Abstract
ABSTRACT. WNK1 is a member of a novel serine/threonine kinase family, With-No-K, (lysine). Intronic deletions in the encoding gene cause Gordon syndrome, an autosomal dominant, hypertensive, hyperkalemic disorder particularly responsive to thiazide diuretics, a first-line treatment in essential hypertension. To elucidate the novel WNK1 BP control pathway active in distal nephron, WNK1 expression in mouse was studied. It was found that WNK1 is highly expressed in testis > heart, lung, kidney, placenta > skeletal muscle, brain, and widely at low levels. Several WNK1 transcript classes are demonstrated, showing tissue-, developmental-, and nephron-segment–specific expression. Importantly, in kidney, the most prominent transcripts are smaller than elsewhere, having the first four exons replaced by an alternative 5′-exon, deleting the kinase domain, and showing strong distal nephron expression, whereas larger transcripts show low-level widespread distribution. Alternative splicing of exons 11 and 12 is prominent—for example, transcripts containing exon 11 are abundant in neural tissues, testis, and secondary renal transcripts but are predominantly absent in placenta. The transcriptional diversity generated by these events would produce proteins greatly differing in both structure and function. These findings help further define and clarify the role of WNK1 and the thiazide-responsive pathway relevant to essential hypertension in which it participates. E-mail: Roger.Brown@ed.ac.uk
Gordon syndrome (also known as pseudohypoaldosteronism type 2, PHA 2; Online Mendelian Inheritance in Man 145260) is a familial form of hypertension with an autosomal dominant mode of inheritance (1). Patients have suppressed plasma renin activity and present with symptoms of severe hypertension (attributed to increased renal Na+ reabsorption), hyperkalemia (despite normal glomerular filtration), and metabolic acidosis as a result of reduced renal K+ and H+ excretion, respectively (2). These features are chloride dependent and are particularly well ameliorated by administration of thiazide diuretics (2–4⇓⇓). In common with other chloride transport–related disorders affecting BP, such as Bartter syndrome (5,6⇓), it seemed possible that the etiology lay in a direct mutation in a chloride channel subunit or transporter (7).
However, mutations in human WNK1 and WNK4 genes have recently been associated with Gordon syndrome (8). The encoded proteins of these genes are described as members of a novel family of serine/threonine kinases known as WNK (With No K, lysine) because of the atypical positioning of a conserved lysine residue within the catalytic domain (9). How they relate to ion transport abnormalities is of key importance as this seems to represent a novel BP regulatory pathway (10–12⇓⇓). Moreover, intriguingly, cases attributed to WNK1 mutations in this autosomal dominant disorder are due to intronic deletions and are not recognized as directly affecting WNK1 coding sequence (8).
It is of great interest how such WNK1 mutations, in a widely expressed gene, cause an autosomal dominant disease, with a mechanism seemingly explicable by a distal nephron–limited ion transport defect (2,13⇓). One possibility sees the intronic region involved in transcriptional regulation, and some preliminary evidence maintains this as a reasonable explanation. Alternatively, this apparently “silent” mutation may alter the complement of spliced products transcribed from the gene. Such silent mutations have been shown to cause dominant diseases by altering splicing, e.g., in the fibrillin-1 gene, causing Marfan syndrome (14). Thus, evidence relating to promoter use and alternative splicing of WNK1, especially relating to kidney, is of interest.
The study presented here describes, in some detail, several isoforms produced from the WNK1 gene, some of which show strikingly different tissue-specific distributions, including one showing abundant expression in kidney that is seen at low level, if at all, in other tissues. Such modifications, generating this transcriptional diversity, also produce predicted proteins of very different structure and proposed function. The findings reported thus help elucidate not only probable mechanisms by which the WNK1 intronic deletions cause disease, but also further clarify the novel WNK1 BP control pathway.
Materials and Methods
Northern Blot Analysis
Total RNA was extracted in TRIzol Reagent (Invitrogen, Inchinnan, UK) and separated (15 μg/lane) by denaturing (formamide/formaldehyde) agarose electrophoresis (0.8%), and blotted onto Hybond-N+ membrane (Amersham Biosciences, Little Chalfont, UK). Blots were hybridized with PCR-amplified mouse WNK1 (mWNK1) DNA fragments, radiolabeled with 32P-dCTP (Rediprime II Random Prime Labeling System, Amersham Biosciences, UK), at 65°C overnight in hybridization buffer (0.5 M Na2PO4·NaH2PO4 pH 7.2/7% SDS with 100 μg/ml denatured salmon sperm and 20 μg/ml yeast tRNA). Washes were at 65°C with 50 mM Na2PO4·NaH2PO4 pH 7.2/1% SDS. Blots were exposed to Kodak x-ray film (BioMax MS-1; Sigma, Poole, UK) at −70°C. cDNA templates for Northern blot probe production were PCR amplified with primers as detailed in Table 1 and Figures 1 and 2⇓.
Table 1. Primer sequences
Figure 1. Schematic representation of WNK1 cDNA structure showing primer positions. (a and b) cDNA structure of WNK1 and kidney-specific WNK1 containing exon 4A, respectively, with vertical bars representing splice junctions and dashed vertical bars indicating 3′ polyadenylation sites A1 and A2. Numbers within rectangles refer to exon numbers. Primers are represented by arrows and are numbered P1 to P28. Additional primer details are given in Table 1. Horizontal dashed lines specify the cDNA region amplified by PCR with the corresponding primer pairs. (c) cDNA sequence of exon 4A (accession no. AY319934). The predicted amino acid sequence also shown would continue in frame across the splice site to exon 5.
Figure 2. Northern blot analysis of WNK1 expression. Hybridization with a probe incorporating exons 6 to 9 (a) detects widespread expression of large transcripts approximately 10.2 kb in size. In addition, smaller and more prominent transcripts are detected in kidney. This expression profile is also observed in (b) when hybridizing with the poly-A probe (refers to sequence located between the two 3′ polyadenylation signals A1 and A2). (c) Probing with exon 1 fails to detect the smaller kidney-specific transcripts. (d) Northern blot analysis reveals differential expression of individual exons in WNK1 isoforms in kidney (K) and testis (T). Hybridization with probes against exons 2, 3, 4, 5, and 6 detect the large WNK1 isoform, strongly in testis but weakly in kidney. The smaller and more abundant kidney-specific isoform is only detected by probes to exons 5 and 6. A probe to an alternative exon, 4A, detects the smaller kidney-specific isoform only. Hybridization with a probe to exon 11 detects the large WNK1 isoform in testis but signal is low for either isoform in kidney. In contrast, hybridization with exon 12 readily detects both the kidney-specific isoform and the large WNK1 testis isoform. WNK1 expression was studied across a range of adult mouse tissues and in placenta (E16.5). Probe templates were PCR amplified with the following primer pairs: exons 6 to 9, P5/P6 (860 bp); poly-A, P3/P4 (875 bp); exon 1, P1/P2 (928 bp); exon 2, P15/P16 (180 bp); exon 3, P17/P18 (187 bp); exon 4, P19/P20 (161 bp); exon 4A, P9/P10 (340 bp); exon 5, P21/P22 (96 bp); exon 6, P23/P24 (212 bp); exon 11, P25/P26 (462 bp); and exon 12, P27/P28 (278 bp). For primer positions and sequences, refer to Figure 1 and Table 1, respectively.
RT-PCR
Promega Reverse Transcription System (random primed) (Promega, Southampton, UK) generated PCR templates (5 μl, diluted 1:20), used in standard PCR reactions, denaturing at 95°C for 3 min, then incubating on ice, and adding 15 pmol each primer, 250 μmol dNTP, and 2 U Taq DNA Polymerase (Promega) to a final volume of 50 μl in 1× PCR buffer. The PCR program used was as follows: 3 min at 95°C, then 35 to 40 cycles of 60 s at 95°C, 60 s at 60°C, 120 s at 72°C, and finally 10 min at 72°C. For negative controls, template was replaced with water. Products were visualized by agarose gel electrophoresis and purified with QIAquick PCR Purification Kit (Qiagen, Crawley, UK).
Production of Single-Stranded RNA Probes
RNA probes for in situ hybridization (ISH) analysis, produced as described previously (15) to specific regions of mWNK1, used a nested PCR method with primers, including 5′ extensions containing phage polymerase consensus sites, with sense and antisense primer pairs incorporating T3 (TCTAGATTAACCCTCACTAAAGGGA) and T7 (GGATCCTAATACGACTCACTATAGGG) sites, respectively. The PCR program used was as follows: 5 cycles of 45 s at 95°C, 45 s at 55°C, 120 s at 72°C, followed by 30 cycles of 45 s at 95°C, 45 s at 69°C, and 120 s at 72°C, and finally 10 min at 72°C. The required DNA-dependent RNA phage polymerase (T3-sense, T7-antisense) was then used on these purified PCR products to produce single-stranded 35S-UTP–labeled RNA probes of the corresponding inserts for ISH, as described elsewhere (15).
ISH Analysis
ISH, as described previously (15,16⇓), used cryostat sections (10 μm) cut from adult mouse kidney and mouse embryo samples (gestation day 16.5) mounted on silane-coated glass microscopic slides. Slides were fixed in 4% paraformaldehyde, incubated with prehybridization buffer at 50°C, hybridized with denatured 35S-UTP–labeled RNA probes (at a final concentration of approximately 10×106 cpm/ml in hybridization buffer at 50°C for approximately 16 h), followed by washes in 2× SSC, treatment with RNase A, and further washes to a maximum stringency of 0.1× SSC at 60°C. After ethanol series dehydration, slides were exposed to Kodak x-ray film (BioMax MR-1; Sigma). Slides were then emulsion dipped and exposed in a light-tight box for 3 to 5 wk before being developed.
Results
The mouse WNK1 gene (mWNK1) is large, spanning >100 kb, with a coding region showing 86% identity with human WNK1 (hWNK1). mWNK1 produces large transcripts (>10 kb) with cDNA encoded by 28 exons (Figure 1a), having a predicted 2377 amino acid protein. Northern blot hybridization with exons 6 to 9 revealed widespread distribution of these large transcripts (Figure 2a) with high expression seen in testis > heart, lung, kidney, placenta > skeletal muscle, brain, and low-level expression elsewhere. Additional Northern blots show placental transcripts are of similar size to the major widely found isoform. In kidney, a smaller, more abundant mWNK1 transcript is seen differing by at least 1.5 to 2 kb (Figure 2a). A 3′-polyadenylation (poly-A) site A1 (Figure 1) is positioned approximately 800 bp downstream of the open reading frame (ORF). To assess whether the use of an alternative poly-A site could account for this size difference, we isolated cDNA sequence encoding a second poly-A site approximately 2400 bp downstream of the ORF (Figure 1, A2). Northern blot hybridization with cDNA between these poly-A sites revealed a similar expression profile to that described above (Figure 2b). Both transcripts in kidney were detected implying most WNK1 transcripts terminate at the second 3′ poly-A site (Figure 1, A2). Having excluded this possibility, an exon 1 probe, overlapping the ORF (Figure 1), was designed against the 5′ region. Northern blot hybridization with exon 1 revealed a similar expression profile for the large mWNK1 transcript, but the smaller kidney-specific transcript remained completely undetected, indicating it lacks exon 1 (Figure 2c).
To investigate whether this kidney-specific isoform lacked further 5′ exons, nested PCR was used to produce probes against exons 2, 3, 4, 5, and 6 (Figure 1), ensuring amplification of mWNK1 cDNA only (confirmed by sequencing) across this highly conserved kinase domain region. Northern blot analysis of kidney and testis RNA showed probes to exons 2, 3, and 4 only detect the large mWNK1 transcript and not the smaller kidney-specific isoform (Figure 2d). In contrast, exon 5 and 6 probes detect both mWNK1 transcripts. Thus, the smaller mWNK1 transcript in kidney lacks exons 1 to 4. Screening submitted expressed sequence tag (EST) databases suggested an alternative exon preceding exon 5 in kidney, positioned between exons 4 and 5 in the genomic sequence and therefore termed exon 4A. Northern blot hybridization with this exon only detects the smaller mWNK1 transcript (Figure 2d), confirming inclusion of exon 4A in the kidney-specific isoform (Figure 1b).
ISH to exons 6 to 9 allowed detailed study of mWNK1 expression in mouse kidney. Figure 3 reveals clear WNK1 expression above background in distal nephron extending from early distal convoluted tubule (DCT) into connecting tubule and at lower level into cortical collecting duct (lower expression in medullary rays; Figure 3). Strong expression in DCT continues adjacent to glomeruli and is seen looping close to their vascular pole (Figure 3 right panels) the site of the macula densa. Additionally, lower expression, above background, is distributed more extensively. This appears to have a different origin from the higher expression seen in DCT (i.e., different WNK1 transcripts; see below).
Figure 3. In situ hybridization (ISH) of WNK1 in mouse kidney. The ISH study used a probe to WNK1 exons 6 to 9. Images are of emulsion-dipped slides (eosin/cresyl violet counterstain; original magnification: left, ×40; right, ×200). View of renal cortex with a medullary ray (MR) demarcated between the parallel lines and cortical labyrinth lateral to them. Below the dashed line is the outer medulla. Bright field on top (black, strong expression), corresponding dark field view on bottom (white, expression). The very highest expression—showing black on bright field—obscures light, and hence the strongest expressing tubules appear as black holes with ring outlines on dark field view (bottom left). Short black arrows indicate glomeruli; a, examples of distal convoluted tubules (DCT) largely adjacent to glomeruli; b, example of connecting tubules (CNT) in midcortical labyrinth arcades (adjacent to radial vessels); c, cortical collecting duct (CCD) in medullary rays; and GA, glomerular arteriole. (left) Study with long exposure (5 wk) demonstrating wide range of expression level. Note regional expression at 3 levels: (1) widespread low level (low level white seen in darkfield), (2) higher in CCD (faint but distinct tubular outline seen on dark field, pale gray on bright field); and (3) highest in DCT and CNT (darker gray-black on bright field and bright ring with black hole center when silver grains confluent). (right) Three-week exposure: note the proximity of strongly expressing DCT to glomerulus and its vascular pole (between arterioles) and stronger expression than in collecting duct.
To investigate the distribution of the two kidney isoforms, adult kidney sections were hybridized with exons 1, 4A, and 6 to 9 (Figure 4, top left panel). Probing with exon 1 detects widespread signal at a low level (Figure 4a), whereas with exon 4A, only strong punctate signal in cortex is detected (Figure 4c). The expression pattern seen with exons 6 to 9 is thus a combination of that seen with exons 1 and 4A (Figure 4e). These probes were also hybridized to embryo tissue slices (E16.5) to examine developmental mWNK1 expression. Probing with exons 6 to 9 revealed a wide mWNK1 distribution, with high expression in tissues—for example, placenta, nasal epithelium, lung, intestine, regions of the brain, and developing renal cortex (Figure 4f). As expected, the expression patterns revealed by probing with exons 1 and 4A were subtypes of that seen for exons 6 to 9, with exon 1 widely expressed, particularly in placenta, lung, kidney, intestine, thymus, and forebrain (Figure 4b), whereas exon 4A revealed much lower expression, with high signal detected in restricted regions (e.g., nasal epithelium, forebrain, thymus, and kidney; Figure 4d). Sense control sections showed no specific hybridization (Figure 4).
Figure 4. Analysis of WNK1 expression by in situ hybridization (ISH) of mouse adult kidney and fetal sections (E16.5). Cryostat sections were subjected to ISH analysis; a range of probes against WNK1 were used. Left columns of top panels show local distribution within the kidney (a, c, e, g, i, and k). Right columns of top panels show corresponding developmental expression patterns (b, d, f, h, j, and l). Bottom panel shows sense controls for both kidney (m and n) and fetal (o and p) sections. Probes were constructed against exon 1 (434 bp), exon 4A (280 bp), exons 6 to 9 (552 bp), exon 11 (460 bp), exon 12 (283 bp), and exon 4B (108 bp). Exposure times for detection of ISH signal were as follows: exon 1, (a and b) 1 d; exon 4A, (c) 1 d, (d) 4 d, (n) 1 wk; exon 6 to 9, (e and f) 1 d, (m) 3 d; exon 11, (g, h, and o) 3 d; exon 12, (i and j) 3 d; exon 4B, (k, l, and p), 2.5 wk. For details of probe lengths and production, see Materials and Methods. B indicates brain; K, kidney; L, lung; and P, placenta.
EST sequences suggest that alternative splicing of mWNK1, primarily concerning exons 11 and 12, takes place in some tissues (Figure 1). To investigate this, we subjected kidney and testis RNA to Northern blot analysis with probes specific for each exon. Hybridization with exon 11 detects the large mWNK1 isoform in testis, but signal is greatly diminished for either isoform in kidney (Figure 2d), suggesting exon 11 is usually spliced out in both kidney mWNK1 transcript classes. In contrast, hybridization with exon 12 shows strong signal for mWNK1 in testis and for the smaller kidney-specific transcript. The large mWNK1 transcript in kidney is also detected at a lower level (Figure 2d). To investigate tissue-specific and developmental variations in these splicing events, we subjected adult mouse kidney and fetal (E16.5) sections to ISH analysis. The expression patterns seen when probing adult kidney sections with either exon are similar to that described above for exons 6 to 9, showing low-level, widespread expression throughout the kidney, overlaid with strong punctate cortical expression (Figure 4, g and i). However, developmental expression studies that use these probes indicate that although many fetal tissues express both exons similarly, striking tissue-specific splicing occurs in some developing organs. For example, transcripts containing exon 11 are abundant in some neural tissues but are rare or absent in placenta (Figure 4, h and j). Sense-control sections showed no specific hybridization (Figure 4).
Reverse transcriptase–PCR (RT-PCR) studies across exons 11 and 12 confirm these splicing events (Figure 5a). Amplification in kidney by means of primers spanning the exon 7/8 splice site down to the exon 15/16 splice site shows two major bands, corresponding in size to PCR products having either exon 11, or both exons 11 and 12 spliced out (Figure 5b). Additional RT-PCR studies that use alternative primers spanning this region also show major bands, similarly representative of these splice variants. However, it is difficult to interpret the importance of additional large weak products occasionally amplified by RT-PCR.
Figure 5. Detection of WNK1 alternative splicing. (a) Ethidium bromide–stained agarose gel, showing reverse transcriptase–PCR–amplified fragments in duplicate. These products were amplified with mWNK1 primer pair P7 and P8, which span a region from the exon 7/8 splice site down to the exon 15/16 splice site (lanes 3 and 4). A negative control is seen in lane 2. Lane 1 shows a 100-bp DNA ladder. (b) Schematic depiction of WNK1 alternative splicing involving exons 11 and 12. The major products seen correspond to (b) (iii) and (b) (iv). (c) Schematic representation of the major predicted WNK1-derived proteins. mWNK1 is 2377 amino acids in length and is particularly rich in serine, glutamine, and proline, having 26 PXXP sites potentially recognized by SH3 domains. Black bars denote four putative coiled coil domains, and a conserved WNK autoinhibitory domain is represented by horizontal stripes. Black arrowheads indicate the positions of all potential phosphorylation sites; intriguingly none of which overlap with the region encoded by exons 11 and 12. Exon 11 encodes a leucine zipper (LXXLL) motif. Kinase-deficient (KDP) WNK1 has a truncated N-terminus, lacking one coiled coil domain and deleting a major portion of the kinase domain. This region is substituted with a highly cysteine-rich stretch of 30 amino acids.
EST sequences also suggest that in addition to exons 4 and 4A, a third exon positioned between exons 4A and 5 in the mouse genomic sequence and therefore termed 4B, may precede exon 5 in some mWNK1 transcripts (accession no. AK052468). The EST evidence suggests that unlike the kidney-specific transcripts described above containing exon 4A, which lack all known upstream exons, transcripts containing 4B splice directly from exon 4 to 4B to 5. ISH analysis probing with exon 4B shows a similar pattern to that seen previously in kidney, with high cortical expression overlaying a low widespread distribution (Figure 4k). However, splicing events producing transcripts containing exon 4B appear to be rare, judged by the lower ISH signal (requiring severalfold longer for clear detection). Furthermore, signal detected on fetal sections was low and widespread, lacking the striking tissue-specific differences in expression levels seen with probes to other exons (Figure 4l), and was only marginally higher than that seen for sense controls (Figure 4p).
As described above, Northern blot studies examining exon 11 expression detect very weak signal for both isoforms in kidney, despite high expression in testis. In contrast, ISH studies suggest exon 11 is expressed at levels comparable with exon 12 in kidney. This discrepancy led to further Northern blot analysis looking for evidence of further novel mWNK1 kidney transcripts containing exon 11. These studies revealed at least two novel mWNK1 transcripts in kidney and testis, evidently several kilobases smaller than the two isoforms described above (Figure 6). Intriguingly, transcripts of similar size were detected in additional Northern blot analysis of exon 4A expression in both tissues (Figure 6).
Figure 6. Detection of smaller WNK1 transcripts in kidney and testis by Northern blot (short exposure). Probes to exon 4A detect the kidney-specific WNK1 transcript (B), but also reveal two additional transcripts several kilobases smaller (C and D) in both kidney (K) and testis (T). These transcripts are also seen when probing with exon 11, which also detects the large WNK1 transcript (A) as expected.
Discussion
In this study, we show multiple WNK1 mRNA species are expressed in both adult mouse and during development, with some showing striking tissue-specific expression differences. Large transcripts—greater than 10 kb in size—were seen by Northern blot in virtually all tissues examined, with expression highest in testis > heart, lung, kidney, placenta > skeletal muscle, brain. Transcripts detected in testis appeared larger than transcripts common to most other tissues, attributed to the inclusion of exon 11 (462 bp). The major transcripts seen in kidney, however, were smaller than elsewhere. This phenomenon has been reported in previous studies (8) and appears to be conserved between species. We have demonstrated that the difference between these transcripts is not the result of the use of an alternative 3′ polyadenylation site, as previously suggested (8,9,17,18⇓⇓⇓). Instead, we show that the first four exons in the smaller, more prominent kidney transcripts are replaced by an alternative exon, exon 4A (19). The sequence of this novel exon and the resulting predicted amino acid sequence are presented for the first time (Figure 1c). ISH analysis revealed strong punctate expression of these smaller transcripts restricted to cortex, localizing to distal nephron, whereas the large transcripts showed uniform low-level expression throughout the kidney. These expression patterns are combined in that seen for exons 6 to 9, found in both transcript classes. In addition to exons 4 and 4A, another novel exon, exon 4B, was found to precede exon 5 in some WNK1 kidney transcripts; however, weak ISH signal suggests that this is a rare splicing event.
Use of exon-specific primers identified two relatively abundant, alternatively spliced mWNK1 mRNAs in kidney, corresponding in size to transcripts lacking either exon 11 or both exon 11 and 12. Alternative splicing of these exons has also been reported for hWNK1 (17). Furthermore, the published rat WNK1 sequence (accession no. NM_053794) corresponds to the splice variant lacking both exons. Exon-specific ISH detected expression in kidney, with probes to both exons showing a similar pattern to that described above for exons 6 to 9. In contrast, Northern blot analysis showed that although exon 12 is usually included in the small abundant kidney isoform, exon 11 is largely absent in both kidney isoforms. Further studies indicate the production of substantial levels of novel smaller alternatively spliced transcripts from the WNK1 gene, with Northern blot analysis implicating the inclusion of both exons 11 and 4A. These smaller transcripts are seen in kidney but also in testis, a tissue in which exon 4A expression was previously unsuspected because testis lacks obvious expression of the much larger “kidney-specific” 4A transcript. Preliminary investigations suggest that these transcripts differ from this largely kidney-specific 4A transcript, described above, being much smaller in size. However, RT-PCR in testis easily amplifies a product from exon 4A to exon 11, suggesting these smaller transcripts contain this region.
This study was the first to assess WNK1 expression in development by means of ISH analysis, demonstrating that WNK1 is widely expressed in the mouse embryo (E16.5), in both epithelial tissues (e.g., developing renal cortex, intestine, lung, nasal epithelia, and placenta) and in nonepithelial tissues (e.g., regions of the CNS). Again, there are tissue-specific differences in the pattern of transcripts expressed in developing organs, with transcripts containing exon 1 showing high widespread but nonuniform expression and exon 4A–containing transcripts showing high expression only in restricted sites. Exon-specific ISH also showed striking tissue-specific differences in expression of exons 11 and 12—for example, transcripts containing exon 11 are abundant in some neural tissues but are very low or absent in placenta.
It is appropriate to suggest that WNK1 regulates ion transport through the kinase domain. However, the major WNK1 transcript in kidney lacks this functional entity as a result of the replacement of exons 1 to 4 with exon 4A (Figure 5c). Additionally, a coiled coil motif predicted just N-terminal to the kinase domain is lost, possibly disrupting interactions with molecular targets. This implies that the remainder of the WNK1 protein must contribute functionally to the WNK1 regulatory pathway. Exons 11 and 12 are of key interest because splicing of this region is conserved between species, is clearly tissue specific, and would produce a repertoire of proteins likely to be coexpressed in the same tissues and probably the same cells (Figure 4). The predicted amino acid sequence encoded by these exons is proline rich (approximately 15%), and analysis suggests a potential transmembrane span flanked by a flexible conformation. This region also shows homology to a number of extracellular matrix proteins (e.g., mucins, glycosaminoglycans, and sialoproteins). Clearly, this is of interest in the context of proteins that may play a role in tight junction regulation. Conversely, when this stretch of amino acids is removed, these features disappear. The resulting juxtaposed sequence is predicted to form an exposed loop region with homology to proteins that tend to bind ligands and/or act as transcription factors. Intriguingly, although WNK1 contains numerous potential phosphorylation sites, no such site occurs within the exon 11 to 12 region, implying that signaling through protein kinase pathways acting on WNK1 is not affected by such splicing events. Clearly, experimental studies are required to fully investigate these possibilities.
Also of interest is the addition of thirty amino acids to the N-terminus of kinase-deficient (KDP) WNK1, contributed by exon 4A. This exon may have no major functional effect other than deleting the kinase domain. However, this sequence is strikingly cysteine rich. Within a cluster of six likely very reactive cysteine residues, at least three have a high predictive index for forming either disulfide bonds or bonds with other molecules (e.g., metal-containing moieties). Moreover, the N-terminal positioning of this cysteine cluster may promote this region as a potential point of anchorage to other structures. Furthermore, the novel small bands seen by Northern blot clearly indicate the production of WNK1 transcripts lacking several kilobases compared with the kidney-specific 4A band. Therefore, it is very likely that additional changes in WNK1 protein structure exist that have profound effects on WNK1 function.
The transcriptional modifications described above would greatly influence the potential complement of proteins produced from the WNK1 gene and may have evolved to regulate WNK1 function. KDP-WNK1 proteins may act to inhibit other WNK1 proteins, having “active” kinase domains, via interactions through the remaining coiled-coil motifs (Figure 5c). This is supported by the recent report of a WNK1 autoinhibitory domain, positioned between residues 515 to 569 in the rat sequence, which is conserved between species and also within the WNK family (Figure 5c). Preliminary evidence was also reported for WNK1 tetramer formation via the coiled-coil motif C-terminal to the autoinhibitory domain (Figure 5c) (20).
This work provides a number of insights into the cause of Gordon syndrome. We have examined WNK1 expression in some detail, further elucidating kidney-specific and distal-nephron–specific WNK1 transcripts. Our findings imply the use of alternative promoters, one initiating transcription in exon 1 and the second giving rise to transcripts having exon 4A in place of exons 1 to 4. It is therefore reasonable to suggest that cis elements within intron 1 affect the second promoter regulating such 4A transcripts. Dominant negative regulation would imply that intronic deletions causing Gordon syndrome lead to abnormally high expression of KDP-WNK1 transcripts in kidney, in turn causing excessive inhibition of “normal” WNK1 function. Alternatively, the intron 1 deletions could interrupt splicing enhancer or silencer sequences, thereby affecting the inclusion or exclusion of exons and disrupting the complement of alternatively spliced WNK1 transcripts. A similar effect is seen in Marfan syndrome, another autosomal dominant disorder affecting connective tissue, where a silent mutation results in exon skipping (14). In ataxia-telangiectasia, an intronic deletion in the ATM gene results in aberrant inclusion of a cryptic exon (21).
The pathway disrupted in Gordon syndrome involving WNK1 is regarded as different from other signaling pathways known to regulate BP. The findings presented here reveal the central importance of the transcriptional control of this gene in generating a complement of kidney-specific, WNK1-derived proteins. The correct balance within this complement of proteins must mediate the effects on ion transport that constitute this novel BP regulatory pathway. Clearly the regulation of alternative promoter use and splicing is likely to participate in the control of BP by WNK1-derived proteins.
Acknowledgments
We thank Susan K. Coan for her technical assistance, and we thank the Wellcome Trust (grant 065616; PhD Studentship to MOR) and the British Heart Foundation (grant PG2001075) for their support.
- © 2003 American Society of Nephrology
References
Jump to section
More in this TOC Section
Cited By...
- Potassium-regulated distal tubule WNK bodies are kidney-specific WNK1 dependent
- Src-family protein tyrosine kinase phosphorylates WNK4 and modulates its inhibitory effect on KCNJ1 (ROMK)
- The WNK-SPAK/OSR1 pathway: Master regulator of cation-chloride cotransporters
- Activation of PI3-Kinase Stimulates Endocytosis of ROMK via Akt1/SGK1-Dependent Phosphorylation of WNK1
- WNK1 is required for mitosis and abscission
- WNK4 Kinase Stimulates Caveola-mediated Endocytosis of TRPV5 Amplifying the Dynamic Range of Regulation of the Channel by Protein Kinase C
- Src family protein tyrosine kinase (PTK) modulates the effect of SGK1 and WNK4 on ROMK channels
- Regulation of ROMK Channel and K+ Homeostasis by Kidney-specific WNK1 Kinase
- The regulation of salt transport and blood pressure by the WNK-SPAK/OSR1 signalling pathway
- WNK Kinases and Renal Sodium Transport in Health and Disease: An Integrated View
- Dietary Electrolyte-Driven Responses in the Renal WNK Kinase Pathway In Vivo
- WNK1 and OSR1 regulate the Na+, K+, 2Cl- cotransporter in HeLa cells
- WNK1 Affects Surface Expression of the ROMK Potassium Channel Independent of WNK4
- WNK1 kinase isoform switch regulates renal potassium excretion
- Antagonistic regulation of ROMK by long and kidney-specific WNK1 isoforms
- Familial Hyperkalemic Hypertension
- Association of Blood Pressure With Genetic Variation in WNK Kinases in a White European Population
- WNK1 activates SGK1 to regulate the epithelial sodium channel
- The kidney-specific WNK1 isoform is induced by aldosterone and stimulates epithelial sodium channel-mediated Na+ transport