The renal H⁺-K⁺-ATPases: physiology, regulation, and structure

Abstract

The H⁺-K⁺-ATPases are ion pumps that use the energy of ATP hydrolysis to transport protons (H⁺) in exchange for potassium ions (K⁺). These enzymes consist of a catalytic α-subunit and a regulatory β-subunit. There are two catalytic subunits present in the kidney, the gastric or HKα₁ isoform and the colonic or HKα₂ isoform. In this review we discuss new information on the physiological function, regulation, and structure of the renal H⁺-K⁺-ATPases. Evaluation of enzymatic functions along the nephron and collecting duct and studies in HKα₁ and HKα₂ knockout mice suggest that the H⁺-K⁺-ATPases may function to transport ions other than protons and potassium. These reports and recent studies in mice lacking both HKα₁ and HKα₂ suggest important roles for the renal H⁺-K⁺-ATPases in acid/base balance as well as potassium and sodium homeostasis. Molecular modeling studies based on the crystal structure of a related enzyme have made it possible to evaluate the structures of HKα₁ and HKα₂ and provide a means to study the specific cation transport properties of H⁺-K⁺-ATPases. Studies to characterize the cation specificity of these enzymes under different physiological conditions are necessary to fully understand the role of the H⁺-K⁺ ATPases in renal physiology.

the h⁺-k⁺-atpases use the energy of ATP hydrolysis to pump hydrogen (H⁺)¹ and potassium (K⁺) ions against their concentration gradients. Because they form a high-energy phosphorylated intermediate during the catalytic cycle, these enzymes are classified as P-type ATPases. They consist of two subunits. The large, ∼100-kDa, α-subunit contains 10 transmembrane helices and houses the catalytic site and the ion translocating pathways. A second, smaller glycosylated β-subunit, ∼30 kDa, is required for proper trafficking and processing of the enzyme.

The first evidence for an H⁺-K⁺-ATPase in the kidney came from studies performed in the outer medullary collecting ducts (OMCD) of rabbits maintained on low-K⁺ diets (93). Total CO₂ flux (a measure of H⁺ secretion) and K⁺ flux were recorded, and both were sensitive to omeprazole, an H⁺-K⁺-ATPase-specific inhibitor. Since then, two renal H⁺-K⁺ ATPases have been identified that contain the catalytic subunits HKα₁ or HKα₂. The HKα₁ H⁺-K⁺-ATPase, or gastric H⁺-K⁺-ATPase, is well known for its role in acidifying the stomach contents. The HKα₂ H⁺-K⁺-ATPase has been termed the nongastric or colonic H⁺-K⁺-ATPase and was found to be abundant in and cloned from the colon (20). The HKα₁ H⁺-K⁺-ATPase is sensitive to micromolar concentrations of imidazopyridine, 2-methyl-8-(phenylmethoxy)imidazol[1,2a]-pyridine-3-acetonitrile (Sch-28080) (87), and the seminaphthoquinone A80915A (21),² whereas relatively high concentrations of ouabain, a classic inhibitor of the Na⁺-K⁺-ATPase, have been used to assess the physiological contribution of the HKα₂ H⁺-K⁺-ATPase (16, 85). Both enzymes acidify the tubular fluid and reabsorb K⁺ in the kidney. For more information on the molecular regulation and pharmacology of these enzymes, the reader is referred to several recent review articles (5, 15, 67, 74, 107). In the following sections, we summarize novel information about the physiological role, regulation, and structure of renal H⁺-K⁺-ATPases.

Physiology of Renal H⁺-K⁺-ATPases

Localization of HKα₁ and HKα₂ expression in the kidney.

Studies to localize HKα₁ and HKα₂ mRNA and protein expression throughout the kidney have been performed in rat, rabbit, and mouse (Table 1). While some differences in the activity and ion affinity profiles may be attributed to species-specific factors (discussed below), the localization of renal HKα₁ and HKα₂ is largely consistent between species. Both isoforms primarily localize to the collecting duct (CD), although several reports suggest HKα₂ expression in the thick ascending limb (TAL) (47, 83) and connecting tubule (CNT) (2, 24, 83). HKα₁ mRNA and protein were identified in the cortex of rat (1, 6) and rabbit (95), in the outer medulla of mouse (55), rat (1, 6), and rabbit (95), and in the inner medulla of mouse (55) and rat (1, 6). Specifically, HKα₁ was localized to intercalated cells of the OMCD of rat (1, 6) and rabbit (95). HKα₂ is expressed in the CD of rat (2), rabbit (83), and mouse (99) as well.

Table 1.

Localization of HKα₁ and HKα₂ expression in kidney

	Segment	Cell Type	Species	Reference
HKα1
mRNA	Whole kidney		Rat	23
	CCD		Rat	1
	OMCD	IC	Rat	1
	IMCD		Rat	1
	OM		Mouse	55
	IM		Mouse	55
Protein	Whole kidney		Rat	38
	CCD	IC	Rat, rabbit	6, 95
	OMCD	IC	Rat, rabbit	6, 95
	iIMCD, IMCD		Rat	6, 98
HKα2
mRNA	Whole kidney		Rat, mouse	20, 23, 99
	TAL		Rat	2, 47
	DCT		Rat	2
	CNT		Rat	2
	CD		Rat	2, 47
	CD	PC	Mouse	100
Protein	Whole kidney		Rat	38, 39
	TAL		Rabbit	83
	CNT	IC, PC	Rabbit	24, 83
	CCD	IC, PC	Rabbit	83
	OMCD		Rabbit	83
	IMCD		Rat	98

CCD, cortical collecting duct; OMCD, outer medullary collecting duct; IMCD, inner medullary collecting duct; OM, outer medulla; IM, inner medulla; iIMCD, initial IMCD; TAL, thick ascending limb; DCT, distal convoluted tubule; CNT, connecting tubule; CD, collecting duct; IC, intercalated cell; PC, principal cell.

Role of H⁺-K⁺-ATPases in K⁺ reabsorption and hypokalemia.

The majority of reports indicate that HKα₂ H⁺-K⁺-ATPase mRNA and protein expression are increased with dietary K⁺ depletion as shown in Table 2. This effect is well established in studies from rat and has been confirmed at the whole kidney level in mouse. Only one study (3) reported increased HKα₁ mRNA expression with dietary K⁺ depletion. HKα₂ mRNA and protein have been shown to be increased particularly in the medulla of rats fed a K⁺-depleted diet.

Table 2.

Effect of K⁺ restriction or depletion on HKα₁ and HKα₂ expression in kidney

	Effect of K+ Restriction or Depletion	Segment	Species	Reference
HKα1
mRNA	Present	Whole kidney	Rat	40
	Increased	Ctx		3
	Decreased	IMCD		53
Protein	Unknown
HKα2
mRNA	Present	Whole kidney	Rat	40
	Increased	Whole kidney		23
		Medulla		2
		CCD, MCD		47
		Ctx, OM, IM		37
		OMCD		56
		IMCD		53
		Whole kidney	Mouse	61
Protein	Increased	Whole kidney	Rat	38
		OM		65
		Medulla		13

MCD, medullary collecting duct; Ctx, cortex.

In each of the dietary K⁺ depletion studies, a Sch-28080-sensitive component, indicative of HKα₁ H⁺-K⁺-ATPase, is present (see Tables 3 and 4). Additionally, luminal ouabain sensitivity, indicative of HKα₂ H⁺-K⁺-ATPase, was reported in the rat OMCD and inner medullary collecting duct (IMCD) but was not additive with the effect of Sch-28080 (53, 56). In the terminal IMCD (tIMCD) of K⁺-restricted rats, ouabain inhibited steady-state acid secretion both in the presence and absence of Sch-28080 (85). Overall, the data suggest that both HKα₁ and HKα₂ H⁺-K⁺-ATPases participate in K⁺ reabsorption during K⁺ restriction, and the relative roles of both these H⁺-K⁺- ATPases in hypokalemia are discussed further below.

Table 3.

K⁺-ATPase activities in kidney

	Type I K+-ATPase	Type II K+-ATPase	Type III K+-ATPase
IC50 Sch-28080	0.25 μM	1.7 μM	0.85 μM
IC50 ouabain	Insensitive	6.4 μM	20 μM
Cation activation	K+	K+	K+, Na+
Effect of K+ depletion	Decreases	Decreases	Increases
Localization	CD	PT, TAL	CD
HKα1 knockout	Absent	Unknown	Present
HKα2 knockout	Present	Unknown	Absent

Data derived from Ref. 22. PT, proximal tubule.

Table 4.

Effect of pathophysiological conditions on H⁺-K⁺-ATPase Activity

Pathophysiological Condition	Effect on H+-K+-ATPase Activity	Nephron Segment	Cell Type	Species	Reference
Hypokalemia	Increased/Induced	DT		Rat	88
		tIMCD		Rat	85
		IMCD		Rat	53
		OMCD		Rat	56
		OMCD	IC/OMCDi	Rabbit	44
	Present	CCD		Rabbit	102, 103
		OMCD			93
		OMCDis			94
Acute NH4Cl loading	Present	CCD		Rabbit	71
		CCD	B-type IC		91
		CCD	A-type IC		49
		CCD	IC		72
		OMCDis	PC and IC		90
Metabolic acidosis	Increased	CCD	IC	Rabbit	72, 73
	Present	tIMCD		Rat	86
Metabolic alkalosis	Induced	CCD		Rat	30
		DT			92

H⁺-K⁺-ATPase activity is often defined as K⁺-dependent H⁺ transport and pharmacologically defined as omeprazole-, Sch-28080-, and/or ouabain-sensitive H⁺ or K⁺ flux. The reader should be aware that inhibitor sensitivities do not strictly define the activity as mediated by HKα₁ or HKα₂. DT, distal tubule; tIMCD, terminal IMCD; OMCDi or OMCDis, OMCD of the inner stripe.

K⁺-stimulated ATPase activity has been reported in the CD of both rabbit and rat (74). As shown in Table 3, the most extensive studies of these separate types of K⁺-ATPase activities comes from the Doucet laboratory (22), examining the distinct activities of the nephron segments and the CD in animals adapted to a normal or a K⁺-depleted diet. K⁺-ATPase activity has been identified in isolated microdissected nephron segments of rat, and all of the activities are sensitive to Sch-28080, which suggests that they arise from HKα₁ H⁺-K⁺-ATPase. Three distinct enzymatic subtypes of K⁺-ATPase activity were described, which exhibit different characteristics (Table 3). In brief, K⁺-dependent ATPase activity in the CD was classified as either type I or type III. Type I was Sch-28080 sensitive and ouabain insensitive and was present in the cortical collecting duct (CCD) and OMCD of rats maintained on a normal-K⁺ diet (9). This activity exhibited a pharmacological profile that was virtually identical to the classic K⁺-ATPase obtained from microsomes from the gastric mucosa. Type III activity, defined as partially Sch-28080- and ouabain sensitive, replaced type I activity in rats fed a low-K⁺ diet (9). A recent study reported the existence of type I and type III activities in mouse and further characterized both activities in HKα₁ or HKα₂ H⁺-K⁺-ATPase-null mice (22). Dherbecourt et al. (22) reported that type I activity was no longer detectable in the HKα₁-null mice, whereas type III activity was preserved. These data suggest that type I activity requires the expression of the HKα₁ subunit. In contrast, type III activity was no longer detectable in the HKα₂-null mice, whereas type I activity was preserved, which suggests that type III activity requires the expression of the HKα₂ subunit. It should be noted that the pharmacological maneuvers used in these studies could result in a nonspecific pharmacological effect of Sch-28080 as has been shown previously (12). Codina et al. (12) demonstrated a decline in intracellular ATP after Sch-28080 incubation in mIMCD-3 cells. Thus the common Sch-28080 sensitivity reported for each of the K⁺-ATPase activities described in Table 3 may require further clarification.

Another pharmacologically distinct K⁺-ATPase activity (type II) was discovered in the early proximal tubule (PT) and TAL of rat (97). This unique activity was Sch-28080- and ouabain sensitive and was markedly reduced during K⁺ depletion (9). Type II activity was similar to type III in pharmacology, yet different because 1) it was almost completely absent during K⁺ depletion and 2) it was not stimulated by Na⁺. Beltowski and Wójcicka (8) reported the existence of type II activity in the cortex and medulla by an independent method; however, it was not possible to conclude in which nephron segment(s) the activity existed. The presence of type II activity has not been described in HKα₁- or HKα₂-null mice.

Role of H⁺-K⁺-ATPases in H⁺ secretion and acid-base disturbances.

The CD exhibits a significant Sch-28080-sensitive or K⁺-dependent acid secretion mechanism in response to acute acid loading (10, 49, 57, 71, 90, 91, 96). Apical localization of H⁺-K⁺-ATPase in immunohistochemical studies supports the concept that the enzyme is important in both K⁺ reabsorption and H⁺ secretion (24, 65, 83). Data from K⁺ tracer and HCO₃⁻ flux experiments clearly demonstrate that H⁺-K⁺-ATPase plays an important role in transportation of both cations in all segments of the CD (4, 85, 94, 102, 103). Additional measurements of H⁺-K⁺-ATPase-mediated HCO₃⁻ flux in isolated CDs revealed that the enzyme's activity could be controlled by the application of barium, a K⁺ channel blocker (101, 105). Barium was shown to block HCO₃⁻ reabsorption in the CCD when applied either to the luminal perfusate, under K⁺-replete conditions, or to the peritubular fluid, under K⁺-restricted conditions. K_Rb reabsorption was also inhibited by peritubular barium under K⁺-restricted conditions. From these data, a model was proposed coordinating the H⁺-K⁺-ATPase with apical and basolateral channels, suggesting coordinated H⁺ secretion with apical membrane recycling of K⁺ during K⁺ repletion and K⁺ reabsorption via basolateral K⁺ channels during K⁺ depletion (Fig. 1).

Fig. 1.

Physiological model of the renal H⁺-K⁺-ATPase. The renal H⁺-K⁺-ATPase secretes H⁺ and recycles K⁺ via apical K⁺ channels during K⁺ repletion and reabsorbs K⁺ via basolateral K⁺ channels during K⁺ depletion.

Indeed, the mRNA expression of H⁺-K⁺-ATPases has been studied in states of metabolic acidosis as well. A recent study of CD genes regulated by chronic metabolic acidosis in mice showed that, after 3 days of 0.7 M NH₄Cl loading, HKα₁ and HKα₂ expression, determined by real-time PCR, in isolated OMCD increased ∼15 fold and ∼2-fold, respectively, compared with control levels (11). After 14 days of acid loading, OMCD expression of HKα₁ was not different from control, whereas HKα₂ gene expression was approximately threefold greater than control levels. Analysis of intracellular pH (pH_i) recovery in split-open CCD from rabbits subjected to chronic metabolic acidosis (0.075 M NH₄Cl for 10–14 days) demonstrated a threefold stimulation of Sch-28080-inhibitable and K⁺-dependent H⁺ secretion, suggesting that H⁺-K⁺-ATPases may play a role in the kidney's response to chronic metabolic acidosis (72, 73).

Ammonia³ increases HCO₃⁻ reabsorption in the rat CCD and IMCD (35, 84). The effects of ammonia on acid/base transport was investigated in the rabbit CCD (26–28). Ammonia stimulates a luminal Sch-28080- and ouabain-sensitive net HCO₃⁻ flux, suggesting that both isoforms of the H⁺-K⁺-ATPase are involved in mediating this response. Ammonia also stimulates the CCD H⁺-K⁺-ATPase through an intracellular calcium-dependent, microtubule-dependent and a Golgi vesicle transport-dependent process indicative of membrane vesicle insertion into the apical plasma membrane (27). These observations are consistent with studies on the stimulation of K⁺ reabsorption by 10% CO₂ to mimic respiratory acidosis in the CCD of K⁺-deplete rabbits that was blocked by calmodulin inhibition or attenuating changes in intracellular Ca²⁺ concentration ([Ca²⁺]_i) (102). Combined, these findings suggest that both ammonia and respiratory acidosis stimulate a calcium-dependent mechanism of H⁺ extrusion via an H⁺-K⁺-ATPase.

Additionally, HKα₂ mRNA expression in the rabbit CCD has been shown to be significantly increased during acute metabolic alkalosis induced by dietary HCO₃⁻ loading (25). Sch-28080 also significantly inhibited the increased HCO₃⁻ reabsorption observed in distal tubules (DT) and CCD from alkalotic rats, indicating that the HKα₁ H⁺-K⁺-ATPase may be stimulated as well (30, 92).

Table 4 provides a summary of the effects of both hypokalemia and acid-base disturbances on H⁺-K⁺-ATPase activity in the kidney.

Na⁺ and NH₄⁺ transport and H⁺-K⁺-ATPase activity.

Substantial clinical evidence supports the assertion that diets with little K⁺ content can either worsen hypertension in humans with preexisting hypertension (34, 42) or predispose normotensive individuals to NaCl-sensitive hypertension (43). Moreover, mild K⁺ depletion appears to worsen hypertension in part by promoting renal Na⁺ retention (41). Given the effect of chronic hypokalemia to reduce plasma aldosterone concentrations, such observations suggest that an as yet not fully understood renal adaptive mechanism exists that exerts greater control of Na⁺ excretion than plasma aldosterone. Conversely, diets with a large K⁺ content have been associated with improvement in systemic hypertension (50, 64). At least for HKα₁, there appears to be a significant interaction of Na⁺ and K⁺ for reabsorption by an apparent common mechanism. This observation is supported by work performed in native renal (104) and gastric (77) tissues. Interestingly, additional evidence from heterologous expression studies by Swarts et al. (78) suggests that HKα₂ may have a greater affinity for NH₄⁺ than K⁺. Specifically regarding HKα₁, studies under different physiological conditions supported a significant interaction of Na⁺ and K⁺ with a Sch-28080-sensitive absorptive mechanism and concluded that K⁺ and Na⁺ compete for the same reabsorptive pathway (103, 104). Competitive binding between Na⁺ and K⁺ on the H⁺-K⁺-ATPase is also supported by the observation that two structurally different H⁺-K⁺-ATPase inhibitors, Sch-28080 and A80915A, can inhibit Na⁺ reabsorption in the rabbit CCD under conditions of dietary Na⁺ depletion (106). This observation suggests that the HKα₁ isoform can mediate Na⁺ absorption since Sch-28080 is the classic gastric-type inhibitor. Dherbecourt et al. (22) reported K⁺-ATPase activity (type III) in HKα₁-null mice that is sensitive to both Sch-28080 and ouabain. Direct evidence distinguished type III activity from type I activity by several characteristics, but of particular note, type III activity can be stimulated by Na⁺, which suggests direct binding of Na⁺ on this enzyme (9). If the HKα₁ H⁺-K⁺-ATPase does transport Na⁺ under physiological conditions, then the disruption of the apical Na⁺/H⁺ exchanger 3 (NHE-3) would lead to enhanced distal luminal Na⁺ delivery and plausibly an upregulation of HKα₁ as observed by Nakamura et al. (54). Swarts et al. (77) concluded that Na⁺ as well as K⁺ could support the dephosphorylation of the HKα₁-containing H⁺-K⁺-ATPase by comparing the steady-state amount of the phosphorylated enzyme in tight and leaky gastric membrane vesicles.

Additionally, physiological evidence shows that one or more H⁺-K⁺-ATPase isoforms are stimulated by dietary Na⁺ depletion, consistent with a role for H⁺-K⁺-ATPase in Na⁺ reabsorption. The intercalated cells (IC) from the split-open CCD of rats on a low-Na⁺ diet had twofold higher K⁺-dependent pH_i recovery from an acid pulse compared with control (70). The control K⁺-dependent pH_i recovery was blocked by Sch-28080, but abolishment of pH_i recovery in the IC from the low-Na⁺ animals required the application of both Sch-28080 and ouabain. These results suggested that NaCl depletion induces an additional H⁺-K⁺-ATPase isoform. However, other studies did not observe that dietary Na⁺ depletion upregulated kidney HKα₂ mRNA (65, 75) or protein (75). These findings suggest at least two explanations, either that yet a third isoform (possibly an α/β pair that has not been characterized) is induced under these conditions or that the increase in pump activity occurred as a posttranslational event. Supporting the former, Petrovic et al. (61) proposed that the kidneys of HKα₁-null mice possess a novel H⁺-K⁺-ATPase that is neither HKα₁ or HKα₂.

The H⁺-K⁺-ATPase has been implicated as a compensatory mechanism to other renal perturbations. HKα₂ H⁺-K⁺-ATPase has been demonstrated to be upregulated by ischemia-reperfusion injury or acetazolamide treatment. Rats subjected to ischemia-reperfusion injury substantially upregulate cortical HKα₂ H⁺-K⁺-ATPase, and not HKα₁, perhaps to compensate for the (∼75%) downregulation of the cortical and medullary NHE-3 (89). NHE-3 is an apical Na⁺/H⁺ exchanger that facilitates over half of the acid secretion that occurs in the PT. Rats treated with acetazolamide, which causes renal HCO₃⁻ wasting, also exhibited upregulated cortical HKα₂ (89). Of particular interest, NHE-3-null mice have profoundly reduced HCO₃⁻ reabsorption in the PT, with only slightly decreased serum HCO₃⁻ levels compared with wild type (WT) (66). In an attempt to resolve the compensatory acid secretion mechanism limiting the development of substantial metabolic acidosis in NHE-3-null mice, Nakamura et al. (54) studied net HCO₃⁻ absorption in the OMCD and concluded that the HKα₁ H⁺-K⁺-ATPase is involved in the adaptive changes in these mice. Specifically, there was a greater Sch-28080-sensitive acid secretion and upregulation of HKα₁ mRNA in the OMCD from NHE-3-null compared with control mice. Nakamura et al. reported that renal HKα₂ H⁺-K⁺-ATPase does not play a compensatory role in the NHE-3-null mice since the renal HKα₂ mRNA was undetectable in kidney and there was no effect of luminal 1.0 mM ouabain. However, the axial heterogeneity of each CD segment and the small mass of CD in the whole kidney or cortical sections limits the interpretation of these data.

As noted above, dietary K⁺ depletion has been shown to produce a volume- and Na⁺-dependent form of hypertension despite reduction of plasma aldosterone concentration. Such evidence suggests that certain H⁺-K⁺-ATPase isoforms could participate in NaCl uptake, and experimental evidence supports the functional coupling of an H⁺-K⁺ ATPase and an apical Cl⁻/HCO₃⁻ exchanger in the CCD, possibly in the B-type IC (106). Thus the precise cation specificity of the luminal and cytosolic binding sites of H⁺-K⁺-ATPase isoforms may have profound importance if the binding affinities differ between species. These observations could help to explain not only the well-known interaction of K⁺ intake on Na⁺ intake but also the difference in the adaptive responses to K⁺ depletion of humans versus commonly studied laboratory animals.

Knockout studies.

HKα₁-null mice have been generated (76) and have survived, although they exhibited achlorhydria, hypergastrinemia, and metaplasia of the gastric epithelium. The secretory membranes and gastric mucosa of the HKα₁-null mice were also abnormal. Although no overt renal phenotype was observed in the HKα₁-null mice, no detailed study has been published on the renal function of these mice. Subsequent studies indicate that such investigations are essential to fully understand the role of these two H⁺ pumps in solute transport (46). Petrovic et al. (61) reported the possible existence of a novel and compensatory H⁺-K⁺-ATPase activity. Measurement of K⁺-dependent H⁺ secretion in the CCD of HKα₁-null animals revealed the presence of an apparently unique Sch-28080- and ouabain-insensitive transporter. Review of the human genome for a third P-type ATPase α-subunit gene did not reveal a likely candidate gene (http://biobase.dk/∼axe/patbase.html). However, such an activity could be produced via alternative splicing of existing genes or posttranslational modification of an ion pump subunit. Since the H⁺-K⁺- ATPases require an additional β-subunit to function, the activity detected by Petrovic et al. may be due to a previously unrecognized HKα₂/β pair that confers these unique pharmacological properties. P-type ATPase α-subunits can exhibit modified profiles depending on the identity of the coassembled β-subunit (29). Specifically, the work of Geering and coworkers (19) has clearly demonstrated that different β-subunits confer different protease sensitivities to HKα₂ and that these structural changes can be observed in response to K⁺ and ouabain.

The ability of HKα₂-null mice to retain fecal and urinary K⁺ has been studied under control and K⁺-depleted conditions (48). Under control conditions, no dramatic differences were observed between WT and HKα₂-null mice. However, when fed a K⁺-free diet for 18 days, the null mice lost four times more fecal K⁺ than the WT animals. However, these knockouts did not display a significant renal phenotype. The HKα₂-null mice also lost twice as much body weight as WT and had smaller plasma and muscle K⁺ concentration ([K⁺]) than WT mice. These studies clearly demonstrated that the HKα₂ H⁺-K⁺-ATPase has a pivotal role in maintaining K⁺ balance, at least in the colon.

A previous report suggested that the HKα₂ H⁺-K⁺ ATPase could function as a Na⁺-K⁺-ATPase in the colon (62). A subsequent study involving the HKα₂-null animals examined the effect of dietary Na⁺ depletion in such mice versus WT control animals from the same strain (75). The HKα₂-null mice had a greater rate of fecal K⁺ excretion than the WT animals when fed a Na⁺-restricted diet. In addition, regardless of diet, the null mice exhibited smaller values of epithelial Na⁺ channel (ENaC)-mediated short-circuit current in the distal colon. The authors postulated that HKα₂ plays an important role not only in K⁺ conservation but also in Na⁺ absorption by the colon, which suggests that HKα₂ H⁺-K⁺-ATPase may also mediate Na⁺ reabsorption by the kidney.

A recent report on the HKα₂-null mice provided in vivo evidence for the importance of the HKα₂ H⁺-K⁺-ATPase as a H⁺ pump (60). Previously, Pestov et al. (58) had shown that the HKα₂ H⁺-K⁺-ATPase was expressed in the apical membrane of the anterior prostate epithelium. In a convincing study, this group evaluated the acidification of anterior prostate fluid in the HKα₂-null mice (60). They found that the pH of prostate fluid in the HKα₂-null animals was 6.96 compared with 6.38 in WT mice. This loss of acidification strongly suggested that the HKα₂ H⁺-K⁺-ATPase functions as a H⁺ pump in vivo and is required for acidification of luminal prostate fluids. Additionally, mice with disruption for Atp12a (encoding for HKα₂) have loss of apical immunohistochemical localization for NaKβ₁, supporting the role of this β-subunit as the authentic in vivo subunit in the mouse anterior prostate for the apically localized HKα₂-NaKβ₁ H⁺-K⁺-ATPase (59) similar to the work of Codina et al. (14) in rat medulla. Indeed, we recently investigated (46) rates of pH_i recovery and acid extrusion in response to acute acid loading in two types of IC from the CCD of WT mice and in these null mice. The results confirm a significant and separate contribution of both HKα₁ and HKα₂ to acid extrusion in both A-type and B-type IC in mice fed a normal diet. Moreover, it is now clear that both HKα₁ and HKα₂ independently contribute to H⁺ secretion in mice fed normal mouse chow and that combined disruption of both catalytic subunits (HKα_1,2) reveals independent effects of both gene products at least in A-type IC (46). Such observations suggest that these two pumps independently secrete H⁺ in exchange for cellular K⁺/cation reabsorption, which supports the role of both isoforms in renal solute transport.

In summary, HKα₁- and HKα₂-null mice have been generated, but the renal phenotype of these mouse models has not been fully characterized. Examining the HKα₁- and HKα_1,2-null mice during K⁺ and Na⁺ depletion, as well as acid and alkali loading, will provide valuable insight into the physiological function of renal H⁺-K⁺-ATPases in pathophysiological states.

Regulation of Renal H⁺-K⁺-ATPases

Signaling molecules.

Regulation of the Na⁺-K⁺-ATPase by accessory subunits and signaling molecules is considerably better understood compared with that of the H⁺-K⁺-ATPases. In an effort to characterize more fully these mechanisms of regulation for the HKα₁ H⁺-K⁺-ATPase, Cornelius and Mahmmoud (18) undertook a profiling study to explore the acute regulation mechanisms of the gastric H⁺-K⁺-ATPase. It is well established that the activity of the gastric enzyme depends on trafficking of the enzyme from intracellular vesicles to the apical membrane of gastric parietal cells. The signals regulating this trafficking event are not fully understood but most likely involve second messenger molecules such as cAMP, Ca²⁺, and diacylglycerol. The protein kinases involved in these signaling pathways include protein kinases A and C (PKA, PKC). Consensus sequence sites for these kinases are present in the HKα₁ protein. Using a combination of proteolytic fingerprinting and ³²P-kinase assays, these investigators demonstrated the presence of a PKC phosphorylation site at the NH₂ terminus of HKα₁ and a PKA phosphorylation site at the COOH terminus. However, the latter site was only phosphorylated in the presence of detergent, indicating that it may or may not be physiologically relevant. The PKC phosphorylation site is likely Ser 27, a highly conserved PKC site. The catalytic activity of the gastric H⁺-K⁺-ATPase was significantly increased by NH₂-terminal PKC (Ca²⁺-dependent isoforms) phosphorylation. These investigators concluded that, since the activation of catalytic activity occurred at saturating ATP concentrations, the most likely reaction step affected by PKC phosphorylation was the E1∼P to E2-P transition. Phosphorylation of HKα₁ at the NH₂-terminal PKC site may therefore serve as a “conformational switching signal,” with implications for regulation of H⁺ secretion.

Another group of investigators used rat CCD to study the mechanism behind the PKA-dependent activation of the HKα₁ H⁺-K⁺-ATPase by isoproterenol (45). Using a [³²P]ATPase assay in combination with various inhibitors, Laroche-Joubert et al. (45) investigated a β-adrenergic receptor pathway specifically in β-IC of the CCD. They found that isoproterenol activated a G protein-coupled receptor that in turn led to the release of cAMP and activation of PKA. This was followed by stimulation of a pathway involving a tyrosine kinase-Ras-Raf-1-MEK-ERK cascade, ultimately resulting in activation of the HKα₁ H⁺-K⁺-ATPase as characterized by its sensitivity to Sch-28080.

HKα₂: PKA.

Consensus sites for PKA are present in the amino acid sequence of HKα₂. In an effort to determine the effects of PKA signaling on the ouabain-sensitive H⁺-K⁺-ATPase in the kidney, Beltowski et al. (7) treated male Wistar rats with a nonhydrolyzable analog of cAMP. Analysis of ouabain-sensitive H⁺-K⁺-ATPase activity in the microsomal fractions from either renal cortex or medulla revealed 30% or greater increases in activity. This effect was sensitive to brefeldin A, a compound that prevents trans-Golgi processing of plasma membrane proteins, suggesting that the PKA-mediated activation of the HKα₂ H⁺-K⁺-ATPase was dependent on insertion of a protein into the plasma membrane. Whether this protein is the enzyme itself or another regulatory protein required for H⁺-K⁺-ATPase activity remains to be determined.

A recent report from the Codina and DuBose laboratories may shed some light on the above question. Codina et al. (17) mutated the COOH-terminal PKA consensus site, Ser 955, to either an alanine or an aspartic acid. A thorough analysis of these mutants was performed by assaying HKα₂ protein expression and functionality (by ⁸⁶Rb uptake assay) at the plasma membrane, total HKα₂ levels by immunoblot, and ³⁵S-met labeling to determine the rate of HKα₂ protein synthesis and degradation. These studies were conducted in HEK293 cells and revealed that phosphorylation of HKα₂ at Ser 955 promoted maturation of the HKα₂ protein. Inhibition of PKA activity by expression of the PKA inhibitor resulted in decreased expression of HKα₂ and a subsequent decrease in ⁸⁶Rb uptake. These results indicate that phosphorylation of HKα₂ by PKA may contribute to the enzyme being properly trafficked to the plasma membrane.

Structure of H⁺-K⁺-ATPases

To date, no high-resolution structure of the HKα₁ or the HKα₂ H⁺-K⁺-ATPase has been published. However, several structures of a related P-type ATPase, the single-subunit Ca²⁺-ATPase, have been published (79–82). The Ca²⁺-ATPase has been crystallized with bound analogs, representing various stages of the catalytic cycle, thereby allowing the pumping mechanism of a P-type ATPase to be visualized (see supplemental video in Ref. 80). Resolution of the Ca²⁺-ATPase structures has led to several modeling exercises in order to apply this knowledge to related enzymes, including the HKα₁ and HKα₂ H⁺-K⁺-ATPases.

Our laboratory (31) published the first such study, using the first high-resolution structure of the Ca²⁺-ATPase (Protein Data Bank ID: 1eul) as template to build a homology model of HKα₂. The overall shape of the HKα₂ H⁺-K⁺-ATPase model appeared to be conserved compared with the Ca²⁺-ATPase. Important features shared between the HKα₂ model and the Ca²⁺-ATPase included the A (actuator), P (phosphorylation), N (nucleotide binding), and M (transmembrane) domains of the subunit. Extensive modeling of the HKα₁ H⁺-K⁺-ATPase has also been performed based on several of the Ca²⁺-ATPase structures. HKα₁ H⁺-K⁺-ATPase models in both the E1 and E2 conformations have been used together with biochemical experiments to evaluate the docking of K⁺-competitive inhibitors in the pump (5) as well as to investigate a salt bridge required for K⁺ binding (36). The most comprehensive modeling studies on HKα₁ came from the laboratory of Sachs and were based on the E1-Ca²⁺, E2-thapsigargin (51), and E2-magnesium fluoride (52) structures. In these two elegant studies, Munson et al. presented homology models of the HKα₁ H⁺-K⁺-ATPase and investigated them based on a large body of biochemical data concerning structure/function relationships in this ion pump.

The Sachs model of HKα₁ bound to ADP-Mg²⁺ is pictured in Fig. 2. The model is colored from dark blue at the NH₂ terminus to red at the COOH terminus. The A, N, P, and transmembrane domains are labeled. In their two reports, Munson et al. used the HKα₁ models to consider the mechanism for counterion transport, the binding of K⁺ competitive inhibitors, and the pathways for K⁺ entry and exit. For further review of these modeling studies, the reader is referred to Sachs et al. (63). This group used a combination of energy minimization and molecular dynamics to generate their models. Alignment of the HKα₁ and Ca²⁺-ATPase sequences revealed large gaps and insertions in the cytoplasmic N domain. To address this issue, Munson et al. replaced the Ca²⁺-ATPase coordinates in this region with those of the Na⁺-K⁺-ATPase α₂-subunit crystal structure (pdb: 1q3i). In their 2005 report (51), Munson et al. modeled hydronium ions (H₃O⁺) into the E1 HKα₁ model and compared it to their hypothetical E2-K⁺ form of HKα₁ in order to consider the mechanism for counterion transport. They observed that H₃O⁺ interacted with Asp 824, Glu 820, and Glu 795 and was displaced by insertion of Lys 791 into this region. Binding of K⁺ caused movement of Lys 791. This resulted in occupation of the former H₃O⁺ site by K⁺, dephosphorylation of the pump, and release of K⁺ to the cytosol. A model of this mechanism was proposed by Shin and Sachs (69) for the HKα₁ H⁺-K⁺-ATPase as a drug target. Binding sites for the covalent class of proton pump inhibitors (i.e., pantoprazole) and for the K⁺-competitive class of inhibitors (i.e., Sch-28080) were considered as well. Subsequently (63), Munson and coworkers improved on the previous modeling exercise by further refining the binding properties of both classes of inhibitors. Indeed, they were able to visualize a hydrated vestibule formed by transmembrane helices 4, 5, 6, and 8 that allowed for entry of K⁺-competitive inhibitors. Docking of the inhibitor prevented access of K⁺ to the ion binding site. With the use of homology models of the E1-K⁺ and E2-K⁺ catalytic states, the positions of K⁺ entry and exit pathways were postulated. K⁺ entry was predicted to occur near the extracellular loop between transmembrane helices 5 and 6. K⁺ exit was suggested to occur between the intracellular surfaces of transmembrane helices 1 and 2. The stability of their HKα₁ model was demonstrated during a 10-ns molecular dynamics simulation that included water and lipids. Further work by this group and refinement of the proposed mechanism of ion transport includes the entry pathway of K⁺ and the role of the arrival of K⁺ to destabilize the interaction of Lys 794 and Glu 820 and 795. These events are proposed to initiate the conformation changes that result in the hydrolysis of the phosphate at the active site, which results in the conformational change to the E₁ form of the enzyme (67). Of note, K⁺ access from the lumen depends on K⁺ conductance through the KCNQ1/KCNE2 complex, whereas CLIC6 is proposed to conduct Cl⁻ (67). These theoretical models are consistent with site-directed mutagenesis studies in which the WT and mutant (E820A) HKα₁ subunit were expressed with the HKβ subunit in Sf9 cells. The WT and mutant enzymes both showed a biphasic activation by K⁺ on ATP hydrolysis, but with the mutant having a significantly reduced K⁺ affinity (33), and showed that Glu 795 is essential for HKα₁ H⁺-K⁺-ATPase activity (32).

Fig. 2.

Molecular model of HKα₁. HKα₁ is shown bound to ADP-Mg²⁺. The model is colored from dark blue at the NH₂ terminus to red at the COOH terminus. The A or actuator domain is shown in light blue, the N or nucleotide binding domain in green, and the P or phosphorylation domain in white. The 10 transmembrane helices in the membrane domain are labeled M1–M10. For more detail, see Ref. 52.

The application of these modeling exercises to the kidney could potentially answer important questions about the ion affinities and transport properties of HKα₁ and HKα₂. For example, given the extensive data that suggest a role for either or both of the HKα₁ and HKα₂ H⁺-K⁺-ATPases in Na⁺ transport, a molecular dynamics simulation should be performed to test whether or not Na⁺ can be directly transported on the enzyme. The transport of other ions and the accessibility of proposed regulatory sites could be evaluated as well. As discussed below, future studies on the HKα₁ and HKα₂ H⁺-K⁺-ATPases will likely be focused on delineating their role in the transport of ions other than H⁺ and K⁺.

Conclusions and the Future

Despite the well-established catalytic cycle of the H⁺-K⁺-ATPase family of ion-motive pumps, much further investigation into the function and regulation of H⁺-K⁺-ATPases is needed to ascertain the specific physiological roles of these ion pumps. The available evidence, albeit imperfect, supports the assertion that progress on the functional importance of these pumps has been hampered by the label as an “H⁺-K⁺- ATPase.” Our knowledge of the role of these pumps will remain incomplete until reliable data can be obtained about the cation specificity of both the cytosolic and the extracellular cation binding sites and their rank order affinity for the physiological (and pathophysiological) abundant cations such as H₃O⁺, Na⁺, K⁺, and NH₄⁺. Various approaches will likely prove essential, and mutual verification by physiological and biochemical approaches should provide important new understanding of these pumps during normal and physiologically stressed conditions. The characterization of mice lacking functional expression of either or both H⁺-K⁺-ATPase α-subunits, HKα₁ and HKα₂, in pathological dietary conditions will be imperative to assess the specific role of the enzyme in both electrolyte and acid-base homeostasis. A recent report demonstrates the impairment of H⁺ secretion in either or both H⁺-K⁺-ATPase α-subunit-null mice on a normal diet. Lynch et al. (46) found that both HKα₁ and HKα₂ play a crucial role in acid secretion in the CCD. The regulation of H⁺-K⁺-ATPases by ammonia and the possible transport of NH₄⁺ in place of K⁺ bear implications for acid-base handling. The response of the HKα₁-, HKα₂-, and HKα_1,2-null mice to acid and base loading will elucidate the enzyme's role in acid-base balance.

The importance of HKα₂ in K⁺ homeostasis was previously studied. HKα₂-null mice displayed a reduced conservation of fecal K⁺ during K⁺ depletion, leading to lower plasma [K⁺] than K⁺-depleted WT mice. Both WT and HKα₂-null mice conserved urinary K⁺, and only a trend toward increased urinary K⁺ excretion in the null mice was observed. This retention of urinary K⁺ may be due to the presence and activation of the HKα₁ subunit or a considerable reduction of K⁺ secretion. Investigation of the HKα₁-null and HKα_1,2-null mice on a K⁺-depleted diet will provide insight into the functional importance of H⁺-K⁺-ATPases in K⁺-depleted conditions.

Evidence implicates H⁺-K⁺ ATPases in Na⁺ transport. HKα₂-null mice excrete 2.8 times more fecal K⁺ and slightly more fecal Na⁺ on a Na⁺-depleted diet than WT mice (75). The lack of an overt renal phenotype in HKα₂-null mice during Na⁺ depletion may be a result of compensation by the HKα₁ subunit or differences in specific tissues. The upregulation of H⁺-K⁺-ATPases in the kidneys of NHE-3-null mice also suggests that this enzyme may have a physiological role in both H⁺ and Na⁺ transport. To fully investigate the role of H⁺-K⁺-ATPases in Na⁺ transport, both the HKα₁-null and HKα_1,2-null mice need to be examined similarly under Na⁺-depleted conditions. HKα_1,2-null mice provide an invaluable tool that can be used to assess the physiological significance of H⁺-K⁺-ATPases, especially within pathological dietary conditions.

GRANTS

These studies were supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants R01-DK-049750 (to C. S. Wingo) and T32-DK-07518 (M. L. Gumz), American Heart Association Grant 0825467E (M. L. Gumz), and by funds from the Medical Research Service of the North Florida/South Georgia Veterans Health System.

DISCLOSURES

No conflicts of interest are declared by the author.