| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Departments of Urology (W.Y.C., L.B., G.S.P.) and Physiology and Biophysics (G.S.P.), University of Illinois College of Medicine, Chicago, Illinois 60612; and VA Medical Center and Departments of Laboratory Medicine, Pathology and Surgical Oncology (M.J.W.), University of Minnesota, Minneapolis, Minnesota 55417
Address all correspondence and requests for reprints to: Gail S. Prins, Ph.D., Department of Urology, M/C 955, 820 South Wood Street, Chicago, Illinois 60612. E-mail: gprins{at}uic.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-actin, fibronectin, and vimentin identified this
multicellular layer of periductal cells as differentiated fibroblasts.
Peripheral to these fibroblasts, actin-positive smooth muscle formed a
second layer of periductal stromal cells. PCNA labeling showed that
estrogen exposure increased the fibroblast proliferation. Because many
periductal fibroblasts were positive for estrogen receptor (ER)
in estrogenized rats, a direct effect of estradiol on their
proliferation is suggested. Gelatinolytic gels revealed that estrogen
exposure did not alter the activity of matrix metalloproteinases
associated with tissue remodeling during prostate morphogenesis.
However, the periductal fibroblast layer in estrogenized prostates was
devoid of urokinase- and tissue-plasminogen activator, which may
potentially alter the localized proteolysis involved in matrix
remodeling. It is proposed that proliferation of a multicellular layer
of periductal fibroblasts in estrogenized prostates results in a
physical barrier that constrains branching morphogenesis and blocks
paracrine communications between smooth muscle and epithelial cells
which normally regulate differentiation. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Mesenchymal-epithelial interactions play an important role in prostate
development and are believed to reciprocally mediate androgen-induced
growth and differentiation of the gland (12). The initial development
of the prostate is regulated by androgens acting through AR in
mesenchymal cells that direct proliferation and differentiation of
adjacent epithelial cells through diffusible mediators (13). Recent
observations also suggest that the prostate epithelium directs smooth
muscle differentiation and orientation along the length of the
developing ducts and that the initiation and maintenance of the smooth
muscle phenotype is androgen dependent (12, 14). Rats that are
neonatally estrogenized exhibit a rapid loss of epithelial and stromal
AR (10); thus, one might predict that this could disrupt essential
androgen-driven morphogenetic signals between these two cell types.
Additional evidence that stromal-epithelial communications may be
mechanistically involved in developmental estrogenization comes from
our findings that ER in the mesenchymal cells are up-regulated by
neonatal estrogens, which allows for amplification of estrogenic
signals in periductal stromal cells at the time of neonatal exposure
(15).
In the rat prostate, the periductal stroma is composed predominantly of smooth muscle cells that form a sheath 2–4 cell layers thick outside the basement membrane (16). We have identified the smooth muscle as the predominant AR+ stromal cell type in the rat ventral prostate (17). Because these cells are in close proximity to the epithelial ducts, it has been proposed that smooth muscle is the critical mediator of the androgenic signal for epithelial growth and differentiation (10, 18). The other major stromal cell type in the rat prostate is fibroblasts which populate the interductal spaces and are approximately 30% AR+ in the adult ventral lobe (16, 17). In addition, an extremely thin single-cell layer of fibroblasts as narrow as 300 Å is interspersed between the basement membrane (BM) and smooth muscle cells and has generally been attributed with a structural support for the epithelial and muscular components of the rat prostate (16). The location of these fibroblasts immediately outside the basement membrane also positions them strategically to affect passage of diffusible factors from the smooth muscle cells as well as modulate the composition of the extracellular matrix (ECM).
In addition to a structural role, the ECM and related molecules of many branched structures are recognized as key regulatory components during developmental morphogenesis and cell differentiation (19, 20, 21). For example, collagen deposition determines cleft formation during branching morphogenesis in the salivary gland (20). Specific ECM molecules can also bind and sequester growth factors and have been postulated to form cross-links between epithelial and stromal cells and, in effect, couple cell signaling systems (22). Two major components of the ECM are collagens and substrate adhesion molecules, including laminin and fibronectin. These molecules transmit signals through interactions between their binding site, such as an RDG sequence, and specific cellular integrin receptors which transduce signals intracellularly (23). During prostate development, extensive branching morphogenesis during the first 15 days of life is facilitated by a highly dynamic ECM compartment that is actively remodeled through controlled localized proteolysis (24). Proteases implicated in this developmental process include plasminogen activators (PA) and specific metalloproteinases such as MMP-2, a type IV collagenase (25, 26). Thus, changes in the ECM composition through altered production of these molecules, alterations in the cell types that produce these components or their modification by proteases, could significantly impact on branching patterns as well as growth and differentiation of the prostate. Support that neonatal estrogens may be acting through these pathways comes from studies that have shown that estrogens can alter prostatic collagen and fibronectin deposition (27, 28, 29) and stromal cell cytology (30, 31, 32).
The overall goal of the present study was to determine if neonatal
estrogenization of the prostate is mediated in part through alterations
in the composition of either the ECM and its related molecules, or in
the stromal cells adjacent to the epithelial ducts. To accomplish that
end, laminin, collagen III and IV, fibronectin, integrin 6ß1, and
plasminogen activators (uPA and tPA) were examined by
immunocytochemistry in developing and neonatally estrogenized rat
prostates. Stromal cytology was identified by immunostaining for
-actin, vimentin, and von Willebrand factor as well as by
electron microscopy and cell proliferation was assessed by
proliferating cell nuclear antigen (PCNA) labeling. To determine
whether enzymes involved in developmental ECM remodeling are influenced
by estrogen exposure, the activity of previously identified
metalloproteases and plasminogen activator in the rat prostate were
assessed by zymography and enzyme assay.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Immunocytochemistry
Immunocytochemistry was performed according to previously
published methods (17). Briefly, frozen prostatic complexes or
individual lobes were mounted on precooled chucks (-20 C) in a
Reichert-Jung cryostat and sections (6 µm) were thaw-mounted on
gelatin-coated glass slides. Whenever possible, individual lobes were
sectioned longitudinally to reveal the proximal-distal orientation. At
4 C, the sections were fixed in 2% paraformaldehyde, rinsed, incubated
with appropriate 2% blocking serum (goat or horse), and subsequently
incubated overnight with primary antibody. The specific antibodies,
sources and concentrations used are presented in Table 1
. As a negative control, normal rabbit
IgG (Vector Laboratories, Inc., Burlingame, CA) or normal mouse ascites
fluid (Sigma Chemical Co.) was substituted for primary
antibody on separate sections of all tissues analyzed to determine
nonspecific binding.
|
For PCNA immunostaining, 6 µm paraffin sections were mounted on glass slides, deparaffinized in xylene, gradually hydrated with decreasing concentrations of alcohol, rinsed, incubated with 2% blocking serum (horse) and incubated overnight at 4 C with monoclonal antibody raised against rat PCNA (Upstate Biotechnology, Inc., Waltham, MA). The rest of the assay was as described for frozen sections.
Transmission electron microscopy
Control and estrogenized pups were anesthetized on day 10 with
ketamine/xylazine (50 mg/kg Ketaset, Bristol Laboratories, Syracuse,
NY; 10 mg/kg Rompum, Mobay Corp., Shawnee, KS) and perfused through the
left ventricle with 150 ml paraformaldehyde fixative (2.5%
paraformaldehyde, 2% sucrose, 0.1% glutaraldehyde in 0.01
M PBS at pH 7.3). Accessory sex gland complexes were
removed, placed in paraformaldehyde fixative, and ventral prostate
lobes were microdissected under a dissecting microscope. The ventral
prostates were fixed overnight in paraformaldehyde fixative, washed
twice in 0.1 M PBS (30 min each), postfixed for 1 h in
2% osmium tetroxide in 0.1 M PBS at pH 7.3 and gradually
dehydrated in alcohol. The lobes were cleared twice in propylene oxide
(20 min each) and embedded in epon resin. The resin was allowed to dry
at room temperature overnight and was then baked at 60 C in a vacuum
for 2 days. Sections (600 angstroms thick) were mounted on copper grids
and stained with 2% uranyl acetate and lead citrate. Photographs were
taken on a Jeol JEM-100S electron microscope.
Tissue extract preparation for proteinase analysis
Because the ventral prostate lobes were quite small in the early
postnatal and young animals, it was necessary to pool tissues from
several animals for homogenization. Ventral prostates were pooled at
day 10 from 10 oil-treated rats and 18 estrogenized rats, at day 21
from 4 oil-treated rats and 6 estrogenized rats, and at day 35 from 2
oil-treated rats and 4 estrogenized rats. The frozen tissues were
thawed on ice, minced, and homogenized (1 g/10 ml medium) in 0.1%
Triton X-100 (pH 6.5) using a Teflon glass homogenizer (15 strokes).
The homogenates were centrifuged at 12,000 x g for 20
min. The supernatants were removed and the protein concentration was
estimated using bicinchoninic acid (Pierce Chemical Co.,
Rockford, IL) with BSA as the standard. The extracts were frozen in
powdered solid CO2, and stored frozen at -20 C until used
for proteinase zymography or plasminogen activator activity. This was
repeated three times for both treatment groups at the different time
points.
Zymography of metalloproteinases
Aliquots (10 µg protein) of prostate extracts were subjected
to electrophoresis in gelatin-containing polyacrylamide (10%
acrylamide) gels in the presence of SDS under nonreducing conditions as
previously described (25). The gelatin substrate was present at 0.1%
final concentration in the gel. The gels (0.75 mm thick) were
electrophoresed for 35 min at 200 V in a Bio-Rad MiniProtean II system
(Bio-Rad Laboratories, Inc., Richmond, CA). Following
electrophoresis, the gels were rinsed with distilled water and washed
with gentle shaking at room temperature with 2.5% Triton X-100 (2
changes) for 1 h. The gels were again rinsed with distilled water
and incubated overnight (18–20 h) in 50 mM Tris-HCl (pH
8.4) containing 5 mM CaCl2 at 37 C. Following
the incubation, gels were stained with amido black. Areas of
proteolysis appeared as clear zones against a blue background.
Molecular mass determinations were made with reference to prestained
protein standards (Bio-Rad Laboratories, Inc.)
coelectrophoresed in the gels.
Plasminogen activator assay
Plasminogen activator activity was monitored with an enzyme
assay that quantitates total tissue PA activity including uPA and tPA.
Three independent assays were performed, and each assay contained
tissue extracts from oil and estrogenized prostates at day 10, 21, and
35. The enzyme assays were carried out in 96-well multiwell plates
(Linbro, Flow Laboratories, McLean, VA) as previously described using
H-D-val-leu-lys-pNA (S-2251) as plasmin substrate (26). The final
reaction volume was 0.10 ml containing 100 mM
Tris(hydroxymethyl)-aminomethane (Tris)-HCl (pH 8.8), 0.5% Triton
X-100, 2.8 µg plasminogen, and 50 µg S-2251 (Kabi, Franklin, OH).
The reaction was carried out by incubating 10 µl of extract with
plasminogen overnight at room temperature. The assay was started with
the addition of S-2251 (25 µl) and the change in A405 was
monitored with a Titertek multiscan plate reader. Control incubations
included those of S-2251 with buffer alone or plasminogen alone.
Plasminogen activator activities were determined after subtraction of
values found for extracts incubated without plasminogen from those
incubated with plasminogen. Results are presented as Plough Units
determined using a standard urokinase preparation
(Calbiochem, La Jolla, CA).
Due to interassay variability, the PA activities were analyzed by three-way ANOVA comparing treatments, days, and assays. Post test calculation of the t value was determined and significance was accepted at P < 0.05.
Assay of 26-kDa protease activity
We have isolated (unpublished data) the 26-kDa protease that
strongly cleaves gelatin in zymograms from ventral prostate extracts
and secretions (25). It selectively cleaves the chromogenic peptide
H-D-Ile-Pro-Arg-pNA (S-2288), and this property is the basis for its
assay. The enzyme assay was carried out in 96-multiwell microtiter
plates in 0.10 ml final volume containing ventral prostate extract
(4.8–13.1 µg protein), 0.9 mM peptide (Kabi) and 50
mM Tris-HCl (pH 7.5). The reaction was started by the
addition of substrate, incubated at RT, and the changes in
OD405 monitored over time with a Titertek multiscan plate
reader. Control incubations included those without substrate or without
extract and these ODs were subtracted from those found with extract and
substrate together. One unit of enzyme activity is equal to the change
of one unit OD405/min. Specific activities are expressed as
mU/mg tissue extract protein. Data were analyzed by one way ANOVA with
post t test where appropriate.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-actin staining, a basal
lamina was produced by these cells. Likewise, endothelial cells
produced a basal lamina as they developed soon after birth. Thus by day
5 and onward, laminin was immunolocalized to the basement membrane and
into the immediate periductal smooth muscle cells in what appeared as a
continuum in the control animals (Fig. 1A). At this time, the
laminin receptor, integrin 6ß1, immunolocalized to the basal
aspect of the basal epithelial cells and, to a lesser degree, to the
smooth muscle cells surrounding the ducts (Fig. 1B). The localization
pattern observed for collagen IV was similar to the laminin staining
(Fig. 1C). As the ducts elongated and the smooth muscle layer thinned,
the periductal stain for laminin and collagen IV thinned as well (Fig. 1H). In contrast, immunostain for fibronectin and collagen III
exhibited a diffuse localization throughout the mesenchymal/stromal
compartment during the early days of prostate morphogenesis (Fig. 1D),
and this began to concentrate to the periductal stromal cells around
day 10.
|
). However, the
estrogenized prostates possessed a distinct layer of cells immediately
superficial to the basement membrane that was free of laminin and
collagen IV immunostain. Peripheral to this laminin-free zone, a layer
of cells immunostained for both laminin and collagen IV, resulting in a
halo effect around each duct in cross-section (Fig. 1E). The
laminin-free cells were also negative for 6ß1 integrin, which
resulted in an 6ß1 integrin-negative zone between the basal and
smooth muscle cells (Fig. 1F). The laminin-free zone surrounding the
ducts was most prominent in the proximal region and thinned toward the
distal tips. This abnormal pattern of laminin and collagen stain
persisted out to day 30 in many of the tissues examined (Fig. 1, I–L).
Interspersed within the laminin-free zone were individual cells that
immunostained for laminin (Fig. 1, I–L). Immunostaining for
endothelial cells with von Willebrand factor antibody identified these
individual cells as capillaries. When the estrogenized prostates were
stained for fibronectin, collagen III, and vimentin, all stromal cells
were positive at day 6 similar to the control prostates (Fig. 1G).
Because these molecules are produced exclusively by stromal cells, this
finding indicated that the laminin-free cells were of mesenchymal
origin.
Stromal cell characterization
Because the stromal cell cytology appeared to be altered by
neonatal estrogenization, we next sought to characterize these cells.
Adjacent sections of estrogenized prostates were immunostained for
laminin and -actin to identify smooth muscle cells. Laminin and
-actin always colocalized to the same periductal cells, whereas the
laminin-free cells were devoid of this smooth muscle marker (Fig. 2, A and B). In
contrast, all periductal cells in the estrogenized prostates
immunostained for vimentin (Fig. 2C), again indicating that the
laminin-free cells were of mesenchymal origin. In addition, both
periductal stromal cell-types in the estrogenized prostates were
positive for estrogen receptor- (Fig. 2F).
|
). The majority of organelles in smooth
muscle were located in the conical ends of the nucleus, and the cell
membrane contained multiple invaginations and small vesicles at its
margins. As previously noted at the light microscopy level (10), the
periductal smooth muscle layer was several cells thick proximally and
thinned to 1–2 cells in the distal region of the ducts. In the
proximal regions of the control ventral prostate, a thin single-cell
layer of fibroblasts was identified between the basement membrane and
the smooth muscle cell layer as previously observed by Flickinger (16)
(Fig. 3A). These differentiated fibroblasts contained numerous free
ribosomes and an extensive network of rough endoplasmic reticulum
throughout the elongated cytoplasmic processes. Adjacent fibroblasts
did not form tight junctions, but rather ECM molecules were observed
traversing the intercellular space from the smooth muscle cells to the
basement membrane. This single cell layer of fibroblasts thinned toward
the central and distal aspects of the ducts (Fig. 3B). In contrast to
the thin, single cell layer of fibroblasts in control prostates, a
thick layer of differentiated fibroblasts was present in the
estrogenized prostates between the smooth muscle cells and the basement
membrane (Fig. 4, A and B). This layer
was several cells thick in the proximal ducts and thinned to a 1–2
cell layer in the distal tips (Fig. 4B). There was an abundance of
extracellular space between the multiple fibroblasts, and this space
contained numerous striated collagen fibrils. These data indicate that
the laminin-free zone observed by light microscopy in estrogenized
prostates is comprised of differentiated fibroblasts. Similar to the
findings observed with von Willebrand staining, capillaries were often
seen within the fibroblast zone in the estrogenized ventral prostate
(Fig. 4B).
|
|
). In day 10 estrogenized
prostates, the frequency of epithelial PCNA staining was dramatically
reduced whereas the periductal stromal cells immediately adjacent to
the basement membrane were PCNA positive (Fig. 2E). While interacinar
fibroblasts were also frequently PCNA positive in estrogenized
prostates, the periductal stromal cells peripheral to the fibroblast
layer, presumably the smooth muscle cells, were largely PCNA negative.
These findings suggest that the periductal fibroblasts were actively
proliferating in response to estrogen exposure.
Plasminogen activators
In the prostate, there are two forms of PA, tissue plasminogen
activator (tPA) and urokinase plasminogen activator (uPA), the latter
being the major form secreted into seminal plasma (24). Because the
cleaved product, plasmin, is proteolytic for the ECM and can activate
other ECM proteases, PAs have been implicated in the localized
proteolysis associated with remodeling during development. In the
present study, immunocytochemistry for tPA and uPA was performed on
control and estrogenized ventral prostates. tPA immunostained to the
basement membrane and periductal smooth muscle cells between days 5 and
10 (Fig. 2G), and this pattern persisted through day 30. A similar
staining pattern was observed in estrogenized prostates; however, the
periductal zone of fibroblasts was devoid of tPA (Fig. 2H). uPA
localized primarily to epithelial cells and, to a lesser extent, to the
periductal smooth muscle region. On day 10, immunostaining for both
cell types was weak and intensity gradually increased as the cells
differentiated (Fig. 2I). By day 30, staining for uPA was primarily
localized to the apical epithelium in control prostates (Fig. 2K). In
contrast, epithelial uPA immunostaining was significantly reduced
following neonatal estrogen exposure, and this effect was most
pronounced in the central ducts. (Fig. 2, J and L). While low amounts
of uPA persisted in the periductal smooth muscle region of estrogenized
prostates at day 10, it did not immunolocalize to the periductal
fibroblast zone (Fig. 2J).
Total PA activity in the prostates at day 10, 21, and 35 is
presented in Fig. 5. In oil-treated
control ventral prostates, the total PA activity exhibited a U-shaped
curve with high activity at day 10, reduced activity at day 21
(P < 0.05), and higher levels again at day 35
(P < 0.05 vs. day 21). The same response
curves were observed in the dorsal and lateral prostate extracts. This
change in prostatic PA activity is thought to reflect the two types of
PAs present with tPA activity higher during morphogenesis (day 10) and
secretory uPA expressed following puberty (day 35). There were no
differences in total PA activity between the control and estrogenized
ventral prostates at day 10 or day 21; however, by day 35, there was a
significant reduction in total PA activity in the estrogenized
prostates compared with controls. Based on the immunocytochemical data,
this decrease most likely reflects the reduced production of secretory
uPA by the epithelial cells.
|
). This complex of activities is
considered to correspond to MMP-2 (type IV collagenase) and is thought
to be involved in developmental remodeling of the rat prostate (24, 25). The 76- and 71-kDa activities are proenzyme forms and the 64-kDa
is the active form of MMP-2. These activities were sharply diminished
at day 21 and by day 35, minimal active MMP-2 existed while low levels
of the proenzyme 71 kDa form were present. Neonatal estrogen exposure
did not alter the prostatic activities of these three forms of
MMP-2 at any of the time points examined. Minor gelatinolytic
activities were also observed in the day 10 ventral prostates at
approximately 85, 43, and 35 kDa, which did not appear in the older
tissues and may reflect proteases involved with morphogenesis. These
activities were not affected by estrogen exposure in the day 10
prostate. A previously identified secretory protease of 26 kDa was
initially observed in the day 21 prostate in the present study with
higher amounts present at day 35. This activity was consistently
suppressed by neonatal estrogen exposure (Fig. 6). The suppression of
the 26 kDa protease activity by estrogen treatment observed in
gelatinolytic gels was substantiated by quantitative assays of its
activity (Table 2).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
6ß1 integrin (laminin
receptor), tPA and uPA and are positive for fibronectin and collagen
III. Thus, the composition of the ECM immediately adjacent to the
basement membrane as well as the stromal cell type in close proximity
to the epithelium are markedly altered following neonatal
estrogenization. This structural perturbation of the prostate early in
development may be responsible for some of the permanent branching,
growth, and differentiation defects observed as the animals age. Previous work by Flickinger (16) has shown that a thin layer of periductal fibroblast cells resides immediately adjacent to the basement membrane in the adult rat ventral prostate. We observe that this thin single cell layer of differentiated fibroblasts is also present in the developing rat prostate at day 5 primarily encircling the proximal ducts. PCNA staining of control prostates demonstrates that periductal fibroblasts are normally growth quiescent during development. In contrast, periductal fibroblasts in neonatally estrogenized prostates are actively proliferating and form a multicellular layer four or more cells thick proximally, which thins toward the distal tips. Our previous studies have shown that the proximal and central prostatic ducts in estrogenized prostates retain permanent imprints throughout adulthood characterized by epithelial differentiation defects (10). It is noteworthy that the presence of the thick periductal fibroblast zone in the proximal-to-central ducts directly corresponds to this same region, which suggests a potential relationship between these events. Increased PCNA labeling was also observed in interductal fibroblasts of estrogenized prostates, which may account for the relative increase in overall stromal mass observed in these glands.
We have previously shown that ER is autoinduced in periductal smooth
muscle cells immediately following neonatal estrogen treatment and have
proposed that smooth muscle cells are a direct target of estrogenic
action in this gland (15). Herein, we show that ER is also
up-regulated in the periductal fibroblasts of estrogenized prostates,
thus, it is possible that estrogen may directly regulate fibroblast
proliferation. Alternatively, because the periductal smooth muscle
cells express ER, estrogen stimulation of fibroblast proliferation
may be mediated indirectly through those cells. For example, prostatic
smooth muscle cells secrete TGF-ß1 (33) and, in turn, TGF-ß1 is a
mitogen for fibroblasts in other systems (34, 35). Importantly, we have
reported that active TGF-ß1 is increased in the periductal smooth
muscle region of rat ventral prostates immediately following neonatal
estrogen treatment (36). Thus, an estrogen-induced increase in active
TGF-ß1 may stimulate proliferation of periductal fibroblasts.
Previous examination of ERß in the developing prostate showed minimal
epithelial expression during the neonatal period that was not initially
affected by estrogen exposure (11). While it is possible that an
estrogenic effect on prostatic fibroblasts may be indirectly mediated
through epithelial cell ERß, the likelihood is considered low due to
limited ERß levels present at that time.
The ECM is an important regulator of epithelial cell polarity,
proliferation, and differentiation (20). ECM molecules mediate
cell-to-cell interactions via specific receptors (22); thus, it is
possible that in the developing prostate, smooth muscle and epithelial
cells communicate through laminin (or collagen IV) cross-links with
specific integrin molecules expressed on both cell types. In this
manner, the ECM may transduce critical cell-to-cell signals and link
intracellular signaling pathways. As support of this possibility, we
identified an integrin that specifically binds laminin, 6ß1
integrin, on the basal aspect of basal epithelial cells as well as on
periductal smooth muscle cells. A similar observation has been reported
for human prostate carcinoma samples (37). The proximity of the smooth
muscle cells to the basement membrane could allow for cross-links
through extracellular laminin molecules. Following estrogenization,
however, this potential intercellular bridge is physically displaced by
a thick layer of fibroblasts that do not produce laminin or express
6ß1 integrin. Thus, by changing the ECM composition outside the
epithelium, the fibroblast zone may be disrupting smooth
muscle-epithelial communication and altering epithelial cell
differentiation.
The ECM also plays an important role in regulating branching morphogenesis and acts as a restraint on bud formation in many branched structures (20, 38). During rat prostatic development, extensive ductal branching occurs during the first 15 days of life (39). In order for branching to initiate, local proteolysis of ECM components by tissue proteases must occur to allow for penetration of epithelial buds through the basement membrane and the surrounding ECM (20, 24). In the present study, tPA and uPA were absent from the ECM immediately outside the basement membrane in estrogenized rats, whereas the PAs were in intimate contact with the basement membrane in controls. Neither the intensity of staining nor activity level of the PAs were affected by neonatal estrogenization during the branching period. Because tPA and uPA are not found in the periductal fibroblast zone after estrogenization, these cells may function to insulate the ECM adjacent to the basement membrane from protease degradation and act as a constraint on ductal branching during a critical developmental period. The absence of PAs in this region is more significant when considering their ability not to only act as general proteases but to also activate MMPs.
Specific MMPs have been previously identified that are potentially important during prostatic morphogenesis, the most notable one being a temporally expressed MMP-2 (24). Interestingly, a recent report demonstrated that MMP-2 activity is prolonged during prostatic development in neonatally hypothyroid rats, a condition that leads to enlarged prostates in adulthood (40). In the present study we examined MMP activity using gelatinolytic gels and again observed a high level of MMP-2 activity (76, 71, 64-kDa collagenase IV complex) at day 10 with decreasing activity as the prostate matured. In addition, minor collagenase activity was observed at 85-, 43-, and 35-kDa on day 10 that was not observed thereafter; thus, it is possible that these proteinases contribute to localized proteolysis during development. However, none of these activities were influenced by neonatal exposure to estrogens; therefore, we conclude that regulation of protease levels is not an apparent mechanism of action in the reduced branching and prostate size observed following neonatal estrogenization. It is important to keep in mind that while MMP levels are not changed, their localization with regards to basement membrane proximity may be altered by the presence of a thick layer of fibroblasts similar to PAs.
Neonatal estrogen exposure caused a significant reduction in secretory proteases as the prostates matured. Thus, during normal development, secretory uPA is localized to the apical aspect of epithelial cells, which accounts for the high level of total PA activity at day 35. Likewise, a 26-kDa secretory protease specific to the ventral lobe has been identified on gelatinolytic gels (25), which is first observed on day 21 as the epithelium undergoes functional cytodifferentiation. Reduced apical immunostain for secretory uPA in epithelial cells was noted in the central ducts of estrogenized prostates. Similarly, total PA activity was significantly reduced in the day 35 estrogenized prostates compared with controls, which was most likely a result of reduced secretory uPA. On gelatinolytic gels and in a quantitative assay with a substrate-specific peptide, the major 26 kDa secretory proteinase was markedly reduced in estrogenized prostates at day 21 and 35 compared with controls. Because these activities are functions of a fully cytodifferentiated epithelium, these data provide further support that neonatal estrogenization alters the differentiation of the prostatic epithelium.
In summary, neonatal estrogen stimulates the proliferation of periductal fibroblasts forming a zone that can potentially disrupt cellular and biochemical interactions essential to normal morphogenesis and differentiation of the prostate epithelium. The disruption during the critical developmental period that occurs in the first 15 days of life may lead the prostate toward abnormal growth and function that persists to adulthood. Stimulation of the fibroblast zone and subsequent alterations in ECM composition and paracrine interactions may be key mechanisms through which changes associated with neonatal estrogenization are precipitated.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received July 10, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
6-integrins in the normal and neoplastic
prostate. Am J Pathol 145:167–174[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  J. J. Bianco, S. J. McPherson, H. Wang, G. S. Prins, and G. P. Risbridger Transient Neonatal Estrogen Exposure to Estrogen-Deficient Mice (Aromatase Knockout) Reduces Prostate Weight and Induces Inflammation in Late Life Am. J. Pathol., June 1, 2006; 168(6): 1869 - 1878. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Atanassova, C. McKinnell, J. Fisher, and R. M Sharpe Neonatal treatment of rats with diethylstilboestrol (DES) induces stromal-epithelial abnormalities of the vas deferens and cauda epididymis in adulthood following delayed basal cell development Reproduction, May 1, 2005; 129(5): 589 - 601. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Huang, Y. Pu, S. Alam, L. Birch, and G. S. Prins Estrogenic Regulation of Signaling Pathways and Homeobox Genes During Rat Prostate Development J Androl, May 1, 2004; 25(3): 330 - 337. [Full Text] [PDF]
|
|
|
|
 |
|
 J. G. Ramos, J. Varayoud, L. Kass, H. Rodriguez, L. Costabel, M. Munoz-de-Toro, and E. H. Luque Bisphenol A Induces Both Transient and Permanent Histofunctional Alterations of the Hypothalamic-Pituitary-Gonadal Axis in Prenatally Exposed Male Rats Endocrinology, July 1, 2003; 144(7): 3206 - 3215. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. J. Bianco, D. J. Handelsman, J. S. Pedersen, and G. P. Risbridger Direct Response of the Murine Prostate Gland and Seminal Vesicles to Estradiol Endocrinology, December 1, 2002; 143(12): 4922 - 4933. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. S. Prins, W. Y. Chang, Y. Wang, and R. B. van Breemen Retinoic Acid Receptors and Retinoids Are Up-Regulated in the Developing and Adult Rat Prostate by Neonatal Estrogen Exposure Endocrinology, September 1, 2002; 143(9): 3628 - 3640. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Nilsson, S. Makela, E. Treuter, M. Tujague, J. Thomsen, G. Andersson, E. Enmark, K. Pettersson, M. Warner, and J.-A. Gustafsson Mechanisms of Estrogen Action Physiol Rev, October 1, 2001; 81(4): 1535 - 1565. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. G. Ramos, J. Varayoud, C. Sonnenschein, A. M. Soto, M. Munoz de Toro, and E. H. Luque Prenatal Exposure to Low Doses of Bisphenol A Alters the Periductal Stroma and Glandular Cell Function in the Rat Ventral Prostate Biol Reprod, October 1, 2001; 65(4): 1271 - 1277. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. S. Prins, L. Birch, J. F. Couse, I. Choi, B. Katzenellenbogen, and K. S. Korach Estrogen Imprinting of the Developing Prostate Gland Is Mediated through Stromal Estrogen Receptor {alpha}: Studies with {alpha}ERKO and {beta}ERKO Mice Cancer Res., August 1, 2001; 61(16): 6089 - 6097. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Atanassova, C. McKinnell, K. Williams, K. J. Turner, J. S. Fisher, P. T. K. Saunders, M. R. Millar, and R. M. Sharpe Age-, Cell- and Region-Specific Immunoexpression of Estrogen Receptor {{alpha}} (But Not Estrogen Receptor {beta}) during Postnatal Development of the Epididymis and Vas Deferens of the Rat and Disruption of This Pattern by Neonatal Treatment with Diethylstilbestrol Endocrinology, February 1, 2001; 142(2): 874 - 886. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Habermann, W. Y. Chang, L. Birch, P. Mehta, and G. S. Prins Developmental Exposure to Estrogens Alters Epithelial Cell Adhesion and Gap Junction Proteins in the Adult Rat Prostate Endocrinology, January 1, 2001; 142(1): 359 - 369. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
