Constitutively Active SPAK Causes Hyperkalemia by Activating NCC and Remodeling Distal Tubules

BASIC RESEARCH www.jasn.org

† P. Richard Grimm,* Richard Coleman,* Eric Delpire, and Paul A. Welling*

*Department of Physiology, Maryland Kidney Discovery Center, University of Maryland Medical School, Baltimore, Maryland; and †Department of Anesthesiology, Vanderbilt University Medical School, Nashville, Tennessee

ABSTRACT Aberrant activation of with no lysine (WNK) kinases causes familial hyperkalemic hypertension (FHHt). Thiazide diuretics treat the disease, fostering the view that hyperactivation of the thiazide-sensitive sodium-chloride cotransporter (NCC) in the distal convoluted tubule (DCT) is solely responsible. However, aberrant signaling in the aldosterone-sensitive distal nephron (ASDN) and inhibition of the potassium- excretory renal outer medullary potassium (ROMK) channel have also been implicated. To test these ideas, we introduced kinase-activating mutations after Lox-P sites in the mouse Stk39 gene, which encodes the terminal kinase in the WNK signaling pathway, Ste20-related proline-alanine–rich kinase (SPAK). Renal expression of the constitutively active (CA)-SPAK mutant was specifically targeted to the early DCT using a BASIC RESEARCH DCT-driven Cre recombinase. CA-SPAK mice displayed thiazide-treatable hypertension and hyperkalemia, concurrent with NCC hyperphosphorylation. However, thiazide-mediated inhibition of NCC and consequent restoration of sodium excretion did not immediately restore urinary potassium excretion in CA-SPAK mice. Notably, CA-SPAK mice exhibited ASDN remodeling, involving a reduction in connecting tubule mass and attenuation of epithelial sodium channel (ENaC) and ROMK expression and apical localization. Blocking hyperactive NCC in the DCT gradually restored ASDN structure and ENaC and ROMK expression, concurrent with the restoration of urinary potassium excretion. These findings verify that NCC hyperactivity underlies FHHt but also reveal that NCC-dependent changes in the driving force for potassium secretion are not sufficient to explain hyperkalemia. Instead, a DCT-ASDN coupling process controls potassium balance in health and becomes aberrantly activated in FHHt.

J Am Soc Nephrol 28: 2597–2606, 2017. doi: https://doi.org/10.1681/ASN.2016090948

With no lysine (WNK) kinase signaling cascades The thiazide-sensitive sodium chloride cotrans- form a molecular switch that adjusts the aldoste- porter (NCC) is believed to be the primary effector of rone response of the kidney to either retain sodium WNK signaling in the kidney,7 although WNK kinases or excrete potassium depending on physiologic may also directly regulate potassium transport in the need. According to current understanding, WNK ASDN.8 Theideaevolvedfromtheseminaldiscovery signaling pathways differently regulate electrolyte that activating mutations in WNK1 and WNK4 cause a transport processes in the distal tubule segments, allowing sodium to be absorbed with chloride in the distal convoluted tubule (DCT) or exchanged Received September 6, 2016. Accepted February 27, 2017. with potassium in the aldosterone-sensitive distal Published online ahead of print. Publication date available at nephron (ASDN). WNK activation by Angiotensin www.jasn.org. II in the high aldosterone state of intravascular vol- Correspondence: Dr.PaulA.Welling,UniversityofMarylandSchool 1–4 ume depletion supports salt absorption in the of Medicine, Department of Physiology, Bressler Research Building DCT, whereas potassium-dependent inhibition of 5-029, 655 West Baltimore Street, Baltimore, MD 21201. Email: the WNKs5,6 favors potassium excretion by the [email protected] ASDN in potassium excess. Copyright © 2017 by the American Society of Nephrology

J Am Soc Nephrol 28: 2597–2606, 2017 ISSN : 1046-6673/2809-2597 2597 BASIC RESEARCH www.jasn.org

hypertension and hyperkalemia in transgenic WNK4FHHt mice,10 confirming that NCC is necessary for FHHt. However, it has not been proven that hyperactivation of NCC is sufficient to explain the disease. According to this view, WNKs influence potassium secretion solely by activating NCC and limiting sodium delivery to the ASDN for electrogenic sodium-potassium exchange. Until now, rigorous investiga- tion of the “sodium-delivery hypothesis” in mouse models of WNK-inducedFHHthasnotbeenpossible,becauseWNK kinases are expressed in the ASDN,11 where they may have direct effects on the potassium secretory machinery.12–15 To test these ideas, we developed a DCT-specific knock-in mouse model of constitutively active (CA) Ste20-related proline-alanine–rich kinase (SPAK) of the STK39 gene, the terminal kinase in the WNK-NCC signaling pathway.10,16,17 On activation by WNKs, phospho-SPAK binds to and phos- phoactivates NCC.18–21 SPAK knockout (KO)18,19,21 or loss of kinase-function17 mice display a salt-wasting phenotype identical to NCC null mice, phenocopying Gitelman syndrome in humans, which is caused by loss of function mutations in SLC12A3, encoding NCC.22 Here, we show that specificex- pression of CA-SPAKwithin the early distal convoluted tubule (DCT1) drives FHHt. Surprisingly, tubular remodeling of the ASDN and inhibition of renal outer medullary potassium (ROMK) and epithelial sodium channel (ENaC) accompany

Figure 1. CA-SPAK is targeted to DCT1. (A) Representative Western blots of CA-SPAK (HA tag) and WT SPAK in the kidney cortex of WT and CA-SPAK knockin mice with anti-HA or -SPAK antibodies. (B) Confocal microscopy and colocalization of WT or CA-SPAK (red) with Parv (green) to identify the DCT1. *Antibodies to NKCC2 (not shown) identified the thick ascending limb. DAntibodies to NCC and calbindin (not shown) identified the DCT2/CNT. (C) Tubule segment expression of SPAK expression was quantified as the percentage of cells within a specific nephron segment that has SPAK (n=4 animals per genotype, .100 cells per nephron segment per animal counted). rarefamilial disorder of hyperkalemia and hypertension that is Figure 2. NCC is hyperphosphorylated in CA-SPAK mice. (A) treatable with thiazide diuretics (familial hyperkalemic hy- Representative Western blots and (B) quantitative analysis of 9 pertension [FHHt] or pseudohypoaldosteronism type II). tNCC, pNCC (T58), and the ratio of pNCC to tNCC in WT (red) Because thiazide diuretics specificallyinhibitNCC,ithas versus CA-SPAK (blue) mice. Bars and wickers are means6SEM, been argued that FFHt is manifested as a sole consequence and circles are individual data points from each mouse (n=6 per of aberrant NCC activation. Genetic ablation of NCC blocks genotype). *P,0.05.

2598 Journal of the American Society of Nephrology J Am Soc Nephrol 28: 2597–2606, 2017 www.jasn.org BASIC RESEARCH

used to specifically drive kidney Cre recombinase expression in DCT1 (Supplemental +/+ Figure 1D). CA-SPAK mice (CA-SPAK 3 +/2 Parv-Cre )werecomparedwithSPAKKO +/+ 2/2 littermates (CA-SPAK 3 Parv-Cre ) and wild-type (WT) mice expressing Parv- +/2 Cre (WT, Parv-Cre ). As detected with anti-HA antibodies, the HA epitope–tagged knock-in protein was found in the kidney cortex but was not found Figure 3. Increased NCC and pNCC (T58) abundance is restricted to the DCT1 in CA- in the medulla of CA-SPAK mice, and it was SPAK mice. (A) Colocalization of tNCC (red) and pNCC (green) with Parv (cyan). Dashed absent in WT and SPAK KO kidneys as ex- lines demarcate the transition between the DCT1 and the DCT2 in contiguous nephron pected. Anti-SPAK immunoblots corrobo- segments. Mean P,0.05, WT vs. CA-SPAK (B) Quantification of the apical membrane rated that CA-SPAK proteins have the same delimited pixel intensity of pNCC (T58) and tNCC. Bars and wickers are means6SEM, molecular masses as WT SPAK, including and circles are individual data points from each mouse (n=6 animals per genotype the full-length protein (60 kD)18 and smaller with .60 cells per segment measured per animal). *P,0.05. species (Figure 1A). Because CA-SPAK is en- gineered from the entire SPAKopen-reading frame, the smaller SPAK proteins are likely NCC activation in the DCT1, revealing a mechanism to ex- proteolytic fragments26 rather than alternative SPAK gene prod- plain urinary potassium retention in FHHt. ucts.19 Quantitative confocal microscopy with tubule-specific markers (Figure 1, B and C) revealed that CA-SPAK expression is restricted to Parv-positive DCT cells, contrasting WT SPAK, RESULTS which is also found in nearly all DCT2, thick ascending limb, connecting tubule, and cortical collecting duct cells. These A CA-SPAK, containing phosphomimetic mutations at the key findings verify that CA-SPAK is appropriately expressed and activation sites (T243E and S383D),23 was introduced in the specifically targeted to DCT1 on an SPAK null background. mouse SPAK gene (Supplemental Figure 1A). CA-SPAK was NCC phosphorylation was assessed at the key activation site HA epitope tagged for detection and placed after an Lox-P flanked (T58).20,27 As shown in the representative immunoblot selection cassette (Supplemental Figure 1B), allowing targeted (Figure 2A) and quantified by densitometry (Figure 2B), phos- expression with a cell-specificCrerecombinase(Supplemental phosodium chloride cotransporter (pNCC) abundance is sig- Figure 1C). Taking advantage of DCT1-specific expression of par- nificantly greater in CA-SPAK mice than WT (and SPAK KO) valbumin (Parv) within the kidney,24,25 the Parv promoter was mice (Supplemental Figure 2). Total sodium chloride cotransporter (tNCC) protein abundance is also increased, presumably a consequence of phosphorylation-dependent stabilization Table 1. Plasma electrolytes and renal function assessment in WT, CA-SPAK, 28 and SPAK KO on control diet of the transporter and hypertrophy of the DCT1 (see below). An increase in the Electrolytes WT, n=12 CA-SPAK, n=12 SPAK KO, n=10 pNCC-to-tNCC ratio was observed, con- Plasma sistent with enhanced SPAK activity. Be- + 6 6 6 Na , mmol/L 144.4 0.8 143.8 0.9 144.7 0.8 cause CA-SPAK mice display hyperkalemia K+, mmol/L 3.8660.32 4.8260.44a 3.4660.35a 2 and hypertension (see below) that should Cl , mmol/L 108.660.8 115.261.2a 107.460.6 29,30 pH 7.39460.019 7.29860.016a 7.41560.018 suppress pNCC, the sustained increase 2 a a in pNCC is indicative of constitutive SPAK HCO3 , mmol/L 23.9760.47 19.8960.71 25.3960.46 Hematocrit 45.5160.32 43.3260.29a 50.2160.54a activity. Confocal microscopy (Figure 3A) BUN, mg/dl 23.660.3 21.660.4a 28.060.3a and quantitative image analysis (Figure 3B) Plasma renin activity, ng/ml per h 27.962.7 15.863.2a 50.864.2a confirmed that the increases in apical 24-h Urine aldosterone, pmol 16.761.6 19.362.4 22.661.1a membrane pNCC and tNCC are confined Urine to DCT1, where CA-SPAK is specifically 6 6 6 FENa,% 0.69 0.13 0.62 0.12 0.64 0.18 expressed. In neighboring DCT2 cells, 6 6 a 6 a FEK,% 21.47 3.11 10.33 1.72 29.62 2.38 pNCC is suppressed, presumably a result 6 6 a 6 FECl,% 1.18 0.27 1.01 0.32 1.85 0.29 of the SPAK null background. These find- Creatinine clearance, ml/min 251.8617.2 233.5619.9 230.7618.9 ings indicate that targeted expression of One-way ANOVA with post hoc multiple group comparisons via Tukey tests. FE , fractional excretion Na CA-SPAK in DCT1 is sufficient to drive of sodium; FEK, fractional excretion of potassium; FECl, fractional excretion of chloride. aP,0.05 relative to the WT is considered significant. NCC hyperphosphorylation.

J Am Soc Nephrol 28: 2597–2606, 2017 SPAK Activation Causes Familial Hyperkalemic Hypertension 2599 BASIC RESEARCH www.jasn.org

Figure 4. CA-SPAK mice develop HCTZ-remediated salt-sensitive hypertension and hyperkalemia. (A) Average systolic BP (SBP) during the awake period in the presence of vehicle or HCTZ (25 mg/kg body wt) after 3 days of treatment on control diet or high-sodium diet (HNa). (B) Plasma potassium concentration under basal conditions or after dietary potassium loading (HK) in the presence of vehicle or

HCTZ (3 days treatment). (C) Urinary potassium excretion relative to amount ﬁltered (fractional excretion of potassium [FEK]) in the presence of vehicle or HCTZ (3 days treatment). WT (red) and CA-SPAK (blue) bars and wickers are means6SEM, and circles are individual data points from each mouse (n=6 per genotype per treatment). *P,0.05, WT versus CA-SPAK undergoing same treatment; **P,0.05, vehicle versus HCTZ in same genotype; ***P,0.05, control diet versus experimental diet in the same genotype.

CA-SPAK mice exhibit the salient electrolyte abnormalities Renal responses to dietary potassium loading and HCTZ of FHHt (Table 1), identical to mice bearing FHHt mutations were evaluated to explore the mechanism underlying hyper- in WNK131 and WNK4.10,32 Hyperkalemia and metabolic ac- kalemia. In contrast to WT mice, which tolerate dietary idosis are observed without changes in plasma aldosterone or creatinine clearance. They also display a reduction in hematocrit and BUN, consistent with intravascular fluid expan- sion from unrestrained renal salt absorption. By contrast, +/+ 2 /2 CA-SPAK littermates without Cre (Parv-Cre ) exhibit a phenotype resembling Gitelman syndrome and SPAK KO mice,18,19,21 with metabolic alkalosis, hypokalemia, and increased BUN and hematocrit. Thus, DCT1-specificex- pression of CA-SPAK in an SPAK KO background is sufficient to reverse the electrolyte disorder of Gitelman syndrome to FHHt. Telemetric measurements of BP in conscious, unrestrained mice (Figure 4, Supplemental Figure 3) revealed that CA-SPAK mice (Supplemental Figure 3A) have significantly higher systolic and diastolic BPs than WT mice. Hypertension was especially pro- nounced during the active, awake period (Figure 4A, Supplemental Figure 3). Dietary salt loading exacerbated hypertension in CA- SPAK mice but had no effect on BP in WT mice (Figure 4A). Salt-sensitive hypertension in the face of low renin (Table 1) is consistent with an aberrant gain in renal sodium absorption. Remarkably, hydrochlorothiazide (HCTZ; 25 mg/kg body wt intraperitoneally) quickly normalized (12 hours) BP to WT levels, coincident with correction of the HCT and BUN. Thiazides can Figure 5. Delayed restoration of urinary potassium excretion and plasma K+ in CA-SPAK mice. Fractional excretion of urinary so- have modest off-target effects, including inhibition of carbonic an- dium (FE ; blue; left axis) and fractional excretion of potassium hydrase33,34 and inhibition of pendrin/NDCBE in B-intercalated Na (FEK; red; right axis) in (A) WT and (C) CA-SPAK mice after HCTZ 35 fi cells. However, HCTZ speci cally inhibits NCC with the dose treatment with corresponding changes in plasma potassium used here. These findings indicate NCC hyperactivation, and ex- concentrations in (B) WT and (D) CA-SPAK. Dots and wickers are aggerated salt absorption in DCT1 is sufficient to cause hyperten- means6SEM (n=6). *P,0.05 for treatment versus day 0 within sion (Figure 4A). genotype.

2600 Journal of the American Society of Nephrology J Am Soc Nephrol 28: 2597–2606, 2017 www.jasn.org BASIC RESEARCH

excretion should abruptly increase in CA- SPAK mice when sodium excretion rapidly increases with NCC inhibition. Surpris- ingly, we found that urinary potassium excretion and plasma potassium are slowly restored to WT levels (Figure 5, A and B) over3daysofHCTZtreatmentinCA- SPAK mice (Figure 5, C and D), long after the prompt natriuresis is induced (Figure 5, A, WTand C, CA-SPAK, Supplemental Fig- ure 5). Thus, decreased sodium delivery does not explain compromised potassium excretion in CA-SPAK mice. The distal nephron undergoes extensive cellular remodeling in response to inhibition of NCC,18,19,36–38 characterized by DCT atrophy and ASDN hypertrophy. To determine if opposite structural changes occur in the CA-SPAK mice, confocal microscopy with distal segment–specific markers was performed, and distal tubule morphology was evaluated quantitatively by stereology. As summarized in Figure 6, we observed that the tubule length and cross-sectional area (Figure 6, A and B) of the DCT1 (Parv/NCC-positive tubules) (green in Figure 6A) expand in CA-SPAK mice, similar to the increase in DCT1 mass 31 Figure 6. ASDN mass is reduced in DCT1-targeted CA-SPAK mice. (A) Representative in FHHt-WNK1 mice and FHHt-WNK4 images of distal nephron marker staining (DCT1 in green and CNT in red) showing transgenic mice.10,32 This was paralleled changes structural changes. (B) Quantitative assessment of cortical distal nephron by a decrease in length and cross-sectional segments (DCT1, DCT2, CNT, and CCD) revealed that the length and area of DCT1 area of calbindin/AQP2-labeled CNT (red segments increased in CA-SPAK mice, with a commensurate decrease in the length in Figure 6A) but not the DCT2 (calbindin and area of CNT segments. (C) The structural changes in CA-SPAK DCT1 and CNT are NCC positive), CNT1 (calbindin positive n normalized after 3 days of HCTZ ( =6 animals per genotype with area of all labeled but NCC and AQP2 negative),39 or CCD P, nephron segments measured). * 0.05. (AQP2 positive),39 consistent with a specific decrease in CNT mass. Because potassium secretion is most avid in this early potassium loading, hyperkalemia in CA-SPAK is exacerbated part of the ASDN,40 the reduction in the CNT is expected to by a potassium-rich diet (Figure 4B). CA-SPAK mice exhibit severely compromise potassium excretion in the CA-SPAK lower rates of urinary potassium excretion (fractional excretion mice. Importantly, HCTZ treatment reversed the structural of potassium in Figure 5C or UKVinSupplementalFigure4B) remodeling (Figure 6, B and C), indicating that NCC over- than WT mice and SPAK KO littermates (Figure 5A, Table 1) activity is necessary to drive the remodeling response. and are unable to develop a significant transtubular potassium The reduction in CNT mass was paralleled by a decrease in gradient (Supplemental Figure 5A), indicative of impaired po- ENaC, which provides a large electrical driving force for potassium secretion from the ASDN. HCTZ corrects the hyper- tassium secretion through ROMK and BK channels in the kalemia (Figure 4B) and restores urinary potassium excretion CNT.41,42 Indeed, benzamil-sensitive (4.3 mg/kg body wt in- (Figure 4C) and transtubular potassium gradient (Supple- traperitoneally) sodium reabsorption was significantly re- mental Figure 5B) to WT levels. Taken together, these findings duced (Figure 7A) in CA-SPAK mice. Biochemical assessment indicate that NCC hyperactivity underlies urinary potassium of ENaC revealed a decrease in a- (full-length) and g-ENaC retention in CA-SPAK mice. but not the b-subunit (Figure 7, B and C). Furthermore, the To test whether reduced distal sodium delivery limits po- cleaved form of a-ENaC was also reduced,43 raising the tassium excretion, we monitored the temporal natriuretic and possibility that post-translational activation may also be com- kaliuretic responses to HCTZ (Figure 5, Supplemental Figure promised. Thus, ENaC attenuation likely contributes to po- 6). The sodium delivery hypothesis predicts that potassium tassium retention in CA-SPAK mice.

J Am Soc Nephrol 28: 2597–2606, 2017 SPAK Activation Causes Familial Hyperkalemic Hypertension 2601 BASIC RESEARCH www.jasn.org

Figure 7. CA-SPAK mice have blunted sensitivity to benzamil and reduced ENaC expression. (A) Although the rate of urinary sodium excretion is not different between WT and CA-SPAK mice, benzamil (4.3 mg/kg body wt intraperitoneally) signiﬁcantly increases WT urinary sodium excretion rate (UNaV) above that in CA-SPAK (n=6). (B and C) Quantitative assessment of ENaC subunit expression was performed by Western blotting and showed reduced a-andg-ENaC abundance (n=4 animals per genotype). WT (red) and CA-SPAK (blue) bars and wickers are means6SEM, and circles are individual data points of each mouse. *P,0.05.

We also evaluated the two potassium channels that mediate transgenic mice provided compelling evidence that enhanced potassium secretion, BK (Figure 8) and ROMK (Figure 9). Western NCC activity in the DCT is necessary to drive the FHHt phe- blot analysis revealed that the accessory B1 subunit is decreased, but notype. Our studies reveal that aberrant activation of the NCC the pore-forming a-subunit and accessory B4 subunit are surprisingly increased in CA-SPAK kidney cortical homogenates compared with the WT. Thus, urinary potassium retention in CA-SPAK mice cannot be explained by BK downregulation. By contrast, ROMK channels, which are especially regulated in DCT2 and CNT to maintain potassium excretion,39 were signiﬁcantly reduced in CA-SPAK mice. A striking reduction in the abundance of the ROMK protein was observed within the CA-SPAK renal cortex compared with WT (Figure 9, A and B) or SPAK KO littermates (Supplemental Figure 6, A and B) but not in the Loop of Henle– enriched medulla (Figures 9A and B, Supplemental Figure 6, C and D). Confocal microscopy revealed that apical ROMK localization and abundance along all segments of the ASDN were strikingly reducedintheCA-SPAKmicecomparedwithWT(Figure9C)or SPAK KO mice (not shown). It might be argued that WNKs could be activated in CA-SPAK mice as part of a compensatory response that predisposes the mice to inhibit ROMK. However, we found that WNK1, pWNK1, and WNK4 are actually reduced (Supple- mental Figure 7), indicating that ROMK must be inhibited by a WNK-independent pathway in CA-SPAK mice. Signiﬁcantly, HCTZ restored ROMK expression to normal levels within 3 days of treatment, coincident with gradual correction of hyperkalemia and potassium excretion (Figure 9, E and F) and restoration of CNT mass (Figure 6C). These observations reveal that enhanced NCC activity in the DCT1 not only drives dystrophic CNTremod- eling but also, inhibits apical ROMK localization and expression along the entire ASDN.

Figure 8. Cortical big potassium channel (BK) expression is reduced DISCUSSION in CA-SPAK mice. Cortical BK-a,-b1, and -b4 protein abundance was assessed by (A) Western blots and (B) quantiﬁcation (n=4). WT The ﬁnding of Lalioti et al.10 that genetic ablation of NCC (red) and CA-SPAK (blue) bars and wickers are means6SEM, and abolishes hypertension and hyperkalemia in WNK4-FHHt circles are individual data points of each mouse. *P,0.05.

2602 Journal of the American Society of Nephrology J Am Soc Nephrol 28: 2597–2606, 2017 www.jasn.org BASIC RESEARCH

balance, intravascular volume, and BP. They also reinforce the concept that renal potassium excretion from the ASDN is strongly and negatively influenced by NCC activity but not by the suspected mechanism of reduced sodium delivery. We found that an unexpected distal tubule (DCT-ASDN) coupling process in- versely links NCC activity to urinary potassium excretion. Traditionally, it has been thought that NCC activation negatively influences potassium secretion by limiting sodium delivery to the ASDN. Contrary to the textbook view, we found that targeted CA- SPAK expression in the DCT1 is accompa- nied by an ASDN remodeling program characterized by ASDN dystrophy and inhibition of ENaC and ROMK. Normally, ENaC42 and ROMK are exquisitely regulated in the ASDN,8 allowing urinary potassium excretion to match dietary potassium intake.40 In CA-SPAK mice, ROMK expression becomes so compromised that the ASDN is unable to efficiently secrete potassium, even when sodium delivery is acutely activated. Gradual correction of urinary potassium excretion and the hyperkalemia in the CA-SPAK mice with thiazide diuretics coincides with the return of ENaC and ROMK expression to WT levels. Taken together, these finding indicate that structural changes in the ASDN and diminished ENaC and ROMK apical membrane expression limit urinary potassium excretion when NCC becomes hyperactive, providing a new mechanism to explain Figure 9. ROMK expression within the ASDN is reduced in DCT1-targeted CA-SPAK how aberrant activation of NCC causes mice. Cortical ROMK protein abundance was assessed by (A) Western blots and (B) thiazide-correctable hyperkalemia in FHHt. quantification (n=6). *P,0.05. (C) Immunolocalization of ROMK in typical DCT2, CNT, The distal tubule coupling process likely and CCD along with (D) quantification of apical ROMK abundance along individual influences potassium balance in health. Re- segments of the ASDN (n=4 animals per genotype and .75 cells per nephron seg- cent breakthrough studies revealed that low P, ment). * 0.05. (E) Representative cortical ROMK Western blots and (F) quantitative plasma potassium directly activates WNK- n analysis in CA-SPAK and WT animals after 3 days of HCTZ treatment ( =6). WT (red) SPAK signaling in the DCT to stimulate and CA-SPAK (blue) bars and wickers are means6SEM, and circles are individual data NCC,46 but it has been mysterious how po- points of each mouse. *P,0.05. tassium sensing in the DCT is communi- cated downstream to limit potassium in the DCT1 is sufficient to initiate FHHt. Using a Parv pro- secretion by the ASDN. Contrary to the widely held view, moter–driven Cre recombinase, we specifically targeted renal careful kinetic determination of sodium-dependent potas- expression of a CA-SPAK mutant to DCT1. In contrast to the sium secretion indicates that physiologic changes in sodium negligible effects of overexpressing NCC in the DCT,44 CA- delivery are not sufficient to significantly alter the driving SPAK mice phenocopy FHHt. Although Parv-positive cells forces for potassium secretion47 or explain NCC-dependent outside the kidney may express CA-SPAK, hypertension and modulation of potassium secretion from the ADSN.48 We hyperkalemia were completely reversed by inhibiting NCC. propose that NCC-dependent changes in sodium delivery These findings reinforce the importance of the early distal tu- shape longer-term “remodeling” of the potassium secretory bule and phosphoactivation of NCC45 in the control of salt machinery in the ASDN, reminiscent of the way that sodium

J Am Soc Nephrol 28: 2597–2606, 2017 SPAK Activation Causes Familial Hyperkalemic Hypertension 2603 BASIC RESEARCH www.jasn.org modulates DCT structure and transport.49 It will be interesting to DISCLOSURES learn if CNT-specificENaCKOmicehavedystrophicCNTaspre- None. dicted if sodium is the sole mediator of the remodeling process. Other transtubular coupling mechanisms, such as paracrine com- munication, which has been implicated in ASDN hypertrophy REFERENCES in SPAK KO mice,38 should also be considered. WNK-SPAK– dependent activation of NCC by low potassium50 together with the 1. Castañeda-Bueno M, Cervantes-Pérez LG, Vázquez N, Uribe N, Kantesaria S, Morla L, Bobadilla NA, Doucet A, Alessi DR, Gamba G: DCT-ASDN inverse coupling mechanism described here would pro- 1 2 vide an effective means to conserve sodium and potassium, especially Activation of the renal Na :Cl cotransporter by angiotensin II is a WNK4-dependent process. Proc Natl Acad Sci USA 109: 7929–7934, in fasting states when dietary sodium and potassium are limiting. 2012 It is important to recognize that the ASDN phenotype of CA- 2. Gonzalez-Villalobos RA, Janjoulia T, Fletcher NK, Giani JF, Nguyen MT, SPAK mice may be different than that of some of the WNK FFHt Riquier-Brison AD, Seth DM, Fuchs S, Eladari D, Picard N, Bachmann S, mouse models. Altered regulation of ROMK and ENaC has been Delpire E, Peti-Peterdi J, Navar LG, Bernstein KE, McDonough AA: The J Clin Invest reported, but the phenotypes are surprisingly divergent manners absence of intrarenal ACE protects against hypertension. 123: 2011–2023, 2013 depending on the WNK FFHt model. Recently, Zhang and co- 3. Takahashi D, Mori T, Nomura N, Khan MZ, Araki Y, Zeniya M, Sohara E, 51 workers found that ROMK and ENaC activities are suppressed Rai T, Sasaki S, Uchida S: WNK4 is the major WNK positively regulating inthelateDCT/earlyCNTofmicetransgenicforWNK4FFHt, NCC in the mouse kidney. Biosci Rep 34: e00107, 2014 identical to CA-SPAK mice. By contrast, WNK4 FFHt mice were 4. van der Lubbe N, Lim CH, Fenton RA, Meima ME, Jan Danser AH, Zietse reported to have increased ENaC activity and no change in ROMK R, Hoorn EJ: Angiotensin II induces phosphorylation of the thiazide- protein abundance, although ROMK was misidentified with a sensitive sodium chloride cotransporter independent of aldosterone. Kidney Int 79: 66–76, 2011 fi 32 nonspeci cantibody. WNK1 FFHt mice exhibit another 5. Rengarajan S, Lee DH, Oh YT, Delpire E, Youn JH, McDonough AA: 1 ASDN phenotype characterized by a decrease in ROMK abun- Increasing plasma [K ] by intravenous potassium infusion reduces NCC dance in the late DCT and no change in ENaC.31 On the basis of phosphorylation and drives kaliuresis and natriuresis. Am J Physiol these variable observations and findings in vitro that WNKs have Renal Physiol 306: F1059–F1068, 2014 the capacity to increase ENaC activity52,53 and decrease ROMK,8 it 6. Sorensen MV, Grossmann S, Roesinger M, Gresko N, Todkar AP, Barmettler G, Ziegler U, Odermatt A, Loffing-Cueni D, Loffing J: Rapid seems likely that disparate ASDN phenotypes are shaped by effects dephosphorylation of the renal sodium chloride cotransporter in re- of different WNK signaling pathways in the ASDN. Our results in sponse to oral potassium intake in mice. Kidney Int 83: 811–824, 2013 CA-SPAK should not be interpreted that WNK signaling in the 7. Hadchouel J, Ellison DH, Gamba G: Regulation of renal electrolyte ASDN does not play a role in FHHt or physiologic regulation of transport by WNK and SPAK-OSR1 kinases. Annu Rev Physiol 78: 367– ROMK and ENaC. 389, 2016 8. Welling PA: Roles and regulation of renal K channels. Annu Rev Physiol Our understanding of renal potassium balance has been – fl 78: 415 435, 2016 transformed by a con uence of recent discoveries, revealing a mo- 9. Wilson FH, Disse-Nicodème S, Choate KA, Ishikawa K, Nelson-Williams lecularswitch,orchestratedbyWNK-SPAKsignaling,thatshiftsthe C, Desitter I, Gunel M, Milford DV, Lipkin GW, Achard JM, Feely MP, balance of distal nephron transport activities between the DCTand Dussol B, Berland Y, Unwin RJ, Mayan H, Simon DB, Farfel Z, the ASDN. This study reveals that WNK-SPAK signaling in the Jeunemaitre X, Lifton RP: Human hypertension caused by mutations in Science – DCT1 is sufficient to drive the switch mechanism to activate salt WNK kinases. 293: 1107 1112, 2001 10. Lalioti MD, Zhang J, Volkman HM, Kahle KT, Hoffmann KE, Toka HR, transport from the DCTand suppress potassium secretion from the Nelson-Williams C, Ellison DH, Flavell R, Booth CJ, Lu Y, Geller DS, ASDN through a surprising transtubule coupling mechanism. Lifton RP: Wnk4 controls blood pressure and potassium homeostasis via regulation of mass and activity of the distal convoluted tubule. Nat Genet 38: 1124–1132, 2006 11. O’Reilly M, Marshall E, Macgillivray T, Mittal M, Xue W, Kenyon CJ, CONCISE METHODS Brown RW: Dietary electrolyte-driven responses in the renal WNK kinase pathway in vivo. J Am Soc Nephrol 17: 2402–2413, 2006 Methods were performed as previously reported18,38 and are detailed 12. Cai H, Cebotaru V, Wang YH, Zhang XM, Cebotaru L, Guggino SE, in the Supplemental Methods and Supplemental Tables 1 and 2. An- Guggino WB: WNK4 kinase regulates surface expression of the human Kidney Int imal studies were performed in adherence to the National Institutes sodium chloride cotransporter in mammalian cells. 69: 2162–2170, 2006 of Health Guide for the Care and Use of Laboratory Animals and 13. Lazrak A, Liu Z, Huang CL: Antagonistic regulation of ROMK by long approved by the University of Maryland School of Medicine Institu- and kidney-specific WNK1 isoforms. Proc Natl Acad Sci USA 103: tional Animal Care and Use Committee. 1615–1620, 2006 14. Murthy M, Cope G, O’Shaughnessy KM: The acidic motif of WNK4 is crucial for its interaction with the K channel ROMK. Biochem Biophys Res Commun 375: 651–654, 2008 ACKNOWLEDGMENTS 15. Fang L, Garuti R, Kim BY, Wade JB, Welling PA: The ARH adaptor protein regulates endocytosis of the ROMK potassium secretory channel in mouse kidney. JClinInvest119: 3278–3289, 2009 Funding for this project was provided National Institutes of Health 16. Moriguchi T, Urushiyama S, Hisamoto N, Iemura S, Uchida S, Natsume Grants DK63049, DK54231, and DK093501 (to E.D. and P.A.W.). T, Matsumoto K, Shibuya H: WNK1 regulates phosphorylation of

2604 Journal of the American Society of Nephrology J Am Soc Nephrol 28: 2597–2606, 2017 www.jasn.org BASIC RESEARCH

cation-chloride-coupled cotransporters via the STE20-related kinases, analysis of a Wnk4(D561A/+) knockin mouse model. Cell Metab 5: 331– SPAK and OSR1. JBiolChem280: 42685–42693, 2005 344, 2007 17. Rafiqi FH, Zuber AM, Glover M, Richardson C, Fleming S, JovanovicS, 33. Puscas I, Coltau M, Baican M, Pasca R, Domuta G: The inhibitory effect JovanovicA,O’Shaughnessy KM, Alessi DR: Role of the WNK-activated of diuretics on carbonic anhydrases. Res Commun Mol Pathol Phar- SPAK kinase in regulating blood pressure. EMBO Mol Med 2: 63–75, macol 105: 213–236, 1999 2010 34. Schaeffer P, Vigne P, Frelin C, Lazdunski M: Identification and phar- 18. Grimm PR, Taneja TK, Liu J, Coleman R, Chen YY, Delpire E, Wade JB, macological properties of binding sites for the atypical thiazide di- Welling PA: SPAK isoforms and OSR1 regulate sodium-chloride co- uretic, indapamide. Eur J Pharmacol 182: 503–508, 1990 transporters in a nephron-specificmanner.J Biol Chem 287: 37673– 35. Leviel F, Hübner CA, Houillier P, Morla L, El Moghrabi S, Brideau G, 37690, 2012 Hassan H, Parker MD, Kurth I, Kougioumtzes A, Sinning A, Pech V, 19. McCormick JA, Mutig K, Nelson JH, Saritas T, Hoorn EJ, Yang CL, Riemondy KA, Miller RL, Hummler E, Shull GE, Aronson PS, Doucet A, 1 Rogers S, Curry J, Delpire E, Bachmann S, Ellison DH: A SPAK isoform Wall SM, Chambrey R, Eladari D: The Na -dependent chloride- switch modulates renal salt transport and blood pressure. Cell Metab bicarbonate exchanger SLC4A8 mediates an electroneutral Na+ 14: 352–364, 2011 reabsorption process in the renal cortical collecting ducts of mice. J 20. Richardson C, Rafiqi FH, Karlsson HK, Moleleki N, Vandewalle A, Clin Invest 120: 1627–1635, 2010 Campbell DG, Morrice NA, Alessi DR: Activation of the thiazide-sensitive 36. Loffing J, Loffing-Cueni D, Hegyi I, Kaplan MR, Hebert SC, Le Hir M, 1 2 Na -Cl cotransporter by the WNK-regulated kinases SPAK and OSR1. Kaissling B: Thiazide treatment of rats provokes apoptosis in distal tu- J Cell Sci 121: 675–684, 2008 bule cells. Kidney Int 50: 1180–1190, 1996 21.YangSS,LoYF,WuCC,LinSW,YehCJ,ChuP,SytwuHK,UchidaS, 37. Loffing J, Vallon V, Loffing-Cueni D, Aregger F, Richter K, Pietri L, Sasaki S, Lin SH: SPAK-knockout mice manifest Gitelman syndrome Bloch-Faure M, Hoenderop JG, Shull GE, Meneton P, Kaissling B: Al- and impaired vasoconstriction. JAmSocNephrol21: 1868–1877, tered renal distal tubule structure and renal Na(+) and Ca(2+) handling 2010 in a mouse model for Gitelman’s syndrome. JAmSocNephrol15: 22. Simon DB, Nelson-Williams C, Bia MJ, Ellison D, Karet FE, Molina AM, 2276–2288, 2004 Vaara I, Iwata F, Cushner HM, Koolen M, Gainza FJ, Gitleman HJ, Lifton 38. Grimm PR, Lazo-Fernandez Y, Delpire E, Wall SM, Dorsey SG, Weinman RP: Gitelman’s variant of Bartter’s syndrome, inherited hypokalaemic EJ, Coleman R, Wade JB, Welling PA: Integrated compensatory net- alkalosis, is caused by mutations in the thiazide-sensitive Na-Cl co- work is activated in the absence of NCC phosphorylation. JClinInvest transporter. Nat Genet 12: 24–30, 1996 125: 2136–2150, 2015 23. Gagnon KB, Delpire E: On the substrate recognition and negative 39. Wade JB, Fang L, Coleman RA, Liu J, Grimm PR, Wang T, Welling PA: 1 1 2 regulation of SPAK, a kinase modulating Na -K -2Cl cotransport Differential regulation of ROMK (Kir1.1) in distal nephron segments by activity. Am J Physiol Cell Physiol 299: C614–C620, 2010 dietary potassium. Am J Physiol Renal Physiol 300: F1385–F1393, 2011 24. Bindels RJ, Timmermans JA, Hartog A, Coers W, van Os CH: Calbindin- 40. Welling PA: Regulation of renal potassium secretion: Molecular D9k and parvalbumin are exclusively located along basolateral mem- mechanisms. Semin Nephrol 33: 215–228, 2013 branes in rat distal nephron. JAmSocNephrol2: 1122–1129, 1991 41. Rubera I, Loffing J, Palmer LG, Frindt G, Fowler-Jaeger N, Sauter D, 25. Loffing J, Loffing-Cueni D, Valderrabano V, Kläusli L, Hebert SC, Rossier Carroll T, McMahon A, Hummler E, Rossier BC: Collecting duct-specific BC, Hoenderop JG, Bindels RJ, Kaissling B: Distribution of transcellular gene inactivation of alphaENaC in the mouse kidney does not impair calcium and sodium transport pathways along mouse distal nephron. sodium and potassium balance. JClinInvest112: 554–565, 2003 Am J Physiol Renal Physiol 281: F1021–F1027, 2001 42. Christensen BM, Perrier R, Wang Q, Zuber AM, Maillard M, Mordasini 26. Markadieu N, Rios K, Spiller BW, McDonald WH, Welling PA, Delpire E: D, Malsure S, Ronzaud C, Stehle JC, Rossier BC, Hummler E: Sodium Short forms of Ste20-related proline/alanine-rich kinase (SPAK) in the and potassium balance depends on aENaC expression in connecting kidney are created by aspartyl aminopeptidase (Dnpep)-mediated tubule. JAmSocNephrol21: 1942–1951, 2010 proteolytic cleavage. JBiolChem289: 29273–29284, 2014 43. Hughey RP, Bruns JB, Kinlough CL, Harkleroad KL, Tong Q, Carattino 27. Pacheco-Alvarez D, Cristóbal PS, Meade P, Moreno E, Vazquez N, MD, Johnson JP, Stockand JD, Kleyman TR: Epithelial sodium channels Muñoz E, Díaz A, Juárez ME, Giménez I, Gamba G: The Na+:Cl- co- are activated by furin-dependent proteolysis. JBiolChem279: 18111– transporter is activated and phosphorylated at the amino-terminal 18114, 2004 domain upon intracellular chloride depletion. JBiolChem281: 28755– 44.McCormickJA,NelsonJH,YangCL,CurryJN,EllisonDH:Overex- 28763, 2006 pression of the sodium chloride cotransporter is not sufficient to cause 28. Yang SS, Fang YW, Tseng MH, Chu PY, Yu IS, Wu HC, Lin SW, Chau T, familial hyperkalemic hypertension. Hypertension 58: 888–894, 2011 Uchida S, Sasaki S, Lin YF, Sytwu HK, Lin SH: Phosphorylation regulates 45. Hoorn EJ, Walsh SB, McCormick JA, Fürstenberg A, Yang CL, Roeschel NCC stability and transporter activity in vivo. J Am Soc Nephrol 24: T, Paliege A, Howie AJ, Conley J, Bachmann S, Unwin RJ, Ellison DH: 1587–1597, 2013 The calcineurin inhibitor tacrolimus activates the renal sodium chloride 29. Castañeda-Bueno M, Cervantes-Perez LG, Rojas-Vega L, Arroyo-Garza cotransporter to cause hypertension. Nat Med 17: 1304–1309, 2011 I, Vázquez N, Moreno E, Gamba G: Modulation of NCC activity by low 46. Wade JB, Liu J, Coleman R, Grimm PR, Delpire E, Welling PA: SPAK- 1 and high K(+) intake: Insights into the signaling pathways involved. Am mediated NCC regulation in response to low-K diet. Am J Physiol J Physiol Renal Physiol 306: F1507–F1519, 2014 Renal Physiol 308: F923–F931, 2015 30. Frindt G, Palmer LG: Effects of dietary K on cell-surface expression of 47. Good DW, Velázquez H, Wright FS: Luminal influences on potassium renal ion channels and transporters. Am J Physiol Renal Physiol 299: secretion: Low sodium concentration. Am J Physiol 246: F609–F619, F890–F897, 2010 1984 31. Vidal-Petiot E, Elvira-Matelot E, Mutig K, Soukaseum C, Baudrie V, Wu 48. Hunter RW, Craigie E, Homer NZ, Mullins JJ, Bailey MA: Acute in- S, Cheval L, Huc E, Cambillau M, Bachmann S, Doucet A, Jeunemaitre hibition of NCC does not activate distal electrogenic Na+ reabsorption X, Hadchouel J: WNK1-related familial hyperkalemic hypertension re- or kaliuresis. Am J Physiol Renal Physiol 306: F457–F467, 2014 sults from an increased expression of L-WNK1 specifically in the distal 49. Stanton BA, Kaissling B: Adaptation of distal tubule and collecting duct 1 1 nephron. Proc Natl Acad Sci USA 110: 14366–14371, 2013 to increased Na delivery. II. Na and K transport. Am J Physiol 255: 32. Yang SS, Morimoto T, Rai T, Chiga M, Sohara E, Ohno M, Uchida K, Lin F1269–F1275, 1988 SH, Moriguchi T, Shibuya H, Kondo Y, Sasaki S, Uchida S: Molecular 50. Terker AS, Zhang C, McCormick JA, Lazelle RA, Zhang C, Meermeier pathogenesis of pseudohypoaldosteronism type II: Generation and NP, Siler DA, Park HJ, Fu Y, Cohen DM, Weinstein AM, Wang WH, Yang

J Am Soc Nephrol 28: 2597–2606, 2017 SPAK Activation Causes Familial Hyperkalemic Hypertension 2605 BASIC RESEARCH www.jasn.org

CL, Ellison DH: Potassium modulates electrolyte balance and blood 53. Merz WA: Die Streckenmessung an gerichteten Strukturen im Mikros- pressure through effects on distal cell voltage and chloride. Cell Metab kop und ihre Anwendung zur Bestimmung von Oberﬂächen-Volumen- 21: 39–50, 2015 Relationen im Knochengewebe. Mikroskopie 22: 132–142, 1968 51. Zhang C, Wang L, SU XT, Zhang J, Lin DH, Wang WH: ENaC and ROMK activity are inhibited in the DCT2/CNT of TgWNK4PHAII mice. Am J Physiol Renal Physiol 312: F682–F688, 2017 52. Heise CJ, Xu BE, Deaton SL, Cha SK, Cheng CJ, Earnest S, Sengupta S, See related editorial, “Nephron Remodeling Underlies Hyperkalemia in Fa- Juang YC, Stippec S, Xu Y, Zhao Y, Huang CL, Cobb MH: Serum and milial Hyperkalemic Hypertension,” on pages 2555–2557. glucocorticoid-induced kinase (SGK) 1 and the epithelial sodium channel are regulated by multiple with no lysine (WNK) family mem- This article contains supplemental material online at http://jasn.asnjournals. bers. JBiolChem285: 25161–25167, 2010 org/lookup/suppl/doi:10.1681/ASN.2016090948/-/DCSupplemental.

2606 Journal of the American Society of Nephrology J Am Soc Nephrol 28: 2597–2606, 2017 Complete Methods

Animal Studies

Animal studies were performed in adherence to the NIH Guide for the Care and Use of

Laboratory Animals and approved by The University of Maryland School of Medicine Institutional

Animal Care and Use Committee. Mice were housed in groups of two to five per cage on a 12:12 h light/dark cycle; with lights on at 6 a.m. Tissues, blood and urine samples were collected between 8 and

10 am. Male mice were studied. Food and water were available ad libitum.

Creating Mice Carrying the Floxed CA-SPAK Allele

Two mutations, T243E and S383D, which render the kinase constitutively active,(23) were inserted into the full length N-terminal HA-epitope SPAK cDNA, and the mutant SPAK was incorporated into a construct that targets the mouse SPAK gene, STK39. The construct consisted of a 3.5 kb genomic DNA fragment as the 5’ arm of recombination, followed by a loxP site, the neomycin resistance gene cassette, a second loxP site, the SPAK mutant, the mouse phosphoglycerate kinase-1 polyadenylation sequence (pA), and a 6.5 kb fragment as the 3’ arm of recombination (Supplement

Figure 1A). Arms of recombination were dropped into the construct vector using BAC clone bMQ-

410m10 and recombineering. The loxP site is in frame with the remainder of the SPAK protein and encodes a short 12 amino acid peptide. Successful insertion of construct into the STK39 gene creates a

SPAK KO, and allows CA-SPAK to expressed under control of the native SPAK promoter following recombination with Cre-recombinase.

129/SvEvTac-derived ES cells were electroporated with the targeting vector, and ES cells were grown on fibroblast feeder cells in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with

15% fetal bovine serum, gentamicin (50 μg/ml), LIF (1,000 U/ml), β-mercaptoethanol (90µM) and G418

(0.2 mg/ml). Two hundred and sixty neomycin-resistant colonies were picked and analyzed by Southern blot analysis. Recombination reduces the size of an XbaI fragment from 14 kb (wild-type allele) to 7 kb (mutant allele) as an additional XbaI site is introduced upstream of the second loxP site (Supplemental

Figure 1A, C). Three positive clones containing the floxed CA-SPAK allele were identified and injected into C57BL6 blastocysts, thereby generating 4 chimeric males of high chimeric percentage. Two of these males gave germline transmission of the floxed CA-SPAK allele. Male mice carrying the floxed

CA-SPAK allele were backcrossed with C57BL6/J females

Creating DCT1 specific and Global CA-SPAK Mice

To drive early DCT (DCT1)-specific expression within the kidney, male mice homozygous for floxed CA-SPAK were bred with female mice that express Cre recombinase under the control of the parvalbumin promoter (The Jackson Laboratory, B6.129P2-Pvalbtm1(Cre)Arbr/J). Within the kidney, parvalbumin is specifically expressed within the DCT1(24, 25). The progeny of these crosses (PV-CRE +

CA-SPAK, one copy each) were backcrossed for 10 generations with wild-type C57BL/6J mice, then interbred to produce mice homozygous for the floxed CA-SPAK alleles and Cre recombinase. During the course of our studies, we found animals heterozygous for Cre recombinase expressed CA-SPAK as efficiently as animals having two copies of Cre recombinase. Thus to avoid any off-target effects of Cre protein expression, all studies were performed on animals having a single copy of Cre recombinase.

With Cre expression, transcription of the CA-SPAK targeting cassette occurs using the native SPAK gene promoter. However, in the absence of Cre, the CA-SPAK cassette prevents expression of the native gene with these animals being effectively SPAK-null (See Supplemental Figure 1). Neither genotype exhibited any obvious physical or behavioral abnormalities and both were produced in the expected Mendelian ratio and normal litter size (6-8 pups per litter).

To facilitate the molecular and physiological comparison of CA-SPAK to WT animals, several mice homozygous for floxed CA-SPAK and homozygous for Cre recombinase were crossed with WT

C57Bl/6J (purchased from Jackson Laboratory) mice. The resulting progeny were further crossed to yield animals homozygous for either floxed CA-SPAK or WT SPAK, and heterozygous for Cre recombinase. Neither genotype exhibited any obvious physical or behavioral abnormalities and both were produced in the expected Mendelian ratio and normal litter size (6-8 pups per litter).

For global CA-SPAK expression, male mice homozygous for CA-SPAK alleles were bred with female mice that express Cre recombinase under the control of the E2a promoter (The Jackson

Laboratory, B6.FVB-Tg(Ella-cre)C5379Lmgd/J). Interestingly, these crosses never produced offspring that were homozygous for CA-SPAK, suggesting whole body expression of CA-SPAK is lethal (See

Supplemental Table 1).

Genotyping

Genomic DNA was harvested from tail snips collected at weaning using Extract-N-Amp Tissue

PCR Kit (Sigma). Quantitative PCR was then used to genotype CA-SPAK offspring by tracking the Cre recombinase gene (Fwd primer - ACCTGAAGATGTTCGCGATTATCT; Rev primer -

ACCGTCAGTACGTGAGATATCTT), the 2LoxP allele of SPAK (Fwd primer -

TACACTTCATTCTCAGTATTGTTTTGCC; Rev primer - TGATGATATCCAACATGGAACCTCC), and the WT allele of SPAK (Fwd primer - GTACGAGCTCCAGGAGGTTATCG; Rev primer -

TTACTGGGTTCCAGCTCCGCC). At the end of each experimental protocol genomic DNA was collected from a piece of kidney to insure Cre recombinase had properly excised the one loxP site and enabling expression of the CA-SPAK coding cassette (1loxP allele; Fwd –

AAAAGGCCCACAGCAGCAGAAC; Rev – ACCGAGATCTCTGAGTTCTCTTC). A Roche LightCycler

480 qPCR system and LightCycler 480 SYBER Green I Master reagents were used to quantify copy number. All reactions were performed in triplicate for each animal. Relative transcript abundance was calculated using the Pfaffl equation,(38) a derivation of the ΔΔCt method. Unlike the latter, the Pfaffl equation accounts for actual efficiency of doubling within the linear range of amplification, a value that can vary depending on template concentration. The values represented in the figures are normalized

Pfaffl values relative to WT. Roche Cyber Green Master Mix was used to conduct PCR reaction. Animals found to be homozygous for the 2loxP alleles and heterozygous for Cre were considered CA-

SPAK, while animals found to be homozygous for the 2loxP alleles but lacking Cre were considered

SPAK KO. Animals lacking the 2loxP alleles but found to be heterozygous for Cre were considered WT.

Dietary Manipulation

At approximately 6-7 weeks of age, all mice being studies were switched from the house diet to our control diet containing 1% potassium, 0.32% sodium, 0.9% chloride (TD.88238 from Harlan

Teklad). The control diet is matched in composition to the various experimental diets except for salt content. The animals were acclimated to the control diet for at least 10 days prior to the beginning of the study. At 8-10 weeks of age the animals were assigned to a dietary regimen that consisted of either: 1) the control diet, 2) high-salt diet (1% potassium, 1.6% sodium, 2.9% chlorine; TD.10432), or

3) high-potassium diet (5.2% potassium, 0.3% sodium, 0.9% chlorine; TD.10432). All diets were purchased from Harlan Teklad and designed with assistance of a Teklad certified dietician.

Metabolic Cage Studies

Renal function was evaluated in metabolic cage clearance studies. After acclimation (3 days) in metabolic cages (Nalgene (Thermo Scientific), # NALGE650-0322), food and water consumption was assessed and urine samples were collected. Urine samples were collected several times a day to prevent contamination from food, water, and fecal matter. Urine samples were collected in tubes containing mineral oil to prevent evaporation. Kidney electrolyte handling was assessed in 24 hour measurements of Fractional Excretion (FE), which are calculated as the rate of urinary excretion of a solute (UxV, where Ux is the urinary concentration of substance, x, and V is the urinary flow rate) relative to the filtered load (FEx = UxV/eGFR*Px, where eGFR is the estimated glomerular filtration rate calculate by creatinine clearance (see below), and Px is the plasma concentration). Trans-Tubular Potassium Gradient (TTKG), an index of distal potassium secretion, was calculated as follows; TTKG =

+ (UK/PK)/(UOsm/POsm), where UK and PK are the urinary and plasma [K ] and UOsm and POsm are the osmolarity of urine and plasma respectively.

Blood Pressure Measurements

Blood pressure measurements were made in conscious mice using a DSI (Data Sciences

International) telemetry based system. The catheters from PA-C10 telemeters were surgically inserted into the internal carotid artery following the manufacturer's instruction. Following the surgery, the animals were allowed to recover for 4 days before beginning measurements. Dataquest ART 4.2 software was used to create a sampling program that measured the blood pressure of each animal for

20 s every 10 min.

Sample Collection, Preparation, and Analysis

Animals were anesthetized by intraperitoneal injection with ketamine/xylazine (100 mg/kg of ketamine, 10 mg/kg of xylazine). Once an animal was unconscious, a kidney was removed; cortex and medulla were separated by free-hand dissection and flash frozen in liquid nitrogen. Blood samples were

+ + - - collected from the carotid artery. Blood chemistry and gases (Na , K , Cl , HCO3 , pH, hematocrit, and

BUN) were measured using from a 100ul aliquot of whole blood using an i-STAT EC8+ cartridge and an i-STAT1 Handheld Analyzer (Axaxis). The remaining fraction of blood was immediately spun-down to separate formed elements and plasma, the latter was subsequently isolated and frozen for later analysis of renin activity, aldosterone, and creatinine levels. Urine sodium, potassium and chloride analysis was performed in our lab using an Easylyte Analyzer (Medica Corporation). Plasma and urine creatinine levels were measured using the QuantiChrom Creatinine Assay Kit (BioAssay Systems) following the manufacturer’s protocol. Plasma aldosterone levels were assayed using the Aldosterone

EIA Kit manufactured by Cayman Chemical. Plasma renin activity was determined using a Renin Assay Kit from Sigma-Aldrich (MAK157) following the manufacture’s protocol.

Sample Preparation for Western Blotting

Mouse kidney tissue (cortex or medulla) was sonicated in HEENG buffer (20mM Hepes (pH

7.6), 125mM NaCl, 1mM EDTA, 1mM EGTA, 10% glycerol) containing 1% Triton and 0.5% SDS with protein and phosphatase inhibitor. Samples were taken from the freezer and immediately placed into

HEENG buffer and then cut into small pieces using small scissors. The tissue was sonicated 2 times on ice at 8-second pulses (20 seconds between pulses) and allowed to sit at room temperature for 15 minutes before being rotated at 4°C for 1 hour followed by high-speed centrifugation (15,000 rpm) for

10 min to pellet insoluble material. The supernatant was collected and quantified for protein yield using a bicinchoninic acid protein assay reagent kit (Pierce). After incubating in Laemmli buffer supplemented with 2-mercaptoethanol (room temperature for 30 min), 20 µg of kidney protein per sample/well was resolved on precast TGX SDS-PAGE gels (4-20% gradient) purchased from BioRad, and transferred using BioRad TurboBlot system. The membranes were blocked in Tris-buffered saline with 0.1% Tween

20 (TBS-T) containing 5% nonfat dry milk for 1 h at room temperature. Membranes were then incubated in 5% nonfat dry milk containing primary antibody (4°C, overnight), washed in TBS-T 10 minutes (3 times), incubated in 5% nonfat dry milk containing HRP-conjugated secondary antibody, and then washed for 10 minutes (3 times) in TBS-T. Bound antibodies were then revealed using enhanced chemiluminescence reagent (Pierce) and fluorography.

Protein quantification was performed by scanning autofluorograms and measuring the integrated density of protein bands using ImageJ software. Bands were measured in the linear range of the fluorographic signal. Duplicate gels were processed and developed in parallel. After detection of the target, membranes were stripped and reprobed for tubulin as a loading control. Unless otherwise stated, each protein signal was divided by its own tubulin signal to yield a tubulin-normalized signal.

These data are presented and analyzed as the relative abundance, denoting the tubulin-normalized test signal relative to the average of the wild-type normalized signal.

Immunolocalization

Anesthetized mice were fixed by perfusion with 2% paraformaldehyde in PBS via the left ventricle for 5 min at room temperature. The kidneys were then removed and fixed for additional 24 h at

4 °C, rinsed in PBS, and embedded in paraffin. Cross-sections 3-μm thick, cut at the level of the papilla, were picked up on chrome-alum gelatin-coated glass coverslips and dried on a warming plate. The sections were then deparaffinized in two xylene baths and two absolute ethanol baths, 5 min each, and rehydrated in a graded ethanol series to distilled water.

For epitope retrieval, the coverslips were placed in a pH 8 aqueous solution containing Tris (1 mM), EDTA (0.5 mM), and SDS (0.02%). The retrieval solution was heated to boiling in a microwave oven, transferred to a conventional boiling water bath for 15 min, and then allowed to cool to room temperature before the sections were thoroughly washed in distilled water to remove the SDS.

Sections were preincubated for 30 min with Image-iT blocking solution (Invitrogen), rinsed in

PBS, and then preincubated an additional 30 min in a solution of 2% BSA, 0.2% fish gelatin, 5% normal donkey serum, and 0.2% sodium azide in PBS. Tissues were thoroughly rinsed with Tris-buffered saline

(TBS) to remove PBS. Incubations with specific antibodies (as described above), diluted in TBS containing 1% BSA, 0.2% fish gelatin, 0.1% Tween 20, 10 mM CaCl2, and 0.2% sodium azide, took place overnight in a humid chamber at 4 °C. After thorough washing in high salt wash (incubation medium plus added sodium chloride at 0.5 M), various Alexa Fluor 405-, 488-, 568-, and 649- conjugated donkey anti-mouse, anti-rabbit, anti-chicken, and anti-guinea pig IgG antibodies (Jackson

Laboratories) were used to visualize specific target protein.

Quantitative analysis of confocal images (apical fluorescence intensity and co-localization) was performed using Improvision Volocity 5 by a trained investigator who was blind as to identity of the sample groups.

Quantification of ROMK intracellular localization in kidney sections.

ROMK intracellular localization was determined by measuring the pixel intensity from the tubule lumen toward the intracellular space at 0.4 μm–increments using Volocity 5 3D Image Analysis

Software (PerkinElmer). A plot profile line was drawn exactly perpendicular to the cell apical membrane at the point to be measured, and the density profile was plotted. The peak intensity value was taken along with the pixel intensity three pixels from the peak in the direction of the cytoplasm. This later value provided a measure of background label and ROMK label not associated with the apical membrane and was subtracted from the peak intensity and taken as the apical membrane signal. A total of fifty cells from random selection of 5 WT and 5 CA-SPAK KI mice (n ≥ 250 cells per group) were measured and compared.

Morphometric Analysis of DCT and ASDN in Kidney Sections

Nephron segments (TAL, DCT1, DCT2, CNT1, CNT, and CCD) were identified in coronal kidney sections by confocal microscopy (Zeiss LSM 510, ×10 objective lens) and segment specific antibody labeling (NKCC2, TAL; NCC, total DCT; parvalbumin, DCT1; calbindin + NCC, DCT2; calbindin along

CNT1; calbindin + AQP2, CNT; AQP2 alone, CCD). Images were tiled, and the entire cortical area was analyzed for each section. Nephron segment length were determined with a curvilinear stereometric system developed by Merz,(53) and used previously by our group.(38, 39) For these measurements, a curve linear test grid composed of evenly spaced curved lines was superimposed over the images of the cortex. Tubule length, L, was determined by L = I × D, where L is the linear boundary length being measured, I is the number of intersections between the curvilinear test grid and the boundary, and D is the diameter of the grid semicircles (21 μm). The boundary in this case was one outer edge of the tubule being measured, but limited by the axial extent of the apical membrane. The length of each identifiable DCT1, DCT2, CNT1, CNT, and CCD was measured and normalized to the area of the cortex in the image (μm/mm2). The curve linear grid also contained a small dot at the midpoint of each semicircle. The dots were used in point counting to measure the total tubular area of each nephron segment. The fractional-area of each tubule type was calculated by dividing the number of dots falling on each tubule type by the total number of dots falling on the cortex. The dots on the grid were 21 μm apart, and thus each dot represented 421 μm2 of area. Using this conversion, the fractional-area could be expressed as μm2/mm2. Sections from 6 animals of each genotype were evaluated in this manner, and the figures represent the average of each genotype (n = 6).

Image Processing

As described in the morphometric analysis section above, four different markers were used to identify the DCT and ASDN segments. Quantitative analysis revealed changes in DCT1 and CNT abundance but these are not entirely obvious by eye when all markers are included in the image. For this reason,

Figure 6 only shows DCT1 and CNT segments in a two-color image, against a bright field image. For imaging processing, imageJ64 software (NIH) was used. To identify and visualize DCT1, overlapping signals from the Parvalbumin and NCC channels were selected, and merged to single color (green). To identify and visualize the CNT overlapping signals from AQP2 and Calbindin signals were selected and merged to another color (red). Selected DCT1, CNT, and bright field images were then merged.

Statistical Analysis

Data are presented as means ± SEM. Statistical analysis was performed using GraphPad PRISM version 6. Statistical significance was determined using One-way ANOVA when comparing a single dependent variable in all three genotypes. Two-way ANOVA was used in HCTZ and dietary manipulation studies to determine if there is an interaction between two variables (e.g., diet and genotype or drug and genotype) and the dependent variable (e.g., blood pressure, plasma potassium concentration, Fractional excretion). For both One-way and Two-way ANOVA analyses, a multiple comparisons analysis was performed post hoc using Tukey’s Multiple Comparison test. P < 0.05 was considered significant.

Supplemental Table 1. Genotyping Results for Global CA-SPAK (G-CA-SPAK) Expression Litters Total Number of Pups Delivered Pups/Litter 9 59 6.6 Male Female G-CA-SPAK-/- G-CA-SPAK+/- G-CA-SPAK+/+ G-CA-SPAK-/- G-CA-SPAK+/- G-CA-SPAK+/+ 10 22 0 8 19 0

Supplemental Table 2. List of Antibodies Used During Study Antibody Use Dilution Source or Company w/ product ID Reference (species-antigen) (PMID) ms = mouse sh = sheep WB = rb = rabbit ch = chicken western blot gp = guinea-pig IF = immunolocalization ch – Aquaporin2 IF, 1:100, James Wade, U. of Maryland S.O.M. 25893600 WB 1:1000 rb – ROMK IF, 1:80, James Wade, U. of Maryland S.O.M. 21454252 WB 1:3000 gp – NCC IF 1:200 James Wade, U. of Maryland S.O.M. 22977235 rb – NCC WB 1:500 James Wade, U. of Maryland S.O.M. 22977235 ch – pNCC (T58) IF, 1:50, James Wade, U. of Maryland S.O.M. 22977235 WB 1:100 ch – NKCC2 IF 1:100 James Wade, U. of Maryland S.O.M. 12145305 rb – NKCC2 WB 1:6000 Mark Knepper - NIH 11053048 rb – pNKCC2 (R5) WB 1:1000 Biff Forbush – Yale University 12145305 rb – SPAK IF, 1:250, Eric Delpire – Vanderbilt S.O.M 12386165 WB 1:700 rb - parvalbumin IF 1:500 Swant - #PV27 22977235 ms – calbindin D-28kDa IF 1:600 Sigma – C8666 25893600 rb - tubulin WB 1:3000 Cell Signaling - #2144 25893600 rb - HA WB 1:1000 Cell Signaling - #3724 rb- WNK1 WB 1:1000 Arohan Subramanya – U. of Pittsburgh 26241057 Sh - pWNK1 (S382) WB 1:500 MRC-PPU Reagents – Dundee Scotland 25565204 rb - WNK4 WB 1:1000 Novus Biologicals – NB600-284 19401467 ms - ENaC (alpha) WB 1:1000 Stress Marq Biosciences – 14E10 20966128 rb – EnaC (beta) WB 1:500 Larry Palmer – Cornell University 16554417 rb - ENaC (gamma) WB 1:1000 Larry Palmer – Cornell University 16554417 rb - BK-alpha WB 1:1000 Alamone – APC-021 26537348 rb - BK-B1 WB 1:750 Alamone – APC-036 19458125 rb - BK-B4 WB 1:500 Alamone – APC-061 20299355

Supplement Figure 1

A. WT SPAK Gene (Stk39) ATG

Promoter Intron 1 Intron 2 Chr. 2 Exon 1 Exon 2

B. CA-SPAK KI (w/o Cre Expression) = SPAK KO

ATG HA Tag T243E S383D STOP

Promoter Intron 1 Intron 2 Chr. 2 PGK: neo pA Remaining SPAK ORF pA Exon 2 Partial loxP loxP Exon 1

C. CA-SPAK KI (w/ Cre Expression) = CA-SPAK

ATG HA Tag T243E S383D STOP

Promoter Intron 1 Intron 2 Chr. 2 Remaining SPAK ORF pA Exon 2 Partial loxP Exon 1

D. Pvlb Promoter Drives Cre Expression

ATG STOP

Chr. 15: 78.2 Pvlb IRES CRE pA

Supplemental Figure 1. Targeted CA-SPAK Knockin Strategy. (A) WT SPAK gene, showing promoter and first two exons. (B) Targeting knockin cassette, containing floxed Neo, followed by the SPAK reading frame with phospho-mimic mutations (T243E & S383D), an HA-tag and an in-frame stop codon, was inserted after the start codon of SPAK. In the absence of Cre-recombinase, the cassette renders the native gene inactive and thus the animals are functionally SPAK-null. (C) In the presence of Cre-recombinase, the Neo selection cassette is excised, allowing CA-SPAK transcription under the control of the native SPAK gene promoter. (D) DCT1 specific expression of CA-SPAK was driven by Cre-recombinase under the control of the parvalbumin promoter for these studies. In the kidney, parvalbumin is exclusively expressed in the early DCT. The individual elements of the above genes are not draw to scale. Supplemental Figure 2

A. MW (kDa) WT CA-SPAK SPAK KO IB 250 150 NCC 100 250 pNCC 150 (T58) 100 75

50 tubulin 37

B. NCC C. pNCC (T58) D. pNCC/Total NCC * * * * ** * ** * ** 1.5 2.5 2.0 e e e c c c n n n 2.0 1.5 bund a bund a bund a 1.0 A A A 1.5 in in in e e e t t t 1.0 o o o r r r P

1.0 P P 0.5 ve ve ve i i i t t t 0.5 a a a l l l 0.5 e e e R R R 0.0 0.0 0.0 WT CA-SPAK SPAK KO WT CA-SPAK SPAK KO WT CA-SPAK SPAK KO

Supplemental Figure 2. NCC and pNCC (T58) abundance in WT, CA-SPAK and SPAK KO littermates. (A) Representative western blot and quantification of (B) total NCC, (C) pNCC and (D) the ratio of pNCC to total NCC in WT, CA-SPAK, and SPAK KO mice. Dots and wickers are mean + SEM (n=4, *P< 0.05 vs WT, **P< 0.05 CA-SPAK vs SPAK KO). Supplemental Figure 3

A. Hourly Systolic BP - control diet B. Effect of HCTZ on Diastolic BP

Day 1 Day 2 WT CA-SPAK WT 145 ** ** CA-SPAK * * 140 95 135 90 130

mmH g 125 85 120 DBP (mmH g) 80 115

110 75 Time (Hours) Baseline HCTZ Vehicle

Supplemental Figure 3. Characterization of WT and CA-SPAK blood pressure. (A) Hourly telemetric measurements of blood pressure (dark bar, awake period at night; open bar, day). CA-SPAK animals retain the typical diurnal variation in systolic blood pressure, but at all time points CA-SPAK mice are hyperten- sive relative to WT mice. Each data point and wicker is the average hourly SBP + SEM (n=8). (B) Diastolic blood pressure is also elevated in CA-SPAK and is normalized with HCTZ treatment. (n=8, *P< 0.05 WT vs CA-SPAK for the same treatment, **P< 0.05 different treatments within the same genotype). Supplemental Figure 4

WT A. UNaV B. UKV CA-SPAK * * * 600 ** 800

600 400 da y da y 400 umol/ umol/ 200 200

0 0 0 1 2 3 0 1 2 3 Days of HCTZ Days of HCTZ

Supplemental Figure 4. Delayed Restoration of Urinary Potassium Excretion in CA-SPAK mice. The temporal response of urianry sodium (A) and potassium excretion (B) in WT and CA-SPAK mice follwing HCTZ treatement. Dots and wickers are mean ± SEM (n=6, *P < 0.05 WT vs CA-SPAK undergoig same treatment) . Supplemental Figure 5 TTKG A. Control B. HCTZ Treated * *

12 12

9 9

6 6

3 3 WT CA-SPAK SPAK KO WT CA-SPAK SPAK KO

Supplemental Figure 5. Distal Nephron Potassium Secretion as Assessed by Trans-Tubular Potas- sium Gradient (TTKG). (A) Relative to WT, CA-SPAK mice display a decreased TTKG indicative of impaired potassium secretion from the ASDN (n=12 per genotype). (B) Chronic HCTZ treatment (3 days) restores TTKG of CA-SPAK to elevated level of SPAK KO mice. (n=6, *P< 0.05). WT (red), CA-SPAK (blue), and SPAK KO (gray) bars and wickers are mean + SEM, circles are individual data points from each mouse. Supplemental Figure 6 A. B. Cortex ROMK * IB WT CA-SPAK SPAK KO MW (kDa)

e 3 * ** 100 c n 75 ROMK

50 bund a 2 A

37 in e t o r Tubulin 75 P 1 ve i

50 t a l e

R 0 WT CA-SPAK SPAK KO C. D. ROMK Medulla *

e 1.5 IB WT CA-SPAK SPAK KO MW (kDa) c n 100

75 bund a 1.0 ROMK A

50 in e t o 37 r P 0.5 ve i

75 t a

Tubulin l

50 e R 0.0 WT CA-SPAK SPAK KO

Supplemental Figure 6. ROMK Abundance is Reduced in Kidney Cortex of CA-SPAK. (A) Western blots and (B) quantification of ROMK in CA-SPAK renal cortex, compared to WT and SPAK-KO littermates, and (C & D) Loop of Henle-enriched renal medulla (n=6 for each genotype, *P < 0.05 vs WT, **P < 0.05 CA-SPAK vs SPAK KO). WT (red), CA-SPAK (blue), and SPAK KO (gray) bars and wickers are mean + SEM, circles are individual data points from each mouse. Supplemental Figure 7

A. B. MW (kDa) WT CA-SPAK IB:

250

WNK1 150 WNK1 pWNK1 WNK4

e 1.5 100 * * * 250 bundan c

A 1.0

pWNK1 n

150 i e t 100 Pr o

e 0.5 v

250 i t a l e

150 WNK4 R 0.0 100 WT CA-SPAK WT CA-SPAK WT CA-SPAK 75

50 Tubulin 37

Supplemental Figure 7. WNK expression is decrease in CA-SPAK mice. (A) Western blots and (B) quantification of WNK1. pWNK1, and WNK4 in CA-SPAK renal cortex, compared to WT littermates, (n=4 for each genotype, *P < 0.05 vs WT, bars and wickers are mean + SEM, circles are individual data points from each mouse.