University of Southern California Dissertations and Theses

Kinase activity of the pseudohypoaldosteronism type II gene product, WNK4

(USC Thesis Other)

Kinase activity of the pseudohypoaldosteronism type II gene product, WNK4

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content KINASE ACTIVITY OF THE PSEUDOHYPOALDOSTERONISM TYPE II
GENE PRODUCT, WNK4
by
Erik Robert Ahlstrom
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PHYSIOLOGY AND BIOPHYSICS)
May 2008
Copyright 2007 Erik Robert Ahlstrom
ii
Table of Contents
List of Figures iiv
Abstract vi
Chapter 1: Introduction 1
Pseudohypoaldosteronism type II (PHAII) 1
PHAII is linked to mutations in WNK kinases 2
Table 1. PHAII associated mutations 3
WNK4 and regulation of electrolyte flux over epithelial cell barriers 5
Surface expression of NCC, ROMK1 and ENaC is inhibited by
coexpression with WNK4 5
WNK4 reduces surface expression of various extrarenal transport
proteins 8
WNK4 increases paracellular chloride permeability of MDCKII
monolayers 9
Two mouse models reproduce the PHAII phenotype 10
Kinase activity of WNK4 13
WNK1 behaves like a serine-threonine kinase 14
Structural elements of WNK4 16
Studies of WNK4 kinase activity 17
Chapter 2: Aims of this project 20
Chapter 3: Materials and Methods 21
Reagents 21
WNK4 Expression Constructs 21
pCSPT-mWNK4-KXX 21
pNTAP-WNK4 22
Truncated pNTAP-WNK4 23
Protein Expression and Recovery 25
Protein Expression in Mammalian Cells HEK293, COS-7 and CHO 25
Recovery of c-myc-tagged WNK4 by Immunoprecipitation 25
Protein Recovery by Tandem Affinity Purification (TAP) 26
OSR1 Expression and purification from Escherichia Coli 27
Protein Detection by SDS-PAGE and Western Blot 28
In vitro kinase assay 28
Immunoprecipitated WNK4 28
iii
TAP-WNK4 29
In-gel kinase assay 29
Chapter 4: Results 31
Assay of kinase activity in WNK4 protein expressed and
immunoprecipitated from HEK293 cells 31
Immunoprecipitation of WNK4 expressed in HEK293 cells 31
In vitro phosphorylation of MBP by immunoprecipitated WNK4 33
Assay of kinase activity in WNK4 protein expressed in HEK293 cells
and recovered by tandem affinity purification 35
Tandem affinity purification of WNK4 expressed in HEK293 35
In vitro kinase assay of TAP-WNK4 with MBP and OSR1 as
substrates 38
In-gel kinase assay of TAP-WNK4 with OSR1 as substrate 39
TAP-WNK4 expressed and purified from COS-7 and CHO cells and
assayed for in-gel kinase activity 41
Identification of WNK4 associating proteins by peptide mass
fingerprinting 41
Studies of truncated WNK4 43
In-gel kinase activity of WNK4 with c-terminal truncations Δ593,
Δ500 and Δ443 43
Kinase activity of WNK4 Δ593 mutants 45
N-terminal WNK4 truncation 47
High salt and detergent protocol for recovery of full-length and Δ593
truncated TAP-WNK4 47
Chapter 5: Discussion 52
Full-length WNK4 has kinase activity 52
Kinase domain mutants abolish WNK4 kinase activity 53
WNK4 autophosphorylation cannot be detected 54
PHAII associated mutations do not affect WNK4 kinase activity 55
WNK4 associates with a 40 kD kinase when expressed in mammalian
cells 57
Chapter 6: Overall Conclusions 61
Chapter 7: Future Directions 62
References 64
iv
List of Figures
Figure 1. Chloride shunt hypothesis. 3
Figure 2. Studies of WNK4 coexpressed with various ion transporters in
Xenopus oocytes. 6
Figure 3. Studies of WNK4 expressed in MDCKII cells. 6
Figure 4. Mouse models of PHAII. 12
Figure 5. Schematic representations of WNK1 and WNK4. 15
Figure 6. Crystal stuctures of WNK1 and PKA. 15
Figure 7. Assay of kinase activity in WNK4 protein expressed and
immunoprecipitated from HEK293 cells. 32
Figure 8. Kinase assyas used to measure WNK4 substrate phosphorylation 34
Figure 9. Tandem affinity purification. 36
Figure 10. Assay of kinase activity in WNK4 protein expressed in HEK293
cells and recovered by tandem affinity purification. 37
Figure 11. In-gel phosphorylation of OSR1. 40
Figure 12. WNK4 expressed in different cell lines. 40
Figure 13. C-terminal truncations of TAP-WNK4. 44
Figure 14. C-terminally truncated WNK4 proteins expressed and assayed
for in-gel kinase activity. 44
v
Figure 15. Kinase activity of TAP-WNK4 Δ593 mutants. 46
Figure 16. In-gel kinase activity of n-terminally truncated TAP-WNK4. 48
Figure 17. High salt and detergent protocol for recovery of full-length and
Δ593 truncated TAP-WNK4. 48
Figure 18. In vitro kinase assay of full-length and Δ593 truncated
TAP-WNK4 mutants prepared by HSD protocol. 50
vi
Abstract
Mutations in WNK1 (with no K [lysine] 1) and WNK4 protein kinases cause
pseudohypoaldosteronism type II (PHAII), a rare genetic disorder that features high
blood pressure and elevated serum potassium levels. Potential targets of WNK4
regulation have been identified by various approaches, but the mechanism by which
it influences its targets is still poorly understood. In an effort to characterize its
kinase activity, we expressed and purified full-length and truncated forms of WNK4
from HEK293 cells. Due to endogenous kinases binding non-specifically to the
protein G resin when immunoprecipitating WNK4 from HEK293 cells, we decided
to recover the protein by the tandem affinity purification (TAP) method. Our
phosphorylation experiments identified a 40 kD kinase that associates specifically to
the C-terminal half of WNK4 and is able to phosphorylate WNK4 and the WNK4
substrate, oxidative stress response kinase 1 (OSR1). This kinase copurifies with
WNK4 in the mammalian cell lines HEK293, COS-7 and CHO, but could not be
identified by peptide mass fingerprinting. By modifying the TAP protocol to include
high salt and detergent washes, we were able to dissociate a large fraction of the 40
kD kinase and measure specific in vitro phosphorylation of OSR1 by full-length and
Δ593 truncated WNK4. The PHAII associated mutations E559K, D561A and Q562E
had no significant effect on this phosphorylation. This study contributes important
information on the mechanism by which WNK4 may influence its targets by
showing that full-length WNK4 has kinase activity and also what effects the PHAII
mutations have on this activity.
1
Chapter 1: Introduction
Nine out of ten people with high blood pressure are diagnosed with essential
hypertension, which means that the underlying cause is not known. The mechanisms
controlling blood pressure are complex and involve multiple molecular pathways,
which is why monogenic disorders associated with dysregulation of blood pressure
can serve as tools for identifying important components of these pathways.
Pseudohypoaldosteronism type II (PHAII)
PHAII, also known as Gordon’s syndrome, is a rare autosomal dominant disease
characterized by salt sensitive hypertension, hyperkalemia, renal tubular acidosis
(RTA, variable finding) and otherwise normal kidney function (6). Plasma renin
activity is usually low and circulating aldosterone levels are low-normal. The name
pseudohypoaldosteronism type II was suggested by Schambelan (22) after he
observed that the administration of exogenous mineralocorticoids failed to correct
the hyperkalemia, resembling the mineralocorticoid resistant state of PHA type I.
Early investigators proposed two major pathophysiological mechanisms involving
defects in electrolyte handling by the distal nephron to explain the clinical features of
PHAII, the first one involving overactivity of the thiazide sensitive sodium chloride
cotransporter (TSC or NCC) and the second one stemming from excessive
paracellular chloride permeability in the distal nephron, also known as the “chloride
shunt” hypothesis. Supporting the theory of an overactive NCC was the observation
2
that treatment with low doses of thiazide diuretics return blood pressure and
potassium levels to normal (6) and the fact that PHAII is the mirror image of
Gitelman’s syndrome, which is characterized by hypotension, hypokalemia and
metabolic alkalosis due to inactivating mutations in NCC. The chloride shunt
hypothesis was introduced by Schambelan to explain his observation that all the
clinical features of PHAII could be reversed by infusing sodium together with a non-
reabsorbable anion like sulfate or bicarbonate (22). According to the hypothesis
excessive reabsorption of sodium with chloride in the distal nephron results in
volume expansion (hypertension) and diminishes the lumen-negative potential that
drives potassium and proton excretion (hyperkalemia and RTA) (Figure 1). By
increasing the chloride permeability of the distal nephron epithelium, the lumen-
negative potential that is normally established by electrogenic sodium reabsorption
through the epithelial sodium channel (ENaC), is “short-circuited” by parallel
reabsorption of chloride.
PHAII is linked to mutations in WNK kinases
Genome wide linkage analysis of affected families resulted in PHAII associated loci
being mapped to chromosomes 1 (PHAIIA), 17 (PHAIIB) and 12 (PHAIIC), but no
candidate genes were identified in these studies (2, 15). Wilson et al performed a
linkage analysis in PHAII kindred K22 and were able to demonstrate linkage of
PHAII to the telomeric region of chromosome 12p in this kindred (26).
3
Figure 1. Chloride shunt hypothesis. Sodium reabsorption without chloride sets up
the lumen negative potential that drives potassium excretion in the cortical collecting
duct. Increased chloride permeability short-circuits this potential thereby inhibiting
potassium excretion.
Table 1. PHAII associated mutations.
4
Further analysis of this region revealed a large deletion found in all affected
members. This deletion was located within the large first intron of the WNK1 gene,
which encodes a recently identified serine-threonine kinase (29). Members of the
WNK (with no lysine(K)) family of kinases are unique among serine-threonine
kinases in that a highly conserved lysine in protein kinase subdomain II (7) is
replaced by a cysteine. Northern blot analysis showed the presence of WNK1
transcripts in a variety of human tissues, including a 10 kb transcript in the kidney
and a 12 kb transcript in heart and skeletal muscle (26). The PHAII associated
deletions in the WNK1 gene were shown to increase transcript levels in leukocytes,
consistent with a gain of function mutation. WNK1 protein expression in kidney was
localized to the distal part of the nephron by immunofluorescence, mainly distal
convoluted tubule (DCT) and cortical collecting duct (CCD). Within the cells WNK1
staining is predominantly cytoplasmic. In an effort to pin down the affected genes in
other PHAII associated loci mapped to different chromosomes, Wilson et al searched
genomic sequences and EST databases for paralogs of WNK1 and found that WNK4
on chromosome 17 lies within the minimum genetic interval containing the PHAIIB
locus identified by Mansfield (15). Closer examination of the WNK4 gene in PHAII
kindreds revealed missense mutations in affected members of four kindreds (26).
Three kindreds had charge changing mutations that were clustered in a highly
conserved, negatively charged, ten amino acid segment. The PHAII associated
mutations reported from Wilson’s study are summarized in Table 1. Screening RNA
from a variety of human tissues with a WNK4 probe, Wilson et al identified kidney
5
as the only site of expression. Immuno-localization of WNK4 within the kidney
showed staining in the distal part of the nephron, mainly in the DCT and CCD. Later
reports suggest that WNK4 transcript and protein is also present in the biliary and
pancreatic ducts, the endothelium of the blood brain barrier, epididymis (testis) and
colonic crypts (9). The subcellular localization of WNK4 is tight junctional with
some staining of the cytoplasm.
WNK4 and regulation of electrolyte flux over epithelial cell barriers
WNK4 immuno-localization in the distal nephron fits well with the proposed
pathophysiological mechanisms of PHAII, suggesting that this kinase may play a
role in the regulation of electrolyte handling in this segment. A long list of potential
targets of WNK4 regulation has emerged from in vitro studies where WNK4 has
been co-expressed with various transport proteins in Xenopus laevis oocytes (Figure
2) and by itself in madin-darby canine kidney type II cells (Figure 3).
Surface expression of NCC, ROMK1 and ENaC is inhibited by
coexpression with WNK4
The first obvious candidate transporter to be investigated in the oocyte system was
NCC. This protein is located in the apical membrane of DCT cells where it mediates
electroneutral sodium and chloride uptake. Two independent studies show that NCC
mediated sodium influx is markedly reduced by co-expressing wild type mouse
6
Figure 2. Studies of WNK4 coexpressed with
various ion transporters in Xenopus oocytes.
Figure 3. Studies of WNK4 expressed in MDCKII cells.
7
WNK4 (27, 32). WNK4 harboring the PHAII associated mutation Q562E was found
to lack this inhibitory capacity in both studies. Introducing the kinase domain
mutation D318A also abolished inhibition by WNK4 in Wilson’s study. Surprisingly
Yang et al found that PHAII mutations E559K and D561A had no effect on WNK4
inhibition of NCC. Employing different methods both groups showed that WNK4
exerted its effects on NCC by reducing its surface expression via a non-clathrin
mediated mechanism.
Renal outer medullary potassium channel (ROMK1) expression overlaps
with WNK4 in CCD where it mediates potassium excretion driven by the lumen
negative potential produced by electrogenic sodium reabsorption by ENaC in this
segment. In an effort to identify possible mechanisms by which WNK4 is directly
involved in distal nephron potassium handling, Kahle et al co-expressed ROMK1
with either wild type, Q562E or E559K mWNK4 in Xenopus oocytes (11). Wild type
WNK4 caused an inhibition of ROMK1 mediated potassium current and a
proportional decrease in EGFP-ROMK surface expression. Both PHAII mutations
augmented this inhibition while kinase domain mutant D318A had no effect on
WNK4 inhibition of ROMK1. Further studies showed that the mechanism of surface
retrieval is clathrin mediated and that the endocytic scaffold protein intersectin is
involved in the process (8, 11).
The in vitro data with NCC and ROMK1 suggest that PHAII could result
from overactivity of the former and impairment of the latter, resulting in salt
retention followed by volume expansion and hyperkalemia. Interestingly the PHAII
8
mutations appear to have very different effects on these two regulatory processes in
that they relieve the inhibition of NCC in a kinase dependent manner and at the same
time increase the inhibition of ROMK1 in a kinase independent manner.
ENaC expression overlaps with WNK4 and ROMK1 in the CCD, making it a
candidate for WNK4 regulation. When coexpressed with WNK4 in the oocyte
system, surface expression of ENaC was reduced (21). This effect of WNK4 was
abolished by PHAII mutation Q562E. The authors of this study had access to a
mouse transgenic for WNK4 harboring the PHAII mutation Q562E (further
discussed below) and were able to show that amiloride sensitive current was
increased in the colonic epithelium of these animals. No data was presented on the
effects on ENaC in CCD, but this in vivo observation illustrates another way in
which PHAII triggers salt retention. The overall contribution of colonic salt retention
to the pathology is not clear.
WNK4 reduces surface expression of various extrarenal transport
proteins
The observation that WNK4 is expressed in chloride transporting epithelia outside
the kidney led investigators to expand the oocyte studies and include other channels
and transporters. Coexpression with WNK4 causes reduced surface expression of the
basolateral bumetanide sensitive sodium-potassium-2-chloride co-transporter
subtype 1(NKCC1) (9), the apical chloride anion exchanger (CFEX) (9), the cystic
fibrosis transmembrane conductance regulator (CFTR) (33) and the recently
9
characterized chloride channel SLC26A9 (3). Each one of these transport proteins
has an expression pattern that includes several chloride transporting epithelia and the
significance of these WNK4 actions in the development of PHAII is unclear. The
CFTR study was the only one that included coexpression of PHAII mutant WNK4
Q562E and the effect was an augmentation of the surface retrieval (33).
WNK4 increases paracellular chloride permeability of MDCKII
monolayers
Two studies have been published in which WNK4 was transfected into MDCKII
cells and paracellular transport characteristics measured (10, 31). Both groups report
that WNK4 is localized to the tight junction when expressed in MDCKII cells.
Yamauchi et al were able to determine that WNK4 harboring the PHAII mutation
D561A increased paracellular chloride permeability by measuring
22
Na and
36
Cl
fluxes over MDCKII monolayers (31). The authors go on to show that the increases
in chloride permeability coincide with increases in claudin phosphorylation. Kahle et
al used electrophysiological methods to record an increase in paracellular chloride
permeability in cells expressing wild type WNK4. Chloride permeability was
increased further in cells expressing PHAII mutants Q562E and E559K (10). In
addition Kahle showed that WNK4 harboring the kinase domain mutation D318A
had no effect on paracellular chloride permeability at all. These observations lend
support to the chloride shunt hypothesis by suggesting that the volume expansion
10
seen in PHAII is due to an increase in chloride leakiness of the distal nephron
epithelium.
Two mouse models reproduce the PHAII phenotype
In an effort to translate the sometimes conflicting in vitro data into an in vivo setting,
two groups generated mouse models of PHAII. Lalioti et al produced one strain of
mice transgenic for WNK4 harboring the PHAII associated mutation Q562E
[Tg(Wnk4
PHAII
)] and one strain transgenic for wild type WNK4 [Tg(Wnk4
WT
)] and
analyzed the physiology of these strains (12). Compared to non-transgenic
littermates the Tg(Wnk4
PHAII
) mice had a 100% increase in the total amount of
WNK4 mRNA and the Tg(Wnk4
WT
) mice had a 50% increase, indicating that these
transgenic strains overexpress PHAII mutant and wild type WNK4 in a background
of endogenous WNK4. The physiology of the Tg(Wnk4
PHAII
) mice closely resembled
that of human PHAII patients with elevated blood pressure and serum potassium
levels accompanied by depressed serum bicarbonate (acidosis). Tg(Wnk4
WT
) mice on
the other hand had lower blood pressure than non-transgenic littermates and showed
signs of hypokalemia when challenged with a low K
+
diet. Striking changes in the
morphology of the DCT and the surface expression of NCC were observed in both
transgenic strains with a marked DCT hyperplasia and increase in surface NCC in
Tg(Wnk4
PHAII
) mice and the opposite in Tg(Wnk4
WT
) mice. No morphological
changes were reported for other nephron segments. Also noteworthy was the fact that
there were no differences in ROMK1 expression between Tg(Wnk4
PHAII
) mice and
11
non-transgenic littermates. The data suggest that the PHAII phenotype is the result of
an overactive NCC reabsorbing a larger proportion of filtered Na
+
and Cl
-
in the DCT
causing volume expansion and decrease in NaCl delivery to the collecting duct
where K
+
excretion through ROMK goes down due to the drop in electrogenic Na
+
reabsorption through ENaC (Figure 4). In support of this hypothesis is the
observation that the hyperkalemia is corrected by either treatment with
hydrochlorothiazide or knocking out the NCC gene in the Tg(Wnk4
PHAII
) mice.
Yang et al generated Wnk4
D561A/+
knock-in mice by replacing one of the
endogenous WNK4 alleles with an allele containing the PHAII mutation D561A
(34). These animals faithfully reproduced the PHAII phenotypes of hypertension,
hyperkalemia and metabolic acidosis, all corrected by hydrochlorothiazide treatment.
Total amount of NCC was increased in the DCT according to western blot.
Immunofluorescence and immunoelectron microscopy also indicated that he luminal
surface area and apical surface density of NCC expression in the DCT were greater
in the Wnk4
D561A/+
mice than in their wild type littermates. Using antibodies raised
against phosphorylated NCC and SPAK/OSR1 (kinases shown to be substrates of
WNK4 (24) and in turn may potentially phosphorylate and activate NCC (17),
discussed further below) the authors showed an increase in the phosphorylation of
both SPAK/OSR1 and NCC in the Wnk4
D561A/+
mice, with p-NCC strongly enriched
in the apical membrane. ROMK1 protein levels and subcellular localization in the
CCD were not affected in the Wnk4
D561A/+
mice, but an increase in ENaC density and
activity was reported. In vitro microperfusion studies revealed that amiloride
12
Figure 4. Mouse models of PHAII. Data from two independent mouse models
suggest that the PHAII phenotype is the result of an overactive NCC. Increased NaCl
reabsorption in the hyperplasic DCT results in reduced NaCl delivery in the CCD
and dissipation of the lumen negative potential, established by electrogenic Na
uptake by ENaC, that normally drives K secretion through ROMK.
13
sensitive sodium permeability, but not chloride permeability, was enhanced in the
CCD. The fact that this upregulation of ENaC could be reversed by the addition of
hydrochlorothiazide led the investigators to the conclusion that the pathophysiology
of these mice resulted from an upregulation of NCC in the DCT and that the effects
on ENaC represented compensatory mechanisms in the CCD in response to
diminished NaCl delivery. The overall conclusions are similar to the transgenic study
by Kahle et al, but the phospho-antibody data adds a mechanism by which the NCC
upregulation may be accomplished.
Kinase activity of WNK4
The mechanism of WNK4 action in the studies summarized above is ambiguous, in
some cases attributed to protein-protein interaction and in others to WNK4 acting as
a kinase. Including the kinase domain mutant D318A or looking at phosphorylation
levels of proteins believed to be WNK4 substrates addresses the importance of
kinase activity for the observed regulatory process. However, these are indirect
methods and cannot make the distinction between whether the observed effects are
mediated directly by WNK4 phosphorylation or by some other mechanism, possibly
involving WNK4 interacting with another protein kinase. It is of fundamental
importance to characterize the kinase activity of WNK4 in order to fully understand
its mechanism of action. WNK4 was originally identified as a paralog of WNK1, but
as a serine-threonine kinase WNK4 is not as well characterized as WNK1. However,
14
with 80% homology between the kinase domains of these two proteins (23, 26), the
basic elements are most likely the same (Figure 5).
WNK1 behaves like a serine-threonine kinase
WNK1 was originally identified during an attempt to isolate novel serine-threonine
kinases of the MEK family (29). Their approach entailed using nested degenerate
primers, designed based on conserved features of MEK kinases, for PCR on
mammalian cell mRNA isolates and screening resulting products by using them as
probes and pick out candidate cDNAs from libraries. From such experiments carried
out on the rat derived cell line PC-12 and rat cDNA libraries, a novel kinase was
isolated and named WNK1 because of its peculiar lack of a critical lysine residue in
kinase subdomain II. The full-length rat cDNA was 7.2 kb and encoded a protein of
2126 amino acids, apparent MW 230 kD. WNK1 shares approximately 50%
sequence homology with other serine-threonine kinases such as STK2 and MEKK3
in its catalytic domain (23). Most of the residues that have been identified in
previous studies to be crucial for kinase activity (7) are conserved with the exception
of the aforementioned lysine, which has been replaced by a cysteine (C250 in rat
WNK1). In the three dimensional structure of the serine-threonine kinase PKA,
crystallized with ATP, this highly conserved lysine (K72 in PKA) projects into the
ATP binding pocket and interacts with the α and β phosphates on the ATP molecule
(Figure 6). Despite the functional importance of this lysine in other serine-threonine
kinases, as shown by mutational studies (35), WNK1 was able to phosphorylate
15
Figure 5. Schematic representations of WNK1 and WNK4. WNK1 and WNK4 share
80% homology in their kinase domains. Also shown is the truncated WNK4
construct employed by Vitari et al to assay WNK4 kinase activity (24).
Abbreviations: FL, full-length; Δ593, truncated at amino acid 593; AID,
autoinhibitory domain; CC, coiled-coil domain; GST, glutatione S-transferase
affinity tag.
Figure 6. Crystal structures of WNK1 (left) and PKA (right). Lysine 72 in PKA
represents an invariant lysine residue found in subdomain II of all known serine-
threonine kinases, except for the WNK family of kinases where it has been replaced
by a cysteine (Cys 250 in WNK1). Lysine 233 in subdomain I is believed to assume
the role of the missing lysine in WNK1.
16
myelin basic protein (MBP), a common serine-threonine kinase substrate, in an in
vitro kinase assay (29). Homology modeling of WNK1 kinase domain on to the
crystal structure of PKA Cα catalytic domain, combined with mutational studies, led
Xu et al to propose that the functional role of K72 in PKA is assumed by another
lysine (K233 in rat WNK1), which occupies a nearby position in kinase subdomain I
normally taken up by a glycine (G55 in PKA) in the highly conserved G-X-X-G-X-G
motif. When lysine 233 is mutated to alanine WNK1 loses its ability to
phosphorylate MBP and itself (29). The crystallization of the catalytic domain of
WNK1 further substantiated the notion that K233 in WNK1 assumes the role of K72
in PKA as it can be seen pointing into the catalytic cleft at approximately the same
spatial position (Figure 6)(16). The overall configuration of the catalytic domains of
WNK1 and PKA are remarkably similar and most of the residues conserved
throughout the serine-threonine kinase superfamily can be found at their expected
positions.
Structural elements of WNK4
Human WNK4 encodes a protein of 1243 amino acids and apparent MW 135 kD.
Mouse WNK4 is slightly shorter at 1222 amino acids. The high sequence identity
between the catalytic domains of WNK1 and WNK4 suggests that functionally
important residues in WNK4 may be predicted by aligning the amino acid sequence
of the WNK4 catalytic domain with that of the WNK1 kinase domain. The
previously discussed residues K233, C250 and D368 of rat WNK1 correspond to
17
K183, C200 and D318 in mouse WNK4. A rational approach to producing a kinase-
dead WNK4 mutant may be taken by combining this information with the kinase
studies performed on WNK1 by Xu et al (29). A linear representation of the WNK4
protein sequence shows the catalytic domain roughly 300 residues long stretching
from 150 to 450 near the amino (N) terminal (Figure 5). Two coiled-coil domains
were predicted based on the algorithm introduced by Lupas (14). From the N-
terminal the first one starts about 50 residues downstream of the catalytic domain
between residues 500-530 and the second one can be found relatively close to the
carboxyl (C) terminal between residues 1110-1140. Coiled-coil domains are
typically the sites of interaction with other proteins. The three most investigated
PHAII associated mutations are located in a negatively charged segment just 30
residues downstream from the first coiled coil domain. It is unclear whether this is
close enough to have an effect on a putative protein interaction and predictions are
made more difficult because no structural models have been proposed for the
sequence outside of the catalytic domain (75%) due to the lack of conserved
structural elements.
Studies of WNK4 kinase activity
Yamauchi et al were first to describe in vitro phosphorylation by WNK4 using a
GST fusion protein of the WNK4 catalytic domain (amino acids 65-457) and the c-
terminal cytoplasmic domain of claudin 4 as substrate (31). No negative controls
were included and the construct was too short to contain the segment that harbors the
18
PHAII mutations. In a study by Wang et al suggestions were made that WNK4 may
contain an autoinhibitory domain located between the C-terminal end of the kinase
domain and the beginning of the first putative coiled-coil, comprising amino acids
444-518 (25). This assumption was based on the homology with WNK1, which was
reported to contain an autoinhibitory domain in the corresponding region (30).
Expressing a number of different GST-WNK4 constructs containing the catalytic
domain with or without its C-terminal autoinhibitory domain, the authors were
unable to show in vitro phosphorylation of the common serine-threonine kinase
substrate histone.
In an effort to identify interaction partners of the WNK kinases, Vitari et al
showed that a protein called sterile20-related, proline-, alanine-rich kinase (SPAK)
coprecipitated with WNK1 from rat testis lysate (24). SPAK and the closely related
oxidative stress-responsive kinase (OSR) 1 are Ste20p-related protein kinases that
bind to and phosphorylate members of the solute carrier (SLC) 12 family including
NKCC1, NKCC2 (4, 20) and potentially NCC (17). Truncated GST-fusion proteins
of WNK1 (amino acids 1-661) and WNK4 (amino acids 1-593) (Figure 5), expressed
and recovered from HEK293 cells, are able to phosphorylate both SPAK and OSR1
in vitro (24). Phosphorylation of MBP, claudin-4 or histone under similar conditions
was very weak or non-existent. Introducing kinase domain mutation D368A into
WNK1 Δ661 and the double mutation K183A/D318A into WNK4 Δ593 reduced in
vitro phosphorylation of SPAK and OSR1 to almost zero, indicating that these
observations stem from specific WNK kinase activity. The authors also confirmed
19
that phosphorylation of OSR1 and SPAK led to the activation of these proteins, as
measured by their ability to in turn phosphorylate NKCC1. With previous reports
indicating that phosphorylation of NKCC1 leads to its activation (4), the end result
of WNK4 acting via this pathway would seem to be the activation of sodium,
potassium and chloride entry via NKCC1. This amounts to the opposite of the
findings in oocytes, where cotransfection of WNK4 and NKCC1 leads to inhibition
of NKCC1 carried current by reduction of surface expression (9). Two additional
studies confirm the finding that WNK1 associates to and phosphorylates SPAK and
OSR1, which in turn activate members of the SLC12 family of transporters (1, 17).
These reports illustrate a molecular pathway by which WNK kinases may control ion
transport processes and identify novel substrates of WNK phosphorylation.
All the above mentioned in vitro phosphorylation experiments were
performed with truncated forms of the kinases and it is therefore still an open
question whether full length WNK4 has kinase activity and if this activity is
influenced by the PHAII associated mutations.
20
Chapter 2: Aims of this project
Our goal is to characterize the kinase activity of the full-length WNK4 protein in an
effort to address the mechanism by which WNK4 influences its targets. It is of
importance to investigate how the full-length protein behaves as compared to
truncated versions that may lack structural elements that are important for the
physiological functions that are summarized above. In addition, working with full-
length WNK4 allows us to determine what, if any, effects disease causing mutations
found outside the kinase domain may have on kinase activity. This gives us the
following specific aims for the project:
1. Express full-length WNK4 and characterize its kinase activity.
2. Determine the effect of kinase domain and PHAII mutations on the ability of
the expressed protein to phosphorylate substrates in vitro.
21
Chapter 3: Materials and Methods
Reagents
Mouse anti-c-myc antibody was purchased from Zymed (San Diego, CA).
Lipofectamine™ 2000, rProtein G Agarose were from Invitrogen (Carlsbad, CA).
Complete Mini protease inhibitors were purchased in the form of tablets from Roche,
one tablet per 10 ml of solution. Glutathione sepharose 4B and enhanced
chemiluminescence (ECL) western blot detection kit including Horseradish
peroxidase (HRP) conjugated secondary antibody and ECL substrate solution were
from Amersham Biosciences. Polyvinyl difluoride (PVDF) membrane was from
Millipore. Protein concentration was determined by BCA protein assay with kit from
Pierce biotechnology. Recombinant Extracellular signal-regulated kinase 2 (Erk2)
was purchased from Calbiochem.
WNK4 Expression Constructs
pCSPT-mWNK4-KXX
A cDNA, identified by expressed sequence tag [EST 602105958F1], from a mouse
kidney cDNA library was found to contain the full mouse WNK4 open reading
frame (ORF). The clone was provided in the pCMV-SPORT6 vector and PCR was
used to introduce a 5’ Kozak sequence for increased protein expression level and a 3’
c-myc tag for easy detection and recovery with anti-c-myc antibody, creating the
22
pCSPT-mWNK4-KXX construct. The sequence was confirmed by DNA sequencing.
PHAII associated mutations E559K, D561A and Q562E were introduced by PCR
into a BstBI/SpeI restriction cassette of pCSPT-mWNK4-KXX and “kinase dead”
mutation D318A was introduced by PCR into a BglII/SphI restriction cassette. The
presence of each individual desired mutation was confirmed by DNA sequencing
before the cassette was reinserted into pCSPT-mWNK4-KXX. Two other kinase
domain mutants, K183M and G180A were introduced by PCR into a KpnI/BbvCI
restriction cassette and were sequenced before being reinserted into pCSPT-
mWNK4-KXX.
pNTAP-WNK4
An XhoI fragment of the pCSPT-mWNK4-KXX plasmid containing the WNK4
ORF with its c-terminal c-myc tag was cloned into XhoI digested pNTAP-B vector
(Stratagene) to produce the pNTAP-WNK4 construct with an n-terminal tandem
affinity purification (TAP) tag in frame with the WNK4 ORF. This tandem affinity
tag consists of a streptavidin binding peptide in tandem with a calmodulin binding
peptide coupled to the n-terminus of the expressed protein. The orientation of the
insert was confirmed by restriction digestion. PHAII associated mutations E559K,
D561A and Q562E and kinase domain mutation D318A were introduced into the
pNTAP-construct by subcloning a BglII/SpeI restriction fragment of the pCSPT-
construct containing each corresponding mutation. Kinase mutants G180A and
K183M were subcloned similarly using a NheI/BglII fragment of the pCSPT-
23
construct. The K183M and D318A mutations were combined by inserting a
BbvCI/SacII fragment from the pNTAP-WNK4 K183M plasmid into BbvCI/SacII
digested pNTAP-WNK4 D318A producing the kinase double mutant construct
pNTAP-WNK4 KD.
Truncated pNTAP-WNK4
C-terminal truncations: Three constructs based on the pNTAP-WNK4 plasmid were
generated by PCR with truncations at amino acid 443 (Δ443), 500 (Δ500) and 593
(Δ593). Due to the presence of inconvenient restriction sites in the pNTAP-WNK4
plasmid, a template suitable for the PCR protocol had to be generated in which three
n-terminal fragments of WNK4 could be amplified and inserted in front of the c-
terminal c-myc tag. In the original pCSPT-mWNK4-KXX construct a BamHI site
was engineered in between the WNK4 ORF and the c-myc tag. The presence of two
other BamHI sites inside the WNK4 ORF makes it impossible to use BamHI to
excise the intact ORF without the tag. To obtain a unique restriction site between the
WNK4 ORF and the c-myc tag, two sets of primers were designed for site directed
mutagenesis with the pCSPT-mWNK4 construct as template. By mutagenic PCR the
c-terminal BamHI site was changed to an ApaI site and another ApaI site found
inside the WNK4 ORF was abolished with a silent mutation, producing the construct
pCSPT-mWNK4-BA. An additional ApaI site located 3’ of the c-myc tag in the
pcSPORT6 vector multiple cloning site results in the separation of the tag from the
vector when pCSPT-mWNK4-BA is digested with ApaI. In order to circumvent this
24
problem a KpnI/XbaI cassette, containing the WNK4 ORF and the c-myc-tag but
without the 3’ ApaI site, was subcloned from pCSPT-mWNK4-BA into pBlueScript
KS+, creating the construct pKS-Kpn/Xba. This construct was used as the template
for a PCR where the left primer was complimentary to a sequence just 5’ of a unique
BbvCI site and three different right primers were complimentary to the sequences
ending with the codons for D443, Y500 and D593. Each right primer was designed
with an ApaI site at the 5’ end. PCR products were digested with BbvCI and ApaI
and cloned into BbvCI/ApaI digested pKS-Kpn/Xba. Sequences were confirmed by
DNA sequencing. A BbvCI/XbaI cassette from each truncated pKS-Kpn/Xba
construct was subcloned into BbvCI/XbaI digested pNTAP-WNK4 producing the
three constructs pNTAP-WNK4 Δ443, Δ500 and Δ593. Kinase domain mutation
K183M was introduced into pNTAP-WNK4 Δ593 by subcloning a BbvCI/EcoRI
fragment of pNTAP-WNK4 K183M into BbvCI/EcoRI digested pNTAP-WNK4
Δ593. PHAII associated mutations E559K, D561A and Q562E and kinase domain
mutation D318A were introduced into the Δ593 construct by amplifying a fragment
of each corresponding pCSPT-mWNK4-KXX plasmid using the left primer together
with the right primer ending with D593 described above. Each PCR product
harboring its mutation was digested with BbvCI and ApaI and cloned into
BbvCI/ApaI digested pKS-Kpn/Xba. Sequences were confirmed by DNA
sequencing. A BbvCI/XbaI cassette from each truncated pKS-Kpn/Xba construct
was subcloned into BbvCI/XbaI digested pNTAP-WNK4 Δ593.
25
N-terminal truncation: A c-terminal fragment of WNK4 containing amino acids 594-
1222 was amplified by PCR with right primer complimentary to a sequence starting
with A594 and with a 5’ HindIII restriction site and with left primer annealing 3’ of
c-myc tag and stop codon. The unique HindIII site in pNTAP-WNK4 is located in
the multiple cloning site 5’ of the XhoI site used to clone the WNK4 ORF into
pNTAP-B. By digesting the PCR product with HindIII and XbaI and inserting it into
HindIII/XbaI digested pNTAP-WNK4, a construct with an n-terminal TAP-tag in
frame with amino acids 594-1222 of WNK4 followed by a c-terminal c-myc tag is
produced. Sequence was confirmed by DNA sequencing.
Protein Expression and Recovery
Protein Expression in Mammalian Cells HEK293, COS-7 and CHO
Purified plasmid DNA was transfected into mammalian cells, grown on P10 tissue
culture plates until 95% confluency, using Lipofectamine™ 2000 according to
manufacturers protocol. Cells were harvested and proteins recovered 24 hours post
transfection.
Recovery of c-myc-tagged WNK4 by Immunoprecipitation
Cells transfected with the pCSPT-mWNK4-KXX plasmid were lysed directly on
culture plate in 1 ml IP lysis buffer per P10 plate (50 mM Tris-HCl, pH 7.4, 150 mM
NaCl, 1% Triton X-100 supplemented with complete mini protease inhibitors) and
cell debris was cleared by 10 min centrifugation at 16,100 x g. Lysates were
26
precleared with 20 µl protein G agarose per ml of lysate and then incubated with 2.5
µg mouse anti-c-myc antibody for 2 hours. Immunocomplexes were precipitated
with 40 µl protein G agarose per ml lysate, washed once in IP lysis buffer, twice in
IP wash buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl) and resuspended in 40 µl
50 mM HEPES pH 8. Samples were divided into four aliquots and stored at -80°C.
Protein Recovery by Tandem Affinity Purification (TAP)
Cells transfected with pNTAP-WNK4 plasmid were harvested by scraping in 1 ml of
lysis buffer (provided with TAP kit, supplemented with complete mini protease
inhibitors) and lysed by three cycles of 20 min incubation at -80°C followed by 10
min in ice-water. TAP-tagged protein was recovered according to manufacturers
protocol, briefly summarized below. Lysates were cleared by 20 min centrifugation
at 16,000 x g and TAP-WNK4 was precipitated from the supernatant with 50 µl
streptavidin resin slurry for 2 hours at 4°C. Precipitates were washed twice with
streptavidin binding buffer (provided with TAP kit, supplemented with complete
mini protease inhibitors and β-mercaptoethanol) and TAP-WNK4 was eluted 100 µl
streptavidin elution buffer (provided with TAP kit, supplemented with complete mini
protease inhibitors and β-mercaptoethanol). Eluates were diluted to 500 µl with
calmodulin binding buffer (provided with TAP kit, supplemented with complete mini
protease inhibitors and β-mercaptoethanol) and TAP-WNK4 was precipitated again
with 25 µl calmodulin resin slurry. Precipitates were washed twice with calmodulin
27
binding buffer and TAP-WNK4 was eluted in 50 µl of calmodulin elution buffer.
Eluates were divided into five aliquots of 10 µl and stored at -80°C.
Where indicated, the kit-provided lysis buffer was replaced with a modified
lysis buffer used by Vitari et al (24) containing 50 mM Tris-HCl, pH 7.5, 1 mM
EDTA, 1 mM EGTA, 1% NP-40, 1 mM Na
3
VO
4
, 50 mM NaF, 5 mM sodium
pyrophosphate, 0.27 M sucrose and 1 mM DTT. The manufacturers protocol was
followed as normal with the exception of four extra 5 min washes of the streptavidin
resin precipitate in the modified lysis buffer supplemented with 0.15 M NaCl.
OSR1 Expression and purification from Escherichia Coli
A pGEX-6P construct containing the kinase inactive OSR1 D164A (kind gift from
the Dario Alessi lab (24)) was transformed into BL21 E. coli cells and grown in 0.5
L Luria Broth with 100 µg/ml ampicillin until reaching an OD
600
of 0.7. GST-OSR1
expression was induced for 2 hours by the addition of Isopropyl β-D-
thiogalactopyranoside (IPTG) to a final concentration of 0.5 mM. Cells were
harvested by 15 min centrifugation at 6000 x g and lysed by sonication in 25 ml PBS
supplemented with protease inhibitors. Cell debris was pelleted by a 20 min
centrifugation at 23,000 x g and the supernatant was applied to a column with 250 µl
glutathione sepharose 4B. Column resin was washed twice with 5 ml of ice cold PBS
and protein was eluted with three ten minute incubations at room temp with 300 µl
glutathione elution buffer (10 mM glutathione in 50 mM Tris-HCl, pH 8).
28
Protein Detection by SDS-PAGE and Western Blot
Denaturing sodiumdodecylsulfate polyacrylamide gel electrophoresis was performed
according to standard protocol with 7.5% Biorad mini protean gels. Proteins were
transferred to PVDF membranes and detected using either ECL method or Odyssey
LiCor system, both according to manufacturers protocol as described briefly below.
ECL: Membranes were blocked with 5% milk in TBST (10 mM Tris-HCl, pH 7.5,
150 mM NaCl and 0.1% Tween 20) and proteins were detected with the indicated
antibody diluted in primary antibody solution (TBST with 1% BSA and 0.1%
sodium azide). HRP-conjugated secondary antibody was diluted 1:1000 in TBST.
Odyssey LiCor: Membranes were blocked with Odyssey blocking reagent and the
indicated primary antibody was diluted in primary antibody solution. Secondary
Alexafluor 680 antibody was diluted 1:5000 in TBST.
In vitro kinase assay
Immunoprecipitated WNK4
Immunoprecipitates consisting of 10 µl liquid volume and 10 µl protein G agarose
were thawed and mixed with 10 µl of 2x reaction buffer (20 mM MgCl
2
, 2 mM DTT,
0.2 mg/ml BSA, 100 µM Na
2
ATP, 1 µCi/µl γ
32
P-ATP and 1 µg/µl [5 µM] myelin
basic protein). Phosphorylation was carried out for 30 min at 30°C and the reaction
was stopped by boiling the samples in SDS sample buffer (62.5 mM Tris-HCl, pH
6.8, 2% SDS, 20% glycerol, 0.01% Bromophenol Blue). Phosphorylated proteins
29
were separated on 10% polyacrylamide gels and phosphorylation was visualized by
autoradiography of dried coomassie blue stained gels. Phosphate incorporation into
protein bands, excised from the gel, was quantified on a Beckmann LS6000 liquid
scintillation counter.
TAP-WNK4
The in vitro assay was slightly modified to compensate for the presence of 2 mM
EGTA and unknown buffer strength of the calmodulin elution buffer. 10 µl 2 x
reaction buffer (40 mM HEPES, pH 8, 24 mM MgCl, 2 mM DTT, 0.2 mg/ml BSA,
100 µM Na
2
ATP, 1 µCi/µl γ
32
P-ATP and 0.1 µg/µl [5 µM] MBP or 0.5 µg/µl [5 µM]
OSR1) was mixed with one 10 µl aliquot of TAP-WNK4 and incubated for 30 min at
30°C. Otherwise as above.
In-gel kinase assay
A method for assaying in-gel phosphorylation of kinases was adapted from Wooten
(28). A 7.5 % polyacrylamide separating gel was cast with 1% glycerol and with
substrate (MBP at 0.5 mg/ml or OSR1 at 0.2 mg/ml in separating gel buffer [375
mM Tris-HCl, pH 8.8, 0.1% SDS]) copolymerized into gel. Kinase samples were
boiled for 3 min in SDS-sample buffer and then separated on the substrate containing
gel. SDS was removed from the gel by four washes with SDS-removal solution I (50
mM Tris-HCl, pH 8 and 20% 2-propanol) and three washes in SDS-removal solution
II (50 mM Tris-HCl, pH 8 and 1 mM DTT) for a total of two hours. Proteins were
denatured for two hours in denaturation solution (50 mM Tris-HCl, pH 8, 20 mM
30
DTT and 6 M Guanidine-HCl) and then renatured over night in renaturation solution
(50 mM Tris-HCl, pH 8, 5 mM DTT, 0.04% Tween 20, 100 mM NaCl and 5 mM
MgCl
2
). Phosphorylation was carried out by submerging the gel in kinase buffer (25
mM HEPES, pH 7.4, 20 mM MgCl
2
, 1 mM MnCl
2
, 5 mM NaF, 100 µM Na
3
VO
4
, 2.5
µg/ml pNPP, 1 mM DTT, 50 µM Na
2
ATP and 20 µCi/ml γ
32
P-ATP) for 2 hours at
30°C, after which excess γ
32
P-ATP was removed with gel wash solution (5%
trichloro-acetic acid and 1% sodium pyrophosphate) for 1 hour. Phosphorylated
bands were visualized by autoradiography of the dried gel.
31
Chapter 4: Results
Assay of kinase activity in WNK4 protein expressed and
immunoprecipitated from HEK293 cells
Measuring in vitro phosphorylation of model substrates is a useful and commonly
employed method for investigating the activity of predicted kinases (25, 29). In an
effort to determine whether the predicted serine-threonine kinase WNK4 has kinase
activity, we expressed full-length WNK4 and recovered the protein by
immunoprecipitation. Immunoprecipitated WNK4 was assayed for in vitro
phosphorylation of the model substrate MBP.
Immunoprecipitation of WNK4 expressed in HEK293 cells
Full-length mouse WNK4 (1222 amino acids), fused to a c-terminal c-myc-tag, was
expressed in HEK293 cells and recovered by immunoprecipitation with the c-myc-
antibody. Western blot analysis using the same anti-c-myc antibody at 1:200 dilution
revealed a band at ~170 kD corresponding to the myc-tagged WNK4 (Figure 7A)
(corresponds well with the 170 kD band for myc-tagged WNK4 reported by Wilson
et al (27)). WNK4 protein harboring kinase domain mutation D318A and PHAII
mutations E559K, D561A and Q562E was detected at the same level of expression
as wild type protein.
32
Figure 7. Assay of kinase activity in WNK4 protein expressed and
immunoprecipitated from HEK293 cells. (A) HEK293 cells were transfected with
the indicated pCSPT-mWNK4-KXX construct and 24 hours post transfection WNK4
protein was recovered from cell lysates by immunoprecipitation with anti c-myc
antibody. Immunoprecipitates were probed with anti c-myc antibody and WNK4 is
detcted at relative molecular mass of 170. (B) In vitro kinase assay. The indicated
WNK4 immunoprecipitate was incubated with myelin basic protein in the presence
of Mg
2+
and γ
32
P-ATP. Phsophorylated proteins were separated by SDS-PAGE (10%
gel) and visualized by autoradiography. Extracellular regulated kinase 2 was used as
positive control for MBP phosphorylation and HEPES buffer as negative control.
MBP migrates at approximately 20 KDa with, what we assume to be, a number of
smaller degradation products. (C) In vitro kinase assay. As in (B) but with additional
kinase domain mutants K183M and G180A. The following controls were included:
MBP incubated in the presence of γ
32
P-ATP and Mg
2+
with 1) anti c-myc precipitate
from HEK293 cells transfected with empty expression vector [Vector], 2) Protein G
sepharose purification resin in HEPES buffer [PrG], 3) Protein G sepharose
preincubated with immunoglobulin [PrG+Ig], 4) HEPES buffer [HEPES].
33
In vitro phosphorylation of MBP by immunoprecipitated WNK4
An in vitro kinase assay was set up to study WNK4 mediated phosphorylation of the
serine-threonine kinase model substrate MBP (Figure 8). Two major phosphorylated
products appeared on the gel when immunoprecipitated WNK4 was assayed (Figure
7B). The top band, running at approximately 170 kD, corresponds to phosphorylated
WNK4 and the bottom band at 18 kD is MBP and its degradation products.
Surprisingly, kinase domain mutation D318A and PHAII mutations E559K, D561A
and Q562E appear to have no effect on this phosphorylation. 5 U of recombinant
Erk2 was included as a positive control and can be seen phosphorylating MBP
strongly. In an attempt to address the fact that the kinase domain mutation D318A
appears to have no effect on kinase activity, another two mutations were introduced
into the kinase domain. K183M substitutes the lysine thought to assume the role of
the missing lysine in the active site and G180A replaces an absolutely conserved
lysine in the kinase domain. Neither one of these mutations appear to affect the in
vitro phosphorylation of MBP (Figure 7C). Importantly, MBP phosphorylation is
observed for anti-c-myc precipitates from cells transfected with the empty expression
vector pCMV-SPORT6. These observations suggest that phosphorylation is not
mediated by WNK4, but rather by endogenous kinases in HEK293 cells that
coprecipitate with the resin during the recovery process.
34
Figure 8. Kinase assays
35
Assay of kinase activity in WNK4 protein expressed in HEK293 cells
and recovered by tandem affinity purification
The substrate phosphorylation observed in the in vitro kinase assay appears to be
caused by endogenous kinases binding to the immunoprecipitation resin and
therefore we decided to use another method to recover our expressed WNK4 protein.
We subcloned the WNK4 ORF into the pNTAP vector, which couples a dual affinity
tag to the n-terminus of the protein. The tandem affinity purification (TAP)
expression system allows the expressed protein to be precipitated and then
specifically eluted from the purification resin in two successive rounds (Figure 9). As
opposed to the immunoprecipitated WNK4 protein, which is still associated to the
protein G resin when assayed for in vitro kinase activity, TAP-WNK4 is freely
dissolved in the elution buffer, eliminating the possibility of interference in the
kinase assay by kinases that stick to the purification resin.
Tandem affinity purification of WNK4 expressed in HEK293
We were able to express and recover TAP-WNK4 protein at a typical concentration
of 50 µg/ml. Kinase domain and PHAII mutant TAP-WNK4 was recovered at levels
similar to wild type and were visualized as ~170 kD bands on a coomassie stained
gel (Figure 10A).
36
Figure 9. Tandem affinity purification. Expressed WNK4 is fused to two affinity
tags, streptavidin binding peptide and calmodulin binding peptide. In the first round
the fusion protein is precipitated with a streptavidin conjugated purification resin and
eluted with biotin. Another precipitation, this time with calmodulin resin, followed
by elution under low Ca
2+
conditions concludes the second round of purification.
37
Figure 10. Assay of kinase activity in WNK4 protein expressed in HEK293 cells
and recovered by tandem affinity purification. (A) HEK293 cells were transfected
with the indicated TAP-WNK4 construct and 24 hours post transfection WNK4
protein was recovered from cell lysates by tandem affinity purification. TAP-WNK4
proteins were visualized as ~170 kD bands on a coomassie stained gel. (B) and (C)
Kinase domain and PHAII mutant TAP-WNK4 was assayed for in vitro
phosphorylation of MBP and OSR1.
32
P-phosphorylated proteins were separated by
SDS-PAGE on 10% (B) or 7.5% (C) gels and visualized by overnight (B) or 3 hr (C)
autoradiography. As observed by other investigators, OSR1 migrates as a doublet
band (24). (D) and (E) As in (B) and (C) but with the K183M/D318A double mutant
KD. Negative controls for the phosphorylation experiments consisted of substrate
incubated in the presence of Mg
2+
and γ
32
P-ATP with samples derived from cells
transfected with empty TAP vector (Vector) or with HEPES buffer.
38
In vitro kinase assay of TAP-WNK4 with MBP and OSR1 as substrates
TAP-WNK4 proteins were assayed for in vitro phosphorylation of MBP and the
WNK4 substrate OSR1 (24). Phosphorylated bands corresponding to TAP-WNK4
and MBP were detected when the wild type protein was assayed (Figure 10B).
Neither one of the PHAII mutations or kinase domain mutations appear to have any
effect on the strength of this phosphorylation. Similar results were obtained with
OSR1, although OSR1 phosphorylation in general appears to be stronger than MBP
phosphorylation (Figure 10C). No phosphorylation of either MBP or OSR1 was
observed when samples derived from cells transfected with empty TAP vector were
assayed. These results suggest that either the individual kinase domain mutations
G180A, K183M and D318A have no effect on WNK4 kinase activity or that another
kinase, that is able to phosphorylate MBP and OSR1, is specifically associated to
WNK4 expressed in HEK293 cells. To address the question regarding the effect of
individual kinase domain mutations on WNK4 kinase activity, another construct was
produced in which K183M and D318A were combined into a kinase double mutant.
In a previously published report the combination of these two mutations prevented
the in vitro phosphorylation of OSR1 by a kinase domain fragment of WNK4 (24).
In our experiments TAP-WNK4 harboring kinase domain mutations K183M and
D318A, separately or together (KD), was able to in vitro phosphorylate MBP (Figure
10D) or OSR1 (Figure 10E) to the same extent as the wild type protein. The fact that
none of the kinase domain mutations appear to have any effect on in vitro
phosphorylation of MBP, OSR1 and TAP-WNK4 indicates that this phosphorylation
39
is most likely mediated by a kinase that copurifies with TAP-WNK4 rather than by
TAP-WNK4 itself.
In-gel kinase assay of TAP-WNK4 with OSR1 as substrate
In an effort to identify kinases that copurify with TAP-WNK4, an assay was set up
that allows the separation and detection of individual protein kinases in a mixture.
The in-gel kinase assay is based on separating the components of the kinase
preparation by denaturing SDS-PAGE and then reconstituting the proteins to their
native conformations in the separating gel (Figure 8). By incubating the gel in kinase
assay buffer containing γ
32
P-ATP, autophosphorylation can be detected as a
phosphorylated band at the expected molecular weight of the kinase. Substrate
phosphorylation can be assayed by copolymerizing the substrate into the gel,
allowing the renatured kinase to phosphorylate substrate molecules in its immediate
vicinity, which results in a phosphorylated band at the expected molecular weight of
the kinase. When the TAP-WNK4 samples were assayed for in-gel phosphorylation
of OSR1, a prominent doublet band appears at 40 kD (Figure 11) in all samples. No
signal is detected at 170 kD, the expected molecular weight of WNK4, or in the lane
containing recombinant Erk2. This suggests that a 40 kD kinase associated with
WNK4 is responsible for the majority of the observed in vitro phosphorylation of
OSR1 and MBP. The introduction of point mutations in the kinase domain or in the
40
Figure 11. In-gel phosphorylation of OSR1. The indicated TAP-WNK4 samples
were separated by SDS-PAGE on a 7.5% gel with 0.2 mg/ml OSR1 copolymerized
into gel. After renaturation procedure, in-gel phosphorylation was carried out in the
presence of Mg
2+
and γ
32
P-ATP and phosphorylated bands were visualized by 3 hrs
of autoradiography. Erk2 was used as negative control for phosphorylation of OSR1.
Figure 12. WNK4 expressed in different cell lines. (A) Chinese hamster ovary
cells (CHO) and african green monkey kidney cells (COS-7) were transfected as
indicated with either full-length (FL) or Δ593 truncated (Δ593) TAP-WNK4
constructs. Proteins were recovered by TAP and visualized on a silverstained 7.5%
polyacrylamide gel. (B) In-gel phosphorylation with indicated samples was carried
out as described in legend for Figure 11.
41
segment harboring the PHAII mutations does not appear to significantly change the
interaction with this kinase.
TAP-WNK4 expressed and purified from COS-7 and CHO cells and
assayed for in-gel kinase activity
To determine whether the 40 kD copurifying kinase is associated to WNK4 when
expressed in other cell lines than HEK293 cells, TAP-WNK4 was also expressed and
purified from COS-7 and CHO cells. TAP-WNK4 was recovered from COS-7 cells
at similar levels to HEK293, but low transfection efficiency in the CHO cells
resulted in much lower recovery from this cell line (Figure 12A). The in-gel kinase
assay reveals that the 40 kD kinase copurifies with WNK4 from both COS-7 and
CHO cells, confirming that this kinase can be found in cells originating from other
organisms and tissues (Figure 12B). The weak signal observed in the CHO sample is
likely due to the low yield of TAP-WNK4 from this cell line.
Identification of WNK4 associating proteins by peptide mass
fingerprinting
Two attempts at identifying proteins associated with WNK4 by peptide mass
fingerprinting were made. In the first experiment, approximately 0.5 µg of purified
TAP-WNK4 was separated on a 7.5% polyacrylamide gel and stained with
coomassie blue. Eight coomassie stained bands were excised from the gel and sent
off to the Biomolecular Mass Spectrometry Core Laboratory at University of
42
Louisville for identification by MALDI-TOF MS analysis of tryptic peptides. Four
of these bands appear to be degradation products of WNK4, two corresponded to
heat shock cognate protein 70 (HSC70), a molecular chaperone not believed to
possess kinase activity, and the two remaining bands were identified as tubulin and
adenylate kinase, a phosphotransferase which catalyzes the formation of ATP from
ADP. None of these proteins have been previously identified as protein kinases. In a
second experiment, the amount of purified TAP-WNK4 loaded on to the 7.5%
polyacrylamide gel was increased to approximately 12 µg. The gel was stained with
coomassie blue and a gel slice encompassing the size interval 35-55 kD was excised
and sent off for identification by nano-bore LC/MS analysis of tryptic peptides in the
Pasarow Mass Spectrometry Laboratory at UCLA. Initial sample separation on a
nano-bore HPLC instrument allows very low flow rates resulting in better separation
and, in combination with subsequent mass spectrometry analysis, greater overall
sensitivity in detecting low abundance peptides. A majority of the peptides identified
with this method originated from different parts of WNK4. They were not confined
to the kinase domain as might be expected if the 40 kD copurifying kinase was in
fact a degradation product of WNK4 itself. The second most abundant source of
peptides was elongation factor 1-alpha, an abundant protein that plays an important
part in protein translation and is also believed to interact with the cytoskeleton by
binding and bundling actin filaments and microtubules (18). There is no evidence to
suggest that EF1-α has kinase activity of its own. The peptide mass fingerprinting
43
experiments have thus not been able to identify a likely candidate for the 40 kD
copurifying kinase that can be observed in the in-gel kinase assay.
Studies of truncated WNK4
In-gel kinase activity of WNK4 with c-terminal truncations Δ593, Δ500
and Δ443
In an effort to characterize the interaction between WNK4 and the copurifying
kinase, three c-terminal truncations were introduced into the TAP-WNK4 ORF and
the expressed proteins were assayed for in-gel kinase activity. In addition, removing
parts of the protein that are potentially involved in autoinhibition may activate kinase
activity, as suggested in a previously published report on WNK1 (30). The Δ593
truncation creates a protein that consists of amino acids Met1-Asp593 and includes
the n-terminal kinase domain, a putative autoinhibitory domain (AID), the first
coiled-coil region and the stretch of acidic residues that harbors the PHAII mutations
(Figure 13). The Δ500 truncation cleaves WNK4 behind the putative autoinhibitory
domain and the Δ443 truncated protein consists solely of the
kinase domain. These three proteins were expressed in HEK293 cells and recovered
by tandem affinity purification (Figure 14A). In-gel phosphorylation experiments
with OSR1 as substrate showed a marked reduction in the signal for the 40 kD
copurifying kinase for Δ593 and Δ500 and no signal for Δ443 (Figure 14B). The c-
44
Figure 13. C-terminal truncations of TAP-WNK4. The Δ-symbol followed by a
number indicates the amino acid in the WNK4 polypeptide sequence behind which
the protein has been truncated. FL denotes full-length protein.
Figure 14. C-terminally truncated WNK4 proteins expressed and assayed for
in-gel kinase activity. (A) The indicated TAP-WNK4 protein, expressed and
recovered from HEK293 cells, was separated on a 7.5% polyacrylamide gel and
visualized by silverstaining. (B) Assay of in-gel phosphorylation of OSR1 (0.2
mg/ml copolymerized into gel) by full-length and truncated TAP-WNK4 proteins.
32
P-phosphorylated proteins were visualized by 3 hours of autoradiography.
45
terminal amino acids 594-1222 appear to be important for strong interaction with the
copurifying kinase, but amino acids 443-593 allow limited interaction to take place.
Kinase activity of WNK4 Δ593 mutants
A weak signal at roughly 90 kD was observed when TAP-WNK4 Δ593 was assayed
for in-gel phosphorylation of OSR1 (Figure 14B). This corresponds well with the
position of the Δ593 protein in the silver stained gel (Figure 14A) and additional
experiments were set up to pursue the possibility that this truncation has resulted in
the activation of WNK4 kinase activity. PHAII associated mutations E559K, D561A
and Q562E and kinase domain mutations K183M and D318A, individually or
combined, were introduced into the Δ593 expression construct. The expressed
proteins (Figure 15A) were assayed for both in vitro (Figure 15B) and in-gel
phosphorylation (Figure 15C) of OSR1. The 90 kD band in the in-gel kinase assay
appears not to be affected by any of the introduced kinase domain or PHAII
mutations, suggesting that this low intensity signal is the result of non-specific
32
P
associated with the Δ593 protein rather than stemming from WNK4 kinase activity.
Surprisingly in vitro phosphorylation of OSR1 is observed for all Δ593 mutants
assayed, including the double kinase domain mutant reported to be kinase dead (24).
The most likely explanation is that the 40 kD copurifying kinase, detected at low
levels for all mutants in the in-gel assay, mediates the majority of this
phosphorylation. It is possible that specific WNK4 kinase activity cannot be
46
Figure 15. Kinase activity of TAP-WNK4 Δ593 mutants. (A) The indicated TAP-
WNK4 Δ593 mutant protein, expressed and recovered from HEK293 cells, was
separated on a 7.5% polyacrylamide gel and visualized by silverstaining. (B) Δ593
mutants were assayed for in vitro phosphorylation of OSR1 and phosphorylated
proteins, separated by SDS-PAGE, were visualized by 3 hours of autoradiography.
The upper band of the OSR1 doublet migrates at the same apparent molecular mass
as the Δ593 protein. A sample derived from vector transfected cells was used as
negative control (Vector). (C) In-gel kinase assay was carried out with 0.2 mg/ml
OSR1 copolymerized into a 7.5% polyacrylamide gel. Phosphorylated proteins were
visualized by 3 hours of autoradiography.
47
reconstituted in the in-gel assay, but is present in the in vitro assay at levels not
detected in the background of the copurifying kinase.
N-terminal WNK4 truncation
The experiments with the c-terminally truncated WNK4 proteins suggest that amino
acids 594-1222 are important for strong interaction with the copurifying kinase. A
TAP expression construct with an n-terminal TAP-tag followed by amino acids 594-
1222 of WNK4 was created (Figure 16A). To determine whether these residues
alone are enough to coprecipitate the 40 kD kinase, TAP-WNK4 594-1222,
expressed and recovered from HEK293 cells (Figure 16B), was assayed for in-gel
phosphorylation of OSR1. The signal for the copurifying kinase is 20 times stronger
in the sample containing residues 594-1222 as compared to full length WNK4
(Figure 16C). It appears that this c-terminal part of WNK4, which lacks the kinase
domain, the putative autoinhibitory domain, the first coiled coil domain and the
acidic stretch harboring the PHAII mutations, is able to achieve even stronger
interaction with the 40 kD kinase than the full-length protein.
High salt and detergent protocol for recovery of full-length and Δ593
truncated TAP-WNK4
Agents known to disrupt protein-protein interaction were introduced into the tandem
affinity purification protocol in an effort to dissociate the copurifying kinase from
WNK4 and potentially enable us to assay WNK4 specific kinase activity. Cell lysis
48
Figure 16. In-gel kinase activity of n-terminally truncated TAP-WNK4. (A) The
594-1222 construct encodes the c-terminal 629 amino acids of WNK4. (B) Indicated
TAP-WNK4 proteins, expressed inHEK293 were visualized by silverstaining. (C)
TAP-WNK4 594-1222 was assayed for in-gel phosphorylation of OSR1 (0.2 mg/ml)
in parallel with samples of full-length and Δ593 truncated TAP-WNK4.
32
P-
phosphorylated proteins were visualized by 3 hours of autoradiography.
Figure 17. High salt and detergent protocol for recovery of full-length and Δ593
truncated TAP-WNK4. (A) TAP-WNK4 protein, expressed in HEK293 cells from
the indicated construct, was recovered by regular (Reg) or high salt and detergent
(HSD) protocol and visualized by silverstaining. (B) Full-length (FL) and Δ593
truncated TAP-WNK4, prepared in parallel by regular or HSD protocol, was assayed
for in-gel phosphorylation of OSR1 (0.2 mg/ml). TAP-WNK4 proteins harboring the
KD double mutation, but otherwise treated the same, were assayed in parallel.
32
P-
phosphorylated bands were visualized by overnight autoradiography.
49
was performed in a buffer containing detergent and high sucrose concentration (see
methods) and extra wash steps were added using the lysis buffer supplemented with
high NaCl concentration. The levels of TAP-WNK4 recovered with the high salt and
detergent (HSD) protocol were similar to the regular protocol (Figure 17A). In-gel
phosphorylation experiments revealed a greatly reduced signal for the 40 kD
copurifying kinase in the samples treated with the HSD protocol, both for full-length
and Δ593 truncated WNK4 (Figure 17B). This indicates that high salt and detergent
disrupts the interaction between WNK4 and the associated kinase, without
significantly affecting the interaction with the purification resin.
Led by the observation that the HSD protocol significantly reduced the
amount of copurifying kinase, we decided to repeat the in vitro kinase assay of full-
length WNK4 with OSR1 as substrate. WNK4 harboring kinase domain mutations
K183M, D318A and KD, phosphorylates OSR1 to a similar but significantly lower
extent than the wild type protein (Figure 18A). Specific kinase activity of WNK4
was represented by a signal that was roughly 3-fold higher for wild type than for
either kinase domain mutant (Figure 18C). This observation provides strong
evidence in support of the hypothesis that full-length WNK4 has intrinsic kinase
activity and that OSR1 is a target of full-length WNK4 phosphorylation.
WNK4 specific kinase activity towards OSR1 was also observed in the Δ593
truncated protein (Figure 18B), with the wild type signal roughly 5-fold higher than
the kinase domain mutants (Figure 18D). The most likely explanation for the higher
50
Figure 18. In vitro kinase assay of full-length and Δ593 truncated TAP-WNK4
mutants prepared by HSD protocol. (A) and (B) The indicated TAP-WNK4
proteins were assayed for in vitro phosphorylation of OSR1.
32
P-phosphorylated
bands were visualized by 1 hour (A) or 3 hours (B) of autoradiography. As noted
before, the upper band of the OSR1 doublet migrates at the same apparent molecular
mass as the Δ593 protein. (C) and (D) OSR1 phosphorylation was quantified by
scintillation counting of phosphorylated bands (both bands of OSR1 doublet in (A)
and lower OSR1 band in (B)), excised from the dried gel.
32
P-incorporation was
calculated in fmols of phosphate per µg of protein and statistical significance was
determined by repeated measures ANOVA from three sets of experiments for each
full-length and Δ593 protein. OSR1 phosphorylation by WNK4 mutants was plotted
as a percentage of wild type for each set of experiments and represented by mean ±
standard deviation.
51
signal to noise ratio observed when measuring WNK4 specific phosphorylation of
OSR1 by the Δ593 protein is that it appears to have less of the copurifying kinase
associated to it than the full length protein (Figure 17B).
The effects of the PHAII mutations on the WNK4 specific in vitro
phosphorylation of OSR1 were also measured. Full-length and Δ593 truncated
WNK4 harboring the PHAII associated mutations E559K, D561A and Q562,
prepared with the HSD protocol, was subjected to the in vitro kinase assay. No
statistically significant effect of the PHAII mutations could be detected for either the
full-length (Figure 18C) or Δ593 protein (Figure 18D), suggesting that changes to the
WNK4 mediated phosphorylation of OSR1 is not one of the primary
pathophysiological mechanisms of PHAII.
All three kinase domain mutations K183M, D318A and KD reduced OSR1
phosphorylation by the full-length protein as observed on the autoradiogram, but the
170 kD phosphorylated band corresponding to full-length WNK4 is not affected by
these mutations (Figure 18A). This suggests that the majority of this phosphorylation
is carried out by residual copurifying kinase rather than by WNK4 itself
(autophosphorylation).
52
Chapter 5: Discussion
Full-length WNK4 has kinase activity
Several groups have performed in vitro phosphorylation studies employing truncated
forms of WNK4 with varying results (1, 16, 24, 25, 31), but there are no reports
where the full-length protein has been successfully expressed and assayed. We report
here that full-length WNK4, expressed and purified from HEK293 cells, specifically
phosphorylates the substrate OSR1 (Figure 18A), providing the first evidence that
full-length WNK4 has kinase activity. There are several reasons for choosing to
employ truncated forms of kinases when performing in vitro studies. These include
1) the fact that smaller proteins are easier to express heteorologously and 2) the
removal of potential autoinhibitory domains that might interfere with activity studies.
The full-length WNK4 sequence includes a predicted autoinhibitory domain (24), but
this does not prevent us from observing substrate phosphorylation, calling in to
question what, if any, role it plays in the activation of WNK4. In the first report
identifying WNK1 as a serine-threonine kinase, full-length WNK1 also exhibits
substrate and autophosphorylation in vitro (29) despite the presence of an
autoinhibitory domain. In a later study from the same group three truncated WNK1
proteins are assayed for kinase activity (30): 1) WNK1(1-491), consisting solely of
the kinase domain, produces strong MBP and autophosphorylation, 2) WNK1(1-
555), which includes the autoinhibitory domain, has almost no activity and 3)
WNK1(1-639), which includes the first coiled-coil domain, displays intermediate
53
kinase activity. The report suggests that WNK1 kinase activity is suppressed by the
autoinhibitory domain, but that this suppression is alleviated by the presence of the
first coiled-coil domain, thus explaining why the WNK1(1-693) protein appears
more active than WNK1(1-555). These three WNK1 proteins correspond well with
our WNK4 Δ443, Δ500 and Δ593 proteins respectively. Our full-length and Δ593
proteins are both able to phosphorylate OSR1 in vitro (Figure 18A and B), well in
agreement with the WNK1 data. However, all three truncated WNK4 proteins failed
to exhibit in-gel kinase activity (Figure 14B) leaving the question of potential
activation by removing the predicted autoinhibitory domain unanswered.
Kinase domain mutants abolish WNK4 kinase activity
In light of the fact that kinase domain mutants K183M, D318A and KD all reduce
OSR1 phosphorylation to a similar extent, we conclude that these mutations all
abolish kinase activity and that the residual signal stems from background
phosphorylation by the copurifying kinase. It is of interest to investigate the effects
of mutating catalytically conserved residues, individually and in combination, in
order to validate their predicted importance for the proper function of the protein. He
et al show that the individual WNK1 kinase domain mutations K233D and D386K
abolish the ability of WNK1 to phosphorylate OSR1 in vitro and inhibit ROMK1 in
oocytes (8). The double mutant K233D/D386K was also unable to phosphorylate
OSR1 in vitro, but it was able to inhibit ROMK1 to the same extent as the wild type
protein. The authors suggest that these two residues form a salt bridge that is
54
important for the structural integrity of the kinase domain and that the individual
kinase domain mutants fail to inhibit ROMK1 as a result of a conformational change,
disrupting protein-protein interaction, rather than loss of kinase activity. In this
scenario the double mutation reconstitutes the salt bridge and the conformation of the
kinase domain, allowing WNK1 to participate in protein-protein interaction but not
phosphorylate OSR1. Our results confirm the predictions made in various published
reports that the D318A kinase domain mutation abolishes WNK4 kinase activity, but
this does not necessarily mean that the loss of kinase activity is the primary cause for
the reported observation in these reports.
WNK4 autophosphorylation cannot be detected
As noted in the results section, we have been unable to detect significant
autophosphorylation by WNK4 in our preparation in the in vitro assay (Figure 18A).
It is possible that WNK4 autophosphorylates to some extent, but that the
phosphorylation by the copurifying kinase is so much stronger that WNK4 specific
activity cannot be detected over background, even when the protein is prepared with
the HSD protocol. Lenertz et al report that a kinase domain fragment of WNK4
autophosphorylates moderately at serine 332 (13), which corresponds to a serine
residue in the activation loop of WNK1 found to be critical for WNK1 activation
(29). It is possible that the coprecipitating kinase phosphorylates our full-length
WNK4 protein at this residue, contributing to the activation of the protein, but the
observation that the kinase domain mutants do not change the strength of this band
55
suggests that there is no significant autophosphorylation in our protein.
Autophosphorylation has been reported for expressed full-length WNK1 (29)
suggesting that these two proteins behave differently with respect to their
autophosphorylation activities.
PHAII associated mutations do not affect WNK4 kinase activity
Our findings exclude any strong activating or deactivating effect of PHAII mutations
(Figures 18C and 18D), though very small effects may not be detectable over the
background of the copurifying kinase.. The lack of detectable changes in OSR1
phosphorylation by introducing the PHAII mutations argues against changes to the
intrinsic kinase activity of WNK4, however we cannot discount potential effects on
the interaction with and phosphorylation of other, as of yet, undiscovered
physiological substrates. Many protein kinases form active complexes with their
substrates and PHAII mutations may alter the way that WNK4 participates in such
complexes. Three out of four PHAII associated mutations, corresponding to the ones
investigated in this study, are found in an acidic 10 amino acid stretch starting at
glutamic acid 554 (‘EPEEPEADQH’). Clusters of acidic amino acids may confer
various properties on a protein domain including coordination of cations and
electrostatic interactions within the protein or with other proteins. It is conceivable
that this region acts as a sensor for the presence of cation species, adopting specific
conformations with functional implications for the protein depending on the
concentration of the cation in question. This could render the protein sensitive to
56
changes in pH as well as osmolyte levels and mutations to the acidic motif might
impair this property. Another possibility is that the acidic cluster is of importance for
the interaction with other protein domains. Electrostatic attraction or repulsion may
for example be critical for the specific interaction between WNK4 and another
protein. Disrupting the charge composition of the motif may change such an
interaction. There is data to support the notion that WNK4, in addition to
phosphorylating OSR1 and the closely related SPAK, also associates through direct
protein-protein interaction with these kinases (24). OSR1 and SPAK have been
reported to regulate various members of the SLC12 family of ion transport proteins
(4, 20) and it has been suggested that WNK4 may function as an upstream kinase in
a cascade with OSR1 and SPAK with the ultimate effect of altering ion transport
processes mediated by these transport proteins (24). In such a scenario the
introduction of the PHAII mutations may affect the interaction between WNK4 and
other proteins in the complex, e.g. enhancing the recruitment of some downstream
targets and/or preventing the recruitment of others. Immunoprecipitation experiments
indicate that WNK4 interacts with the SLC12 transporter NCC (27), supporting the
hypothesis that these proteins may form a complex.
The roughly 100 amino acid long conserved c-terminal domain of OSR1 and
SPAK has been identified as a substrate recognition sequence that mediates the
interaction with both upstream and downstream targets that contain a consensus
RFxV motif, including NKCC1 (20) and WNK4 (19). Two groups report that
WNK4, via phosphorylation of OSR1 (24) or SPAK (5, 24), is able to mediate the
57
phosphorylation and activation of NKCC1, but there are inconsistencies regarding
the importance of the above mentioned RFxV motif. In WNK4 the RFxV motif is in
the c-terminal part of the protein, starting at R996. Gagnon et al report that WNK4
F997A is unable to activate NKCC1 when coexpressed with SPAK in oocytes, while
Vitari et al are able to show in vitro phosphorylation and activation of OSR1 and
SPAK by WNK4 Δ593, which lacks the entire c-terminal half of the protein. In our
experimental setting WNK4 Δ593 phosphorylates OSR1, implying that the RFxV
motif is not essential for the ability of WNK4 to phosphorylate its substrates.
WNK4 associates with a 40 kD kinase when expressed in mammalian
cells
We found that an unidentified 40 kD kinase, which strongly phosphorylates OSR1
(Figure 11), is associated to the c-terminal part of WNK4 (Fig 16C) expressed in
various cell lines (Figure 12B). Its presence was detected by employing the in-gel
kinase assay, where the components associated with purified WNK4 were separated
by denaturing SDS-PAGE and then renatured in the gel. The 40 kD kinase was the
only component of the WNK4 preparation, which could be reconstituted into an
active conformation that allowed it to phosphorylate OSR1 (Figure 11). Truncated
WNK4 proteins also failed to produce a phosphorylated band (Figure 14B),
indicating that WNK4 either cannot regain an active conformation or that it is unable
to phosphorylate OSR1 in this assay. In previous experiments with the in-gel
phosphorylation assay we were able to successfully renature recombinant Erk2 in a
58
gel containing MBP and detect a phosphorylated band at the expected molecular
weight of Erk2 (data not shown). A WNK4 sample in the same gel failed to produce
any phosphorylated bands.
The interaction between WNK4 and the coprecipitating kinase appears to be
strong because we still detect the kinase after extended washing procedures with
high salt and detergent. No phosphorylated bands can be detected in the 40 kD
region in the in vitro kinase assay (Figures 10B and 10C), indicating that this kinase
neither autophosphorylates nor is phosphorylated by WNK4 or other WNK4
associated kinases. Our experiments with truncated WNK4 proteins show that the
amount of coprecipitating protein is greatly reduced when the c-terminal half of
WNK4 is not present (Figure 14B). SDS-PAGE gels with full-length and truncated
WNK4 in parallel lanes fail to show coomassie stained bands in the 40 kD size range
that are stronger for the full-length protein, which interacts strongly with the 40 kD
kinase, and weaker for the truncated protein, which interacts weakly. Not even with
the more sensitive silver staining protocol can clear candidate bands be discerned in
the purified full-length WNK4 preparation. Attempts at scaling up the amounts of
WNK4 separated by SDS-PAGE and subsequently analyzing the 40 kD size range
by ultra sensitive peptide mass fingerprinting experiments also failed to produce any
clear candidates in this region.
The importance of the c-terminal part of WNK4 for strong interaction with
the copurifying kinase is illustrated by the ability of amino acids 594-1222 to
precipitate this protein (Figure 16C). This part of WNK4 lacks any conserved
59
structural motifs except for the coiled-coil domain between residues 1110-1140 and
the RFxV motif starting at 996. Coiled-coil domains are often involved in protein-
protein interaction and protein oligomerization. The first coiled-coil domain of
WNK1 (analogous to the one found in WNK4 between residues 500-530) has been
suggested to be involved in tetramerization of WNK1 (30), but there are no reports
implicating the c-terminal coiled coil domain of WNK4 in oligomerization or protein
interaction. The RFxV consensus motif is essential for interaction with SPAK and
OSR1 and was originally described in cation chloride cotransporters of the SLC12
family (20). This motif is recognized for its specific ability to interact with the
conserved c-terminal domain of SPAK and OSR1 and not as a general protein
interaction motif. There is a possibility that the WNK4 RFxV motif may mediate
interaction with proteins containing domains similar to the conserved c-terminal
domain of SPAK and OSR1, but nothing has been reported to that effect. The
likelihood that the 40 kD copurifying kinase corresponds to endogenous SPAK or
OSR1 is not very high, because these proteins have considerably higher molecular
weights.
As to the identity of this coprecipitating kinase, our efforts have not yet
produced any potential candidates. Many serine-threonine kinases can be found in
the 40 kD size range and the likelihood is that this kinase is ubiquitous as it is found
to be associated to WNK4 in cell lines derived from different tissues (kidney and
ovary) and organisms (human, green monkey and Chinese hamster). The proteins
identified in the peptide mass fingerprinting experiments were either abundant
60
cellular proteins or peptides originating from WNK4 itself. If the 40 kD copurifying
protein were in fact a degradation product of WNK4, it would be expected that the
kinase domain mutations abolish the signal detected in the in-gel kinase assay.
61
Chapter 6: Overall Conclusions
In conclusion we have shown that full-length WNK4 has specific protein kinase
activity and is able to phosphorylate OSR1 in vitro. We also demonstrate that the
PHAII associated mutations E559K, D561A and Q562E have no significant effect on
WNK4 phosphorylation of OSR1. A 40 kD unidentified protein kinase, that
phosphorylates OSR1, is strongly associated with the C-terminal half of WNK4
when it is expressed in various mammalian cell lines. Overall this work has
contributed to the understanding of the mechanisms by which WNK4 exerts
influence over its targets.
62
Chapter 7: Future Directions
Our aim with this study was to characterize WNK4 kinase activity and the effects of
disease causing mutations on this activity. The experiments presented here have
provided initial insight into the functional mechanisms governing substrate
phosphorylation by WNK4. In order to gain a complete picture additional
experiments are necessary.
The importance of the predicted autoinhibitory domain in WNK4 could be
further investigated by performing in vitro phosphorylation studies with our Δ443,
Δ500 an Δ593 proteins and compare the levels of kinase activity.
Further analysis of the kinase domain mutants may allow us to differentiate
between effects stemming from loss of interaction rather than loss of kinase activity.
By producing the kinase domain mutants K183D and D318K, individually or in
combination, and combining coimmunoprecipitation studies of WNK4 and OSR1
with in vitro phosphorylation data, we may determine the importance of protein-
protein interaction for the phosphorylation reaction.
The identity of the residue(s) on WNK4 that is phosphorylated in the in vitro
phosphorylation experiments may be determined by a candidate site directed
mutagenesis approach, starting with serine 332, which has been implicated
previously (13).
A rational approach to determine the identity of the 40 kD copurifying kinase
may include designing a c-terminal domain protein of WNK4 that can be
63
immobilized on a column and perform large scale recovery experiments from
mammalian cell lysates. An immunochemical candidate approach, using antibodies
directed against candidate proteins is a possibility, but is made difficult by the large
number of kinases that exist in the 40 kD size range.
64
References
1. Anselmo AN, Earnest S, Chen W, Juang YC, Kim SC, Zhao Y, and Cobb
MH. WNK1 and OSR1 regulate the Na+, K+, 2Cl- cotransporter in HeLa
cells. Proc Natl Acad Sci U S A 103: 10883-10888, 2006.
2. Disse-Nicodeme S, Achard JM, Desitter I, Houot AM, Fournier A,
Corvol P, and Jeunemaitre X. A new locus on chromosome 12p13.3 for
pseudohypoaldosteronism type II, an autosomal dominant form of
hypertension. Am J Hum Genet 67: 302-310, 2000.
3. Dorwart MR, Shcheynikov N, Wang Y, Stippec S, and Muallem S.
SLC26A9 is a Cl channel regulated by the WNK kinases. J Physiol 584: 333-
345, 2007.
4. Dowd BF and Forbush B. PASK (proline-alanine-rich STE20-related
kinase), a regulatory kinase of the Na-K-Cl cotransporter (NKCC1). J Biol
Chem 278: 27347-27353, 2003.
5. Gagnon KB, England R, and Delpire E. Volume sensitivity of cation-Cl-
cotransporters is modulated by the interaction of two kinases: Ste20-related
proline-alanine-rich kinase and WNK4. Am J Physiol Cell Physiol 290:
C134-142, 2006.
6. Gordon RD, Geddes RA, Pawsey CG, and O'Halloran MW. Hypertension
and severe hyperkalaemia associated with suppression of renin and
aldosterone and completely reversed by dietary sodium restriction. Australas
Ann Med 19: 287-294, 1970.
7. Hanks SK, Quinn AM, and Hunter T. The protein kinase family:
conserved features and deduced phylogeny of the catalytic domains. Science
241: 42-52, 1988.
8. He G, Wang HR, Huang SK, and Huang CL. Intersectin links WNK
kinases to endocytosis of ROMK1. J Clin Invest 117: 1078-1087, 2007.
65
9. Kahle KT, Gimenez I, Hassan H, Wilson FH, Wong RD, Forbush B,
Aronson PS, and Lifton RP. WNK4 regulates apical and basolateral Cl-
flux in extrarenal epithelia. Proc Natl Acad Sci U S A, 2004.
10. Kahle KT, Macgregor GG, Wilson FH, Van Hoek AN, Brown D, Ardito
T, Kashgarian M, Giebisch G, Hebert SC, Boulpaep EL, and Lifton RP.
Paracellular Cl- permeability is regulated by WNK4 kinase: Insight into
normal physiology and hypertension. Proc Natl Acad Sci U S A, 2004.
11. Kahle KT, Wilson FH, Leng Q, Lalioti MD, O'Connell AD, Dong K,
Rapson AK, MacGregor GG, Giebisch G, Hebert SC, and Lifton RP.
WNK4 regulates the balance between renal NaCl reabsorption and K+
secretion. Nat Genet 35: 372-376, 2003.
12. Lalioti MD, Zhang J, Volkman HM, Kahle KT, Hoffmann KE, Toka
HR, Nelson-Williams C, Ellison DH, Flavell R, Booth CJ, Lu Y, Geller
DS, and Lifton RP. Wnk4 controls blood pressure and potassium
homeostasis via regulation of mass and activity of the distal convoluted
tubule. Nat Genet 38: 1124-1132, 2006.
13. Lenertz LY, Lee BH, Min X, Xu BE, Wedin K, Earnest S, Goldsmith EJ,
and Cobb MH. Properties of WNK1 and implications for other family
members. J Biol Chem 280: 26653-26658, 2005.
14. Lupas A, Van Dyke M, and Stock J. Predicting coiled coils from protein
sequences. Science 252: 1162-1164, 1991.
15. Mansfield TA, Simon DB, Farfel Z, Bia M, Tucci JR, Lebel M, Gutkin
M, Vialettes B, Christofilis MA, Kauppinen-Makelin R, Mayan H, Risch
N, and Lifton RP. Multilocus linkage of familial hyperkalaemia and
hypertension, pseudohypoaldosteronism type II, to chromosomes 1q31-42
and 17p11-q21. Nat Genet 16: 202-205, 1997.
16. Min X, Lee BH, Cobb MH, and Goldsmith EJ. Crystal structure of the
kinase domain of WNK1, a kinase that causes a hereditary form of
hypertension. Structure (Camb) 12: 1303-1311, 2004.
66
17. Moriguchi T, Urushiyama S, Hisamoto N, Iemura S, Uchida S, Natsume
T, Matsumoto K, and Shibuya H. WNK1 regulates phosphorylation of
cation-chloride-coupled cotransporters via the STE20-related kinases, SPAK
and OSR1. J Biol Chem 280: 42685-42693, 2005.
18. Murray JW, Edmonds BT, Liu G, and Condeelis J. Bundling of actin
filaments by elongation factor 1 alpha inhibits polymerization at filament
ends. J Cell Biol 135: 1309-1321, 1996.
19. Piechotta K, Garbarini N, England R, and Delpire E. Characterization of
the interaction of the stress kinase SPAK with the Na+-K+-2Cl- cotransporter
in the nervous system: evidence for a scaffolding role of the kinase. J Biol
Chem 278: 52848-52856, 2003.
20. Piechotta K, Lu J, and Delpire E. Cation chloride cotransporters interact
with the stress-related kinases Ste20-related proline-alanine-rich kinase
(SPAK) and oxidative stress response 1 (OSR1). J Biol Chem 277: 50812-
50819, 2002.
21. Ring AM, Cheng SX, Leng Q, Kahle KT, Rinehart J, Lalioti MD,
Volkman HM, Wilson FH, Hebert SC, and Lifton RP. WNK4 regulates
activity of the epithelial Na+ channel in vitro and in vivo. Proc Natl Acad Sci
U S A 104: 4020-4024, 2007.
22. Schambelan M, Sebastian A, and Rector FC, Jr. Mineralocorticoid-
resistant renal hyperkalemia without salt wasting (type II
pseudohypoaldosteronism): role of increased renal chloride reabsorption.
Kidney Int 19: 716-727, 1981.
23. Verissimo F and Jordan P. WNK kinases, a novel protein kinase subfamily
in multi-cellular organisms. Oncogene 20: 5562-5569, 2001.
24. Vitari AC, Deak M, Morrice NA, and Alessi DR. The WNK1 and WNK4
protein kinases that are mutated in Gordon's hypertension syndrome
phosphorylate and activate SPAK and OSR1 protein kinases. Biochem J 391:
17-24, 2005.
67
25. Wang Z, Yang CL, and Ellison DH. Comparison of WNK4 and WNK1
kinase and inhibiting activities. Biochem Biophys Res Commun 317: 939-944,
2004.
26. Wilson FH, Disse-Nicodeme S, Choate KA, Ishikawa K, Nelson-Williams
C, Desitter I, Gunel M, Milford DV, Lipkin GW, Achard JM, Feely MP,
Dussol B, Berland Y, Unwin RJ, Mayan H, Simon DB, Farfel Z,
Jeunemaitre X, and Lifton RP. Human hypertension caused by mutations
in WNK kinases. Science 293: 1107-1112, 2001.
27. Wilson FH, Kahle KT, Sabath E, Lalioti MD, Rapson AK, Hoover RS,
Hebert SC, Gamba G, and Lifton RP. Molecular pathogenesis of inherited
hypertension with hyperkalemia: the Na-Cl cotransporter is inhibited by wild-
type but not mutant WNK4. Proc Natl Acad Sci U S A 100: 680-684, 2003.
28. Wooten MW. In-gel kinase assay as a method to identify kinase substrates.
Sci STKE 2002: PL15, 2002.
29. Xu B, English JM, Wilsbacher JL, Stippec S, Goldsmith EJ, and Cobb
MH. WNK1, a novel mammalian serine/threonine protein kinase lacking the
catalytic lysine in subdomain II. J Biol Chem 275: 16795-16801, 2000.
30. Xu BE, Min X, Stippec S, Lee BH, Goldsmith EJ, and Cobb MH.
Regulation of WNK1 by an autoinhibitory domain and autophosphorylation.
J Biol Chem 277: 48456-48462, 2002.
31. Yamauchi K, Rai T, Kobayashi K, Sohara E, Suzuki T, Itoh T, Suda S,
Hayama A, Sasaki S, and Uchida S. Disease-causing mutant WNK4
increases paracellular chloride permeability and phosphorylates claudins.
Proc Natl Acad Sci U S A 101: 4690-4694, 2004.
32. Yang CL, Angell J, Mitchell R, and Ellison DH. WNK kinases regulate
thiazide-sensitive Na-Cl cotransport. J Clin Invest 111: 1039-1045, 2003.
33. Yang CL, Liu X, Paliege A, Zhu X, Bachmann S, Dawson DC, and
Ellison DH. WNK1 and WNK4 modulate CFTR activity. Biochem Biophys
Res Commun 353: 535-540, 2007.
68
34. Yang SS, Morimoto T, Rai T, Chiga M, Sohara E, Ohno M, Uchida K,
Lin SH, Moriguchi T, Shibuya H, Kondo Y, Sasaki S, and Uchida S.
Molecular Pathogenesis of Pseudohypoaldosteronism Type II: Generation
and Analysis of a Wnk4(D561A/+) Knockin Mouse Model. Cell Metab 5:
331-344, 2007.
35. Zoller MJ, Nelson NC, and Taylor SS. Affinity labeling of cAMP-
dependent protein kinase with p-fluorosulfonylbenzoyl adenosine. Covalent
modification of lysine 71. J Biol Chem 256: 10837-10842, 1981.

Abstract (if available)

Abstract Mutations in WNK1 (with no K [lysine] 1) and WNK4 protein kinases cause pseudohypoaldosteronism type II (PHAII), a rare genetic disorder that features high blood pressure and elevated serum potassium levels. Potential targets of WNK4 regulation have been identified by various approaches, but the mechanism by which it influences its targets is still poorly understood. In an effort to characterize its kinase activity, we expressed and purified full-length and truncated forms of WNK4 from HEK293 cells. Due to endogenous kinases binding non-specifically to the protein G resin when immunoprecipitating WNK4 from HEK293 cells, we decided to recover the protein by the tandem affinity purification (TAP) method. Our phosphorylation experiments identified a 40 kD kinase that associates specifically to the C-terminal half of WNK4 and is able to phosphorylate WNK4 and the WNK4 substrate, oxidative stress response kinase 1 (OSR1). This kinase copurifies with WNK4 in the mammalian cell lines HEK293, COS-7 and CHO, but could not be identified by peptide mass fingerprinting. By modifying the TAP protocol to include high salt and detergent washes, we were able to dissociate a large fraction of the 40 kD kinase and measure specific in vitro phosphorylation of OSR1 by full-length and delta593 truncated WNK4. The PHAII associated mutations E559K, D561A and Q562E had no significant effect on this phosphorylation. This study contributes important information on the mechanism by which WNK4 may influence its targets by showing that full-length WNK4 has kinase activity and also what effects the PHAII mutations have on this activity.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Novel functions of the macula densa in renal physiology and disease

PDF

Characterization of rice-derived proinsulin-transferrin fusion protein and its oral bioactivity as a basal insulin analogue

PDF

Integrated regulation of transporters along the nephron to maintain electrolyte and fluid homeostatis

PDF

Vpr-binding protein negatively regulates p53 by site-specific phosphorylation through intrinsic kinase activity

PDF

The pivotal role of AMP-activated protein kinase in the regulation of fatty acid metabolism in skeletal muscle

PDF

Effect of claudins on cyst formation in MDCK cells

PDF

Developing peptide and antibody-mimetic ligands for the cell surface receptors β2AR and DC-SIGN

PDF

Flipped zinc transporter 8 (ZnT8): a novel approach to characterize zinc transport and its possible relevance to type 2 diabetes

PDF

C-jun N-terminal Kinase (JNK) mediated inhibition of Pyruvate Dehydrogenase (PDH) activity and its effect on mitochondrial metabolism during brain aging

PDF

Role of GRP78 and its novel cytosolic isoform GRP78va in regulating the unfolded protein response in cancer

PDF

The effect of low protein intake on IGF-1 and cancer

PDF

Site-specific PEGylation of recombinant immunotoxin αFAP-PE38 for half-life extension

PDF

A novel role for hypoxia-inducible factor-1alpha (HIF-1alpha) in the regulation of inflammatory chemokines and leukotriene expression in sickle cell disease

PDF

Biochemical modifications and regulation of the gastric H,K-ATPase

PDF

Structure-function studies on the mammalian intestinal dipeptide transporter hPEPT1

PDF

Visual arrestin interactions with clathrin adaptor AP-2 regulate photoreceptor survival

PDF

Enhancing the anti-HIV potency of eCD4-Ig by unnatural amino acid mutagenesis

PDF

The molecular mechanism of GRP78 translocation from the endoplasmic reticulum to the cell surface

PDF

Systematic analysis of single nucleotide polymorphisms in the human steroid 5-alpha reductase type I gene

PDF

Characterization of three novel variants of the MAVS adaptor

Asset Metadata

Creator Ahlstrom, Erik Robert (author)

Core Title Kinase activity of the pseudohypoaldosteronism type II gene product, WNK4

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Physiology

Publication Date 01/25/2008

Defense Date 11/21/2007

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag hypertension,OAI-PMH Harvest,protein kinase,pseudohypoaldosteronism type 2

Language English

Advisor Yu, Alan S.L. (committee chair), Farley, Robert A. (committee member), McDonough, Alicia A. (committee member), Peti-Peterdi, Janos (committee member), Shen, Wei-Chiang (committee member)

Creator Email ahlstrom@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m991

Unique identifier UC1470383

Identifier etd-Ahlstrom-20080125 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-594158 (legacy record id),usctheses-m991 (legacy record id)

Legacy Identifier etd-Ahlstrom-20080125.pdf

Dmrecord 594158

Document Type Dissertation

Rights Ahlstrom, Erik Robert

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu