Maternal regulation of inflammatory cues is required for induction of preterm birth

Abstract

Infection-driven inflammation in pregnancy is a major cause of spontaneous preterm birth (PTB). Both systemic infection and bacterial ascension through the vagina/cervix to the amniotic cavity are strongly associated with PTB. However, the contribution of maternal or fetal inflammatory responses in the context of systemic or localized models of infection-driven PTB is not well defined. Here, using intraperitoneal or intraamniotic LPS challenge, we examined the necessity and sufficiency of maternal and fetal Toll-like receptor (TLR) 4 signaling in induction of inflammatory vigor and PTB. Both systemic and local LPS challenge promoted induction of inflammatory pathways in uteroplacental tissues and induced PTB. Restriction of TLR4 expression to the maternal compartment was sufficient for induction of LPS-driven PTB in either systemic or intraamniotic challenge models. In contrast, restriction of TLR4 expression to the fetal compartment failed to induce LPS-driven PTB. Vav1-Cre-mediated genetic deletion of TLR4 suggested a critical role for maternal immune cells in inflammation-driven PTB. Further, passive transfer of WT in vitro-derived macrophages and dendritic cells to TLR4-null gravid females was sufficient to induce an inflammatory response and drive PTB. Cumulatively, these findings highlight the critical role for maternal regulation of inflammatory cues in induction of inflammation-driven parturition.

Keywords: Cellular immune response; Cytokines; Inflammation; Mouse models.

Figure 1. Preterm birth is induced by both systemic and local route of challenge.

(A) A schematic overview of the approach used to study PTB in gravid mice following LPS challenge. (B) Gravid WT mice (n = 4–12/condition) were injected i.p. with saline (US, unstimulated) or LPS (standard or ultrapure) at the indicated doses on day 16 of gestation, and the incidence of PTB was quantified. χ² (2 × 3 matrix): standard LPS P = 0.0108; ultrapure LPS P < 0.0001. (C) Ultrasound image taken of an individual amniotic sac on day 16 of gestation. Contrast (green) was included in the i.a. injection, and all the injected fluid was retained within the amniotic sac. Saline or the concentration of ultrapure LPS used in challenge is provided below each bar and was administered as 2 doses in separate amniotic sacs for each uterine horn. Instance of PTB was quantified (n = 3–4/condition). χ² (2 × 3 matrix) P = 0.0041. (D–G) Gravid WT mice (n = 3–8/condition) were challenged with ultrapure LPS by i.p. (75 μg) or i.a. (5 μg) injection, and mRNA expression in the decidua/myometrium was quantified at 6 and 12 hours postchallenge. Data represent fold change over nonstimulated ± SEM. (D) Cd68, Ccl2, and Ccl4 mRNA expression. (E) Ifnb, Isg15, and Irf7 mRNA expression. (F) Il6, Tnf, and Il1b mRNA expression. (G) Ptgs2 mRNA expression. (D–G) ANOVA followed by Tukey’s correction. *P < 0.05, **P < 0.01.

Figure 2. Maternal TLR4 expression is required to induce preterm birth.

(A) MEFs (n = 3/condition) were isolated on day 13 of pregnancy, and the IL-6 response to ultrapure LPS was measured in WT, TLR4-heterozygous, and TLR4-knockout cells. Data represent average ± SEM. (B) IL-6 levels in the serum and amniotic fluid of WT mice 6 hours after i.p. or i.a. administration of LPS (75 μg or 5 μg, respectively) or saline on day 16 of pregnancy (n = 3/condition). Cytokines were measured in pooled amniotic fluid for each uterine horn and graphed as the average amount per pregnancy. Data represent average ± SEM. (C) Endotoxin levels in amniotic fluid of WT mice 6 hours after i.p. or i.a. administration of LPS (75 μg or 5 μg, respectively) or saline on day 16 of pregnancy (n = 3–4/condition). Amniotic fluid was pooled for each uterine horn and graphed as the average amount per pregnancy. Data represent average ± SEM. (D) Levels of serum IL-6 and TNF determined by in vivo cytokine capture assay (IVCCA) in adult WT, TLR4-heterozygous, and TLR4-knockout mice following i.p. injection with ultrapure LPS (n = 3–5/condition). Data represent average ± SEM. (E) Gravid WT and TLR4-knockout mice (n = 3–7/condition) carrying heterozygous pups were treated at day 16 of pregnancy with i.p. or i.a. ultrapure LPS and instance of PTB was quantified. χ² (2 × 4 matrix) P = 0.0004. (A–D) ANOVA followed by Tukey’s correction. **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 3. Maternal TLR4 expression is sufficient to induce preterm birth.

(A) Schematic overview of the approach used to evaluate inflammation-induced PTB in TLR4-knockout and WT pregnancies following reciprocal embryo transfer. (B) Pregnancy day 16 poly(I:C) challenge in TLR4^–/– mothers carrying WT pups (n = 5); instances of PTB were quantified. (C) Gravid WT and TLR4^–/– mice (n = 5/condition) carrying the indicated genotype of pup following embryo transfer were challenged i.p. with 75 μg ultrapure LPS at D16 of pregnancy and instances of PTB were quantified. χ² (2 × 4 matrix) P = 0.0002. (D) Gravid WT mice (n = 3) carrying TLR4^–/– pups following embryo transfer were challenged i.a. with 5 μg ultrapure LPS at day 16 of pregnancy and instances of PTB were quantified.

Figure 4. Activation of TLR4 on immune cells contributes to preterm birth.

(A) Gravid TLR4^fl/fl Vav1-Cre and WT female mice (n = 5–8/condition) were challenged i.p. with 75 μg ultrapure LPS at day 16 of pregnancy and instances of PTB were quantified. Fisher’s exact test P = 0.0002. (B) TLR4^–/– mice received WT in vitro–derived macrophages and dendritic cells by passive transfer (n = 2–5/condition) as indicated, and 2 hours later mice were challenged with 75 μg ultrapure LPS or saline. Serum levels of IL-6 and TNF were measured by IVCCA. Data represent average ± SEM. ANOVA **P < 0.01, ***P < 0.001, ****P < 0.0001. (C) TLR4^–/– mice received 150 × 10⁶ WT in vitro–derived macrophages and dendritic cells by passive transfer (n = 3), and 2 hours later mice were challenged with 75 μg ultrapure LPS alongside WT controls (n = 6). Serum levels of IL-6 and TNF were measured by IVCCA. (D) Gravid TLR4^–/– mice received 150 × 10⁶ WT in vitro–derived macrophages and dendritic cells by passive transfer (n = 5) on day 16 of pregnancy and 2 hours later were challenged with 75 μg ultrapure LPS alongside gravid TLR4^–/– controls (n = 5). Instances of PTB were quantified. Fisher’s exact test P = 0.0476. (E) Gravid TLR4^fl/fl LysM-Cre, TLR4^fl/fl CD11c-Cre, and WT female mice (n = 7–9/condition) were treated with LPS at day 16 of pregnancy and instances of PTB were quantified. χ² P = 0.1244. (F) WT, TLR4^fl/fl Vav1-Cre, TLR4^fl/fl LysM-Cre, and TLR4^fl/fl CD11c-Cre mice were treated with ultrapure LPS and serum levels of IL-6 and TNF were measured by IVCCA (n = 3–5/condition). Data represent average ± SEM. ANOVA of each Cre⁺ condition compared with Cre^– followed by Tukey’s correction. ****P < 0.001.