SPAK Deficiency Corrects Pseudohypoaldosteronism II Caused by WNK4 Mutation

Stimulation of the OSR1 (Oxidative stress-responsive kinase-1)/SPAK [STE20 (sterile 20)/SPS1-related proline/alanine-rich kinase]-NCC (Na+-Cl− cotransporter) signaling cascade plays an important role in the WNK [With-No-Lysine (K)] kinase 4 D561A knock-in mouse model of pseudohypoaldosteronism type II (PHA II) characterized by salt-sensitive hypertension and hyperkalemia. The aim of this study was to investigate the respective roles of Osr1 and Spak in the pathogenesis of PHA II in vivo. Wnk4 D561A/+ mice were crossed with kidney tubule-specific (KSP) Osr1 knockout (KSP-Osr1 −/−) and Spak knockout (Spak −/−) mice. Blood pressure, plasma and urine biochemistries, and the relevant protein expression in the kidneys were examined. Wnk4 D561A/+, KSP-Osr1 −/−, and Spak −/− mice recapitulated the phenotypes of PHA II, Bartter-like syndrome, and Gitelman syndrome, respectively. Wnk4 D561A/+.KSP-Osr1 −/− remained phenotypically PHA II while Wnk4 D561A/+.Spak −/− mice became normotensive and lacked the PHA II phenotype. Phosphorylated Spak and Ncc were similarly increased in both Wnk4 D561A/+ and Wnk4 D561A/+.KSP-Osr1 −/− mice while phosphorylated Ncc normalized in Wnk4 D561A/+.Spak −/− mice. Furthermore, Wnk4 D561A/+.KSP-Osr1 −/− mice exhibited exaggerated salt excretion in response to thiazide diuretics while Wnk4 D561A/+.Spak −/− mice exhibited normal responses. Wnk4D561A/+.Spak −/−.KSP-Osr1 −/− triple mutant mice had low blood pressure and diminished phosphorylated Ncc. Both SPAK and OSR1 are important in the maintenance of blood pressure but activation of SPAK-NCC plays the dominant role in PHA II. SPAK may be a therapeutic target for disorders with salt-sensitive hypertension related to WNK4 activation.


Statistical analysis
All results are expressed as mean 6 standard deviation (SD). The significance of differences between groups was examined by K-independent samples Kruskal-Wallis nonparametric test with SPSS 21.0 for Windows (SPSS, Chicago, IL), followed by Mann-Whitnery two-sample test. The slope of Urine Na + and Cl 2 excretion rates between before and after diuretics administrating was analyzed by one-sample Kolmogorov-Smirnov test. A P-value less than 0.05 was considered to be statistically significant.

Response to diuretics
Urine Na + and Cl 2 excretion rates (FE Na and FE Cl ) were measured in mice before and after the administration of hydrochlorothiazide (HCTZ, a Ncc inhibitor) or furosemide (a Nkcc2 inhibitor) to assess the in vivo activities of Ncc and Nkcc2 respectively. Compared with WT and KSP-Osr1 2/2 mice, Wnk4 D561A/+ and Wnk4 D561A/+ .KSP-Osr1 2/2 mice exhibited exaggerated salt excretion in response to a single dose of HCTZ, indicating the Ncc overactivity in both sets of mice ( Figure 6A). When challenged with furosemide, KSP-Osr1 2/2 mice showed blunted response in comparison with WT, suggesting lower Nkcc2 activity ( Figure 6B). However, Wnk4 D561A/+ and Wnk4 D561A/ + .KSP-Osr1 2/2 mice responded similarly to WT.
The Spak 2/2 mice had blunted urine Na + and K + excretion compared to WT in response to HCTZ ( Figure 7A), indicating lower Ncc activity. Interestingly, the response of Wnk4 D561A/+ .Spak 2/2 mice was between that of Spak 2/2 and Wnk4 D561A/+ mice and similar to WT controls, indicating that the Ncc function had normalized in these mice. In the furosemide challenge, Spak 2/2 and Wnk4 D561A/+ .Spak 2/2 mice had significantly increased Na + and Cl 2 excretion, suggesting increased Nkcc2 function ( Figure 7B).

Discussion
In this study, we crossed Wnk4 D561A/+ mice with KSP-Osr1 2/2 and Spak 2/2 mice to investigate the independent roles of Osr1 and Spak in the pathogenesis of Wnk4-PHA II. Wnk4 D561A/+ .KSP-Osr1 2/2 mice preserved the PHA II phenotype with increased abundance of p-Spak and p-Ncc and corresponding exaggerated response to thiazide diuretics. Wnk4 D561A/+ .Spak 2/2 mice exhibited normal blood pressure and blood/urine electrolytes with relatively normal abundance of p-Ncc despite enhanced p-Osr1 expression and a normal response to thiazides. These findings indicated that activation of Spak-Ncc plays the more dominant role in Wnk4-PHA II, which is affirmed by the decreased total expresssion and phosphorylation of Ncc in triple mutant Wnk4 D561A/+ .Spak 2/2 .KSP-Osr1 2/2 mice.
Mutations in the Wnk4 kinase gene have been shown to cause many cases of PHA II. [15,[38][39][40] Recent studies have implicated  the activation of downstream WNK4 substrates, SPAK and OSR1, in the pathogenesis of PHA II. [26,32,36,41,42] Although SPAK has been found to be predominantly expressed in the cortex and OSR1 in the medulla, both are expressed in the TAL and DCT. [29] Unlike the interchangeability of SPAK and OSR1 in peripheral neurons, [43] these two kinases seem to be differentially regulated and have different function in renal tubules. This study clarified the relative contribution of Spak and Osr1 to PHA II in vivo. Wnk4 D561A/+ .KSP-Osr1 2/2 mice had increased expression of total Ncc and p-Ncc in parallel with increased total and p-Spak. Their exaggerated response to thiazide diuretics indicated Ncc hyperactivity, similar to Wnk4 D561A/+ mice, and in line with the immunoblottings. These findings suggest that Osr1 is not essential and can be fully compensated by the increased Spak expression in this modle of PHA II. However, the decreased p-Ncc expression found in heterozygous kinase-dead Osr1 knockin (Osr1 T158A/+ ) mice and Wnk4 D561A/+ .KSP-Osr1 2/2 .Spak 2/2 triple mutant mice suggested that the rescue of Ncc activation in Osr1 deficient states depends on abnormal activation of the Spak pathway. [41] The roles of Osr1 and Spak on Nkcc2 in TAL were also clarified by this study. KSP-Osr1 2/2 mice exhibited reduced p-Nkcc2, indicating that Osr1 is an up-regulator of Nkcc2. Substantively, the Spak 2/2 , Wnk4 D561A/+ , and Wnk4 D561A/+ .Spak 2/2 mice all exhibited increased p-Nkcc2.
These three sets of mice share the commonality of preserved or increased Osr1. However, the Wnk4 D561A/+ .KSP-Osr1 2/2 mice also exhibited increased p-Nkcc2 suggesting that mutant Wnk4 can activate Nkcc2 through activited Spak. This is corrobarated by our finding of decreased p-Nkcc2 when Spak is abolished in the triple mutant Wnk4 D561A/+ .KSP-Osr1 2/2 .Spak 2/2 mice. It would appear that Osr1 is the major activator of Nkcc2 but Spak may play a role in abnormally activated states. Recently, it has reported that the kinase-deficient SPAK variant, so-called kidney specific SPAK (KS-SPAK), functions as an antagonist of OSR/SPAK-NKCC2 pathway and is the major SPAK isoform in renal medulla. [27] Since WNK4 expression in the Henle's loop is primarily in the cortical TAL, [44] the role of KS-SPAK in PHA II with WNK 4 mutation is still questionable.
The furosemide challenge studies generally correlate with the densitometry studies except in the Wnk4 D561A/+ and Wnk4 D561A/+ .KSP-Osr1 2/2 mice, which had normal responses to furosemide despite increased phosphorylated Nkcc2. It is important to note that both of these mice have hyperactive downstream Ncc, which may attenuate the observable response to furosemide. Supporting this theory is the observation that Wnk4 D561A/+ .Spak 2/2 mice, with their increased activated Osr1 and p-Nkcc2 but relatively normal p-Ncc expression, showed an exaggerated response to furosemide, providing direct evidence Role of SPAK and OSR1 in PHAII PLOS ONE | www.plosone.org linking increased Nkcc2 activity through Osr1. Besides Wnk4, other upstream regulator, such as WNK1 or calcium-binding protein 39 (Cab39), also regulate OSR1-NKCC2 pathway. [19,21,[45][46][47].
Regarding the Ncc in the distal nephron, Spak appears to be the dominant player as Wnk4 D561A/+ .Spak 2/2 mice became virtually normal phenotype with expression levels of total and p-Ncc similar to WT littermates, indicating that the PHA II phenotype could be effectively corrected by Spak deficiency. The response to thiazide diuretics in Wnk4 D561A/+ .Spak 2/2 mice was similar to WT mice (and higher than Spak 2/2 mice) suggesting that Spak is a major but not sole activator of Ncc in Wnk4-PHA II. Osr1 is a likely accomplice and increased Osr1 activity through activated Wnk4 may compensate enough to sustain normal Ncc expression and activity. This borne out by the finding of decreased expression and phosphorylation of Ncc in triple mutant Wnk4 D561A/+ . KSP-Osr1 2/2 .Spak 2/2 mice, where Osr1 has been abolished. The phenotype and Ncc phosphorylation level of our Wnk4 D561A/+ .Spak 2/2 mice resembled those of the recentlyreported Wnk4 D561A/+ .Spak T243A/T243A mice (kinase-dead knockin), [41] which also support the importance of SPAK kinase activity in PHA II. Another recent study has demonstrated that WNK4-SPAK-dependent signaling is the primary mechanism behind angiotensin II induced Ncc stimulation. [42] The WNK4-NCC signaling pathway is also regulated by other hormones (aldosterone and insulin) and drugs (tacrolimus, cyclosporine) associated with salt-sensitive hypertension. [36,[48][49][50][51] Whether those mechanisms are principally mediated through SPAK warrants further investigation.
Historically, WNK4 was reported to inhibit membrane trafficking of NCC based on oocyte experiments. [52][53][54] However, this Ncc inhibitory mechanism has not been found in vivo. Similarly, in vitro studies proposing various mechanisms WNK4related for Ncc degradation are equally suspect. [55,56] How WNK4 directly affects NCC in vivo merits further study.
Thiazide diuretics are commonly and effectively used to treat human PHA II disease. However, the chronic use of thiazide also cause several side effects, such as insulin resistance with hyperglycemia, hyperlipidemia, hyperuricemia with gout, chronic kidney injury and even renal failure. These side effects can be independent of volume status and plasma K + concentration. [57] Because both Spak deficiency and inhibition of Spak kinase activity corrected the phenotype of PHA II due to Wnk4 mutation, specific inhibition of SPAK may be a plausible therapy for patients with salt-sensitive hypertension related to WNK4 activation. Since human PHA II is also linked to the mutations in WNK1, Kelchlike 3 or cullin 3 genes, the SPAK in those gene mutations will need to be clarified first. [16,17,20,[58][59][60].