Progesterone is essential for protecting against LPS-induced pregnancy loss. LIF as a potential mediator of the anti-inflammatory effect of progesterone

Abstract

Lipopolysaccharide (LPS) administration to mice on day 7 of gestation led to 100% embryonic resorption after 24 h. In this model, nitric oxide is fundamental for the resorption process. Progesterone may be responsible, at least in part, for a Th2 switch in the feto-maternal interface, inducing active immune tolerance against fetal antigens. Th2 cells promote the development of T cells, producing leukemia inhibitory factor (LIF), which seems to be important due to its immunomodulatory action during early pregnancy. Our aim was to evaluate the involvement of progesterone in the mechanism of LPS-induced embryonic resorption, and whether LIF can mediate hormonal action. Using in vivo and in vitro models, we provide evidence that circulating progesterone is an important component of the process by which infection causes embryonic resorption in mice. Also, LIF seems to be a mediator of the progesterone effect under inflammatory conditions. We found that serum progesterone fell to very low levels after 24 h of LPS exposure. Moreover, progesterone supplementation prevented embryonic resorption and LPS-induced increase of uterine nitric oxide levels in vivo. Results show that LPS diminished the expression of the nuclear progesterone receptor in the uterus after 6 and 12 h of treatment. We investigated the expression of LIF in uterine tissue from pregnant mice and found that progesterone up-regulates LIF mRNA expression in vitro. We observed that LIF was able to modulate the levels of nitric oxide induced by LPS in vitro, suggesting that it could be a potential mediator of the inflammatory action of progesterone. Our observations support the view that progesterone plays a critical role in a successful pregnancy as an anti-inflammatory agent, and that it could have possible therapeutic applications in the prevention of early reproductive failure associated with inflammatory disorders.

Figure 1. Effect of LPS treatment on serum progesterone levels.

Mice were injected on day 7 of gestation and sacrificed at 6, 12 and 24 h after LPS (1 µg/g) or vehicle (PBS) administration. Progesterone was evaluated in serum samples by radioimmunoassay. Progesterone levels are expressed as ng/ml serum. n = 6–9. Values are means ± SEM. Bars with different superscript letters denote significant differences (P<0.05). a≠b≠c≠d.

Figure 2. Effect of LPS and progesterone treatment on uterine nitric oxide levels.

Mice were injected on day 7 of gestation with LPS (1 µg/g) or LPS and progesterone (2 mg/mouse, 2 h before LPS injection). They were sacrificed 6 h after LPS treatment and uterine tissue was cultured for 24 h. Nitric oxide was measured as nitrate (NO₃ ⁻) plus nitrite (NO₂ ⁻) accumulation in culture supernatants by Griess technique. n = 5. Values are means ± SEM. Results are expressed as µM NO₂ ⁻/mg tissue. Bars with different superscript letters denote significant differences (P<0.05). a≠b.

Figure 3. Effect of LPS treatment on progesterone receptor (PR) protein levels in uterus.

Mice were injected on day 7 of gestation with vehicle (PBS), LPS (1 µg/g), progesterone (2 mg/mouse) or LPS+progesterone and sacrificed at 6 and 12 after treatment. PRA and PRB protein expression were evaluated in uterine samples by western blot. (A) PRA 6 h, (B) PRA 12 h and (C) PRB 6 h. PR levels are expressed as relative optical density (PR/Actin). n = 6–5. Values are means ± SEM. Bars with different superscript letters denote significant differences (P<0.05). a≠b.

Figure 4. Immunolocalization of PR (A) and LIF (B) in uteri from day 7 of pregnancy in mice.

Tissue sections were processed by the immunoperoxidase technique using specific antibodies. (A) Negative controls for LIF, (B) Immunostaining for LIF, (C) Negative controls for PR and (D) Immunostaining for PR. Arrows indicate glands. The scale bar indicates 10 µm. Magnifications: x40 and x100.

Figure 5. Effect of progesterone on uterine LIF expression.

Mice were sacrificed on day 7 of gestation and uterine tissue was cultured for 2 h with and without LPS (1 µg/ml) and progesterone (50 ng/ml). LIF mRNA levels was determined by RT-PCR. n = 5. Values are means ± SEM. Results are expressed as relative optical density (LIF/Actin). Bars with different superscript letters denote significant differences (P<0.05). a≠b.

Figure 6. Effect of LPS and exogenous LIF on uterine nitric oxide levels.

Mice were sacrificed on day 7 of gestation and uterine tissue was cultured for 24 h. Assays were performed using LPS (1 µg/ml) and murine recombinant LIF (50 and 100 ng/ml). Nitric oxide was measured as the accumulation of nitrate (NO₃ ⁻) and nitrite (NO₂ ⁻) in culture supernatants by Griess technique. n = 6. Values are means ± SEM. Results were expressed as µM NO₂ ⁻/mg tissue. Bars with different superscript letters denote significant differences (P<0.05). a≠b≠c.

Figure 7. Effect of endogenous LIF and progesterone on uterine nitric oxide levels.

Mice were sacrificed on day 7 of gestation and uterine tissue was cultured for 24 h. Assays were performed using progesterone (50 ng/ml), LPS (1 µg/ml) and an antibody raised against murine LIF (anti-LIF: 5 µg/ml). Nitric oxide was measured as the accumulation of nitrate (NO₃ ⁻) and nitrite (NO₂ ⁻) in culture supernatants by Griess technique. n = 5. Values are means ± SEM. Results were expressed as µM NO₂ ⁻/mg tissue. Bars with different superscript letters denote significant differences (P<0.05). a≠b≠c.

Figure 8. Effect of RU-486 on uterine nitric oxide levels.

Mice were sacrificed on day 7 of gestation and uterine tissue was cultured for 24 h. Assays were performed using progesterone (50 ng/ml), LPS (1 µg/ml) and RU-486 (0.1 µM). Nitric oxide was measured as the accumulation of nitrate (NO₃ ⁻) and nitrite (NO₂ ⁻) in culture supernatants by Griess technique. n = 11. Values are means ± SEM. Results were expressed as µM NO₂ ⁻/mg tissue. Bars with different superscript letters denote significant differences (P<0.05). a≠b≠c.