Molecular regulation of NKCC2 in blood pressure control and hypertension

Abstract

Purpose of review

The apical cotransporter NKCC2 mediates NaCl reabsorption by the thick ascending limb (TAL), contributing to maintenance of blood pressure. Despite effective NKCC2 inhibition by loop diuretics, these agents are not viable for long-term management of blood pressure due to side effects. Novel molecular mechanisms that control NKCC2 activity reveal an increasingly complex picture with interacting layers of NKCC2 regulation. Here we review the latest developments that shine new light on NKCC2-mediated control of blood pressure and potential new long-term therapies to treat hypertension.

Recent findings

Emerging molecular NKCC2 regulators, often binding partners, reveal a complex overlay of interacting mechanisms aimed at fine tuning NKCC2 activity. Different factors achieve this by shifting the balance between trafficking steps like exocytosis, endocytosis, recycling and protein turnover, or by balancing phosphorylation vs. dephosphorylation. Further molecular details are also emerging on previously known pathways of NKCC2 regulation, and recent in vivo data continues to place NKCC2 regulation at the center of blood pressure control.

Summary

Several layers of emerging molecular mechanisms that control NKCC2 activity may operate simultaneously, but they can also be controlled independently. This provides an opportunity to identify new pharmacological targets to fine-tune NKCC2 activity for blood pressure management.

Introduction

The apical Na⁺/K⁺/2Cl^- cotransporter NKCC2 (SLC12A1) is the gateway for NaCl reabsorption in the thick ascending limb (TAL) of the loop of Henle (1, 2). NKCC2 allows the TAL to generate a hyperosmotic renal medulla via the countercurrent multiplier mechanism, playing a key role in control of blood pressure, as demonstrated by loss of function NKCC2 mutations in mice (3) and humans, referred to as Bartter syndrome (4–8). NKCC2 is also the target of loop diuretics, which lower blood pressure (9–12). More recently, NKCC2 has been linked to salt-sensitive hypertension in animal models (13–17) and humans (18, 19) in particular African-Americans (20), a population susceptible to salt-sensitivity. The importance of NKCC2 in regulation of blood pressure has been long recognized and reviewed in depth recently (21). In recent years, efforts to understand NKCC2 function have made clear that regulation of the co-transporter at the molecular level is paramount to control of blood pressure under physiological and pathological circumstances. Several recent reviews can provide an overview on different aspects of NKCC2 regulation (22–25). In the present article, we will discuss recent developments in established molecular mechanisms of NKCC2 regulation such as trafficking to the apical membrane and phosphorylation, as well as novel NKCC2 interactions with regulatory factors that open up exciting new avenues in efforts to control blood pressure.

Apical trafficking of NKCC2

At steady state, approximately 3–5% of NKCC2 is at the TAL apical surface (26) while the rest is at an intracellular sub-apical location (27). In the last decade it has been established that this small percentage of surface NKCC2 is tightly regulated via trafficking into and out of the apical membrane (17, 26, 28–33), with profound consequences for blood pressure (13, 31, 34, 35). NKCC2 has a half-life of approximately 1 hour at the apical membrane maintained by exocytosis, endocytosis and recycling (36). However, an additional mechanism of regulation acts on degradation of the internalized NKCC2 pool (29). Recent evidence shows that internalized NKCC2 degradation is accelerated by the second messenger cGMP via the ubiquitin-proteasome system (37). Although the details of this process (e.g. ubiquitin ligases and signaling involved) remain to be determined, these observations add to an emerging picture where NKCC2 apical trafficking and protein turnover are subject to fine tuning by the opposing actions of inhibitory second messengers like cGMP, and stimulatory factors such as cAMP and reactive oxygen species, which enhance NKCC2 at the apical surface. These second messengers provide the platforms for physiological regulation of TAL-mediated NaCl absorption by hormones, neurotransmitters and variations in luminal flow. A recent example of the latter is luminal flow-induced superoxide, a reactive oxygen species produced by NADPH oxidases, which was found to stimulate NKCC2-mediated Na⁺ absorption by the TAL (38). Luminal flow-stimulated NKCC2 activity was prevented by tetanus toxin, a protease that cleaves vesicle fusion proteins VAMP2 and VAMP3, and previously shown to mediate cAMP stimulation of NKCC2 (26). These results show that different pathways converge on mechanisms that control trafficking of NKCC2 to the apical surface. However, the precise molecular mediators may differ, as we have shown that individual silencing of VAMP2 and VAMP3 mediate different routes of NKCC2 trafficking to the membrane. Specifically, VAMP2 mediates a cAMP-stimulated pathway of NKCC2 exocytic insertion (32), whereas VAMP3 mediates a separate constitutive pathway of NKCC2 exocytosis (31). This separation of trafficking routes may allow different mechanisms of regulation of blood pressure since we observed lower blood pressure in VAMP3 knockout mice (31). We anticipate that additional molecular mediators of NKCC2 trafficking will provide the opportunity to modulate NKCC2 trafficking to develop more effective loop diuretics.

Another novel regulator of NKCC2 surface expression is the melanoma-associated antigen D2 (MAGE-D2), recently identified by whole-exome sequencing in families affected by transient antenatal Bartter’s syndrome with polyhydramnios (39). This study found that patients with a MAGE-D2 mutation had decreased NKCC2 presence at the apical membrane of TALs. In vitro studies in HEK293T cell cultures confirmed that the presence of MAGE-D2 increased surface NKCC2 expression and activity, presumably by protecting the maturation of NKCC2 in the biosynthetic pathway (39). These results are exciting although the stimulatory role of MAGE-D2 on NKCC2 trafficking still remains to be confirmed in native TALs. However, this novel regulator of NKCC2 certainly opens up new possibilities for control of NKCC2 activity. Based on co-immunoprecipitation results, the authors speculate that possible mechanisms of action of MAGE-D2 may include interactions with at least two proteins, the chaperone Hsp40 and the G-protein Gs-alpha (40). Hsp40 may protect from endoplasmic reticulum–associated degradation, a process described by the same group to regulate NKCC2 exit from the endoplasmic reticulum in OK and HEK-293 cell lines (41–43). Gs-alpha stimulates cAMP production by adenylyl cyclases, which we have shown is an intermediate step in hormonal stimulation of NKCC2 exocytic delivery in TALs (30). Although these links between MAGE-D2 and NKCC2 still need to be confirmed in native TALs, we anticipate that future research on MAGE-D2 will deepen our understanding on the molecular mechanisms that control NKCC2 abundance at the apical membrane.

As we progress towards understanding the complex machinery involved in the dynamic balance between NKCC2 apical trafficking and protein turnover, we should also focus our attention on identifying new dietary factors, hormones and metabolites that affect NKCC2 function as they may be relevant to blood pressure. In our laboratory, we have focused on the effects of high salt intake previously and on the effects of dietary sugars more recently. The reason for this is the reported increase in added sugar consumption and its relationship to cardiovascular disease and hypertension, in particular fructose, which has seen a dramatic increase in western diets due to its presence in high-fructose-corn-syrup. We focused on fructose given its potential to induce salt-sensitive hypertension in normal rats (44–46). We have recently shown that fructose itself, directly applied to isolated perfused TALs, stimulates NKCC2-mediated NaCl transport by increasing the presence of NKCC2 at the apical membrane (34). This increase was specific to fructose, since glucose had no effect on NKCC2 activity. The molecular and trafficking mechanisms that mediate the fructose-induced increase in surface NKCC2 remain to be determined. As pointed out earlier, it is likely that regulation of NKCC2 trafficking and protein turnover are at the core of the process. Interestingly, NKCC2 phosphorylation at key threonines (Thr-96,101) was not affected by fructose (34). In addition to fructose, thousands of urinary metabolites, some derived from bacterial metabolism in the gut, are concentrated in the loop of Henle, and some of these may directly influence NKCC2 trafficking and activity. This is an area of research that has received little attention but is likely to play a role in blood pressure control through direct action in the nephron.

Altogether, recent reports highlight the importance of understanding the several steps in the process of NKCC2 trafficking as they may control separate aspects of regulation of NKCC2 activity in the TAL. This emerging picture of NKCC2 trafficking and protein turnover may provide points of action for regulatory factors from endogenous pathways like second messengers and regulatory proteins, but also from external sources like dietary fructose.

Regulation of NKCC2 by phosphorylation

NKCC2 can be phosphorylated at the cytoplasmic tails at Thr96, Thr101, Ser87, Ser126 (amino-terminus) and Ser874 (carboxy-terminus) (47–50) via PKA (50), SPAK (Ste20-related proline alanine rich kinase [STK39]) or OSR1 (oxidative stress responsive kinase 1 [OXSR1]) (51–53). We refer the reader to a previous review (36). In general terms, hormones like Arginine-vasopressin (AVP) phosphorylate Ser126,874 via cAMP and PKA (50), whereas SPAK and OSR1 phosphorylate Thr96,101 (51–53).

The importance of these pathways in control of blood pressure is well known. We have shown enhanced baseline NKCC2 phosphorylation at Thr96,101 in TALs from Dahl salt-sensitive rats preceding the development of hypertension (17). NKCC2 stimulation by AVP not only relates to phosphorylation, but also to enhanced apical membrane trafficking as recently confirmed in vivo in TAL-specific vasopressin V2 receptor transgenic rats (54). Consistent with NKCC2 downregulation, these mutants had polyuria and reduced ability to concentrate the urine, but unfortunately the blood pressure was not measured. The importance of SPAK and OSR1 in maintenance of blood pressure was established in SPAK and kidney-specific OSR1 knockout mice, both of which had reduced blood pressure (55, 56). However, although these models indicated that OSR1 is the predominant kinase regulating NKCC2, while SPAK acts mainly on the sodium-chloride co-transporter (NCC) in the distal convoluted tubule, these two kinases are emerging as components of a more complex system of NKCC2 regulation. Recently developed double SPAK/OSR1 knockout mice had an unexpected less severe phenotype than the individual knockouts and enhanced NKCC2 Thr96,101 phosphorylation (57). The authors attributed these surprising observations to compensation by a yet-to-be identified kinase.

Additional pieces of evidence further support in vivo convergence of signaling pathways that may involve unidentified kinases. For instance, baseline NKCC2 Thr96,101 phosphorylation is observed in isolated rat or mice TALs in the absence of any stimulation (17, 48). The kinases responsible are unknown. The Thr96,101 residues are not within any PKA consensus site, but their phosphorylation is enhanced by AVP (48, 54). Similarly, the PKA site Ser126 becomes phosphorylated in cell lines placed under hypotonic low-chloride (53), a condition known to activate SPAK and OSR1. In fact, SPAK and OSR1 mediate an intracellular Cl^- sensing mechanism (51, 53, 58) as downstream effectors of with-no-(K) lysine kinases WNK1, WNK3 and WNK4. The relevance of WNKs to blood pressure control is a topic with open questions. Blood pressure in WNK1 knockout mice is only slightly reduced (59) or unchanged (60). A dominant-negative kidney-specific WNK1 splice variant does not alter the ratio of total NKCC2 vs. phospho-Thr96,101 NKCC2 (61). WNK3 deficient mice do not have a blood pressure phenotype (62, 63). WNK4 was first recognized as a regulator of NCC in the distal convoluted tubule, due to mutations causing familial hyperkalemic hypertension that stimulate NCC via SPAK (56, 64). Probably because of the strong effect of WNK4 in the distal tubule, it was difficult to discern its role in the TAL, as early reports in WNK4 knockouts could not find any change in NKCC2 (65, 66). It was not until very recently that WNK4 was shown to stimulate NKCC2 phosphorylation and activity in vivo (67), contributing to part of the familial hyperkalemic hypertension phenotype. WNK4 is regulated via proteasomal degradation as part of a multiprotein complex containing the scaffold protein Cullin 3 and E3 ubiquitin ligase. A recent study described a mechanism by which a truncated form of Cullin 3 known to cause familial hyperkalemic hypertension fails to target WNK4 for degradation, thereby upregulating WNK4 and SPAK/OSR1 and raising blood pressure (68). The transgenic mice used to describe this model had increased phosphorylation of NCC and also NKCC2 at Thr96,101, further suggesting that mechanisms of regulation of SPAK/OSR1 have the potential of modulating blood pressure via NKCC2 phosphorylation.

In addition to upstream WNK kinases, little is known about hormonal and autacoid-mediated regulation of SPAK and OSR1. One such factor may be the natriuretic autacoid Nitric Oxide (NO), which in the TAL is primarily produced by NO Synthase 3 (NOS3) and inhibits NKCC2 (69). Recently developed nephron-specific NOS3 knockout mice have distinguished the renal role of NOS3 from extrarenal tissues (70). These mice had increased phosphorylation of NKCC2, NCC and SPAK/OSR1, and they had difficulties excreting a salt load, which also raised their blood pressure higher than wild-type mice. Although the authors did not elaborate on possible molecular links between NOS3, NO and SPAK/OSR1, this study further highlights the overlap among pathways of NKCC2 phosphorylation.

NKCC2 regulation by protein-protein interactions

The array of NKCC2-binding proteins has expanded since establishing that SPAK and OSR1 bind NKCC2 (52). We already discussed these kinases as part of a growing network of regulation. The most recently identified member of this network may be the phosphatase Calcineurin, which binds the amino- and carboxy-terminal tails of NKCC2 and dephosphorylates Thr96,101 directly (71), therefore counteracting SPAK and OSR1. A recent study showed that chronic, systemic inhibition of calcineurin with cyclosporin A enhanced NKCC2 phosphorylation, salt retention and hypertension in rats (72). Because the overall effect of cyclosporin A may include actions in the TAL, the distal tubule and the renin-angiotensin axis, future research will need to address the specific contribution of calcineurin to NKCC2-dependent regulation of blood pressure.

Additional NKCC2-binding proteins identified in the last decade regulate different steps in NKCC2 trafficking and are discussed elsewhere (25, 36). These interacting proteins can either increase the presence of NKCC2 at the plasma membrane, like MAL (73), Annexin A2 (74), moesin (75), VAMP2 (32) and VAMP3 (31), or alternatively decrease surface NKCC2, e.g. Aldolase B (76), SCAMP2 (77), OS9 (41) and the most recently identified ALMS1 (35). Recent research in whole-animal models highlights the physiological relevance of NKCC2 binding partners. We showed that NKCC2-interacting vesicle fusion protein VAMP3 is required for normal fluid and ion excretion and maintenance of blood pressure (31). Moesin, a protein that links cargoes to the actin cytoskeleton, interacts with the carboxy-terminus of NKCC2 and increases NKCC2 surface expression in cell lines (75). A recent study in moesin knockout mice confirmed the moesin-NKCC2 interaction and stimulation in TALs, but unfortunately blood pressure was not measured (78). Interestingly, although both studies conclude that moesin promotes surface NKCC2 expression, the mechanisms proposed are different, probably due to the assays used to measure NKCC2 trafficking. The original report used a cell line expressing a NKCC2/NKCC1 chimera and determined that moesin stimulates NKCC2 exocytosis (75). In contrast, the most recent work concluded that moesin decreases NKCC2 internalization as measured in freshly isolated TAL preparations (78).

A novel regulator of NKCC2 trafficking and activity is the Alström syndrome 1 (ALMS1) protein, which binds the carboxy-terminus of NKCC2 (35) at a sequence known to target NKCC2 to the apical membrane (79). We studied the role of ALMS1 in NKCC2 regulation and blood pressure in ALMS1 knockout rats, and via adenovirus-mediated shRNA gene silencing in vivo. We found that knocking down ALMS1 caused NKCC2 accumulation at the apical membrane of TALs due to impaired endocytosis, therefore increasing NKCC2 activity (35). ALMS1 knockout rats were hypertensive on a normal salt diet, exhibited salt-sensitivity and decreased ability to excrete a Na⁺ load (35). This study confirmed the relationship between ALMS1 and hypertension that had been only described in association genetic studies in Alström syndrome patients, but it also provided a renal mechanism by which ALMS1 may control blood pressure. While we found ALMS1 was abundant in the TAL, it was also detectable at lower levels, by immunofluorescence, in other nephron segments. Thus, we cannot rule out a role of ALMS1 in other nephron segments or extra renal tissues. This limitation also applies to other NKCC2 binding partners and future studies should benefit from the generation of TAL-specific knockouts.

Conclusion

Given the importance of NKCC2 in TAL-mediated NaCl transport and salt-sensitive hypertension, there has been a recent interest in the molecular mechanisms that control NKCC2 activity by trafficking or phosphorylation, and the molecular partners and signaling cascades involved (Figure 1). We have discussed how NKCC2 regulation is the result of the convergence of multiple layers of control. An open question that has not been directly addressed in the TAL is whether trafficking and phosphorylation are independently controlled, as the available literature offers examples of both co-dependent and independent regulation. The two processes are not necessarily mutually exclusive. However, answering this open question may rely on knowing all possible phosphorylation sites and the subcellular localization where this event occurs. This is a task that may benefit from phospho-proteomic studies and molecular imaging with enhanced resolution.

Figure 1: Novel mechanisms of NKCC2 regulation by phosphorylation and trafficking to the apical membrane.

The most recently identified regulators and mechanisms discussed in this review are bolded. Abbreviations: TAL: Thick ascending limb, Ad Cy: adenylyl cyclase, AVPR2: Vasopressin receptor 2, ß2AR: Beta-2 adrenergic receptor, ERAD: endoplasmic reticulum associated degradation, NOX4: NADPH oxidase 4, WNK: With-no-lysine kinase, MAGE-D2: melanoma-associated antigen D2, OS9: amplified in osteosarcomas, VAMP: Vesicle-associated membrane protein, CnAβ: calcineurin Aβ. Dotted lines represent mechanisms that affect NKCC2 but the proposed pathway has not been verified experimentally.

As the different layers of NKCC2 regulation come together, their relevance will continue to emerge from in vivo studies with assessment of renal and blood pressure phenotypes. Ultimately, any gene or interacting protein that affects NKCC2 trafficking, phosphorylation or ubiquitination may influence TAL NaCl reabsorption and blood pressure. We anticipate that future research will benefit from the discovery of additional regulators via high-throughput techniques like proteomics and next-gen sequencing, and from use of transgenic models that dissect the molecular mechanisms operating in a tissue specific fashion.

Key points.

1. Apical trafficking of NKCC2 is a mechanism subject to regulation by recently characterized factors like VAMPs, MAGE-D2, flow-stimulated NOX4, and dietary fructose.

2. NKCC2 endocytosis and protein turnover provide additional layers of control mediated by cGMP-stimulated proteasomal degradation of ubiquitinated NKCC2.

3. NKCC2 phosphorylation at Thr-96,101 may result from the convergence of several pathways mediated by SPAK, OSR1, WNKs, PKA, NOS3, the phosphatase calcineurin, and yet-to-be identified kinases.

4. Major advances on mechanisms of NKCC2 regulation are being pushed forward by the discovery of novel NKCC2 binding partners like ALMS1 and moesin, which control surface NKCC2 expression in vivo by mediating endocytic retrieval from the plasma membrane.

5. Recent experiments in animal models emphasize the central role of NKCC2 regulation in control of blood pressure including salt-sensitive and transgenic rats, and knockout mouse models for VAMP3, WNK4, NOS3 and combined knockout of SPAK/OSR1.