Association of Familial Hyperkalemia and Hypertension with Proximal Renal Tubular Acidosis and Epileptic Seizures

Introduction: Familial hyperkalemic hypertension (FHHt) is an inherited disease characterized by hyperkalemia, hypertension, and hyperchloremic acidosis (HCA). The primary defect is a hyperactive sodium chloride co-transporter, expressed in the renal distal tubule. FHHt is caused by mutation in either WNK1, WNK4, KLHL3, or Cul3. The mechanism of HCA is not completely understood. Methods: Clinical and genetic data were collected from the largest family with FHHt described in the literature. Urine ammonia was measured in 26 family members. Epilepsy was diagnosed clinically. Results: Of the 85 family members, 44 are affected by the Q565E WNK4 mutation, and 28 are newly described. In genetically engineered mice, urinary ammonium was decreased. In our study, urine ammonium did not change. In 11 unaffected subjects, urine ammonia per creatinine was 8.013 ± 3.620 mm/mm, and in 15 subjects affected by FHHt, it was 8.990 ± 4.300 mm/mm (p = 0.546, not significant). Due to the large family size and prolonged follow-up, rare conditions can be identified. Indeed, two children have genetic generalized epilepsy and one child has migraine. The prevalence of epilepsy is 4.545% (2/44) much higher than in the general population (0.681%). This difference is statistically significant (χ2 with Yates correction = 5.127, p = 0.023). Conclusions: We provide further evidence that the origin of HCA in FHHt lies in the proximal renal tubule. The association of FHHt with epilepsy leads us to speculate that the raised serum K in susceptible subjects may cause a rise in CSF K, and extracellular cerebral K, leading to epilepsy.


Introduction
Familial hyperkalemic hypertension (FHHt) is a rare inherited disease characterized by hyperkalemia, hypertension, and hyperchloremic acidosis (HCA) [1,2].The primary defect is a hyperactive sodium chloride co-transporter, expressed in the renal distal tubule.FHHt is caused by mutation in one of four genes, WNK1, WNK4, KLHL3, and Cul3 [1,2].Here, we provide a description of the largest family described karger@karger.com in the literature to date.This family now consists of 85 subjects, of whom 44 are affected by the Q565E WNK4 mutation, i.e., consistent with autosomal dominant inheritance.
The prolonged follow-up and the large family size enabled us to extend our investigation of the nature of the hyperchloremic metabolic acidosis.A mechanism of distal renal tubular acidosis in a transgenic mouse involving WNK4-induced activation of pendrin, an electroneutral 2Cl − /2HCO 3exchanger, expressed in collecting duct β-intercalated cells has been proposed [3].However, the significance in humans is uncertain since mutations in the pendrin gene producing the "Pendred syndrome" (hearing loss accompanied by goitre, thyroid enlargement, with or without hypothyroidism, inherited in an autosomal recessive pattern) do not display a renal phenotype except for rare cases associated with hypokalemia [4] rather than the hyperkalemia in FHHt.By contrast, a bicarbonate loading test in subjects with FHHt displayed proximal renal tubular acidosis, i.e., hyperbicarbonaturia [5].As a further test of this hypothesis, we have now measured urinary ammonium in 26 members of the family, 15 of them affected by the FHHt gene mutation.In addition, we have made the observation that 3 of the 44 affected subjects had epileptic seizures -2 of grand mal type and one of migraine seizures.

Materials and Methods
Subjects were studied clinically and genetically as described in our previous publications [6,7].Spot urinary ammonia in all new and in some previously described untreated subjects was measured using a Sigma-Aldrich urine ammonia assay kit [8].Briefly, this assay is based on the o-phthalaldehyde method in which the reagent reacts with ammonia/ammonium ion producing a fluorometric result proportional to the ammonia concentration in the sample.Urine samples were stored at −80°C and were thawed to room temperature before use.All samples and standards were run in duplicates.NH4Cl standards concentrations were used for a standard curve ranging from 0 to 1.0 mM.Urine samples were diluted 150-fold with ultrapure water prior to assay.Dead cells and debris were removed by centrifugation.The working reagent mix was prepared according to the instructions and was added to each sample and standard well.The reaction was incubated for 15 min in the dark at room temperature.Then, the fluorescence intensity was measured (λex = 360 nm/ λem = 450 nm).A standard curve was created, and the slope was determined.Ammonium values were obtained by subtracting the background (assay blank) from all readings and dividing by the slope generated from the standard curve.Epilepsy was diagnosed according to clinical and EEG criteria.

Results
Figure 1 shows the family pedigree up to and including the seventh generation.The family now consists of 85 subjects, characterized clinically and genetically, 28 subjects are newly diagnosed.A total of 44 subjects have the Q565E WNK4 mutation and are phenotypically affected 1 .Two aspects of the disease are dealt with, renal ammonia metabolism and occurrence of epilepsy.The patients examined in this study are all from this large family reported in previous publications [6,7].

Renal Ammonia Metabolism
Table 1 shows the clinical characteristics of the patients examined.They are virtually identical to the data reported previously in affected and unaffected members of this family with FHHt [7].As seen in the Table 1, no significant difference in urinary anion gap (UAG) was detectable.Previously, it has been suggested that in patients with HCA, an indirect measure of urinary NH4 + may be the UAG [9].However, a more recent paper has cast doubt on the notion that the UAG is a reliable measure of urinary ammonium [10], which should be measured directly.We have measured both parameters, UAG as an initial rough measure and then ammonium directly (as suggested recently by Rehman et al. [11]).
The hypothesis proposed in [12] that the HCA is distal in origin would predict that the renal ammonium secretion would be decreased in the affected FHHt patients.Figure 2 shows urinary NH 4 + in mM per urinary creatinine in mM, in 11 unaffected, and in 15 affected subjects.It was 8.013 ± 3.620 in unaffected and 8.990 ± 4.300 in affected subjects.This difference is not significant (p = 0.546).The two groups did not differ in their age (10.27 ± 6.59 vs. 16.2 ± 12.8 years in unaffected and affected subjects, respectively, p = 0.173).The result is inconsistent with a distal renal origin but supports the other arguments for a proximal origin of the HCA.

Epilepsy
We have made the surprising observation that among the 85 subjects shown in Figure 1, two had epilepsy and one had migraine.All three were children between the ages of 6 and16 years.All three were among the 44 subjects carrying the Q565E WNK4 mutation.
Their clinical details are as follows: female subject VI 6: FHHt was diagnosed at age 9. Generalized tonicclonic seizures (GTCSs) and myoclonic seizures, mostly after awakening from sleep, appeared 7 years later at age 16 (2011).Serum K was 5.77, Cl: 111.The mean of previous serum K determinations was 5.8 ± 0.4 mM, (n = 8).Blood pressure was 110/70 mm Hg (percentile 70/60).Neurological examination was normal.EEG showed a general interictal epileptiform abnormality, in the form of bisynchronous 4 Hz spike and slow wave complexes, maximal anteriorly.Brain MRI was normal.The electro-clinical diagnosis of juvenile myoclonic epilepsy (JME) was made.Depression was also present.She was treated by thiazides.Her seizures proved resistant to lamotrigine, levertiracetam, topiramate, and clobazam in monotherapy and in various combinations.Major depression was treated with escitalopram.Her seizures remained poorly controlled, and valproate was not prescribed due to teratogenic concerns.Male subject VI 23: FHHt was diagnosed at age 3. Generalized seizures (GTCSs) appeared 6 years later at age 9 (2012).The mean of previous serum K determinations was 6.6 mM, (range: 5.9-6.9,n = 6).Blood pressure was 115/65 mm Hg (percentile 97/74).Neurological examination was normal.Several EEG examinations were normal.Brain MRI was normal.Seizures were controlled by a combination of thiazides and antiepileptic medications including valproic acid and clobazam.In 2020, he discontinued medications, and GTCS recurred.Serum K was 7.1 mM.Seizures were again controlled after resuming antiepileptic and thiazide therapy.
Female subject VI 17: at the age of 6 years (2019), episodes of headache appeared, usually once weekly, lasting about 2 h, and terminated by vomiting.At least once, uncontrolled movements of hands and lips were seen.Serum K was 6.95, 6.22, and 6.16 mM.Blood pressure was 132/75 mm Hg.She was put on hydrochlorothiazide 12.5 mg per day, and her symptoms improved significantly.A diagnosis of migraine was made.
In the affected members of the family, there is increased prevalence of epilepsy (2/44 = 5.545%), compared to that in the general population, which is 0.681% [13].This difference is statistically significant (χ 2 with the Yates correction = 5.1272, p = 0.0235).Although this value is significant, we believe that the reported cases may be less than the number of cases, due to a reluctance of this ultraorthodox family to reveal the illness because of social and family reasons.

Proximal Renal Tubular Acidosis
The mechanism of HCA in FHHt is not completely understood.Since HCA is accompanied by hyperkalemia, it is claimed to be caused by hyperkalemia.
In a genetic mouse model with renal distal convoluted tubule having constitutively activated SPAK (Ste20/SPS1-related proline-alanine-rich kinase), serum K was 5.1 mM compared to 3.9 mM in wild-type mice.The mechanism of the accompanying metabolic acidosis was proposed to be a defect in ammonia metabolism.Urine ammonia in the mutant was decreased to 5.0 mM/mM creatinine, compared to 7.2 mM/ mM creatinine in wild-type mice [12].Although the correlation between urine ammonia and serum K is strong in the range of serum K 2.3-4.5 mM, in the range of 4.5-6.2mM, this correlation is weak [12] if existing at all.
It was shown that in humans with FHHt, and the WNK4 Q565E mutation, serum K is raised from 4.2 ± 0.2 mM (n = 16) in unaffected subjects to 5.6 ± 0.2 mM (n = 18) in affected subjects [7].By contrast to the basic finding in DCT-CA-SPAK mice, hyperkalemia in man (subjects affected by FHHt), was not accompanied by a decrease in urine ammonia.In subjects unaffected by FHHt (n = 11), urine ammonia/creatinine was 8.013 ± 3.62, mm/mM, and in affected subjects (n = 15), it was 8.99 ± 4.30 mM/mM (p = 0.536).This is not inconsistent with the data on CA-SPAK mice, for the higher K, between 4.5 and 6.0 mM.Shirin/Rabinowitz/Blatt/Karlish/Farfel/ Mayan Thus, the overall conclusion is that the hyperchloremic acidosis in human subjects with FHHt represents proximal tubular acidosis.The distal tubular mechanisms of acidosis, associated with hyperkalemia in one mouse model (Harris et al. [12]) or in a FHHt transgenic mouse model, with a Q562E mutation in WNK4, the same mutation found in human FHHt [3] do not necessarily contradict our conclusion on proximal tubular acidosi.As mentioned above, this rests on the findings of bicarbonaturia [5,14] and absence of detectable changes in urinary ammonium excretion and also a virtual lack of renal phenotype in Pendred syndrome, and the latter's recessive inheritance mode, compared to the dominant pattern in FHHt.It is, nevertheless, noteworthy that the acidosis in FHHt in humans is mild, and it is certainly not excluded that under different pathophysiological conditions, genetic backgrounds, or animal species, distal mechanisms of acidosis are more prominent [3].

Epileptic Seizures
The finding of epilepsy in FHHt has not been described previously.A likely reason is that most of the described families with FHHt are small, making it difficult to reliably identify subjects affected by rare conditions such as epilepsy [1,2,15,16].
Despite prior observations that epilepsy has not been described under other conditions associated with hyperkalemia, such as advanced renal failure, we propose here that the hyperkalemia characteristic of FHHt is a likely cause of epilepsy.What might be the mechanism of epilepsy associated with hyperkalemia?In recent years, a great interest has been expressed in the dramatic phenomenon of cortical spreading depression (CSD), which was described only in 1940 [17][18][19].CSD is a phenomenon characterized by electrical excitability followed by an advancing wave of depolarization and depression of neural activity that can occur in various parts of the brain.CSD is associated with and is thought to trigger various pathological phenomena such as epilepsy and migraine.The detailed pathogenic mechanism is not completely understood, although it is significant that CSD can be induced experimentally by application of high local concentration of extracellular potassium [20].High CNS extracellular potassium is a core feature of the pathogenetic mechanism of familial hemiplegic migraine (type 2), a genetic disease in which inactivating mutations were found in the α2 sub unit of the Na,K-ATPase in astrocytes.The mutations reduce the ability of the astrocytes to regulate extracellular K, resulting from activity of CNS neurons [21,22].The normal K concentration in the CSF (2.7-3.9 mM) is lower than K in the blood (3.5-5.0 mM), and changes in serum K are not thought to affect CSF K concentrations [23].In FHHt serum, K concentration is elevated, and in affected children, it reaches 5.98 ± 0.42 mmol/L [24].However, we propose that in susceptible children affected by epilepsy in FHHt, there is a mechanism, perhaps genetic, or physiological such as acidosis, that produces a disruption in the blood-brain barrier and allows the elevated blood K to enter the CNS and raise extracellular K, thus triggering epileptic seizures.It may be relevant that WNK4 is highly expressed in the specialized endothelium of the blood-brain barrier, and overlaps with the tight junction protein ZO-1 [25].Obviously, this proposed mechanism is speculative but it implies that control of blood K levels will effectively treat the epilepsy in these children and might be tested in the clinic.

Statement of Ethics
This study protocol was reviewed and approved by the Sheba Medical Centre Ethics committee, approval number 6805-20-SMC.Participation was approved by all subjects described in this study.The authors have no conflicts of interest to declare.