Introduction
Liddle syndrome (LS) was first reported at 1963 by Liddle et al. (1). It is a monogenic hypertension caused by mutations in genes encoding epithelial sodium channels (ENaCs) (2). ENaCs are composed of three subunits: α, β, and γ (encoded by SCNN1A, SCNN1B, and SCNN1G, respectively) (3). Proline-rich segments, referred to as Pro-Pro-Pro-X Tyr (PY) motifs, in the carboxyl terminal regions of β and γ subunits are binding sites for WW domains of neural precursor cell expressed, developmentally Down-regulated 4 (NEDD4), an E3 ubiquitin-protein ligase (4). By binding the ubiquitin-protein ligase domain, PY motifs inactivate ENaCs (5). LS is characterized with increased sodium reabsorption in collecting tubules resulted from ENaCs dysfunction, including inactivation disorders, and abnormal opening frequencies (3, 6–8). The function of ENaCs usually is examined by patch clamp techniques.
The typical manifestations of LS include early-onset hypertension, hypokalemia and suppression of plasma renin activity (PRA), and plasma aldosterone concentration (PAC) (3, 9). Some LS patients lack a typical presentation and are not diagnosed and treated in time, which can lead to major complications such as stroke and early death because of poor control of hypertension (10–13). Genetic sequencing is recommended for LS screening for SCNN1A, SCNN1B, and SCNN1G mutations, especially in families with a history of early-onset hypertension (3). Through family screening, it often screens for children who carried mutations in family (11, 14). Genetic sequencing screened LS patients, but the pathogenicity of most variants has not been verified by functional experiments. For LS patients diagnosed based on sequencing results and silico analysis prediction, the usual follow-up after beginning amiloride treatment is approximately 1 or 2 months and the long-term outcomes of LS patient remains unclear.
In this study, we explored the pathogenicity of a nonsense mutation (c.1171G > T) in exon 13 of SCNN1G reported in 2015 year by our team and efficiency of blood pressure control by amiloride. We used patch-clamp to verify functional mechanism of this mutation, and found that amiloride-sensitive currents differed significantly between wild-type and mutant groups. All members harboring the c.1171 G > T mutation in SCNN1G with hypertension and/or abnormal biochemical presentations were diagnosed with LS. Over a 7-year follow-up, the results show efficacy with tailored treatment of amiloride in blood pressure control and prevention of target organ damage.
Discussion
In this study, the pathogenicity of a nonsense mutation (c.1171 G > T) in SCNN1G was confirmed by functional experiments. The blood pressure and electrolyte concentrations of LS patients in this family returned to normal after 1 month of amiloride treatment, and no target organ damage and cardiovascular events occurred over a 7-year follow-up.
LS is a rare autosomal dominant disease, the prevalence reported as 0.91 and 1.52%, respectively (2, 17). LS is caused by mutations in SCNN1A, SCNN1B, and SCNN1G, which encode ENaCs in kidney tubules, affecting sodium reabsorption (18). Proline-rich segments in the C terminals of β and γ subunits, which are critical for function, are known as PY motifs (19). The ubiquitin ligase NEDD4 binds to ENaC PY motifs, leading to the ubiquitination and degradation of ENaC (20). Mutations typically prevent the ubiquitination of subunits, thus inhibiting the rate at which they are internalized from the membrane, resulting in increased ENaC enrichment and elevated channel activity (3). However, a minority of mutations increase ENaC activity by changing the open probability in the membrane (20, 21). In 2017, a SCNN1A mutation, a gain-of-function mutation in the extracellular domain of the α subunit, was reported, which primarily increased channel open probability rather than channel surface density (8). By altering ENaC activity, sodium reabsorption increases, which leads to volume expansion and hypertension.
There are 29 mutation sites linked to LS, and nearly all mutations delete or alter PY motifs, and include frameshift, nonsense, or missense mutations (12, 20, 22–24). Among the SCNN1G genetic spectrum, 10 sites have been detected (3, 13, 22, 23), including four nonsense mutations, three missense mutations, and three frameshift mutations (Table 2). The first identified SCNN1G mutation was a nonsense mutation (p.Trp573*) reported by Hansson et al.; Xenopus oocytes expressing mutant p.Trp573* displayed 7.5-fold increase in amiloride-sensitive sodium current than wild-type (25). In the following years, a nonsense mutation site (p.Trp575*) was detected in a sporadic case with hypertension in Japan (26), and Shi et al. and Zhang et al. reported a p.Gln567* mutation in a Chinese family and sporadic case, respectively (27, 28). However, pathogenic functional experiments were not conducted for the p.Gln567* mutation. Three frame mutations and one missense site have been reported in different families, and their clinical features are summarized in Table 2. Most mutation sites impact PY motifs, except for p.Asn530Ser. The p.Asn530Ser mutation is in the extracellular domain and does not influence the PY motif, but it does cause an LS phenotype (21). In patch-clamp experiments, the p.Asn530Ser mutation presented twofold higher amiloride-sensitive currents compared with the wild type, by increasing the open probability of channels (21). The p.Glu571* mutation was first reported by our team in a Chinese family, and this variant has also been identified in another Chinese family (2, 17). In the present functional experiments, cells with the p.Glu571* mutation displayed 3.7-fold amiloride-sensitive currents of wild-type cells.
LS is a kind of channelopathy, and patch-clamp experiments are very important for verifying ion channel diseases. According to previous research, pathogenic mutations in SCNN1G mostly affect the PY motif, but pathogenicity has only been functionally verified in three sites. The physiological characteristics of ENaCs expressed in Xenopus oocytes are similar to those of ENaCs in human distal renal tubules (29); thus, functional experiments usually use Xenopus oocytes for detecting current variations in patch-clamp experiments. In the present study, however, we use CHO cells to detect current variations. CHO cells are used in patch-clamp electrophysiological experiments because of their low endogenous expression of ion channels (30). Amiloride-sensitive-sodium currents were suppressed after adding amiloride in the present study, indicating drug efficacy from an ion channel aspect. Different variants display different degrees of amiloride-induced sodium current changes, which may be associated with the effects of mutation sites on PY motifs and the cell model. Compared with extracellular domain mutations, PY motif mutations expressed higher amiloride-sensitive currents, indicating that PY motif mutations have a greater impact on ENaC activity.
Phenotypes vary greatly in patients harboring different mutations, as well as for the same mutation in a pedigree (Table 2). A systematic review conducted in 2018 reported that 92.4% of patients with LS presented hypertension (3). In the current pedigree, all members carrying the p.Glu571* variant presented hypertension, and some even had hypertensive stroke. Moreover, one of the typical features of LS is early-onset hypertension; most patients develop hypertension before 30 years. The youngest reported patients are a 10-week-old baby, who was a sporadic case, and a 2-year-old child from an LS family (9, 31). According to the largest retrospective analysis of LS patients, the average onset age of hypertension is 15.5 ± 3.3 years (7). In our research, the index case was diagnosed with hypertension at 18 years old with left ventricular concentric hypertrophy. We speculated that the proband probably had hypertension in child, and she had no obvious symptoms and had not been diagnosed in time. Another 7-year-old pediatric patient had hypertension, hypokalemia, and no target organ damage, as well he had no obvious symptoms and his onset age is unclear. The remaining family members diagnosed with hypertension are almost less than 30 years old.
In this family, there was heterogeneity not only in the occurrence of high blood pressure, but also in the severity of hypertension and complications. Reported complications of persistent hypertension in LS patients include cerebrovascular accidents, early death, renal insufficiency, nose bleeds, and retinal damage (7, 12, 32). Cerebrovascular accidents before the age of 40 were common in this LS family, especially stroke, which causes a heavy burden to patients and their families. The reason for the high incidence of stroke in this family may include poor management of blood pressure, genotype, and environmental factors.
Hypokalemia is present in 71.8% of cases of LS (3). In the current study, all mutation-carrying participants exhibited hypokalemia. Tamura et al. and Fan Peng et al. reached a similar conclusion: that some LS patients have hypertension but exhibit normokalemia (13, 33). The suppression of PRA and PAC is considered a typical presentation, and can be used to differentiate other possible causes of secondary hypertension (34). In our study, all members presented with suppression of PRA, which is in line with the typical characteristics of LS. However, the PAC levels of all members were within the normal range, and members harboring the same variant in another LS pedigree also presented normal PAC (17). The presentation of PAC shows clear heterogeneity compared with other variant-related phenotypes. The reasons for the heterogeneity of LS may be related to aspects including environmental factors and gene polymorphisms. LS patients with SCNN1A mutation may presented with mild phenotype than typical LS, which may attribute to different pathogenetic mechanisms (8). Therefore, for patients with suspected LS phenotype but milder symptoms, the possibility of carrying SCNN1A mutation should be considered.
Clinical presentations and laboratory findings in LS are heterogeneous, which can hamper its diagnosis. Family history can provide clues, and combined with early-onset hypertension and laboratory results, can be used to identify suspected LS patients. Inadequate symptoms and insufficient awareness of children with monogenic hypertension leading to fail to intervene in time, and most of them have developed complications in youth (12). According to the previous study, 90% pediatric patients had LS family history, so genetic screening can provide definitive confirmation of adult, and pediatric LS (22, 35). Additionally, investigation of the pathogenicity of variants based on amiloride-sensitive currents is needed.
There were several limitations in the present study. This site was not a novel mutation—it has been briefly mentioned in a review by our team. Furthermore, we only compared the amiloride-sensitive sodium currents of γGlu571* mutations with wild-type channels, and not with other truncation mutations of the γ subunit, which may explain the heterogeneous results to some extent. The tailored treatment with amiloride efficacy in blood pressure control and avoiding cardiovascular events has only been validated in the long-term follow-up results of this family, which cannot be representative of all LS patients. A cohort study of LS patients is needed to explore the efficacy of amiloride in improving prognosis and blood pressure control.