| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
ARTICLES |
Medical Research Council Reproductive Biology Unit (J.S.F., K.J.T., H.M.F., P.T.K.S., R.M.S.), Centre for Reproductive Biology, Edinburgh EH3 9EW, Scotland, United Kingdom; Renal Unit (D.B.), Massachusetts General Hospital, Charlestown, Massachusetts 02129
Address all correspondence and requests for reprints to: Richard M. Sharpe, Medical Research Council Reproductive Biology Unit, Centre for Reproductive Biology, Edinburgh EH3 9EW, Scotland, United Kingdom.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
knockout (ERKO) mice are infertile (1) due to problems with
spermatogenesis and changes in sexual behavior (2). The existence of a
viable mouse carrying the ERKO phenotype also prompted reevaluation of
the estrogen receptor status in some patients. As a result, a few
reports have emerged in the literature of men with estrogen resistance
due to either a mutation in the estrogen receptor- (ER) gene (3)
or in the P450 aromatase gene that encodes the enzyme required for the
synthesis of estrogens (4, 5). These men show several notable
abnormalities though, as yet, only hints of abnormal reproductive
function attributable to loss of estrogen action.
Studies using the ERKO mouse have highlighted the efferent ducts of the
epididymis as an important site of estrogen action. These ducts appear
abnormally distended in the ERKO males due to impairment of fluid
resorption, although the exact mechanism responsible for this change is
not known (6). Recent studies using immunocytochemistry have identified
this area as having the strongest localization of ER in the male
reproductive tract in several species [rat and marmoset (7), goat (8),
and rooster (9)]. Expression of the recently identified ERß has also
been identified in the efferent ducts using RT-PCR, although at much
lower levels than the expression of ER (10).
Anatomically, the efferent ducts extend from the rete region of the testis to the initial segment of the epididymis (11). They are composed of several ducts (the exact number depending on the species), which in the rat join together to form one duct which enters the epididymis (12). The ducts are lined with a simple epithelium which is composed largely of two cell types, ciliated and nonciliated cells that normally have a high columnar appearance with either an extensive brush border or cilia. The efferent ducts are related embryologically to the kidney with both forming from the developing mesonephros (13). The major role ascribed to the efferent duct, like parts of the nephron, is fluid resorption and to a lesser extent protein synthesis, secretion, and resorption. Studies suggest that as much as 89% of the fluid entering the ducts from the rete testis along with spermatozoa, is resorbed before reaching the caput epididymis (14). The cellular mechanisms responsible for this resorption, and their regulation, are largely unstudied. The vast amount of work published regarding kidney function in this area may yield some potential clues.
A family of proteins termed Aquaporins have been identified and cloned from renal tissue, and these have been shown to act as selective water channels by expression studies in Xenopus oocytes (15). Aquaporin-1, formerly known as CHIP-28, has been localized to the apical brush border and basolateral membranes of both the proximal tubule and the long thin descending limb of the nephron (16). Freeze fracture studies have shown that AQP-1 forms tetramers in membranes (17), but each individual subunit is believed to function as a water channel (18). Hydropathy analysis of the complementary DNA has shown that AQP-1 consists of six membrane-spanning regions with both the -COOH and -NH2 regions intracellular, but how water passes through the molecule is not known (18). AQP-1 has been immunolocalized in the efferent ducts (19) and other resorptive epithelia of the adult rat (see Ref. 17).
The initial aim of the present study was to establish whether the
developmental expression of AQP-1 in the efferent ducts was comparable
with that for ER, which we have shown previously to be expressed at
high levels in the efferent ducts of the rat and marmoset in
fetal/neonatal life and thereafter throughout life (7). Having
established this, the aim was then to determine whether the
immunoexpression of AQP-1 was subject to hormonal modulation during
this period. It is well established that FSH and androgens, stimulated
by LH secretion, play vital roles in development of the testis and
reproductive tract, so the consequences of withdrawal of gonadotropin
support during the neonatal period on AQP-1 expression was assessed by
administration of a potent GnRH antagonist (GnRHa). The role of
estrogens in development of the male reproductive system is unknown,
though it is established that estrogens can suppress gonadotropin
(especially FSH) secretion in the neonatal period (20). We therefore
assessed whether neonatal administration of DES, at a dose known to
cause abnormal development of the testis (21), was able to alter
immunoexpression of AQP-1. By comparison of the results obtained in
DES-treated rats with those treated with GnRHa, it was also hoped that
distinction between direct and indirect (gonadotropin-suppression)
effects of DES would be possible. The results obtained demonstrate that
immunoexpression of AQP-1 is reduced by estrogen exposure and that this
coincides with evidence of fluid accumulation in this part of the
reproductive tract.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In animals up to 25 days of age, the testis and epididymis were removed together and immersion fixed in Bouin’s fluid for 5.5 h at room temperature before being transferred into 70% ethanol. Before processing, the tissue was cut into 2–6 pieces with a razor blade to allow more accurate sectioning of the efferent ducts. After 35 days of age, rats were anesthetized with halothane and perfusion fixed via the thoracic aorta, first with 0.9% saline containing 0.01% heparin until the testicular blood vessels cleared and then with Bouin’s fixative for 30 min, as described previously (22). The testis and epididymis were then dissected apart to leave the proximal efferent ducts attached to the testis. These tissues were then postfixed for a further 5.5 h in Bouin’s fluid before being transferred into 70% ethanol.
Testes with the epididymides attached were removed from captive bred marmoset monkeys (Callithrix jacchus) which were classed as being either neonates, infants, prepubertal or adults (aged 1 day, 8, 18–24, 54–62, and 92–112 weeks, respectively; n = 2–4 per group). Tissues were immersion-fixed for 5.5 h in Bouin’s fluid before being processed as described for rat tissues. Other tissues were also collected for use in other experiments to be described elsewhere.
Tissue processing
Fixed tissue was processed for 17.5 h in an automated
Shandon processor and embedded in paraffin wax. Sections were cut at 5
µm and floated onto coated slides (2% 3-aminopropyltriethoxy-saline;
Sigma) and dried at 50 C overnight before being used for
immunocytochemistry.
Antibody production
Immunolocalization of AQP-1 was determined using a previously
validated polyclonal antibody raised to human erythrocyte AQP-1 as
described elsewhere (16). The antibody and preimmune serum were used at
a dilution of 1:500.
Western analysis
Protein was extracted from adult male kidney and from the pooled
efferent ducts of 5 adult male rats. Each sample was homogenized in 300
µl chilled PBS containing 50 µl protease inhibitor cocktail
(complete protease inhibitor cocktail, Boehringer Mannheim, East
Sussex, UK). The efferent ducts were also removed from 6 control and 6
DES-treated rats on postnatal day 18 and protein extracted as described
above except that the homogenization volume was halved. To remove fat
and tissue debris from homogenates, samples were centrifuged at
3000 x g for 10 min at 4 C. The supernatant from below
the fat layer was decanted and stored at -20 C.
Protein samples were separated using SDS-PAGE. Gels contained 12% acrylamide (14 ml 30% acrylamide, 8 ml 1.5 M Tris-HCl (pH 8.8) and 9.6 ml distilled water). Gels were degassed and polymerized using 300 µl 1% ammonium persulfate and 9 µl TEMED (Sigma Chemical Co., St. Louis, MO). Gels were loaded with 75 µg of each protein sample and one lane contained 10 µl biotinylated molecular weight markers (Amersham, Buckinghamshire, UK). Each gel was run at 38 mA for approximately 3–4 h before being blotted onto a PVDF membrane (Immobilon-P, Millipore, Watford, UK) for 90 min at 150 mA. Membranes were blocked overnight at 4 C in 5% normal swine serum (NSS; SAPU Laboratories, Carluke, Scotland) in TBS-Tween (Tris-buffered saline, pH 7.4, containing 50 mM Tris-HCl, 150 mM NaCl, and 0.05% Tween 20). The AQP-1 primary antibody and preimmune sera were added at a dilution of 1:5000 in TBS-Tween containing 5% NSS and incubated for 2 h. After repeated washing in TBS-Tween, swine antirabbit peroxidase was added at a dilution of 1:5000 in TBS-Tween containing 5% NSS. The strip containing the molecular weight markers was incubated with streptavidin peroxidase (1:5000 in TBS-Tween in 5% NSS). Specific signals were detected using ECL (Amersham) on hyperfilm (Amersham) following the manufacturers’ instructions.
Immunocytochemistry
Slide-mounted sections were dewaxed in Histoclear (National
Diagnostics, Fleet Business Park, Hull, UK), rehydrated through a
graded series of ethanol and washed in distilled water. The sections
were then pretreated with 3% hydrogen peroxide in methanol to block
endogenous peroxidase activity. At this point an additional step to
block endogenous biotin (23) production was included for sections of
marmoset tissue, for reasons described elsewhere (7). The marmoset
sections were washed twice in PBS and then incubated with 0.01% avidin
(Sigma). After a further two 5-min washes in PBS, the sections were
incubated with 0.001% biotin (Sigma) and then washed twice for 5 min
in TBS (pH 7.4). Thereafter, the protocol for both rat and marmoset
tissues was identical. To block nonspecific binding sites, the sections
were incubated in NSS diluted 1:5 in TBS. The immune and preimmune sera
were prepared by diluting in the NSS block, and 100 µl were added to
each slide, which was coverslipped and incubated overnight in a
light-proof box at 4 C. Coverslips were then removed and the slides
washed in TBS (2 x 5 min) before incubation for 30 min with a
linking antibody, biotinylated swine-antirabbit serum (Dako Ltd.,
Cambridge, UK; diluted 1:500 in NSS). After two washes in TBS,
avidin-biotin conjugated horseradish peroxidase (Dako) was applied to
the slides for 30 min. The slides were given two final rinses in TBS
before the addition of a diaminobenzidine (DAB) based chromogen, which
left a brown colored precipitate at the sites of antibody localization.
Alternatively, for rat kidney sections, avidin-biotin conjugated
alkaline phosphatase (Dako) was applied for 30 min, and after further
washes in TBS (2 x 5 min), the slides were given a final wash in
100 mM Tris-MgCl buffer (100 mM NaCl and 50
mM MgCl; pH 9.5) before the addition of nitroblue
tetrazolium (NBT 337.5 mg/ml), 5-bromo-4 chloro-3-indolylphosphate (175
µg/ml) and 0.001% levomisole in 10 µl Tris-MgCl buffer to develop
color (blue) at the sites of antibody localization. The slides were
developed until the color reached the required intensity in controls,
before the reaction was stopped by immersing the slides in distilled
water. The slides were then counterstained with hematoxylin before
being dehydrated by immersion in a graded series of ethanol and then
being cleared in xylene. A coverslip was fixed over the sections using
Pertex mounting medium (Cell Path, Hemel Hempstead, UK).
Evaluation of immunostaining
Slides were examined and photographed using an Olympus Provis
microscope (Olympus Optical, London, UK) fitted with a Kodak DCS420
camera (Eastman Kodak, Rochester, NY). Captured images were stored on
an 8100 PowerPC computer (Apple MacIntosh) and compiled using Photoshop
3.0 before being printed using a Kodak XLS 8600 PS printer (Eastman
Kodak). To enable accurate comparison of immunostaining, sections of
tissue from animals at each age and/or treatment were processed for
immunolocalization in parallel. Where differences in the level of
immunostaining between treated and control animals was evident,
confirmation was sought by performing the immunocytochemistry for
control and treated tissues on the same slide. Tissue from three to six
animals at each age and treatment were evaluated on at least three
occasions to ensure the reproducibility of the results.
Efferent duct epithelial cell height
To confirm the impression that DES treatment had altered the
shape/height of epithelial cells in the efferent ducts, cross-sections
of efferent ducts were evaluated by image analysis from three to five
rats from control, DES- and GnRHa-treated groups at 18 days of age, and
further analysis was performed between DES and control rats at days 10,
25, 35, and 75 postnatal. The height of the efferent duct epithelium
was measured using an Olympus BH2 microscope fitted with a 40x plan
achromat objective and a 3.3x phototube (Olympus Optical Co.). The
image was captured using a Sony XC77CE video camera linked to a
personal computer with frame grabber and image pro image analyses
software (Media Cybernetics, Silver Spring, MD). The height of
the epithelium was measured using the length tool at right angles from
the base of the cell adjacent to the basement membrane to the luminal
surface of the cell. After measuring the length, the angle of the line
was measured to ensure that it was at 90 degrees. Any line was excluded
from the analyses if the angle was >10 degrees from 90. For each
animal, a total of 50 cells were measured with sampling from a number
of different ducts.
Hormone assays
Plasma levels of FSH were measured using a kit provided by the
NIDDK, (Bethesda, MD) and results expressed in terms of the rFSH-RP-2
standard.
Statistics
Comparison of testis weights, FSH levels, and efferent duct
epithelial cell height for the three groups at each age-point was made
using ANOVA and, where significant differences between groups were
indicated, subgroup comparisons used the same test with the variance
from the experiment as a whole as the measure of error.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
a, b, d, f, and g, respectively). In
rats up to the age of 18 days, when the ciliated and nonciliated cells
of the efferent ducts had not differentiated, all cells exhibited
strong immunostaining for AQP-1 confined mainly to the apical brush
border (Fig. 1, b and e). However, in animals aged 25 days or older it
was clear that AQP-1 was immunolocalized to the apical brush border of
the nonciliated cells and that the ciliated cells did not express AQP-1
(Fig. 1h). The intensity of AQP-1 immunoexpression appeared less marked
in efferent ducts of adult rats when compared with neonatal and
peripubertal rats (compare Fig. 1d and 1g), but under higher power
magnification it was clear that AQP-1 was immunoexpressed at similar
intensity in adult and younger animals (compare Fig. 1, b, e, and h).
At all ages, there was faint and more diffuse localization of AQP-1 on
the basal and lateral surface of the epithelial cells (Fig. 1, b, e,
and h).
|
, j–p), results comparable to those in the rat
were found, with AQP-1 immunolocalization to the efferent ducts being
evident from postnatal day 1 and at all stages of development checked
until adulthood. As in the rat, the most intense immunostaining for
AQP-1 occurred on the apical brush border, but lower intensity
immunoexpression was also evident on the basal and lateral surfaces of
the epithelial cells. Due to the intensity of the immunostaining
reaction in marmosets, it was difficult to identify whether ciliated
cells did not exhibit AQP-1 localization.
For both the rat and marmoset, the developmental pattern of expression
of AQP-1 in the efferent ducts was comparable with that found
previously by us for ER (7). In both species, it was noted that the
lumen of the efferent ducts was open at all ages but that its size
increased substantially in the rat from the time (
day 18) when
seminiferous tubule fluid (STF) starts to flow (compare Fig. 1 d,
f, and g); a similar change was evident in the marmoset at 62 weeks
(Fig. 1n) and is presumably indicative of STF production at this
age.
At no age assessed was AQP-1 localized to the testis in either the rat
(results not shown) or the marmoset (see Fig. 1p). However, in contrast
to the rat, in which there was no localization of AQP-1 to the rete
testis epithelium (not shown), in the marmoset there was a narrow
apical band of AQP-1 localization to the rete testis in both neonatal
(not shown) and peripubertal animals (Fig. 1p). In both the rat (Fig. 1i) and marmoset (Fig. 1j), immunolocalization of AQP-1 to the thin
descending limb of the loop of Henle was demonstrated as a positive
control.
Effect of hormone treatments on testis size and gonadotropin
levels
In the rat, administration of GnRHa caused a major reduction in
testicular size at days 18 and 25, and this decrease was comparable to
that induced by neonatal DES treatment (Table 1). In GnRHa-treated rats, FSH levels
were suppressed by >50% at 18 and 25 days, when
compared with controls, whereas FSH levels were only reduced
significantly in DES-treated rats at the earlier of these two
time-points (Table 1).
|
, f and i) than in their age-matched
controls (Fig. 2, d and g), but the intensity of AQP-1 immunoexpression
was either unchanged or was reduced only marginally. At both time
points, the epithelial cells of the efferent ducts in GnRHa-treated
rats exhibited a columnar appearance similar to controls, though
measurement of the apical height of the epithelial cells revealed a
significant reduction compared with controls (see below).
|
). First, immunoexpression of AQP-1 along the
brush border of the efferent ducts was reduced. Second, the morphology
of the epithelial cells in the efferent ducts changed from highly
columnar (in control animals) to cuboidal. Third, the lumens of the
efferent ducts were enlarged. Immunostaining for AQP-1 was clearly
evident in control animals at all ages (Fig. 2, a, d, g, and j) but was
reduced markedly in intensity in DES-treated rats on day 10 (Fig. 2, b
and c) and was virtually abolished at day 18 (Fig. 2e). The intensity
was still greatly reduced at day 25 (Fig. 2h), and to a lesser extent
at day 35 (Fig. 2, k and l). By day 75, apical immunostaining for AQP-1
was similar to control levels (data not shown). One additional
observation was that the more prominent AQP-1 immunostaining in
DES-treated rats at day 25 (Fig. 2h) compared with those at day 18
(Fig. 2e), was due mainly to immunostaining around the basal surface of
the efferent duct epithelium, rather than the apical localization
evident in controls (Fig. 2); this was also evident at day 75 (data not
shown). None of the changes described above were observed in rats at 18
or 25 days of age that were treated with GnRHa (Fig. 2, f and i,
respectively), despite this treatment causing similar or greater
reductions in testis size and FSH levels to that induced by DES (Table 1).
Effect of DES treatment on AQP-1 protein levels in efferent
ducts
Confirmation of the DES-induced reduction in immunoexpression of
AQP-1 was obtained by Western blot analysis of proteins extracted from
the efferent ducts of control and DES-treated rats on postnatal day 18
(Fig. 3). Lane 7 (Fig. 3) indicates the
control level of AQP-1 in the efferent ducts at 18 days, whereas lane 8
contains protein from DES-treated rats at 18 days and shows a marked
reduction in the level of AQP-1 protein. Lanes 5 and 6 show control
adult efferent ducts and kidney respectively. The higher molecular
weight band in lanes 5 and 6 are probably indicative of glycosylated
forms of the protein (16).
|
). Measurement of epithelial cell height
in efferent ducts from DES-treated rats demonstrated a 33% decrease in
cell height at 10 days, more than 50% at 18 and
25 days, and an approximately 20% decrease at 35 and 75 days when
compared with controls (Fig. 4). Unlike in DES-treated rats, epithelial
cells in GnRHa-treated rats were still columnar in shape but their
height (11.55 ± 1.85 µm; mean ±SD, n = 5) was
reduced by 29% (P < 0.01) at 18 days of age, compared
with 18 day controls (16.31 ± 1.48 µm), but was similar in
height to that found in 10-day controls (12.98 ± 1.83 µm; Fig. 4), consistent with delayed development. Epithelial cell height in
efferent ducts from DES-treated rats at 18 days of age was
significantly (P < 0.01) reduced compared with
GnRHa-treated rats.
|
). The latter change was most evident
at days 10 and 18 compared with controls (contrast panels a and d with
b and e in Fig. 2) but was still remarkable at day 25 when considering
the small size of the testis and the absence of lumens in most
seminiferous tubules (data not shown). Indeed, at day 25 perhaps the
more appropriate comparison is with the GnRHa-treated rats (compare h
and i in Fig. 2), which had similarly sized testes and which did
not exhibit distension of the efferent ducts (Table 1). Even at 35 days
of age, the efferent ducts of DES-treated rats (Fig. 2k) still appeared
to have consistently larger lumens than in controls (Fig. 2j). In view
of the distension of the efferent ducts in DES-treated animals
associated with reduced expression of AQP-1 protein, attention was
switched to the testis to assess whether there was any evidence of
fluid accumulation in the rete testis or seminiferous
tubules/cords.
Effect of DES treatment on the rete testis
In none of the DES-treated rats at any of the time points was
there evidence of distension of the seminiferous tubules, though at
days 10–25, few if any of the tubules had formed a lumen (data not
shown). In contrast, the rete testis was grossly enlarged at all ages
in DES-treated, compared with control, animals (Fig. 5) and appeared to extend more deeply
into the testis than in controls (see Fig. 5f). These changes were
evident in every treated animal (Table 1) but were most pronounced at
day 18 (Fig. 5d) and occurred despite the markedly smaller size of the
testes in DES-treated, compared with control, rats. There was no
evidence of comparable changes in either the control (Fig. 5) or
GnRHa-treated (not shown) rats (Table 1). Exfoliated germ cells were
evident in the lumen of the rete in DES-treated, but not control, rats
at day 35 (Fig. 5, g vs. h), and smaller numbers of
exfoliated germ cells were also evident in some of the DES-treated
males at 18 and 25 days (not shown) but were not observed in control or
GnRHa-treated animals. Similar findings were evident at day 75 with
DES-treated animals showing both distension of the rete testis and
exfoliated germ cells in the rete testis lumen (data not shown). It is
emphasized that, with the exception of animals sampled at day 10, all
other DES-treated rats that showed abnormalities of the rete testis and
efferent ducts had ceased DES treatment 6–63 days earlier.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
from early neonatal life to
adulthood. Having shown that this was the case, we then sought evidence
that immunoexpression of AQP-1 in the efferent ducts of the rat during
prepubertal development was subject to hormonal modulation,
particularly by estrogens. The results obtained suggest strongly that
AQP-1 is one of the factors involved in fluid resorption from the
efferent ducts and that estrogens play a developmental role in
regulating the expression of this water channel protein, though it is
unclear whether this is a direct effect or is the consequence of some
other effect of the estrogen (e.g. on differentiation of
epithelial cells in the efferent ducts). The data also raise important
questions concerning the role(s) of fluid flow in the developing
excurrent duct system of the male during fetal and neonatal life and
its potential modulation by hormones. Considering that fluid flow through the rete testis and efferent ducts is generally considered as being coincident with puberty and the production of seminiferous tubule fluid (24), it was somewhat surprising to find that AQP-1 immunoexpression in the efferent ducts was evident during fetal and/or early postnatal life in both the rat and marmoset, and thereafter was maintained through to adulthood. These findings are in agreement with those already published for the adult rat (19). AQP-1 localization in the rat was limited to the efferent ducts with no immunolocalization to either the testis or rete testis. During the neonatal period, before ciliated epithelial cells had differentiated in the efferent ducts, all of the epithelial cells showed a strong immunoprecipitate along the apical brush border and a weak reaction in the lateral and basal membranes. After day 25 postnatal, it was clear that ciliated cells did not express AQP-1. The marmoset showed a similar pattern of expression of AQP-1 in the efferent ducts to the rat, but in addition some immunolocalization of AQP-1 was evident along the apical surface of the rete testis epithelium.
In addressing the potential hormonal modulation of AQP-1 expression during early development, our first approach was to assess the consequences of withdrawing the gonadotropin support essential for pubertal development of the testis (25). Administration of GnRHa neonatally to rats is well established to result in near-complete gonadotropin suppression (26, 27), and results obtained in the present and other (unpublished) studies using the current treatment protocol confirm this. GnRHa treatment clearly delayed pubertal development of the testis, as would be expected from the suppression of FSH and LH levels. The cross-sectional size of the efferent ducts in treated rats was smaller than in the controls, especially at day 18, and the height of the epithelial cells was reduced, though their shape remained cuboidal. In all GnRHa-treated animals, AQP-1 was still immunoexpressed in the epithelium of the efferent ducts, though the intensity of immunoexpression was reduced marginally but consistently at day 18. This change and the minor change in efferent duct size induced by GnRHa treatment are probably symptomatic of delayed pubertal development. These findings suggest that gonadotropin support is probably not essential for the expression of AQP-1 in the efferent ducts during neonatal/pubertal development, and this would fit with data for the kidney in which AQP-1 expression is constitutively activated (28).
In contrast to the effects of gonadotropin withdrawal, neonatal administration of DES to rats resulted in major and prolonged changes in efferent duct morphology, size, and the immunoexpression of AQP-1. On postnatal days 18 and 25, and to a lesser extent at days 10 and 35, after administration of DES, immunoexpression of AQP-1 in the efferent ducts was absent or reduced in intensity. The immunocytochemical data do not enable consideration of whether this effect represents a direct effect of the DES on AQP-1 expression at the messenger RNA level, but Western blotting confirmed a reduction in the level of AQP-1 protein at day 18. The effects of DES on AQP-1 immunoexpression were evident at least 6–23 days after DES treatment had ceased and this, combined with the pattern of its effect (i.e. more marked at 18 than at 10 days), perhaps makes transcriptional regulation of the AQP-1 gene by estrogen an unlikely explanation for our findings, though an effect on programming of AQP-1 gene expression remains a possibility. As estrogen treatment neonatally is able to suppress FSH secretion and delay puberty (29), as confirmed in our studies, it is possible that reduced immunoexpression of AQP-1 is simply a consequence of this general change. This possibility can be discounted as the changes observed in expression of AQP-1 in GnRHa-treated rats, which were sampled at 18 and 25 days and which showed equivalent delays in pubertal development, were radically different to those observed after DES treatment. Also, the fact that the DES-induced reduction in AQP-1 immunoexpression coincided with distension of the efferent ducts could mean that distension per se results in reduced AQP-1 expression. This is considered unlikely, primarily because our studies in adult rats indicate that experimental induction of efferent duct distension results in unaltered immunoexpression of AQP-1 (Piner, Fisher & Sharpe, unpublished data).
A more feasible explanation for the DES-induced reduction in AQP-1 immunoexpression is that it is a consequence of other morphological changes produced in the epithelial cells of the efferent ducts. At all ages studied after DES treatment, the normally high columnar epithelial cells displayed a cuboidal appearance with loss of much of the brush border and apical cytoplasm; at all ages, including in adulthood, epithelial cell height was reduced significantly compared with controls. The apical cytoplasm of the nonciliated cells normally houses the endocytotic apparatus (30), which is implicated in fluid resorption. Epithelial cells within the efferent ducts of the ERKO mouse also exhibit a loss in cell height, accompanied by a large reduction in the number of vesicles and granules that comprise the endocytotic apparatus, and this is associated with failure of fluid resorption (6). It is therefore possible that neonatal overexposure to DES either retards or permanently impairs the morphological differentiation of the epithelial cells of the efferent ducts and/or their endocytotic apparatus. More detailed study of the efferent ducts of DES-treated rats through to adulthood may indicate whether or not this is a serious possibility, but such a change might explain the permanent impairment of spermatogenesis which occurs in some DES-treated rats (21, 31).
The present finding that neonatal overexposure of rats to an estrogen causes distension of the efferent ducts and rete testis seems, at face value, incongruous when considering that these changes are similar to those described for the male ERKO mouse in which estrogen action in the efferent ducts is presumably impaired. In both instances there is a (possibly permanent) reduction in height of the epithelial cells of the efferent ducts, which might imply that the correct level of estrogen is required for normal development of these cells and that either too little or too much estrogen action will impair development. On the other hand, there are notable differences between the DES-treated rats and ERKO mice. For example, distension of the efferent ducts and rete in the DES-treated rats is first evident neonatally/prepubertally rather than postpubertally as in ERKO males (2, 6) and, unlike in the latter, does not become progressively worse with age. Again, in DES-treated rats there is overgrowth of the rete testis before puberty, whereas no such change has been reported for ERKO males (2, 6). Our tentative conclusion from these and other comparisons is that the similarities between ERKO males and neonatally DES-treated males are probably coincidental and have resulted from treatment-induced abnormalities in different, and presently unknown, pathways.
DES treatment produced gross distension with apparent hyperplasia of the epithelium of the rete testis at all five ages, with the most pronounced change at day 18 when AQP-1 immunoexpression in the efferent ducts was most reduced. The distension induced by DES treatment was confined to the rete testis and efferent ducts with no apparent distension of the seminiferous cords/tubules at days 10–25. This is in contrast to other situations in adult males in which fluid resorption from the efferent ducts is impaired, for example in ERKO mice (6) or after ligation of the efferent ducts (24), when distension of seminiferous tubule lumens also occurs. This difference may be a consequence of the delayed development of the seminiferous tubules in DES-treated rats and the presumed delay in formation of Sertoli-Sertoli cell tight junctions (32); this delay would presumably allow any excess fluid to escape to the interstitial area.
Although the temporal association between reduced im-munoexpression
of AQP-1 in the efferent ducts and distension of the rete testis in
DES-treated rats could be interpreted rationally as cause and effect,
it is perhaps more likely that direct effects of DES on the rete testis
may have occurred, as ER is expressed in the rete epithelium in the
rat (7). In several DES-treated animals, the rete appeared to invade
the testicular parenchyma, rather than remaining restricted to its
normal superficial position underlying the tunica, a change that seems
difficult to ascribe simply to distension. In this respect,
abnormalities of the rete have also been reported in mice and in human
males who were exposed to DES in utero. (33). In a study
that examined 233 DES-exposed male mice, 56% showed various degrees of
papillary proliferation and hyperplasia of the rete testis epithelium.
Of these mice, 5% developed adenocarcinoma of the rete testis, a rare
malignant neoplasm (33).
Immunolocalization of AQP-1 to the apical surface of nonciliated cells within the efferent ducts supports the morphological findings that the resorptive apparatus is confined to this cell type (19, 34, 35). This finding updates the current model of fluid resorption proposed by Ilio and Hess (12). Water resorption occurs secondary to ion transport, and during transit through the efferent ducts the fluid concentration of Na+ is lowered, whereas that of H+ is increased (36), suggesting the existence of ion channels within the epithelial membranes. The immunolocalization of AQP-1 to the apical membrane implicates this protein as a primary candidate in the absorption of water into non ciliated cells. However, AQP-1 acts exclusively as a water channel and does not carry ions (15); therefore, other channels must exist to facilitate the entry of Na+ and maintain the osmotic environment. How water exits the cell is not known, but this may occur via Na+/K+-ATPase, which has been localized to the basal and baso-lateral membranes of the nonciliated cells of the efferent ducts (37, 38). There is little immunoreactive AQP-1 along the basal membrane of the nonciliated cells, suggesting that AQP-1 is of less importance in water transport across this membrane.
The present finding of immunoexpression of AQP-1 during perinatal life
in the rat and marmoset suggests indirectly that fluid is flowing
through the reproductive tract during its development. This idea is
supported by the fact that the testis and Wolffian duct develop from
the mesonephric tubules that function as a kidney (39). It has been
proposed that, during sexual differentiation, testosterone from the
fetal Leydig cells may be transported along the Wolffian duct by fluid
flow rather than by simple diffusion (39). The present demonstration of
AQP-1 immunoexpression in the duct system supports this line of
thinking and raises the question of what induces AQP-1 expression in
the male reproductive tract during fetal life. It has been shown that
AQP-1 is switched on by maternal corticosteroids in the rat fetal lung
(40), but the present findings, together with the demonstration of
androgen receptors and ER in the developing ducts (7), raises the
possible involvement of these sex steroids in induction of AQP-1. If
fluid flow in the excurrent duct system is an important factor in the
development of the reproductive tract of the male, alteration of fluid
flow/resorption as a consequence of abnormalities in exposure to
estrogens (or androgens) in utero or neonatally could
provide one indirect mechanism via which abnormalities in downstream
tissues (epididymis, vas deferens, seminal vesicle) might be
induced.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received December 4, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
& ß)
expression in the excurrent ducts of the adult male rat reproductive
system. J Androl 18:602–611This article has been cited by other articles:
 |
|
 |
|
  A. Wellejus, H. E. Jensen, S. Loft, and T. E. Jonassen Expression of Aquaporin 9 in Rat Liver and Efferent Ducts of the Male Reproductive System After Neonatal Diethylstilbestrol Exposure J. Histochem. Cytochem., May 1, 2008; 56(5): 425 - 432. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. H. Aneck-Hahn, G. W. Schulenburg, M. S. Bornman, P. Farias, and C. De Jager Impaired Semen Quality Associated With Environmental DDT Exposure in Young Men Living in a Malaria Area in the Limpopo Province, South Africa J Androl, May 1, 2007; 28(3): 423 - 434. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Assinder, R. Davis, M. Fenwick, and A. Glover Adult-only exposure of male rats to a diet of high phytoestrogen content increases apoptosis of meiotic and post-meiotic germ cells Reproduction, January 1, 2007; 133(1): 11 - 19. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H.-F. Huang, R.-H. He, C.-C. Sun, Y. Zhang, Q.-X. Meng, and Y.-Y. Ma Function of aquaporins in female and male reproductive systems Hum. Reprod. Update, November 1, 2006; 12(6): 785 - 795. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A Glover and S J Assinder Acute exposure of adult male rats to dietary phytoestrogens reduces fecundity and alters epididymal steroid hormone receptor expression. J. Endocrinol., June 1, 2006; 189(3): 565 - 573. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Da Silva, C. Silberstein, V. Beaulieu, C. Pietrement, A. N. Van Hoek, D. Brown, and S. Breton Postnatal Expression of Aquaporins in Epithelial Cells of the Rat Epididymis Biol Reprod, February 1, 2006; 74(2): 427 - 438. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Hermo, D. L. Chong, P. Moffatt, W. S. Sly, A. Waheed, and C. E. Smith Region- and Cell-specific Differences in the Distribution of Carbonic Anhydrases II, III, XII, and XIV in the Adult Rat Epididymis J. Histochem. Cytochem., June 1, 2005; 53(6): 699 - 713. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C. Branes, B. Morales, M. Rios, and M. J. Villalon Regulation of the immunoexpression of aquaporin 9 by ovarian hormones in the rat oviductal epithelium Am J Physiol Cell Physiol, May 1, 2005; 288(5): C1048 - C1057. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Hermo, D. Krzeczunowicz, and R. Ruz Cell Specificity of Aquaporins 0, 3, and 10 Expressed in the Testis, Efferent Ducts, and Epididymis of Adult Rats J Androl, July 1, 2004; 25(4): 494 - 505. [Abstract] [Full Text] [PDF]
|
|
|
|
