Glucocorticoid-Remediable Aldosteronism (Familial Hyperaldosteronism Type I)

Glucocorticoid-Remediable Aldosteronism (GRA) is an autosomal dominant disease featuring hypertension with inappropriate aldosterone secretion despite suppression of the renin–angiotensin system. Patients typically present with hypertension in the first two decades of life, and are found to have elevated aldosterone secretion despite suppressed plasma renin activity, indicating autonomous production of aldosterone. Hypertension is often severe, and affected subjects are at a markedly increased risk of cerebral hemorrhage at young ages. Hypokalemia and metabolic alkalosis are variable associated findings. Consistent with autosomal dominant transmission, the family history is usually positive, with one parent and about half of siblings and offspring having early diagnosis of hypertension, and there is a frequent history of early cerebral hemorrhage.

Unlike normal individuals, aldosterone secretion in GRA shows sustained increase with administration of the cortisol secretagogue ACTH. Moreover, affected subjects produce so-called hybrid steroids, 18-hydroxy- and 18-oxocortisol, which are present in negligible amounts in normal subjects. These hybrid steroids have hydroxylation at C-17, characteristic of metabolism by 17-alpha hydroxylase (CYP17 gene) in the adrenal fasciculata, and oxidation at C-18, characteristic of metabolism by aldosterone synthase, which is normally confined to the adrenal glomerulosa ( Figure 36.3 ). Suppression of cortisol secretion from the adrenal gland by administration of exogenous glucocorticoids also causes rapid and sustained suppression of aldosterone secretion, a finding that does not occur in normal individuals. These features suggest that aldosterone in GRA is produced in the adrenal fasciculata under the control of ACTH, rather than in adrenal glomerulosa under the control of the normal secretagogue angiotensin II (Ang II).

Adrenal steroid synthesis.

Hormone synthesis in the adrenal cortex occurs in distinct compartments: aldosterone is synthesized in the outermost zona glomerulosa; cortisol in adrenal fasciculata; and sexual steroids in the innermost zona reticularis. Enzymes required for biosynthesis are noted (3β-HSD: 3β-hydroxysteroid dehydrogenase).

Analysis of linkage in a large GRA kindred demonstrated complete linkage of early hypertension to chromosome 8q, which contains the gene aldosterone synthase (CYP11B2) and the closely related gene steroid 11-β hydroxylase (CYP11B1). These two genes recently evolved from a common ancestor, and are highly similar at the DNA sequence level. Patients with GRA prove uniformly to have mutations that result from unequal crossing-over recombination between these two homologous genes ( Figure 36.4 ). This recombination event produces two mutant chromosomes, one with normal copies of aldosterone synthase and 11-β hydroxylase, and a third chimeric gene between them that fuses 5′ regulatory sequences from 11-β hydroxylase to coding sequences of aldosterone synthase. This is the mutation uniformly found in patients with GRA. The other mutant chromosome has no normal gene, and only a single hybrid gene that fuses 5′ sequences from aldosterone synthase to a coding sequence that produces 11-β hydroxylase enzymatic activity. This mutation is found in some patients with 11-β hydroxylase deficiency. These events occur in pre-meiotic germ cell development, hence only one is transmitted to a zygote. All of these recombination events seen in GRA occur upstream of exon 5, with the result that the hybrid genes in patients with GRA include the two amino acids in exons 5 and 6 that are critical for the resulting enzyme having aldosterone synthase rather than 11-β hydroxylase activity. These hybrid genes bear the 5′ regulatory elements of 11-β hydroxylase, and are consequently expressed in adrenal fasciculata under the control of ACTH, but encode aldosterone synthase enzymatic activity. As a consequence, aldosterone secretion in these patients is constitutive, driven by ACTH and the maintenance of normal cortisol levels. The renin–angiotensin system is suppressed, but this fails to turn off aldosterone production. These mutations thus account for the ectopic production of aldosterone in adrenal fasciculata and its control by ACTH, and explain the resulting hypertension.

Unequal crossing-over recombination between aldosterone synthase and 11-β hydroxylase causes GRA.

Shown is the segment of chromosome 8 that contains the two highly homologous genes. Unequal crossing-over between these two genes produces a chromosome with normal copies of aldosterone synthase and 11-β hydroxylase, and a third chimeric gene between them that fuses 5′ regulatory sequences from 11-β hydroxylase to coding sequences of aldosterone synthase, conferring aldosterone synthase function on the encoded gene product

(reproduced with permission from ref. ).

GRA should be suspected by the finding of hypertension in young individuals with elevated aldosterone level (best measured in 24 hour urine samples), despite suppressed plasma renin activity, particularly if there is a history of early hypertension among first degree relatives. Hypokalemia and metabolic alkalosis are common but by no means invariant, and the absence of these findings should not be used to exclude the diagnosis. Molecular genetic testing by either Southern blotting or PCR provides a sensitive and specific means for establishing the diagnosis. Because of the autosomal dominant transmission, there are frequently many affected members among extended families of index cases, and case finding can be performed by sequential sampling, testing all first degree relatives of affected subjects, then all first degree relatives of positive cases in subsequent rounds. Sampling in these families is likely to prevent morbidity and mortality from uncontrolled hypertension and early cerebral hemorrhage. Because of the availability of genetic screening, determination of urinary levels of hybrid steroids or dexamethasone suppression test is no longer recommended. Of note, family screening has demonstrated that not all individuals carrying the gene fusion are hypertensive, and first degree relatives of patients with confirmed GRA should be genetically screened even in the absence of hypertension.

Treatment options for GRA include use of mineralocorticoid receptor antagonists (spironolactone or eplerenone), amiloride or triamterene (which inhibit the epithelial sodium channel that drives both hypertension and hypokalemic alkalosis in GRA). Exogenous glucocorticoids can be used to shut down aldosterone production from the adrenal fasciculata. Careful attention to dosage in children is essential to maintain normal growth and to avoid glucocorticoid side-effects. Potassium-wasting diuretics such as hydrochlorothiazide and furosemide should be used with caution, because of the risk of severe hypokalemia.

Aldosterone-Producing Adrenal Adenoma and Familial Hyperaldosteronism Due to KCNJ5 Mutations

Aldosterone-producing adenoma, also known as Conn’s syndrome, is a common cause of severe hypertension, found in about 5% of patients in hypertension referral clinics worldwide, and in about half of patients diagnosed with primary aldosteronism. Patients typically present with worsening hypertension, and are found to have elevated serum and 24 hour urinary aldosterone levels in conjunction with suppressed plasma renin activity, consistent with these tumors having renin-independent aldosterone secretion. Hypokalemia and metabolic alkalosis are frequent, but not invariant findings. The finding of an adrenal mass by CT scan with increased aldosterone levels in ipsilateral adrenal vein plasma is considered diagnostic, and removal of these tumors corrects or improves blood pressure in the large majority of patients.

Exome sequencing of four aldosterone-producing adrenal adenomas (APAs) and matched blood DNA enabled identification of somatic mutations in the tumors. The results showed a low somatic mutation rate, with only 2–5 somatic mutations per tumor. Surprisingly, the K ⁺ channel encoded by KCNJ5 was mutated in two of these tumors. Examination of this channel in 22 tumors identified somatic KCNJ5 mutations in 8 tumors, and either of the same two mutations were found in all, substituting arginine for glycine at position 151 or arginine for leucine at position 168. These two positions lie in or abut the highly conserved K ⁺ channel selectivity filter that enables these channels to allow passage of K ⁺ but not other ions through the channel pore. Electrophysiologic studies demonstrated that these mutations cause markedly increased Na ⁺ conductance of the mutant channel, sufficient to depolarize the cell.

These findings explain the pathogenesis of APA in these tumors. The normal adrenal glomerulosa cell is hyperpolarized owing to constitutively open K ⁺ channels. Angiotensin II signaling results in closure of these channels, resulting in depolarization and activation of voltage-gated calcium channels, which raises intracellular calcium. Hyperkalemia produces the same results, potentially via increased frequency of depolarizing membrane potential oscillations. Increased intracellular Ca ²⁺ is the acute signal for increased expression of aldosterone synthase and other rate-limiting steps in aldosterone biosynthesis. Chronic signaling provides the stimulus for increased cell replication. Thus, these single mutations can explain both the cell-autonomous aldosterone secretion and cell proliferation that are the hallmarks of these benign tumors.

Subsequent work has confirmed these findings. The study of 287 tumors by Björklund et al. found the G151R or L168R mutations in 47% of APAs, and a markedly higher prevalence of these mutations among women with APA (63%) than men (22%). Only one additional mutation (E145Q) was found in two cases. Similar findings were observed by Boulkroun et al., who found either of these two mutations in 34% of all APAs with a similar bias for female subjects. These tumors are more prevalent among younger individuals, and tend to be slightly larger compared to non-KCNJ5-mutant tumors.

In addition to these mutations causing APAs, they also account for a rare inherited form of primary aldosteronism. In 2008, Geller et al. reported a father and two daughters with a new familial form of severe early-onset hypertension due to primary aldosteronism. Hypertension in these patients was diagnosed between the ages of 4 and 7 years; it was resistant to aggressive antihypertensive therapy, including spironolactone and amiloride. There was massive adrenocortical hyperplasia. Hybrid steroids 18-oxocortisol and 18-hydroxycortisol were elevated, however, in clear contrast to patients with GRA, there was a significant increase in aldosterone levels upon dexamethasone administration, and affected subjects did not have the gene fusion characteristic of GRA. Due to unrelenting hypertension, all three subjects underwent bilateral adrenalectomy in childhood, demonstrating massive adrenal hyperplasia (with paired adrenal weights up to 81 g, normal is less than 12 g) and diffuse hyperplasia of the zona fasciculata by light microscopy (with transitional zone morphology by electron microscopy). Screening of candidate genes revealed no pathogenic mutations.

Affected members of this family proved to have a T158A mutation in KCNJ5 which also modified channel selectivity, resulting in Na ⁺ conductance. In this case, since the mutation is present in every cell, rather than acquired somatically by a single cell, every cell in the adrenal glomerulosa is receiving the signal for aldosterone production and cell proliferation, accounting for the massive adrenal hyperplasia and severe aldosteronism at young ages. This Mendelian form of disease provides strong evidence that these single mutations are sufficient for cell proliferation and constitutive aldosterone secretion.

Subsequent studies have identified additional families with early severe aldosteronism and mutations in KCNJ5. Most interestingly, these include two different mutations at the same amino acid – G151 – that result in markedly different phenotypes. One of these inherited mutations is G151R, the same mutation found as a somatic mutation in APA. Patients with inherited G151R mutations develop massive adrenocortical hyperplasia, have difficult-to-control hypertension, and virtually invariably come to bilateral adrenalectomy for control of hypertension, similar to the T158A mutation described above. The other mutation is G151E. Patients with this mutation also present with early hypertension and aldosteronism; however, they do not develop adrenal hyperplasia, are typically responsive to antihypertensive therapy, and do not come to adrenalectomy. Most interestingly, the milder human phenotype resulting from G151E is associated with a much more severe electrophysiologic phenotype. G151E results in dramatically greater Na ⁺ conductance than G151R. The consequence of this is markedly increased sodium-dependent lethality. This suggests a model in which cells expressing the G151E mutation differentiate from a stem cell pool, produce aldosterone as they are born, but die rapidly, preventing development of hyperplasia. Their continuous renewal from a stem cell pool provides a long-term source for excessive aldosterone production, and a milder hypertension than in subjects with the G151R mutation.

This phenotype of primary aldosteronism without adrenal hyperplasia due to KCNJ5 mutation is as yet unique to the G151E mutation, while massive adrenal hyperplasia requiring adrenalectomy has also been reported with another KCNJ5 mutation, I157S.

While KCNJ5 mutations unequivocally explain the pathogenesis of a large fraction of APAs and a rare form of familial hypertension, the role of the wild-type channel in human adrenal function is less clear. Its activation by dopamine, an inhibitor of aldosterone secretion, suggests that the normal role of this channel might be to hyperpolarize cells, contributing to inhibition of aldosterone secretion.

Primary aldosteronism due to mutations in aldosterone synthase and KCNJ5 both produce constitutive secretion of aldosterone and hypertension. Patients with KCNJ5 mutations tend to present in the first several years of life with severe hypertension, while those with GRA are more frequently, though not exclusively, diagnosed later in the first or second decade. Consistent with greater disease severity among KCNJ5 families and impaired reproductive fitness, no families with more than four affected members have been reported to date, while a number of large, multigenerational families with more than 20 members with GRA have been studied.

Congenital Adrenal Hyperplasia Due to 11-β-Hydroxylase Deficiency

Congenital adrenal hyperplasia (CAH) is a recessive disease, featuring cortisol deficiency caused by inherited defects of enzymes required for cortisol biosynthesis in the adrenal gland. Affected patients have increased ACTH levels due to impaired feedback inhibition, which stimulates excessive production of steroids proximal to the impaired step in cortisol biosynthesis. Precursor steroids can have androgenic effects, and female patients may present with virilization. In the most common form of CAH (21-hydroxylase deficiency), impaired synthesis of mineralocorticoids leads to salt-wasting (see below), while other defects can result in accumulation of precursors which have mineralocorticoid effects, leading to hypertension.

The most common form of congenital adrenal hyperplasia associated with hypertension is 11-β-hydroxylase deficiency ( Figure 36.3 ). Affected females may present with ambiguous genitalia at birth or with menstrual irregularities and hirsutism in adolescence or adulthood, while male subjects may present with penile enlargement or precocious puberty. Hypertension and hypokalemia are variable associated features, and are thought to occur due to the accumulation of 11-deoxycorticosterone, a moderately potent mineralocorticoid. However, there is no clear correlation between 11-deoxycorticosterone levels, the degree of virilization and the presence of hypertension. Renin and aldosterone are typically low.

In most cases, 11-β-hydroxylase deficiency is caused by recessive point mutations that cause loss-of-function of CYP11B1, although loss-of-function mutations due to the reciprocal product of unequal crossing-over between aldosterone synthase and 11-β-hydroxylase found in GRA (see above) can also produce loss-of-function mutations (in this case there is a chromosome that only expresses 11-β hydroxylase in adrenal glomerulosa).

11-β-hydroxylase deficiency should be suspected in female infants with ambiguous genitalia and male infants with penile enlargement. In the late onset form, hirsutism and menstrual irregularities in female patients and precocious puberty in boys may be the only signs of presentation. The abnormal hormone profile, in particular the finding of elevated basal and ACTH-stimulated 11-deoxycortisol concentrations, is diagnostic. Treatment with exogenous glucocorticoids suppresses ACTH secretion and the accumulation of precursor steroids. Precursor steroids should be monitored, and glucocorticoid dosage has to be carefully adjusted to avoid growth inhibition in children and exogenous Cushing’s syndrome. Persistent hypertension may require additional treatment with aldosterone antagonists or amiloride.

Congenital Adrenal Hyperplasia Due to 17-α-Hydroxylase Deficiency

17-α-hydroxylase deficiency is a rare cause of congenital adrenal hyperplasia. Patients typically present in adolescence with lack of pubertal development. Genetic females have primary amenorrhea and do not develop secondary sexual characteristics; genetic males typically have complete pseudohermaphroditism with female external genitalia and intra-abdominal testes, although ambiguous genitalia have been reported. Hypertension is a common finding, and may be associated with hypokalemia.

17-α-hydroxylase deficiency is caused by mutations in CYP17. The encoded enzyme is expressed in adrenal gland and gonads, and has both 17-hydroxylase and 17,20-lyase activities ( Figure 36.3 ). Most patients with CYP17 mutations thus have combined deficiencies of both enzymatic functions, although isolated deficiency of 17,20-lyase activity has been reported. Hydroxylation at carbon 17 of the steroid nucleus is required for cortisol production, and in its absence normal activation of the glucocorticoid receptor is not achieved. In addition, 17,20-lyase activity is needed for the generation of androgen and estrogen precursors from cortisol precursors.

Patients with CYP17 mutations consequently have cortisol deficiency leading to elevated ACTH levels, as well as androgen and estrogen deficiency, accounting for sexual infantilism and pseudohermaphroditism. Hypertension is caused by activation of the mineralocorticoid receptor (MR) due to increased levels of 11-deoxysteroids (corticosterone, 11-deoxycorticosterone, and 18-hydroxy-deoxycorticosterone), similar to the mechanism of hypertension in 11-β-hydroxylase deficiency. As a result, renin and aldosterone are typically suppressed.

The diagnosis is based on the clinical presentation and characteristic hormone profile, and treatment relies on replacement of glucocorticoids and sex steroids, the latter starting in adolescence.

Liddle Syndrome

Liddle syndrome is an autosomal dominant trait featuring hypertension with variable hypokalemia and alkalosis. While these features recapitulate the phenotype of the diseases with mineralocorticoid excess discussed above, there are no circulating mineralocorticoids in patients with Liddle syndrome. Hypertension is responsive to treatment with blockers of the epithelial sodium channel ENaC (amiloride or triamterene), but not to spironolactone.

Genetic studies identified rare mutations in subunits of ENaC as the cause of Liddle syndrome ( Figures 36.5 and 36.6 ). ENaC is composed of subunits encoded by three homologous genes, alpha, beta, and gamma ; each has cytoplasmic amino and carboxy termini, and traverses the plasma membrane twice. Virtually all mutations that cause Liddle syndrome eliminate a sequence required for removal of the channel from the cell surface by clathrin-coated pit-mediated endocytosis. This internalization sequence, PPPXY (where X is variable among different ENaC subunits) is present in the carboxy terminus of each subunit. Disease causing mutations introduce premature termination codons after the second transmembrane domain and before the PPPXY sequence, or alternatively produce missense mutations at one of the cognate positions of the internalization motif. These mutations result in reduced clearance and prolonged half-life of the channel at the cell surface ( Figure 36.6 ). Nedd4 was subsequently identified as having domains that specifically bind the PPPXY domains in ENaC, and contribute to its clearance from the cell surface. This protein contains ubiquitin-ligase domains, suggesting that ubiquitination and subsequent degradation plays a role in ENaC regulation. The likely candidate for this regulation in human kidney is Nedd4-2. Together, these findings suggest that mutations in ENaC in Liddle syndrome impair binding of Nedd4 to the mutant channel, and prevent ubiquitination and degradation. In addition to the mutations deleting or altering the PY motif, a mutation in the extracellular domain of γ-ENaC has been reported as causing an increased open probability of the channel. Consistent with these genetic findings, renal transplantation can cure the hypertension of Liddle syndrome.

Salt reabsorption in the cortical collecting duct.

Na ⁺ reabsorption in the principal cells of the cortical collecting duct is dependent on aldosterone and happens via ENaC. The resulting lumen-negative potential promotes K ⁺ secretion via ROMK and H ⁺ secretion.

Increased surface expression of the epithelial sodium channel ENaC causes Liddle syndrome.

The β- and γ-subunits of the channel contain a PPPXY motif in the cytoplasmic C-terminus. In normal individuals, binding of Nedd4-2 (not shown) leads to internalization in clathrin-coated pits. Channels are then either degraded or recycled back to the membrane. ENaC lacking the PPPXY motif in Liddle syndrome does not interact with Nedd4-2, and as a result is retained at the membrane. This causes increased Na ⁺ reabsorption and hypertension in affected individuals

(reproduced with permission from ref. ).

A mouse model of Liddle syndrome does not develop hypertension or hypokalemic alkalosis on a standard diet, despite increased ENaC activity. However, a Liddle phenotype develops when these animals are fed a high-salt diet.

Liddle syndrome should be suspected in patients with early hypertension and suppressed PRA and aldosterone levels; hypokalemia and metabolic alkalosis are supportive, but not essential findings. As discussed above, these features are shared with AME. A family history of hypertension consistent with autosomal dominant transmission, however, is suggestive of Liddle syndrome. Genetic testing is recommended. Treatment with amiloride or triamterene can correct hypertension and hypokalemia.

The Syndrome of Hypertension with Hyperkalemia (Pseudohypoaldosteronism Type II)

Aldosterone is secreted in two distinct physiologic conditions: volume depletion and hyperkalemia. Restoration of normal intravascular volume in the former condition requires increased salt and water reabsorption, while K ⁺ secretion must be increased in the latter state. How the kidney achieves these alternative results has been obscure. Classical explanations have suggested that this is a result of alterations in fluid flow and/or delivery of salt to the distal nephron. The notion that this is an incomplete explanation comes from a disease in which the kidney appears to be unable to use aldosterone to direct K ⁺ secretion, but instead constitutively activates salt reabsorption. This disease, referred to as pseudohypoaldosteronism type II (also known as syndrome of hyperkalemia with hypertension, Gordon syndrome) invariably features hyperkalemia, and has variable levels of hypertension and distal renal tubular acidosis. Aldosterone levels are normal to slightly elevated, and plasma renin activity is suppressed. Serum K ⁺ levels are virtually always greater than 5.0 mM without therapy, and can be greater than 7.0 mM. The disease was first described in 1964 by Paver and Pauline, who reported a 15-year-old boy with blood pressure 180/120 mmHg, serum potassium 7.0–8.2 mM, but normal glomerular filtration, and suppressed PRA and normal aldosterone. Since early descriptions, the disease has been recognized to be familial, initially described as an autosomal dominant trait.

Subsequent studies have demonstrated that K ⁺ cannot be corrected by supplemental aldosterone; however, both hypertension and hyperkalemia can be corrected by elimination of chloride intake or by low doses of thiazide diuretics. These findings have suggested that increased activity of the thiazide-sensitive Na-Cl co-transporter may play a role in this disease.

Analysis of genetic linkage in families followed by molecular studies identified mutations in two members of a novel family of serine-threonine kinases, PRKWNK1 and PRKWNK4, as causes of PHAII ( Figure 36.7 ). Mutations in WNK4 were novel missense mutations that predominantly clustered in a short acidic domain of unknown function in the protein. Mutations in WNK1 were large deletions of 20,000 to 40,000 base pairs in the first intron of WNK1 that caused increased expression of WNK1 mRNA. These genes were both found to be expressed in the distal convoluted tubule and collecting duct of the kidney.

Salt reabsorption in the distal convoluted tubule.

Salt reabsorption is driven by the activity of the Na ⁺ /K ⁺ -ATPase on the basolateral suface. K ⁺ entering the cell via the sodium pump recirculates through Kir4.1 (“pump leak coupling”). NaCl enters the cell via NCCT following the favorable Na ⁺ gradient, and Cl ⁻ exits via basolateral ClC-Kb channels. The basolateral NCX transporter uses the Na ⁺ gradient for Ca ²⁺ reabsorption. Ca ²⁺ enters on the apical site via the TRPV5 channel. On the apical surface, Mg ²⁺ enters the cell via TRPM6. The molecular identity of a putative basolateral Mg ²⁺ transporter is unknown. NCCT activity is regulated by the WNK1 and WNK4 protein kinases.

At the time of this discovery, the normal and disease-related functions of WNK1 and WNK4 were entirely unknown. Much subsequent work has been done to determine the roles these WNK kinases play in physiology. It appears that both play a role in coordinating the functions of diverse electrolyte flux pathways, and that a key function in the kidney is to regulate the balance between Na-Cl reabsorption and K ⁺ secretion. The WNK kinases regulate ion transport by several mechanisms, among them phosphorylation via intermediary kinases (SPAK, OSR1, and SGK1), and regulation of protein trafficking.

Early studies addressed the role of WNK4 in the regulation of NCCT using the oocyte expression system. Co-expression of NCCT with WT-WNK4 resulted in inhibition of thiazide-sensitive Na ⁺ uptake due to reduced surface expression. This effect was completely abrogated when a mutation seen in PHA II (Q562E) or kinase-dead WNK4 was used, suggesting that NCCT inhibition is lost in patients with PHA II, causing hypertension. Subsequent studies in mammalian cells suggested that this inhibitory WNK4 effect is mediated by suppression of membrane delivery via lysosomal accumulation.

Co-expression with WNK4 also dramatically reduces ROMK activity in oocytes, by interaction with an endocytic scaffold protein, intersectin. However, in contrast to the effect on NCCT, PHA II-mutant WNK4 produces increased inhibition of ROMK, accounting for hyperkalemia in PHA II. While WT WNK4 led to a modest increase of paracellular chloride permeability in MDCK cells, this effect was greatly augmented in PHA II-mutant WNK4, which may contribute to the chloride retention observed in PHA II. In addition, WNK4 inhibited ENaC activity in oocytes, and this inhibition was relieved by PHA II mutations. The inhibitory effect of WNK4 on ENaC and ROMK appears to require phosphorylation of WNK4 by SGK1, a kinase which translates aldosterone effects in the kidney.

Similar studies as for WNK4 were performed for WNK1. While a direct effect of WNK1 on NCCT was not identified, a long isoform of WNK1 (L-WNK1) inhibits WNK4 (which inhibits NCCT, see above), and a kidney-specific isoform (KS-WNK1) inhibits L-WNK1. In addition, L-WNK1 inhibits ROMK, an effect reversed by KS-WNK1 and synergistic with, but not dependent on, the WNK4 effect on ROMK ; L-WNK1 also activates ENaC. Finally, similar to WNK4, WNK1 increases paracellular chloride flux. ATII, which is produced in volume depletion, appears to be an upstream regulator of WNK4, attenuating its inhibitory effect on NCCT and thus stimulating salt reabsorption via NCCT. In contrast, the inhibitory effect of WNK4 of ROMK is retained. These findings suggest that PHA II mutations mimic a state of volume depletion with increased angiotensin II levels.

Regulation of several additional proteins (e.g., TRP channels, SLC26A6) by WNK1 and WNK4 kinases has been reported, and studies of the related kinase WNK3 has revealed a role in regulation of NKCC, as well as KCC co-transporters.

Taken together, in vitro studies suggest a key role of the WNK kinases in a functional switch among three states governing the balance between salt reabsorption and K ⁺ secretion. In a basal equilibrium state, both NaCl reabsorption and K ⁺ secretion are inhibited by the WNK kinases. In the setting of low intravascular volume, AII signaling inhibits WNK4’s inhibition of NCC, but increases inhibition of ROMK; this effect is mimicked by PHA II mutations, accounting for constitutive NaCl reabsorption and hyperkalemia seen in affected patients. In a third state, hyperkalemia leads to increased aldosterone production by the adrenal gland and, potentially via SGK, to increased ROMK activity and increased ENaC activity, with sustained inhibition of electroneutral salt reabsorption via NCCT.

Animal models have confirmed the role of WNK1 and WNK4 in generating hyperkalemia and hypertension. A transgenic mouse model carrying an additional copy of PHA II-mutant WNK4 shows hypertension, hyperkalemia, and hypercalciuria and, in addition, increased mass of the distal convoluted tubule and increased NCCT expression. In contrast, an additional copy of WT WNK4 leads to decreased DCT mass and reduced NCCT abundance. A knockout mouse model of WNK4 displays a phenotype similar to Gitelman syndrome (see below), with increased plasma renin activity, and decreased NCC activity. Deletion of the first WNK1 intron as found in PHA II in mice leads to ubiquitous ectopic KS-WNK1 expression, as well as overexpression of L-WNK1 in the DCT, which explains increased activity of NCCT through alleviation of WNK4 inhibition.

Importantly, mutations in WNK1 and WNK4 explained only about 15% of families with PHA II. Exome sequencing of unrelated PHA II cases led to the identification of two additional genes for PHA II, which together account for nearly all of the remaining cases. These two genes are partners in a ubiquitin ligase complex. Cullin 3 (CUL3) is a scaffold that serves to assemble a complex containing a RING E-3 ubiquitin ligase and one of many BTB domain-containing proteins that target specific proteins for ubiquitination. One of these BTB domain proteins is Kelch-like 3 (KLHL3). KLHL3 has a C-terminal BTB domain, and an N-terminal kelch domain, a six-bladed propeller that is likely used to bind and target proteins for ubiquitination. Surprisingly, different families showed either autosomal dominant or autosomal recessive transmission of mutations in KLHL3. Recessive mutations are distributed throughout the protein, and include typical loss-of-function mutations such as premature termination mutations. In contrast, dominant mutations cluster at positions in the kelch domain that are clustered at parts of the domain inferred to be involved in target binding. Because these dominant mutations produce a phenotype very similar to that caused by homozygous loss-of-function mutations in KLHL3, the dominant mutations likely have a dominant-negative effect.

The mutations in CUL3 that cause PHA II all cause skipping of exon 9, resulting in an in-frame deletion of 57 amino acids between the BTB-binding and RING-binding domains. These mutations are all dominant, and nearly all appear to be de novo mutations. Because these mutations phenocopy recessive loss-of-function mutations in KLHL3, it is inferred that these CUL3 mutations selectively prevent ubiquitination at targets bound by KLHL3. While CUL3 is ubiquitously expressed, KLHL3 is selectively expressed in the distal nephron.

Comparison of the phenotypes of patients with mutations in these different genes reveals strong genotype–phenotype correlations. Patients with CUL3 mutations have the most severe hyperkalemia, are nearly always hypertensive at young ages, and commonly have acidosis. Patients with recessive and dominant KLHL3 mutations have intermediate phenotypic severity, and those with WNK1 and WNK4 mutations often have only hyperkalemia without hypertension at early ages, and do not uniformly develop hypertension at later ages. The molecular mechanisms that link mutations in KLHL3 and CUL3 to PHA II are presently unknown; however, it is highly likely that these mutations will intersect with the WNK/NCC/ROMK pathway.

Autosomal Recessive Pseudohypoaldosteronism Type I (PHAI)

Infants with autosomal recessive PHA I frequently present with life-threatening episodes of hypovolemia, sometimes leading to neonatal death. Affected subjects have marked renal salt-wasting and striking hyperkalemia (as high as 12 mM), and renal tubular acidosis despite drastic elevations of PRA and aldosterone. There is also salt-wasting from salivary and sweat glands, as well as the colon. These patients require life-long sodium supplementation and often chronic treatment with resin-binding agents for hyperkalemia. Frequent pulmonary infections somewhat reminiscent of cystic fibrosis have been reported in some patients, and markedly increased ciliary clearance of lung water has been reported.

Genetic studies led to identification of recessive loss-of-function mutations in the alpha, beta or gamma subunits of the epithelial Na ⁺ channel (ENaC) in virtually all affected subjects. Functional studies in Xenopus oocytes confirmed loss of channel activity. In summary, these findings demonstrated that mutations in all ENaC subunits can cause PHA I. All the features of this disease are explained by loss of ENaC function. Loss of ENaC function results in dramatic salt-wasting, leading to activation of the renin–angiotensin system; however, because ENaC is absent, this fails to adequately augment salt reabsorption in the kidney. Impaired salt absorption in the colon may also contribute to disease severity. K ⁺ and H ⁺ secretion are also markedly impaired, owing to loss the lumen-negative potential resulting from loss of ENaC, causing hyperkalemia and acidosis.

These findings have been confirmed in animal studies. Mice deficient for the β- and γ-subunits show a phenotype characteristic of PHA I, with early death due to Na ⁺ -wasting and K ⁺ retention, and a PHA I-like phenotype is observed in α-ENaC knockout mice with transgenic expression of rat α-ENaC for rescue. α-ENaC knockout mice show early postnatal demise with a mean survival of 24 hours postnatally due to respiratory distress, suggesting a role of the mouse channel in pulmonary liquid clearance after birth that is apparently not observed at birth in humans but is found later in life.

Autosomal Dominant PHA I

Autosomal dominant PHA I has a milder phenotype that typically largely resolves in the early years of life. Patients may be asymptomatic or can present with renal salt-wasting and mild-to-moderate hyperkalemia and acidosis. NaCl supplementation can typically be discontinued after early childhood, as the disease remits with age. Geller et al. identified heterozygous mutations in the mineralocorticoid receptor gene in five kindreds with autosomal dominant PHA I. All mutations were inferred to be loss-of-function, inferred to cause diminished ENaC activity, with resulting salt loss. Since Na ⁺ reabsorption via ENaC is required for the lumen-negative potential that drives K ⁺ and H ⁺ secretion in the cortical collecting duct, diminished ENaC activity can cause hyperkalemic acidosis in these kindreds. The improvement with age is consistent with a decreasing requirement for aldosterone action in older children, presumably due to increasing salt intake and a self-selected high-salt diet after infancy. Adult carriers show no abnormalities of serum and urinary electrolytes or blood pressure, but consistently demonstrate markedly elevated serum aldosterone levels, consistent with homeostasis being achieved by compensating for the reduced level of MR by increased aldosterone levels.

A case of autosomal recessive PHA I due to mutations in the mineralocorticoid receptor was recently reported – a newborn with very severe PHA I who was compound heterozygous for S166X and W806X deletions. Within her family, carriers of the S166X mutation had typical autosomal dominant PHA I, while carriers of the W806X mutation were clinically asymptomatic, but had elevated renin and aldosterone levels.

A mouse model of PHA 1 shows increased renin and aldosterone, but develops normally in the heterozygous state. The homozygous knockout is lethal, but can be rescued by subcutaneous NaCl injection before weaning.