Molecular Pathogenesis
Pathogenic variants in the genes encoding two members of the WNK protein family of serine-threonine kinases, WNK1 and WNK4, have been implicated in the pathogenesis of pseudohypoaldosteronism type II (PHAII) [Wilson et al 2001]. Members of this kinase family are named WNK (or with no lysine [K]) kinases because of their unique substitution of cysteine for lysine at a highly conserved residue within the catalytic kinase domain [Xu et al 2000]. Over the past decade, members of the WNK kinase family have been shown to regulate the coordinated transport of Na+, K+, and Cl- ions across epithelia in a variety of tissues [Kahle et al 2008].
Alterations in WNK1 and WNK4 are present in only a minority of individuals with PHAII, motivating the search for additional genetic contributions to the disorder in other families. Recently, pathogenic variants in kelch-like 3 (KLHL3) and cullin 3 (CUL3) have been identified in the majority of families with PHAII [Boyden et al 2012, Louis-Dit-Picard et al 2012]. The protein products of CUL3 and KLHL3 function together as part of the cullin-RING E3 ubiquitin ligase complex, which has a role in ubiquitin-mediated protein degradation.
The electrolyte and blood pressure abnormalities in individuals with PHAII are readily corrected with thiazide diuretics, inhibitors of the Na-Cl cotransporter (NCC; encoded by SLC12A3) expressed in the renal distal convoluted and connecting tubules (see Management, Treatment of Manifestations). This clinical observation led to the initial hypothesis that increased activity of NCC could play a role in the pathogenesis of PHAII [Gordon 1986]. However, to date, no PHAII-causing variants in the gene encoding NCC have been demonstrated.
CUL3
Gene structure.
CUL3 is widely expressed and has several alternatively spliced isoforms. For a detailed summary of gene and protein information, see Table A, Gene.
Pathogenic variants.
CUL3 alterations causing PHAII occur within intron 8, exon 9, or intron 9 and disrupt splicing of exon 9. This results in a 57-amino acid in-frame deletion.
Normal gene product. CUL3 is a component of the cullin-RING E3 ubiquitin ligase complex, which functions in ubiquitin-mediated proteolysis. Immunostaining of mouse kidney demonstrates that CUL3 is present throughout the nephron [Boyden et al 2012].
Abnormal gene product. It has been suggested that CUL3 pathogenic variants in PHAII may disrupt ubiquitination of at least a subset of KLHL3 targets [Boyden et al 2012]. The observation that all reported CUL3 pathogenic variants impair splicing of exon 9 and result in an in-frame deletion of a segment of CUL3 may hint at the functional specificity of these pathogenic variants [Boyden et al 2012].
KLHL3
Gene structure.
KLHL3 is widely expressed with at least three alternatively spliced isoforms reported. For a detailed summary of gene and protein information, see Table A, Gene.
Pathogenic variants. Dominant KLHL3 pathogenic variants that cause PHAII are missense alternations that cluster in segments connecting beta-strands of Kelch propeller blades within the C-terminal Kelch propeller domain and disrupt substrate binding.
Recessive KLHL3 pathogenic variants occur throughout the gene and include frameshift and premature termination variants.
Normal gene product. KLHL3 is a component of the cullin-RING E3 ubiquitin ligase complex, which functions in ubiquitin-mediated proteolysis. KLHL3 contains an N-terminal bric-a-brac tramtrack broad complex (BTB) domain that has a role in binding to CUL3 and a BTB and C-terminal Kelch (BACK) domain. KLHL3 also contains a C-terminal six-bladed Kelch propeller domain that functions in substrate binding.
Wild type KLHL3 has been shown to bind to WNK4 and WNK1 [Ohta et al 2013, Shibata et al 2013, Wakabayashi et al 2013]. Furthermore, binding of KLHL3 to WNK4 has been shown to result in ubiquitination and degradation of WNK4 in vitro, suggesting that KLHL3 normally regulates WNK4 levels via ubiquitin-mediated proteolysis [Shibata et al 2013, Wakabayashi et al 2013, Wu & Peng 2013]. Expression of KLHL3 together with binding of KLHL3 to WNK1 has also been shown to result in ubiquitination of WNK1 [Ohta et al 2013].
Immunostaining of mouse kidney demonstrates that KLHL3 is predominantly present in the distal convoluted tubule and collecting duct [Boyden et al 2012, Louis-Dit-Picard et al 2012].
Abnormal gene product. Missense variants in KLHL3 that cause PHAII disrupt binding to WNK4, WNK1, or CUL3 and lead to decreased ubiquitination and increased levels of WNK4 [Mori et al 2013, Ohta et al 2013, Shibata et al 2013, Wakabayashi et al 2013]. Mutated KLHL3 has also been shown to decrease WNK4-mediated clearance of ROMK from the cell membrane [Shibata et al 2013].
WNK1
Gene structure.
WNK1 transcript variant 1 (reference sequence NM_018979.3) has 30 exons and encodes the most common protein isoform. Alternatively spliced transcript variants have been described; the full-length nature of all of them has yet to be determined. For a detailed summary of gene and protein information, see Table A, Gene.
Pathogenic variants. Pathogenic variants in WNK1 were initially reported in two families with PHAII [Wilson et al 2001]. Both are large deletions (41 kb and 21 kb) that occur within the 60-kb intron 1 of WNK1. The deletions do not affect the coding sequence of the flanking exons. Recently, 6 kindreds (9 individuals) were reported to have pathogenic missense variants in WNK1, coupled with a phenotype that includes hyperkalemia and acidosis; the hypertension was minimal. The pathogenic missense variants clustered in the region homologous to the acidic domain mutated in WNK4 [Picard et al 2015b].
Table 2.
Selected WNK1 Pathogenic Variants
View in own window
| DNA Nucleotide Change | Predicted Protein Change | Reference Sequences |
|---|
| g.18538_59810del 1, 2 | See footnote 3. |
NG_007984.2
|
| g.28500_50277del 2 | See footnote 3. |
Variants listed in the table have been provided by the author. GeneReviews staff have not independently verified the classification of variants.
GeneReviews follows the standard naming conventions of the Human Genome Variation Society (varnomen.hgvs.org). See Quick Reference for an explanation of nomenclature.
- 1.
Deletion within intron 1 (also known as NG_007984.2:c.759+12272_760-5774del)
- 2.
- 3.
No amino acid change is expected.
Normal gene product.
WNK1 encodes at least four different alternatively spliced WNK1 transcripts (See see Table A, Gene further details). The interplay between these isoforms is complex. Two variants were initially identified: a longer isoform with ubiquitous tissue expression (L-WNK1) and a smaller isoform that lacks the 5' kinase domain of the larger isoform and appears to be kidney-specific (KS-WNK1) [Delaloy et al 2003, O'Reilly et al 2003]. A neuronal isoform of WNK1 (termed WNK1/HSN2) that is highly expressed in the cell body of sensory ganglia neurons and neuronal projections has been identified. Pathogenic variants in an alternatively spliced exon of the transcript that encodes this isoform cause autosomal recessive hereditary sensory and autonomic neuropathy type 2, a disorder of progressive sensory deficit to touch, temperature, and pain [Shekarabi et al 2008].
The L-Wnk1 isoform can activate the kinase SPAK (encoded by STK39), which in turn can phosphorylate and activate NCC (the Na-Cl cotransporter) encoded by SLC12A3. The KS-WNK1 isoform, expressed predominantly along the distal nephron, binds to and inhibits the activity of L-Wnk1 in Xenopus [Subramanya et al 2006] and transgenic mice [Liu et al 2011]. WNK1 can also interact, through a conserved "HQ" domain, with other WNK kinases, including WNK4. These interactions have complex functional effects [Yang et al 2003, Chávez-Canales et al 2014].
As a multifunctional regulator of ion channels and transporters, WNK1 also inhibits the K+ channel ROMK1, encoded by KCNJ1 [Liu et al 2009] and (probably via phosphorylation of downstream kinase SGK-1) activates the amiloride-sensitive Na+ channel ENaC encoded by SCNN1A, SCNN1B, and SCNN1G [Xu et al 2005, Hadchouel et al 2010].
Abnormal gene product. Both of the known WNK1 deletions causing PHAII occur within the first intron of the gene and do not affect the amino acid structure of the gene product(s). It was initially shown that deletion within the first intron increases WNK1 transcription in peripheral leukocytes [Wilson et al 2001]. Subsequent work demonstrated that the intron deletion leads to increased L-WNK1 expression in the distal convoluted tubule [Vidal-Petiot et al 2013]. These findings support the hypothesis that the deletion within the first intron leads to increased expression of L-WNK1, which should phosphorylate and stimulate NCC.
A knockout mouse model of L-WNK1 has also been generated, and mice with a heterozygous targeted disruption of the L-Wnk1 transcript have significantly decreased blood pressure compared to wild type [Zambrowicz et al 2003]. Mice with targeted disruption of KS-Wnk1 exhibited increased activity of NCC, altered function of the ROMK (encoded by KCNJ1) and BKCa potassium channels, and decreased ENaC expression [Hadchouel et al 2010], confirming previous in vitro observations.
WNK4
Gene structure.
WNK4 has 19 exons. For a detailed summary of gene and protein information, see Table A, Gene.
Pathogenic variants. All reported pathogenic variants are missense alterations. All pathogenic variants reported to date have been identified in exons 7 and 17 [Wilson et al 2001, Golbang et al 2005, Brooks et al 2012].
Table 3.
Selected WNK4 Pathogenic Variants
View in own window
| DNA Nucleotide Change | Predicted Protein Change | Reference Sequences |
|---|
| c.1679A>G | p.Glu560Gly |
NM_032387.4
NP_115763.2
|
| c.1682C>T | p.Pro561Leu |
| c.1684G>A | p.Glu562Lys |
| c.1690G>C | p.Asp564His |
| c.1691A>C | p.Asp564Ala |
| c.1693C>G | p.Gln565Glu |
| c.3505A>G | p.Lys1169Glu |
| c.3553C>T | p.Arg1185Cys |
Variants listed in the table have been provided by the authors. GeneReviews staff have not independently verified the classification of variants.
GeneReviews follows the standard naming conventions of the Human Genome Variation Society (varnomen.hgvs.org). See Quick Reference for an explanation of nomenclature.
Normal gene product.
WNK4 product is a 1,243-amino acid protein encoded by a 3,732-nucleotide open reading frame within a 4-kb cDNA transcript. Northern blot analysis showed expression primarily within the kidney, but immunofluorescence studies have shown the protein is present in the epithelial lining of a variety of tissues including the colon, liver, and pancreas [Kahle et al 2004a].
WNK4 has been shown to regulate the activity of a number of ion transporters through heterologous expression in Xenopus oocytes and in mammalian cell systems [Kahle et al 2008]. The major effect of WNK4 in vivo appears to be stimulation of NCC; it may also inhibit ROMK [Kahle et al 2003, Wilson et al 2003]. WNK4 has also been reported to increase paracellular chloride permeability in mammalian kidney cells via claudin phosphorylation [Kahle et al 2004b, Yamauchi et al 2004, Tatum et al 2007]. Like WNK1, WNK4 also regulates other structurally diverse but functionally related ion channels including ENaC [Ring et al 2007a, Ring et al 2007b] and the cation nonselective TRP channels TRPV4 and TRPV5 [Fu et al 2006]. Deletion of WNK4 results in a phenotype that resembles Gitelman syndrome, indicating that the predominant WNK4 effect in vivo is stimulation of NCC [Castañeda-Bueno et al 2012, Takahashi et al 2014].
Abnormal gene product. The effects of PHAII-associated WNK4 pathogenic variants on the above targets have been evaluated in vitro in oocytes and mammalian cells and in vivo in mouse models [reviewed in Chávez-Canales et al 2014].
Important information about the mechanisms underlying WNK4-related PHAII came from the development and characterization of mouse models of PHAII [Lalioti et al 2006, Yang et al 2007]. Mice transgenic for a chromosomal segment encoding the murine Wnk4 with a Gln562Glu pathogenic variant (orthologous to the human p.Gln565Glu pathogenic variant) had hyperkalemia, higher blood pressure, and hypercalciuria compared to mice transgenic for a chromosomal segment encoding wild type Wnk4 [Lalioti et al 2006]. In addition, marked hyperplasia of the distal convoluted tubule (DCT) and increased expression of NCC were noted in mutated Wnk4 transgenic mice but not in wild type Wnk4 transgenic mice. All abnormalities were entirely corrected when mutated Wnk4 transgenic mice were crossed with mice harboring a targeted disruption of the gene encoding NCC, indicating that the effect of mutated Wnk4 on NCC activity alone is sufficient to cause the PHAII phenotype. Essentially similar findings were reported in a mouse Wnk4 mutated knock-in model of PHAII [Yang et al 2007].
Most PHAII-causing variants in WNK4 cluster within a highly conserved noncatalytic domain just distal to the kinase domain. Recent in vitro studies have demonstrated that this segment is critical for binding to KLHL3, and PHAII-associated variants within this domain disrupt interactions between WNK4 and KLHL3 [Ohta et al 2013, Shibata et al 2013, Wakabayashi et al 2013]. Due to the role of KLHL3 in ubiquitin-mediated proteolysis (see KLHL3), this should lead to increased WNK4 levels. Indeed, increased WNK4 levels are reported in the WNK4 mouse models of PHAII discussed above [Shibata et al 2013, Wakabayashi et al 2013].
It has also been suggested that pathogenic variants within this segment disrupt a calcium-sensing mechanism important in the regulation of WNK4 kinase activity [Na et al 2012]. Similarly, the PHAII-associated p.Arg1185Cys pathogenic variant (located near the C-terminus of WNK4 and separate from this domain) has also been implicated in impaired calcium sensing and altered phosphorylation by SGK1 [Na et al 2013].
