The research presented here demonstrates that the secretion of CCL20/MIP3α and TNF-α by rat uterine epithelial cells is influenced by estradiol. We show that E2 significantly inhibits the constitutive release of both CCL20/MIP3α and TNF-α. In contrast, when both E2 and PAMP are present, E2 increases CCL20/MIP3α production beyond that seen with PAMP stimulation alone. These studies indicate that E2 has an inhibitory effect on the basolateral release of TNF-α in the presence as well as the absence of PAMP. Moreover, this work demonstrates that ICI 182,780, which has little or no effect on reversing the inhibitory action of E2 treatment on the constitutive release of TNF-α or CCL20/MIP3α, reverses the stimulatory effect of E2 on CCL20/MIP3α release in response to LPS and Pam3Cys as well as partially reverses the inhibitory effect of E2 in the presence of LPS.
Production of CCL20/MIP3α at other mucosal surfaces, including the lungs and gastrointestinal tract, is enhanced by pathogenic challenge (36
). Previously, we demonstrated that polarized uterine epithelial cells in culture produce CCL20/MIP3α in response to live and heat-killed E. coli
as well as selected PAMP (8
). To the best of our knowledge, our finding that estradiol exerts a stimulatory effect on CCL20/MIP3α is the first demonstration that this chemokine is under hormonal control. Moreover, these studies demonstrate a unique role of estradiol, in that constitutive release of CCL20/MIP3α is inhibited by estradiol, whereas the release of CCL20/MIP3α in response to PAMP is stimulated by estradiol. That release of CCL20/MIP3α in response to PAMP is ER mediated is suggested by our finding that when added along with estradiol, ICI 182,780, a receptor antagonist of estradiol, reverses the stimulatory effects of estradiol on CCL20/MIP3α secretion. In contrast, when estradiol and ICI 182,780 were added to cell cultures releasing CCL20/MIP3α under constitutive release conditions, the effect of estradiol was not reversed.
The complexities of regulation observed in these studies suggest that estradiol acts though different pathways to differentially control CCL20/MIP3α and TNF-α production. This may explain the apparent contradiction of our findings with those of others who have shown that TNF-α plays an important role in stimulating CCL20/MIP3α production (14
). For example, estradiol is able to bind to different isoforms of the same receptor and may either increase or decrease mRNA expression, the net result being either enhanced or suppressed cytokine secretion. Alternatively, since cytokines such as TNF-α are known to exist as procytokines, estradiol may act on matrix metalloproteinase to either enhance or suppress processing and/or release from epithelial membranes (15
Our findings that the constitutive release of TNF-α and CCL20/MIP3α is inhibited by E2
suggest that in the absence of potential pathogens, E2
, produced during the reproductive cycle, acts in the uterus to suppress these two proinflammatory molecules. Unexpected was our finding that under conditions of PAMP stimulation, E2
increased CCL20/MIP3α release while continuing to inhibit TNF-α secretion. Estradiol levels in blood vary with the stage of the reproductive cycle, with highest levels measured just prior to ovulation and mating (50
). Our finding that CCL20/MIP3α is differentially regulated by estradiol suggests that CCL20/MIP3α production in the uterus may vary with the stage of the cycle. If transferable to in vivo situations, CCL20/MIP3α production may be inhibited in the uterus by estradiol if mating does not occur, while under conditions of mating and the presence of bacteria, estradiol may stimulate the release of CCL20/MIP3α, thus affording protection against infection. The observed influx of leukocytes that occurs in the uterus following mating (46
) may in part be attributable to putative increases in CCL20/MIP3α. Since an additional dimension of CCL20/MIP3α immune protection is its ability to act as a microbicide (22
), elevated levels of CCL20/MIP3α stimulated by estradiol at mating might protect by limiting bacterial growth until such time that recruitment of immune cells can occur.
Estradiol influences on CCL20/MIP3α and TNF-α are complex and may be associated with genomic and/or nongenomic effects of E2
. Estradiol is known to influence genes when the liganded ERα binds to an Sp1 promoter site (12
). In other studies, it has been demonstrated that the promoter region of the CCL20/MIP3α gene contains an Sp1 promoter sequence (35
). This offers an explanation for a mechanism whereby estradiol could influence the transcription of CCL20/MIP3α. In contrast to the stimulatory role of estrogen on the transcription of CCL20/MIP3α, the TNF-α promoter contains an estrogen inhibitory element (2
). Nongenomic effects of E2
in the endometrium have also been reported. For example, treatment of endometrial cells in culture with E2
results in Ca2+
influx within 10 min, and treatment of ovariectomized rats with E2
results in changes in the morphology of uterine epithelial cells within 1 min (13
). Some of the nongenomic effects of E2
are reported not to be antagonized by the ER antagonist ICI 182,780 (13
). For example, in a study of neurite growth, the effects of E2
could be inhibited by an inhibitor of cyclic AMP/PKA and Ca2+
signaling pathways but not by ICI 182,780. This and other studies have shown that ICI 182,780 antagonism of the effects of E2
is complex. Further studies are needed to define the mechanism(s) whereby E2
exerts its effects on TNF-α and CCL20/MIP3α production by uterine epithelial cells.
In conclusion, estradiol influences the production of CCL20/MIP3α and TNF-α by uterine epithelial cells both constitutively and in response to PAMP. Estradiol regulation of the release of CCL20/MIP3α and TNF-α suggests that the mucosal immune cell response to the presence of bacteria in the reproductive tract is precisely regulated and coordinated with the reproductive cycle to optimize the potential for successful mammalian reproduction.