2-chloride (NKCC) and sodium chloride cotransporters (NCC), are targets of WNK regulation via the intermediary kinases SPAK (Ste20-related proline/alanine-rich kinase) and OSR1 (oxidative stress response). The pathway is activated by low dietary potassium intake, resulting in increased phosphorylation and activity of NCC. Chloride regulates WNK kinases in vitro by binding to the active site and inhibiting autophosphorylation and has been proposed to modulate WNK activity in the distal convoluted tubule in response to low dietary potassium. WNK-SPAK/OSR1 regulation of NKCC-dependent ion transport is evolutionarily ancient, and it occurs in the Drosophila Malpighian (renal) tubule. Here, we review recent studies from the Drosophila tubule demonstrating cooperative roles for chloride and the scaffold protein Mo25 (mouse protein-25, also known as calcium-binding protein-39) in the regulation of WNK-SPAK/OSR1 signaling in a transporting renal epithelium. Insights gained from this genetically manipulable and physiologically accessible epithelium shed light on molecular mechanisms of regulation of the WNK-SPAK/OSR1 pathway, which is important in human health and disease.

WNK-SPAK/OSR1 SIGNALING IN MAMMALIAN RENAL PHYSIOLOGY

WNK [with no lysine (K)] kinases play an important role in maintaining homeostasis of electrolyte concentrations, extracellular volume, and blood pressure by regulating renal epithelial ion transport. Human gain-of-function mutations in WNK1 or WNK4 result in hypertension and hyperkalemia (69), a phenotype that has been recapitulated in animal models (24, 63, 68, 75). Conversely, WNK4 knockout mice are hypokalemic and have decreased blood pressure (8, 32, 58). WNK kinases phosphorylate and activate two related downstream Ste20 kinases, SPAK (Ste20-related proline/alanine-rich kinase) and OSR1 (oxidative stress response) (30, 64). SPAK and OSR1 in turn phosphorylate conserved serines and threonines in the NH₂ termini of sodium-dependent SLC12 family cation-chloride cotransporters, including the sodium chloride cotransporter (NCC) and sodium-potassium-2-chloride cotransporters NKCC1 and NKCC2 (1, 11, 15, 43, 44, 65). NCC and NKCC2 reabsorb sodium chloride in the distal convoluted tubule and thick ascending limb of the mammalian nephron, respectively, whereas NKCC1 has a more widespread distribution, including in secretory epithelia (16).

CHLORIDE REGULATION OF THE WNK-SPAK/OSR1-NCC/NKCC PATHWAY

Cation-chloride cotransporters, including NKCCs, are evolutionarily ancient (18). Studies in shark rectal gland, squid giant axon, and rat salivary acinar gland showed that lowering of intracellular chloride results in NKCC1 NH₂-terminal threonine phosphorylation by SPAK (also called PASK) and cotransporter activation (6, 9, 11, 28, 45, 48). WNK-SPAK/OSR1 pathway activation also occurs in cultured cells in conditions resulting in lowering of intracellular chloride (30, 43, 44). This raised the possibility that WNKs (or SPAK/OSR1) could be chloride-sensitive kinases (23, 36), an idea supported by the demonstration that chloride ion binds directly to the active site of the human WNK1 kinase domain (38). The structure of the chloride-binding pocket also explains the atypical placement of the WNK catalytic lysine (73), which would otherwise interfere with chloride binding. Chloride binding inhibits autophosphorylation required for kinase activation (38), and chloride also inhibits the activity of WNK3 and WNK4 (59).

More recently, intracellular chloride has been proposed to regulate NCC activity in response to changes in dietary potassium. High-potassium diet promotes renal potassium excretion through aldosterone-dependent and -independent pathways (13, 20, 54, 70, 77). One mechanism is by inhibiting proximal reabsorption of sodium, allowing increased delivery to the distal nephron, where potassium is secreted in exchange for sodium. Early studies showed that this occurs in the proximal tubule and thick ascending limb (3, 5, 19, 55, 61). Low potassium also stimulates NCC phosphorylation and activation in a WNK-SPAK/OSR1-dependent manner, whereas high potassium results in NCC dephosphorylation (12, 37, 41, 52, 60, 62, 66, 67, 76). Terker et al. proposed that low dietary potassium hyperpolarizes the basolateral membrane of the distal convoluted tubule, resulting in chloride exit through basolateral channels and lowering of intracellular chloride. Relief of chloride inhibition of WNK signaling could then lead to increased NCC phosphorylation and activity. This model was supported by studies in cultured cells, mathematical modeling, and studies in isolated distal convoluted tubules (37, 60).

THE WNK-FRAY-NKCC PATHWAY IN DROSOPHILA RENAL TUBULE ION TRANSPORT

We have explored the role of intracellular chloride in the regulation of WNK signaling and transepithelial ion transport in the Drosophila melanogaster (Dm) renal epithelium, which is easily amenable to genetic manipulation, including cell-specific gene knockdown and expression (4, 53), analysis of transepithelial water, and ion flux in isolated tubules (10, 50) and measurement of intracellular chloride concentrations in live tubules (57). The four fly renal (Malpighian) tubules are located in the abdominal cavity and are bathed in hemolymph (the insect equivalent of plasma). Because the tubules are blind-ended, urine generation occurs by the transepithelial movement of ions and water across the epithelial cells of the main segment of the tubule (10, 34), which consists of cation-conducting principal cells and chloride-conducting stellate cells (7, 26, 33, 35, 42) (Fig. 1).

Fig. 1.

Model of WNK-SPAK/OSR1 [with no lysine (K) kinase-Ste20-related proline/alanine-rich kinase/oxidative stress response 1] signaling in the Drosophila renal tubule. Flies have four renal tubules, an anterior pair and a posterior pair, that lie in the abdominal cavity bathed in hemolymph. Urine generation by the tubule main segments occurs by the transepithelial secretion of cations through principal cells and chloride through stellate cells. The apically located vacuolar H⁺-ATPase drives fluid secretion by generating a lumen-positive transepithelial potential that allows exchange of protons for cations (primarily potassium in the Drosophila tubule). Chloride secretion is also driven by the lumen-positive charge (7, 10, 33). In the principal cell, the sodium-potassium-2-chloride (NKCC) encoded by Ncc69, and inwardly-rectifying potassium channels encoded by Irk1 and Irk2, are required for normal transepithelial ion flux. Sodium entering through the NKCC is recycled by the basolateral Na⁺/K⁺-ATPase (46, 71). Chloride may be recycled as well, as there is an outwardly-directed basolateral chloride conductance, whose molecular identity is unknown (22). Drosophila WNK and Fray (the fly SPAK/OSR1 homolog) are positive regulators of transepithelial ion flux through the NKCC (72). Hypotonic bathing medium, which stimulates transepithelial ion flux in a WNK-, Fray-, and NKCC-dependent manner (72), results in a decrease in principal cell intracellular chloride concentration over ten to sixty minutes, and an increase in tubule Drosophila melanogaster (Dm)WNK activity at thirty to sixty minutes. Drosophila Mo25 is also required for stimulation of transepithelial ion flux by hypotonic bathing medium. The overexpression of Mo25 with a chloride-insensitive WNK mutant in adult tubules is sufficient to increase transepithelial ion flux in isotonic conditions, indicating a cooperative role for chloride and Mo25 in regulation of WNK pathway activation (57). Reprinted with permission from the Journal of the American Society of Nephrology.

We found that the NKCC-encoding Drosophila gene Ncc69 (25, 56) is required in the principal cells of the fly tubule for normal fluid secretion and transepithelial ion flux (46). Drosophila has a single WNK homolog, DmWNK, as well as a single SPAK/OSR1 homolog, Fray (47). As in mammals, DmWNK phosphorylates Fray, and Fray phosphorylates the NH₂ terminus of the fly NKCC (49, 51, 72). Furthermore, we demonstrated that DmWNK and Fray positively regulate transepithelial ion flux in an NKCC-dependent manner (72). Thus, ion transport regulation via WNK-SPAK/OSR1 modulation of SLC12 cotransporters is conserved from Drosophila to mammals.

ROLE OF CHLORIDE IN DROSOPHILA WNK PATHWAY REGULATION

Bathing cultured cells in hypotonic low-chloride medium results in activation of WNK-SPAK/OSR1 signaling, and NKCC1 is also activated in shark rectal gland bathed in hypotonic medium (27, 30, 43, 44). Vertebrate renal epithelial cells exposed to hypotonic medium swell and then undergo regulatory volume decrease, with a fall in intracellular chloride concentration as a result of efflux of potassium and chloride (21, 29). We found that bathing Drosophila renal tubules in hypotonic medium resulted in DmWNK-, Fray-, and NKCC-dependent stimulation of transepithelial ion flux (72). We were therefore interested in whether 1) intracellular chloride was decreased in these conditions, and 2) DmWNK was acutely activated. Using the transgenic chloride sensor, ClopHensor (2, 31), we made paired measurements in tubules bathed in isotonic standard bathing medium and then switched to hypotonic medium and found that intracellular chloride fell from 27 to 30 mM to 15–16 mM by 30–60 min in hypotonic medium (57).

The chloride-binding site of human WNK1 is perfectly conserved in DmWNK, and we demonstrated that autophosphorylation of the DmWNK kinase domain is inhibited by chloride in vitro. To determine whether endogenous DmWNK is acutely activated in low intracellular chloride conditions, we introduced transgenic kinase-dead rat SPAK into the principal cells to serve as a substrate for DmWNK and monitored its phosphorylation by immunoblotting for phosphorylated and total SPAK. Increased tubule DmWNK activity was seen at 30 and 60 min after hypotonic medium exposure, corresponding to the nadir of intracellular chloride concentration (57). Thus, tubule WNK activity was acutely stimulated when intracellular chloride concentrations decreased.

Similar to the distal convoluted tubule, the fly principal cell has basolateral potassium and chloride conductances (22). Therefore, changes in bath potassium concentrations likely also result in a change in intracellular chloride. DmWNK activity increased in an isotonic low-potassium bath and decreased in an isotonic high-potassium bath (57). Similar results were seen in a mouse kidney slice preparation, although the high-potassium effect did not reach statistical significance (37).

To test whether relief of DmWNK chloride inhibition is sufficient to increase transepithelial ion flux, we made transgenic flies to allow expression of a chloride-insensitive WNK mutant, DmWNK^L421F [homologous to human WNK1^L369F (38)] in tubule principal cells. Whereas expression of chloride-insensitive DmWNK results in gain-of-function phenotypes in other physiological processes in the fly (unpublished observations), expression of DmWNK^L421F did not alter transepithelial ion flux in isotonic conditions, suggesting a requirement for additional factors for maximal pathway activation.

ROLE OF Mo25 IN DROSOPHILA ION TRANSPORT REGULATION

Mo25 (Mouse protein 25, also known as calcium-binding protein 39, or Cab39) acts synergistically with WNKs to activate SPAK/OSR1 in vitro (14) and activates NKCC1 and NKCC2 in Xenopus oocytes together with WNK-SPAK/OSR1 pathway components (39, 40). Cooperative roles for Drosophila Mo25 (DmMo25) and Fray have also been demonstrated in Drosophila development (74). We (57) showed that, similarly to its mammalian homolog, DmMo25 potentiated Fray kinase activity in vitro. Furthermore, principal cell knockdown of DmMo25 decreased transepithelial ion flux in hypotonic conditions, but not in isotonic conditions, suggesting that DmMo25 is required in conditions of maximal pathway activation.

Overexpression in adult tubules of wild-type DmWNK, DmMo25, or both together did not alter ion flux, nor, as mentioned above, did overexpression of DmWNK^L421F. However, overexpression of DmWNK^L421F together with DmMo25 resulted in increased ion flux in isotonic conditions, similar to the increase seen in hypotonic conditions. Thus, relief of chloride inhibition, together with Mo25, is able to recapitulate activated pathway conditions and increase transepithelial ion flux (57), suggesting a cooperative role for chloride and Mo25 in WNK pathway activation.

CONCLUSION AND PERSPECTIVES

We have thus demonstrated for the first time that conditions that lower measured intracellular chloride concentration in a transporting renal epithelium result in acute activation of WNK signaling and a WNK-SPAK/OSR1/Fray-dependent increase in transepithelial ion transport (57, 72). Our study also highlights an important role for Mo25. While Mo25 is expressed in the thick ascending limb and distal convoluted tubule in the mammalian nephron (17), its physiological role in the kidney is unknown, as is the relative importance of chloride and Mo25 in the regulation of WNK signaling in different cell types or nephron segments. Our study suggests that DmMo25 is dispensable under baseline conditions but is required under conditions of maximal pathway activation. In contrast, in HEK-293 cells, Mo25α knockdown decreased NKCC1 activity in both baseline and stimulated conditions (14). This suggests that Mo25 could allow adaptive regulation of ion transport depending on physiological conditions in different cell types. In addition, WNK4 and Mo25 can directly activate NKCC1 and NKCC2 independently of SPAK or OSR1 in the Xenopus oocyte heterologous expression system (40), which could add additional flexibility to the system. How Mo25 itself is regulated is also unknown. The fly renal tubule affords an opportunity to carry out detailed mechanistic studies in a transporting renal epithelium, yielding insights that could further inform understanding of ion transport regulation in humans.

GRANTS

A. R. Rodan is supported by grants from the National Institute of Diabetes and Digestive and Kidney Diseases (DK-110358 and DK-106350) and the American Heart Association (16CSA28530002).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author.

AUTHOR CONTRIBUTIONS

A.R.R. prepared figures, drafted manuscript, edited and revised manuscript, and approved final version of manuscript.