the role of the electroneutral cation chloride cotransporters (SLC12 family) Na⁺-K⁺-2Cl⁻ cotransporter 2 (NKCC2) and Na⁺-Cl cotransporter (NCC) in modulating arterial blood pressure has gained a lot of attention in the last decade. These cotransporters, expressed in the thick ascending limb of Henle's loop and the distal convoluted tubule, respectively, are responsible for a considerable amount of salt reabsorption. Their activity is important to define the final salt concentration in urine, thus affecting blood volume and arterial pressure. Inhibition of these cotransporters with loop or thiazide-type diuretics is the basis for the pharmacologic management of edematous states and arterial hypertension. Inactivating mutation of NKCC2 or NCC results in Bartter or Gitelman syndromes, respectively, featuring hypokalemic metabolic alkalosis and arterial hypotension. Additionally, lost of appropriate regulation of NCC, and perhaps NKCC2, by mutations in genes encoding kinases with-no-lysine[K] (WNK)1 or WNK4 or ubiquitin ligases CUL3 or KELCH3 results in pseudohypoaldosteronism type II (PHAII), also known as familiar hypertension with hyperkalemia (FHH). This is a mirror image disease of Gitelman syndrome exhibiting arterial hypertension and hyperkalemia, with metabolic acidosis.

The study of NCC regulation by WNK kinases revealed a complex network of kinases that modulates the activity of all members of the SLC12 family and other ion channels such as the epithelial Na⁺ channel (ENaC), the renal outer medulla potassium channel ROMK, and the paracellular claudins (4). Soon after the discovery of WNKs, it was demonstrated that in their effect towards SLC12 cotransporters, WNKs lie upstream of other serine/threonine kinases known as the Ste-20 proline alanine-rich kinase (SPAK) and the oxidative stress-responsive 1 (OSR1; Ref. 8) and that these last kinases are the responsible for direct phosphorylation of NCC or NKCC2 in the amino-terminal serine/threonine regulatory sites (1). Later on, genetically altered animal models revealed that regulation of NCC and NKCC2 by these kinases is more complex than expected (Table 1). Elimination of the particular threonine on SPAK that is phosphorylated by WNKs to activate SPAK towards SLC12 phosphorylation resulted in a knockin mice model SPAK^T243A/T243A that features a Gitelman like phenotype with a significant reduction of both NCC and NKCC2 expression and phosphorylation (7). In contrast, targeted disruption of SPAK gene expression resulted in the SPAK-knockout mice that also feature a Gitelman-like disease, with a decrease of NCC expression and phosphorylation, but surprisingly, with increased expression and phosphorylation of NKCC2 (2, 5, 10). In addition, a conditional kidney tubule-specific elimination of OSR1 resulted in a mice colony that developed a Bartter like syndrome with severe reduction of NKCC2 but increased expression/phosphorylation of SPAK and NCC (3). These works together suggested the intriguing idea that NKCC2 is under regulation by OSR1, while NCC is regulated by SPAK.

The discovery that SPAK gene expresses at least three distinct isoforms in the kidney began to shed light into the molecular mechanisms for the diversity of NKCC2 and NCC regulation by SPAK/OSR1 (5). In addition to the full-length SPAK, there are two variants known as SPAK2 and KS-SPAK. SPAK2, a shorter version of SPAK, lacking the initial PAPA box (which is also absent in OSR1), is possibly derived from an alternative translation start site within the transcript encoding full-length SPAK. KS-SPAK is a more extensively truncated variant that not only lacks the PAPA box but also the entire kinase domain, rendering it noncatalytic. This variant is derived from an alternative promoter and is only expressed in kidney and so was named KS-SPAK for kidney specific. Thus the difference between the SPAK-knockin mice and the SPAK-knockout mice will be that in the first, SPAK is not active, but the isoforms will be present along the nephron, while in the second one, all three isoforms will not be present. In addition, the observation that SPAK truncated variants are predominantly expressed in the thick ascending limb of Henle's loop suggested that their effect could be to reduce the activity of NKCC2, perhaps by interacting with OSR1, exerting a dominant negative effect.

In an issue of the American Journal of Physiology–Renal Physiology, Park et al. (6) went back to the functional expression system of Xenopus laevis oocytes to analyze the effect of SPAK isoforms on NKCC1 and NKCC2 activity. They observed that KS-SPAK is indeed a potent inhibitor of both NKCC1 and NKCC2 cotransporters. The effect of KS-SPAK requires the extreme COOH terminus that is known to mediate the interaction among WNKs, SPAK, and the cotransporter. It was observed that KS-SPAK coprecipitates with NKCC2 and that both protein-protein interaction and the negative effect on NKCC2 could be dramatically reduced by mutating key residues on KS-SPAK (D238 or L252) previously shown to be important for SPAK, NKCC1, and WNK1 interaction (9). SPAK-2 also exhibits some inhibitory effect on NKCC2, but it is significantly lower to that observed for KS-SPAK (16 vs. 70%), suggesting that SPAK-2 is probably able to interact with endogenous OSR1 in oocytes, without disrupting its effects on NKCC2. This is supported by the observation that the presence of the catalytic loop in SPAK-2, in contrast with KS-SPAK, makes the difference, because removal of 108 residues including the catalytic T-loop in SPAK2 turns it into a potent inhibitor of NKCC2.

The observations of Park et al. (6) provide strong evidence that indeed KS-SPAK is a truncated version of SPAK that is able to interact with OSR1-NKCC2, disrupting the positive effect of OSR1 on this cotransporter. Thus NCC is under the control of SPAK, while NKCC2 is under the control of OSR1 and the truncated version of SPAK, KS-SPAK, modulates the effect of OSR1 on NKCC2 (Fig. 1). These results support the hypothesis that in the SPAK^T243A/T243A knocking mice NKCC2 is decreased (7), while in the SPAK-knockout mice is increased (2, 5, 10), due to the presence of KS-SPAK and SPAK2 in the first, as well as a dominant negative full-length SPAK, but absence of inhibitory SPAK isoforms in the second model. In addition, because it is known that KS-SPAK is predominantly expressed in renal medulla (5), these studies together add a molecular mechanism for the fine regulation of NKCC2 activity. It will be interesting in the future to find out what specific physiological conditions, known to require NKCC2 regulation, produce a modification of isoform expression levels to promote or reduce the production of KS-SPAK isoform. In this regard, salt restriction and NCC knockout, both of which reduce extracellular fluid volume and activate NKCC2, decrease the ratio of KS-SPAK/full-length SPAK and increase full-length OSR1 (5).