[CANCER RESEARCH 37, 2843-2853, August 1977]
Malignant Transformation of Urinary Bladder in Humans and in
N-(4-(5-Nitro@2-furyI)-2-thiazolyl]forma
mide-exposed
Fischer Rats: Ultrastructure of the Major
Components of the Permeability
l
Frederick B. Merk, Bendicht U. Pauli, Jerome B. Jacobs, Joseph Alroy, Gilbert H. Friedell, and Ronald S.
Weinstein2
Department of Pathology, Tufts University School of Medicine, Boston, Massachusetts 02111 (F. B. M.J, Department of Pathology, Rush Medical College,
Chicago, Illinois 60612 (B. U. P., J. A. , R. 5. W.J,and Department of Pathology, the St. Vincent Hospital, Worcester, Massachusetts 01604 (J. B. J. , G. H. F.)
Summary
Normal urinary bladder transitional epithelium (umothelium)
contains a permeability barrier to water and solutes. Elec
tron microscopy studies show that three structural elements
of the plasma membranes of superficial urothelial cells, i.e.,
asymmetrical unit membrane plaques and interplaque
membrane at the luminal surface and tight junctions at
lateral surfaces, provide the actual barmier. In spontaneous
transitional cell carcinomas in humans and in similar ap
peaning tumors in Fischer rats fed the carcinogen N-[4-(5nitno-2-furyl)-2-thiazolyl]fonmam ide, ultrastructume of the lu
minal membrane is altered, but the observed changes do
not appear to adequately account for alterations in pemmea
bility. However, modifications in tight junctions are suffi
cient to account for drastic alterations in transepithelial
permeability. In control humans and rats, tight junctions
consist of a network of three to five intramembrane strands.
In Grade I and II human tumors and in early N-[4-(5-nitro-2furyl)-2-thiazolyl]fommamide-induced rat tumors (26 weeks),
tight junctions are focally attenuated to a single strand. In
higher-grade human tumors and in late rat tumors (61
weeks), intramembrane strands are often discontinuous.
Since the number and distribution of intramembrane
strands at tight junctions correlates well with transepithelial
permeability in most epithelia (note that themeare important
exceptions), it is reasonable to conclude that either the
incomplete formation and/on rapid turnover of tight junc
tions is the structural basis of the increased permeability in
transitional cell carcinomas.
Introduction
In normal mammalian urinary bladder, the osmotic gra
dient that is established between urine and plasma by the
kidney remains essentially unchanged as the urine is col
lected and stoned by the urinary bladder before excretion.
Many studies have shown that the transitional epithelium
(urothelium) which lines the urinary tract contains the phys
I
Presented
at
the
National
Bladder
Cancer
Conference,
November
28
to
December 1, 1976, Miami Beach, Fla. Supported by Grants CA-14447 and CA
16377 from the National Cancer Institute and by Grant CA-15945 from the
National Cancer Institute through the National Bladder Cancer Project.
2 Presenter.
AUGUST 1977
ical barrier that is capable of maintaining this gradient
without an appreciable expenditure of metabolic energy (5,
11). Although it is generally agreed that the plasma mem
branes of the superficial urothelial cells form the actual
barmier (14), tracer studies have suggested that the barmier
may be complex, and the different substances may pene
trate the barrier at different locations (4, 5). These pnedic
tions correlate well with ultrastructural studies that show at
least 2 components of the barmier: (a) tight junctions (zonu
lae ocdludentes), which are apically located areas of mem
brane union at the lateral surfaces of superficial cells (3, 35)
and which provide an intercellular barmierto diffusion (1, 21,
27); and (b) the luminal portion of the plasma membrane of
superficial cells (i.e., Iuminal membrane), which provides
an intracellular barrier (11, 13, 14, 18, 28). Luminal mem
brane contains 2 distinct components, unique plaques of
thickened AUM3that are embedded in and surrounded by
interplaque membrane(6, 14, 18, 24, 27, 31). When the AUM
plaques were first discovered, they were believed to endow
the luminal membrane with its permeability characteristics.
However, recent evidence fails to support this point of view
(28).
With malignant transformation, the urothelial barrier to
diffusion is compromised (4, 14). In the current study, we
examined the ultrastructure of the components of the
permeability barrier in spontaneous human transitional cell
carcinomas and in morphologically similar tumors induced
in Fischer rats with the carcinogen, FANFT (2). Our observa
tions support the hypothesis that attenuation of tight junc
tions is the primary cause of the increased epithelial perme
ability that is observed in carcinogenesis (5, 8, 34, 35). Our
comparative ultrastructural studies on spontaneous human
tumors and the FANFT-Fischer tumor model demonstrate
that ultrastructural changes during tumonigenesis in the rat
model closely correspond to those observed in human uro
thelium undergoing malignant transformation.
Materials and Methods
Rat Urinary Bladder. Weanling male Fischer rats (Charles
River Breeding Laboratory, Wilmington, Mass.) were fed
3 The
abbreviations
used
are:
AUM,
asymmetrical
N-[4-(5-nitro-2-furyl)-2-thiazolyljformamide;
croscopy.
unit
membrane;
SEM , scanning
FANFT,
electron
mi
2843
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1977 American Association for Cancer Research.
F. B. Merk et a!.
FANFT obtained from Saber Laboratories, Morton Grove,
Ill. This drug produces a high incidence of umothelial tumors
formaldehyde (17) and fixed for 1 hr.
with
mouse bladder samples for thin-section electron micros
copy were rinsed briefly in 0.1 M cacodylate buffer, pH 7.4;
postfixed with 1% osmium tetroxide [email protected] M cacodylate
buffer, pH 7.4; dehydrated with graded ethanol solutions;
and embedded in Epon 812. Thin sections were prepared
with diamond knives and were double stained with uranyl
acetate and lead citrate.
a clinical
course
similar
to the disease
in man
(2, 25).
FANFT was administered ad libitum as 0.2% of a powdered
Charles River rat chow for 25 weeks. Twenty-six or 61 weeks
after the initiation of FANFT feeding, matswere anesthetized
by i.p. injection of 0.5 to 1.0 ml of Nembutal, 50 mg/mI
(Abbott Laboratories, North Chicago, Ill.). The urinary blad
den was then exposed and immediately covered externally
with a 2% glutaraldehyde fixative in 0.1 M cacodylate buffer,
pH 7.4. Simultaneously, the bladder was inflated through
the urethra with buffered glutaraldehyde at a constant pmes
sure of 50 mm Hg. The urethra was ligated when the bladder
was fully inflated. The bladder was excised and placed in
fixative for an additional 15 mm. It was then opened, and
representative
pieces
of epithelia
were
removed
for study.
Thirty-week-old Fischer rat controls as well as several adult
BALB/c mice, all having been maintained on standard diets,
were anesthetized, and bladder specimens were prepared
as outlined
above.
HumanUrinaryBladder.Studieson humanurinaryblad
den ultrastructure
nature
are particularly difficult because of the
of standard
cystoscopy
procedures
which
must
be
used to obtain human bladder specimens. During cystos
copy, the bladder is filled either with distilled water on a
nonelectrolyte
solution,
such
as glycine.
These
Thin-SectionElectron Microscopy.All human, rat, and
Freeze-FractureElectronMicroscopy.Specimensforthe
freeze-fracture technique were prepared according to
methods described elsewhere (23, 32). In brief, glutaralde
hyde-fixed, 1-cu mm tissue blocks were soaked overnight in
20% glycerol-0.1 M cacodylate. They were then quenched
in Freon 22 cooled with liquid nitrogen. Replicas were pre
pared at —
100°in a Balzems Model BAF 301 freeze-etch
device.
sections
and
freeze-fracture
replica
speci
SEM. For SEM, postfixedspecimenswere rapidlydehy
drated through ascending ethanol and amyl acetate solu
tions and then dried in a Denton DCP-1 critical point drying
apparatus (Denton Vacuum Inc. , Cherry Hill, N. J.). The
tissues were cemented to specimen stubs, coated in a vac
uum
unphysio
All thin
mens were examined by transmission electron microscopy
using either a Philips Model 300 or Philips Model 301 elec
tron microscope.
with
gold,
and
photographed
either
in a JEOL
Model
logical solutions undoubtedly introduce artifacts (30) and,
JSMU-3 or an ETEC Autoscan SEM microscope at an accel
therefore,
orating
must
be carefully
considered
in an interpretation
voltage
of 25 kV.
of ultrastructume (32). Further, cystoscopies are generally
performed
on patients
who
are suspected
of having
a dis
ease in the urogenital system, since urologists are under
standably reluctant to cystoscope asymptomatic patients. In
general, so-called “control―
specimens in this type of study
are from patients with disease in the genitouninary tract but
whose bladder urothelium appears normal by light micro
scopic histopathological criteria. We feel justified in using
cystoscopy specimens as controls in this study, since the
membrane structures under consideration are known to be
stable
and to resist
artifactual
modification.
Human tumors were obtained at cystoscopy from 10 pa
tients with Grade I on II transitional cell carcinoma of the
urinary
bladder.
Controls
consisted
of biopsies
obtained
at
cystoscopy from 4 adult humans. These were from 2 pa
tients with incomplete urethral obstruction and from 2 pa
tients with benign prostatic hyperplasia and relatively mild
urethral obstruction (Fig. 16). The cystoscopy and biopsy
procedures were performed very rapidly using standard
surgical methodology which included filling the bladder
with distilled water. Histopathological evaluation of the bi
opsy specimens by light microscopy showed essentially
normal epithelium. So-called “water
artifact― (30) was
judged to be minimal (Fig. 16). The identity of superficial
cells in the electron microscope was based on location
and the presence of tight junctions, since tight junctions in
normal
margins
unothelium
are present
exclusively
at the lateral
of superficial
cells. In addition
to these biopsies,
2
specimens of human bladder were obtained at autopsy
from postnatal infants who died from unrelated causes and
who had grossly normal appearing urothelium at autopsy
and by light microscopy evaluation. All specimens were
diced into 1- to 2-mm cube blocks in Kamnovsky's para
2844
Results
Luminal Membrane in Normal Rat Bladder. Superficial
cells line the luminal surface of normal Fischer rat bladder.
SEM demonstrates that the luminal faces of individual cells
are polygonal in the plane of the bladder wall. Single seam
like folds are frequently observed at the boundaries be
tween adjacent luminal faces (16) (Fig. 1). These are similar
to those reported by Noack et a!. (24) in expanded Wistamrat
bladder. A characteristic SEM feature of the luminal cell
surface is a complex system of micmomidges(16) (Fig. 2).
In all thin-section studies on control urothelium and tu
moms in humans
and rats,
luminal
membrane
was identified
by its proximity to tight junctions. This is especially impor
tant in studies of human umothelium, since it is possible for
superficial cells to slough during cystoscopy or with proc
essing for electron microscopy. The morphology of thin
sectioned luminal membrane in Fischer matumothelium is
essentially identical to that reported elsewhere for other rat
strains (14, 18). The luminal membrane has concave
plaques that appear as troughs when viewed in profile (Fig.
7). The plaques appear as a tripI@-layered AUM approxi
mately 11 to 12 nm in thickness. The outer leaflet is thicker
than the inner one but is less well defined. The plaques ap
pear to be rigid, but they are joined by flexible “hinges―
of
symmetrical interplaque membrane measuring 8 to 9 nm in
thickness. The combination of rigid and flexible mem
branes results in prominences that apparently correspond
to the micronidges. When micronidges are present, asym
metrical membrane may form the lateral walls, but at the
crest only symmetrical interplaque membrane is observed
(Fig. 8).
The cytoplasm of superficial cells contains many dilated
CANCER RESEARCHVOL. 37
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1977 American Association for Cancer Research.
Ultrastructure of Bladder Permeability Barrier
fusiform vesidles of which some are found in close proxim
ity to Golgi complexes and others are in contact with the
luminal membrane (12, 18). These vesidles are often par
tially lined by AUM plaques with the thick leaflet facing
inward. At the poles of each fusiform vesidle, segments of
unthickened symmetrical membrane link the plaques to
gether. Fusiform vesidles that bean plaques are frequently
encountered in deeper (intermediate) cells, but the plasma
membranes of these cells are symmetrical.
Luminal Membrane in ControlHuman Bladder. The su
perficial cells of the control adult human bladders are
smaller and more variable in size than are those of rat
bladder. SEM of the luminal surface reveals that, although
the texture of the luminal surface is smooth, many irregular
furrows and ridges are present. Microvilli are uniform, and
many are present at the surface, but the number varies from
cell to cell (Fig. 14).
Thin sections of normal bladder urothelium, obtained
from human infants at autopsy, reveal many AUM plaques
measuring 11 to 12 nm thick at the luminal surface (Fig. 17).
There are numerous fusifomm vesidles in the periphery of the
cytoplasm (Fig. 18). On the other hand, Iuminal membranes
of our adult human control bladders are not usually thicker
than
10 to 1 1 nm (Figs.
encountered
19 to 21 ). Although
AUM
plaques
are
in the urothelium of the adult controls (Fig.
21 ), they are very infrequent,
although
umbrella
(superficial)
cells are present and appear fully differentiated in the light
microscope (Fig. 16).
A surface “fuzzy―
coat (i.e., glycocalyx) is often present
on the luminal surface of adult human bladder (22), espe
cially where the membrane profile displays gently rounded
contours (Fig. 19). In areas where the profile is more angu
Ian on the membrane thickness more pronounced, a glyco
calyx is less likely to be observed. Sometimes the luminal
membrane profile in humans appears to have a reversed
asymmetry, the inner leaflet having a greater electron den
sity and thickness than the outer leaflet (Fig. 19). This
appearance may be the result of tonofilaments, closely ap
posed to the inner leaflet and running parallel to the mem
brane face. When transversely sectioned, the cross-sec
tioned filaments are indistinguishable from the inner leaflet
(Fig. 19), but when the plane of section is tangential, tono
filaments are seen (not illustrated).
Vesidles are often present in the cytoplasm of superficial
cells near the luminal surface (Fig. 23). They are usually
quite round but occasionally are fusiform (Fig. 18). We have
been able to demonstrate a clear association between these
vesicles and the Golgi apparatus in human bladder, a find
ing also reported by others (9).
LuminalMembrane in Rat Tumors. Luminalsurfacesof
Fischer rat bladder tumors are found to be composed of
many cell faces that vary greatly in size and shape (Fig. 3).
Micromidges are occasionally encountered, but they are
generally obscured by masses of pleomorphic microvilli (16,
25) (Fig. 4). Between adjacent faces, theme are enlarged
seam-like folds (Fig. 5) that often appear split by an intercel
lular cleft (Fig. 6), thus forming paired ridges. These folds
and ridges overlie the tight junctions at the lateral surfaces
of neighboring umothelial cells.
Thin sections of rat bladder tumors reveal many cell sun
face projections extending into the lumen. Although a small
AUGUST 1977
percentage of the surface area of the luminal membrane
does retain the thickened rigid appearance of typical AUM
plaques (Fig. 12), the luminal surface ofthe most superficial
tumor cells consists exclusively of interplaque membrane
(Figs. 11 and 13). The ultrastructure of luminal membrane
profiles of normal superficial cells differs at straight and
sharply bent regions (Fig. 8). On the other hand, profiles of
tumor luminal membranes that are devoid of AUM appear
the same at gently contoured and acutely angulated me
gions (Fig. 13). Luminal membranes of tumors generally
appear thin and flexible and are often associated with a
conspicuous glycocalyx (Fig. 11).
Lange numbers of vesidles are present in mat umothelial
tumor cells. Some of these vesidles are fusiform, but unlike
the fusiform vesicles of normal rat bladder, they often ap
pear collapsed and are not observed to coalesce with the
luminal membrane (Figs. 9 and 10). Other vesidles have
round profiles consisting of symmetrical membrane (Fig. 9).
Luminal Membrane in Human Tumors. The superficial
cell surfaces of human bladder tumors are shown by SEM to
appear similar to the controls, except that the microvilli are
generally more numerous and somewhat pleomorphic (Fig.
15).@
The ultmastmuctureof thin-sectioned luminal membranes
in tumors
is similar
in every
biopsy
we
have
examined,
regardless of histopathological grade of the tumor. Their
morphology also resembles that of the luminal membranes
usually present in control biopsies (compare Figs. 19 and
22). The luminal membrane of tumors is of uniform thick
ness throughout its length and is usually covered with a
well-formed glycocalyx. We have not observed asymmetni
cal plaques associated with either the luminal membrane or
the surfaces of numerous round vesidles present in the
cytoplasm of human tumor cells.
Freeze-fractured luminal membrane of human transi
tional cell carcinomas resembles the interplaque membrane
in normal
human
urothelium.
It differs
from
the
normal
luminal membrane in that there are many more microvilli.
These are often freeze fractured near their base (Fig. 24).
The luminal membrane in noninvasive tumors can be distin
guished from plasma membrane at the surfaces where there
is cell-cell apposition by the relatively low numerical density
of intramembrane particles on the protoplasmic fracture
face of the luminal membrane.
Tight Junctions in Normal and Malignant Fischer Rat
Bladders.Tightjunctions(zonulaeocdludentes)arecontin
uous
belt-like
regions
of intimate
contact
(union)
between
plasma membranes of adjacent cells. These junctions,
which girdle bladder superficial cells near their luminal
faces, consist of a network of branching intramembrane
fibmils that are visualized by freeze-fracture electron mi
croscopy (21) (Fig. 25). In urothelium of normal Fischer rats,
the network has a width of 3 to 5 strands (Fig. 25). In
FANFT-induced tumors examined 26 weeks after initiation
of FANFT feeding (i.e. , 1 week after FANFT is withdrawn
from the diet), tight junctions are often focally attenuated to
a single
invasive
4 Because
strand
(Fig. 26). We have
transitional
cell carcinomas
of
preparative
artifacts
in
human
examined
obtained
specimens,
replicas
of
61 weeks
a
definitive
state
ment on occurrence of pleomorphic microvilli cannot be made at the present
time.
2845
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1977 American Association for Cancer Research.
F. B. Merk
after
et a!.
initiation
of
FANFT
tumors the tight junctions
feeding
and
are frequently
find
that
in these
discontinuous
(not
illustrated). Our thin-section data on control and tumor
material corroborate the freeze-fracture observations.
Three to 5 focal contacts or “kisses'
‘
(8) are generally pres
ent at tight junctions between normal superficial cells, but
this number is usually reduced to 1 or less in tumors. A
detailed study of junctional modification during the induc
tion of tumors
in Fischer
mats by FANFT
will be the subject
of
a separate paper (B. U. Pauli et a!. , manuscript in prepara
tion).
Tight Junctions in Control and Malignant Urothelia of
Human Bladder. Our resultswith freeze-fractureelectron
microscopy confirm and extend the thin-section obsemva
tions of Fulken et al. (9) and others (34) that tight junctions
are complete
and well developed
in control
human bladder
urothelium. These junctions consist of a network of intma
membranous fibnils (33) (Fig. 27) in freeze-fracture prepara
tions.
Tight
junctions
are complete
but attenuated
in low
grade noninvasive human transitional cell carcinomas (Fig.
28) (34, 35), but are discontinuous in high-grade invasive
human transitional cell carcinomas (34) in which they ap
pear as maculae occludentes (21).
Discussion
studied individually, it can nevertheless be assumed that
each component has, as a minimum, the permeability char
actenistics of the entire intracellular barrier, since both oc
cupy a significant fraction of the membrane surface area.
The possibility remains that plaque regions may be rela
tively impermeable compared to interplaque regions, but
the physiological significance of the differences is probably
trivial (4). Observations on human bladder unothelium, pre
sented in the current study, provide additional evidence that
AUM-plaque membrane does not endow the urothelium
with unique permeability characteristics. In normal human
bladder urothelium obtained from infants, AUM plaques are
prominent and occupy the majority of the urothelial surface
area, as previously described for other species (11, 18, 27).
However, in bladder specimens obtained from adults with
mild on minimal urinary obstruction, and with histopatho
logically “normal―
urothelium, AUM plaques are rarely en
countered.
The vast
majority
of the
luminal
surface
mem
brane consists of intemplaque symmetrical unit membrane.
Although we have not measured urothelial permeability in
these particular patients, it is well known that the permea
bility barrier is intact in comparable patients. Absence of the
AUM plaques in these specimens essentially rules out a
specific permeability barmier function for the plaques (5).
Staehelin et al. (28) have proposed an alternative function
for the plaques. These authors
suggest that AUM plaques
A coherent
the excretory
encloses the
to diffusion.
and continuous sheet of superficial cells lines
system from the renal pelvis to the urethra and
urinary space within a semipermeable barmier
There are 3 major structural elements of the
serve as an attachment site for cytoplasmic filaments (28).
In the transitional cell carcinomas of humans and rats,
luminal surfaces of the superficial tumor cells share a simi
permeability
barrier,
has a symmetrical unit membrane appearance and, on
some cells, a conspicuous “fuzzy―
coat (i.e., glycocalyx).
Also, many surface projections including pleomorphic mi
all of which
are components
of the
plasma membrane of urothelial superficial cells. Two of the
barrier elements, AUM plaques and intemplaque membrane,
are present at the luminal surface of urothelial superficial
cells (6, 11-14, 18, 24, 28, 31). The 3rd component is zonu
lae occludentes
(‘
‘tight―),intercellular
structurally
specialized
zones located
junctions
that are
at the apical-lateral
surfaces of these superficial cells. They form a continuous
belt-like region completely encircling the cells (3, 21, 26,
27). The luminal membrane elements retard the intracellular
(transcellular)
diffusion
of water
across
the epithelium,
whereas the tight junctions provide a barrier to intercellular
(pamacellular or bypass) diffusion of water across the blad
den wall (1). Tracer studies have suggested that the barrier
may be complex and that the different substances may
penetrate the barrier at different locations; for example,
water may pass through the intracellular barrier, and so
dium may pass through the intercellular barrier (5).
It has been widely speculated that the AUM plaques may
endow the intracellular barrier with special permeability
properties. Early support for this concept came from thin
section and negative-stain electron microscopy observa
tions
that
were
interpreted
as showing
that
the AUM
plaques covered most or all of the luminal surface of super
ficial cells (13). The results of recent freeze-fracture studies
challenge this concept. Staehelin et a!. (28) have shown that
in species in which AUM plaques are particularly
prominent,
the plaques occupy less than 80% of the luminal surface with
interplaque membrane accounting for the remainder of the
surface area. Although the permeability characteristics of
the AUM
2846
plaques
and intemplaque
membrane
have not been
lam ultrastructure.
cnovilli
extend
In each species,
into the bladder
the surface
lumen.
We have
membrane
no data on
the permeability characteristics of the luminal membrane of
the superficial tumor cells. However, it is noteworthy that
there is no evidence in the literature that membrane perme
ability per unit area to either water or small ions is altered by
changes in the contour of membranes, by the formation of
pleomorphic microvilli, or by the overdevelopment of the
cell surface glycocalyx. Therefore, alterations in the ultra
structure of the tumor luminal membrane cannot be related
to alterations in transurothelial permeability at the present
time. On the other hand, there is strong evidence that al
terations in zonulae occludentes junctions can account for
the altered permeability in transitional cell carcinomas (4)
as has previously been reported for a number of other types
of tumors (19, 20, 29, 34, 35).
In normal epithelium, extracellular bypass diffusion of
small molecules is blocked by zonulae ocdludentes cell-to
cell junctions (tight junctions) (7, 10), and it is generally
held that functional differences between various tissues
stem in part from structural differences in these junctions
(1, 7, 8). Claude and Goodenough (1) have classified epithe
ha ranging
from
‘
‘veryleaky― to “verytight'
‘
on the basis
of
measurements of electrical resistances across the epithe
hum. They have correlated the number of interconnected
strands located within the membranes at these intercellular
junctions with transepithelial resistance to electric current,
a measure of bypass diffusion. In our study, we have shown
that in both human and rat urinary bladder, tight junctions
CANCER RESEARCH VOL. 37
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1977 American Association for Cancer Research.
Ultrastructure of Bladder Permeability Barrier
generally have 3 to 5 intramembrane strands (Figs. 25 and
27). According to the criteria of Claude and Goodenough
(1), it might be expected that these junctions could account
for relatively tight seals between cells.
The tight junctions between superficial tumor cells are
remarkably similar in moderately low-grade (Grade I and II)
human transitional cell carcinomas, as reported previously
(34, 35), and in the FANFT-induced
transitional
cell camcino
mas harvested at 26 weeks. In tumors of both species, the
tight junctions are focally attenuated and often have only a
solitary strand (Figs. 26 and 28). Attenuation of the tight
junctions to a single strand can account for an increase in
epithelial permeability (1). In higher-grade human tumors
and in invasive FANFT-induced tumors harvested at 61
weeks, the intramembrane fibrillan network of the tight
junctions is frequently discontinuous (i.e. , they become
maculae or fasciae ocdludentes). This may further increase
permeability. Similar observations have been made in intra
cranial germinomas (29). On the basis of our findings and
the results of other studies (for reviews see Refs. 21 and 27),
it is reasonable
to conclude
that transepithelial
leakiness
in
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21 .
Intercellular Junctions in the Mouse Liver. J. Cell Biol., 45: 272-283,
1970.
Hicks, R. M. The Permeability of Rat Transitional Epithelium. J. Cell
Biol., 28: 21-31, 1966.
Hicks, A. M. The Function of the Golgi Complex in Transitional Epithe
hum. J. Cell Biol., 30: 623-643, 1966.
Hicks, R. M., and Ketterer, B. Isolation of the Plasma Membrane of the
Luminal Surface of Rat Bladder Epithelium and the Occurrence of a
Hexagonal Lattice of Subunits Both in Negatively Stained Whole Mounts
and in Sectioned Membranes. J. Cell Biol., 45: 542-553, 1970.
Hicks, R. M., Ketterer, B., and Warren, R. C. The Ultrastructure and
Chemistry of the Luminal Plasma Membrane of the Mammalian Urinary
Bladder: A Structure with Low Permeability to Water and Ions. Phil.
Trans. Roy. Soc. London Part B., 268: 23-38. 1974.
Hoyes, A. 0., Ramus, N. I., and Martin, B. G. H. Fine Structure of the
Epithelium of Human Fetal Bladder. J. Anat., 111: 415-425, 1972.
Jacobs, J. B., Arai, M., Cohen, S. M., and Friedell, G. H. Early Lesions in
Experimental Bladder Cancer: Scanning Electron Microscopy of Cell
Surface Markers. Cancer Res., 36: 2512-2517, 1976.
Karnovsky, M. J. A Formaldehyde-Glutaraldehyde Fixative of High Os
molality for Use in Electron Microscopy (Abstract). J. Cell Biol., 27: 137,
1965.
Koss, L. G. The Asymmetric Unit Membranes of the Epithelium of the
Urinary Bladder of the Rat. Lab. Invest., 21: 154-158, 1969.
Martinez-Palomo, A. Ultrastructural Modifications of Intercellular Junc
tions in Some Epithelial Tumors. Lab. Invest., 22: 605—614,1970.
Martinez-Palomo, A. Intercellular Junctions in Normal and Malignant
Cells. Pathobiol. Ann. 1: 261-279. 1971.
McNutt, N. S. , and Weinstein, R. S. Membrane Ultrastructure at Mamma
han lntercellularJunctions. Progr. Biophys. Mol. Biol., 26: 47-101 , 1973.
Monis, B., and Dorfman, H. D. Some Histochemical Observations on
Transitional Epithelium of Man. J. Histochem. Cytochem., 15: 475—481,
1967.
Moor, H., and MUhlethaler, K. Fine Structure in Frozen-etched Yeast
bladder tumors can be explained on the basis of tight junc
tional attenuation.
22.
Acknowledgments
23.
We thank Dr. Samuel M. Cohen for providing us with FANFT-induced
bladder tumors and for his advice; Marjorie DeVeau, Stephen Halpern, and
William Leonard for expert technical assistance; and Lela Silverstein for
assistance in preparing the manuscript.
24. Noack, W., Jacobson, M., Schweichel, J. U., and Jayyousi, A. The
Superficial Cells of the Transitional Epithelium in the Expanded and
Unexpanded Rat Urinary Bladder. Acta Anat., 93: 171-183, 1975.
25. Psi, S. H., Amaral, L., Werthamer, S., and zak, F. G. Ultrastructure and
Reversibility of Bladder Carcinoma of Rats Produced by Feeding of N-[4(5-Nitro-2-furyl)-2-thiazolyl]formamide. Invest. Urol., 11: 125—135,1973.
26. Staehelin, L. A. Further Observations on the Fine Structure of Freeze
Cells.J. Cell Biol., 17: 609-628,1963.
References
1. Claude, P., and Goodenough, D. Fracture Faces of Zonulae Occludentes
from “Tight―
and “Leaky―
Epithelia. J. Cell Biol., 58: 390—400,1973.
2. Ertürk,
E., Cohen, S. M., Price, J. M., and Bryan,G. T. Pathogenesis,
Histology and Transplantability of Urinary Bladder Carcinomas Induced
in Albino Rats by Oral Administration of N-[4-(5-Nitro-2-furyl)-2-thiazo
lyljformamide.CancerRes.,29: 2219-2229,1969.
3. Farquhar, M. G., and Palade, G. E. Junctional Complexes in Various
Epithelia. J. Cell Biol., 17: 375—412,1963.
4. Fellows, G. J., and Marshall, D. H. The Permeability of Human Bladder
Epitheliumto Waterand Sodium. Invest.Urol.,9: 339—344,
1972.
5. Fellows, G. J., and Turnbull, G. J. The Permeabilityof Mammalian
Urinary Bladder Epithelium. Rev. European Etudes dIm. Biol., 16: 303-
310, 1971.
6. Firth, J. A., and Hicks, R. M. Interspecies variation in the Fine Structure
and Enzyme Cytochemistry of Mammalian Transitional Epithelium. J.
Anat.,116:31-43,1973.
7. Friend, D. S., and Gilula, N. B. Variations in Tight and Gap Junctions in
Mammalian Tissues. J. Cell Biol. , 53: 758-776, 1972.
8. Frômter,E., and Diamond,J. Routeof PassiveIon Permeationin Epithe
ha. Nature New Biol., 235: 9-13, 1972.
9. Fulker,M.J., Cooper,E. H., andTanaka,T. Proliferationand Ultrastruc
ture of Papillary Transitional Cell Carcinoma of the Human Bladder.
Cancer, 27: 71-82, 1971.
10. Goodenough, D. A., and Revel, J. P. A Fine Structural Analysis of
CleavedTight Junctions.J. Cell Sci., 13: 763-786,1973.
27. Staehelin, L. A. Structure and Function of lntercellularJunctions. Intern.
Rev. Cytol., 191-283, 1974.
28. Staehelin, L. A., Chlapowski, F. J., and Bonneville, M. A. Luminal
Plasma Membrane of the Urinary Bladder. I. Three-dimensional Recon
struction from Freeze-etch Images. J. Cell Biol., 53: 73-91 , 1972.
29. Tani, E., Ikeda, K., Kudo, S., Yamogata,
S., Nishiura,
M., and Higashi, N.
Specialized Intercellular Junctions in Human Intracranial Germinomas.
Acta Neuropathol.,27: 139-151,1974.
30. Tannenbaum, M. Improper Pathologic Assessment of Bladder Tumor.
FilteredWaterArtifact. Urology,6: 627-630,1975.
31 . Turnbull, G. J. Ultrastructural Basis of the Permeability Barrier in Urothe
hum. Invest. Urol., 11: 198-204, 1973.
32. Weinstein, R. S. Changes in Plasma Membrane Structure Associated
with Malignant Transformation in Human Urinary Bladder Epithelium.
Cancer Res., 36: 2518-2524, 1976.
33. Weinstein,
R. S., McNutt, N. 5., Nielsen, J. L. , and Pinn, v. W. Intramem
branous Fibrils at Tight Junctions. Proc. Electron Microsc. Soc. Am.,
28: 108-109, 1970.
34. Weinstein,
R. S. , Merk, F. B. , and Alroy, J. The Structure
and Function
of
Intercellular Junctions in Cancer. Advan. Cancer Res., 23: 23-89,
1976.
35. Weinstein, R. 5. , Zel, G., and Merk, F. B. Quantitation of Occludens,
AdherensandNexusCellJunctionsin HumanTumors.In: J. Schultzand
R. E. Block (eds.), Membrane Tranformations in Neoplasia. Miami Winter
Symposia, Vol. 8, pp. 127-150. New York: Academic Press, Inc., 1974.
AUGUST 1977
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1977 American Association for Cancer Research.
2847
F. B. Merk et a!.
Fig. 1. Scanning electron micrograph of the luminal surface of a distended urinary bladder from a normal Fischer rat. The polygonal faces of superficial
cellsarecoveredwith irregularfolds. Narrow,seam-likeelevations(arrow)arefrequentlyobservedat boundariesbetweenthe superficialcellfaces.x 2,500.
Fig. 2. Luminal surface of a distended normal Fischer rat bladder at high magnification. The honeycombed network of microridges covering a superficial
cell is demonstrated. x 11,000.
Fig. 3. Scanning electron micrograph of the surface of a Fischer rat FANFT-induced bladder tumor. The superficial cell faces are irregular in size and in
some areas are separated by seam-like elevations. Long slender microvilli (arrows) are prominent on faces of underlying cells recently exposed by sloughing
superficial cells. x 3,500.
Fig. 4. Surface of FANFT-induced Fischer rat bladder tumor. Pleomorphic microvilli cover the luminal membranes of several superficial cells. Microridges
are rare. x 5,500.
Fig. 5. Detail of seam-like elevation between bladder tumor cells of a Fischer rat. In tumors the seams usually appear wider than in bladders of control
animals. x 7,500.
Fig. 6. Split seam in Fischer rat bladder tumor. The seams are often partially or entirely split along the midline by a cleft. x 9,000.
Fig. 7. Thin section of normal Fischer rat bladder urothelium. Two superficial cells are joined together by a junctional complex (upper left). The profile of
the luminal surface has a scalloped appearance and includes asymmetrical membrane plaques. Cytoplasmic vesicles, some of which are fusiform in shape,
frequently appear in superficial cells. x 65,000.
Fig. 8. Sagittal section of a microridge in normal mouse bladder. Rigid asymmetrical plaque membrane (12-nm thick) is located at the lateral walls of the
ridge. At the crest of the ridge only flexible symmetrical membrane (8-nm thick) is found. Arrows, regions where the plaques join interplaque membrane. X
102,000.
Fig. 9. Superficial cell of a FANFT-induced Fischer rat bladder tumor. Microvilli are prominent, and the contour of the luminal membrane profile is
generally less angular than that found in controls. Collapsed fusiform vesicles are prominent in the cytoplasm. x 36,000.
Fig. 10. Detail of collapsed fusiform vesicles in a FANFT-induced Fischer rat tumor. Vesicle walls contain AUM plaques. x 100,000.
Fig. 11. Detail of a Fischer rat bladder tumor. The luminal membrane (8-nm thick) is symmetrical throughout its length and is covered with a delicate
glycocalyx. x 115,000.
Fig. 12. Detail of Fischer rat bladder tumor. The luminal membrane has a scalloped appearance and includes asymmetrical membrane plaques. x 115,000.
Fig. 13. Sagittal section of a microvillus from a FANFT-induced Fischer rat bladder tumor. The luminal membrane, approximately 9 nm thick, is uniform in
appearance throughout its length. x 102,000.
Fig. 14. Scanning electron micrograph ofthe luminal surface of a control human bladder. Many ridges and furrows are present on the surfaces of the cells,
but the surface appears smooth in areas not covered with microvilli. Numbers of microvilli vary greatly from cell to cell. x 3,250.
Fig. 15. Scanning electron micrograph of human bladder tumor luminal cells. Although smooth areas at the surface are visible, most regions of the cell
surfaces are covered with microvilli. The microvilli are usually more numerous in human tumors than in controls. Compare with Fig. 14. x 5,000.
Fig. 16. Light micrograph of Epon-embedded, toluidine blue-stained, typical control urothelium. This is a cystoscopy specimen from a patient with
bladderneck constriction.Theepitheliumis contractedand slightly dysplastic.“Water-artifact―
(30)is minimal. x 345.
Fig. 17. Luminal membrane of postnatal human bladder obtained at autopsy. Asymmetrical membrane plaques have an appearance similar to those found
in normal rat urothelium. x 115,000.
Fig. 18. Fusiform vesicles (FV) in superficial cell of postnatal human bladder. The vesicles, located near the luminal membrane (LM), bear AUM with the
thick leaflet facing inward. x 115,000.
Fig. 19. Superficial cell of control bladder from a patient with benign prostatic hyperplasia. The luminal membrane is approximately 9 nm in thickness and
is covered at the free surface with a delicate glycocalyx. At some regions, the inner leaflet of the membrane appears thicker and more electron dense than
does the outer leaflet, but plaques are not observed. x 175,000.
Fig. 20. Superficial cell of control bladder from a patient with bladder neck constriction. The symmetrical luminal membrane is about 10 nm in thickness. X
185,000.
Fig. 21. Superficial cell of control bladder from a patient with benign prostatic hyperplasia and very mild obstruction. Transversely sectioned luminal
membrane (between arrowheads) appears to be an asymmetrical plaque; the inner leaflet has greater density than does the outer leaflet and in some regions
is thicker as well. x 160,000.
Fig. 22. Superficial cell of patient with noninvasive transitional cell carcinoma (Grade I to II). The luminal membrane is covered with a glycocalyx and in
some areas shows reversed asymmetry (compare with control, Fig. 19). Thin sections of all the human bladder tumors we examined have luminal membranes
with a similar appearance. x 185,000.
Fig. 23. Luminal surface of adult human bladder obtained from a patient with incomplete urethral obstruction. A cytoplasmic “seam―
projects into the
lumen (upper right) where a tight junction (TJ) joins 2 superficial cells. Tight junctions may provide support for the prominences which frequently appear in
the SEM as seam-likefolds betweenluminal facesof superficial cells. A heterogeneouspopulationof vesiclesis presentin the cytoplasmof the cells.
x 42,000.
Fig. 24. Freeze-fracture replica of a superficial tumor cell in a patient with transitional cell carcinoma (Grade I). The luminal membrane displays numerous
cross-fractured pleomorphic microvilli. No asymmetrical membrane plaques are observed. Intramembrane particles are more numerous on the lateral surface
(from Ref. 35). x 68,000.
Fig. 25. Freeze-fractured tight junction in a normal Fischer rat bladder. The junctions consist of an interconnected fibrillar network that is 3 to 5 strands in
depth. x 96,000.
Fig. 26. Tight junction in a FANFT-induced bladder tumor of the rat. At the right, the full complement of junctional strands is present. Elsewhere the
network is narrowed and is focally attenuated to a single strand (between arrows). x 96,000.
Fig. 27. Freeze-fractured tight junction in normal human bladder. The tight junction is 3 to 5 strands deep. X 96,000.
Fig. 28. Freeze-fractured tight junction of patient with transitional cell carcinoma (Grade I). The junction is focally attenuated to a single strand (between
arrows). x 96,000.
2848
CANCER RESEARCH VOL. 37
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1977 American Association for Cancer Research.
@
4@
@
.@
Ultrastructure of Bladder Permeability Barrier
,pw@
@. ,
-—@
I ..-@
@
@d ‘
@
@.
—‘
@y'@-@
.
@- .
@.. .‘
.
@:‘ @‘.
,@,
@
,.‘:-
. . . ,.-
@
@
@
..,,@
‘5.
..
‘..‘
.@
.‘:
..
i,;'..
a
t_@aa@
@@f#'@:―
,@
@ cD
‘wI,r―
•
.,
4:H.Tza;?
@@t;;@
@-
@
@:. -
,,
@
.@-
i:@
‘-\
@‘-@
.
‘ . :*‘r.@f:.-1,
-
-:@@@:
•
y
.. .
@.
.
,‘@@“vI'
@
.:‘@
@::@
°
-‘
k:@,
-:
.@
@_:
@
fr@ :
@
‘
./‘
S
@
V
@
@
‘@.
.@/
@12r-@;;‘:. ‘. .
.
AUGUST
1977
—
.@.
.@
.
@-T
@@:-
2849
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1977 American Association for Cancer Research.
@
@
@I
N@:
F. B. Merk et a!.
...
@
.@-,-
@
-4.:@:.
:,.
,‘
@;
‘-.-.
@
.
.
@
@
.;.@
.(@.
@%
,::@
.,q,
@
@
.
‘..
1?'\
-z'
“>
‘ @jT
‘-‘-a.
\
•““@.
.
@° ...
‘;. . .
@‘::
-.
£
@
A1@
1z1\
. , 4. -
@
...
@
@
@
@
@
@
.
‘ ‘
.
.
@.
.
‘
f@
.@
2850
p ‘•
@-.,
.
‘
,0,@
.@
,
.•.@
i
S
,‘
‘0
‘@
,
.
.tdt@?
5@
...
,‘@.,‘.,
CANCER RESEARCH VOL. 37
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1977 American Association for Cancer Research.
@
@
@
@P@LJ@:
‘
4 @:‘@
@.
@.
:
-
@@:: ‘I
@;
Ultrastructure of Bladder Permeability Barrier
@
j
1@.
@â€S̃
@
...
@
,.-
@
@
@‘
.,.
r
@
:
@,.
4,-...
-
.•
“_‘
.. ,@
@:“-‘@-@L@-
..
. ‘
.@
.@ :
•, {@
:â€.̃
0,.@
@.-
:
,_@
@
@
@
@
@
@
@
@..
.
@-
-.@:‘,
F
L.@
::---‘
.
Us.@,y,
-
:‘@,
:
—,_,#.,p,;@
@
@
@:1' ,
dj@
‘- -
@@z;―
,@.‘
@,:‘- @it
@i'@
:@
‘U
_
.
#
-r
‘@
..
‘.@a@―
—
‘ , -
-
,@
“- •
:‘-;
@@@-‘;@\:-‘ ,,‘,@‘
“Pa
,.
,-
.
@.0
@. 0
.•-
@
@
A,
•4.4@
<@:
II@4-@
‘@
:@L
@
@
@
.-
‘,@‘
“I_f'
, 4L*t.:'@.
..
p
@,-tg―.'
-
Jr'.
•A1@' .
4.
—4.t―@
@‘
,,‘-@ _,..‘_..‘
-,::@_â€ẫ€¢j.@
-@@.
-.
—
:
*
. @.
.
%
@
-
@
.,. ,@. . .@z'
AUGUST
1977
2851
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1977 American Association for Cancer Research.
@
,r@c@v
@
@
“@
@j
.
:
:
.@
‘
F. B. Merk et a!.
?-
@
@v
“5
@
/
@
-.3+:@
@
£•@
—
-C-‘@
(@l
@
@
-
.‘
@.-
@
_.
1'@@‘:@
@j.*.‘@ @:
@v
tt@—
p'@s@._
@-
.
@-‘@
@..t.
@\f
..•
.•-.
@.
-@
@.T;:t@
@
‘@r'@iV.@ .
..;@
@
-
@
h@: . -
@
-
@
@ai..@4• :
-
.
..@. .
. .4..
‘- ‘.‘
@..
.
.
,
@urf@c@
.0'
2852
5,@
.
-
.-@
U,'.
CANCER RESEARCHVOL. 37
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1977 American Association for Cancer Research.
@
@
w
@
@
s@
;
.-
@-
‘@‘.
:\_@.@
“
\
, -@
Ultrastructure of Bladder Permeability Barrier
@
@
.;.. -t@_
-
- .
-
.
::..
- :-:
-
-.
-
.@;‘ ‘
-.
@
@
@;@Z―@
-- . ,
@
@
.@
--@‘-/.
,/@@•;i
,1@
‘--
@‘
-
.-@
-.
.
;
:.
,
@:
,—.@
:
‘
.- I
-
:@-‘@
“.4,
.,.
.(@:@E;(:i;;@
‘: @:f:;@J'
@:
‘-U
U
‘
‘5
4...-
@tT
:‘‘-
@
,—.
@.
@
v
@
—
@
-
@
@
@
@
@.“@:e:.―@A.
..
‘,_
-.
-
. ..
@
.-
..
@-@: - -
0 @.: - -
•.@.
‘@.
.@-‘!
@-‘
@:‘
@
@
.‘.,@
.
.
-â€-̃
.
.
@.
.@
-
.
. ..-‘@-r'.
, ,
. -.
.U ,- ,
@v. .‘J@
.•
• U .::.@‘
.:
‘7-.-
‘ -‘
—
:
-
•..
@.-.
-@ - .
•-
.
-
@.-
-
@
:
.-.
-@.\
.@ .“.-, .
,v5W*@@
-@
‘.@
.
U
@
@
@
‘-U--.-.%.-
.
:
@.
,.@‘..
,
.
.
•
..
.
.
S
@“
..
..
.
.
“@-
,
..
.@
-
,-
..,
d
,.@
-,
-
,
,,@
.
@
ta-_s
. ‘ . .
@
@- 4
—ft:-—7,--@,
.
I. •
1r:@
4;,-@
-.
•
.
@! -‘@
‘U@@'
-;
•@-@
@‘
—.—•.‘
1@;a@:
“-U.
.
--@:.--@-
—
,
@
@‘.
-
- . -
.-.“‘•,“
..,.%
“@. .@•U-
:-
. ..
“..
-.;
.‘
@
@
‘.
.
.
7_@_'
.
-@ . -., . ..:N
•,-.
—
.
@
•
.
. - ... ‘
-
-
-..-
‘ ‘. . :“
Lun@e1'@
@
_“@,.,
.
. .
@.. ‘ -
@-
-•.
.%
@-—U,--.
@%_•‘U―'―'\
•@
-
•.
: ..- @—
. -‘@
- 4@@@@::'
-@“
- - - . -
:-‘.
@@:rt!@
@
@
,
.-
“. —-
-
,@.,)‘
.
I
U-――
ItJ@I@
.“
@—.--•@
-@
.: ..:@‘:@-‘
.@
-
@:
U--
@,
U'
-
-
...
.
. ,
.
-.U.
U—.---
—@
i@
.
.
•@
@,-
.
.
,
•.
-
I,'
-
@.I'
AUGUST 1977
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1977 American Association for Cancer Research.
2853
Malignant Transformation of Urinary Bladder in Humans and in
N-[4-(5-Nitro-2-furyl)-2-thiazolyl]formamide-exposed Fischer
Rats: Ultrastructure of the Major Components of the
Permeability Barrier
Frederick B. Merk, Bendicht U. Pauli, Jerome B. Jacobs, et al.
Cancer Res 1977;37:2843-2853.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/37/8_Part_2/2843
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1977 American Association for Cancer Research.