Abstract
WNK1 and WNK4 are unusual serine/threonine kinases with atypical positioning of the catalytic active-site lysine (WNK: With-No-K[lysine]). Mutations in these WNK kinase genes can cause familial hyperkalemic hypertension (FHHt), an autosomal dominant, hypertensive, hyperkalemic disorder, implicating this novel WNK pathway in normal regulation of BP and electrolyte balance. Full-length (WNK1-L) and short (WNK1-S) kinase-deficient WNK1 isoforms previously have been identified. Importantly, WNK1-S is overwhelmingly predominant in kidney. Recent Xenopus oocyte studies implicate WNK4 in inhibition of both thiazide-sensitive co-transporter–mediated Na+ reabsorption and K+ secretion via renal outer medullary K+ channel and now suggest that WNK4 is inhibited by WNK1-L, itself inhibited by WNK1-S. This study examined WNK pathway gene expression in mouse kidney and its regulation in vivo. Expression of WNK1-S and WNK4 is strongest in distal tubule, dropping sharply in collecting duct and with WNK4 also expressed in thick ascending limb and the macula densa. These nephron segments that express WNK1-S and WNK4 mRNA have major influence on long-term NaCl reabsorption, BP, K+, and acid-base balance, processes that all are disrupted in FHHt. In vivo, this novel WNK pathway responds with significant upregulation of WNK1-S and WNK4 with high K+ intake and reduction in WNK1-S on chronic lowering of K+ or Na+ intake. A two-compartment distal nephron model explains these in vivo findings and the pathophysiology of FHHt well, with WNK and classic aldosterone pathways responding to drivers from K+ balance, extracellular volume, and aldosterone and cross-talk through distal Na+ delivery regulating electrolyte balance and BP.
The role of WNK1 and WNK4 in control of electrolyte balance and BP first became apparent with their mutation being associated with familial hyperkalemic hypertension (FHHt; also known as Gordon syndrome and pseudohypoaldosteronism type 2), a human autosomal dominant disorder that features hypertension, hyperkalemia, and acidosis that usually are hyperresponsive to thiazide diuretics (1,2). FHHt can be caused by intronic deletions in WNK1 or missense mutations in WNK4 (3). Mutations in either cause a broadly similar phenotype, suggesting that WNK1 and WNK4 function in a common pathway. Unlike most monogenic disorders that affect BP, which feature reciprocal Na+ and K+ (and/or H+) imbalances and share a relationship to the aldosterone pathway (4), FHHt features concurrent NaCl and K+(and/or H+) retention (1,3,5). This unusual characteristic indicates the existence of a novel “WNK pathway” functioning in normal physiology, which may allow the “independent of aldosterone” regulation of K and Na balance (and extracellular volume) by the kidney, ultimately also maintaining BP within the normal range. The BP-regulatory role of this WNK pathway is conserved in evolution as WNK1+/− mice are hypotensive (6).
Previously, we and others demonstrated that a 5′-truncated kinase-deficient isoform (WNK1-S) predominates in kidney (7), this being conserved between human and mouse (7–9). Isoform-specific probes distinguished ubiquitous low-level expression of full-length WNK1-long (WNK1-L) from abundant WNK1-S expression in distal nephron. Recent Xenopus oocyte studies implicate WNK4 in inhibition of NaCl reabsorption by thiazide-sensitive Na+Cl− co-transporter (NCC) (10,11) and/or K+ transport via renal outer medullary K+ (ROMK) (12), while WNK1-L may in turn inhibit WNK4 (11). Furthermore, very recent in vitro reports show that WNK1-S, the predominant isoform in kidney, may participate in regulation of electrolyte transport. In Xenopus oocyte studies, WNK1-S acted as a dominant negative regulator of WNK1-L (13) (thereby WNK1-S would relieve repression of WNK4). In cultures of mouse cortical collecting duct (CCD) cells, aldosterone induces WNK1-S expression, and this may mediate an increased epithelial Na channel (ENaC) conductance (14).
This evidence suggests that this WNK pathway plays a functional role in normal physiology to control electrolyte homeostasis and BP. The full renal response that maintains normal BP or compensates for altered electrolyte intake involves a substantial component that takes many hours per day to develop fully and is accompanied by significant persistent changes in nephron ultrastructure and gene expression. These current studies progress from valuable insights of Xenopus oocyte work to investigate the physiologic role of the WNK pathway in vivo in mice, localizing it within the nephron and tracking WNK mRNA expression changes in response to chronic variations in dietary electrolyte intake and aldosterone status. This leads to a working model of the distal nephron, explaining the WNK pathway findings and the pathophysiology of FHHt.
Materials and Methods
Animal Treatment
All procedures were carried out under provisions of ethically approved licenses and involved adult, 25- to 30-g, male C57BL/6 mice (Charles River, Margate, UK). Modified electrolyte feeds for mice were obtained from Special Diet Services (Witham, UK).
RNA Extraction
At conclusion of treatments of mice, both kidneys were removed under terminal anesthetic, immediately frozen on dry ice, and stored at −80°C. Frozen kidneys were fragmented and immediately homogenized in TRIzol Reagent (Invitrogen, Paisley, UK), and total RNA was extracted following the manufacturer’s guidelines.
Real-Time PCR
Assays used ABI PRISM 7900 relative quantification real-time methods (Applied Biosystems, Foster City, CA). PCR was performed in 384-well plates (AB Gene) and used 10-μl reactions that contained 5.0 μl of TaqMan Master Mix (Applied Biosystems), 200 nM of each primer, 5 nM of probe, and 4.5 μl of template (1:40 dilution of cDNA synthesized as described previously [7]). PCR conditions involved 95°C for 10 min, then 40 cycles of 95°C for 15 s and 60°C for 60 s. Standard template dilution curves enabled target gene quantification and normalization to the endogenous control TATA-Box Binding Protein (TBP). All group values were calibrated to their control groups.
Validation studies using mouse renal RNA established TBP as an excellent control for these studies, showing less variation than 18S, actin, and several other reputed housekeeping genes and excellent reproducibility against an exogenous control gene. Previous literature (15) reinforces TBP as a particularly good renal endogenous control gene. Real-time PCR assays used the ABI Assays-on-Demand (TBP: Mm00446973_m1) or Assays-by-Design services (all other assays, see Table 1).
Primer and probe sequences for assays by designa
In Situ Hybridization Analysis
RNA probes for in situ hybridization (ISH) were produced to specific gene regions using nested PCR methods (7,16). For primer sequences, see Table 2. ISH used renal cryostat sections (10 μm) from each mouse mounted on silane-coated glass slides, then fixed, hybridized, RNase A treated, washed, and ethanol-dehydrated as described previously (7,16,17). Slides were exposed to Kodak x-ray film (BioMax MR-1; Sigma, Poole, UK), dipped in NTB-2 photographic emulsion (Anachem Ltd., Bedfordshire, UK), exposed (within light-tight box) for up to 4 wk, and developed. Cresyl violet/eosin counterstaining with bright and dark-field illumination was used routinely to visualize and localize silver-grain distribution in emulsion-dipped slides. In counterstained slides, blue dark-field views (dark-field view in blue-filtered light) also were used to limit/eradicate counterstain dark-field artifacts. Serial sections allowed co-localizations and tracing of nephron structures beyond the plane of individual sections.
Primer sequences for PCR construct probes used in ISHa
Image Analysis
Digital image analysis was performed using custom-written applications within the Image Processing Toolkit MATLABR version 7 (The Mathworks, Inc., Natick, MA) environment on 16-bit grayscale TIFF images, having an intensity range from 0 (black) to 65535 (white). This allows semiquantitative image analysis of ISH autoradiographic films and dipped emulsion slides, measuring areas and grayscale intensities (or silver grain densities) as an index of expression level. Measured grayscale levels are converted to equivalent dpm of bound radioactive probe by calibrating from co-exposed radioactive microscales (Amersham, Little Chalfont, UK) that are designed specifically for this purpose. Alignment of consecutive sections involved systematic rotation and linear displacement of one image relative to the next to calculate their normalized cross-correlation and interatively proceeding to the optimal alignment, where this value is a maximum. The aligned images can be displayed as different red and green channels of a single overlaid merged image. Autoradiographic films were scanned on a high-resolution flatbed scanner. Any damaged or inadequate sections were excluded. The software greatly facilitated selection of regions (typically complex shapes) showing expression that was highly significantly above background levels. Background was low, with sense-section backgrounds having a grayscale level that was not significantly different from zero.
Statistical Analyses
Data are expressed as mean ± SEM, P < 0.05 was considered significant, and group comparisons were analyzed by one-way ANOVA and Neuman-Keuls post hoc testing, unless otherwise stated. Least significant difference post hoc test for planned comparisons (ANOVALSD) was used when Neuman-Keuls testing was borderline nonsignificant. Routine changes are expressed as %change (mean ± SEM) relative to control values. Larger scale changes are expressed as fold change (mean ± SEM), relative to control values, and 95% confidence intervals are quoted.
Results
Renal WNK1 and WNK4 Gene Expression
Renal WNK1 and WNK4 mRNA expression is illustrated in Figures 1 through 5⇓⇓⇓⇓, with WNK structure and probe positions shown in Figure 1, an overview shown in Figure 2, the structures involved (and their abbreviations) in superficial and deep distal nephrons in Figure 3, and key expression details in Figures 4 and 5. Both WNK1-S and WNK1-L are expressed in kidney (Figure 2, B, D, and E). WNK1-L shows near background seemingly ubiquitous expression. The great preponderance of WNK1 expression is due to WNK1-S and limited to renal cortex (Figure 2B). High levels of WNK1-S in distal tubules (Figure 2, D and E) fall off sharply distally from connecting tubule (CNT) to cortical collecting tubule (CCT) to CCD (Figure 5C; see Figure 3) (7) and proximally dropping 10-fold at the thick ascending limb of the loop of Henle (TAL)–distal convoluted tubule (DCT) junction (Figure 5, A and B) with expression (including the macula densa) extinguishing in the cortical TAL. Hence, WNK1-S has only a limited weak extension into medullary rays. WNK4 expression levels are lower than WNK1 (more than threefold longer exposures for WNK4 than for WNK1[-S/-T] in Figures 2, 4, and 5). WNK4 expression also is strongest in distal tubule structures (DCT/CNT) but extends beyond distal tubule, at reduced levels, more proximally into TAL, including macula densa and medullary TAL (Figure 4, A, B, D, and G) and more distally at low expression levels (compared with DCT) in collecting duct (CD; Figure 4F). Thus, substantial WNK4 expression extends into medullary rays and outer (but not inner) medulla (Figures 2A and 4G) involving much too high a proportion of tubules to be due to outer medullary CD alone (ENaC γ-subunit versus WNK4; Figure 4G) rather resembling the density of medullary TAL tubules (shown by Na+K+2Cl− co-transporter type 2 [NKCC2]; Figure 4G).
Schematic representation of WNK cDNA structure showing 5′-exon composition and probe positions. The cDNA structures of the 5′-regions of kinase-intact full-length WNK1 (WNK1-L; containing exons 1 through 4), kinase-deficient WNK1 short (WNK1-S; containing exon 4a in place of exons 1 through 4), and WNK4 are illustrated with vertical bars representing exon:exon boundaries. Numbers within rectangles refer to exon numbers. Black boxes specify the cDNA region probed by in situ hybridization (ISH). Double-headed arrows indicate the exon:exon boundary crossed by probes used in real-time PCR studies. Horizontal dashed lines indicate further exons that were not included in this illustration.
WNK kinase gene expression in kidney. ISH for the genes indicated in mouse kidney. (A and B) Serial transverse sections, dark-field views. (C and D) Two-color dark-field views for paired serial sagittal sections, merging showing co-localization (yellow shift). (E) Two-color merge views of D (at higher magnification). WNK1-T expression (E) encompasses the striking distal nephron cortical expression of WNK1-S (B) and additional widespread expression near background levels of WNK1-L. Note: comparing (A)+(B), major coincidence of expression of WNK1-S and WNK4 in cortex, but WNK1-S proportionately very limited in medullary rays and absent in medulla; (E) strong WNK1-S cortical (labyrinth) expression is seen beyond thiazide-sensitive Na+Cl− co-transporter (NCC) overlap (distal convoluted tubule [DCT]), representing particularly extension into connecting tubule (CNT; NCC being DCT restricted).
Nephron structure. Several distal nephrons that drain into one cortical collecting duct (CCD) are shown. Distal nephron begins in the renal medulla in thick ascending limb of the loop of Henle (TAL; numbered), which ascends into cortical medullary rays and enters the cortical labyrinth, contacting its glomerulus (specialized macula densa cells at contact point) then shortly beyond this enters DCT, CNT, and finally initial/cortical collecting tubule (ICT/CCT), which re-enters medullary rays as CCD and re-descends into medulla. Nephrons of superficial glomeruli (e.g., TAL1) typically have a simple structure, with a short usually unbranched CNT. In contrast, in most species, nephrons of mid-cortical (e.g., TAL2) and some juxtamedullary glomeruli (TAL3 + 4) commonly drain through well-developed branched arcades of CNT. A considerably higher proportion of nephrons form arcades in mouse than in human. Arcades are arranged around vascular axes, with CNT being close to arcuate and especially intralobular vessels. The dashed boxes and lettering delimit regional views relevant to panels of Figure 4 (red letters) and Figure 5 (gray letters).
WNK4 gene expression in mouse kidney. ISH studies are shown in bright-field, dark-field, and blue dark-field (see In Situ Hybridization Analysis). Throughout a through g, similar regions of view are illustrated in Figure 3 (red labeled [A through G] boxes). (a) WNK4 expression in TAL(T) in medullary ray, crossing boundary (dashed line) to contact glomeruli (*) in cortical labyrinth. (b) Substantial expression in segment of macula densa (MD) and higher still in DCT (D), also seen in c. U, proximal tubule emerging from urinary pole of glomerulus. (d) WNK4 expression in TAL emerging from medulla, through MD segment, DCT, and CNT. DCT convolutions (dashed arrow) and a segment of TAL (red dashed outline) fall outside the plane of section. (e) Extensive WNK4 expression throughout a well-developed deep cortical arcade. Vascular axis (arcuate artery [AA] and vein [AV] and intralobular artery [IA]) and associated glomeruli are seen in left panel. Note extensive DCT loops adjacent to glomerulus (*) in left panel. Dashed arrow indicates connections of DCT convolutions (out of section plane). (f) Superficial cortex with loops of DCT and ICT/CCT passing near renal capsule. Although both segments have WNK4 expression, note that the level is strikingly higher in DCT. Dashed line indicates renal capsule. (g) Views of ISH studies for the genes indicated in inner stripe of renal outer medulla (having TAL and outer medullary CD [OMCD] as largest tubules). Note that the relative density of WNK4-positive tubules is greater than that for the γ subunit of epithelial Na channel (γ-ENaC; an OMCD marker) but similar to Na+K+2Cl− co-transporter type 2 (NKCC2; a TAL marker). Dashed red line indicates boundary of outer and inner medulla (I). Magnifications: ×50 in g, top (dark-field view); ×100 in g, bottom (blue dark-field view).
WNK1-S gene expression in mouse kidney. ISH studies: in bright-field view in a (top) and b, in blue dark-field view in A (bottom), and in dark-field view in c. Note that WNK1-S gene expression is present in late TAL (T) including at the macula densa (MD) but rises to much higher levels in DCT (D) and CNT (C), reducing somewhat by CCT. *Glomerulus; dashed line indicates position of renal capsule; U, proximal tubule emerging from urinary pole of glomerulus (see also Figure 3, gray-labeled boxes).
Effects of Chronic Variation in Dietary K+In Vivo
Body Weight, Food Intake, and Fluid Balance.
Mice were given group treatments of varying dietary K+ intake (low K+ [LK], normal K+ [NK], and high K+ [HK]; see Table 3 and Figure 6, A through C). The LK group showed borderline lower weight that became significant versus HK (but not NK) at the end (25.6 ± 0.6 versus 27.3 ± 0.6 g, respectively; P = 0.03), but mice seemed healthy throughout. Both HK and LK groups developed a higher fluid intake and urinary output compared with the NK group (HK versus NK significant at conclusion: >3.4-fold higher intake [P < 0.05] and >4.9-fold higher output [P < 0.01]).
Effect of dietary K+ intake on metabolic measurements. The K+ diet study involved adult male C57BL/6 mice, housed in pairs in mouse metabolic cages (Techniplast). After 3 d of acclimation, groups (n = 6) of mice commenced specific diets: (1) Normal (0.33%) K+ feed (NK; control group), (2) low (0.006%) K+ feed (LK), and (3) high-KCl (3.3% K+, 4.4% Cl−) feed (HK). Batch analysis confirmed that diets were well matched for Na+ (0.28 ± 0.05%; all diets) and Cl− content (0.7 ± 0.12%; NK and LK). Mice took the specified diets for 10 d, and daily measurements of food intake (A), fluid intake (B), and urine output (C) were recorded. Urinary electrolyte excretion was measured daily, reflected changes in dietary electrolyte intake, and was expressed as a ratio of creatinine or Na+ excretion (Cl/Cre, K/Cre, K/Na [D through F]). A transient decrease in food intake was observed in the HK group on day 4 immediately after the introduction of treatment diet.
Metabolic measurementsa
Urinary Electrolytes.
Mice were allowed a period of 3 d (days 1 through 3) to acclimate to metabolic cages. After this, during the initial 3 d of active treatment with specific diets (days 4 through 6), K/Cre and Cl/Cre rose 6.8-fold and 4.9-fold, respectively, with HK; remained unchanged with NK; and K/Cre showed a dramatic 24.4-fold decrease with LK. After days 5 to 6, group K/Cre ratios did not change significantly, indicating reestablishment of appropriate electrolyte balance. The HK and LK groups demonstrated a >17-fold increase and a >14-fold decrease, respectively, in K/Na by day 6 (Figure 6, D through F).
Plasma Measurements.
As expected, HK induced a small but significant increase in plasma K+ within the normal range. Plasma aldosterone was elevated with HK but was unchanged with LK (see Table 3).
WNK Expression Responses to Dietary K+ and Na+ Challenges
For investigation of the effect of varied dietary K+ intake on WNK expression, mice were fed diets with a specific K+ content (NK, LK, or HK) for 10 d. Specific real-time PCR assays indicated that renal WNK1-S was downregulated by 20 ± 9.3% with LK (P = 0.04) and upregulated by 30 ± 10.4% with HK (P = 0.01; a 50 ± 10.2% rise from LK to HK), as was total WNK1 (WNK1-T: 24 ± 7% increase; P = 0.0009) compared with NK (Figure 7A). WNK4 was upregulated with HK (48 ± 24.2%; P = 0.01) but unchanged with LK.
WNK expression in response to varied K+ intake. (A) Real-time PCR results from renal RNA from groups of mice (n = 6) with variations in dietary K+. WNK1-S expression is significantly downregulated with LK, whereas WNK1-S and WNK4 are upregulated with HK. No significant changes were observed in WNK1-L expression across the experimental groups. (B and C) ISH analysis for WNK1-S and WNK4 detected any major shifts in distribution and level of WNK expression with K+ intake at the regional level, undiluted by any widespread, invariant low-level expression. (i) Representative sections and (ii) densitometric analysis from ISH studies. WNK1-S (B) and WNK4 (C) have different expression profiles with WNK1-S expression restricted to the cortex, and WNK4 restricted to cortex and outer medulla. WNK1-S and WNK4 expression both are upregulated with HK. *P = 0.009; **P = 0.04; ***P = 0.01; #P = 0.003; ##P = 0.02.
In regions with clear WNK1-S expression (Figure 7B), ISH analysis showed upregulation in cortex by HK (versus NK: 2.1 ± 0.6-fold; 95% confidence interval 1 to 3.3; P = 0.003). WNK1-S distribution remained cortical without striking change. Over kidney regions with clear WNK4 expression, there was a 2.3 ± 0.6-fold (1.1 to 3.5) upregulation in expression with HK (P = 0.02; Figure 7C); in contrast, the LK diet did not affect WNK4 mRNA significantly.
For testing whether varied dietary Na+ induces similar WNK expression changes, mice were fed diets with a specific Na+ content (normal Na+ [control], low Na+ [LNa], or high Na+ [HNa]) for 6 d. WNK1-S showed a borderline significant downregulation, by 39 ± 16% of control levels, between HNa and LNa-diet (P = 0.049, ANOVALSD; Figure 8). Although WNK4 in particular showed a trend to similar changes, no other significant changes in WNK expression were observed across the treatment groups.
WNK expression in response to varied Na+ intake. The Na+ diet study involved adult male C57BL/6J mice, housed in pairs in mouse metabolic cages (Techniplast). Animals were allowed 1 wk to acclimate, after which they commenced specific dietary Na+ treatments that lasted 7 d: (1) low Na+ (0.03%) feed group (LNa), (2) normal/control Na+ (0.3%) feed group (NNa), or (3) high Na+ (3%) feed group (HNa). WNK1-S showed a marginally significant* downregulation (by ANOVALSD) between HNa and LNa diet. No other significant changes in WNK expression were observed across the treatment groups. *P = 0.049.
WNK Expression and Aldosterone Challenge
For examination of the effect of aldosterone levels on WNK expression, mice were given excess aldosterone via minipump (150 μg aldosterone/kg per d [18]) or adrenalectomized (supplemented with 0.9% saline drinking water) to abolish aldosterone production. Real-time PCR showed that aldosterone treatment induced a 32 ± 6.7% (18.9 to 45.2) WNK1-S upregulation (P = 0.0002) without affecting WNK1-L or WNK4 expression (Figure 9). Adrenalectomy had no significant effect on WNK expression, but WNK1-S rose significantly across the adrenalectomy-aldosterone excess range (43 ± 7.6%; P = 0.0002).
WNK expression in response to variations in aldosterone. The aldosterone study involved adult male C57BL/6 mice, housed in pairs in normal cages. Mice that had food and fluid intake and BP monitoring were given (n = 6) subcutaneous (Alzet) minipump (MP) treatments for 6 d: (1) Adrenalectomy (ADX): Bilateral adrenalectomy, 0.3% Na+ feed, 0.9% saline drinking water, saline-only MP; (2) control (CTRL): Normal 0.3% Na+ feed, saline-only MP; and (3) aldosterone excess (ALDO 150): 150 μg aldosterone/kg per d (18) by MP, 0.3% Na+ feed. Plasma renin activity and aldosterone measurements allowed confirmation of adequacy of treatments. WNK1-S expression was upregulated by chronic aldosterone treatment (150 μg/kg per d, 6 d) but was unchanged in the absence of aldosterone after adrenalectomy. No changes in WNK1-L or WNK4 were observed across the experimental groups. *P = 0.0002; **P = 0.02.
Discussion
In beginning to understand the WNK pathway, Xenopus oocyte studies provided invaluable evidence of WNK pathway regulation of key mediators of distal nephron electrolyte transport. This study is one of the first to investigate this pathway in vivo in a much more physiologically relevant system, the mouse, reporting detailed nephron segment localization and WNK expression responses to dietary electrolyte challenges. The relevant aspects of human physiology and their disorders are very well modeled in mice, particularly mechanisms of electrolyte handling and associated effects on long-term BP control (19,20). Very few tools that reliably differentiate WNK isoforms are currently available to quantify expression changes and examine distribution simultaneously. Moreover, this study allows examination of changes within kidney regions that are not easily accessible to micropuncture techniques and have no good, well-validated, cell-line models (e.g., outer medullary CD, deep distal nephron arcades/CNT) and avoids dangers of unequal RNA degradation, a concern that is associated with microdissection.
Here we report strongest WNK1 and WNK4 expression in the distal tubule (DCT, CNT) with WNK1-S dropping to much lower levels by CD, whereas WNK4 extends somewhat diminished into TAL and CD. WNK1-L has widespread, low-level, near-background expression. Figure 10 puts the distribution of WNK pathway expression in context.
Distribution of gene expression of key genes involved in Na+ and K+ balance in distal nephron. Structure illustrated in Figure 3. mTAL, cTAL, medullary and cortical TAL; CNT/ARC, connecting tubule/arcade; CCD/OMCD/IMCD, cortical/outer medullary/inner medullary CD. αβγENaC, αβγ subunits of ENaC; ROMK, renal outer medullary K channel; WNK1-S/-L/-T, WNK1 short (kinase deficient), long (kinase intact), and total; sgk1, serum and glucocorticoid kinase 1. For αENaC/sgk expression, arrows represent regulated expression by rising aldosterone (A).
WNK4 expression in TAL and macula densa (discussed further below) has not been recognized before and may indicate different WNK4 isoforms or posttranslational modifications in TAL (3). These are common in transport pathways (e.g., NKCC2 in TAL), whereas additional bands on WNK4 immunoblots suggest varied posttranslational modification (3). A study of transgenic mice that express a FHHt-WNK4 cDNA in TAL-CNT (and intercalated cells) reported that mutant-WNK4 protein was absent from tight junctions and apically localized in TAL (21). The lack of FHHt phenotype in these FHHt-WNK4 mice leaves some uncertainty. This may relate to transgenic WNK4 expression being driven from a cDNA and so lacking normal in vivo regulation and potential for transcriptional diversity of the WNK4 genomic locus.
Our studies challenged mice with varied dietary K+ and Na+ intake and aldosterone. All three classes of treatment showed notable WNK pathway gene expression responses in vivo to these physiologic determinants of electrolyte balance and BP. Changes with varied dietary K+ intake were particularly clear. WNK1-S expression rose on HK and fell on LK diet, correlating significantly with K+ intake, while ISH findings revealed the importance of upregulation of this isoform with HK diet, in strongly expressing segments (DCT-CNT). HK intake also increased WNK4 expression. These coordinated WNK expression changes seem functionally significant as merely heterozygous changes in WNK1 or WNK4 cause substantial BP and electrolyte abnormalities (1,3,5,6).
WNK1-S was significantly upregulated with chronic aldosterone excess. There were no significant changes in WNK1-L or WNK4 across the aldosterone-adrenalectomy range. With variations in Na+ intake, a fall in WNK1-S expression just reached significance comparing HNa and LNa groups. It is intriguing that this could represent a WNK pathway response to reductions in extracellular fluid volume as Na+ intake falls.
Thus, K+ intake, plasma aldosterone, and dietary Na+ intake/extracellular volume seem to be possible in vivo regulators of the WNK pathway. K+ intake seems to be a relatively robust regulator, whereas in some circumstances, the role of aldosterone may be counterbalanced or secondary to another regulator. Thus, similar WNK1-S expression accompanies very different aldosterone elevations (2.3-fold [HK] and 11.2-fold [aldosterone]) or when aldosterone fell from normal to adrenalectomized levels (Figure 9). Aldosterone-independent regulation of WNK1-S expression clearly is present across LK-HK dietary groups, with individual WNK1-S expression correlating with K+ intake (P < 0.001) but not significantly with aldosterone. Moreover, reduced WNK1-S with LNa intake (lowering volume, stimulating secondary hyperaldosteronism) and increased WNK1-S with aldosterone (raising volume, primary hyperaldosteronism) implies that WNK1-S responds more to extracellular volume than aldosterone and suggests falling WNK1-S as a potential response to conserve volume. Further investigation is required to define fully the in vivo regulatory roles of K+ intake, aldosterone, and volume acting on the WNK pathway, but integrating these findings into a more functional context is of interest. One attempt to do so is outlined in Figure 11 and discussed further next.
Potential WNK pathway and two-compartment model of distal nephron Na+/K+ handling. (A) The WNK pathway is suggested to regulate the extent of NaCl reabsorption in (early) DCT in the manner of the alleged molecular switch proposed in explaining familial hyperkalemic hypertension (FHHt). A high K+ intake switches to lower Na+ reabsorption via NCC in DCT, so distal Na+ delivery rises, allowing K+ to be cleared. (B) Two-compartment model. (i) Normally, Na reabsorption is partitioned between an upper co-transport compartment (under WNK pathway repression), driving NaCl reabsorption and a lower electrogenic compartment (stimulated by aldosterone and Na+ delivery), where Na+ reabsorption via ENaC generates a lumen-negative charge facilitating K+ secretion. (ii) In hypovolemia, both compartments upregulate Na reabsorption, minimizing urinary NaCl loss and facilitating K+ secretion. (iii) In high K intake, the mechanism above delivers increased Na and aldosterone, which strongly drive Na+ reabsorption and K+ secretion. (iv) In FHHt, there is constitutive overreabsorption of NaCl, causing hypervolemia and hypertension with low renin, which limits the distal delivery to the electrogenic compartment without renin suppression and blunted aldosterone, promoting limited K secretion and hyperkalemia. WNK pathway regulation of NCC and the extent of Na+ delivery to the distal electrogenic compartment and inappropriate restriction of this in FHHt fits well with reported features of FHHt, including thiazide-sensitivity and NaCl dependence.
How do these results expand understanding of this novel WNK pathway? The FHHt phenotype and usual hyperresponsiveness to thiazide diuretics highlight the importance of DCT in electrolyte balance and BP control (1,5,22). The work above provides some of the first clues as to the role in vivo of WNK1-S, the predominant WNK1 isoform in kidney, and intriguingly shows that strongest WNK1-S and WNK4 expression co-localizes in DCT-CNT, where they may contribute to a mechanism that regulates K+ homeostasis and BP. Xenopus oocyte work suggests that WNK4 may inhibit NCC in DCT, and WNK1-L prevents this inhibition (10,11). In a preliminary presentation of this work (see Acknowledgments), we proposed the hypothesis that both WNK1-S and WNK4 can limit NCC transport (Figure 11A). Xenopus oocyte work reported since supports a similar mechanism (13). We propose, especially in DCT-CNT, that WNK1-S could bind and counterbalance WNK1-L effects, so shielding WNK4 from inhibition. Although WNK1-S is kinase deficient, it retains domains (coiled-coils) that likely facilitate multimeric/tetrameric WNK1 assembly (23,24). Alternatively, WNK1-S could interact directly with a WNK-binding site on WNK4 or another regulatory kinase to repress NCC.
Figure 11 incorporates this WNK pathway in a two-compartment model. In the upper co-transport compartment, increased WNK1-S or WNK4 (as in our in vivo studies) downregulates NCC-mediated NaCl reabsorption, increasing Na+ delivery to the lower electrogenic compartment (from late DCT distally), where ENaC reabsorbs Na+, facilitating K+ secretion (19) (itself augmented by high distal flow-mediated Maxi-K channel activation). The two compartments will normally overlap in late DCT. LNa diet/hypovolemia and HK diet both will stimulate aldosterone, activating the lower compartment. The in vivo studies above indicate that LNa diet/hypovolemia also will stimulate the upper compartment (Figure 11B[ii]), producing appropriate Na+ retention, whereas HK diet will repress it, producing appropriate K+ secretion (Figure 11B[iii]). The upper co-transport compartment thus interconverts the same aldosterone response between Na+ retention and K+ excretion. We propose that the molecular switch that is alleged to explain FHHt (12) is based on distal delivery of Na+. This hypothesis is supported by careful ultrastructural studies showing that HK diet (25), NCC−/− Gitelman mice, and high-dosage thiazide diuretics (19) all predispose to extensive hypertrophy and increased Na+/K+-ATPase (25) in early CNT (indicating higher Na+ delivery) and predispose to greater K+ clearance and reduced BP (19). In WNK1+/− mice, global WNK1 reduction would repress NaCl reabsorption (in DCT, unopposed WNK4) and promote lower extracellular volume, in keeping with their lower BP.
FHHt jams the switch in the opposite direction (Figure 11B[iv]), inappropriately engaging a response (reducing distal Na+ delivery and K+ secretion) that these studies suggest is normal when body K and/or Na/extracellular volume fall and require conservation. Although aldosterone level is low to normal, its contribution to FHHt pathophysiology and hypertension should not be underestimated (26). Both aldosterone-dependent and -independent mechanisms contribute to K+ secretion (19,27); blunting of both seems likely in FHHt. Thus, distal Na+ delivery is blunted and hypervolemic suppression of renin will restrain circulating aldosterone, blunting the aldosterone response to levels that are inadequate to restore normokalemia despite hyperkalemic drive (28). Thus, FHHt causes hypertension and hyperkalemia.
In vitro evidence suggests the WNK pathway also may directly (i) reduce surface ROMK (12), (ii) promote ENaC conductance (14), and (iii) increase paracellular Cl− flux (29,30) (depleting electrogenic lumen-negative charge). Although the pathway in Figure 11 seems not to require these effects, these processes could impair K+ secretion, promote Na+ reabsorption, or both if they contributed significantly. Effect (ii) depends on WNK1-S and was demonstrated in CCD cells; it is unresolved if it extends to late DCT-CNT, where the expression of WNK1-S and ENaC and the potential physiologic influence all are stronger. The roles and significance in vivo, and in FHHt, of effects (i) through (iii) are not yet clear.
It is intriguing that we have found previously unexpected gene expression of WNK4 in TAL and macula densa. WNK4 potentially could influence TAL NaCl transport via regulation of ROMK, NKCC2, CLC-Kb, or Barttin, because inactivation of any causes severe NaCl wasting in Bartter’s syndromes. ROMK surface localization was unaffected in transgenic mice that expressed FHHt-mutant WNK4 protein, which was apically distributed in TAL (21). In vitro, WNK4 can interact directly or with other kinases (e.g., OSR1, SPAK, other WNK), to regulate proteins of key importance in TAL-DCT transport, including NCC (11,12), ROMK (12), and, it seems, NKCC2 (31,32). Hence, WNK4 might influence NKCC2 co-transport and expand the co-transport compartment (Figure 11) to more powerful proportions. Overactivity of TAL-DCT NaCl reabsorption seems compatible with FHHt. Certainly, increased BP is reported with activating mutation of CLC-Kb (T481S) (33). Moreover, considering phenotypes of Gitelman plus Bartter syndromes (DCT+TAL salt-wasting hypokalemic alkalosis), it seems that the inverse of these may encompass hyperkalemia, acidosis, and low-renin hypertension, all FHHt features. Other WNK4-FHHt features (e.g., degree of thiazide sensitivity, hypercalcuria [34]) might depend on the spectrum of overactivation within DCT-TAL. Clearly, this makes WNK4 expression in TAL of interest, but much needs to be clarified before a role in physiology or FHHt pathophysiology could be attributed.
Acknowledgments
We acknowledge The Wellcome Trust (grant 065616 PhD Studentship to M.D.), British Heart Foundation (grant PG2 0/01075), and Scottish Hospitals Endowment Research Trust (grant 77/00) for support. The Wellcome Trust Clinical Research Facility enabled real-time PCR and image analysis studies.
Portions of this work were presented at the 24th British Hypertension Society Meeting, September 13 through 15, 2004, Cambridge, UK; the 37th annual meeting of American Society of Nephrology, October 29 through November 1, 2004, St. Louis, MO; and the 24th British Endocrine Societies’ Meeting, April 4 through 6, 2005, Harrogate, UK; and appear in abstract form (Endocr Abstr 9: 138, 2005).
Footnotes
Published online ahead of print. Publication date available at www.jasn.org.
- © 2006 American Society of Nephrology