| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Department of Obstetrics and Gynecology, University MacDonald Women’s Hospital, University Hospitals of Cleveland, and Departments of Reproductive Biology and Physiology and Biophysics, Case Western Reserve University School of Medicine, Cleveland, Ohio 44106
Address all correspondence and requests for reprints to: George I. Gorodeski, M.D., Ph.D., University MacDonald Women’s Hospital, University Hospitals of Cleveland, 11100 Euclid Avenue, Cleveland, Ohio 44106. E-mail: gig{at}po.cwru.edu
Abstract
The objective of the study was to understand effects of estrogen
and aging on paracellular permeability of human vaginal-cervical
epithelia. Assays included determinations of transepithelial electrical
conductance across cultures of normal human ectocervical epithelial
cells on filters. Baseline transepithelial electrical conductance
across steroid-deprived cells from postmenopausal women was lower than
across cells of premenopausal women. Short-term (24–48 h) treatment
with 10 nM 17ß-estradiol increased transepithelial
electrical conductance in both groups of cells. In cells of
premenopausal women longer-term treatment with estrogen for up to
14 d had no additional effect on permeability, but in cells of
postmenopausal women it caused a secondary increase in transepithelial
electrical conductance that continued for the duration of the 2-wk
treatment. Binding assays of 17ß-[3H]estradiol revealed
saturable binding to high affinity (1.2–1.3 nM), low
capacity sites (0.2–1.2 pmol/mg DNA) in cells of both premenopausal
and postmenopausal women. In both types of cells treatment with
17ß-estradiol increased 17ß-[3H]estradiol binding
activity in a time- and dose-related manner (EC50 1
nM; maximal effect within 9–12 h), and increased estrogen
receptor-
and -ß mRNA. 8-Br-cGMP, a stable cell-permeant analog of
cGMP, could mimic the estrogen first phase increase in transepithelial
electrical conductance, but not the secondary increase. Treatment with
estrogen augmented the increase in transepithelial electrical
conductance that was induced by hydrostatic gradients (modulator of the
resistance of the lateral intercellular space), and the effect was
independent of woman’s age or baseline transepithelial electrical
conductance. In contrast, the effect of low extracellular calcium
(modulator of the tight junctional resistance) was more potent in cells
of premenopausal women than in cells of postmenopausal women and was
independent of treatment with estrogen. These results suggest that
changes in vaginal-cervical epithelial permeability after menopause are
determined by aging-related increase in tight junctional resistance,
and by low estrogen-dependent increase in lateral intercellular space
that lead to net increase in total paracellular resistance and
decreased permeability and result in reduced lubrication of the lower
genital canal.
THE MAIN FUNCTION of secretory epithelia, such as the vaginal-cervical epithelia, is to regulate movement of fluid and solutes from the blood into the lumen. The end result is generation of a fluid layer on the apical surface of the epithelium that lubricates the lumen. The vaginal-cervical fluid is a mixture of water, water-insoluble components, and water-soluble components (vaginal-cervical plasma). The vaginal-cervical plasma originates by transudation of fluid and solutes from the blood into the genital canal and constitutes 80–99% of the total weight of the vaginal fluid (1, 2). The driving force for transudation is blood pressure in the capillaries that generates a transepithelial hydrostatic gradient between capillaries and the vaginal-cervical canal. Stromal and smooth muscle cells do not significantly restrict the movement of fluid through the vagina and cervix. In contrast, epithelial vaginal-cervical cells do restrict fluid transudation through the intercellular space (3).
Vaginal dryness is a common symptom of postmenopausal women and is usually attributed to hypoestrogenic atrophy of the vaginal-cervical epithelia. Until recently, little was known about cellular mechanisms that lead to decreased lubrication of the lower genital tract in postmenopausal women. An important progress for the understanding of regulation of vaginal-cervical plasma was made by using new systems to culture human vaginal-cervical cells on filters (4, 5). These experimental systems enabled researchers in the field to fit experimental results into mathematical models, such as the Ussing-Zerahn model of fluid transepithelial transport, and to draw novel conclusions (6, 7).
Secretory epithelia, including the vaginal-cervical epithelia, are organized as a layer(s) of confluent cells, where plasma membranes of neighboring cells come into close contact and functionally occlude the intercellular space. Molecules can move across epithelia either through cells (transcellular route), or via the intercellular space (paracellular route). Movement in the transcellular route is restricted by plasma membranes, whereas movement via the paracellular route is determined by the resistance of the intercellular tight junctions, and by the resistance of the lateral intercellular space. Because the resistance of plasma membranes to passive movement of molecules is higher by about 105 than that of the paracellular route (8), the paracellular route determines the overall permeability properties of secretory epithelia. Previous studies confirmed that statement also in vaginal-cervical epithelia (3, 4, 5).
Women after menopause experience a significant decrease in
vaginal-cervical secretions. Traditionally, this phenomenon was
attributed to hypoestrogenism. Estrogens increase vaginal-cervical
secretions, including in postmenopausal women, but not all
postmenopausal women improve on estrogen (9, 10). These
observations raise the possibility that hypoestrogenism may not be the
only factor that contributes to decreased vaginal-cervical secretions
in postmenopausal women. Recent studies showed that short-term
treatment of normal human vaginal-ectocervical epithelial cells
in vitro increases paracellular permeability, but the
overall permeability across cultures from postmenopausal women remains
lower than across cultures from postmenopausal women (11, 12). In cells of premenopausal women the
estrogen-signaling pathway involves activation of estrogen receptor
(ER) , up-regulation of nitric oxide and cGMP, cGMP-dependent
activation of protein kinase, stimulation of ADP-ribosylation of
monomeric G-actin, and fragmentation of the cytoskeleton
(12, 13, 14, 15, 16, 17). Cells become more deformable and tend to
decrease their size more readily in response to stimuli that modulate
the cytoskeleton. Decreases in cell size cause reciprocal increase in
the volume of the intercellular space, and increase the paracellular
permeability.
This novel estrogen-signaling pathway can explain effects of estrogen on permeability, and on vaginal-cervical fluid accumulation 6–48 h after hyperestrogenemia or estrogen administration. For example, preovulatory vaginal and cervical secretions in the woman are apparent within hours after the preovulatory increase in plasma estradiol (1).
During the course of the above experiments we found that in addition to short-term effects, longer treatment with estrogen caused an additional increase in permeability. The main objective of the present study was to characterize the longer-term effect of estrogen on paracellular permeability in normal human vaginal-ectocervical epithelial cells of premenopausal and postmenopausal women. An additional objective was to determine what paracellular mechanisms are being activated by long-term estrogen treatment.
Materials and Methods
Patients and cell cultures
Primary to tertiary cultures of human ectocervical epithelial (hECE) cells were derived from histologically normal ectocervical tissues, using previously published methodology (5, 18, 19, 20). Cells outgrow from the peripheral rim of the ectocervix, grow and differentiate as multilayered squamous stratifying epithelium on filters in vitro, and retain phenotypical and biological characteristics of the native ectocervical and vaginal epithelium. This makes the filter-generated cultures a suitable experimental model for the study of transvaginal and transcervical transport phenomena.
Cells were cultivated from minces of ectocervix of premenopausal women (age, 35–45 yr) and from postmenopausal women (age, 55–65 yr). Women were grouped based on age and hormonal (estrogen) status: premenopausal women in the estrogenic phase of their menstrual cycle (as determined by histology of a proliferative endometrium); postmenopausal women not treated with hormone replacement (estrogen or progestin) for at least 1 yr before the study. Menopause was defined by the treating physicians as amenorrhea for 1 or more years, or amenorrhea of at least 6 months plus climacteric symptoms and plasma levels of FSH greater than 25 mIU/ml. All women were selected among those who underwent a hysterectomy by their treating physician for indications that were unrelated to the present study and had histologically normal cervix. Indications for hysterectomy included uterine fibroids, adnexal mass, abnormal uterine bleeding, and prolapsed uterus; the ectocervical tissues were defined as discarded tissues. The study was carried out according to institutional regulations and had been approved by the hospital’s Institutional Review Board. After their removal, ectocervical tissues were washed with cold and sterile saline to clear mucus, carefully dissected of stroma, and placed in cold saline. Ectocervical tissues were minced under the microscope to the size of 1 mm and plated epithelial face down on culture dishes as described (4, 18, 19).
Cell-culturing methods, including steroid-free conditions were described (11, 12).
Measurements of transepithelial electrical conductance (GTE)
Changes in paracellular permeability were determined as changes
in the GTE, across filters mounted vertically in
a modified Ussing chamber from successive measurements of the
transepithelial electrical current (
I, obtained by measuring the
current necessary to clamp the offset potential to zero, and normalized
to the 0.6 cm2 surface area of the filter) and
the PD (lumen negative), as GTE = I/PD
(6). The experimental design of the electrophysiological
measurements, including calibrations and controls, the significance of
the PD and I, and the conditions for optimal determinations of
GTE across low resistance epithelia
(e.g. hECE cells), were described and discussed (6, 11, 12, 21, 22, 23, 24, 25, 26).
17ß-[3H]estradiol binding assays in cell extracts were described (13, 18).
RNA methods were described (27). For RT-PCR experiments
the following oligonucleotide primers were used: human ER
(28), forward (sense) 5'-CAGGGGTGAAGTGGGGTCTGCTG3';
reverse (antisense) 5'-ATGCGGAACCGAGATGATGTAGC-3'; human ERß
(29), forward (sense) 5'-TGCTTTGGTTTGGGTGATTGC-3'; reverse
(antisense) 5'-TTTGCTTTTACTGTCCTCTGC-3'; human
glyceraldehyde-3-phosphate dehydrogenase (GAPDH; Ref. 26),
forward (sense) 5'-TGAAGGTCGGACTCAACGGATTTGGT-3'; reverse (antisense)
5'-GTGGTGGACCTCATGGCCCACATG-3'. To validate that RT-PCR can yield
semiquantitative estimates of changes in mRNA, the following steps were
taken: 1) the cDNAs of the ER, ERß, and the GAPDH were amplified
in parallel tubes, and changes in ER and ERß RNA were determined
relative to the changes in GAPDH RNA; 2) experiments using DNase-I
before RT-PCR were routinely run, to control for amplification of
genomic cDNA; 3) in control experiments GAPDH was amplified using
different doses of the GAPDH cDNA to determine the semiquantitative
changes in the amplification rates, as we (14, 26, 27) and
others (30) have described.
Generation of hydrostatic gradients
Aliquots of buffer were added to the subluminal compartment to establish hydrostatic gradients of 5–30 mm H2O in the subluminal to luminal direction (4, 7).
Determinations of free calcium
Levels of calcium in the extracellular buffer were manipulated using the calcium chelator EGTA. Concentrations of free calcium were calculated as described (25, 31).
Statistical analysis of the data
Data are presented as means (±SD), and significance of differences among means was estimated by ANOVA. Trends were calculated using GB-STAT V5.3 (Dynamic Microsystems Inc., Silver Spring, MD) and analyzed with ANOVA. Best fit of regression equations (least squares criterion) was achieved with SlideWrite Plus (Advanced Graphics Software, Inc., Carlsbad, CA), which uses the Levenberg-Marquardt Algorithm, and analyzed using ANOVA.
Chemicals and supplies
Anocell (Anocell-10) filters were obtained from Anotec (Oxon, UK). All other chemicals were obtained from Sigma (St. Louis, MO).
Results
Short-term effects of estrogen on GTE
The objective of the first experiment was to compare effects of short-term treatment with estrogen on paracellular permeability across hECE cultures of premenopausal and postmenopausal women. Experiments were carried out on cells grown in steroid-free, phenol red-deficient medium to prevent confounding estrogenic effects (11, 12, 32). The potent ER agonist in hECE cells, 17ß-estradiol (11, 12, 13), was added at concentrations that are in the physiological range for the woman.
Baseline levels of GTE across cultures of cells
from premenopausal women were higher than across cells of
postmenopausal women (46 ± 4 vs. 26 ± 4
mS·cm-2, respectively, Fig. 1
, P < 0.01), confirming
previous reports (11, 12, 13). Treatment with 17ß-estradiol
in vitro increased GTE in cells of
premenopausal and postmenopausal women, in a concentration-dependent
manner. In both groups of cells GTE increased
already with 0.1 nM, and saturation was achieved
with 10 nM 17ß-estradiol. The dose-response
curves of the means could be fitted into modified Hill equation with a
Hill coefficient n of 1.1 for both groups (Fig. 1), suggesting
interaction of estradiol with a single class binding sites. The
calculated EC50 of estradiol was 1.1 and 1.2
nM for cells of premenopausal and postmenopausal
women, respectively, indicating a similar potency of estradiol effect
in both groups of cells. These levels correspond with the reported kD
of the ER (33).
|
indicate that (1)
baseline GTE across steroid-deprived cells from
postmenopausal women is lower than across cells of premenopausal women
(2), that short-term treatment with 10 nM
17ß-estradiol increases GTE in both groups of
cells (3), and that the potency of estradiol effect was
similar in both groups of cells. Effects of longer-term treatment with estrogen on GTE
In cells of premenopausal women, treatment with estrogen for 24 h results in maximal increase in paracellular permeability (11). The objective of the second set of experiments was to compare effects of longer-term treatment with estrogen on paracellular permeability across hECE cultures of premenopausal and postmenopausal women.
Steroid-deprived cells from premenopausal women and from postmenopausal
women were grown on filters and treated with 10 nM
17ß-estradiol. The concentration of 10 nM was chosen
based on the results in Fig. 1. At successive time intervals after
treatment, GTE determinations were done under
sterile conditions. This allowed us to determine changes in
permeability over time across the same filter insert. In cells of
premenopausal women, treatment with 10 nM 17ß-estradiol
increased GTE from 69 ± 12 to 145 ± 7
mS·cm-2 (P < 0.01). The
increase in GTE was time dependent: it reached
plateau already after 12–24 h and persisted for at least 7 d
(Fig. 2, •). In cells of postmenopausal
women treatment with 10 nM 17ß-estradiol
increased GTE from 31 ± 3 to 94 ± 4
mS·cm-2 (Fig. 2, 
, P <
0.01), but the effect was biphasic: an initial increase in
GTE reaching plateau at about 55
mS·cm-2 within 12–24 h, followed by a
secondary time-dependent increase to 94
mS·cm-2 within 7 d of treatment (Fig. 2).
|
, because hECE cells cannot
be maintained on filters for more than 8–9 d. As is shown in Fig. 3A, in cells of premenopausal women the
estrogen increase in GTE remained the same from d
2 to at least d 14. In contrast, in cells of postmenopausal women
GTE continued to increase from d 2 through d 14:
after 2 d GTE increased from 34 ± 3 to
57 ± 11 mS·cm-2; after 7 d it
reached 89 ± 8 mS·cm-2; and after
14 d 112 ± 8 mS·cm-2 (Fig. 3B, trend significant at P < 0.01). We could not measure
changes in GTE across hECE cultures for longer
periods of time because after four to five subcultures normal hECE
cells stop proliferating.
|
and 3 indicate that short-term
treatment with estradiol is sufficient to produce maximal increase in
GTE across cultures of premenopausal women. In
contrast, in cultures of postmenopausal women estradiol induces a
biphasic increase in GTE, and the increase in
permeability may continue even past 2 wk of treatment in
vitro. Involvement of ERs
A possible explanation for the different responses to estrogen in
cells of premenopausal and postmenopausal women is involvement of
different estrogen-dependent signaling mechanism. In cells of
premenopausal women the effect of estrogen on permeability is mediated
by the ER (13). Tamoxifen blocks estrogen increase in
permeability both in cells of premenopausal (11) and
postmenopausal women (submitted for publication), suggesting
involvement of ER also in cells of postmenopausal women. The
objective of the third set of experiments was to determine the degree
to which the effect of estrogen on permeability in cells of
postmenopausal women is mediated by ERs.
Binding of 17ß-[3H]estradiol to lysates of hECE
cells. hECE cells obtained from premenopausal or from
postmenopausal women were treated with 10 nM
17ß-estradiol for 2 d, and binding of
17ß-[3H]estradiol to total cells extracts was
measured. Estradiol binding was saturable, and Scatchard analysis
revealed a single class of binding sites with Kd
of 1.3 nM for cells of premenopausal women and 1.2
nM for cells of postmenopausal women; the binding activity,
respectively, was 1.2 and 1.0 pmol/mg DNA (Fig. 4A). These levels are similar to values
reported in the cervix in vivo (34, 35, 36, 37), and
the results indicate that in extracts of hECE cells from premenopausal
and postmenopausal women estradiol binds in a saturable manner to
high-affinity, low-capacity sites.
|
, B and C). In cells of premenopausal women
treatment with estradiol increased
17ß-[3H]estradiol binding activity from
0.5 ± 0.1 to 1.3 ± 0.2 pmol/mg DNA. The effect of estradiol
required 1–3 h of treatment with the hormone and was maximal after
9–12 h (Fig. 4B, •). In cells of postmenopausal women treatment with
estradiol increased 17ß-[3H]estradiol binding
activity from 0.2 ± 0.1 to 0.9 ± 0.3 pmol/mg DNA
(P < 0.01). The effect required 1–3 h of treatment
with the hormone and was also maximal after 9–12 h (Fig. 4B, ). In
both types of cells binding of
17ß-[3H]estradiol remained stable for at
least 48 h (Fig. 4B).
In both types of cells the increase in
17ß-[3H]estradiol binding activity began with
0.1 nM 17ß-estradiol and reached saturation at 10
nM, with EC50 of estradiol of about 1
nM (Fig. 4C). These results indicate that estradiol
up-regulates 17ß-[3H]estradiol binding
activity with similar potency in cells of both premenopausal and
postmenopausal women, and that the effect is maximal within 9–12
h.
Estrogen increases ER and ERß mRNA. Using oligonucleotide
primers complementary to cloned human ER and human ERß, single
cDNA fragments of 483 bp and 283 bp, respectively, were amplified by
RT-PCR from lysates of hECE cells obtained from premenopausal and
postmenopausal women (Fig. 5). These cDNA
fragments were isolated, amplified, and purified, and the products were
sequenced by the dideoxy chain termination method. Sequence analysis of
the cloned segments revealed homologies of 99% (sense and antisense)
with the human ER and ERß (the differences were sequence errors;
data not shown). Treatment with estrogen had no effect on GAPDH mRNA,
but it increased ER and ERß mRNA. Densitometry results of three
experiments revealed that treatment with 10 nM
17ß-estradiol increased ER/GAPDH mRNA 5- and 12-fold, and
ERß/GAPDH mRNA 3- and 2-fold, respectively, in cells of premenopausal
and postmenopausal women (Fig. 5). These results indicate that hECE
cells of premenopausal and postmenopausal women express mRNA for ER
and ERß and that treatment with estrogen up-regulates both receptor
isoforms.
|
and 5 indicate that hECE
cells of postmenopausal women express smaller amounts of ERs than cells
of premenopausal women, but in both groups up-regulation of the
receptor(s) by estradiol has similar time course and potency
profile. 8-Br-cGMP effects on GTE
In hECE cells of premenopausal women the estrogen increase in
paracellular permeability is mediated by cGMP-dependent fragmentation
of the cytoskeleton (13, 14, 15, 16, 17). A possible explanation for
the different responses to estrogen in cells of premenopausal and
postmenopausal women (Figs. 2 and 3) is that after menopause
cGMP-dependent fragmentation of the cytoskeleton is not efficacious.
The fourth experiment tested this speculation by measuring the combined
effect of estradiol and 8-Br-cGMP on GTE.
8-Br-cGMP is a stable cell-permeant analog of cGMP, and it can mimic
cGMP-dependent intracellular actions. 8-Br-cGMP was used at a
concentration of 25 µM, which produces maximal increase
in GTE across hECE cells of premenopausal women
(14, 15).
In cells of premenopausal women 8-Br-cGMP increased
GTE from 88 ± 9
mS·cm-2 to 129 ± 12
mS·cm-2 (Fig. 6A, P < 0.01). Treatment
with estrogen for 2 d or for 14 d increased
GTE to about 140 mS·cm-2
(Fig. 6A, P < 0.01), similar to the result shown in
Fig. 2. When administered alone, the effect of 8-Br-cGMP did not differ
from that of estrogen (Fig. 6A). Coadministration of 8-Br-cGMP to
estrogen-treated cells had no additional effect on
GTE to that of estradiol alone (Fig. 6A),
confirming previous reports (14, 15, 16). These results
indicate that 8-Br-cGMP can mimic the estrogen increase in
GTE in cells of premenopausal women, but it
cannot increase GTE above levels that are induced
by estradiol.
|
, P < 0.01).
However, the increase in GTE was smaller than
that obtained by estradiol after 14 d of treatment (98 ± 12
mS·cm-2, P < 0.05, Fig. 6B).
Coadministration of 8-Br-cGMP to cells treated with estrogen for 2
d or for 14 d had no additional effect on
GTE (Fig. 6B). These results in cells of
postmenopausal women indicate that 8-Br-cGMP can mimic the first phase
of estrogen-increase in GTE, but not the later,
second phase of increase in permeability. Modulation of tight junctional resistance (RTJ) and resistance of the lateral intercellular space (RLIS)
The objective of the fifth set of experiments was to understand how aging and estrogen affect paracellular mechanisms that control hECE permeability. In epithelia, including cultured hECE cells, the paracellular permeability is determined by the RTJ and by the RLIS, in series (8, 21, 22). In hECE cells of premenopausal women estrogen increases paracellular permeability mainly by decreasing the RLIS (11). After menopause hECE cell permeability decreases due to estrogen deficiency-dependent increase in RLIS (12), but some of the decrease in permeability could be attributed also to aging-dependent increase in RTJ (Ref. 12 , and submitted for publication). Given the different time-dependent effects of estrogen on GTE across hECE cultures of premenopausal and postmenopausal women, the objective was to determine the degree to which 2 or 14 d of treatments with estrogen modulate RLIS and RTJ.
To determine effects on the RLIS, cells were
exposed to hydrostatic gradients in the subluminal to luminal
direction. This condition physically opens the intercellular space; it
dilates the lateral intercellular space and, therefore, decreases the
RLIS (4, 7, 38, 39). In the present
study, hydrostatic gradients increased GTE in
cultures of cells from premenopausal and postmenopausal women,
regardless if cells were treated with estradiol or the vehicle (Fig. 7). In cells incubated in steroid-free
medium, a hydrostatic gradient of 30 mm H2O
increased GTE by about 40
mS·cm-2, and the effect was similar in cells
from premenopausal and from postmenopausal women (Fig. 7, circles). In cells from premenopausal treated with estradiol
for 2 or for 14 d a hydrostatic gradient of 30 mm
H2O increased GTE by about
100 mS·cm-2 (Fig. 7A, triangles and
diamonds). In cells from postmenopausal treated with
estradiol for 2 d a hydrostatic gradient of 30 mm
H2O increased GTE by 68
mS·cm-2 (Fig. 7B, triangles).
However, in cells from postmenopausal treated with estradiol for
14 d a hydrostatic gradient of 30 mm H2O
increased GTE by 80
mS·cm-2 (Fig. 7B, triangles).
|
depicts correlation analysis of
baseline levels of GTE, and changes in
GTE in response to hydrostatic gradients in cells
of premenopausal women (•) and postmenopausal women (). Means of
the three treatment categories could be fitted into straight lines:
r2 = 0.90, P < 0.05 for cells of
premenopausal women, and r2 = 0.98,
P < 0.01 for cells of postmenopausal women, with
similar linear regression coefficients, 0.86 and 0.68, respectively.
These results confirm that treatment with estrogen in vitro
increases baseline GTE and indicate that
treatment with estrogen augments the increase in
GTE in response to hydrostatic gradients. These
results also suggest that treatment with estrogen, rather than younger
age per se, augments the increase in
GTE in response to hydrostatic gradients.
|
Lowering extracellular calcium increased GTE
regardless of menopause status or treatment with estrogen (Fig. 9). A significantly lower level of
extracellular calcium was necessary to decrease
GTE to 600 mS·cm-2 or
greater in cells of postmenopausal women than in cells of premenopausal
women (0.5 mM vs. 0.8 mM,
Fig. 8, P < 0.01). Furthermore, the differences in
extracellular calcium requirements between the two types of cells to
decrease GTE to 600
mS·cm-2 or greater were independent of
treatment with estradiol (Fig. 8). These results suggest that aging,
rather than estrogen, determines the degree of low extracellular
calcium that is necessary to disrupt the RTJ.
|
The present results show that GTE levels across cultured hECE cells from postmenopausal women were lower than across cultures from premenopausal women. Treatment of cells in vitro with estrogen increased the permeability in both groups of cells, but the responses to estrogen differed between the two groups of cells. In cultures from premenopausal women, GTE reached its highest level already 12–24 h after estrogen treatment. In contrast, the effect in cultures from postmenopausal women was biphasic: an initial increase that reached plateau after 48 h, followed by an additional slower increase that continued for the duration of treatment for at least 2 wk.
Three types of experiments were done to provide a better understanding
for the different time course of estrogen effects on paracellular
permeability. One group of experiments asked whether the differences
could be explained in terms of differences in pharmacokinetic and
pharmacodynamic properties of estrogen action. The present results
refute this speculation. The experiments studied expression and
regulation of the ER mechanism(s) in hECE cells of premenopausal and
postmenopausal women. Both types of cells express high-affinity,
low-capacity, estrogen-dependent estradiol-binding proteins, as well as
estrogen-dependent ER and ERß. Binding activity of
17ß-[3H]estradiol, and ER and ERß mRNA
levels were lower in cells from postmenopausal women than in cells from
premenopausal women. However, in both types of cells, treatment with
17ß-estradiol increased significantly
17ß-[3H]estradiol binding activity and mRNA
levels of ER and ERß. EC50 of estradiol for
up-regulation of 17ß-[3H]estradiol binding
activity was about 1 nM, and the time course of the
increase was similar, t1/2 of about 6–9 h.
Treatment with 17ß-estradiol also increased mRNA levels of ER and
ERß, although the effect on ER mRNA was greater than on ERß
mRNA. A possible explanation for the latter finding is that in human
cervical epithelial cells ER is the dominant estrogen-dependent ER
isoform, as we have previously suggested (13). The
increase in ER mRNA in cells of postmenopausal women was relatively
greater than in cells of premenopausal women. A possible explanation is
that continuous exposure to estrogen (e.g. in cells of
premenopausal women) decreases their sensitivity to additional
treatment with the hormone. However, these results need further
exploration at the receptor(s) protein level to validate that
speculation.
Collectively, the results of the first set of experiments can be
interpreted as activation of the classic nuclear estrogen (ER)
receptor mechanism (42). Based on these results, as well
as of data that estradiol does not have an acute effect on
GTE (unpublished results), we propose that the
estradiol increase in GTE in hECE cells of
postmenopausal women is mediated by the ER mechanism. That conclusion
may have mechanistic importance for our understanding of estrogen
regulation of permeability. In hECE cells of postmenopausal women the
time course of estradiol increase in
17ß-[3H]estradiol binding activity correlated
with the first phase of the estradiol increase in
GTE, but not with the subsequent, late increase
in permeability. In cells of premenopausal women the time course of
estradiol increase in 17ß-[3H]estradiol
binding activity correlates with estrogen signaling (present results,
and Refs. 11, 12, 13). In those cells, short-term treatment
with estrogen increases the permeability by decreasing the
RLIS. The effect is mediated by ER
up-regulation of nitric oxide and cGMP; cGMP-activation of
cGMP-dependent protein kinase stimulates ADP-ribosylation of monomeric
G-actin and fragmentation of the cytoskeleton (14, 15, 16, 17).
Cells become more deformable and tend to decrease their size more
readily in response to stimuli that modulate the cytoskeleton.
Decreases in cell size cause reciprocal increase in the volume of the
intercellular space, a decrease in RLIS, and an
increase in paracellular permeability.
The above signaling cascade can explain the initial phase of increase in permeability in hECE cells of postmenopausal women, but not the late phase increase in GTE. That conclusion is also supported by the results of the second group of experiments, namely with effects of 8-Br-cGMP on permeability. Treatment of cells from postmenopausal women with the cell permeant cGMP analog 8-Br-cGMP increased GTE to the same degree as the estrogen-induced first phase increase in permeability, but failed to mimic the estrogen-dependent long-term-increase in GTE. This result, and our previous studies (13, 14, 15, 16, 17), suggests that in cells of postmenopausal women cGMP mediates the estrogen-induced first phase increase in permeability, but not the estrogen-induced late phase increase in GTE.
The third group of experiments studied involvement of RTJ and RLIS in the responses to estrogen. Epithelial paracellular permeability is determined by RTJ and RLIS in series (8, 21, 22). The decrease in vaginal-cervical permeability after menopause can be explained by changes in RTJ and RLIS (11, 12, 13). In hECE cells of premenopausal women RTJ contributes about 84% to the paracellular resistance, in contrast to 97% in cells of postmenopausal (12). RTJ is a high-resistive element that restricts movement of fluid and solutes to a greater degree than the low-resistive RLIS (8, 21, 22), and an increase after menopause in RTJ can decrease significantly the permeability. In cells of premenopausal women estrogen increases paracellular permeability by decreasing the RLIS, whereas hypoestrogenism causes a reciprocal increase in RLIS (11, 12, 13). Based on these findings, it is anticipated that after menopause, as a result of low-circulating estrogens, RLIS will increase. Therefore, the combined increases in RTJ and RLIS result in net increase in total paracellular resistance and decrease the permeability.
The results of the present study support this model. Treatment with estrogen augmented the increase in GTE that was induced by hydrostatic gradients. In those experiments we used hydrostatic gradients in the subluminal to luminal direction that are similar in magnitude to capillary pressure in vivo (43, 44). The effect of estrogen was observed both in cells of premenopausal women, as well as in cells of postmenopausal women, and was independent of woman’s age or the degree of baseline GTE, confirming that estrogen decreases the RLIS. In contrast to estrogen, older age was associated with a greater increase in RTJ. Using low extracellular calcium as modulator of RTJ (25), calcium in the extracellular medium had to be lowered more in cells of postmenopausal women than in cells of premenopausal women. Furthermore, the effect was independent of treatment with estradiol, suggesting that aging, rather than estrogen, increases the RTJ.
At present, relatively little it is known about what molecular mechanisms regulate the changes in RLIS and in RTJ in human vaginal-cervical epithelial cells of postmenopausal women. ADP-ribosylation of monomeric G-actin is a proposed cellular mechanism for the first-phase estrogen increase in permeability in cells of premenopausal women (14, 15, 16, 17). Less is known about signaling distal to the ERs and the involvement of paracellular mechanisms in the estrogen late-phase response in cells of postmenopausal women. Little is also known about mechanisms that regulate changes in RTJ in human vaginal-cervical epithelial cells. The tight junctional apparatus is a complex of strands and grooves formed by plasma membranes of neighboring cells. It involves proteins that face the extracellular space, as well as intracellular proteins that are associated with cytoskeletal proteins (45, 46, 47, 48, 49). We are currently studying what aging-related mechanisms may be involved in the increase in RTJ in hECE cells of postmenopausal women.
Based on the present results, and on our previous studies in the field, we propose that changes in vaginal-cervical epithelial permeability after menopause are the result of aging-related increase in RTJ and of low estrogen-dependent increase in RLIS. Increased RTJ and RLIS lead to a net increase in total paracellular resistance and decreased permeability and result in reduced lubrication of the lower genital canal. We suggest that estrogen decreases RLIS by two mechanisms that are mediated by ERs but involve different signaling pathways distal to the ERs: short-term treatment with estrogen for 24–48 h involves nitric oxide/cGMP-dependent activation of ADP-ribosylation of monomeric G-actin. The mechanism by which longer-term treatment with estrogen decreases RLIS is not clear but seems not to involve cGMP. In view of the longer time that is required for second phase effect, it is possible that it involves changes in cell structure, including cytoskeleton and plasma membranes. The two estrogen-related mechanisms seem to be fundamentally different, based on the different time course of estrogen effects, but they may be related, because changes in cell structure (i.e. estrogen second phase) can augment estrogen-induced fragmentation of the cytoskeleton (i.e. estrogen early phase). This hypothesis may explain the finding that prolonged treatment with estrogen of cells of premenopausal women produced only monophasic increase in permeability. It is possible that cells derived from euestrogenic (premenopausal) women retain estrogen-dependent cell structure.
Elucidating the signaling and molecular mechanisms of estrogen late-phase regulation of permeability may be important for targeting interventions to clinically affect vaginal-cervical fluid secretion in postmenopausal women.
Acknowledgments
The technical support of Kim Frieden, Brian De-Santis, and Dipika Pal is acknowledged.
Footnotes
This study was supported by NIH Grants HD00977, HD29924, and AG15955.
Abbreviations: ER, Estrogen receptor; hECE, human ectocervical epithelial; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GTE, transepithelial electrical conductance; RLIS, resistance of the lateral intercellular space; RTJ, tight junctional resistance.
Received March 8, 2001.
Accepted May 4, 2001.
References
and ß in the regulation of cervical permeability.
Am J Physiol 278:C689–C696
and ß. Endocrinology 138:863–869This article has been cited by other articles:
 |
|
 |
|
  B. C. Timmons, S. M. Mitchell, C. Gilpin, and M. S. Mahendroo Dynamic Changes in the Cervical Epithelial Tight Junction Complex and Differentiation Occur during Cervical Ripening and Parturition Endocrinology, March 1, 2007; 148(3): 1278 - 1287. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. I. Gorodeski Estrogen Decrease in Tight Junctional Resistance Involves Matrix-Metalloproteinase-7-Mediated Remodeling of Occludin Endocrinology, January 1, 2007; 148(1): 218 - 231. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X. Li, L. Zhou, and G. I. Gorodeski Estrogen Regulates Epithelial Cell Deformability by Modulation of Cortical Actomyosin through Phosphorylation of Nonmuscle Myosin Heavy-Chain II-B Filaments Endocrinology, November 1, 2006; 147(11): 5236 - 5248. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Zhu, X. Li, R. Zeng, and G. I. Gorodeski Changes in Tight Junctional Resistance of the Cervical Epithelium Are Associated with Modulation of Content and Phosphorylation of Occludin 65-Kilodalton and 50-Kilodalton Forms Endocrinology, February 1, 2006; 147(2): 977 - 989. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. I. Gorodeski Aging and Estrogen Effects on Transcervical-Transvaginal Epithelial Permeability J. Clin. Endocrinol. Metab., January 1, 2005; 90(1): 345 - 351. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Zeng, X. Li, and G. I. Gorodeski Estrogen Abrogates Transcervical Tight Junctional Resistance by Acceleration of Occludin Modulation J. Clin. Endocrinol. Metab., October 1, 2004; 89(10): 5145 - 5155. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||
