Estradiol and progesterone strongly inhibit the innate immune response of mononuclear cells in newborns

Abstract

Newborns are particularly susceptible to bacterial infections due to qualitative and quantitative deficiencies of the neonatal innate immune system. However, the mechanisms underlying these deficiencies are poorly understood. Given that fetuses are exposed to high concentrations of estradiol and progesterone during gestation and at time of delivery, we analyzed the effects of these hormones on the response of neonatal innate immune cells to endotoxin, bacterial lipopeptide, and Escherichia coli and group B Streptococcus, the two most common causes of early-onset neonatal sepsis. Here we show that at concentrations present in umbilical cord blood, estradiol and progesterone are as powerful as hydrocortisone for inhibition of cytokine production by cord blood mononuclear cells (CBMCs) and newborn monocytes. Interestingly, CBMCs and newborn monocytes are more sensitive to the effects of estradiol and progesterone than adult peripheral blood mononuclear cells and monocytes. This increased sensitivity is associated with higher expression levels of estrogen and membrane progesterone receptors but is independent of a downregulation of Toll-like receptor 2 (TLR2), TLR4, and myeloid differentiation primary response gene 88 in newborn cells. Estradiol and progesterone mediate their anti-inflammatory activity through inhibition of the NF-κB pathway but not the mitogen-activated protein kinase pathway in CBMCs. Altogether, these results suggest that elevated umbilical cord blood concentrations of estradiol and progesterone acting on mononuclear cells expressing high levels of steroid receptors contribute to impair innate immune responses in newborns. Therefore, intrauterine exposure to estradiol and progesterone may participate in increasing susceptibility to infection during the neonatal period.

Fig. 1.

Estradiol and progesterone inhibit TNF and IL-6 production by CBMCs. CBMCs were incubated for 16 h with estradiol (10⁻⁹ to 10⁻⁷ M) (A and B) or progesterone (10⁻⁷ to 10⁻⁵ M) (C and D) prior to stimulation with E. coli (black bars), GBS (gray bars), LPS (white bars), or Pam₃CSK₄ (hatched bars). TNF (A and C) and IL-6 (B and D) concentrations in cell culture supernatants were measured by bioassay. Data represent means ± SEMs of 8 independent experiments performed in triplicate. *, P < 0.05 versus no steroids.

Fig. 2.

Estradiol and progesterone inhibit TNF and IL-6 production by CBMCs as efficiently as hydrocortisone but have no additive inhibitory effect. CBMCs were incubated for 16 h with hydrocortisone (H; 10⁻⁹ to 10⁻⁷ M), estradiol (E; 10⁻⁹ to 10⁻⁷ M), or progesterone (P; 10⁻⁷ to 10⁻⁵ M) either alone (A and B) or in combination (C and D) prior to stimulation with E. coli. TNF (A and C) and IL-6 (B and D) concentrations in cell culture supernatants were measured by bioassay. Data represent means ± SEMs of 5 to 6 independent experiments performed in triplicate. *, P < 0.05 versus no steroids.

Fig. 3.

CBMCs and newborn monocytes are more susceptible to the anti-inflammatory effect of estradiol and progesterone than adult PBMCs and monocytes. (A and B) PBMCs from adult males (black squares) and premenopausal females (black inverted triangles) and CBMCs (gray circles) were incubated for 16 h with estradiol (A) or progesterone (B) prior to stimulation with E. coli. TNF concentrations in cell culture supernatants were measured by bioassay. Results were expressed in percent TNF release versus that for cells stimulated without steroids. Data represent means ± SEMs of 5 to 6 independent experiments performed in triplicate. *, P < 0.05 between PBMCs from adult males and CBMCs; †, P < 0.05 between PBMCs from adult females and CBMCs. (C) Monocytes from newborns were preincubated for 16 h with hydrocortisone, estradiol, or progesterone prior to stimulation with E. coli. Data represent means ± SEMs of 5 to 6 independent experiments performed in triplicate. *, P < 0.05 versus no steroids. (D and E) Monocytes from adult males (black squares), premenopausal females (black triangles), and newborns (gray circles) were incubated for 16 h with estradiol or progesterone prior to stimulation with E. coli. TNF concentrations in cell culture supernatants were measured by bioassay. Data represent means ± SEMs of 5 to 6 independent experiments performed in triplicate. *, P < 0.05 between monocytes from adult males and newborns; †, P < 0.05 between monocytes from adult females and newborns.

Fig. 4.

CBMCs express higher levels of estrogen and progesterone receptors than adult PBMCs and respond to specific receptor agonists. (A) Real-time PCR analysis of ER-α, ER-β, PR, PGRMC1, mPR-α, mPR-β, and mPR-γ mRNA levels in PBMCs from adult males (AD) and CBMCs (NB) from males. Gene-specific expression was normalized to the expression of HPRT. Data (means ± SEMs; n = 5 to 6) are expressed in A.U. relative to expression of the corresponding gene in MCF-7 breast cancer cells used as a positive control. *, P < 0.05 versus adult PBMCs. (B and C) Western blot (B) and densitometric (C) analyses of ER-α and ER-β in adult PBMCs (AD) and CBMCs (NB). Data (means ± SEMs; n = 6) are expressed in A.U. relative to expression of the corresponding protein in MCF-7 cells. (D) TNF concentrations in cell culture supernatants of CBMCs incubated with estradiol, progesterone, or agonists specific for ER-α (PPT), ER-β (DPN), and PR (norgestrel) for 16 h prior to stimulation with E. coli. Data represent means ± SEMs of 6 independent experiments performed in triplicate. *, P < 0.05 versus no steroids.

Fig. 5.

The anti-inflammatory effect of estradiol and progesterone on CBMCs does not depend on modulation of TLR2, TLR4, and MYD88 expression. (A and B) Adult PBMCs (AD) and CBMCs (NB) (n = 5 to 6) were incubated for 16 h with estradiol (E; 10⁻⁷ M) or progesterone (P; 10⁻⁵ M) (A) prior to stimulation for 2 h with E. coli (B). TLR2, TLR4, and MYD88 mRNA expression was analyzed by real-time PCR. Data are expressed in A.U. relative to the expression of the corresponding gene in adult PBMCs (A) or in untreated CBMCs (B) and represent means ± SEMs. *, P < 0.05. (C) Western blot analysis of TLR2, TLR4, and MYD88 expression in CBMCs incubated for 16 h or 48 h with estradiol (10⁻⁷ M) or progesterone (10⁻⁵ M). Densitometric values (mean of 5 to 6 independent experiments) are expressed relative to expression in untreated CBMCs (set at 1). (D) CBMCs were incubated for 16 h with hydrocortisone (10⁻⁷ M), estradiol (10⁻⁷ M), or progesterone (10⁻⁵ M) prior to stimulation with PMA and ionomycin. TNF (black bars) and IL-8 (white bars) concentrations in cell culture supernatants were measured by bioassay and ELISA. Data represent means ± SEMs of 4 independent experiments performed in triplicate. *, P < 0.05 versus no steroids.

Fig. 6.

Estradiol and progesterone inhibit E. coli-induced NF-κB activation. CBMCs were incubated for 16 h with dexamethasone (10⁻⁷ M), estradiol (E; 10⁻⁷ M), or progesterone (P; 10⁻⁵ M) prior to stimulation with E. coli for 15 or 30 min (A and B) or 2 h (C). (A and B) Total and phosphorylated (p) IκBα (A) and NF-κB p65 (p65) (B) in nuclear (n) and cytosolic (c) extracts were detected by Western blotting. Densitometric values (means of 3 to 6 independent experiments) are expressed relative to expression in untreated CBMCs (set at 1). (C) IκBα mRNA expression was quantified by real-time PCR and normalized to the expression of HPRT. Data represent means ± SEMs of 6 independent determinations. * and **, P < 0.05 versus untreated and E. coli-stimulated CBMCs, respectively. (D) CBMCs were preincubated for 1 h with 10⁻⁵ M quinazoline prior to incubation with E. coli for 6 h. TNF concentrations in cell culture supernatants were measured by bioassay. Data represent means ± SEMs of 4 independent experiments performed in triplicate. *, P < 0.05.

Fig. 7.

Estradiol and progesterone do not inhibit E. coli-induced MAPK activation. CBMCs were incubated for 16 h with dexamethasone (10⁻⁷ M), estradiol (10⁻⁷ M), or progesterone (10⁻⁵ M) prior to stimulation for 15 or 30 min with E. coli. Phosphorylated (p) and total p38 (A) and ERK1/ERK2 (B) in cytosolic extracts and nuclear c-Jun (C) were detected by Western blotting. Densitometric values (mean of 3 to 6 independent experiments) are expressed relative to expression in untreated CBMCs (set at 1).