Impaired phosphorylation of Na+-K+-2Cl− cotransporter by oxidative stress-responsive kinase-1 deficiency manifests hypotension and Bartter-like syndrome

Na+-K+-2Cl− cotransporters (NKCCs), including NKCC1 and renal-specific NKCC2, and the Na+-Cl− cotransporter (NCC) play pivotal roles in the regulation of blood pressure (BP) and renal NaCl reabsorption. Oxidative stress-responsive kinase-1 (OSR1) is a known upstream regulator of N(K)CCs. We generated and analyzed global and kidney tubule-specific (KSP) OSR1 KO mice to elucidate the physiological role of OSR1 in vivo, particularly on BP and kidney function. Although global OSR1−/− mice were embryonically lethal, OSR1+/− mice had low BP associated with reduced phosphorylated (p) STE20 (sterile 20)/SPS1-related proline/alanine-rich kinase (SPAK) and p-NKCC1 abundance in aortic tissue and attenuated p-NKCC2 abundance with increased total and p-NCC expression in the kidney. KSP-OSR1−/− mice had normal BP and hypercalciuria and maintained significant hypokalemia on a low-K+ diet. KSP-OSR1−/− mice exhibited impaired Na+ reabsorption in the thick ascending loop on a low-Na+ diet accompanied by remarkably decreased expression of p-NKCC2 and a blunted response to furosemide, an NKCC2 inhibitor. The expression of total SPAK and p-SPAK was significantly increased in parallel to that of total NCC and p-NCC despite unchanged total NKCC2 expression. These results suggest that, globally, OSR1 is involved in the regulation of BP and renal tubular Na+ reabsorption mainly via the activation of NKCC1 and NKCC2. In the kidneys, NKCC2 but not NCC is the main target of OSR1 and the reduced p-NKCC2 in KSP-OSR1−/− mice may lead to a Bartter-like syndrome.

Edited by Maurice B. Burg, National Heart, Lung, and Blood Institute, Bethesda, MD, and approved September 12, 2011 (received for review May 11, 2011) Na + -K + -2Cl − cotransporters (NKCCs), including NKCC1 and renal-specific NKCC2, and the Na + -Cl − cotransporter (NCC) play pivotal roles in the regulation of blood pressure (BP) and renal NaCl reabsorption. Oxidative stress-responsive kinase-1 (OSR1) is a known upstream regulator of N(K)CCs. We generated and analyzed global and kidney tubule-specific (KSP) OSR1 KO mice to elucidate the physiological role of OSR1 in vivo, particularly on BP and kidney function. Although global OSR1 −/− mice were embryonically lethal, OSR1 +/− mice had low BP associated with reduced phosphorylated (p) STE20 (sterile 20)/SPS1-related proline/alanine-rich kinase (SPAK) and p-NKCC1 abundance in aortic tissue and attenuated p-NKCC2 abundance with increased total and p-NCC expression in the kidney. KSP-OSR1 −/− mice had normal BP and hypercalciuria and maintained significant hypokalemia on a low-K + diet. KSP-OSR1 −/− mice exhibited impaired Na + reabsorption in the thick ascending loop on a low-Na + diet accompanied by remarkably decreased expression of p-NKCC2 and a blunted response to furosemide, an NKCC2 inhibitor. The expression of total SPAK and p-SPAK was significantly increased in parallel to that of total NCC and p-NCC despite unchanged total NKCC2 expression. These results suggest that, globally, OSR1 is involved in the regulation of BP and renal tubular Na + reabsorption mainly via the activation of NKCC1 and NKCC2. In the kidneys, NKCC2 but not NCC is the main target of OSR1 and the reduced p-NKCC2 in KSP-OSR1 −/− mice may lead to a Bartter-like syndrome.
Bartter syndrome | electrolytes | hormone | knockout mice | volume R ecent studies have shown that Na + -K + -2Cl − cotransporters (NKCCs) and the Na + -Cl − cotransporter (NCC) play very important roles in the regulation of blood pressure (BP) and extracellular volume. NKCCs consist of ubiquitous NKCC1 and renal-specific NKCC2. NKCC1 can modulate BP through vascular and renal effects (1)(2)(3)(4). NKCC2 and NCC are two renal Na + cotransporters expressed in the thick ascending limbs (TALs) and distal convoluted tubules (DCTs) of the kidney, respectively, accounting for 20% and 10% of filtered Na + reabsorption (5). In human essential hypertension and salt-sensitive or spontaneously hypertensive animal models, activation of NKCC1 and NKCC2 has also been reported to play a pivotal role in the pathogenesis of hypertension (6,7). In addition, activation of NCC by gene mutations in WNK1 and WNK4 leads to an autosomal dominant salt-sensitive hypertension known as pseudohypoaldosteronism type II (PAHII) (8). On the other hand, loss-of-function mutations in the SLC12A1 and SLC12A3 genes encoding NKCC2 and NCC can lead to renal salt-wasting hypotension with hypokalemic metabolic alkalosis, known as Bartter syndrome (BS) (9) and Gitelman syndrome (GS) (10), respectively.
In vitro studies have shown that posttranscriptional phosphorylation of NKCC1/2 and NCC plays a crucial role in the regulation of normal transport activity. Oxidative stress-responsive kinase-1 (OSR1) (11) and STE20 (sterile 20)/SPS1-related proline/alaninerich kinase (SPAK) (12), two downstream substrates of With-No-Lysine kinase (WNK) 1/4, are the upstream phosphorylators of NKCC1/2 and NCC. Threonine or serine residues in their N-terminal conserved domains (T206/96, T211/101, and T224/114 in mouse NKCC1/2; T53, T58 and S71 in mouse NCC) are the phosphorylation sites of OSR1 and SPAK. The docking site on the conserved C-terminal domains of OSR1/SPAK interacts with the RFXV/I motif on the N terminus of NKCC/NCC and then increases NKCC/NCC phosphorylation and function (13)(14)(15)(16). We have also reported that increased phosphorylated (p) OSR1/SPAK abundance can enhance p-NCC expression in the PHAII-causing Wnk4 D561A knock-in mice (17), whereas the reverse is true in the Wnk4 hypomorphic knockout (KO) mice (18). These findings support that OSR1 and SPAK are important regulators of NKCC and NCC in vivo.
Because OSR1 and SPAK share high homology in their catalytical and regulatory domains and their expression in tissues often overlaps, the creation and analysis of distinct OSR1 or SPAK KO mice is warranted to tease apart the role of each kinase in vivo. For this purpose, we first generated SPAK KO mice and found that SPAK +/− mice exhibited hypotension with decreased p-NKCC1 abundance in aortic tissues and SPAK −/− mice presented a GS phenotype caused by reduced total and p-NCC expression (19). In the present study, we generated global and kidney tubulespecific (KSP) OSR1 KO mice to elucidate the physiological role of OSR1 in vivo (SI Text and Figs. S1 and S2). Results to be reported indicate that global OSR1 −/− mice were embryonically lethal and OSR1 +/− mice had low BP associated with reduced p-SPAK expression and p-NKCC1 abundance in aortic tissue and attenuated p-NKCC2 abundance with increased total and p-NCC expression in the kidney. KSP-OSR1 −/− mice manifested Bartterlike syndrome because of impaired NKCC2 phosphorylation and function in the TAL with a compensatory increase in NCC phosphorylation and expression. This study provides in vivo evidence that OSR1 is primarily involved in the regulation of BP and renal tubular Na + reabsorption via the phosphorylation of NKCC1 and NKCC2 but not NCC.
We further evaluated renal Na + and K + handling in the global OSR1 +/− and KSP-OSR1 −/− mice. On normal Na + diets, all these different groups of mice had similar urine Na + excretion rates and fractional excretion of Na + (FE Na ) (Tables 1 and 2). On paired-fed low-Na + diets, the urinary Na + excretion quickly dropped in all three groups (WT, global OSR1 +/− , and KSP-OSR1 −/− ) of mice. There was no significant difference in the degree of reduced urine Na + excretion between the global OSR1 +/− and WT littermates (Fig. 1A). However, KSP-OSR1 −/− mice clearly demonstrated more urine Na + excretion than WT controls on the second, third, and fourth days (Fig. 1B), supporting the presence of a renal tubular defect in KSP-OSR1 −/− mice. Compared with WT controls, the average increased Na + excretion was 115 ± 7 μmol per mouse for the 6 d of the low-Na + diet in the KSP-OSR1 −/− mice (Fig. 1B).
Diuretic Response in Global OSR1 +/− and KSP-OSR1 −/− Mice. To determine the function of NKCC2 and NCC, two substrates of OSR1, in these global OSR1 +/− and KSP-OSR1 −/− mice, we administered the NKCC2 inhibitor furosemide and NCC inhibitor hydrochlorothiazide (HCTZ), respectively. Like WT mice, global OSR1 +/− mice showed a dramatic and similar increase in the FE Na, FE K , and FE Cl in response to furosemide ( Fig. 2A) and HCTZ (Fig. 2B), suggesting that their NKCC2 and NCC functions were not obviously affected. Compared with WT and global OSR1 +/− mice, KSP-OSR1 −/− mice exhibited a blunted response to furosemide (Fig. 2C) but a normal response to HCTZ (Fig. 2D), indicating that their NKCC2 function was reduced and NCC function was well preserved (Fig. 2D).
Renal OSR1, SPAK, NKCC2, and NCC in Global OSR1 +/− and KSP-OSR1 −/− Mice. In the kidneys, we had previously reported that OSR1 is mainly distributed in the TAL to downstream renal tubules and is dominantly expressed in the medulla, colocalizing with NKCC2 (19). Because KSP-OSR1 −/− mice displayed a BS-like phenotype with hypercalciuria and normal magnesemia and a blunted response to furosemide corresponding to a TAL lesion rather than GS physiology (a DCT lesion with hypocalciuria and hypomagnesemia) (20), the expression of total and p-OSR1, SPAK, NKCC2, and NCC in the kidney of both OSR1 +/− and KSP-OSR1 −/− mice was further evaluated by semiquantitative immunoblotting (IB).

Discussion
In this study, we generated and analyzed global and KSP-OSR1 KO mice to elucidate the physiological role of OSR1 in vivo in the regulation of the NKCC1 and kidney-specific NKCC2 and NCC, focusing on BP and renal tubular Na + reabsorption. As previously reported in OSR1 gene-trapped or kinase-dead knock-in mice (21,22), global homozygous OSR1 −/− mice were embryonically lethal. Heterozygous OSR1 +/− mice exhibited hypotension, markedly reduced p-NKCC1 abundance in aortic tissue, and attenuated p-NKCC2 in kidney. KSP-OSR1 −/− mice recapitulated the reduced renal Na + reabsorption on low-Na + diets and remarkably decreased expression of p-NKCC2. They also had a blunted response to furosemide and a parallel increase in NCC expression and phosphorylation, supporting the notion that TAL function was defective. These results indicate that OSR1 is crucial not only in the regulation of BP but in renal tubular Na + reabsorption, primarily in the TAL rather than the DCT.
In the global OSR1 +/− mice, markedly decreased p-NKCC1 in aortic tissue and kidneys may contribute to obvious hypotension. NKCC1, as a downstream target of OSR1, has been known to play a pivotal role in BP control through vascular and renal effects, as shown in NKCC1 KO mice (1)(2)(3)(4). On the one hand, inactivation of NKCC1 in blood vessels causes reduced intracellular Cl − concentration and, consequently, decreased Ca 2+ influx through Ltype Ca 2+ channels, which may lead to vessel relaxation and hypotension (23). On the other hand, defective NKCC1 expression in the basolateral membrane of inner medullary collecting ducts and renin-producing juxtaglomerular (JG) cells (24) may cause the impairment of renal Na + reabsorption. However, hyperreninemia and hyperaldosteronism with increased renal Na + transporters, including NKCC2 and NCC observed in NKCC1 KO mice, could help minimize hypotension (3, 4).
Reminiscent of the reduced NKCC1 phosphorylation with reduced aortic contractility in SPAK +/− mice featuring hypotension but normal serum and urine electrolytes (19), the hypotension observed in global OSR1 +/− mice reiterates the importance of OSR1/SPAK-NKCC1 phosphorylation signaling in the vascular tissue on BP control. Because both OSR1 +/− and SPAK +/− mice had normal total NKCC1 but reduced p-NKCC1, it appeared that intact expression of both OSR1 and SPAK was required for adequate NKCC1 phosphorylation in the aortic tissue. Based on the attenuated rather than increased p-SPAK expression in the aortic tissue of OSR1 +/− mice, SPAK phosphorylation may be dependent on the OSR1 activity in the vessels.
In addition to vascular NKCC1, kidney-specific NKCC2 and NCC, two other OSR1 substrates in the TAL and DCT, respectively (5), were also examined in global OSR1 +/− mice. A significant decrease in p-NKCC2 and a parallel increase in both p-SPAK and p-NCC in the OSR1 +/− mice strongly suggested that a salt-wasting phenotype was present. Based on the Guyton type renal function curve (a plot between mean arterial pressure and urinary Na + intake and excretion) (25), OSR1 +/− mice showed a shift to the left, supporting a defect in renal Na + transport. Their renal tubule Na + defect was mild, however, because they did not exhibit negative renal Na + balance even on low-salt diets. Furthermore, their responses to furosemide and thiazide challenges were also normal. Nevertheless, hypotension in OSR1 +/− mice might help dampen the tendency of impaired renal Na + reabsorption. In response to vascular hypotension or renal Na + wasting, one should expect an increase in PRA and PAC. However, PRA and PAC were similar between the WT and OSR1 +/− mice, suggesting that PRA and PAC were inappropriately low in OSR1 +/− mice. Perhaps the reduced p-NKCC2 observed in OSR1 +/− mice may have blunted tubuloglomerular feedback, leading to the impaired release of renin from JG cells as shown in NKCC2 isoform KO mice (26,27). However, we could not exclude the direct regulation of aldosterone secretion by OSR1, which was also abundantly expressed in adrenal tissues.
To focus on the specific role of OSR1 in the regulation of NKCC2 and NCC in the kidney, we further created the KSP-OSR1 KO mice. NKCC2 has three different full-length splice variants called A, B, and F isoforms (5). Inactivating mutations in NKCC2 can cause antenatal BS, an autosomal recessive renal Na + -losing nephropathy with chronic hypokalemia and hypercalciuria with nephrocalcinosis (9). Growth retardation and impaired urine concentration with polyuria are also prominent features in patients who have BS with NKCC2 mutations (28,29) and in total NKCC2null mice (30). Unlike total NKCC2-null mice, NKCC2 A-or Bisoform null mice exhibit normal growth but a mild defect in the TAL, with slightly reduced urine osmolality and no polyuria (26,27). Our KSP-OSR1 −/− mice had marked diminution in p-NKCC2 expression and a partially blunted response to furosemide, indicating a moderate defective function of the TAL. Because mice have very high Na + intake, a defect in renal tubular Na + absorption can be concealed without modification of dietary Na + . On a low-Na + diet, KSP-OSR1 −/− mice showed more urine Na + excretion than WT mice, supporting the presence of a renal tubular defect. The salt phenotype in KSP-OSR1 −/− mice resembled that in NKCC2 A or B KO mice because they all had residual NKCC2 function. Furthermore, the enhanced total and p-SPAK and total p-NCC expression seen in these mice may be compensatory responses to the Na + reabsorption defect upstream.
In the kidneys, we have reported that OSR1 and SPAK are expressed in both TAL and DCT. On the one hand, SPAK is mainly expressed in the cortex, especially in the DCT, and SPAK −/− mice with reduced total and p-NCC expression manifest a GSlike syndrome. On the other hand, OSR1 is predominantly expressed in the medulla, especially in the TAL (19). It is reasonable to expect that NKCC2 but not NCC might be the major substrate of OSR1 in renal tubules. Consistent with this notion, we observed reduced p-NKCC2 but increased total and p-NCC abundance in KSP-OSR1 −/− mice. However, unlike the decreased NCC phosphorylation in the setting of decreased total NCC abundance in SPAK −/− mice (19), the reduced NKCC2 phosphorylation with unchanged total NKCC2 abundance in KSP-OSR1 −/− mice suggests a different mechanism at work. A recent elegant study has shown that adaptor protein (AP)-3, a lysosomal protein, is involved in the sorting process of NCC from the trans-Golgi network to lysosomes and from endosomes to lysosomes (31). Of interest, the most important SPAK/OSR1 phosphoacceptor site on NCC, the T58 residue in the N terminus, is embedded in one of the putative canonical AP-3 binding motifs (YXXθ) and participates in the sorting mechanism of NCC. However, none of the putative canonical AP-3 binding motifs are found in the NKCC2 N terminus. Furthermore, NKCC2 phosphorylation at S130, another important phosphoacceptor, has been shown to be activated by SPAK/OSR1-independent pathways, perhaps via AMP-activated protein kinase, and may also be involved in the regulation of NKCC2 apical sorting (32). KSP-OSR1 −/− and SPAK −/− mice may be used further to explore the mechanisms of NKCC2 and NCC phosphorylation on their protein trafficking.
In addition to Na + wasting, KSP-OSR1 −/− mice had other electrolyte disturbances, such as hypercalciuria and hypokalemia with renal K + wasting. The TAL also accounts for 20-25% of filtrated K + and Ca 2+ reabsorption in parallel with Na + reabsorption. Although K + is directly reabsorbed by NKCC2, the reabsorption of Na + provides the main driving force for Ca 2+ reabsorption via paracellular routes (5). Reduced NKCC2 function would thus impede normal K + and Ca 2+ reabsorption in the TAL, leading to increased urine K + and Ca 2+ excretion. When these mice were fed low-K + diets, significant hypokalemia was maintained, supporting that KSP-OSR1 −/− mice had renal K + wasting.
In conclusion, the analysis of global OSR1 +/− and KSP-OSR1 −/− mice sheds some light on the physiological role of OSR1 in BP regulation and renal Na + handling. OSR1 +/− mice exhibit hypotension associated with the reduced p-SPAK and p-NKCC1 abundance in aortic tissue and decreased p-NKCC2 with an increase in both p-SPAK and p-NCC in the kidney, which is indicative of a salt-wasting phenotype. KSP-OSR1 −/− mice show markedly decreased p-NKCC2 expression in the TAL with a blunted response to furosemide and enhanced p-NCC expression in the DCT, supporting the notion that NKCC2 is the main target of OSR1 and accounts for the BS-like phenotype. These results show that OSR1 plays a dual role in arterial tonicity and renal Na + reabsorption, primarily through NKCC1 and NKCC2, respectively. The development of OSR1 inhibitors suppressing vascular NKCC1 and renal NKCC2 may be a promising direction for antihypertensive therapy in the future.

Materials and Methods
Blood and Urine Analysis and BP Measurement. The phenotype of male mice was evaluated at the age of 12-14 wk. Mice were kept in metabolic cages for 24-h urine collection. Urine osmolarities under ambient conditions were determined using spot urine samples. Blood pressure, plasma and urine electrolytes, and hormone were obtained and measured as previously described (17,33).
Na + and K + Balance Study. The mice were raised on a 12-h day/night cycle, fed a normal rodent chow diet, and given plain drinking water ad libitum. For the evaluation of renal Na + and K + handing, a low-Na + diet or low-K + diet was fed for 6 d (34).
HCTZ and Furosemide Challenge Studies. HCTZ (12.5 mg/kg) and furosemide (15 mg/kg) were administered i.p., respectively, to the mice (19). Urine samples in the 4 h after a single-dose treatment were collected for analysis.
Statistical Analysis. All results are expressed as mean ± SD. Results obtained for the OSR1 +/− or KSP-OSR1 −/− mice were compared with those from their WT littermates by means of the Student t test or, if the data violated a normal distribution, the nonparametric Mann-Whitney test. A P value less than 0.05 was considered to be statistically significant.