|
|
|
|||||||
|
|||||||||
(Received for publication, July 28, 1995) From the
The transmembrane 4 (TM4) superfamily contains many important
leukocyte differentiation-related surface proteins including CD9, CD37,
CD53, and CD81; tumor-associated antigens including CD63/ME491, CO-029,
and SAS; and a newly identified metastasis suppressor gene R2.
Relatively little is known, however, about the structure and
aggregation state of these four transmembrane-domained proteins. The
asymmetrical unit membrane (AUM), believed to play a major role in
stabilizing the apical surface of mammalian urothelium thus preventing
it from rupturing during bladder distention, contains two TM4 members,
the uroplakins (UPs) Ia and Ib. In association with two other (single
transmembrane-domained) membrane proteins, UPII and UPIII, UPIa and
UPIb form 16-nm particles that naturally form two-dimensional
crystalline arrays, thus providing unique opportunities for studying
membrane structure and function. To better understand how these
proteins interact to form the 16-nm particles, we analyzed their
nearest neighbor relationship by chemical cross-linking. We show here
that UPIa and UPIb, which share 39% of their amino acid sequence, are
cross-linked to UPII and UPIII, respectively. We also show that UPIa
has a propensity to oligomerize, forming complexes that are stable in
SDS, and that UPII can be readily cross-linked to form homodimers. The
formation of UPII homodimers is sensitive, however, to octyl glucoside
that can solubilize the AUMs. These data suggest that there exist two
types of 16-nm AUM particles that contain UPIa/UPII or UPIb/UPIII, and
support a model in which the UPIa and UPII occupy the inner and outer
domains, respectively, of the UPIa/UPII particle. This model can
account for the apparent ``redundancy'' of the uroplakins, as
the structurally related UPIa and UPIb, by interacting with different
partners, may play different roles in AUM formation. The model also
suggests that AUM plaques with different uroplakin compositions may
differ in their assembly, and in their abilities to interact with an
underlying cytoskeleton. Our data indicate that two closely related TM4
proteins, UPIa and UPIb, can be present in the same cell, interacting
with distinct partners. AUM thus provides an excellent model system for
studying the targeting, processing, and assembly of TM4 proteins.
The ``transmembrane four'' (TM4) ( Uroplakins Ia (UPIa; 27-kDa) ( The
asymmetrical unit membrane offers unique opportunities for studying the
detailed structural arrangement and possible function of the two
differentiation-dependent members of the TM4 family, i.e. the
uroplakins Ia and Ib, because AUMs can be purified in milligram
quantities (Wu et al., 1990; 1994). Moreover, uroplakins
interact closely with one another forming highly organized 16-nm
protein particles that naturally form two-dimensional crystalline
arrays thus greatly facilitating a detailed analysis of protein
structure (Hicks and Ketterer, 1969; Vergara et al., 1969;
Chlapowski et al., 1972; Taylor and Robertson, 1984; Walz et al., 1995). We have therefore probed the topographical
relationship among the four major integral membrane protein subunits of
the asymmetrical unit membrane using the chemical cross-linking
approach. Unexpectedly, our results indicate that uroplakins Ia and Ib
are cross-linked to the 15-kDa uroplakin II and the 47-kDa UPIII,
respectively. The fact that the two structurally related uroplakins Ia
and Ib are cross-linked to different partners suggests that the two TM4
proteins play distinct roles in AUM structure. In addition, we present
data showing that, in intact AUMs, uroplakin II can be cross-linked to
form a homodimer, and that UPIa can form oligomers that are stable in
SDS. Taken together, these results suggest a model in which uroplakins
Ia and II occupy the inner and outer domains, respectively, of a 16-nm
protein particle, and raise the possibility that AUMs are composed of
two types of 16-nm particles containing different subsets of uroplakin
molecules.
A prerequisite of this
approach was the availability of antibodies that were monospecific for
individual uroplakin molecules. We therefore raised a panel of rabbit
antisera against synthetic peptides corresponding to sequences of the
four major uroplakins. Immunoblotting established that all these
antisera reacted strongly with their respective uroplakins (Fig. 1). Antisera to uroplakins Ia, II, and III recognized well
defined 27-, 15-, and 47-kDa protein bands, respectively (Fig. 1b). Antisera to uroplakin Ib recognized multiple
bands in the molecular mass range of 25 to 28 kDa; however, this
apparent heterogeneity could be completely accounted for by
glycosylation (Wu et al., 1994; Yu et al., 1994).
Moreover, at least one antiserum for each uroplakin was shown to
recognize, specifically, only the corresponding uroplakin in the crude
urothelial membrane fraction (Fig. 1).
Figure 1:
Specificity of antibodies against
individual bovine uroplakins. a, proteins of crude bovine
urothelial membranes (lane 1), Sarkosyl-washed AUMs (lane
2), and additionally NaOH-washed AUMs (lane 3) were
dissolved in 1% SDS at room temperature, resolved by SDS-PAGE, and
visualized by silver nitrate (AgNO
These monospecific
antisera enabled us to monitor the cross-linking status of the
uroplakins that were present in crude urothelial membranes that had
been treated with various concentrations of EGS (Fig. 2). This
experiment revealed the formation of three well defined, cross-linked
uroplakin species. A new 22-kDa band was recognized only by the
uroplakin II antibody and was therefore presumably a UPII homodimer; a
35-kDa band was recognized by antisera to both UPII and UPIa and was
thus likely a heterodimer of UPIa and UPII; and finally a 72-kDa band
reacted with antisera to both UPIb and UPIII, suggesting a UPIb/UPIII
heterodimer (Fig. 2). Similar results were obtained using a
hydrophilic analogue of EGS, the sulfo-EGS, although the yield of the
UPIa/UPII heterodimer was greatly reduced (Fig. 3).
Figure 2:
Chemical cross-linking of uroplakins that
are present in native urothelial membranes. Crude bovine urothelial
membranes were cross-linked with (lane 1) 0, (2) 0.5, (3) 1, (4) 2, (5) 3, (6) 4, and (7) 5 mM EGS. Their proteins were dissolved in 1%
SDS, resolved by SDS-PAGE, electrophoretically transferred to
nitrocellulose, and then immunoblotted with (a) antibodies
against UPIa (anti-UPIa), (b) anti-UPII, (c)
anti-UPIb, and (d) anti-UPIII. Note in Panels a) and b the formation of a cross-linked 35-kDa band that was
recognized by both anti-UPIa and anti-UPII (labeled Ia/II),
and a 22-kDa band recognized only by anti-UPII (labeled II/II). Also note in Panels c and d the
formation of a 72-kDa band recognized by both anti-UPIb and anti-UPIII
(labeled Ib/III), a 66-68-kDa band recognized by
anti-UPIb, and a 74-kDa band recognized by anti-UPIII. The relative
positions of molecular weight (M.W.) standards, as well as
those of various uroplakin monomers and dimers are shown on the left and right,
respectively.
Figure 3:
Cross-linking of crude urothelial
membranes using hydrophobic versus hydrophilic bifunctional
cross-linking reagents. Crude bovine urothelial membranes were
cross-linked with (a) EGS and (b) its hydrophilic
analog, sulfo-EGS (S-EGS). The proteins of these cross-linked
membranes were separated by SDS-PAGE and subjected to immunoblotting
using anti-UPIa and anti-UPII, as indicated. Note that these two
reagents are equally effective in generating the 22-kDa UPII homodimer (II/II); however, only the hydrophobic EGS yielded the UPIa/II
heterodimer. The positions of various uroplakin monomers and dimers are
marked on the sides.
Figure 4:
Chemical cross-linking of uroplakins that
are present in purified AUMs. Sarkosyl-washed AUMs (lanes 1 and 3) and the AUMs that had been further washed with
NaOH (lanes 2 and 4) were incubated with only a
buffer (lanes 1 and 2) or with 5 mM EGS (lanes 3 and 4). Proteins of these membranes were
immunoblotted with (a) anti-UPIa and with (b)
anti-UPII. Note the generation of the UPII/UPII homodimer, the
UPIa/UPII heterodimer, and a UPIa/UPIa homodimer, thus confirming the
crude membrane results. A 65-kDa species (?) may represent a
UPIa oligomer (see Fig. 5a and
text).
Figure 5:
Identification of the cross-linked
uroplakin complexes by two-dimensional, diagonal gel electrophoreses.
The proteins of EGS cross-linked AUMs were dissolved in 1% SDS and
separated by first dimensional SDS-PAGE. After staining with Coomassie
Blue and destaining, individual gel lanes were excised, incubated in 1 M hydroxylamine to cleave the cross-linked species, and
subjected to a second dimensional (slab) SDS-PAGE. The two-dimensional
gels were then (a) stained with silver nitrate (AgNO
So far we
identified the cross-linked uroplakin species based on their relative
sizes and their immunoreactivities with various antibodies to
uroplakins. To confirm these assignments, we resolved the cross-linked
AUM proteins by SDS-PAGE, cut out the entire gel lane containing the
cross-linked uroplakins, cleaved them by incubating the gel strip in 1 M NH
Figure 6:
Effects of detergents and chain length of
the cross-linking reagents on uroplakin cross-linking. Cross-linking
was carried out on total or crude membranes (a) or highly
purified AUMs (b and c), using either EGS (16
Å; Panels a and b) or DFDNB (3 Å; Panel c). The cross-linked membrane proteins were dissolved in
1% SDS, resolved by SDS-PAGE, and immunoblotted using anti-UPIa or
anti-UPII, as indicated. In Panels a and b, lanes
1 are controls without cross-linking. EGS cross-linking was
carried out in 10 mM Hepes buffer (lanes 2), or in
the same buffer containing 2% octyl glucoside (lanes 3) or 2%
Triton X-100 (lanes 4). Note that the yield of UPIa/UPII
heterodimer is not affected by the detergents (lanes
2-4); however, the formation of UPII homodimer was largely
abolished by octyl glucoside (lanes 3), although unaffected by
Triton X-100 (lanes 4; see Fig. 7for the scanning of
these lanes). Also note, in Panel c, that the short armed
DFDNB failed to produce the UPII homodimer, although it yielded the
UPIa/UPII heterodimer. The relative positions of uroplakin monomer and
dimer are indicated on the sides.
Figure 7:
Octyl glucoside abolishes selectively the
formation of uroplakin II homodimer. Lanes 2-4 of the
immunoblots produced with anti-UPII, as shown in Fig. 6, a and b, were scanned for densitometry using a Universal
Imaging Program. The samples correspond to (a) crude membranes
and (b) purified AUMs that have been EGS cross-linked (1) without detergent(-), (2) with 2% octyl
glucoside (O.G.), or (3) with 2% Triton X-100 (T.X.). The small, white arrow indicates the
direction of SDS-gel electrophoresis, and the large, open
arrows mark the positions of the UPII homodimer. Note that Triton
had relatively little effect on uroplakin cross-linking, while octyl
glucoside greatly reduced the formation of the UPII homodimer in both
crude membrane and purified AUMs.
To assess the relative distance
of the cross-linked We have probed the topographical relationship of the
uroplakins in the asymmetrical unit membrane using bifunctional
cross-linking reagents. The results that we have obtained so far have
several important features. First, we identified the same set of
cross-linked uroplakin dimers, regardless whether we used the purified
AUMs ( Fig. 4and Fig. 5), or the crude urothelial
membranes ( Fig. 2and Fig. 3), as our starting material.
This suggests that the topographical relationships that exist in the
relatively unperturbed, crude urothelial membranes must have been
maintained to a large extent in our purified AUMs. Second, the
cross-linking patterns were highly reproducible over a wide range of
experimental conditions covering different types and concentrations of
the cross-linking reagents (Fig. 2Fig. 3Fig. 4Fig. 5). Moreover, the
cross-linking was highly efficient capable of capturing >30% of the
uroplakin monomers using the reagent concentrations that we have tested (Fig. 2Fig. 3Fig. 4), thus making it less likely
that we are observing the cross-linking of uroplakins entrapped in a
minor AUM conformation. Third, the cross-linking of purified AUMs
resulted in the formation of only a few, major protein complexes that
have been identified as containing purely uroplakins. Thus the
cross-linking of UPIa with EGS yielded almost exclusively the
UPIa
These
results indicate that UPII is involved in two different kinds of
protein:protein interactions. Its binding to UPIa is short-ranged as
they can be cross-linked not only by EGS but also by the 3 Å
DFDNB, and this binding is relatively strong as it is stable in octyl
glucoside. In contrast, the binding of UPII to another UPII is
relatively distant as the cross-linking required a long-armed reagent,
and the binding is relatively weak as it can be disrupted by octyl
glucoside. This raises the possibility that UPII interacts with UPIa
within a 16-nm particle, but with UPII of perhaps another particle (see
below). This also raises the possibility that a detergent's
ability to break the UPII:UPII interaction, which may be involved in
bridging the neighboring 16-nm particles (see below), enables the
detergent to solubilize the AUMs.
With
these questions in mind, it is interesting to note that the four
uroplakins form two pairs as defined by the two known (cross-linked)
heterodimers, i.e. UPIa/UPII and UPIb/UPIII. Each of these
dimers consists of a four transmembrane-domained member (UPIa or UPIb)
plus a single transmembrane-domained protein (UPII or UPIII). Although
all of the uroplakins appear to be able to form oligomers (Fig. 2Fig. 3Fig. 4Fig. 5) (Wu and Sun,
1993; Wu et al., 1994), so far we have not found conditions
under which we can cross-link UPIa to UPIII (instead of UPII), or UPIb
to UPII (instead of UPIII), suggesting a specificity in uroplakin
interaction that was not suspected previously. This specificity raises
the possibility that AUMs, despite the fact that they appear to be
morphologically homogeneous, may actually contain two distinct
populations of 16-nm protein particles, one composed of UPIa and UPII,
and another of UPIb and UPIII (Fig. 8).
Figure 8:
A schematic model showing the possible
existence of two types of AUM plaques containing 16-nm particles that
are composed of (a) uroplakins Ia and II and (b)
uroplakins Ib and III. a, in the UPIa/UPII model, the 27-kDa
UPIa and the 15-kDa UPII are hypothesized to occupy the inner and outer
domains of a 16-nm protein particle. This model can account for (i) the
oligomerization of UPIa, (ii) the cross-linking of UPIa/UPII
heterodimer, (iii) the cross-linking of UPII/UPII homodimer, and (iv)
the selective disruption of the UPII homodimer formation by octyl
glucoside. b, UPIb and UPIII occupy the inner and outer
domains, respectively, of the UPIb/UPIII model. This model can account
for the cross-linking of UPIb/UPIII heterodimer, as well as the
efficient cross-linking of UPIb/UPIb and UPIII/UPIII homodimers. The
stain-excluding map of bovine AUM was taken from Wu et al. (1994) (also see Walz et al.(1995)). For details, see the
text.
This kind of consideration also
suggests that the UPIa/UPII particles are not intermixed, within a
single plaque, with the other kind of UPIb/UPIII particles, because if
that were the case we should see the cross-linking of UPII of one
particle to the UPIII of a neighboring particle, and we have not yet
seen that. This raises the possibility that there are two types of
urothelial plaques, one consists purely of 16-nm particles containing
UPIa and UPII, while the other consists of particles containing UPIb
and UPIII. This hypothesis is schematically depicted in a working
model, shown in Fig. 8, that can account for all of the
available data. This model is attractive because it can solve two
puzzles. It can explain the redundancy of uroplakins, as the two TM4
family members, i.e. the UPIa and Ib, may actually interact
with different partners and thus play related but distinct roles in AUM
formation. This hypothesis can also solve the stoichiometry puzzle,
because it now predicts a molar relationship of UPIa = UPII and
UPIb = UPIII, thus allowing variations in the overall
stoichiometry, depending on the ratio of the two types of AUM plaques.
In addition, this model predicts that the two types of AUMs may play
different biological roles in terminally differentiated urothethelial
cells. For example, since of all the known uroplakins only the UPIII
has a long cytoplasmic domain, this uroplakin may play a role in
anchoring the AUM plaques into a cytoskeletal network (Wu and Sun,
1993). Is it then possible that only the UPIb/UPIII plaques can bind to
the cytoskeleton? Since uroplakin II is the only AUM protein that has a
long preprosequence, we need to consider the possibility that the UPII
prosequence may be involved in regulating AUM assembly in the Golgi
(Lin et al., 1994). Is it then possible that the assembly of
the UPIa/UPII plaques is regulated differently from that of the other
kind of plaques? Additional experiments are obviously needed to further
study the possible heterogeneity of AUMs and to address some of the
questions raised herein.
Imai and Yoshie(1993) have shown that CD81 and
CD82, which coexist in T cells, can be coimmunoprecipitated, suggesting
that they interact with each other forming a complex. Our finding that
UPIa and UPIb, two members of the TM4 family, interact with different
partners in AUM was therefore unexpected. Taken together, these data
indicate that members of the TM4 family, although structurally related,
have diverse structural and functional properties.
Volume 270,
Number 50,
Issue of December 15, 1995 pp. 29752-29759
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
)superfamily is a recently described gene family that
encodes a group of cell surface proteins all possessing four conserved
transmembrane domains. Members of this family are found in lymphocytic,
mesenchymal, and epithelial cells (reviewed by Horejsi and Vlcek(1991)
and Wright and Tomlinson(1994)). Thus the TM4 proteins that have been
identified so far include several tumor-related surface proteins CO-029
(Szala et al., 1990), L6 (Marken et al., 1992), SAS
(Jankowski et al., 1994), and R2 (Gaugitsch et al.,
1991) (the last was recently described as a metastasis suppressor gene
for prostate cancer (Dong et al., 1995));
leukocyte-differentiation markers CD9 (Boucheix et al., 1991;
Lanza et al., 1991), CD37 (Classon et al., 1990),
CD53 (Amiot, 1990; Korinek and Horejsi, 1993), CD63 (Hotta et
al., 1988; Metzelaar et al., 1991), CD81 (also known as
TAPA-1; Oren et al., 1990; Engel and Tedder, 1994), and CD82
(Lebel et al., 1994; Nagira et al., 1994); as well as
major epithelial differentiation products uroplakins Ia and Ib (Yu et al., 1994). Closely related molecules, SM23 and SJ23, have
even been found in parasitic helminth schistosomes (Davern et
al., 1991; Reynolds et al., 1992), indicating that
members of this gene family are conserved during evolution. Since the
intron positions of several of the TM4 genes are conserved, these genes
may have diverged from a common ancestral gene (Horejsi and Vlcek,
1991; Wright et al., 1993; Wright and Tomlinson, 1994). Recent
data indicate that some of these TM4 proteins may play important roles
in cell growth, adhesion, and metastasis (Horejsi and Vlcek, 1991;
Miyake et al., 1991; Schick and Levy, 1993; Wright and
Tomlinson, 1994; Dong et al., 1995). However, many crucial
questions regarding the structure and function of TM4 proteins remain
unanswered. For example, how do these integral membrane proteins, most
of them lacking a significant cytoplasmic domain, perform their
functions on the cell surface? In several cases, more than one TM4
protein exists in the same cell; thus CD81 and CD82 coexist in T cells
(Imai and Yoshie, 1993; Nagira et al., 1994), and uroplakins
Ia and Ib coexist in the differentiated urothelial cells (Yu et
al., 1994). In such cases, do these TM4 proteins always interact
with each other? Or do some of these TM4 proteins interact specifically
with other integral membrane proteins?
)and Ib (UPIb; 28-kDa; also known as TI-1) (Kallin et
al., 1991) are two newly identified TM4 proteins that are
synthesized by the terminally differentiated, superficially located
cells of mammalian urothelium (Yu et al., 1994). Together with
two other proteins, i.e. the 15-kDa uroplakin II (Lin et
al., 1994, 1995) and the 47-kDa uroplakin III (Wu and Sun, 1993),
the UPIa and UPIb are major protein components of the so-called
asymmetrical unit membrane (AUM) (Wu et al., 1990; Yu et
al., 1990), which forms numerous plaques covering about 80% of the
apical surface area of mammalian urothelium (Porter and Bonneville,
1963; Hicks, 1965; Porter et al., 1967; Koss, 1969; Chlapowski et al., 1972). These AUM plaques are believed to play a role
in stabilizing the luminal surface of the epithelium thus preventing it
from rupturing during bladder distention (Staehelin et al.,
1972; Minsky and Chlapowski, 1978; Sarikas and Chlapowski, 1986).
Recent biochemical data indicate that the major hydrophilic loop
interconnecting the third and fourth trans-membrane domains of UPIa and
UPIb is exposed on the extracellular surface, because this domain
becomes protected from protease digestion once the in vitro transcribed and translated UPIa is inserted into dog pancreatic
microsomes (Yu et al., 1994). In addition, in both UPIa and
UPIb, this domain contains an N-glycosylation site which
harbors high mannose type carbohydrates (Wu et al., 1994; Yu et al., 1994). These data strongly suggest that UPIa and UPIb,
like several other members of the TM4 superfamily, assume the so-called
``type III'' transmembrane configuration with the major
hydrophilic domain extending into the extracellular space leaving very
little cytoplasmic domains (Yu et al., 1994).
Isolation of Asymmetric Unit Membranes
To
isolate the crude membranes of bovine urothelium, we obtained bovine
bladder mucosa by scraping, washed the cells three times in
phosphate-buffered saline, and homogenized them in 10 mM Hepes/NaOH (pH 7.5) containing 1 mM each of
phenylmethylsulfonyl fluoride, EDTA, and EGTA. After centrifugation at
2,000 
g at 4 °C for 10 min, the pellet was
homogenized in the same buffer, transferred onto a 1.6 M sucrose cushion, and centrifuged at 28,000 g at 4
°C for 20 min. The membranes located at the interface were
collected, washed with 10 mM Hepes (pH 7.5), and were used as
the so-called ``crude membrane fraction'' for some of the
cross-linking experiments. To further purify the asymmetrical unit
membranes, we suspended these crude membranes in 2% Sarkosyl in 10
mM Hepes buffer (pH 7.5) at room temperature for 10 min and
recovered the (insoluble) asymmetrical unit membranes by centrifugation
at 18,000 g at 4 °C for 30 min (Wu et
al., 1990, 1994). A portion of the AUMs was further treated with
25 mM NaOH, followed by washing with 10 mM Hepes
buffer (pH 7.5). The membrane proteins were dissolved in 1% SDS and
quantitated using the BCA reagent (Pierce).Generation of Monospecific Antibodies to
Uroplakins
Rabbit antisera were raised against oligopeptides
that were synthesized according to the cDNA-deduced amino acid sequence
of bovine uroplakins Ia (DSNQGRELTRLWDRC); uroplakin Ib
(AKDDSSVRSFQGLLIFGNC); uroplakin II (CDSGSGFTVTRLSA and
SAYQVTNLAPGTKYYIC); and uroplakin III (CATSHDSQITQEAVPK). The
underlined, terminal cysteine residues were added so that these
peptides could be conjugated to keyhole limpet hemocyanin or bovine
serum albumin using m-maleimidobenzoyl-N-hydroxysuccinimide ester as a
cross-linking reagent. The carrier-peptide conjugates were used to
immunize rabbits as described earlier (Wu and Sun, 1993).Cross-linking of Uroplakins
Bifunctional
cross-linking reagents, including EGS, sulfo-EGS, and DFDNB, were
purchased from Pierce. They were dissolved immediately before use in
dry MeSO to make a 25 mM stock solution (Abdella et al., 1979), aliquots of which were then added to a membrane
suspension containing 0.1 mg of protein per ml of 10 mM Hepes/NaOH buffer (pH 7.5). After 2 h at room temperature, the
reaction was quenched by adding 1 M Tris/HCl (pH 7.4) to a
final concentration of 50 mM. The cross-linked reaction
mixtures were stored at -20 °C until further analysis.Cleavage and Two-dimensional Electrophoretic Analyses of
the Cross-linked Dimers
EGS cross-linked membrane proteins were
dissolved in 1% SDS, separated by SDS-PAGE on a 17% polyacrylamide gel
(acrylamide/bisacrylamide ratio, 120:1), and stained with Coomassie
Blue followed by destaining in 50% methanol and 7% acetic acid. The
lanes containing cross-linked proteins were incubated in 1 M NHOH in 50 mM sodium phosphate buffer (pH
8.5) at 37 °C for 6 h. The gel piece was then equilibrated in SDS
sample buffer (50 mM Tris/HCl, pH 6.8; 2% SDS; 10% glycerol,
and 5% 
-mercaptoethanol) at room temperature for 30 min and
subjected to a second dimensional SDS-PAGE (same condition as the first
dimension).Silver Nitrate Staining and Immunoblotting
The
polyacrylamide gel was fixed in a solution containing 50% methanol and
7% acetic acid for 2 h, soaked in 10% glutaraldehyde for 30 min, washed
extensively in distilled water, and then exposed to 20% silver nitrate
dissolved in 0.4% NaOH, 0.1% NH
OH, and 2% ethanol for 8
min. After washing in distilled water for 1 h, the gel was incubated in
a solution containing 0.02% formaldehyde, 10% ethanol, and 0.005%
citric acid. For immunoblotting, proteins separated by SDS-PAGE were
electrophoretically transferred to a nitrocellulose paper. After the
unoccupied sites of the paper were blocked with 5% milk in
phosphate-buffered saline, the paper was incubated with rabbit
antibodies against individual uroplakins followed by a
peroxidase-conjugated goat-anti-rabbit antibody (Surya et al.,
1990).
Cross-linking of Crude Urothelial Membranes
We
showed previously that by using a combination of differential
centrifugation and selective removal of contaminating non-AUM vesicles
with certain detergents, we could isolate large quantities of highly
purified bovine urothelial AUMs (Wu et al., 1990, 1994).
Although these AUMs morphologically resembled the urothelial plaques
found in situ, we could not rule out the possibility that the
isolation procedure, which involved relatively harsh treatments such as
washing with Sarkosyl and a high pH solution, might have altered the
AUM structure. We also could not rule out the possibility that certain
AUM-associated proteins might have been stripped off. To alleviate
these problems, we decided to begin by analyzing the neighboring
relationship of the uroplakins in native urothelial membranes that had
not experienced the detergent/alkaline treatments. We could accomplish
this by isolating crude membranes from bovine urothelium, cross-linking
their proteins using EGS, a bifunctional reagent that cross-links
neighboring amino groups, and monitoring the cross-linked status of
individual uroplakins by immunoblotting.
)
staining. b, proteins of crude urothelial membranes (odd-numbered lanes) and Sarkosyl-washed AUMs (even-numbered lanes) were electrophoretically transferred to
nitrocellulose and immunoblotted using (lanes 1 and 2) antibodies against a synthetic oligopeptide of UPIa; (3 and 4) anti-UPIb; (5 and 6) anti-UPII; (7 and 8) another anti-UPII; and (9 and 10) anti-UPIII. For the sequences of the synthetic
oligopeptides, see ``Materials and Methods.'' Numbers on the left denote the molecular weights (M.W.) of standard
proteins. The relative positions of the four major uroplakins (the
27-kDa UPIa, the 28-kDa UPIb, the 15-kDa UPII, and the 47-kDa UPIII)
are marked on the right. Note that most of the antibodies are
monospecific for their respective uroplakin antigens (see
text).
Cross-linking of Purified AUMs
To see whether the
same topological relationships existed in the purified AUMs, we
prepared a batch of bovine AUMs that had been washed with Sarkosyl. As
we showed earlier, these isolated AUMs contained predominantly the
27-kDa UPIa, the 28-kDa UPIb, the 15-kDa UPII, and the 47-kDa UPIII (Fig. 1a, lane 2; Table 1) (Lin et
al., 1994; Wu et al., 1994; Yu et al., 1994). An
additional wash of these AUMs with 25 mM NaOH further reduced
the level of two minor contaminant bands of 68 and 34 kDa (Fig. 1a, lanes 2 and 3).
Immunoblotting of these purified AUM preparations revealed a 48-kDa,
UPIa-related protein (Fig. 4a, lanes 1 and 2), which as we had reported earlier represented a homodimer
of UPIa which was found in various quantities in AUM preparations (Wu et al., 1994) (also see below). Cross-linking of these AUM
preparations using EGS resulted in the formation of a 22-kDa UPII
homodimer, a 35-kDa UPIa/UPII heterodimer, and a 48-kDa UPIa/UPIa
homodimer (Fig. 4), thus confirming and extending some of the
crude membrane data ( Fig. 2and Fig. 3).
), or immunoblotted with (b) anti-UPIa or (c) anti-UPII. Lanes 1 and 2 are side lanes showing the proteins of either control AUMs (lane 1) or EGS cross-linked AUMs (lane 2) that were
resolved only during the second dimensional SDS-PAGE. Arrows marked with 1 and 2 denote the directions of the
first and second dimensional SDS-PAGE. The molecular weights (MW) of the marker proteins are indicated on the right of Panel a. Note the cleavage of a 35-kDa cross-linked
protein (lane 2), giving rise to a 27-kDa UPIa and a 15-kDa
UPII (dashed lines). Also note the cleavage of a 22-kDa
cross-linked protein yielding a 15-kDa UPII (dotted lines). A
series of UPIa-related spots (horizontal arrows), that can be
seen above the diagonal in Panels a and b, represent
oligomerized UPIa (see text).
OH, and resolved the released uroplakins by a
second dimensional SDS-PAGE. In this procedure, only the monomers that
were released from a cross-linked product during the hydroxylamine step
would migrate below the diagonal (Fig. 5a). Such an
analysis revealed the existence of a 35-kDa, EGS cross-linked species
which was cleaved by a hydroxylamine releasing a 27-kDa uroplakin Ia
plus a 15-kDa uroplakin II (Fig. 5; see the circled protein
spots connected by a dashed line), thus confirming the
identity of the UPIa/UPII heterodimer. The results also clearly
established the presence of a 22-kDa cross-linked product that, upon
hydroxylamine treatment, released only a 15-kDa uroplakin II, thus
confirming the identity of the uroplakin II homodimer (Fig. 5; dotted line). Finally, we observed a series of UPIa oligomers
of 48 kDa (dimer) and 70 kDa (trimer), which apparently were formed
during the hydroxylamine treatment (Fig. 5, a, horizontal arrows, and b).Effects of Detergents and the Chain Length of
Cross-linking Reagents on Uroplakin Cross-linking
We have shown
recently using image enhancement techniques that each of the six outer
domains of the 16-nm protein particle of AUMs is actually
``connected,'' via some fine bridges, with an outer domain of
a neighboring particle, and we proposed that this extensive network of
inter-particle connections may account for the remarkable insolubility
of AUMs in a large number of detergents including Nonidet P-40, CHAPS,
deoxycholate, and Sarkosyl (Wu, et al., 1990; Walz et
al., 1995). The AUMs could be solubilized, however, to some extent
by Triton X-100 and almost completely by octyl glucoside (Wu et
al., 1990). It would therefore be of interest to see whether these
latter two detergents can disrupt, perhaps to different degrees,
certain uroplakin interactions as defined by cross-linking. To test
this, we treated both crude urothelial membranes and highly purified
AUMs with these two detergents and cross-linked the uroplakins using
EGS. Immunoblotting showed that Triton X-100 had negligible effects on
uroplakin cross-linkings (Fig. 6, a and b, and
7). In contrast, octyl glucoside greatly reduced the yield of
cross-linking of the uroplakin II homodimer, even though it had almost
no effect on the formation of the UPIa/UPII heterodimer (Fig. 6, a and b, and 7).
-lysine groups, we treated purified AUMs with
DFDNB, which has an arm length of only 3 Å (versus the
16 Å of EGS and sulfo-EGS). Like EGS, DFDNB cross-linked the
UPIa/UPII heterodimer and UPIa/UPIa homodimer. However, it failed to
cross-link the UPII/UPII dimer (Fig. 6c), suggesting
that the
-lysines in UPII/UPII cross-linking were >3 Å
apart.
UPII complex, and UPII yielded predominantly the UPIa/UPII
heterodimer and the UPII homodimer (Fig. 2Fig. 3Fig. 4). Such a relatively simple
cross-linking pattern of the purified AUMs is perhaps to be expected,
given the fact that AUMs are known to contain only four major protein
subunits (Wu et al., 1990, 1994; Yu et al., 1994). It
was unexpected, however, that crude urothelial membranes yielded no
additional cross-linked species, although of course this negative
finding does not rule out additional protein:protein interactions that
may exist in situ. Taken together, our results strongly
suggest that the uroplakin pairs that we have identified so far by the
cross-linking approach reflect important protein:protein interactions
that occur in the asymmetrical unit membrane.Uroplakin Interactions: A Nearest Neighbor Analysis
Our data thus established the existence of several uroplakin
pairs including homodimers and heterodimers. In addition, we found that
UPIa has a tendency to oligomerize, forming complexes that are stable
in SDS. The fact that UPIa and UPIb, two structurally related
molecules, were cross-linked to two different uroplakin partners
indicated a high degree of specificity in uroplakin interaction. The
apparent molecular weights of various uroplakin monomers and
cross-linked dimers are summarized in Table 1.The UPIa/UPII Heterodimer Formation
This cross-linked
pair ran as a well defined, 35-kDa band during SDS-PAGE (Fig. 2Fig. 3Fig. 4Fig. 5Fig. 6Fig. 7; Table 1); it reacted strongly with antisera to uroplakins Ia and
II (Fig. 2Fig. 3Fig. 4); and it was cleaved by
hydroxylamine giving rise to UPIa and UPII (Fig. 5). The
apparent molecular weight of this cross-linked product was slightly
smaller than the sum of its monomers (35 kDa versus 42 kDa),
which is a common occurrence. The fact that this UPIa/UPII heterodimer
was detected as a major cross-linked product of not only the purified
AUMs, but also the crude urothelial membranes (Fig. 2Fig. 3Fig. 4), suggests that the UPIa:UPII
interaction that is detected here via the cross-linking reaction is
likely to be physiological. This uroplakin pair can even be generated
with DFDNB, a much shorter cross-linking reagent with an arm length of
only 3Å (versus 16 Å of EGS; Fig. 6c), suggesting a close proximity of the
cross-linked -amino groups. On the other hand, the hydrophilic
analogue of EGS, i.e. the sulfo-EGS, failed to cross-link this
particular uroplakin pair, even though it could efficiently cross-link
some other uroplakins (see below). This suggests that the reactive
lysines may be embedded at least in part in a hydrophobic environment.
Taken together, these results clearly established that uroplakin Ia, a
member of the TM4 superfamily, interacts closely with uroplakin II, a
15-kDa ``type I'' integral membrane protein that is anchored
into the lipid bilayer via its single transmembrane domain located at
its C terminus (Lin et al., 1994).
The Uroplakin II Homodimer
This cross-linked
species ran as a well defined band of 22-kDa during SDS-PAGE (Fig. 2Fig. 3Fig. 4; Table 1), it reacted
with only anti-UPII (Fig. 2Fig. 3Fig. 4Fig. 5), and it was
cleaved by hydroxylamine yielding exclusively a monomeric UPII (Fig. 5). Both EGS and its hydrophilic analogue, the sulfo-EGS,
worked well in generating this UPII homodimer (Fig. 3),
indicating that the cross-linking reaction was relatively insensitive
to the hydrophobicity of the cross-linking reagent. On the other hand,
this dimer formation was highly dependent on the chain length of the
cross-linking reagent; while EGS and S-EGS (16 Å) worked well,
the shorter DFDNB (3 Å) was ineffective (Fig. 6c). Interestingly, the cross-linking of UPII
homodimer was abolished by octyl glucoside, which could effectively
dissolve the AUMs, but this cross-linking was only marginally affected
by Triton X-100 that only partially solubilized the AUMs.The Uroplakin Ia Oligomers
Some integral membrane
proteins can form complexes that are so stable that they migrate as
well defined oligomers during SDS-PAGE (Lemmon et al., 1992;
Treutlein et al., 1992; Arkin et al., 1994). An
example of this is glycophorin A which forms a dimer. This dimer
formation involves the precise packing of some of the amino acid side
chains of the single transmembrane domain, as it can be abolished by
even relatively conserved mutations of some of these side chains
(Treutlein et al., 1992). Another example is phospholamban, a
cardiac ion channel, which can oligomerize forming up to a pentamer
that is stable in SDS (Arkin et al., 1994). We have described
earlier that heating the AUM proteins can cause the uroplakins Ia and
Ib to form large aggregates, although this process was difficult to
control (Wu et al., 1990). We found now, quite unexpectedly,
that incubating the UPIa monomer in 1 M NHOH
resulted in the formation of well defined UPIa dimers and trimers ( Fig. 4and Fig. 5). The propensity of UPIa, which may
occupy the inner six domains of the 16-nm protein particle (see below),
to oligomerize may play a role in AUM assembly.The UPIb/UPIII Heterodimer
An interesting feature
of UPIb and UPIII, that distinguishes them from UPIa and UPII, is that
the former can be cross-linked much more efficiently than the latter.
Thus 0.5-1 mM EGS, which barely cross-linked UPIa and
UPII, yielded a nearly maximal amount of UPIb/UPIII heterodimer (Fig. 2, c and d). Increasing the EGS
concentration to 2-5 mM led to the complete
cross-linking of UPIb to form a homodimer and higher oligomers, the
significant cross-linking of UPIII to form a homodimer, and the
disappearance of the UPIb/UPIII heterodimer most likely due to its
conversion to higher oligomers (Fig. 2). Although these
cross-linked species had not been characterized as thoroughly as those
involving UPIa and UPII, our data clearly showed that UPIb and UPIII,
which could be cross-linked extremely efficiently to themselves and to
each other, were not cross-linked to UPIa and UPII (Fig. 2Fig. 3Fig. 4). That UPIa and UPIb, two
closely related members of the TM4 family, were cross-linked to
different partners in AUMs raised the interesting possibility that they
played different roles in AUM formation (see below).Possible Heterogeneity of AUM Plaques
Perhaps one of the most intriguing aspects of AUM structure
is: why do AUMs have two closely related UPIa and UPIb which share 39%
of their amino acid sequences (Yu et al., 1994), as well as
two ``type I proteins,'' the UPII and UPIII, which share a
stretch of 12 amino acids located on the extracellular side of their
single transmembrane domains (Wu and Sun, 1993; Lin et al.,
1994)? Another intriguing feature has to do with protein stoichiometry.
Given the highly organized structure of AUM, one expects a precise and
fixed stoichiometry of its protein subunits. Although as we have
pointed out recently, the significantly different color yields of
uroplakins when stained with Coomassie Blue and silver nitrate
complicate their quantitation (Wu et al., 1994), it now
appears that the ratio of various uroplakins is quite variable.
The UPIa/UPII Particle: A Model
As we and others have shown earlier, each 16-nm particle of
the AUM can be resolved into 12 stain-excluding domains, that are
arranged in an inner ring of six and outer ring of six (Hicks and
Ketterer, 1969, 1970; Vergara et al., 1969; Robertson and
Vergara, 1980; Brisson and Wade, 1983; Walz et al., 1995).
Taylor and Robertson(1984) calculated that the volume of each inner
domain is 1.6 times larger than that of an outer domain. If indeed
there exist two types of 16-nm particles, one of them consisting of
UPIa and UPII, it would be interesting to consider a model in which
each (larger) inner domain contains a 27-kDa UPIa, while each (smaller)
outer domain contains a 15-kDa UPII (Fig. 8a). Since
this model depicts a central ring of six interconnected UPIa molecules (Fig. 8a), this attaches a possible significance to the
observation that UPIa has a propensity to form dimers, trimers, and
higher oligomers that are stable even in SDS (Fig. 5a, horizontal arrows). Since, as we have shown recently, each
outer domain is connected via some fine ``bridges'' to an
outer domain of a neighboring 16-nm particle (Walz et al.,
1995), this model predicts the cross-linking of UPIIs from two
neighboring 16-nm particles resulting in the formation of dimers but no
higher oligomers (Fig. 8a), which is indeed what we
observed (Fig. 2Fig. 3Fig. 4Fig. 5). Also,
since this kind of UPII dimer formation involves protein:protein
interactions across two neighboring 16-nm particles (Fig. 8a), one may expect that this requires a longer
armed cross-linking reagent and is more susceptible to detergent
disruption than the formation of UPIa/UPII dimer which involves only
intraparticle interactions; this is indeed what we observed ( Fig. 6and Fig. 7).Complex Formation Involving Other TM4 Proteins
Our finding that UPIa, a TM4 protein, can form a specific
complex with a small integral membrane protein, the UPII, is not unique
among the TM4 proteins. For example, the 26-kDa TAPA-1 is known to
interact with a 16-kDa Leu-13 antigen in leukocytes and activated
endothelial cells (Takahashi et al., 1990; Matsumoto et
al., 1993). In another instance, a 24-kDa CD9 has been shown to
bind to a 14.5-kDa diphtheria toxin receptor (DTR) which can also serve
as the precursor of a heparin-binding EGF-like growth factor
(HB-EGF/DTR) (Mitamura et al., 1992; Brown et al.,
1993; Iwamoto et al., 1994). In addition to enhancing the
diphtheria toxin binding to its receptor, CD9 can potentiate the
juxtacrine growth factor activity of membrane-bound HB-EGF/DTR
(Higashiyama et al., 1995). These results suggest that
although TM4 proteins themselves lack an appreciable cytoplasmic
domain, some of them can modulate the biological function of another
integral membrane protein that has a cytoplasmic domain. Interestingly,
the HB-EGF/DTR and UPII precursor, both of which bind to TM4 proteins,
share some structural features (Naglich et al., 1992; Lin et al., 1994). Both have a cleavable signal peptide, followed
by a prosequence that can potentially be removed by furin-like enzymes,
both possess a single transmembrane domain located near the C terminus
of the mature protein, and both are relatively basic (pro-UPII and
HB-EGF have pI of 11.1 and 9.9, respectively). Whether these limited
structural similarities are significant, or are merely coincidental, is
currently unknown.Concluding Remarks and Perspectives
In conclusion, we have shown that the four major structural
proteins of AUM, i.e. uroplakins Ia, Ib, II, and III, can be
divided into two pairs consisting of UPIa/UPII and UPIb/UPIII. We
propose that these two uroplakin pairs can form two types of 16-nm
protein particles that may assemble into plaques with different
assembly and functional properties. This model can account for all the
existing data, and it can explain several previously puzzling features
of the AUM. Moreover, it makes specific predictions that can be tested
experimentally. For example, it would now be interesting to determine
whether antibodies monospecific for individual uroplakins (see, e.g.Fig. 1) would decorate only a subset of urothelial
plaques, whether antibodies to UPIa and UPIb would preferentially
associate with the inner domains while those against UPII and UPIII
with the outer domains of some of the 16-nm AUM particles, and whether
one can reconstitute the 16-nm protein particles with specific pairs of
uroplakins (e.g. UPIa plus UPII). Regardless to what extent
the current model will prove to be correct, these experiments should
greatly advance our understanding on the structure and function of AUM,
a fascinating and truly unique biomembrane.
))
We thank Drs. Herbert Lepor and Irwin M. Freedberg for
their continued interest in and support of this project.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  D. Chakraborty, X.-Q. Ding, S. M. Conley, S. J. Fliesler, and M. I. Naash Differential requirements for retinal degeneration slow intermolecular disulfide-linked oligomerization in rods versus cones Hum. Mol. Genet., March 1, 2009; 18(5): 797 - 808. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Bari, Y. H. Zhang, F. Zhang, N. X. Wang, C. S. Stipp, J. J. Zheng, and X. A. Zhang Transmembrane Interactions Are Needed for KAI1/CD82-Mediated Suppression of Cancer Invasion and Metastasis Am. J. Pathol., February 1, 2009; 174(2): 647 - 660. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. V. Kovalenko, X. H. Yang, and M. E. Hemler A Novel Cysteine Cross-linking Method Reveals a Direct Association between Claudin-1 and Tetraspanin CD9 Mol. Cell. Proteomics, November 1, 2007; 6(11): 1855 - 1867. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Tu, X.-P. Kong, T.-T. Sun, and G. Kreibich Integrity of all four transmembrane domains of the tetraspanin uroplakin Ib is required for its exit from the ER J. Cell Sci., December 15, 2006; 119(24): 5077 - 5086. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T.-T. Sun Altered phenotype of cultured urothelial and other stratified epithelial cells: implications for wound healing Am J Physiol Renal Physiol, July 1, 2006; 291(1): F9 - F21. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Min, H. Wang, T.-T. Sun, and X.-P. Kong Structural basis for tetraspanin functions as revealed by the cryo-EM structure of uroplakin complexes at 6-A resolution J. Cell Biol., June 19, 2006; 173(6): 975 - 983. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Xie, G. Zhou, S.-Y. Chan, E. Shapiro, X.-P. Kong, X.-R. Wu, T.-T. Sun, and C. E. Costello Distinct Glycan Structures of Uroplakins Ia and Ib: STRUCTURAL BASIS FOR THE SELECTIVE BINDING OF FimH ADHESIN TO UROPLAKIN Ia J. Biol. Chem., May 26, 2006; 281(21): 14644 - 14653. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C.-C. A. Hu, F.-X. Liang, G. Zhou, L. Tu, C.-H. A. Tang, J. Zhou, G. Kreibich, and T.-T. Sun Assembly of Urothelial Plaques: Tetraspanin Function in Membrane Protein Trafficking Mol. Biol. Cell, September 1, 2005; 16(9): 3937 - 3950. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Levy and T. Shoham Protein-Protein Interactions in the Tetraspanin Web Physiology, August 1, 2005; 20(4): 218 - 224. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Sakakibara, K.-i. Sato, K.-i. Yoshino, N. Oshiro, S. Hirahara, A. K. M. M. Hasan, T. Iwasaki, Y. Ueda, Y. Iwao, K. Yonezawa, et al. Molecular Identification and Characterization of Xenopus Egg Uroplakin III, an Egg Raft-associated Transmembrane Protein That Is Tyrosine-phosphorylated upon Fertilization J. Biol. Chem., April 15, 2005; 280(15): 15029 - 15037. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X.-T. Kong, F.-M. Deng, P. Hu, F.-X. Liang, G. Zhou, A. B. Auerbach, N. Genieser, P. K. Nelson, E. S. Robbins, E. Shapiro, et al. Roles of uroplakins in plaque formation, umbrella cell enlargement, and urinary tract diseases J. Cell Biol., December 20, 2004; 167(6): 1195 - 1204. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Duncan, G. Li, J.-S. Shin, J. L. Carson, and S. N. Abraham Bacterial Penetration of Bladder Epithelium through Lipid Rafts J. Biol. Chem., April 30, 2004; 279(18): 18944 - 18951. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. L. Varley, J. Stahlschmidt, W.-C. Lee, J. Holder, C. Diggle, P. J. Selby, L. K. Trejdosiewicz, and J. Southgate Role of PPAR {gamma} and EGFR signalling in the urothelial terminal differentiation programme J. Cell Sci., April 15, 2004; 117(10): 2029 - 2036. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Chen, X. Guo, F.-M. Deng, F.-X. Liang, W. Sun, M. Ren, T. Izumi, D. D. Sabatini, T.-T. Sun, and G. Kreibich Rab27b is associated with fusiform vesicles and may be involved in targeting uroplakins to urothelial apical membranes PNAS, November 25, 2003; 100(24): 14012 - 14017. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Min, G. Zhou, M. Schapira, T.-T. Sun, and X.-P. Kong Structural basis of urothelial permeability barrier function as revealed by Cryo-EM studies of the 16 nm uroplakin particle J. Cell Sci., October 15, 2003; 116(20): 4087 - 4094. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Tu, T.-T. Sun, and G. Kreibich Specific Heterodimer Formation Is a Prerequisite for Uroplakins to Exit from the Endoplasmic Reticulum Mol. Biol. Cell, December 1, 2002; 13(12): 4221 - 4230. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F.-M. Deng, F.-X. Liang, L. Tu, K. A. Resing, P. Hu, M. Supino, C.-C. A. Hu, G. Zhou, M. Ding, G. Kreibich, et al. Uroplakin IIIb, a urothelial differentiation marker, dimerizes with uroplakin Ib as an early step of urothelial plaque assembly J. Cell Biol., November 25, 2002; 159(4): 685 - 694. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. R. Kazarov, X. Yang, C. S. Stipp, B. Sehgal, and M. E. Hemler An extracellular site on tetraspanin CD151 determines {alpha}3 and {alpha}6 integrin-dependent cellular morphology J. Cell Biol., September 29, 2002; 158(7): 1299 - 1309. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Seigneuret, A. Delaguillaumie, C. Lagaudriere-Gesbert, and H. Conjeaud Structure of the Tetraspanin Main Extracellular Domain. A PARTIALLY CONSERVED FOLD WITH A STRUCTURALLY VARIABLE DOMAIN INSERTION J. Biol. Chem., October 19, 2001; 276(43): 40055 - 40064. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Hu, F.-M. Deng, F.-X. Liang, C.-M. Hu, A. B. Auerbach, E. Shapiro, X.-R. Wu, B. Kachar, and T.-T. Sun Ablation of Uroplakin III Gene Results in Small Urothelial Plaques, Urothelial Leakage, and Vesicoureteral Reflux J. Cell Biol., November 20, 2000; 151(5): 961 - 972. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Adachi, K. Okubo, and S. Kinoshita Human Uroplakin Ib in Ocular Surface Epithelium Invest. Ophthalmol. Vis. Sci., September 1, 2000; 41(10): 2900 - 2905. [Abstract] [Full Text]
|
|
|
|
|
|
S. A. Lewis Everything you wanted to know about the bladder epithelium but were afraid to ask Am J Physiol Renal Physiol, June 1, 2000; 278(6): F867 - F874. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Mulvey, Y. S. Lopez-Boado, C. L. Wilson, R. Roth, W. C. Parks, J. Heuser, and S. J. Hultgren Induction and Evasion of Host Defenses by Type 1-Piliated Uropathogenic Escherichia coli Science, November 20, 1998; 282(5393): 1494 - 1497. [Abstract] [Full Text]
|
|
|
|
|
|
R. L. Yauch, F. Berditchevski, M. B. Harler, J. Reichner, and M. E. Hemler Highly Stoichiometric, Stable, and Specific Association of Integrin alpha 3beta 1 with CD151 Provides a Major Link to Phosphatidylinositol 4-Kinase, and May Regulate Cell Migration Mol. Biol. Cell, October 1, 1998; 9(10): 2751 - 2765. [Abstract] [Full Text]
|
|
|
|
|
|
J. Pak, Y. Pu, Z.-T. Zhang, D. L. Hasty, and X.-R. Wu Tamm-Horsfall Protein Binds to Type 1 Fimbriated Escherichia coli and Prevents E. coli from Binding to Uroplakin Ia and Ib Receptors J. Biol. Chem., March 23, 2001; 276(13): 9924 - 9930. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Hemler Specific tetraspanin functions J. Cell Biol., December 24, 2001; 155(7): 1103 - 1108. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |