mLin-7 is localized to the basolateral surface of renal
epithelia via its NH2 terminus
Samuel W.
Straight1,
David
Karnak2,
Jean-Paul
Borg3,
Emmanuel
Kamberov3,
Heidi
Dare3,
Ben
Margolis1,2,3, and
James B.
Wade4
3 Howard Hughes Medical Institute, Departments
of 1 Internal Medicine and
2 Biological Chemistry, University of Michigan
Medical School, Ann Arbor, Michigan 48109; and
4 Department of Physiology, University of
Maryland School of Medicine, Baltimore, Maryland 21201-1559
 |
ABSTRACT |
In
Caenorhabditis elegans, the basolateral localization of the
Let-23 growth factor receptor tyrosine kinase requires the expression
of three genes: lin-2, lin-7, and lin-10.
Mammalian homologs of these three genes have been identified, and a
complex of their protein products exists in mammalian neurons. In this paper, we examine the interaction of these mammalian proteins in renal
epithelia. Coprecipitation experiments demonstrated that mLin-2/CASK
binds to mLin-7, and immunofluorescent labeling showed that these
proteins colocalized at the basolateral surface of Madin-Darby canine
kidney cells and renal epithelia. Although labeling intensity varied
markedly among different renal epithelial cells, those cells strongly
expressing mLin-7 also showed intense mLin-2/CASK labeling. We have
also demonstrated that mLin-2/CASK binding requires amino acids
12-32 of mLin-7 and have shown that this region of mLin-7 is also
necessary for the targeting of mLin-7 to the basolateral surface.
Furthermore, the overexpression of mLin-2/CASK mutants in Madin-Darby
canine kidney cells caused endogenous mLin-7 to mislocalize. In
summary, the NH2 terminus of mLin-7 is crucial for its
basolateral localization, likely through its interaction with
mLin-2/CASK.
protein targeting; Madin-Darby canine kidney cells; kidney; protein
interactions; epithelia
 |
INTRODUCTION |
MANY CELLS HAVE POLARIZED plasma membrane compartments.
Epithelia are polarized into apical and basolateral plasma membrane domains separated by tight junctions (9). Similarly, neurons are
polarized to form the dendrite and axonal components (11). Several
mechanisms appear to be involved in the proper distribution of proteins
to basolateral or apical surfaces. These mechanisms include the sorting
of proteins from the trans-Golgi network and endocytic
compartments to the apical or basolateral surface and selective
retention of proteins at basolateral or apical membranes (8, 24).
The machinery in cells that recognizes these basolateral-targeting
signals and effects targeting of proteins is poorly
understood. However, genetic systems are now beginning to
reveal the mechanisms that appear important in this targeting process
at the plasma membrane (12, 21). Kim and colleagues (21) recently
identified a mechanism for basolateral targeting in Caenorhabditis
elegans. This mechanism involves three proteins, Lin-2, Lin-7, and
Lin-10, that are required for the basolateral targeting of receptors in worm epithelia and neurons (18, 29). Mutations in lin-2,
lin-7, or lin-10 lead to a failure of vulva formation,
presumably due to mislocalization of Let-23 (16, 18, 31). Lin-2 is a
member of the family of proteins known as membrane-associated guanylate kinases (MAGUKs). These proteins contain a region similar to guanylate kinase, an enzyme that converts GMP to GTP. In MAGUKs, however, this
domain appears to be catalytically inactive and functions in protein
interactions (19, 23, 32). Members of this family also contain PDZ
domains, named for three MAGUK proteins,
SD-95,
iscs
Large, and
ona Occludens-1. PDZ domains
bind the extreme COOH terminus of certain proteins, and are best
understood for their role in the ability of MAGUK-family proteins, such
as PSD-95, to bind and cluster ion channels and receptors in synapses
(13, 22, 30). The Zona Occludens proteins are MAGUK proteins localized at tight junctions, but the role of the PDZ domains in these proteins is still under investigation (14). The Lin-10 protein contains two PDZ
domains and one phosphotyrosine binding (PTB) domain (18). PTB domains
were originally identified for their role in phosphotyrosine-dependent interactions in proteins such as Shc and IRS-1. However, recent studies
show that they also have an important role in binding beta-turn
peptides independent of phosphotyrosine (2). Lin-7 is a smaller protein
with a single PDZ domain that binds to the COOH-terminal tail of the
Let-23 protein, a receptor tyrosine kinase related to the mammalian
epidermal growth factor receptor family (31).
Recent work has indicated that mammalian homologs of the Lin-2, Lin-7,
and Lin-10 proteins are present in mammalian neurons in a stable
protein complex (4, 7). Mammalian Lin-10 homologs have been previously
identified in mammalian cells as the X11 family of proteins (3). They
have been also been described as Munc-18 interacting (Mint) proteins
(26). There are at least three forms of X11 in mammals that
have divergent NH2 termini. The mammalian Lin-2 protein
(mLin-2) has been identified in mammalian cells as a protein known
as CASK (15), and the NH2 terminus of X11
(Mint1) binds to the calmodulin kinase-like domain in mLin-2/CASK (4,
7). The NH2 terminus of mammalian Lin-7 (mLin-7) binds to
mLin-2/CASK (4, 7, 18). The X11, mLin-2/CASK and mLin-7
proteins have not been extensively studied in mammalian epithelial
cells, although mLin-2/CASK has been found at the basolateral surface
of epithelia (10). Therefore, we sought to examine the interactions of
mLin-7, mLin-2/CASK, and X11 in MDCK and renal epithelial cells. We
have found that mLin-2/CASK and mLin-7 interact in renal epithelial
cells, but they do not interact with X11
, the form of X11 we have
identified in epithelial cells. Both mLin-2/CASK and mLin-7 are located
at the basolateral surface of epithelial cells. In renal epithelia,
there is close correlation in the expression of mLin-7 and mLin-2/CASK
in different tubule segments. Finally, we have further refined the
region of mLin-7 required for interaction with mLin-2/CASK and
determined that the region of mLin-7 that mediates its interaction with
mLin-2/CASK is also required for its basolateral targeting.
 |
MATERIALS AND METHODS |
DNA constructs.
The cloning of full-length X11
and X11
cDNAs, as
well as the human lin-2 and mouse mlin-7 cDNAs, has
been described previously (4, 5). The lin-2 and mlin-7
cDNAs were used as templates for PCR with appropriate oligonucleotide
primers to create truncated protein constructs. The cDNAs or PCR
products were subcloned into the plasmid pGSTag by using appropriate
restriction endonucleases, where GST denotes
glutathione-S-transferase. This plasmid and the expression and
purification of GST-fusion proteins have been described previously (3).
The mlin-7 cDNA was subcloned into the plasmid pET28a+
(Novagen; Madison, WI) for the creation of a His6-tagged
protein. The vectors pRK5 and pRK5-Myc (3) were used to express
proteins in mammalian cells with or without an NH2-terminal
Myc epitope, respectively. Canine betaine-gamma amino butyric acid
transporter-1 (BGT-1) cDNA (34), obtained from Joseph Handler and H. Moo Kwon, was subcloned into pRK5-Myc. All constructs were sequenced by
using Sequenase version 2.0 DNA sequencing kit (Amersham Life Science,
Cleveland, OH) or by automated sequencing at the University of Michigan
DNA Sequencing Core.
Cell culture and transfection.
Human embryonic kidney 293T (HEK293T) and Madin-Darby canine kidney
(MDCK) cells were grown in DMEM (Life Technologies, Grand Island, NY)
containing 100 U/ml penicillin and 100 µg/ml streptomycin sulfate,
supplemented with 10% FCS.
HEK293T were transiently transfected by using a calcium phosphate
precipitation method. MDCK cells were stably transfected by using
FuGENE 6 transfection reagent (Boehringer Mannheim, Mannheim, Germany)
followed by selection with Geneticin/G-418 (600 µg/ml active; Life
Technologies) or hygromycin B (500 µg/ml; Invitrogen, Carlsbad, CA)
depending on the selectable marker plasmid initially cotransfected.
Studies were conducted with both pools of selected cells and with
established clonal cell lines.
Antibodies.
Polyclonal anti-mLin-7, anti-mLin-2/CASK, and anti-X11 antibodies used
for immunoprecipitation, immunoblotting, and immunostaining were
prepared by injecting rabbits with purified GST-fusion proteins as
follows: GST-mLin-7, GST-mLin-2/CASK (1-275), GST-mLin-2/CASK (578-897), GST-X11
(620-837), and GST-X11
(15-246).
The anti-mLin-7 antibody was subsequently affinity purified and was
used in all applications. For purification, His6mLin-7 was
expressed in DE3 Escherichia coli, which were lysed, and the His-tagged protein was bound to nickel-agarose beads (Qiagen, Santa
Clara, CA) and purified. The bound protein was eluted with 0.5 M
imidazole, dialyzed, and then coupled to Affigel-10 (Bio-Rad Laboratories, Hercules, CA). Anti-mLin-7 antiserum was precipitated with 50% ammonium sulfate, and the precipitated proteins were resuspended and dialyzed against PBS. The antibody solution was passed
several times over an Affigel-His6mLin-7 column, and the bound antibody was eluted under acid (pH 2.5) conditions. The His6mLin-7 protein was also used in immunofluorescence
applications to competitively inhibit immunostaining with anti-mLin-7 antibody.
Anti-Myc monoclonal antibody (clone 9E10) was used for immunostaining
and immunoblotting. The anti-mLin-2/CASK monoclonal antibodies used for
immunolocalization in the kidney were from Transduction Laboratories
(Lexington, KY). Affinity-purified rabbit polyclonal anti-ZO-1
antibodies and rat anti-uvomorulin/E-cadherin monoclonal antibodies
used for immunostaining were purchased from Zymed (San Francisco, CA)
and Sigma Chemical (St. Louis, MO), respectively. Unless otherwise
noted, fluorochrome-conjugated secondary antibodies used in
immunostaining procedures were purchased from Jackson ImmunoResearch
Laboratories (West Grove, PA).
Antibodies to COOH-terminal peptides of aquaporin-2 (AQP2; LC54) and -3 (AQP3; LP45) were raised in chickens to peptides previously utilized
for production of these antibodies in rabbits (33). Rabbit antibody to
NKCC2 (L320) was raised to amino acids 33-55 of the rat
cotransporter (20) and was kindly provided by Dr. Mark Knepper.
Immunostaining of MDCK cells.
For immunostaining procedures, MDCK cells were seeded at high density
onto acid-washed coverglasses or Transwell PTFE membrane filters (0.4 µm pore size; Corning Costar, Cambridge, MA). The cells were allowed
to grow to confluence to form a polarized monolayer. After being washed
with PBS, the cells were fixed with 4% formaldehyde/PBS and
permeabilized with 0.1% Triton X-100/PBS. After being blocked for 1 h
with goat serum, the cells were incubated with primary antibodies
diluted in 2% goat serum/PBS in a humidified chamber for 1 h
(affinity-purified anti-mLin-7 at 1:50; anti-Myc at 1:400; anti-ZO-1 at
1:400; and anti-E-cadherin at 1:1,600). After being washed three times
with 2% goat serum/PBS, the cells were incubated with secondary
antibodies coupled to FITC, Cy3, or Cy5 (diluted at 1:500 in 2% goat
serum/PBS) for 1 h in a humidified chamber. Coverglasses were mounted
on glass slides with ProLong antifade reagent (Molecular Probes,
Eugene, OR). Membrane filters were cut from their plastic casing with a
scalpel and mounted as above. Examination of immunostained cells was
performed on an Olympus BX60 fluorescent microscope, and digital images
were taken with a SPOT charge-coupled device camera (Diagnostic
Instruments). Confocal laser-scanning microscopy was performed on a
Nikon Diaphot 200 microscope paired with a Noran laser and InterVision
software (Noran Instruments, Middleton, WI) at the Morphology and Image Analysis Core of the University of Michigan Diabetes Research Center.
Protein procedures.
Lysates for precipitation experiments were prepared from subconfluent
HEK293T and MDCK cells in 10- or 15-cm tissue culture dishes,
respectively (3). Cells were washed twice with cold PBS and lysed in
0.5-1 ml of lysis buffer [50 mM HEPES (pH 7.5), 10%
glycerol, 150 mM NaCl, 1% Triton X-100, 1.5 mM MgCl2, 1 mM EGTA] supplemented with 1 mM phenylmethylsulfonylfluoride (PMSF), 10 µg/ml aprotinin, 10 µg/ml leupeptin, 100 mM NaF, 10 mM
Na4P2O7, and 200 µM
Na3VO4. The lysates were cleared by
centrifugation at 16,000 g for 20 min at 4°C to remove
insoluble debris. For precipitation assays, 0.2 ml lysate from
transiently transfected HEK293T or 1 ml lysate from MDCK cells was used.
Fractionated lysates were collected from harvested mouse kidneys
snap-frozen in liquid nitrogen. The frozen organ was ground by mortar
and pestle in hypotonic buffer [10 mM Tris · Cl
(pH 7.4), 0.2 mM MgCl2, 5 mM KCl] supplemented with 1 mM PMSF, 10 µg/ml aprotinin, and 10 µg/ml leupeptin (7.5 ml
buffer/0.5 g organ). The organ was then homogenized by 20 strokes in an
ice-cold dounce (B piston). Sucrose was added to the homogenate to a
final concentration of 0.25 mM, and EDTA to 1 mM, and nonhomogenized
debris was removed by centrifugation at 1,000 g for 10 minutes
at 4°C. The supernatant was then centrifuged at 138,000 g
(max) for 1 h at 4°C. The supernatant was collected and was
termed the cytoplasmic fraction. The pellet was resuspended in a volume
of lysis buffer (described above) equivalent to the volume of the
cytoplasmic fraction. After 30 min on ice, insoluble debris was removed
by centrifugation at 16,000 g for 30 min at 4°C. The
resulting supernatant was termed the membrane fraction. Typically, 1 ml
of fractionated lysate was used in precipitation assays.
For immunoprecipitation, lysates were incubated with antibodies
overnight at 4oC. Protein A-agarose was added and immune
complexes bound to beads were recovered after 1 h, washed three times
with HNTG buffer (50 mM HEPES, pH 7.5, 10% glycerol, 150 mM NaCl,
0.1% Triton X-100), boiled in 1× sample buffer, and separated by
SDS-PAGE. Transfer and immunoblotting on nitrocellulose using
HRP-protein A or HRP-anti-mouse antibody were performed as described
(3) by using the Renaissance chemiluminescence reagent (NEN Life
Science Products, Boston, MA).
GST-fusion protein production and GST-binding assays were performed as
previously described (3). Typically, 5 µg of GST-fusion protein were
used per precipitation reaction.
Immunolocalization in the rat kidney.
Renal tissue was obtained from 180- to 250-g male Sprague-Dawley rats
fixed by perfusion via the abdominal aorta. Perfusion was for 2 min in
PBS to clear the kidneys of blood, 5 min in 2% paraformaldehyde, and 2 min in a cryoprotectant of 10% EDTA in 0.1 M Tris. Fixed kidneys were
sliced and further incubated in the cryoprotectant for 60 min, wrapped
in aluminum foil, and frozen on dry ice. Cryostat sections 12-15
µm thick were made and picked up on coverslips coated with HistoGrip
(Zymed). Sections were usually further fixed and attached to the
coverslips by incubation for 5 min with 4% paraformaldehyde and then
treated with either 1% SDS or 6 M guanidine for 10 min to unmask
antigenic sites (6). Sections were then washed three times with
high-salt buffer (50 ml PBS, 0.5 g BSA, 1.13 g NaCl), incubated in
blocking agent (50 ml PBS, 0.5 g BSA, 0.188 g glycine, pH 7.2) for 20 min, followed by incubation with primary antibody overnight at 4°C.
Primary antibodies were diluted to 10 µg/ml with incubation medium
(50 ml PBS, 0.05 g BSA, 200 µl 5% NaN3). After this
incubation, sections were rinsed five times with high-salt buffer
before incubation with secondary antibody for 2 h at 4°C.
Appropriate species-specific antibodies coupled to Alexa 488 or 568 dyes (Molecular Probes) were diluted 1:200 with incubation medium.
These samples were again washed five times with high-salt buffer over
the course of 1 h and then in PBS to remove the excess salt before
mounting and confocal microscopy.
 |
RESULTS |
The structures and binding sites for X11
, mLin-2/CASK, and mLin-7
are shown in Fig. 1A. The form of
mLin-7 utilized in these studies corresponds to the protein Veli-3
described by Butz et al. (7). The calmodulin kinase II-like (CKII)
domain of mLin-2/CASK binds to a region within the NH2
terminus of X11
, whereas the NH2 terminus of mLin-7
binds the region between the CKII and PDZ domain of mLin-2/CASK (1, 4,
7, 18). We have previously generated antibodies that recognize
mammalian mLin-2/CASK and mLin-7 (4). In MDCK cells, we confirmed that
mLin-2/CASK was bound to mLin-7 (Fig. 1B). Both mLin-2/CASK and
mLin-7 were detected in the cytosol and membrane fractions of mouse
kidney lysates in approximately equal quantities (results not shown),
and in both compartments mLin-2/CASK and mLin-7 were bound to each
other. We have previously detected X11
as an epithelial form of X11 in mammalian cells. Unlike mLin-7 and mLin-2/CASK, most of X11
is
present in the cytosolic fraction of kidney lysate (Fig.
2A). When we immunoprecipitated
with antibodies to mLin-7, we detected no X11
in the
mLin-2/CASK-mLin-7 complex of kidney. Similarly, antibodies to X11
did not coimmunoprecipitate mLin-2/CASK or mLin-7. In MDCK cells, the
expression of X11
is relatively low, and we could not detect
mLin-2/CASK or mLin-7 in a complex with X11
(data not shown). To
better test these interactions in tissue culture cells, we
overexpressed X11
or X11
in MDCK cells. There is no endogenous
X11
in MDCK cells, but it was easily detected when overexpressed
(Fig. 2B). As expected, mLin-2/CASK and mLin-7 were found to
coimmunoprecipitate with X11
in these cells when both anti-X11
and anti-mLin-7 were used as immunoprecipitating antibodies. In
contrast, X11
did not coimmunoprecipitate with mLin-2/CASK or mLin-7
even after X11
was overexpressed in MDCK cells (Fig. 2C).
The divergence in the amino acid sequences of their NH2
termini, particularly in the region of X11
that binds to
mLin-2/CASK, might explain the differential inclusion of X11
and
X11
in a complex with mLin-2/CASK and mLin-7 (5). One final
observation was that X11
appears as three immunoreactive bands in
Fig. 2, A and C. These multiple bands are not likely to
be the product of alternate splicing, because the same bands were
observed from both endogenous (Fig. 2A) and exogenous (Fig. 2C) expression. Nor are they likely to be the result of
proteolytic degradation during the course of the experiments, because
proteases were present in the lysis buffer. More likely, the three
immunoreactive bands are the result of a posttranslational
modification, such as phosphorylation, but this remains to be
determined by future investigation.

View larger version (30K):
[in this window]
[in a new window]
|
Fig. 1.
Lin-2, Lin-7 and Lin-10 are evolutionarily conserved proteins.
A: schematics of mammalian homolog of Lin-2, Lin-7, and Lin-10
proteins of Caenorhabditis elegans. Left: names of
proteins; right: relative molecular weights (kDa). Sites of
interaction (BS) between these proteins are demarcated by labeled bars
above or below each illustration. Recognized protein domains are
indicated as follows: PTB, phosphotyrosine binding; PDZ,
PSD-95/Dlg-1/ZO-1; CKII, calmodulin kinase II; SH3, Src-homology 3; GK,
guanylate kinase. Hook indicates region recognized as binding
cytoskeletal protein 4.1 (10). B: coimmunoprecipitation of
mLin-2/CASK and mLin-7. mLin-2/CASK and mLin-7 proteins were
coprecipitated from Madin-Darby canine kidney (MDCK) cells and from
fractionated mouse kidney lysates with affinity-purified anti-mLin-7
antibodies and protein A-Sepharose beads. Preimmune rabbit serum was
used as a control. Immunoprecipitated proteins were separated by
SDS-PAGE, transferred to nitrocellulose, and identified by immunoblot
with antibodies indicated. Left: relevant bands (arrows);
right: relative molecular weight. memb, Membrane; cyto,
cytoplasm.
|
|

View larger version (30K):
[in this window]
[in a new window]
|
Fig. 2.
X11 does not bind mLin-2/CASK or mLin-7. A:
interaction of X11 , mLin-2/CASK, and mLin-7 in kidney extracts.
Mouse kidney cytoplasm and membrane lysate fractions were incubated
with antibodies to either mLin-7 or X11 or with preimmune rabbit
serum as a control. Bound proteins were precipitated with protein
A-Sepharose beads, separated by SDS-PAGE, transferred to
nitrocellulose, and immunoblotted with antibodies indicated. Note that
in immunoblot for mLin-2, a nonspecific band of ~100 kDa appears in
anti-X11 immunoprecipitation lane. B and C:
interaction of X11 protein, mLin-2/CASK, and mLin-7 in transfected MDCK
cells. Triton-soluble lysates were collected from untransfected MDCK
cells (WT) or MDCK cells stably transfected with a plasmid expressing
X11 (+X11 ) as indicated in B (top). Proteins were
immunoprecipitated from these cells with antibodies to mLin-7, X11 ,
or with preimmune rabbit serum as a control, and treated as described
above. In C, lysates were collected from MDCK cells stably
transfected with a plasmid expressing X11 . Proteins in these lysates
were immunoprecipitated with antibodies to mLin-7, X11 , or with
preimmune rabbit serum as a control. Precipitated proteins were treated
as described above. In all panels, relative molecular weight is
indicated to the right (kDa), and arrows to left of
immunoblots indicate relevant bands. In B and C, lysate
indicates a lane loaded with 1/20 the volume of total cell lysate used
for precipitation.
|
|
Previous work had indicated that the NH2-terminal region of
mLin-7 was responsible for binding to mLin-2/CASK (4, 7, 18). We
further refined the region of mLin-7 that binds to mLin-2/CASK by using
GST-fusion proteins to several deletion mutants of mLin-7 (shown in
Fig. 3A) and were able to show that
amino acids 1-43 were sufficient to mediate this interaction (Fig.
3B). Further minimization of the NH2 terminus to
amino acids 1-33 of mLin-7 resulted in greatly reduced binding to
mLin-2/CASK (data not shown). Removal of the first 12 amino acids of
mLin-7 had no effect on the interaction of mLin-2/CASK with mLin-7
while removing the first 32 amino acids, or more, completely eliminated
the interaction (Fig. 3C). Altogether, these results indicate
that amino acids 12-32 of mLin-7 are required for binding to
mLin-2/CASK, although other elements within the NH2
terminus may be necessary for efficient binding.

View larger version (14K):
[in this window]
[in a new window]
|
Fig. 3.
Deletion mutagenesis of mLin-7. A: schematic shows full-length
mLin-7 protein (top) and mLin-7 constructs created for this
study (bottom). Nos. to left correspond to amino acids
of mLin-7 contained within each construct, whereas nos. on
illustrations indicate amino acid sequence position. PDZ,
PSD-95/Dlg-1/ZO-1 protein interaction domain. B and C:
binding of deletion mutants to mLin-2.
Glutathione-S-transferase (GST)-fusion proteins were made with
each mLin-7 construct shown in A, bound to glutathione-agarose
beads, and used to precipitate Myc-mLin-2/CASK from HEK293T lysates.
Precipitated proteins were separated by SDS-PAGE, transferred to
nitrocellulose, and blotted with anti-Myc antibodies as indicated.
Arrows to left of immunoblots indicate bound Myc-mLin-2/CASK.
GST alone is a control construct not containing any mLin-7 sequence.
Lane labeled lysate indicates 1/10 amount of HEK293T total lysate used
in precipitation experiments.
|
|
Next we used our antibodies to examine the localization of mLin-7 in
MDCK cells (Fig. 4). Shown are the
immunolocalization experiments involving MDCK cells affixed to
coverglasses. Comparable results were obtained with cells on PTFE
membrane filters (data not shown). Using rabbit polyclonal antibodies
to mLin-7, we detected endogenous mLin-7 at the lateral surface of MDCK
cells (Fig. 4A). Using anti-Myc antibodies we determined that
expressed Myc-mLin-7 was localized to the same site as the endogenous
protein (Fig. 4B). Neither our polyclonal antibodies nor
commercial antibodies to mLin-2/CASK were suitable for
immunofluorescence studies of MDCK cells, so we were unable to
determine the location of endogenous mLin-2/CASK. However, we also
found Myc-tagged mLin-2/CASK was localized to the lateral surface of
MDCK cells (see below, Fig. 9C). In contrast, overexpressed
X11
localizes to a perinuclear region and not at the cell surface
(Fig. 4C). This perinuclear localization of X11
corresponds
with the localization of X11
in neurons (1). Furthermore, a similar
perinuclear localization has been observed for X11
expressed in MDCK
cells (data not shown). These results suggest that some conserved
region(s) of the X11 proteins might be responsible for their
localization in cells, and that the machinery involved in this
localization is conserved in both neurons and epithelial cells.

View larger version (86K):
[in this window]
[in a new window]
|
Fig. 4.
Localization of mLin-7 and X11 in MDCK cells. MDCK cells were seeded
onto glass coverslips, fixed, permeabilized, and stained with indicated
primary antibodies, followed by staining with secondary antibodies
conjugated to fluorochromes. A: endogenous mLin-7 protein in
untransfected MDCK cells stained with affinity purified anti-mLin-7
antibodies. B: MDCK cells stably expressing Myc-tagged mLin-7
and stained with anti-Myc antibodies. C: localization of X11
in MDCK cells stably transfected with a plasmid expressing full-length
X11 . Although a confluent monolayer is shown, and was confirmed by
staining with anti-ZO-1 antibodies, not all the cells express X11 .
Cells are shown at higher magnification to show detail of perinuclear
staining pattern. D and E: control immunostaining of
MDCK cells with anti-ZO-1 antibodies (D) and anti-E-cadherin
antibodies (E), which demarcate tight junction and adherens
junction respectively. In A-E, top panel shows digital
image of immunostained cells acquired through a charge-coupled device
camera attached to a fluorescent microscope, and bottom panel
shows Z-section of a Z-series taken with a confocal laser-scanning
microscope. Bar, 10 µm.
|
|
In agreement with our results in MDCK cells, we were also able to
localize endogenous mLin-7 and mLin-2/CASK proteins to the basolateral
surface of multiple renal segments in rat kidney. Figure
5 shows the smooth localization of mLin-7
at the basolateral surface of inner medullary collecting ducts and,
with less intensity, in thin limbs of the loop of Henle (Fig.
5A). The papillary epithelial cells that cover the surface of
the renal papilla were also intensely labeled by antibody to mLin-7.
The corresponding labeling with antibody to AQP3 is shown in Fig.
5B. This water channel protein is present in the basolateral
membrane of collecting ducts and papillary epithelium, but not in thin
limbs. Thus mLin-7 has a distinctly basolateral localization in native
renal epithelia as well as cultured MDCK cells. Tubular segments in the
inner medulla that stained brightly for mLin-7 (Fig. 5C) also
stained intensely for mLin-2 (Fig. 5D), demonstrating a close
correlation in the expression of the two proteins in native epithelial
cells. Similar staining for mLin-2/CASK and mLin-7 at the basolateral surface was also seen in the outer medulla of the kidney (Fig. 6, A and B, respectively).
The specificity of the mLin-7 antibody was examined by incubating it
with an excess of His6-mLin-7 protein. This blocked
labeling by the mLin-7 antibody (Fig. 6C) but not by the
mLin-2/CASK antibody (Fig. 6D) and demonstrates that the colocalization of mLin-7 and mLin-2/CASK was not due to cross-reaction of the mLin-2/CASK antibody with shared epitopes present on mLin-7.

View larger version (111K):
[in this window]
[in a new window]
|
Fig. 5.
Immunolocalization of mLin-7 and mLin-2/CASK in renal inner medulla.
A: antibody to mLin-7 shows this protein at basolateral surface
of inner medullary collecting ducts (CD). Labeling of lateral aspects
is indicated by arrowheads. Thin limbs of loop of Henle (TL) and
papillary epithelium (PE) are also labeled. B:
labeling with antibody to basolateral water channel aquaporin-3
(AQP3) in collecting ducts and papillary epithelium corresponds closely
to mLin-7 labeling. C and D:
immunolocalization of mLin-7 (C) compared with labeling by
using a mouse monoclonal antibody to mLin-2/CASK (D)
demonstrated that these proteins shared a very similar distribution in
renal segments. Bar, 25 µm.
|
|

View larger version (141K):
[in this window]
[in a new window]
|
Fig. 6.
Immunolocalization of mLin-7 and mLin-2/CASK in the renal outer
medulla. A and B: antibodies to mLin-7 (A) and
to mLin-2/CASK (B) label basolateral domains of renal tubules
in a similar pattern. Sections labeled with rabbit antibody to mLin-2
and monoclonal mouse anti-mLin-2/CASK demonstrated that both antibodies
gave a similar pattern of mLin-2/CASK immunostaining (data not shown).
C: incubation of mLin-7 antibody with an excess of
His6mLin-7 protein blocked labeling by antibody. D:
labeling by mLin-2 antibody was not diminished by preincubation with
excess His6mLin-7 protein. Bar, 25 µm.
|
|
Immunolocalization of mLin-7 and mLin-2/CASK with respect to
segment-specific transporters showed that they are not uniformly expressed along renal nephron segments. Figure
7 illustrates the labeling pattern observed
by using antibodies to the Na-K-Cl cotransporter 2 (NKCC2) to identify
the apical domains of the epithelial cells in the thick ascending limbs
of the loop of Henle (Fig. 7A). Although the basolateral
domains of cells comprising the thick ascending limb and collecting
duct segments were strongly labeled by mLin-2/CASK and mLin-7 (not
shown), proximal tubules were only weakly labeled (Fig. 7B). Close
examination of collecting duct labeling showed that the labeling
pattern for mLin-7 and mLin-2/CASK varied significantly in intensity,
even within a given segment. Labeling with antibodies to the principal
cell-specific water channel AQP2 (Fig. 7C) showed that this
cell type strongly expressed mLin-7 at the basolateral domain, whereas
there was little or no labeling by mLin-7 antibody detected in adjacent
intercalated cells (arrows, Fig. 7D).

View larger version (98K):
[in this window]
[in a new window]
|
Fig. 7.
Immunolocalization of mLin-2/CASK and mLin-7 in renal outer medulla and
cortex. A and B: comparison of the labeling with
antibody to bumetanide-sensitive cotransporter NKCC2 (A) with
respect to mLin-2/CASK labeling (B) shows that many of the
brightly labeled tubules in the renal outer medulla and cortex are
thick ascending limbs of loop of Henle (TAL). Other structures that are
brightly labeled by mLin-2/CASK are CD, whereas proximal tubules (PT)
are much more weakly labeled on their basal domains. C and
D: comparison of labeling with an antibody to water channel
aquaporin-2 (AQP2; C) with respect to mLin-7 (D) shows
that mLin-7 expression varies between epithelial cell types even within
a segment. Intercalated cells, which are weakly labeled for AQP2 at
their apical domains (C, arrows), also show little or no
labeling by mLin-7 antibody at their basolateral domains (D,
arrows), in contrast to bright staining with both labels seen in
adjacent principal cells. Bar, 25 µm.
|
|
One potential binding target for mLin-7 in the renal medulla is BGT-1.
BGT-1 is expressed in the cells of the inner medulla in response to
osmotic stress, localizes to the basolateral surface of cells, and has
a COOH-terminal sequence, -EKTHL, remarkably similar to that of worm
Let-23, -EKTCL (27, 34). Recent data also suggests that the COOH
terminus of BGT-1 was responsible for the retention of the transporter
at the basolateral surface of cells (28). Figure
8 shows the coprecipitation of BGT-1 with the PDZ domain of mLin-7: BGT-1 was precipitated with GST-fusion proteins to both full-length mLin-7 and the mLin-7 PDZ domain, GST-mLin-7 (79-197), but not the NH2-terminal half of
mLin-7, GST-mLin-7 (1-92). These results taken together support
the suggestion that BGT-1 might be targeted to the basolateral surface
of medullary collecting duct cells by the mLin-7-mLin-2/CASK complex.

View larger version (25K):
[in this window]
[in a new window]
|
Fig. 8.
Betaine -amino butyric acid transporter 1 (BTG-1) binds to the
PDZ domain of mLin-7. GST-mLin-7-fusion proteins were bound to
glutathione-agarose beads, then used to precipitate Myc-BGT-1 from
HEK293T lysates. Precipitated proteins were separated by SDS-PAGE,
transferred to nitrocellulose, and blotted with anti-Myc antibodies as
indicated. Arrow to left of immunoblot indicates Myc-BGT-1, and
asterisks indicate higher order (glycosylated) forms of Myc-BGT-1
protein. GST alone is a control construct not containing any mLin-7
sequence. Amino acids of mLin-7 included in fusion proteins are
indicated, and these constructs are illustrated in Fig. 3. Lane labeled
lysate indicates 1/10 amount of HEK293T total lysate used in
precipitation experiments. Relative molecular weight is shown to
right (kDa).
|
|
In an attempt to dissect the molecular basis for the localization of
mLin-7 to the basolateral surface, we expressed the Myc-tagged PDZ
(amino acids 79-197) or NH2 terminus (amino acids
1-92) of mLin-7 in MDCK cells. Through the examination of these
cells by immunofluorescence, we found that the amino terminus localized to the lateral membrane (Fig. 9A),
whereas the PDZ domain of mLin-7 localized diffusely within cells (Fig.
9B). This suggested that the NH2 terminus was
responsible for the localization of mLin-7 to the basolateral surface
of the cells. Next, we wanted to determine if the mLin-2 binding domain
within the NH2 terminus of mLin-7 was essential for mLin-7
localization. Therefore, MDCK cells were transfected with the deletion
mutants of mLin-7 shown in Fig. 3. We found that the deletion of the
first 12 amino acids of mLin-7 did not alter localization to the
basolateral surface of cells (Fig. 9C), whereas deletion of the
first 32 amino acids abolished basolateral localization (Fig. 9D).
These results indicated that localization of mLin-7 to the basolateral
surface of cells depended on the region of mLin-7 also required for
binding to mLin-2/CASK. As mentioned previously, we were unable to
determine the localization of endogenous mLin-2/CASK in these MDCK cell
lines, because our antibodies, as well as those available commercially
available, proved to be inadequate for detecting the canine protein by
immunofluorescence.

View larger version (144K):
[in this window]
[in a new window]
|
Fig. 9.
Immunolocalization of mLin-7 constructs expressed in MDCK cells. MDCK
cells were stably transfected with Myc-tagged mLin-7 constructs. These
cells were prepared for immunostaining as described in Fig. 4. Primary
antibody was anti-Myc. As described in Fig. 4, top panel in
A-E is a digital photomicrograph, and bottom panel
shows a Z-section. A-D show MDCK cells expressing different
truncated mLin-7 constructs described in Fig. 3. All panels show
confluent monolayers of cells; however, in some cases not all cells
express transfected plasmid (B and D). Amino acids
contained within each construct are given in parentheses above each
panel. Bar, 10 µm.
|
|
To assess the role of mLin-2/CASK in mLin-7 localization, we examined
the effects of the overexpression of mLin-2/CASK mutants on the
localization of endogenous mLin-7. We expressed two
Myc-tagged fragments of mLin-2/CASK in MDCK cells: the
NH2-terminal half (amino acids 1-612), and the
COOH-terminal half (amino acids 578-897). Previous studies have
shown that the NH2-terminal half of mLin-2/CASK binds
mLin-7, whereas the COOH-terminal fragment does not (4, 7, 18). This
was confirmed in the results of experiments involving the
coprecipitation of the Myc-tagged mLin-2/CASK constructs with
endogenous mLin-7 protein in our MDCK cell lines (Fig.
10A). Examination
of these cells by immunofluorescence showed that the Myc-tagged
wild-type mLin-2/CASK localizes to the basolateral surface of cells,
overlapping in part with the localization of mLin-7 (Fig. 10C,
top and bottom panels), as well as with the junctional protein ZO-1 (Fig. 10C, middle panel). However, neither
the NH2-terminus nor the COOH terminus of mLin-2/CASK
localized to the basolateral surface: the mLin-2/CASK(1-612)
localized in a diffusely cytoplasmic pattern (Fig. 10D),
whereas mLin-2/CASK(578-897) entered the nucleus (Fig.
10E). The aberrant localization of these halves of mLin-2/CASK indicates the correct localization of mLin-2/CASK to the basolateral surface is likely to be a more complex process than mLin-7
localization, perhaps involving more than one region of the mLin-2/CASK
protein. Most interestingly, however, the expression of the
NH2-terminal half of mLin-2/CASK was sufficient to
mislocalize a significant fraction of mLin-7 away from the lateral
membrane (Fig. 10D, top and bottom panels),
with no apparent effect on the localization of the junctional protein
ZO-1 (Fig. 10D, middle panel). In contrast, overexpression of the COOH-terminal half of mLin-2/CASK had no significant effect on the localization of mLin-7 in cells (Fig. 10E, top and bottom panels), consistent with
its inability to bind mLin-7. Thus it appears that in MDCK cells the
localization of mLin-7 was dependent on the correct localization of
mLin-2/CASK.

View larger version (67K):
[in this window]
[in a new window]
|
Fig. 10.
Expression and immunolocalization of mLin-2/CASK constructs in MDCK
cells. MDCK cells were stably transfected with Myc-tagged mLin-2/CASK
constructs. A: coimmunoprecipitation and immunoblotting of
transfected mLin-2/CASK constructs with endogenous mLin-7 from MDCK
cells by using anti-Myc antibodies, as described in Fig. 1. Antibodies
used for blotting are indicated below each panel. Left:
relevant bands (arrows); open arrow, immunoglobin heavy chain;
right: relative molecular weight (kDa). B: MDCK cells
transfected with empty RK5-Myc vector as a control. C: MDCK
cells transfected with full-length mLin-2. D and E:
MDCK cells expressing truncated mLin-2/CASK proteins. B-E:
cells prepared for immunostaining as described in Fig. 4, with
anti-Myc, anti-mLin-7 and anti-ZO-1 as primary antibodies. As described
in Fig. 4, top panel in A-E is a digital
photomicrograph, showing cells colabeled with anti-Myc antibodies
(green) and affinity purified anti-mLin-7 antibodies (red);
middle panel shows Z-section of a Z-series taken with a
confocal laser-scanning microscope of cells stained with anti-Myc
(green) and anti-ZO-1 (red) antibodies. ZO-1 is a tight
juction-associated protein used as a control to indicate lateral aspect
of membrane; and bottom panel also shows a Z-section, but here
cells were immunostained with anti-Myc antibodies (green) as well as
affinity purified anti-mLin-7 antibodies (red) to show distribution of
endogenous mLin-7. In all panels, yellow indicates an overlap in
antibody labeling. Amino acids contained within each truncated protein
are indicated above each panel in parentheses. Bar, 10 µm.
|
|
 |
DISCUSSION |
In worms, the Lin-7 PDZ domain interacts with the Let-23 growth factor
receptor and is believed to be essential for the basolateral localization of the receptor (18, 31). In mammalian epithelial cells,
our results have demonstrated that mLin-7 localizes to the basolateral
surface of cells. The finding that mLin-7, like mLin-2/CASK, is at the
basolateral surface suggests that it may be involved in binding and
retaining cell surface proteins at the basolateral membrane, rather
than playing a role in Golgi sorting. Many PDZ domain-containing
proteins have a role in the clustering of receptors in neurons (22,
30). However, a more detailed examination of mLin-7 and mLin-2/CASK
localization by using electron microscopy would be required to
determine whether these proteins were clustered at distinct sites on
the basolateral plasma membrane of epithelial cells, or more generally
localized as the results of our light and confocal microscopy studies
indicate. The mLin-7 protein is strongly expressed in certain tubular
segments of the kidney, such as the inner medullary collecting duct,
where it colocalizes with mLin-2. The PDZ domain of mLin-7 does not appear to bind any members of the mammalian epidermal growth factor receptor family, but it can bind to the COOH terminus of the Let-23 protein (S. W. Straight, J.-P. Borg, D. Karnak, and B. Margolis, unpublished observations). There are proteins within the inner medulla
of the kidney that have COOH-terminal sequences similar to worm Let-23,
such as BGT-1 (27, 34). The coprecipitation of BGT-1 with the mLin-7
PDZ domain (Fig. 8) suggests that BGT-1 might be targeted to the
basolateral surface of kidney epithelial cells by a mLin-7-mediated
mechanism. Furthermore, the identification of the COOH-terminal
sequence of BGT-1 as a binding partner for the mLin-7 PDZ domain,
combined with the knowledge of the COOH-terminal sequence of
Let-23, suggests that a further search for proteins with similar
COOH-termini may identify other binding partners for the PDZ domain of
mLin-7. It should be noted, however, that mLin-7 and mLin-2/CASK are
only weakly expressed in some epithelial cell types of the kidney,
suggesting this complex may not be operative in all epithelia and is
just one of the systems involved in the basolateral localization of proteins.
We have determined that amino acids 1-43 of mLin-7 are sufficient
to mediate binding to mLin-2 and have also demonstrated that deletion
of amino acids 12-32 of mLin-7 eliminates this interaction with
mLin-2. Using the Chou-Fasman method for secondary structure prediction
and the determination of Eisenberg hydropathic moment (Lasergene
Protean program, DNASTAR), we have identified within this region an
amphipathic helix encompassing amino acids 9-27. Our data suggest
that this putative amphipathic helix is a crucial component of the
mLin-7-mLin-2/CASK interaction. However, a peptide containing amino
acids 1-33 of mLin-7 did not significantly block binding of
mLin-2/CASK to full-length mLin-7 (S. W. Straight, E. Kamberov, and B. Margolis, unpublished observations), indicating that this region may
only comprise part of the necessary elements for efficient binding of
mLin-7 to mLin-2/CASK. We also identified the NH2-terminus
of mLin-7 as being essential for the localization of mLin-7 in MDCK
cells to the basolateral surface. The deletion of amino acids
12-32 prevents mLin-7 from localizing properly to the basolateral
surface, as well as negating mLin-2/CASK binding. Furthermore, the
overexpression of the NH2-terminal half of mLin-2/CASK pulls a fraction of endogenous mLin-7 from the lateral membrane. Thus
our data suggest that the interaction of mLin-7 with mLin-2/CASK is
crucial for mLin-7 localization to the basolateral surface of MDCK
cells. Further support for this notion comes from the close correlation
between mLin-2/CASK and mLin-7 expression in various kidney segments.
However, confocal microscopy (Fig. 10) showed only partial overlap in
the localization of mLin-2/CASK and mLin-7, indicating that other
proteins may be involved in the localization of mLin-7. Butz et al. (7)
have shown that DLG2 and DLG3, MAGUK proteins found in neuronal cells,
have a region homologous to mLin-2/CASK that binds the NH2
terminus of the Lin-7 homolog Veli-1. It is thus possible that mLin-7
has multiple binding partners that direct its localization in diverse cell types.
Our findings also indicate that mLin-7 and mLin-2/CASK are peripheral
membrane proteins that interact both in the cytosol and at the
membrane. The processes involved in the targeting of such proteins to
basolateral vs. apical surfaces are just beginning to be understood (8,
17, 25). If mLin-7 is localized by its interaction with mLin-2/CASK,
what factors lead to the localization of mLin-2/CASK at the basolateral
surface? In worm epithelia, lin-10 and lin-2 are
essential for Let-23 localization and vulval formation, suggesting that
Lin-10 interacts with Lin-2 and might be involved in Lin-2
localization. In mammalian cells, it is possible that X11 family
members might also play a role in mLin-2/CASK localization to the
basolateral surface. However, this hypothesis is challenged by several
findings. First, the localization of X11 proteins does not always
correlate with the localization of mLin-2/CASK. In neurons, we have
identified X11
in a perinuclear region, most likely a part of the
Golgi network (1). Similarly, we find that X11
is located in a
perinuclear region in mammalian epithelia. However, although
mLin-2/CASK colocalizes with X11
in neurons (1), mLin-2/CASK
localized to the lateral membrane in epithelia. Second, although
mLin-2/CASK interacts with X11
in the CNS, we could find no
interaction between mLin-2/CASK with X11
in renal epithelia. It is
possible that there is another form of X11 in epithelia that we have
not yet detected that could control mLin-2/CASK targeting in cells.
Another possibility is that that there are membrane components at the
basolateral surface, distinct from X11, that bring mLin-2/CASK to the
basolateral surface in mammalian epithelia. We also find that neither
the NH2-terminal nor the COOH-terminal half of mLin-2/CASK
is sufficient to correctly localize mLin-2/CASK. Thus the localization
process for mLin-2/CASK may be complex and involve multiple
interactions. The identification of the membrane components that
localize mLin-2/CASK will yield important insights into the targeting
of peripheral membrane proteins.
 |
ACKNOWLEDGEMENTS |
We are grateful to Paul Welling for very helpful discussions, Mark
Knepper for providing antibodies for use in the colocalization studies,
Joseph Handler and H. Moo Kwon for BGT-1 cDNA, Juanita Merchant for the
use of her fluorescent microscope, and Thomas Komorowski, manager of
the University of Michigan Diabetes Research Center Morphology and
Image Analysis Core, for assistance with confocal microscopy. We thank
Jie Liu for extremely valuable technical assistance in carrying out
immunolocalizations in the kidney.
 |
FOOTNOTES |
This work was supported by National Institutes of Health Grants
DK-32839 (J. B. Wade), 5-T32-HD-07505 (S. W. Straight), and GM-08353
(D. Karnak). B. Margolis is an investigator of the Howard Hughes
Medical Institute.
Present address: of J.-P. Borg: U119 INSERM, 27 Boulevard Lei Roure,
13009 Marseille, France.
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Address for reprint requests and other correspondence: B. Margolis,
Howard Hughes Medical Institute, Univ. of Michigan Medical Center, 4570 MSRB II, Box 0650, 1150 W. Medical Center Drive, Ann Arbor, MI
48109-0650 (E-mail: bmargoli{at}umich.edu).
Received 26 May 1999; accepted in final form 21 October 1999.
 |
REFERENCES |
1.
Borg, J.-P.,
M. O. Lopez-Figueroa,
M. Taddeo-Borg,
D. E. Kroon,
R. S. Turner,
S. J. Watson,
and
B. Margolis.
Molecular analysis of the X11-mLin-2/CASK complex in brain.
J. Neurosci.
19:
1307-1316,
1999[Abstract/Free Full Text].
2.
Borg, J.-P.,
and
B. Margolis.
Function of PTB domains.
Curr. Top. Microbiol. Immunol.
228:
23-38,
1998[ISI][Medline].
3.
Borg, J.-P.,
J. Ooi,
E. Levy,
and
B. Margolis.
The phosphotyrosine interaction domains of X11 and FE65 bind to distinct sites on the YENPTY motif of amyloid precursor protein.
Mol. Cell. Biol.
16:
6229-6241,
1996[Abstract].
4.
Borg, J.-P.,
S. W. Straight,
S. M. Kaech,
M. de Taddeo-Borg,
D. E. Kroon,
D. Karnak,
R. S. Turner,
S. K. Kim,
and
B. Margolis.
Identification of an evolutionarily conserved heterotrimeric protein complex involved in protein targeting.
J. Biol. Chem.
273:
31633-31636,
1998[Abstract/Free Full Text].
5.
Borg, J.-P.,
Y. Yang,
M. de Taddeo-Borg,
B. Margolis,
and
R. S. Turner.
The X11alpha protein slows cellular amyloid precursor protein processing and reduces Abeta40 and Abeta42 secretion.
J. Biol. Chem.
273:
14761-14766,
1998[Abstract/Free Full Text].
6.
Brown, D.,
J. Lydon,
M. McLaughlin,
A. Stuart-Tilly,
R. Tyszkowski,
and
S. Alper.
Antigen retrieval in cryostat tissue sections and cultured cells by treatment with sodium dodecyl sulfate (SDS).
Histochem. Cell Biol.
105:
261-267,
1996[ISI][Medline].
7.
Butz, S.,
M. Okamoto,
and
T. C. Sudhof.
A tripartite protein complex with the potential to couple synaptic vesicle exocytosis to cell adhesion in brain.
Cell
94:
773-782,
1998[ISI][Medline].
8.
Caplan, M. J.
Membrane polarity in epithelial cells: protein sorting and establishment of polarized domains.
Am. J. Physiol. Renal Physiol.
272:
F425-F429,
1997[Abstract/Free Full Text].
9.
Cereijido, M.,
J. Valdes,
L. Shoshani,
and
R. G. Contreras.
Role of tight junctions in establishing and maintaining cell polarity.
Ann. Rev. Physiol.
60:
161-177,
1998[ISI][Medline].
10.
Cohen, A. R.,
D. F. Wood,
S. M. Marfatia,
Z. Walther,
A. H. Chishti,
and
J. M. Anderson.
Human CASK/LIN-2 binds syndecan-2 and protein 4.1 and localizes to the basolateral membrane of epithelial cells.
J. Cell Biol.
142:
129-138,
1998[Abstract/Free Full Text].
11.
Craig, A. M.,
and
G. Banker.
Neuronal polarity.
Annu. Rev. Neurosci.
17:
267-310,
1994[ISI][Medline].
12.
Drubin, D. G.,
and
W. J. Nelson.
Origins of cell polarity.
Cell
84:
335-344,
1996[ISI][Medline].
13.
Fanning, A. S.,
and
J. M. Anderson.
PDZ domains and the formation of protein networks at the plasma membrane.
Curr. Top. Microbiol. Immunol.
228:
209-233,
1998[ISI][Medline].
14.
Giepmans, B. N.,
and
W. H. Moolenaar.
The gap junction protein connexin43 interacts with the second PDZ domain of the zona occludens-1 protein.
Curr. Biol.
8:
931-934,
1998[ISI][Medline].
15.
Hata, Y.,
S. Butz,
and
T. C. Sudhof.
CASK: a novel dlg/PSD95 homolog with an N-terminal calmodulin-dependent protein kinase domain identified by interaction with neurexins.
J. Neurosci.
16:
2488-2494,
1996[Abstract].
16.
Hoskins, R.,
A. F. Hajnal,
S. A. Harp,
and
S. K. Kim.
The C. elegans vulval induction gene lin-2 encodes a member of the MAGUK family of cell junction proteins.
Development
122:
97-111,
1996[Abstract/Free Full Text].
17.
Hough, C. D.,
D. F. Woods,
S. Park,
and
P. J. Bryant.
Organizing a functional junctional complex requires specific domains of the Drosophila MAGUK Discs large.
Genes Dev.
11:
3242-3253,
1997[Abstract/Free Full Text].
18.
Kaech, S. M.,
C. W. Whitfield,
and
S. K. Kim.
The LIN-2/LIN-7/LIN-10 complex mediates basolateral membrane localization of the C. elegans EGF receptor LET-23 in vulval epithelial cells.
Cell
94:
761-771,
1998[ISI][Medline].
19.
Kim, E.,
S. Naisbitt,
Y.-P. Hsueh,
A. Rao,
A. Rothschild,
A. M. Craig,
and
M. Sheng.
GKAP, a novel synaptic protein that interacts with the guanylate kinase-like domain of the PSD-95/SAP90 family of channel clustering molecules.
J. Cell Biol.
136:
669-678,
1997[Abstract/Free Full Text].
20.
Kim, G. H.,
C. A. Ecelbarger,
C. Mitchell,
R. K. Packer,
J. B. Wade,
and
M. A. Knepper.
Vasopressin increases Na-K-2Cl cotransporter expression in thick ascending limb of Henle's loop.
Am. J. Physiol. Renal Physiol.
276:
F96-F103,
1999[Abstract/Free Full Text].
21.
Kim, S. K.
Polarized signaling: basolateral receptor localization in epithelial cells by PDZ-containing proteins.
Curr. Opin. Cell Biol.
9:
853-859,
1997[ISI][Medline].
22.
Kornau, H. C.,
P. H. Seeburg,
and
M. B. Kennedy.
Interaction of ion channels and receptors with PDZ domain proteins.
Curr. Opin. Neurobiol.
7:
368-373,
1997[ISI][Medline].
23.
Kuhlendahl, S.,
O. Spangenberg,
M. Konrad,
E. Kim,
and
C. C. Garner.
Functional analysis of the guanylate kinase-like domain in the synapse-associated protein SAP97.
Eur. J. Biochem.
252:
305-313,
1998[Abstract].
24.
Le Gall, A. H.,
C. Yeaman,
A. Muesch,
and
E. Rodriguez-Boulan.
Epithelial cell polarity: new perspectives.
Semin. Nephrol.
15:
272-284,
1995[ISI][Medline].
25.
Myat, M. M.,
S. Chang,
E. Rodriguez-Boulan,
and
A. Aderem.
Identification of the basolateral targeting determinant of a peripheral membrane protein, MacMARCKS, in polarized cells.
Curr. Biol.
8:
677-683,
1998[ISI][Medline].
26.
Okamoto, M.,
and
T. C. Sudhof.
Mints, Munc18-interacting proteins in synaptic vesicle exocytosis.
J. Biol. Chem.
272:
31459-31464,
1997[Abstract/Free Full Text].
27.
Perego, C.,
A. Bulbarelli,
R. Longhi,
M. Caimi,
A. Villa,
M. J. Caplan,
and
G. Pietrini.
Sorting of two polytopic proteins, the gamma-aminobutyric acid and betaine transporters, in polarized epithelial cells.
J. Biol. Chem.
272:
6584-6592,
1997[Abstract/Free Full Text].
28.
Perego, C.,
C Vanoni,
A. Villa,
R. Longhi,
S. M. Kaech,
E. Frohli,
A. Hajnal,
S. K. Kim,
and
G. Pietrini.
PDZ-mediated interactions retain the epithelial GABA transporter on the basolateral surface of polarized epithelial cells.
EMBO J.
18:
2384-2393,
1999[Abstract/Free Full Text].
29.
Rongo, C.,
C. W. Whitfield,
A. Rodal,
S. K. Kim,
and
J. M. Kaplan.
LIN-10 is a shared component of the polarized protein localization pathways in neurons and epithelia.
Cell
94:
751-759,
1998[ISI][Medline].
30.
Sheng, M.,
and
M. Wyszynski.
Ion channel targeting in neurons.
Bioessays
19:
847-853,
1997[ISI][Medline].
31.
Simske, J. S.,
S. M. Kaech,
S. A. Harp,
and
S. K. Kim.
LET-23 receptor localization by the cell junction protein LIN-7 during C. elegans vulval induction.
Cell
85:
195-204,
1996[ISI][Medline].
32.
Takeuchi, M.,
Y. Hata,
K. Hirao,
A. Toyoda,
M. Irie,
and
Y. Takai.
SAPAPs: a family of PSD-95/SAP90-associated proteins localized at postsynaptic density.
J. Biol. Chem.
272:
11943-11951,
1997[Abstract/Free Full Text].
33.
Terris, J.,
C. A. Ecelbarger,
S. Nielsen,
and
M. A. Knepper.
Long-term regulation of four renal aquaporins in rats.
Am. J. Physiol. Renal Fluid Electrolyte Physiol.
271:
F414-F422,
1996[Abstract/Free Full Text].
34.
Yamauchi, A.,
S. Uchida,
H. M. Kwon,
A. S. Preston,
R. B. Robey,
A. Garcia-Perez,
M. B. Burg,
and
J. S. Handler.
Cloning of a Na(+)- and Cl(
)-dependent betaine transporter that is regulated by hypertonicity.
J. Biol. Chem.
267:
649-652,
1992[Abstract/Free Full Text].
Am J Physiol Renal Physiol 278(3):F464-F475
0363-6127/00 $5.00
Copyright © 2000 the American Physiological Society