RelB/NF-κB2 regulates corticotropin-releasing hormone in the human placenta

Abstract

Placental CRH may be part of a clock that governs the length of human gestation. The mechanism underlying differential regulation of CRH in the human placenta is poorly understood. We report here that constitutively activated RelB/nuclear factor-κB2 (NF-κB)-2 (p100/p52) acts as an endogenous stimulatory signal to regulate CRH by binding to an NF-κB enhancer of CRH gene promoter in the human placenta. Nuclear staining of NF-κB2 and RelB in villous syncytiotrophoblasts and cytotrophoblasts was coupled with cytoplasmic CRH in syncytial knots of cytotrophoblasts. Chromatin immunoprecipitation identified that CRH gene associated with both RelB and NF-κB2 (p52). Dexamethasone increased synthesis and nuclear translocation of RelB and NF-κB2 (p52) and their association with the CRH gene. In contrast, progesterone, a down-regulator of placental CRH, repressed NF-κB2 (p100) processing, nuclear translocation of RelB and NF-κB2 (p52), and their association with the CRH gene. Luciferase reporter assay determined that the NF-κB enhancer of CRH was sufficient to regulate transcriptional activity of a heterologous promoter in primary cytotrophoblasts. RNA interference-mediated repression of RelB or NF-κB2 resulted in significant inhibition of CRH at both transcriptional and translational levels and prevented the dexamethasone-mediated up-regulation of CRH transcription and translation. These results suggest that the noncanonical NF-κB pathway regulates CRH production in the human placenta and is responsible for the positive regulation of CRH by glucocorticoids.

Fig. 1.

The proteasomes regulate CRH transcription in cytotrophoblasts. Human cytotrophoblasts were incubated with vehicle (lane 1) or proteasomal inhibitors MG132 (lanes 2–5), lactacystin (lanes 6–9), PSI (lanes 10–13) or nonproteasomal inhibitor, or ZLL (lanes 14–17) at a concentration ranging from 0.1 to 3.0 μm as indicated for 24 h. The total RNA were extracted and subjected to Northern blotting analysis with the use of GAPDH as the reference.

Fig. 2.

Immunohistochemical analyses of NF-κB components and CRH in human term placenta. Four sections from the same placenta were stained with antibody to human RelB (A), NF-κB2 (B), CRH (C), or Rel A (p65) (D), respectively. Representative images are shown in the top panel; original magnification, ×100. Bottom panels represent insets; original magnification, ×400. E, Schematic presentation of cross-section of human term placental villus. CT, Cytotrophoblasts; ST, syncytiotrophoblast; SM, syncytium; SK, syncytial knot; FBV, fetal blood vessel.

Fig. 3.

RelB and NF-κB2 associate with the CRH gene in primary cytotrophoblasts. A, Schematic presentation of the human CRH gene and regulatory elements. The CRH gene contains a putative NF-κB enhancer (κB) at −295, in addition to a negative GRE at −277 and a CRE at −226. B, ChIP analyses were performed to determine occupancy of individual NF-κB family member, p65, NF-κB1 (p50), C-Rel, RelB, and NF-κB2, at the CRH gene promoter. Fold enrichment of CRH with each member was calculated with normalization to a human α-satellite. Rabbit IgG served as a nonspecific control. The bars indicate the average of CRH fold enrichment (y-axis) against each member (x-axis), with error bars representing the sd from three independent experiments. The fold enrichment of CRH with IgG was defined as 1. *, P < 0.001, ***, P > 0.05 (no significant difference, vs. the control, IgG), as determined by one-way ANOVA with Dunnett's test. C, ChIP analyses were performed to determine change in fold enrichment of CRH with RelB or NF-κB2 when cytotrophoblasts were treated with vehicle (−), DEX 1 μm, or progesterone (Prog) 10 μm. Fold enrichment of CRH was determined and presented as described in B. The fold enrichment of CRH with antibody to RelB or NF-κB2 in the presence of vehicle but absence of steroids was defined as 1. *, P < 0.001, **, P < 0.01 (vs. the control cells treated with the vehicle), as determined by one-way ANOVA with Dunnett's test.

Fig. 4.

Differential effects of steroids on CRH correspond with their effects on production and/or nuclear translocation of NF-κB2/RelB in cytotrophoblasts. A, Whole-cell lysates of term cytotrophoblasts that were pretreated with vehicle (−) or DEX 1 μm or progesterone (Prog) 10 μm for 24 h were probed with CRH (upper left panel), RelB (lower left panel), or NF-κB2 (right panel) antibodies. β-Actin was used as a loading control. B, Nuclear (top panels) and cytosolic extracts (bottom panels) from the same whole-cell lysates of A were prepared and probed with RelB (left panel) or NF-κB2 (right panel) antibodies. SP1 was used as a loading control for nuclear extracts and β-actin was used as a loading control for both nuclear and cytosolic extracts. These gels are representatives of three repeated experiments performed on three separate placentas. SP, Specificity protein 1.

Fig. 5.

The putative NF-κB enhancer of CRH regulates transcription of a heterologous promoter. A, Schematic presentation of luciferase reporter constructs. pGL4.32 (Promega) harbors five copies of the NF-κB enhancer and a minimal promoter (TATA), which regulate expression of firefly luciferase gene (FL). pGL4.32-CRH1X and -CRH4X contain one and four copies of 5′-GGGAAATCTC-3′, respectively. pGL4.32-CRH1XMut that contains one copy of mutant version (mutant residues underlined) was also created. pCMV-RL (renilla luciferase) was used as the internal control. B, Human cytotrophoblasts were cotransfected with pCMV-RL (200 ng) and the constructs (1 μg) as indicated at the bottom. The bars represent the average of fold change of FL with normalization to renilla luciferase (RL), and error bars represent the sd from three experiments. Fold FL activity by pGL-4.32CRH1XMut was defined as 1. *, P < 0.001, **, P < 0.01 (vs. FL activity of control cells transfected with pGL4.32-CRH1XMut), as determined by one-way ANOVA with Dunnett's test. C, Cytotrophoblasts were cotransfected with pGL4.32-CRH1X (1 μg) and pCMV-RL (200 ng) for 24 h and then treated with vehicle (−) or with DEX or progesterone (Prog) for an additional 24 h. FL activity was determined with normalization to RL activity and presented in bars and errors as described in B. Fold FL activity by pGL-4.32CRH1X in the presence of vehicle but absence of steroids was defined as 1. *, P < 0.001, **, P < 0.01 (vs. FL activity of control cells transfected with pGL4.32-CRH1X in the presence of vehicle), as determined by one-way ANOVA with Dunnett's test.

Fig. 6.

The effects of RelB/NF-κB2 gene silencing on transcriptional and translational levels of CRH in primary cytotrophoblasts. A, Human term cytotrophoblasts were transfected with RelB or NF-κB2 siRNA or nontargeting scramble siRNA (SS) for 24 h. The cells were then treated with vehicle (−) or with DEX 1 μm or progesterone (Prog) 10 μm. After another 24 h, the cells were lysed and probed with antibody to human CRH, RelB (left panel), or NF-κB2 (right panel). β-Actin served as a loading control. B, Total RNA were extracted from A, and quantitative RT-PCR was performed with normalization to GAPDH mRNA. The bars represent average of fold change in mRNA levels, and error bars represent sd from three experiments. The mRNA levels of RelB or NF-κB2 in cells treated nontargeting scramble siRNA in the absence of steroids were defined as 100%. *, P < 0.001, **, P < 0.01 (vs. that of control cells transfected with SS and treated with vehicle), as determined by one-way ANOVA with Dunnett's test. C, The cells pretransfected with siRNA indicated in A were cotransfected with pGL4.32-CRH1X (1 μg) and pCMV-RL (200 ng) and with vehicle or DEX or progesterone (Prog) for an additional 24 h. FL activity was determined with normalization to renilla luciferase (RL) activity. The bars represent average of percentage change in FL activity, and error bars represent sd from three experiments. FL activity in cells treated nontargeting scramble siRNA in the presence of vehicle but absence of steroids was defined as 100%. *, P < 0.001, **, P < 0.01 (vs. that of control cells transfected with SS and treated with vehicle), as determined by one-way ANOVA with Dunnett's test.