Sie sind auf Seite 1von 12

Copyright # Blackwell Munksgaard 2004

Traffic 2004; 5: 117–128

Blackwell Munksgaard doi: 10.1046/j.1600-0854.2003.00156.x


The Uroepithelium: Not Just a Passive Barrier

Gerard Apodaca with diameters of 25–250 mm (Figure 1). In some species,

such as rat and guinea pig, multinucleate cells are com-
Renal-Electrolyte Division of the Department of Medicine, mon, and like intermediate cells they can also have thin
Laboratory of Epithelial Cell Biology, and Department of projections that connect them to the basement membrane
Cell Biology and Physiology, University of Pittsburgh, (3). Although umbrella cells are long lived, they are rapidly
Pittsburgh, PA 15261, USA, regenerated when the uroepithelium is damaged. This
regeneration can result from cell division within any of
The uroepithelium lines the inner surface of the renal the three cell layers, and generation of the multinucleate
pelvis, the ureters, and the urinary bladder, where it umbrella cells is likely the result of intermediate cell–cell
forms a tight barrier that allows for retention of urine, fusion (2,3).
while preventing the unregulated movement of ions,
solutes, and toxic metabolites across the epithelial bar- A primary function of epithelia, including the uroepithe-
rier. In the case of the bladder, the permeability barrier lium, is to form a barrier that prevents entry of pathogens
must be maintained even as the organ undergoes cyclical
and selectively controls the passage of water, ions,
changes in pressure as it fills and empties. Beyond
furthering our understanding of barrier function, new
solutes, and large macromolecules across the mucosal
analysis of the uroepithelium is providing information surface of the cell into the underlying tissue. Barrier func-
about how detergent-insoluble membrane/protein tion depends, in part, on the presence of specialized mem-
domains called plaques are formed at the apical plasma brane domains that form a seal between the plasma
membrane of the surface umbrella cells, how mechanical membranes of adjacent epithelial cells. In the case of the
stimuli such as pressure alter exocytic and endocytic uroepithelium, high resistance tight junctions are found in
traffic in epithelial cells such as umbrella cells, and how the umbrella cell layer that effectively divide the cell sur-
changes in pressure are communicated to the underlying face of these cells into apical and basolateral membrane
nervous system.
domains (4). In addition, the apical membrane of umbrella
cells has a unique lipid and protein composition that also
Key words: bladder, endocytosis, epithelium, exocytosis,
mechanical stimuli, neural–epithelial interactions, uro- contributes to the low permeability of this membrane
epithelium, uroplakin, pressure domain to water and solutes (4–6).

Received and accepted for publication 3 November 2003 The uroepithelium was long treated as an impermeable
cellular plastic coating that allowed for urine storage, and
the majority of analyses of the lower urinary tract have
Epithelial cells cover external surfaces or line the inner focused on the bladder musculature and innervation. A
surfaces of organ systems such as the gut, blood vessels, renewed interest in the uroepithelium indicates that it
and urogenital tract. In the case of the lower urinary tract can alter the ion and protein composition of the urine
(including the renal pelvis, ureters, and bladder), the sur- (4,7,8), and study of the uroepithelium is providing new
face is coated by a specialized epithelium called the uro- insight into how specialized membrane domains are
epithelium. The uroepithelium is stratified and is comprised assembled, how epithelial cells sense and respond to
of three cell types including basal cells, intermediate cells, mechanical stimuli such as pressure, and how epithelial
and umbrella cells. Basal cells are small ( 10 mm in dia- cells may communicate mechanical stimuli to the nervous
meter), form a single layer that contacts the underlying system. This review will summarize new information per-
connective tissue and capillary bed, and serve as precursors tinent to each of these areas of inquiry.
for the other cell layers. Their estimated half-life is
3–6 months, although estimates are hard to make because
their mitotic index is very low (on the order of 0.1–0.5%)
(1,2). Intermediate cells are pyriform in shape (10–25 mm in Specialized Membrane Domains at the Apical
diameter), sit on top of the underlying basal cells, and form a Surface of the Umbrella Cell
layer that appears in cross-section anywhere from one to
several cell layers thick. In some species the intermediate The apical surface of umbrella cells contains unique struc-
cells have long, thin cytoplasmic processes that connect tural and biochemical features. When examined by scan-
them to the basement membrane (1–3). The outermost ning electron microscopy the surface of the umbrella cells,
umbrella cell layer is comprised of very large polyhedral cells such as those from rabbit bladders, appears pleated and





IC 10 µm

20 µm

Figure 1: The umbrella and intermediate cell layers. A

transmission electron micrograph of rabbit uroepithelium shows a
bi-nucleate umbrella cell and underlying intermediate cells. The location
of the tight junction between adjacent umbrella cells is marked with H
an arrow. Legend: UC, umbrella cell; IC, intermediate cell.

each cell is surrounded by a tight junctional ring (Figure 2A). 5 µm

Higher magnification views show that the surface is cov-
ered by raised ridges, also called hinges or microplicae,
Figure 2: Ultrastructure of umbrella cell apical membrane. (A)
and intervening areas called plaques (Figure 2B). The
Scanning electron micrograph of mucosal surface of rabbit
arrangement of hinges and plaques give the apical surface umbrella cell layer. The tight junction ring of an individual
its characteristic scalloped appearance, which is apparent umbrella cell is demarcated by arrows. (B) High magnification
when the apical surface of cross-sectioned umbrella cells view of apical surface of umbrella cell. Examples of hinges (‘H’)
is viewed by transmission electron microscopy (Figure 3A). are marked with arrows.
The hinge areas are not well understood, but contain at
least one unique protein called urohingin (9), and presum- subdomains may provide important clues to how
ably all other non-plaque proteins. Plaques are thought to detergent-insoluble rafts are formed and how they function.
occupy approximately 70–90% of the surface of the
umbrella cell (2,10,11). An additional characteristic of the membrane associated
with the plaque regions is that the outer leaflet appears to
The membrane associated with the hinge and plaque be twice as thick as the inner leaflet, thus forming an
regions is highly detergent insoluble, even in relatively asymmetric unit membrane (AUM; Figure 3B) (10,15,16).
harsh detergents like sarkosyl (12). The detergent insolu- The AUM is composed of a paracrystalline array of 16-nm
bility, described some 30 years ago, may reflect the diameter AUM particles that are apparent when detergent-
unusual lipid composition of this membrane, which is rich solubilized membranes are negatively stained, or when the
in cholesterol, phosphatidyl choline, phosphatidyl ethanola- apical membrane is examined using high-resolution, quick-
mine, and cerebroside – a lipid profile similar to myelin freeze, deep-etch techniques (Figure 4). AUM particles
(13). Cholesterol-rich and detergent-insoluble membranes exhibit six-fold symmetry, are composed of an inner ring
such as ‘rafts’ and caveolae have recently received containing six large particles and an outer ring containing
considerable attention (14). Essentially, the entire apical six small particles, and each subunit forms a twisted ribbon
surface of the umbrella cell is composed of two lipid raft structure (16). A plaque is comprised of approximately
subdomains: plaques and hinges, and study of these 1000–3000 AUM particles.

118 Traffic 2004; 5: 117–128

The Uroepithelium

50 nm

100 nm

Figure 4: Asymmetric unit membrane (AUM) particles at the

apical surface of the umbrella cell. A rapid-freeze, deep-etch
micrograph shows hinge areas (‘H’) and plaques (‘P’). Inset: higher
magnification view of AUM particles found in plaque region.

(17,18). The latter was only recently described (18). UPIa,

UPII, UPIIIa, and UPIIIb are only expressed in the uro-
epithelium and are concentrated in the umbrella cell layer.
UPIb is also expressed in the cornea and conjunctiva (19).
UPIa serves as a receptor for uropathogenic Escherichia
coli (20,21). How UPs combine to generate the six-fold
1 µm symmetry of the AUM particles is currently unknown.
Furthermore, the currently described UPs may not be the
sole constituents of AUM particles, as there are other
proteins associated with plaques that have not been
B characterized. These include the antigen recognized by
the AE-31 monoclonal antibody, an uncharacterized 27 kDa
protein localized to plaques and originally thought to be
UPI (22).

Assembly of plaque proteins

The membrane trafficking and assembly of UPs into AUM
particles is only now being analyzed. In general, this ana-
lysis has been hampered by the lack of a polarized uro-
epithelial cell line that forms AUM particles, although a
recently described primary uroepithelial culture model
that is highly polarized, expresses UPs, and forms AUMs
0.5 mm
may be useful in this regard (23). Biochemical experiments
using purified proteins established that UPIa pairs with
UPII and UPIb pairs with UPIIIa or UPIIIb (18,24). The
Figure 3: Surface specializations at the apical surface of the
physiological significance of these paired interactions
umbrella cell. (A) Transmission electron micrograph of cross-
sectioned rat umbrella cell apical membrane demonstrating hinge was recently elucidated when UPs were ectopically
and plaque regions. Hinge areas are marked with arrows. The expressed alone or in combination in human embryonic
intervening areas of plasma membrane are plaques. The boxed area kidney 293T cells (18,25). The only singly expressed UP
is magnified in panel B. (B) High magnification view of asymmetric that can exit the endoplasmic reticulum (ER) and be deliv-
unit membrane (AUM). Hinges are marked with arrows, and a ered to the cell surface is UPIb. Exit from the ER and
plaque with readily identifiable AUM is marked with an arrowhead. surface delivery of UPII is dependent on coexpression
with UPIa, and exit and delivery of UPIIIa or UPIIIb is
Potential constituents of the AUM particles include the dependent on coexpression of UPIb. These results indi-
uroplakins (UPs), a family of at least five proteins including cate that the initial steps in AUM particle assembly involve
the tetraspan family members UPIa and UPIb, and the type heterodimer formation within the ER. It is unknown if
I single-span proteins UPII, UPIIIa, and UPIIIb (Figure 5) heterodimer formation occurs in umbrella cells or if ectopic

Traffic 2004; 5: 117–128 119



Figure 5: The uroplakins. UPIa and UPIb are members of the tetraspanning family of proteins that cross the plasma membrane four times.
UPII, UPIIIa, and UPIIIb are single-pass type I membrane proteins. They share a conserved domain (shown in red) at the start of the
transmembrane region that in UPIIIa and UPIIIb extends toward the N-terminus. The region of the extracellular domain of UPIIIb colored in
yellow is >90% identical to a portion of the human DNA mismatch repair enzyme-related PMSR6 protein. The small green circles represent
potential sites for N-linked glycan addition. The brackets denote known pairing interactions between UP isoforms. Figure adapted from
reference (18).

expression of all five known UPs leads to AUM particle apical membrane of umbrella cells. Consistent with such a
formation. Also unknown is if AUM assembly occurs solely role, the apical membranes of umbrella cells from UPIIIa
in the ER or whether assembly and maturation occur in knockout animals show increased permeability to the nor-
other organelles as well. Early studies by Hicks indicated mally membrane impermeant dye methylene blue (27).
that packaging of AUM particles into vesicles might occur Recent biophysical analyses confirm these findings and
in the Golgi apparatus of umbrella cells (26). demonstrate that bladders from knockout animals have a
significant increase in water and urea permeability across
An additional system that may provide information con- the umbrella cell layer; however, junctional permeability
cerning plaque assembly is the recent generation of UPIIIa appeared to be unchanged (6). These results are the first
knockout mice (27). The umbrella cells of these mice con- indication that integral membrane proteins may contribute
tain few plaques, and those plaques remaining are small in to the apical membrane permeability barrier of the uro-
size. The ability to form any plaques may reflect the ability epithelium. Because UPIIIa deficiency is associated with
of UPIIIb to substitute, albeit poorly, for UPIIIa, although several growth and developmental defects that are
this remains to be established. Alternatively, the UPIa– described below, it is important to confirm that the lipid
UPII complex may be able to form AUM particles ineffi- composition of the apical membrane is not affected by
ciently in the absence of UPIII. In fact, UPIa and UPII are UPIIIa depletion, and that UPIIIa alone or in combination
detected at the apical surface of the UPIII-deficient with other plaque proteins can change the permeability
umbrella cells. In contrast, UPIb is not found to accumulate across other membranes including reconstituted lipo-
at the surface of these cells. This latter observation is somes of a defined lipid composition.
somewhat inconsistent with the cell surface delivery of
UPIb observed in HEK293T cells (25), and may indicate Lack of UPIIIa expression has other dramatic conse-
that traffic of UPs in umbrella cells may be different than quences, including formation of a hyperplastic uroepithe-
that observed in heterologous non-polarized cell systems. lium with small umbrella cells (consistent with increased
It will be intriguing to determine which steps in plaque turnover of the epithelium), enlarged ureters, and vesicour-
assembly can occur in the absence of UPIIIa, and whether eteral reflux (leakage of urine from the bladder back to the
plaque formation occurs in animals lacking expression of kidneys) (27). Although vesicoureteral reflux is a hereditary
both UPII and UPIII. disease that affects 0.5–1.0% of the population, the
genetics of this condition are not well understood (28).
Function of plaques The knockout animals indicate that deficiencies in plaque
Although plaques and attendant AUM particles have been subunits or plaque formation may account for some of
known for decades, the role of these structures is not well these genetic abnormalities. Why UPIIIa deficiency leads
understood. However, the UPIIIa knockout mice described to these growth and developmental defects is unknown. It
above are providing some intriguing and somewhat could simply reflect the altered permeability barrier, or
unexpected roles for UPIIIa and plaque function. there may be other causes. For example, the dystrophin
complex links plasma membrane proteins to the under-
A previously ascribed function to plaques is that they may lying cytoskeleton of muscle cells, and disruption of this
help maintain the permeability barrier associated with the complex leads to muscular dystrophy (29). This disease is

120 Traffic 2004; 5: 117–128

The Uroepithelium

characterized by growth defects in muscle fibers including

small size, hypertrophy, and leaky plasma membranes. A
Compared to the other UPs, UPIIIa has a relatively long
cytoplasmic domain that has been proposed to interact
with the cytoskeleton (27). Although this interaction may
be indirect, and remains to be demonstrated, UPIIIa inter-
actions with the cytoskeleton could stabilize membrane
domains important for normal plasma membrane function
in umbrella cells.

Another function ascribed to plaques is that they may

modulate the apical surface area of the umbrella cell by DV
regulating insertion (during filling) and recovery (during
voiding) of plaque membrane (15). In addition to the apical
plasma membrane, UPs are also found in a population of
vesicles that occupy the cytoplasm of the umbrella cell and
have, depending on species, either a fusiform or discoidal
appearance in cross-section. In mice, the mature vesicles
have a fusiform appearance, while in rabbits the vesicles 0.5 µm
are discoidal in shape (Figure 6). Fusiform/discoidal vesi-
cles are thought to be formed in the Golgi apparatus (26),
and as will be discussed below are likely involved in
transport of UPs and other secretory cargo to the apical B
surface of the umbrella cells (30,31). Whether UPs are
required for fusiform/discoidal vesicle formation or fusion
with the apical plasma membrane is an open question.
Although beyond the scope of this review, there is
emerging evidence that cargo proteins may modulate
their intracellular transport (32). The cytoplasm of umbrella
cells from UPIIIa knockout mice is replete with ‘immature’
vesicles that have a discoidal instead of fusiform appear-
ance (27), indicating that significant vesicle formation FV
is occurring even in the absence of UPIIIa expression.
However, it is unknown if these vesicles contain other UP
cargo and whether they undergo exocytosis in a normal
manner. These questions can be readily answered using
the morphological and electrical techniques described below.

Response of the Uroepithelium to Bladder 0.5 µm


An area of cell biology that is again receiving attention is Figure 6: Discoidal and fusiform vesicles in the apical
the response of the uroepithelium to cyclical changes in cytoplasm of rabbit and mouse umbrella cell. (A) Discoidal
vesicles (DV) are found in rabbit umbrella cells. (B) Fusiform
hydrostatic pressure as the bladder fills and empties. In the
vesicles (FV) are more common in mouse umbrella cells.
mammalian bladder, pressure rises in a tri-phasic manner
as the organ fills with urine. The first rise occurs rapidly and
then pressure remains relatively constant for an extended The increased urine volume is accommodated by the uro-
period of time called the storage phase. In rabbits, for epithelium in at least two ways. The chief mechanism is
example, this phase lasts 4–5 h on average but can extend likely to be unfolding of the mucosal surface, which is
to 12–15 h (33,34). The storage phase is followed by the highly wrinkled in the empty bladder (Figure 7). The other
micturition phase, which is characterized by a rapid rise in mechanism occurs at the cellular level and involves transi-
bladder pressure, punctuated by large spikes in pressure tions in the morphology and function of the uroepithelium.
as the smooth muscle contracts. Upon voiding, the pres- As the bladder fills, the uroepithelium becomes thinner,
sure returns to baseline and the process begins anew. A apparently the result of intermediate and basal cells being
crucial aspect of the barrier function of the uroepithelium is pushed laterally to accommodate the increased urine
that it must be maintained in the face of these changes in volume (2,3). The umbrella cells undergo a large shape
hydrostatic pressure. change that involves progression from a roughly cuboidal

Traffic 2004; 5: 117–128 121


with a significant decrease in vesicle surface area (30).

This change occurs gradually (over a 5-h time period),
indicating that exocytosis is a graded process. It was also
noted that little change in surface area or secretion of
S-labeled proteins occurred in the absence of pressure,
and that pressure-induced exocytosis of 35S-labeled
proteins was sensitive to brefeldin A (30). No significant
changes were observed in basolateral surface area.
Fourth, as described above, UPs are major constituents
of discoidal/fusiform vesicles (30,38) and, as predicted,
increased amounts of UPIII are found at the cell surface
of umbrella cells exposed to hydrostatic pressure (30).

It remains to be formally shown in a rigorous pulse-chase

100 µm analysis that discoidal/fusiform vesicle membrane proteins
(e.g. UPs) are indeed fusing with the plasma membrane in
Figure 7: Scanning electron micrograph of mucosal surface
response to increased pressure. It is also possible that
of the rabbit bladder. Note the large folds. other membrane-bound organelles are fusing in response
to pressure. Also unknown is whether intermediate and
morphology in the empty bladder to one that is flat and basal cells undergo adjustments in cell volume or surface
squamous in the filled bladder (2,3). In the classical model area to accommodate increased urine volume. Intriguingly,
for vesicle dynamics, the umbrella cell shape transforma- the 1 or 2 layers of intermediate cells closest to the
tion is hypothesized to be accompanied by discoidal/ umbrella cell layer also possess small vesicles that have
fusiform vesicle exocytosis (Figure 8A). This would a discoidal-like appearance (Figure 9). Whether these vesi-
increase the apical surface area of the umbrella cell and cles contain UPs and whether they turnover during bladder
the overall surface area of the bladder, allowing the filling have not been determined.
bladder to accommodate additional urine volume (30,35).
Upon voiding, it is hypothesized that apical membrane
Pressure-induced endocytosis in umbrella cells
added during filling is rapidly internalized, replenishing
Two pieces of data are inconsistent with the classical
the pool of discoidal vesicles. An alternative model pro-
model described above (30). First, the amount of mem-
poses that there are no changes in umbrella cell surface
brane added to the apical membrane of umbrella cells
area and, instead, changes in umbrella cell shape are
when the epithelium is exposed to pressure
accomplished by folding/unfolding of the apical plasma
( 1500 mm2) is significantly less than the amount of
membrane (36).
membrane present in the steady-state pool of discoidal
vesicles ( 7000 mm2). Second, the total amount of cell-
Exocytosis in response to increased hydrostatic associated UPIII decreases significantly after exposing
pressure the uroepithelium to pressure for 5 h. This conundrum
The current data are most consistent with the classical was solved when biotin protection assays and lectin inter-
model’s tenet that pressure induces exocytosis of nalization assays were used to demonstrate that
discoidal/fusiform vesicles, resulting in increased umbrella increased hydrostatic pressure not only stimulates exo-
cell apical surface area. First, serial sectioning and electron cytosis, but also stimulates rapid endocytosis (30). The
microscopic analysis demonstrates that discoidal/fusiform endocytosed membrane components including UPs are
vesicles are in fact distinct entities that are not connected likely delivered to lysosomes, where they are degraded
to one another or to the cell surface (30). Second, (30), although this has not been formally proven. Whether
discoidal/fusiform vesicles move from a scattered distribu- recycling of internalized membrane is occurring is also
tion in the cell cytoplasm to a position just underneath the unknown. Apparently, the rates of endocytosis and exo-
apical plasma membrane of umbrella cells exposed to cytosis are such that the net effect is to add membrane to
hydrostatic pressure (2,26,30). Third, early studies demon- the apical surface of the cell. These data lead to a refine-
strated that the volume fraction and numerical density of ment of the classical model for vesicle transport to
discoidal/fusiform vesicles are significantly decreased in include an endocytic pathway that operates simulta-
filled vs. voided bladders (37). However, this analysis did neously with the exocytic pathway (Figure 8B) (30).
not take into account changes in cell volume that might
occur as the bladder fills. Stereology and estimates At first glance, it seems counterintuitive that hydrostatic
of apical membrane capacitance (where 1 mF 1 cm2 of pressure would simultaneously induce exocytosis and
surface area) were recently used to demonstrate in endocytosis; however, hydrostatic pressure-induced
isolated tissue that increased hydrostatic pressure stimu- endocytosis would modulate the increase in apical sur-
lates a 50% increase in apical surface area that is coupled face area brought about by exocytosis, and it would

122 Traffic 2004; 5: 117–128

The Uroepithelium

A Classical model
Increase in apical surface
area accompanied by
change in cell shape
of vesicles Filling

Re-establishment of Endocytosis of
vesicle population membrane

B Revised model Figure 8: Models for vesicle

Increase in apical surface dynamics in umbrella cells. (A)
Exocytosis of area coupled with In the classical model bladder filling
vesicles coupled delivery of endocytosed is accompanied by exocytosis of
with endocytosis discoidal/fusiform vesicle, thereby
vesicles to lysosomes increasing the apical surface of the
umbrella cells. Upon voiding, the
added membrane is endocytosed,
thus re-establishing the population
of vesicles. (B) In the revised model
filling stimulates both exocytosis
lysosome and endocytosis. The net rates of
these processes are such that
membrane is added to the apical
Voiding surface. Endocytosed membrane
is delivered to lysosomes where
contents are degraded. Upon
Endocytosis of added voiding the added membrane is
Re-establishment of membrane coupled internalized and reestablishment of
vesicle population with vesicle formation the vesicle pool may result from
in the Golgi both endocytosis and de novo
synthesis along the biosynthetic

ensure turnover of membrane components such as AUM There is limited evidence that mechanical stimuli can
particles. As described earlier, AUM particles may play alter endocytosis in other cell types, including endothe-
important roles in barrier function and plasma membrane lial cells (41,42); however, the underlying machinery
events. Furthermore, endocytosis and exocytosis are that transduces mechanical stimuli to changes in endo-
coupled in other cell types, such as neurons, where cytosis is completely unexplored, and almost nothing is
these two processes maintain the unique composition known in the umbrella cell system. Recent evidence
of the presynaptic membrane (39). Intriguingly, neither indicates that exposing the apical surface of the epithe-
clathrin-coated pits nor caveolae are detected at the lium to hypertonic solutions stimulates apical endocyto-
apical surface of the umbrella cells, indicating that internal- sis (43), but the mechanism is unknown, and endocytosis
ization may be via a nonclathrin-dependent pathway. that accompanies return from hypotonic medium (which
Because so little is understood about these pathways for causes cell swelling) to isotonic medium is blocked in
endocytosis (40), umbrella cells may provide a model sys- cells treated with the actin disrupting agent cytochalasin
tem to define the machinery that drives them and study B (44), indicating a role for actin in umbrella cell apical
how they are regulated. endocytosis.

Traffic 2004; 5: 117–128 123


inner ear (42). How mechanical stimuli are sensed and how
they are transduced into downstream changes in vesicular
traffic is not well understood, although the cytoskeleton,
stretch-activated channels, increased cytoplasmic Ca2þ,
UC integrins, phospholipases, tyrosine kinases, ATP, and
cAMP have been implicated in these events (42).

Secondary messenger cascades and other regulatory
pathways involved in pressure-induced exocytosis
In the umbrella cell, we are only just beginning to under-
stand the mechano-transduction pathways involved in
discoidal/fusiform vesicle exocytosis, although a requirement
for metabolic energy (in the form of ATP) and the actin,
intermediate filament, and microtubule cytoskeleton was
1 µm proposed some years ago (44,47,48). Like other regulated
secretory events, Ca2þ signaling is likely to play an import-
ant role in modulating exocytosis in umbrella cells (35).
Figure 9: Presence of discoidal-like vesicles in intermediate Raising cytoplasmic Ca2þ, by treating the epithelium either
cells of rabbit bladder. Legend: DV, examples of discoidal-like with the Ca2þ ionophore A23187 or with the ER Ca2þ uptake
vesicles; IC, intermediate cell, N, nucleus; UC, umbrella cell.
inhibitor thapsigargin, stimulates exocytosis in the umbrella
cell layer. As expected, pressure-induced exocytosis in
Events following voiding umbrella cells is blocked when tissue is bathed in a nominally
The classical model proposes that, upon voiding, apical Ca2þ-free solution or if cells are treated with inhibitors of
membrane added during filling is rapidly endocytosed inositol 1,4,5-trisphosphate (IP3)-inducible Ca2þ release path-
(Figure 8A). Although highly likely, the current evidence ways from the ER (35). The latter results indicate that both
is scant. Early experiments to demonstrate post-voiding extracellular and intracellular Ca2þ may play a role in this
endocytosis involved instilling fluid-phase endocytic process. One possible upstream signal for Ca2þ release (pur-
markers into bladder just after urination (15,26). How- inergic signaling) is discussed below. The downstream target
ever, filling the bladder increases hydrostatic pressure of Ca2þ is unknown, although possible effectors include
and induces the hydrostatic pressure-stimulated endo- Ca2þ- sensitive isoforms of adenylyl cyclase, as well as phos-
cytosis described above (30). Other evidence comes pholipase Cd, and other channels and pumps (49). Because
from studies in which endocytosis was studied in tissue the epithelium is multilayered, Ca2þ imaging techniques
placed in hypotonic, then isotonic buffers (45); however, will need to be used to confirm that pressure increases
the physiological significance of this experimental cytoplasmic Ca2þ in the umbrella cell layer. An alternative is
manipulation is unclear. Finally, there is evidence that that Ca2þ induces the release of a secretagogue from inter-
short-term application of hydrostatic pressure (for 5 min) mediate/basal cells, which then stimulates umbrella cell
across isolated uroepithelium increases surface area, exocytosis.
and this increase returns to baseline when the pressure
is released, presumably the result of endocytosis (44). An additional secondary messenger that is produced when
The intracellular fate of membrane internalized after the uroepithelium is exposed to hydrostatic pressure is
voiding is an open question. The classical model pro- cAMP (30). Artificially raising cAMP in umbrella cells, by
poses that it serves to reestablish the population of treating tissue in the absence of pressure with forskolin,
discoidal vesicles (Figure 8A) (15,26). However, there causes a dramatic increase in exocytic activity, but not
are few data that support this conclusion. In fact, endo- endocytic activity (30). Unopposed by endocytosis, forsko-
cytosed marker proteins (including fluid-phase and lin induces change in the apical surface area of umbrella
membrane-bound lectins) only label a small fraction of cells of greater than 120%. Forskolin also stimulates exo-
the total discoidal vesicle pool (15,26,46), indicating that cytosis in isolated uroepithelial cells (7). Treatment with
the majority of discoidal vesicles may be formed de H-89, an inhibitor of the downstream cAMP effector
novo along the biosynthetic pathway (Figure 8B). protein kinase A (PKA), partially inhibits (by 50%) hydro-
static pressure-induced changes in the apical surface area
of umbrella cells (30). These data indicate that cAMP,
acting through PKA, modulates hydrostatic pressure-
Regulation of Discoidal Vesicle Exocytosis induced exocytic traffic in umbrella cells. The isoform of
PKA involved and the targets of this kinase are presently
Mechanical stimuli alter exocytic traffic in many cell types, unknown. Again, it is possible that cAMP stimulates secre-
including endothelial cells, myocytes, kidney epithelia, tagogue release from basal/intermediate cells, which then
type II alveolar cells, osteocytes, and the hair cells in the acts upon umbrella cells to induce exocytosis.

124 Traffic 2004; 5: 117–128

The Uroepithelium

Obviously a great deal of work is necessary to understand the activity of the amiloride-sensitive epithelial sodium
the regulation of exocytosis in the uroepithelium. For channel, which is proposed to act as a mechanosensor in
example, the role of specific Rab GTPases in this process this system (57). ATP release may also require expression
is an open question. Rab GTPases are important in regulat- of the transient receptor potential channel, vanilloid
ing tethering/docking of vesicles to their target compart- subfamily member (TRPV1), an ion channel expressed by
ment leading to membrane fusion, and have also been nociceptive (pain sensing) afferent neurons and the uro-
implicated in cargo selection, vesicle budding, and organ- epithelium (62). Isolated bladders from TRPV1 knockout
elle motility (50). Preliminary data indicate that Rab27b is animals fail to release ATP in response to pressure and do
expressed in the bladder (51), and we find that it is not increase umbrella cell apical surface area in response
expressed in the umbrella cell layer (G. Apodaca, D. Barral, to pressure (62). The link between TRPV1 and ATP release
and M. Seabra, unpublished observations). The other iso- is unknown, but could reflect TRPV1 conductance of extra-
form of Rab27, Rab27a, has been implicated in melano- cellular Ca2þ into the cell.
some traffic and exocytosis of lytic granules in cytotoxic
T-cells (52). Rab27a links melanosomes, through melanophi- Intriguingly, addition of apyrase (a membrane impermeant
lin, to the unconventional myosin Va, which interacts with exonucleotidase) to the serosal, but not mucosal, side of
the actin cytoskeleton and serves to retain melanosomes isolated uroepithelial tissue prevents pressure-induced
at the cell periphery (52). As described earlier, discoidal exocytosis in umbrella cells (E. Wang, L. Birder, and
vesicles move from a random position in the cytoplasm to G. Apodaca, unpublished observations), indicating that
below the apical membrane as the bladder fills, and actin extracellular ATP release acts as an upstream signal for
depolymerization inhibits exocytosis (10,30,53). In a man- exocytosis (Figure 10). Furthermore, we observe that
ner analogous to melanocytes, Rab27b may interact with a general inhibitors of P2 receptors inhibit pressure-induced
myosin motor to tether discoidal/fusiform vesicles to the exocytosis, pointing to P2 receptor involvement in this
actin cytoskeleton, ultimately promoting vesicle exocyto- process. P2X2 and P2X3 knockout mice were recently
sis. Localizing Rab27b to discoidal/fusiform vesicles, and generated (63) (D. Cockayne, Roche Palo Alto, unpub-
generation of Rab27b knockout mice would be important lished observations) and pressure-induced exocytosis
first steps in understanding any potential role for this is blocked in bladder tissue taken from these animals
GTPase in vesicle exocytosis. (E. Wang, L. Birder, and G. Apodaca, unpublished observa-
tions), implicating these particular purinergic receptors in
discoidal vesicle exocytosis. ATP-stimulated exocytosis is
blocked by treatments that remove extracellular Ca2þ or
Upstream signals governing discoidal vesicle prevent release of Ca2þ from intracellular stores, indicating
exocytosis: role of ATP that Ca2þ is an important secondary messenger in this
Most cells release ATP when exposed to mechanical sti- process. Any role for P2Y receptors is unknown at present.
muli, and ATP is known to modulate exocytosis in several Again, it is unknown whether ATP directly acts on umbrella
cell systems (54). Mechanisms of release can include ATP cells, or if the effect is indirect.
transporters, exocytosis, hemi-gap junctions, ATP-
conducting ion channels, and cell damage or death (54).
Released ATP can bind to one of two cell-associated pur-
inergic receptors: P2X and P2Y (55). At least six P2Y Sensing Bladder Fullness: Cross-talk Between
receptors (P2Y1, P2Y2, P2Y4, P2Y6, P2Y11, and P2Y12) the Uroepithelium and the Nervous System
and eight P2X receptors (P2X1-7 and P2XM) have been
described (55). P2Y receptors are seven-transmembrane A growing body of evidence indicates that epithelia
domain receptors that upon ATP binding couple through exposed to mechanical stimuli, such as those lining the
heterotrimeric G-proteins to modulate the activity of gut, blood vessels, airways of the lung, and lower urinary
adenylyl cyclases, which generate cAMP, and phos- tract, receive and transmit signals to submucosal neurons
pholipases, leading to generation of diacylglycerol and (64,65). In the case of the bladder, there is evidence that
IP3. In turn, IP3 binds to IP3 receptors that stimulate the uroepithelium may communicate bladder fullness to
increased cytosolic Ca2þ. P2X receptors are similar in the underlying nervous system through a paracrine signal-
structure to the subunits of the epithelial sodium channel ing pathway involving ATP release (63,65,66).
and act as ligand-gated channels that directly stimulate
Ca2þ entry into the cell (55). As described above, Ca2þ Transmission of signals between uroepithelial cells
and cAMP are secondary messengers that stimulate exo- and afferent nerves
cytosis in umbrella cells. Afferent nerve processes (axons) have an intriguing distri-
bution in the bladder. In addition to abutting blood vessels
The uroepithelium releases ATP when exposed to hydro- and surrounding muscle bundles, these axonal processes
static pressure (56–58), and P2X2, P2X3, P2X4, and P2X5 are found within the uroepithelium and in a nerve plexus
receptors have been localized to the serosal surface of just below the basal cell layer (63,67). These nerve pro-
uroepithelial cells (59–61). ATP release may depend on cesses express P2X3 purinergic receptors (60), however

Traffic 2004; 5: 117–128 125


infrequently, have increased bladder capacity, and their

bladders fail to undergo contractions when experimentally
filled. We have noted that the bladders are enlarged in
these animals (E. Wang, L. Birder, and G. Apodaca, unpub-
lished observations). Bladder pressures are normal in P2X3
knockout animals (63). The enlargement of the bladder
may be an adaptation to maintain low pressures, in part
because the umbrella cells of P2X3-deficient animals are
unable to undergo pressure-induced exocytosis. The
above data indicate that ATP, possibly released from the
uroepithelium, binds in a paracrine fashion to P2X3 recep-
tors on afferent nerve processes, signaling bladder full-
ness, and also binds in an ‘autocrine’ fashion to P2X2 and
P2X3 receptors on the umbrella cell to stimulate exocyto-
sis (Figure 10). Thus, ATP has a dual function, allowing for
the epithelium to accommodate more urine and also com-
municating filling to the nervous system.

Communication between the uroepithelium and the

nerves that innervate the bladder is likely to be bidirec-
tional. In addition to releasing neurotransmitters such as
ATP and NO (56–58,70,71), uroepithelial cells express ion
channels and receptors characteristic of sensory neurons.
These include the P2X receptors described above (59–61),
Figure 10: Role of ATP in signaling exocytosis and bladder as well as TRPV1 and the b-adrenoceptor (62,71). Because
fullness. Bladder filling increases hydrostatic pressure,
bladder afferents can release both ATP as well as peptide
stimulating release of ATP from the uroepithelium (steps 1 & 2)
through an unknown mechanism. The released ATP can bind to
neurotransmitters, binding of these neurotransmitters to
P2X3 receptors present on afferent nerve processes (step 3), the uroepithelium could act in a paracrine fashion to stimu-
increasing nerve firing and relaying bladder filling to the central late changes in umbrella cells, including enhanced exocy-
nervous system (CNS) (step 4). Afferents can also release ATP tosis. There is some evidence that neurotransmitters
through a mechanism likely involving exocytosis of synaptic released from efferent nerves can alter barrier function in
vesicle (SV) content (step 5). ATP released from either the the uroepithelium (72), and it was recently observed that
afferents or the uroepithelium can bind to P2X2 and/or P2X3 blocking nerve transmission prevents the acute disruption
receptors, and perhaps P2Y receptors (dashed line in step 6), on
of the uroepithelium observed after spinal cord injury (73).
the umbrella cells (step 6). Ligand binding causes increased
cytoplasmic Ca2þ (a result of Ca2þ influx from outside the cell and
efflux from intracellular stores), which signals discoidal/fusiform
vesicle exocytosis (step 7). ATP may also bind to purinergic
receptors present on basal/intermediate cells (step 8), which Summary and Perspectives
stimulate release of ‘secretagogues’ that act upon umbrella cells
to stimulate vesicle exocytosis (step 9). The composition of urine is markedly different from
plasma, with urine osmolality ranging from 50 to 1200
see (59) for an alternative view, and have recently been Osmol, a pH ranging between 4.5 and 10, and containing
implicated in some forms of nociception (pain sensation) high concentrations of ammonia, urea, as well as other
and warmth perception (63,66,68). It has been proposed toxins. In mammals, the bladder must store this urine for
that ATP released from the uroepithelium during bladder prolonged periods of time without permitting the passage
filling binds to P2X receptors on afferent nerve processes of highly permeable molecules such as ammonia into the
signaling bladder fullness (Figure 10) (63,65). bloodstream. The barrier to ion, solute, and toxin flux is
formed by the uroepithelium, which lines the inner surface
Consistent with this model, the uroepithelium releases of the bladder and must also adapt to large variations
ATP in response to pressure (56–58); however, ATP in pressure as the bladder fills and empties. In addition,
release from other cell types including endothelial cells to providing important information on how barriers
and smooth muscle cells may also contribute to this sig- are formed and maintained, recent analysis of the uro-
naling. Also consistent with the model, blocking P2X epithelium is providing insight into the assembly of protein
receptors significantly decreases nerve firing when blad- particles into specialized membrane domains called pla-
ders are filled in a urinary bladder/pelvic nerve preparation ques. Defects in plaque assembly increase membrane
(69). Even stronger evidence comes from studies in which permeability and may lead to diseases such as vesicour-
bladder function was examined in the P2X3 knockout mice eteral reflux (6,27). In addition, study of the uroepithelium
described above (63). The P2X3-deficient mice urinate is providing clues to how epithelial cells sense mechanical

126 Traffic 2004; 5: 117–128

The Uroepithelium

stimuli such as pressure, and transduce changes in these hingin association, and in vitro formation. Differentiation 1999;65:
stimuli into cellular events such as membrane traffic. 59–69.
13. Hicks M, Ketterer B, Warren R. The ultrastructure and chemistry of the
Increased pressure, for example, stimulates exocytosis
luminal plasma membrane of the mammalian urinary bladder: a struc-
and endocytosis in umbrella cells (30). Exocytosis is modu-
ture with low permeability to water and ions. Philos Trans R Soc Lond B
lated by purinergic signaling cascades, Ca2þ, and cAMP Biol Sci 1974;268:23–38.
(30,35). Nothing is known about regulation of pressure- 14. Simons K, Ehehalt R. Cholesterol, lipid rafts, and disease. J Clin Invest
induced endocytosis, but it may not involve clathrin. Future 2002;110:597–603.
analysis of this pathway may lead to a better understand- 15. Porter K, Kenyon K, Badenhausen S. Specializations of the unit mem-
ing of clathrin-independent pathways of internalization. brane. Protoplasma 1967;63:262–274.
Finally, the uroepithelium interfaces with an underlying 16. Walz T, Häner M, Wu X-R, Henn C, Engel A, Sun T-T, Aebi U. Towards
nervous system, and bidirectional signaling between the molecular architecture of the asymmetric unit membrane of the
these two systems may communicate the degree of blad- mammalian urinary bladder epithelium: a closed ‘twisted ribbon’ struc-
ture. J Mol Biol 1995;248:887–900.
der filling, and may allow the nervous system to modulate
17. Wu X-R, Lin J-H, Walz T, Haner M, YuJ. Aebi, Sun T-T. Mammalian
uroepithelial barrier function.
uroplakins. A group of highly conserved urothelial differentiation-
related membrane proteins. J Biol Chem 1994;269:13716–13724.
Acknowledgments 18. Deng F-M, Liang F-X, Tu L, Resing KA, Hu P, Supino M, Hu C-C, Zhou G,
Ding M, Kreibich G, Sun TT. Uroplakin IIIb, a urothelial differentiation
I thank Dr Lori Birder, Chris Guerriero, Asli Oztan, Raul Rojas, Dr. Debbie marker, dimerizes with uroplakin Ib as an early step of urothelial plaque
Cockayne and Edward Wang for their constructive and helpful comments assembly. J Cell Biol 2002;159:685–694.
during preparation of this manuscript. The transmission and scanning elec- 19. Adachi W, Okubo K, Kinoshita S. Human uroplakin Ib in ocular surface
tron micrographs were prepared by W. Giovanni Ruiz, and the rapid-freeze, epithelium. Invest Ophthalmol Vis Sci 2000;41:2900–2905.
deep-etch micrograph was provided by Dr John Heuser. This work was 20. Min G, Zhou G, Liang F, Sebbel P, Stoffler D, Glockshuber R, Sun TT,
supported by a grant to GA from the NIH (RO1DK54425). Aebi U, Kong XP. Localization of uroplakin Ia, the urothelial receptor for
bacterial adhesin FimH, on the six inner domains of the 16 nm urothelial
plaque particle. J Mol Biol 2002;317:697–706.
References 21. Mulvey MA, Lopez-Boado YS, Wilson CL, Roth R, Parks WC, Heuser J,
Hultgren SJ. Induction and evasion of host defenses by type
1. Martin BF. Cell replacement and differentiation in transitional epithe- 1-piliated uropathogenic Escherichia coli. Science 1998;282: 1494–1497.
lium: a histological and autoradiographic study of the guinea-pig bladder 22. Yu J, Manabe M, Wu X-R, Xu C, Surya B, Sun T-T. Uroplakin I: a 27-kd
and ureter. J Anat 1972;112:433–455. protein associated with the asymmetric unit membrane of mammalian
2. Hicks M. The mammalian urinary bladder: an accommodating organ. urothelium. J Cell Biol 1990;111:1207–1216.
Biol Rev 1975;50:215–246. 23. Truschel ST, Ruiz WG, Shulman T, Pilewski J, Sun T-T, Zeidel ML,
3. Petry G, Amon H. Licht- und Elecktronenmikroskopiche Studien über Apodaca G. Primary uroepithelial cultures: a model system to
Struktur und Dynamik des Übergangsepithels. Z Zellforsch 1966;69: analyze umbrella cell barrier function. J Biol Chem 1999;274:
587–612. 15020–15029.
4. Lewis SA. Everything you wanted to know about the bladder epithe- 24. Wu X-R, Medina JJ, Sun TT. Selective interactions of UPIa and UPIb,
lium but were afraid to ask. Am J Physiol 2000;278:F867–F874. two members of the transmembrane 4 superfamily, with distinct single
5. Negrete HO, Lavelle JP, Berg J, Lewis SA, Zeidel ML. Permeability transmembrane-domained proteins in differentiated urothelial cells.
properties of the intact mammalian bladder epithelium. Am J Physiol J Biol Chem 1995;270:29752–29759.
1996;271:F886–F894. 25. Tu L, Sun TT, Kreibich G. Specific heterodimer formation is a prerequi-
6. Hu P, Meyers S, Liang F-X, Deng F-M, Kachar B, Zeidel M, Sun TT. Role site for uroplakins to exit from the endoplasmic reticulum. Mol Biol Cell
of membrane proteins in permeability barrier function: uroplakin 2002;13:4221–4230.
ablation elevates urothelial permeability. Am J Physiol 2002;283:F 26. Hicks R. The function of the Golgi complex in transitional epithelium.
1200–F1207. J Cell Biol 1966;30:623–644.
7. Deng F-M, Ding M, Lavker RM, Sun TT. Urothelial function reconsid- 27. Hu P, Deng F-M, Liang F-X, Hu C-M, Auerbach A, Shapiro E, Wu X-R,
ered: a role in urinary protein secretion. Proc Natl Acad Sci USA Kachlar B, Sun T-T. Ablation of uroplakin III gene results in small
2001;98:154–159. urothelial plaques, urothelial leakage, and vesicoureteral reflux. J Cell
8. Wang E, Lee J-M, Johnson JP, Kleyman T, Bridges R, Apodaca G. Biol 2000;151:961–971.
Hydrostatic pressure-regulated ion transport in bladder uroepithelium. 28. Dillon MJ, Goonasekara CD. Reflux nephropathy. J Am Soc Nephrol
Am J Physiol 2003;285:F651–F663. 1998;9:2377–2383.
9. Yu J, Manabe M, Sun TT. Identification of an 85–100 kDa glycoprotein 29. Spence HJ, Chen Y-J, Winder SJ. Muscular dystrophies, the cytoske-
as a cell surface marker for an advanced stage of urothelial differenti- leton and cell adhesion. Bioessays 2002;24:542–552.
ation: association with the interplaque (‘hinge’) area. Epithel Cell Biol 30. Truschel ST, Wang E, Ruiz WG, Leung SM, Rojas R, Lavelle J, Zeidel M,
1992;1:4–12. Stoffer D, Apodaca G. Stretch-regulated exocytosis/endocytosis in
10. Hicks RM. The fine structure of the transitional epithelium of rat ureter. bladder umbrella cells. Mol Biol Cell 2002;13:830–846.
J Cell Biol 1965;26:25–48. 31. Kerr DE, Liang F, Bondioli KR, Zhao H, Kreibich G, Wall RJ, Sun TT. The
11. Kachar B, Liang F, Lins U, Ding M, Wu XR, Stoffler D, Aebi U, Sun TT. bladder as a bioreactor: urothelium production and secretion of growth
Three-dimensional analysis of the 16 nm urothelial plaque particle: hormone into urine. Nat Biotechnol 1998;16:75–79.
luminal surface exposure, preferential head-to-head interaction, and 32. Smythe E. Direct interactions between Rab GTPases and cargo. Mol
hinge formation. J Mol Biol 1999;285:595–608. Cell 2002;9:205–211.
12. Liang F, Kachar B, Ding M, Zhai Z, Wu X-R, Sun TT. Urothelial hinge 33. Levin R, Wein A. Response of the in vitro whole bladder (rabbit)
as a highly specialized membrane. detergent-insolubility, uro- preparation to autonomic agonists. J Urol 1982;128:1087–1090.

Traffic 2004; 5: 117–128 127


34. Levin RM, Longhurst PA, Monson FC, Kato K, Wein AJ. Effect of 54. Hamill OP, Martinac B. Molecular basis of mechanotransduction in
bladder outlet obstruction on the morphology, physiology, and pharma- living cells. Physiol Rev 2001;81:685–740.
cology of the bladder. Prostate Suppl 1990;3:9–26. 55. Schwiebert EM. ATP release mechanisms, ATP receptors and purinergic
35. Wang E, Truschel ST, Apodaca G. Analysis of hydrostatic pressure- signalling along the nephron. Clin Exp Pharmacol Physiol 2001;28: 340–350.
induced changes in umbrella cell surface area. Methods 2003;30: 56. Knight GE, Bodin P, de Groat WC, Burnstock G. ATP is released from
207–217. guinea pig ureter epithelium on distention. Am J Physiol 2002;282:
36. Koss LG. The asymmetric unit membranes of the epithelium of the F281–F288.
urinary bladder of the rat: an electron microscopic study of a mechan- 57. Ferguson DR, Kennedy I, Burton TJ. ATP is released from rabbit urinary
ism of epithelial maturation and function. Lab Invest 1969;21:154–168. bladder epithelial cells by hydrostatic pressure changes – a possible
37. Minsky BD, Chlapowski FJ. Morphometric analysis of the translocation sensory mechanism? J Physiol 1997;505:503–511.
of lumenal membrane between cytoplasm and cell surface of transi- 58. Vlaskovska M, Kasakov L, Rong W, Bodin P, Bardini M, Cockayne DA,
tional epithelial cells during the expansion–contraction cycles of mam- Ford PDW, Burnstock G. P2X3 knock-out mice reveal a major
malian urinary bladder. J Cell Biol 1978;77:685–697. sensory role for urothelially released ATP. J Neurosci 2001;21: 5670–5677.
38. Liang FX, Riedel I, Deng FM, Zhou G, Xu C, Wu XR, Kong XP, Moll R, 59. Elneil S, Skepper JN, Kidd EJ, Williamson JG, Ferguson DR. Distribu-
Sun TT. Organization of uroplakin subunits: transmembrane topology, tion of P2X (1) and P2X (3) receptors in the rat and human urinary
pair formation and plaque composition. Biochem J 2001;355:13–18. bladder. Pharmacology 2001;63:120–128.
39. Gundelfinger ED, Kessels MM, Qualmann B. Temporal and spatial 60. Lee HY, Bardini M, Burnstock G. Distribution of P2X receptors in
coordination of exocytosis and endocytosis. Nat Rev Mol Cell Biol the urinary bladder and the ureter of the rat. J Urol 2000;163:
2003;4:127–139. 2002–2007.
40. Johannes L, Lamaze C. Clathrin dependent or not: is it still the ques- 61. Ferguson DR. Urothelial function. BJU Int 1999;84:235–242.
tion? Traffic 2002;3:443–451. 62. Birder LA, Nakamura Y, Kiss S, Nealen ML, Barrick S, Kanai AJ, Wang E,
41. Davies PF, Dewey JCF, Bussolari SR, Gordon EJ, Gimbrone JMA. Ruiz G, De Groat WC, Apodaca G, Watkins S, Caterina MJ. Altered
Influence of hemodynamic forces on vascular endothelial function. In urinary bladder function in mice lacking the vanilloid receptor TRPV1.
vitro studies of shear stress and pinocytosis in bovine aortic cells. J Clin Nature Neurosci 2002;5:856–860.
Invest 1984;73:1121–1129. 63. Cockayne DA, Hamilton SG, Zhu Q-M, Dunn PM, Zhong Y, Novakovic S,
42. Apodaca G. Modulation of membrane traffic by mechanical stimuli. Am Malmberg AB, Cain G, Berson A, Kassotakis L, Hedley L, Lachnit WG,
J Physiol 2002;282:F179–F190. Burnstock G, McMahon SB, Ford APDW. Urinary bladder hyporeflexia
43. Burton TJ, Edwardson JM, Ingham J, Tempest HV, Ferguson DR. and reduced pain-related behaviour in P2X3-deficient mice. Nature
Regulation of Na (þ) channel density at the apical surface of rabbit 2000;407:1011–1015.
urinary bladder epithelium. Eur J Pharmacol 2002;448:215–223. 64. Cooke HJ, Wunderlich J, Christofi FL. ‘The force be with you’: ATP in
44. Lewis S, de Moura J. Incorporation of cytoplasmic vesicles into apical gut mechanosensory transduction. News Physiol Sci 2002;18:43–49.
membrane of mammalian urinary bladder epithelium. Nature 1982;297: 65. Burnstock G. Purine-mediated signalling in pain and visceral percep-
685–688. tion. T Pharm Sci 2001;22:182–188.
45. Chang A, Hammond T, Sun T, Zeidel M. Permeability properties of 66. Cook SP, McCleskey EW. ATP, pain, and a full bladder. Nature
the mammalian bladder apical membrane. Am J Physiol 1994;267: 2000;407:951–952.
C1483–C1492. 67. Gabella G, Davis C. Distribution of afferent axons in the bladder of rats.
46. Amano O, Kataoka S, Yamamoto T. Turnover of asymmetric unit J Neurocytol 1998;27:141–145.
membranes in the transitional epithelial superficial cells of the rat 68. Souslova V, Cesare P, Ding Y, Akopian AN, Stanfa L, Suzuki R,
urinary bladder. Anat Rec 1991;229:9–15. Carpenter K, Dickenson A, Boyce S, Hill R, Nebenuis-Oosthuizen D,
47. Sarikas S, Chlapowski F. Effect of ATP inhibitors on the translocation of Smith AJ, Kidd EJ, Wood JN. Warm-coding deficits and aberrant
luminal membrane between cytoplasm and cell surface of transitional inflammatory pain in mice lacking P2X3 receptors. Nature 2000;407:
epithelial cells during the expansion–contraction cycle of the rat urinary 1015–1017.
bladder. Cell Tissue Res 1986;246:109–117. 69. Namasivayam S, Eardley I, Morrison JFB. Purinergic sensory neuro-
48. Sarikas SN, Chlapowski F. The effect of thioglycolate on intermediate transmission in the urinary bladder: an in vitro study in the rat. BJU Int
filaments and membrane translocation in rat urothelium during the 1999;84:854–860.
expansion–contraction cycle. Cell Tissue Res 1989;258:393–401. 70. Birder L, Apodaca G, De Groat W, Kanai A. Adrenergic- and capsaicin-
49. Berridge MJ, Lipp P, Bootman MD. The versatility and universality of evoked nitric oxide release from urothelium and afferent nerves in
calcium signalling. Nat Rev Mol Cell Biol 2000;1:11–21. urinary bladder. Am J Physiol 1998;275:F226–F229.
50. Zerial M, McBride H. Rab proteins as membrane organizers. Nat Rev 71. Birder LA, Nealen ML, Kiss S, de Groat WC, Caterina MJ, Wang E,
Mol Cell Biol 2001;2:107–117. Apodaca G, Kanai AJ. Beta-adrenoceptor agonists stimulate endothelial
51. Chen Y, Samaraweera P, Sun TT, Kreibich G, Orlow SJ. Rab27b asso- nitric oxide synthase in rat urinary bladder urothelial cells. J Neurosci
ciation with melanosomes: dominant negative mutants disrupt mela- 2002;22:8063–8070.
nosomal movement. J Invest Dermatol 2002;118:933–940. 72. Veranic P, Jezernik K. The response of junctional complexes to
52. Seabra MC, Mules EH, Hume AN. Rab GTPases, intracellular traffic and induced desquamation in mouse bladder urothelium. Biol Cell 2000;
disease. T Mol Med 2002;8:23–30. 92:105–113.
53. Lewis S, de Moura J. Apical membrane area of rabbit urinary bladder 73. Apodaca G, Kiss S, Ruiz WG, Meyers S, Zeidel M, Birder L. Disruption
increases by fusion of intracellular vesicle: an electrophysiological of bladder epithelium barrier function after spinal cord injury. Am
study. J Membr Biol 1984;82:123–136. J Physiol 2003;284:F966–F976.

128 Traffic 2004; 5: 117–128