| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
From the Department of Medicine,*
Mount Sinai School of
Medicine, New York, New York; the Division of
Cardiology,
University of Maryland School of
Medicine, Baltimore, Maryland; and the Department of
Biology,
The Johns Hopkins University,
Baltimore, Maryland
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1, 3,
ß1, and ß2 subunits of NaK-ATPase. High levels of
expression of 1 and ß1 messenger RNA were detected in ADPKD and
age-matched normal adult kidneys in vivo, whereas
ß2 messenger RNA was detected only in ADPKD kidneys. Western blot
analysis and immunocytochemical studies showed that, in normal
adult kidneys, peptide subunit-specific antibodies against 1
and ß1 localized to the basolateral membranes of normal renal
tubules, predominantly thick ascending limbs of Henle’s loop.
In ADPKD kidneys, 1 and ß2 subunits were localized to the
apical epithelial cell membranes, whereas ß1 was distributed
throughout the cytoplasm and predominantly in the endoplasmic
reticulum, but was not seen associated with cystic epithelial
cell membranes or in cell membrane fractions. Polarizing,
renal-derived epithelial Madin Darby canine kidney cells,
stably expressing normal or N-terminally truncated chicken ß1
subunits, showed selective accumulation in the basolateral
Madin Darby canine kidney cell surface, whereas
c-myc epitope-tagged chicken ß2 or human ß2
subunits accumulated selectively in the apical cell surface.
Similarly, human ADPKD epithelial cell lines, which
endogenously expressed 1 and ß2 NaK-ATPase subunits,
showed colocalization at the apical cell surface and coassociation by
immunoprecipitation analysis. These results are consistent with a model
in which the additional transcription and translation of the ß2
subunit of NaK-ATPase may result in the apical mislocalization of
NaK-ATPase in ADPKD cystic epithelia.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The kidney is an organ that functions to reabsorb essential fluid and ions, and this is facilitated by the strictly polarized distribution of numerous transporters, enzymes, and antigens distributed along the 10 distinct segments of the nephron in an epithelial cell type-specific fashion. The polarization of membrane proteins is a critical component in the differentiation of renal tubule epithelial cells and is largely established in the human metanephric kidney before birth.
NaK-ATPase is an important, complex membrane protein of all cells including polarized renal epithelial cells. It exchanges sodium for potassium and functions as the ubiquitous electrogenic sodium pump in all animal cells and is responsible for the active transport of sodium ions out of cells. NaK-ATPase is very highly expressed in renal tubule epithelia, where it is restricted to the basolateral plasma membranes. This segregated distribution is associated with vectorial (apical-to-basal) transport of sodium and is a critical feature in the generation of sodium ion gradients, which in turn regulate other Na-coupled ion movements as well as osmotic fluid reabsorption. A major and pathophysiologically significant abnormality in ADPKD cystic epithelia is the mispolarization of a fully functional NaK-ATPase to the apical membranes of cyst epithelia. This in turn is associated with basal-to-apical sodium ion transport leading to luminal fluid accumulation.8,12,16 Some but not all other membrane and membrane-associated proteins show aberrant polarization in ADPKD cystic epithelia, including the epidermal growth factor receptor (EGFR), calpactin, ankyrin, fodrin, and E-cadherin.5,6
In its mature form, NaK-ATPase is a heterodimeric, complex, integral
membrane protein composed of two and two ß subunits. NaK-ATPase
subunits are multimembrane-spanning, nonglycosylated proteins of 98
to 100 kd, and they contain the catalytically significant
phosphorylation, ATP, and ouabain-binding sites. NaK-ATPase ß
subunits, by contrast, span the membrane only once, are variably
glycosylated on their exterior aspect, accordingly range in molecular
mass from 35 kd to 55 kd, and are absolutely required for the
functioning of the assembled NaK-ATPase complex. At least three
different isoforms of NaK-ATPase subunits (1, 2, 3) and
three of NaK-ATPase ß subunits (ß1, ß2, ß3) have been cloned
from a variety of species, organs, and developmental
stages.17-21
Although the functional implications of these
different isoforms remain obscure, the 1 and ß1 subunits are the
most ubiquitous and are characteristic of normal kidneys, whereas 3
expression is characteristic in brain, and 2 and 3 are frequently
expressed in fetal tissues.22,23
Of particular interest,
the NaK-ATPase ß2 (ß-adhesion molecule on glia) subunit is
not only tissue restricted and developmentally regulated with high
levels in human fetal liver and rat brain but it also functions as a
neural adhesion molecule on glia and promotes neurite outgrowth,
suggesting a role in normal differentiation of neurons in
vivo.24-27
The specific mechanisms underlying the sorting and polarized delivery of NaK-ATPase in renal epithelia are not fully understood. Early studies suggested direct targeting of NaK-ATPase to basolateral membranes of a polarized renal epithelial cell line, Madin Darby canine kidney (MDCK) cells, whereas others implicated indirect targeting followed by stabilization with the insoluble components of the membrane cytoskeleton, ankyrin, and fodrin.28,29 More recent studies implicate both specific basolateral sorting signals in targeting of NaK-ATPase, the lipid environment in the trans-Golgi network, and complex formation with ankyrin, fodrin, and E-cadherin at the membrane.30-32
In the kidney epithelial cell line LLC-PK1, the N-terminal region has been implicated in sorting of HK-ATPase or NaK-ATPase,33 but interactions between the ß subunits and a region in the C terminal of NaK-ATPase are also necessary for targeting.34,35 Further studies show that proteins or protein complexes may contain multiple sorting determinants (eg, 36).
The current studies have been undertaken to examine the mechanisms underlying the important mispolarization of NaK-ATPase in ADPKD epithelia. Our previous studies suggested that NaK-ATPase is specifically mistargeted to the apical plasma membranes of cystic epithelia in human ADPKD, that it is fully catalytically active and binds 125I-labeled ouabain, and that the mispolarization leads to basal-to-apical net transport of Na+ ions in membrane-grown ADPKD epithelia in vitro.8,12 In this study we examine NaK-ATPase isoform expression and distribution in vivo and in vitro and suggest that revertant or persistent expression of the fetal kidney ß2 isoform is a potentially important mechanism for the apical mispolarization in ADPKD.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cell Culture
Primary monolayer cultures of renal tubule segment-specific and ADPKD cyst epithelia were derived from normal adult proximal tubules, thick ascending limb of Henle, and collecting tubules and from ADPKD cyst-lining epithelia as previously described.37-39 Cells were grown in segment-selective media in 25 cm2 flasks coated with type I (rat tail) collagen (Collaborative Research, Lexington, MA), dissociated with collagenase (Life Technologies, Inc., Grand Island, NY), and seeded at confluent density on translucent permeable membrane supports (Transwel-COL, Costar, Cambridge, MA). For immunocytochemistry, confluent cell cultures on membranes were washed with PBS at 4°C for 3 minutes, fixed for 5 minutes in 4% paraformaldehyde in PBS, pH 7.4 at 4°C, and washed three times for 3 minutes each in cold PBS before immunostaining.
Cells and Transfections
MDCK cells were obtained from the American Type Culture Collection (Gaithersburg, MD), and a subclone, A4, was used to generate MDCK stable cell lines. Standard growth medium consisted of high-glucose Dulbecco’s modified essential medium supplemented with 10% fetal bovine serum and 50 µg/ml gentamicin.
For use in transfection, cDNAs were cloned into expression plasmid
pCB6. cDNAs encoding chicken NaK-ATPase 1 and ß2 subunits have
been described.40,41
The chicken ß2 subunit cDNA was
modified to encode a c-myc epitope11
at the ß2
subunit C terminus, which is in the ßsubunit ectodomain. DNA encoding
the human NaK-ATPase ß2 subunit was cloned from a human retinal cDNA
library in 
gt10, kindly provided by Jeremy Nathans (The
Johns Hopkins University School of Medicine, Baltimore, MD). A probe
for the human ß2 subunit cDNA clones was generated by polymerase
chain reaction (PCR), using primers based on the nucleotide sequence
encoding the human ß2 subunit.33
This probe was used to
isolate a full-length clone (~1.5 kb) from the library. Identity of
the clone was confirmed by nucleotide sequencing. The 5' end of the
clone began 80 bp upstream from the reported nucleotide sequence of
human ß2 subunit cDNA (GenBank accession number M81181) and ended in
the 3' untranslated region.
MDCK cells were transfected in 60-mm tissue culture dishes. Lipofectin (Life Technologies) was used to introduce the plasmids into the MDCK cells. Clones of cells resistant to G418 (400 µg/ml) were isolated and screened for expression of exogenous NaK-ATPase ß subunits. For screening, the cells were plated on cover slips at high density and induced for 48 h in 10 mmol/L butyrate. Induced cells were then screened for expression of the exogenous ß subunit by immunofluorescence microscopy.
Immunocytochemistry
For light microscopy, paraffin-embedded tissue sections on glass
slides were first dewaxed and rehydrated through a graded series of
ethanols. Cells and tissues were then incubated in 0.3%
H2O2 in methanol to block endogenous peroxidase
activity, followed by incubation with 10% normal goat serum in PBS for
20 minutes at room temperature in a humidified atmosphere. Cells and
sections were incubated for 45 minutes at room temperature in a
humidified chamber with the following primary antibodies: polyclonal
anti-NaK-ATPase subunit raised in rabbits against the whole
molecule purified from dog, by courtesy of WJ Nelson; and antichicken
NaK-ATPase subunit monoclonal antibodies (mAbs) (5; 2F,
F12 available through the Development Studies Hybridoma Bank, Iowa
City, IA); or polyclonal anti-1 (#1300), 3 (#1301), ß1 (#1303),
and ß2 (#1305) raised in rabbits against isozyme-specific peptides
and affinity purified on a peptide column (Immuno Dynamics, La Jolla,
CA). All anti-1 antibodies recognized a single 98-kd band and showed
identical staining patterns in immunocytochemistry of kidney tissues.
Primary antibodies were diluted in PBS containing 2% bovine serum
albumin (BSA) (1:100–1:500), washed three times in PBS-Tween 20
(0.02%), incubated for 45 minutes with biotinylated goat anti-rabbit
immunoglobulin G (IgG; Vector Laboratories, Burlingame, CA), washed
twice for 5 minutes each in PBS-Tween and once for 5 minutes in PBS,
incubated for 45 minutes with avidin-biotin peroxidase (Vectastain
Elite, Vector Laboratories), and washed for 5 minutes in PBS, followed
by two washes of 5 minutes each in Tris-buffered saline. Color
development was carried out for 10 to 45 minutes, using
aminoethylcarbazole as substrate. Sections and cells were mounted in
Aquamount (Polysciences, Niles, IL) and viewed under a Nikon
FXA-Microphot equipped with Nomarski optics.
For electron microscopy and enzyme and immunocytochemistry, tissue
sections and cells grown on membranes were fixed with 2%
paraformaldehyde or 1% glutaraldehyde in PBS and incubated either for
immunoperoxidase localization of antigens or in reaction media for
localization of NaK-ATPase enzymatic reaction products after
postcoupling with lead.42
Pre-embedding immunoperoxidase
reactions were carried out by 1) incubation for 45 minutes at room
temperature in 1:50 to 1:500 dilutions of 1, ß1, and ß2 peptide
antibodies in PBS with BSA; 2) washing in three changes of PBS for 5
minutes each at 4°C followed by 45 minutes at room temperature in
peroxidase-conjugated anti-rabbit IgG; 3) two washes of 5 minutes each
at 4°C in PBS and one wash in Tris-buffered saline; and 4) incubation
in diaminobenzidine to allow visualization of the electron-dense
peroxidase reaction product. Tissue sections and cells on membranes
were then dehydrated and embedded in Araldite (Taab) resin,
ultrathin sections were cut on an ultramicrotome (Sorvall), and
sections were viewed, with and without uranyl acetate counterstain,
using a JEOL electron microscope.
Immunofluorescence Microscopy
For immunofluorescence microscopy, cells were first fixed in a 2%
formaldehyde solution in PBS. Antibodies were diluted to 5 µg/ml in a
Hanks’ balanced salt solution containing 2% horse serum and 20
mM Tris-Cl, pH 7.5. For basolateral labeling of polarized MDCK
cells, 0.25% saponin was added to permeabilize the surface membrane.
Apical membrane labeling was performed in the absence of saponin. Three
primary antibodies were used: mAb 9E10 to the c-myc
epitope43
that was engineered to be a C-terminal epitope on
the chicken NaK-ATPase ß2 subunit, mAb b24 to the chicken NaK-ATPase
ß1 subunit,44
and the polyclonal antipeptide antibody
(#1305) to the human NaK-ATPase ß2 subunit. All three antibodies
recognized extracellular epitopes on their target ß subunits. Primary
antibody solution was applied to fixed cells for one hour at room
temperature. Cells were rinsed and then incubated in an fluorescein
isothiocyanate-labeled secondary antibody (goat anti-mouse IgG or goat
anti-rabbit IgG; Kirkegaard and Perry, Gaithersburg, MD). Cells were
viewed and photographed with a Zeiss Axioskop epifluorescence
microscope and Kodak Tmax 400 film. An additional set of MDCK cells
stably transfected with ß2, as well as ADPKD epithelial cells, was
subjected to double labeling with anti-1 mAb and anti-ß2 peptide
antibody, and then the labeled MDCK cells were visualized by
goat-anti-rabbit-fluorescein isothiocyanate and anti-mouse Texas red.
Western Immunoblot Analysis and Immunoprecipitation
Membrane extracts were prepared from ADPKD and normal kidneys by the method of Jorgensen.45 After washing in ice-cold PBS, pH 7.4, the kidneys were finely minced in ice-cold homogenization buffer (300 mmol/L sucrose, 25 mmol/L Hepes, made to pH 7.0 with 1 mol/L Tris) containing the protease inhibitors 1 mmol/L 4-(2-aminoethyl)-benzenesulfonyl fluoride (Boehringer Mannheim, Indianapolis, IN), 1 mmol/L benzamidine (Sigma Chemical Co., St Louis, MO), 10 µg/ml leupeptin (Boehringer), 1 µg/ml pepstatin A (Boehringer), 1 µg/ml aprotinin (Boehringer), and 1 µg/ml chymostatin (Boehringer) and then homogenization was carried out in the cold, using a Potter apparatus. The homogenate was centrifuged at 1000 x g for 20 min at 4°C to remove nuclei and cell debris. The supernatant was further centrifuged at 80,000 x g for 30 min at 4°C. The pellet (whole-cell membranes) was suspended in the ice-cold homogenization buffer, and protein concentrations were determined with the bicinchoninic protein assay (Pierce, Rockford, IL), using BSA as standard. In some experiments, the pellet was then resuspended in ice-cold detergent extraction buffer (20 mmol/L Tris-HCl, 120 mmol/L NaCl, 2 mmol/L ethylenediaminetetraacetic acid (EDTA), 2 ethylene glycol bis(ß-aminoethyl ether)-N,N,N',N',-tetraacetic acid, 0.1 mmol/L dithiothreitol, pH 7.4) containing the protease inhibitors described above, incubated for 15 min on ice with one of the following detergents: 1% octylglucoside (Pierce), 0.5% 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS; Pierce), 1% Nonidet P-40 (NP-40) (Boehringer), 1% deoxycholate (Boehringer), 0.5% digitonin (Boehringer), or 0.5% Triton X-100 (Boehringer), and then centrifuged at 100,000 x g for 1 hour at 4°C. Protein concentrations were determined on the supernatant, which contained the solubilized membrane proteins. The extracts were used immediately or flash-frozen in liquid nitrogen and stored at -80°C until further use.
Cell lysates from primary monolayer cultures were also used. After aspiration of the culture media, monolayers were washed for 5 minutes with PBS (pH 7.4) at room temperature, scraped, and centrifuged at 8000 x g for 90 seconds. The cell pellet was then flash-frozen in liquid nitrogen and stored at -80°C until use. Whole-cell extracts were prepared from frozen cell pellets resolubilized in ice-cold lysis buffer (250 mmol/L sucrose, 1 mmol/L EDTA, 20 mmol/L imidazole, pH 7.2) containing the protease inhibitors described above. After 30 minutes incubation on ice, the suspension was briefly sonicated (Branson sonifier 250) and then centrifuged at 14,000 x g for 30 seconds at 4°C. The supernatant (200 µl) was transferred into tubes containing 5 µl of 10% sodium dodecyl sulfate (SDS) and heated at 95°C for 90 seconds. After protein concentration determination, the samples were stored at -80°C until use.
For SDS-polyacrylamide gel electrophoresis (PAGE), all extracts were
solubilized by heating either at 95°C for 2 minutes or at 60°C for
12 minutes in sample buffer (1.5% SDS, 10 mmol/L Tris/HCl, pH 6.8,
0.6% dithiothreitol, 6% (v/v) glycerol). Proteins (20 µg/lane) were
separated by electrophoresis through 0.1 x 9 x 6-cm 12%
acrylamide slabs and transferred to nitrocellulose. The membranes were
briefly stained with Ponceau Red (Sigma) to check the efficiency of
transfer. Destained membranes were blocked for 30 minutes at room
temperature in blotting buffer (50 mmol/L NaPO4, 150 mmol/L
NaCl, 0.05% Tween 20, pH 7.4) containing 5% nonfat dry milk, followed
by incubation with the primary antisera and affinity purified
antibodies (anti-1, anti-ß1, or anti-ß2 diluted to 1:1000,
1:5000 or 1:10,000) in the blotting buffer containing 2% BSA at 4°C
for 18 hours. In addition, a polyclonal antibody against a full-length
human ß2 fusion protein (26; generously supplied by P.
Martin-Vassallo) was used. The membranes were then washed in several
changes of blotting buffer, incubated for 1 hour with
peroxidase-labeled goat anti-rabbit IgG (Kirkegaard & Perry), washed
again, and visualized after 1 minute of incubation with enhanced
chemiluminescence (Amersham, Arlington Heights, IL) at room
temperature.
For immunoprecipitation, protein A/G PLUS Agarose (30 µl) was added
to 100 µg cell protein lysate in the presence of NaK-ATPase mAb 5
or 2F and was incubated for 1 hour at room temperature before
centrifugation at 8000 rpm for 2 minutes. After three washes with
radioimmunoprecipitation assay buffer containing protease
inhibitors (as above), the pellet was resuspended in 2x
electrophoresis sample buffer, boiled, and centrifuged at 8000 rpm for
three minutes, and the supernatant was subjected to SDS-PAGE.
The specificity of the antibodies was demonstrated by comparison with preimmune sera, by competition after preadsorption with the appropriate peptide, and by affinity purification. Additional controls were performed by omitting primary antibodies.
Reverse Transcription-PCR Analysis
Total RNA was isolated from cells or tissues46
and
stored in 70% ethanol, 0.3 M sodium acetate at -80°C until needed.
An aliquot of RNA (0.2–1.0 µg) was solubilized in 1 mmol/L Tris, pH
7.5, EDTA on ice, and the solution was adjusted to 1x in reverse
transcriptase buffer (50 mmol/L Tris-Cl, pH8, 3.75 mmol/L KCl, 3 mmol/L
MgCl2; Bethesda Research Laboratories, Bethesda, MD), 10
mmol/L dithiothreitol, 0.5 mmol/L each of dATP, dCTP, dGTP, and dTTP,
and 10,000 to 50,000 U/ml Maloney murine leukemia virus (M-MLV)
SuperScript reverse transcriptase (Bethesda Research Labs), 100 µg/ml
random hexamer, 0.67 U/100 µl Inhibit-Ace (5 Prime-3 Prime, Inc.).
The RNA and reaction components without enzyme were heated to 65°C
for 3 minutes and placed at 37°C for 5 minutes. Enzyme was added and
incubated at 37°C for 2 hours and purified by ultrafiltration through
Centricon 100 filters three times with 2-ml washes of 10 mmol/L Tris,
pH 7.5, to a final volume of 50 µl. PCR was performed by using 1 to
10 µl of purified sample with conditions appropriate for each primer
pair (primer concentration 100–1000 nmol/L). The following
specific 18-nucleotide primers were designed for NaK-ATPase 1, 3,
ß1, and ß2 isoforms and were positioned towards the 5' end of the
coding region, because the N-terminal regions of the proteins are the
most divergent28
:
1) 1 sense,5'-GAT TAC AAC GGC TGA
TAG-3';
2) 1 antisense, 5'-TTG GAC GTG ATA AGT
ATG-3';
3) 3 sense,5'-GAA AAG AAG GTG ATG
TTC-3';
4) 3 antisense, 5'-GAC AAG AAA GAT GAC
AAG-3';
5) ß1 sense,5'-TTC ATC TGG AAC TCA GAG-3';
6) ß1 antisense, 5'-TCA AGC TTG AAT CTG CAG-3';
7) ß2 sense,5'-AGA CTG AGA ACC TTG ATG-3';
8) ß2 antisense, 5'-ATG CGA CAT TCT ACA TTC-3'.
These primer pairs delineate PCR products of 411 (1), 701
(3), 419 (ß1), and 527 (ß2), bp respectively.
Thermostable DNA polymerase was Ampli-TAQ (Cetus P-E) 19–25 U/ml in
20, mmol/L Tris (pH 8.3), and 35 mmol/L KCL, 10 µg/ml gelatin, 1.25
to 2.5 mmol/L MgCl2, and 200 µmol/L each of dATP, dGTP,
dCTP, and dTTP. Annealings were carried out at 35°C to 55°C for 1
minute, extension was at 72°C for 1 minute, and denaturation was at
94°C for 45 seconds, carried out for 30–35 cycles. Products were
analyzed by 1.4 to 4.0% agarose gel with ethidium bromide staining and
were subcloned into PBSIIKS-(Stratagene, La Jolla, CA). Human
1, 3, ß1, and ß2 probes were sequenced to confirm insert
orientation and identity before use in subsequent RNase protection and
in situ hybridization experiments.
RNase Protection Assay
CsCl gradient-purified plasmid DNA containing T3 and T7 phage
promoter sites flanking the cDNA insert was digested with a restriction
endonuclease selected to yield an antisense riboprobe of 100 to 400 bp.
After enzyme digestion and DNA purification, 0.5 µg of the plasmid
template were transcribed in vitro with either T3 or T7 RNA
polymerase (2.5 U/µl) in 40 mmol/L Tris-Cl (pH 8.0), 25 mmol/L NaCl,
8 mmol/L MgCl2, 2 mmol/L spermidine-(HCl3), 10
mmol/L dithiothreitol, 400 µmol/L ATP, 400 µmol/L CTP, 400 µmol/L
GTP, 12.5 µmol/L [32P]UTP (400 Ci/mmol), 2
U/µl placental RNase inhibitor (Boehringer) in a reaction volume of
20 µl for 30 minutes at 37°C followed by a 15-minutes at 37°C
incubation with 0.5 U/µl RNase-free DNase (Boehringer). The
riboprobes were purified with phenol/CHCl3/isoamyl-OH
extraction and precipitated three times in 75% EtOH, 0.5 mol/L
NH4OAC. For hybridizations, 10 to 30 µg of total RNA (or
control transfer RNA) were resuspended in 40 mmol/L Pipes (pH 6.4), 400
mmol/L NaCl, 1 mmol/L EDTA, 80% formamide in the presence of 2 x
105
cpm of antisense probe and 2 x 105
cpm of internal control 18S riboprobe (Ambion) in 30 µl for 16 hours
in a 45°C bath. Then 350 µl of 10 mmol/L Tris-Cl (pH 7.5), 300
mmol/L NaCl, 5 mmol/L EDTA containing 40 µg/ml RNase A (Boehringer),
and 0.2 µg/ml RNase T1 (Boehringer) were added to the hybridization
mixture and RNA digestion was performed at 30°C min followed by
proteinase-K digestion in 0.5% SDS,
phenol/CHCl3/isoamyl-OH extraction, and ethanol
precipitation. The assay products were fractionated using a 6%
urea-polyacrylamide gel electrophoresis system, and dried gels were
examined by autoradiography with an intensifying screen for 12 to
72 h.
In Situ Hybridization
Tissue sections mounted on silanized Probe-On Plus slides (Fisher, Springfield, NJ) were deparaffinized, dehydrated, treated with proteinase K (1 µg/ml, 30 minutes at 37°C), 0.1 mol/L triethanolamine, 0.25% acetic anhydride, dehydrated, prehybridized for 3 hours at 42°C, and then hybridized with digoxigenin-substituted probe overnight. Stringency washes were carried out in decreasing concentrations of standard saline citrate (optimized for each probe) and incubated in anti-digoxigenin coupled to alkaline phosphatase for 4 hours at room temperature. Color development was carried out with NBT/Xphos (Boehringer), and sections were viewed under a Nikon FXA-Mikrophot with Nomarski optics. Endogenous alkaline phosphatase was inhibited by levamisole (1 mmol/L).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Immunocytochemistry with polyclonal antibodies against the
NaK-ATPase subunit detects clear basolateral distribution of
reaction product in tissue sections of normal human renal epithelia
(Figure 1A)
). As has been described in
other species, there is distinct heterogeneity of intensity of staining
that corresponds with differential activities of NaK-ATPase in
different segments of the nephron. The most intense staining was seen
in thick ascending limbs of Henle’s loop, whereas proximal tubules and
collecting ducts showed lesser intensities of staining and thin limbs
of Henle and glomeruli were negative. In cystic tubules from ADPKD
kidneys, however, virtually all immunodetectable staining for
NaK-ATPase subunit is seen on the apical plasma membranes lining
expanded, fluid-filled lumens (Figure 1, B and C)
. This is the case,
even in the early stages of cystic expansion, although at this stage
some intracellular staining and staining of secreted products in the
lumen are also occasionally detected (Figure 1B)
. Identical patterns of
distribution were seen in normal and ADPKD kidneys with all polyclonal
and monoclonal anti-1 antibodies described in the Materials
and Methods section above.
|
Transcriptional Regulation of NaK-ATPase Isoforms
Since NaK-ATPase is a heterodimer of and ß forms, a
PCR-based strategy was devised to identify NaK-ATPase isoforms in
normal and ADPKD kidney tissues.11,47
Specific primers were
designed that would amplify 1, 3, ß1, and ß2 isoforms of
NaK-ATPase. NaK-ATPases 1 and ß1 were amplified from both normal
and ADPKD kidney RNA by reverse transcription; ß2 was amplified from
ADPKD RNA, whereas 3 was not detected from either source. All PCR
fragments were then isolated, sequenced to confirm their identity, and
cloned into pBluescript; 32P-labeled riboprobes were made,
and RNase protection analysis was carried out (Figure 2)
. Tissue RNAs
(10 µg per lane) from age-matched normal and ADPKD kidneys were
examined for the expression of 1-(411 bp), ß1-(419 bp), and
ß2-(527 bp) protected fragments and compared with that of
constitutively expressed 18S ribosomal RNA (76 bp) (Figure 2A)
. As predicted from previous Northern
analysis,63 the 1 isoform of NaK-ATPase was highly
expressed in all normal and ADPKD kidneys (Figure 2B
, lanes 3–8), but
the highest expression was seen in early-stage ADPKD kidneys, which is
the most active stage of the disease for epithelial proliferation,
fluid secretion, and cystic expansion (Figure 2B
, lanes 5–6). The
pattern of ß1 expression was identical to that of 1 (Figure 2B
,
lanes 11–16). A different pattern of ß2 mRNA expression was seen,
however, in that it was detected only in ADPKD kidneys, not in
age-matched normal kidney tissue (Figure 2C
, cf. lanes 4–6 with lanes
2 and 3).
|
. All reaction products, when present,
were restricted to epithelial cells. In normal kidneys, 1 and ß1
isoforms were localized to the medullary thick ascending limb epithelia
(Figure 3, A and C)
, whereas ß2 was not seen in normal kidney tubule
epithelia (Figure 3E)
. By contrast, in ADPKD kidneys, not only were the
1 and ß1 isoforms localized to cyst-lining epithelia (Figure 3, B and D)
, but intense staining for ß2 was also seen and was also
restricted to cystic epithelia (Figure 3F)
.
|
To relate these findings to protein expression, specific peptide
antibodies were designed against the human NaK-ATPase 1, 3, ß1,
and ß2 isozymes, which corresponded precisely to the regions of the
isoforms used for specific riboprobe design. A schematic representation
of the positioning of these peptides used for antibody generation is
given in Figure 4
. In each case, the
sequence of choice was close or relatively close to the N terminal,
because these were the regions of the N terminal of maximal sequence
divergence between isoforms and therefore those most suitable for the
generation of specific and discriminatory probes.
|
1 peptide antiserum
recognized a single band at 97 kd in total membrane preparations of
normal adult kidneys (Figure 5)
and
showed immunostaining patterns identical to those seen with polyclonal
antiserum raised against dog and mAbs raised against chicken (Figure 5B)
. The human ß1 and ß2 peptide antisera recognized more than one
band, but peptide competitions, comparisons with preimmune sera, and
affinity purification determined that ß1 antibody was specific for a
protein of 43 kd (Figure 6)
and that the
ß2 antibody was specific for a protein of 50 kd in normal adult and
ADPKD epithelia (Figure 7)
. The
additional presence of higher-molecular-weight immunoreactive bands
suggests that glycosylated forms of ß subunits exist in kidney as in
many other tissues. It is interesting that there seemed to be a
decrease in the higher glycosylated forms in ADPKD samples and that, in
human fetal kidneys, where the NaK-ß2 subunit is highly expressed,
higher-molecular-weight immunoreactive bands suggest that this protein
is differentially glycosylated during development (Figure 7B)
. During
these studies it was noted that the mode of membrane extraction was
important; whereas SDS or Nonidet P-40 extractions were sufficient for
blotting 1 and ß2 in membranes, CHAPS or octylglucoside membrane
extraction was necessary for blotting of ß1. Also, the narrow band
width noted was consistent with a minigel system and has been shown
previously, possibly reflecting the recognition of certain glycosylated
forms (48; A McDonough, personal communication). Immunoprecipitation
results suggest a core protein of 45 kd, which was present in ADPKD but
not normal adult kidney epithelia (Figure 7D)
.
|
|
|
1 protein,
although there was occasional detection of a lower-molecular-weight
band in endstage ADPKD kidneys (Figure 5)
. By contrast, expression of
ß1 (43 kd), which was consistent in normal human kidney membrane
preparations extracted with octylglucoside or CHAPS, was absent from
plasma membrane preparations of endstage ADPKD kidneys and reduced in
membrane preparations from early-stage ADPKD kidneys (Figure 8)
. The reverse situation was seen for
ß2, whose expression (50 kd) was consistently detected in membrane
fractions from both endstage and early ADPKD kidneys, but was absent
from normal adult kidneys (Figure 9)
.
Similar patterns of expression were seen in cell cultures derived from
normal and ADPKD epithelia in that normal renal tubule epithelia
expressed membrane 1 and ß1 isozymes in their plasma membranes,
whereas ADPKD epithelia expressed plasma membranes 1 and ß2.
|
|
1 peptide antibodies showed
basolateral membrane localization in normal renal thick-ascending-limb
epithelia but apical membrane localization in ADPKD cyst epithelia
(Figure 10, A and B)
, which is an
identical finding to our previous studies with polyclonal antibodies
derived against the whole protein (8; Figure 1
). NaK-ATPase ß1 and
ß2 immunostaining patterns, however, showed distinctive and different
distribution patterns. As predicted, the NaK-ATPase ß1 antibody
staining colocalized precisely with 1 subunit staining on
basolateral plasma membranes of normal human adult renal epithelia
(Figure 10C)
. In ADPKD cyst-lining epithelia, however, although
staining was intense and widely distributed throughout different cysts,
the reaction product was diffusely distributed throughout the cell
cytoplasm and apparently not associated with plasma membranes (Figure 10D)
. This result was in agreement with the RNase protection and
Western immunoblot data, which had suggested that the ß1 subunit was
highly expressed in ADPKD, but not in plasma membrane extracts.
NaK-ATPase ß2 immunostaining also correlated with RNase protection
and Western immunoblot analysis. Little if any ß2 immunoreaction
above background was detectable in normal human renal tubule epithelia
(Figure 10E)
. By contrast, some clearly membrane-associated reaction
product was seen in cyst-lining epithelia in ADPKD kidneys, and this
membrane reactivity was at the apical surface (Figure 10F)
.
|
and showed that the majority of the reaction product was confined to
linear profiles of the endoplasmic reticulum (Figure 11B)
. This was
contrasted with the localization of 1 and ß2 to the apical
membrane and submembranous vesicles (Figure 11, C and D)
.
|
As discussed above, the formation of cysts in polycystic kidney
disease correlates with misexpression of the ß2 subunit of the
NaK-ATPase and its appearance in the apical surface of cyst cells. To
determine whether the NaK-ATPase ß2 subunit shows any propensity for
apical targeting, we expressed both chicken and human NaK-ATPase ß2
subunits in the kidney-derived, polarizing, epithelial MDCK cells.
Stable MDCK cell lines were generated, expressing 1) chicken ß1
subunits, 2) chicken ß1 subunits lacking their N-terminal cytosolic
domain,49
3) c-myc epitope-tagged chicken ß2
subunits, and 4) human ß2 subunits. Cells of each line were plated at
high density on coverslips, treated with butyrate to induce higher
expression of the exogenous ß subunit, and examined by
immunofluorescence microscopy (Figure 12)
. Both the normal and the
N-terminally truncated chicken ß1 subunits accumulated selectively in
the basolateral surface of the polarized cells (Figure 12A)
. However,
both the epitope-tagged chicken ß2 subunit and the human ß2 subunit
accumulated selectively in the apical membrane (Figure 12, B and D)
,
and no accumulation at the basolateral surface could be found in
permeabilized cells (Figure 12C)
.
|
1 in Human ADPKD
Epithelial Cell Lines
We then turned our attention to human kidney-derived epithelial
cell lines. Many human renal epithelial cell lines derived from normal
collecting tubules, and thick ascending limb epithelia show high-level
expression and basolateral polarization of NaK-ATPase 1 (Figure 13, A
; lanes 1 and 2, and B) and ß1
subunits, but express little if any ß2 subunit (Figure 13C
, lanes 1
and 2). By contrast, primary and immortalized cell lines derived from
human ADPKD epithelia express 1 and ß2 subunit proteins (Figure 13, A and C
, lanes 3 and 4), both of which are apically located
(Figures 11C and 13D)
. In addition to colocalization in membrane
fractions by Western blot and by immunocytochemistry, physical
association between 1 and ß2 subunits was demonstrated by
immunoprecipitation analysis (Figure 13E)
. Also, double
immunofluorescence analysis of a1 and b2 subunits in showed
complete overlapping of images in a punctate and fusiform distribution
consistent with apical membrane localization in ADPKD epithelia (Figure 13F)
.
|
The results obtained suggest that the additional expression and
synthesis of NaK-ATPase ß2 subunit are associated with the complex
formation with 1, which then exists the endoplasmic reticulum, is
processed through the Golgi, and is sorted, inappropriately, to the
apical membrane (Figure 14)
. It is
interesting that ß1 subunit protein, although synthesized, does not
appear to exit the endoplasmic reticulum or assemble with 1 and,
presumably, gets degraded. This model implies that the 1ß2
heterodimers are preferentially produced in comparison with 1ß1
heterodimers in cystic epithelial cells. The cellular and molecular
basis for this preference could involve 1) selective ß subunit
folding, 2) selective ß2 subunit assembly, 3) selective ß
subunit retention in the endoplasmic reticulum, and/or 4) differential
interaction with the protein-targeting machinery.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
6ß1
integrin, retain their normal polarized (apical or basal distribution),
whereas others are mistargeted to apical or basolateral
membranes.6,12,15
Of equal importance, we have implicated
inappropriate fetal gene expression as an important mechanism in ADPKD
cyst generation relevant to the failure to acquire proper polarization
of membrane transport proteins.11,12 Normal nephrogenesis results in a differentiated epithelial cell phenotype, an essential feature of which is the appropriate polarization of membrane proteins such as the NaK-ATPase and EGF receptor. It is of considerable interest that, at stages when polycystin-1 expression in the fetal kidney is high, NaK-ATPase and EGFR are also seen at the apical plasma membranes of renal epithelial cells, whereas, when polycystin-1 expression is low in adult kidneys, basolateral localization is seen.5,12,15 Most importantly, the ß2 subunit of NaK-ATPase is also expressed in fetal kidneys.6,12,47
Our initial studies of human ADPKD epithelia identified a mistargeting and reversal of polarized membrane distribution of NaK-ATPase8 and subsequently of EGFRs5 to the apical epithelial cell membranes of ADPKD epithelia in vivo and in vitro. These modifications were shown to have potentially detrimental physiological consequences for normal renal function and growth, including reversal of Na+ ion and fluid vectorial transport from an absorptive to a secretory mode and the establishment of an autocrine loop for EGF-driven proliferation.7,8,52,57
NaK-ATPase plays a major and essential role in the establishment of the
electrochemical and osmotic ion gradients that direct vectorial
transport of ions and fluid, resulting in the reabsorption of fluid and
essential nutrients. In normal adult kidneys, the vectorial transport
of Na+ ions from the renal-tubule lumen (apical cell
surface) to the blood space (basal cell surface) is associated with the
segregation of NaK-ATPase complexes to the basolateral membrane
domains. These complexes are 1ß1 heterodimers in normal adult
kidneys, the subunit being responsible for the catalytic functions
of the complex, whereas the ß subunit is essential for active complex
assembly and trafficking to the plasma membrane. The studies reported
here, using specific polyclonal peptide antibodies made against human
subunit sequences, confirm the presence of 1 and ß1 subunits of
NaK-ATPase in the basolateral membrane domains of renal-tubule
epithelia. In normal human kidneys, there was no detectable ß2
subunit protein in these basolateral membrane NaK-ATPase complexes.
This was in sharp contrast to ADPKD kidneys, in which cystic epithelia
showed expression of not only 1 and ß1 mRNA but also ß2 subunit
mRNA. It is of interest that this combination of subunit expression has
been reported previously, but in the secretory, not absorptive
epithelium of the choroid plexus, where the NaK-ATPase is apically
located.48
Immunocytochemical and Western analysis showed a similar complement of
1, ß1, and ß2 protein products, but ß1 could not be detected
in membrane fractions and was localized to the cytoplasm by light and
electron immunocytochemistry, predominantly to the endoplasmic
reticulum. By contrast, both 1 and ß2 were detected in the
membrane fractions of ADPKD kidneys and, by immunocytochemistry, was
localized to the apical plasma membranes of ADPKD epithelia in
vivo and in vitro. It is interesting that the
differences in patterns of glycosylation of ß subunits between normal
adult and ADPKD kidneys might be related to alterations with impact on
processing and possibly trafficking, but further studies are needed to
elucidate such a relationship.
In vitro, the targeting of both human and chicken ß2
subunits to the apical surface of MDCK cells was an unequivocally
robust observation. The ß1 and ß2 subunits share only about 40%
amino acid sequence identity, even within a species. However, there is
likely a much higher conservation of tertiary structure, and it is
known that each of these ß subunits is capable of assembly with each
of the three known NaK ATPase -subunits.58-61
Determining what aspect of molecular structure is responsible for the
divergent sites of accumulation of NaK-ATPase ß1 and ß2 subunits in
polarized epithelial cells will require rigorous analysis.
Although the findings reported here do not prove a causal relationship
between ß2 subunit expression and apical targeting of NaK-ATPase
molecules in polycystic kidney cysts, the results do suggest a possible
link between the genetic defect in polycystic kidney disease and the
reversal of cell polarity that involves altered genetic expression.
Work is in progress to examine whether the mutant ADPKD gene product
may cause a loss of repression of fetal gene transcription, including
the ß2 subunit of NaK-ATPase, and to determine the molecular basis of
the abnormal apical targeting of 1ß2 complexes in ADPKD cystic
epithelia.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by National Institutes of Health grants R01 DK44833 (to P. W.) and R01 NS23241 (to D. F.). S. R. is a Robert Wood Johnson Foundation University Scholar.
Accepted for publication September 15, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
