Early growth response protein-1 mediates lipotoxicity-associated placental inflammation: role in maternal obesity

Abstract

Obesity is associated with low-grade chronic inflammation, which contributes to cellular dysfunction promoting metabolic disease. Obesity during pregnancy leads to a proinflammatory milieu in the placenta; however, the underlying causes for obesity-induced placental inflammation remain unclear. Here, we examine the mechanisms by which saturated fatty acids and inflammatory cytokines induce inflammation in placental trophoblasts. We conducted global transcriptomic profiling in BeWo cells following palmitate and/or TNFα treatment and gene/protein expression analyses of MAPK pathways and characterized downstream transcription factors directly regulating inflammatory cytokines. Microarray analysis revealed increased expression of genes regulating inflammation, stress response, and immediate early response in cytotrophoblasts in response to palmitic acid (PA), TNFα, or a combination of both (PA + TNFα). Both gene ontology and gene set enrichment analysis revealed MAPK and EGR-1 signaling to be upregulated in BeWo cells, which was confirmed via immunoblotting. Importantly, activation of JNK signaling was necessary for increased proinflammatory cytokine (IL-6, TNFα, and IL-8) and EGR1 mRNA. Consistent with the requirement of JNK signaling, ChIP analysis confirmed the recruitment of c-Jun and other MAPK-responsive immediate early factors on the EGR1 promoter. Moreover, recruitment of EGR-1 on cytokine promoters (IL-6, TNFα, and IL-8) and an impaired proinflammatory response following knockdown of EGR-1 suggested it as a central component of the mechanism facilitating inflammatory gene expression. Finally, akin to in vitro findings, term placenta from obese women also had both increased JNK and p38 signaling and greater EGR-1 protein relative to lean women. Our results demonstrate that lipotoxic insults induce inflammation in placental cells via activation of JNK/EGR-1 signaling.

Trial registration: ClinicalTrials.gov NCT01104454.

Keywords: developmental programming; early growth response protein-1; fatty acids; gestational obesity; mitogen-activated protein kinase.

Figures

Fig. 1.

Transcriptome analysis of BeWo cells challenged with palmitic acid (PA), TNFα, or PA + TNFα. A and B: Venn diagram (A) and hierarchical clustering (B) of differentially expressed genes (±1.5-fold change, P ≤ 0.05) in BeWo cells following 24-h exposure to PA, TNFα, and PA + TNFα. C: gene ontology (GO) analysis of biological processes of genes altered by PA + TNFα treatment. D: correlation-based clustering of genes with known functions in response to chemical stimulus. E and F: gene set enrichment analysis of molecular functions showing MAPK signaling (E) and transcription factor motifs showing genes regulated by serum response factor (SRF), ATF3, and CREB/c-Jun enriched in BeWo cells treated with PA or PA + TNFα (F). Heat map colors red, white, and blue represent upregulation, no relative effect, and downregulation of transcripts, respectively. G: mRNA expression of genes involved with inflammation, immediate early response, and lipid transport assessed via real-time RT-PCR (n = 6/group). Values were normalized to cyclophillin mRNA expression and are expressed as mean fold change relative to the vehicle-treated group. Statistical significance (P ≤ 0.05) was assessed using a 2-way ANOVA, followed by all-pairwise comparison by the Student-Neuman-Keuls method. Different lowercase letters (a, b, or c) indicate significantly different values. Con, control.

Fig. 2.

Temporal analysis of MAPK activation and cytokine gene expression in response to PA or PA + TNFα. A and B: immunoblot anlaysis for total and phosphorylated p38, JNK1/2, and ERK1/2 over a 24-h time course following treatment with PA (A) or PA + TNF-α (B). C and D: densitometric quantitation of immunoblots (normalized to α-tubulin) for phosphorylated protein (◇), total protein (gray circle), or the ratio of phosphorylated to total protein (■) for PA (C) or PA + TNF-α treated cells (D). Values are expressed as mean fold change over 0 h. *Significant difference for the ratio of phosphorylated to total protein (P ≤ 0.05) compared with 0-h time point. E: mRNA expression of cytokine genes at 24 h following exposure to PA or PA + TNFα with and without preexposure to MAPK inhibitors SP-600125 (JNK inhibitor), PD-98059 (ERK1/2 inhibitor), and SB-203580 (p38 inhibitor). Gene expression was normalized to cyclophillin mRNA (n = 6/group). Statistical differences were determined using Student's t-test. *Significance, P ≤ 0.05.

Fig. 3.

Expression of ATF3, c-Jun, and SRF following exposure to PA and PA + TNFα. A: mRNA expression of the immediate early-response genes via real-time PCR following exposure to PA or PA + TNFα over 24 h (n = 6/group). B: immunoblot analysis and densitometry quantitation of ATF3, c-Jun, and SRF nuclear protein levels 8 h following PA or PA + TNFα treatment. Densitometry values were normalized to TATA-binding protein. C: mRNA expression of immediate early-response genes at 24 h following exposure to PA or PA + TNFα with and without preexposure to MAPK inhibitors SP-600125 (JNK inhibitor), PD-98059 (ERK1/2 inhibitor), and SB-203580 (p38 inhibitor). Gene expression was normalized to cyclophillin mRNA (n = 6/group). Statistical differences were determined using Student's t-test. *Significance, P ≤ 0.05.

Fig. 4.

Transcriptional regulation of early growth response protein-1 (EGR-1) in response to PA or PA + TNFα. A: EGR1 mRNA expression following exposure to PA or PA + TNFα over 24 h. Values were normalized to cyclophillin mRNA (n = 6/group). B: GO term (biological processes) enrichment of genes with differential H3K4 trimethylation (me3) following PA treatment (500 μM, 24 h). Genome-wide H3K4me3 was assayed using chromatin immunoprecipitation (ChIP)-seq. C: alignment of H3K4me3 reads following ChIP-seq showing greater read density at the EGR1 gene locus following PA treatment. From top, tracks showing gene, mRNA, and coding DNA sequence (CDS) are in red, followed by control and PA groups. Read densities are shown as a heat map (red being high, blue being low) showing greater H3K4me3 in the PA group. D: a schematic of the human EGR1 promoter upstream of the transcriptional start site (+1). Response elements from SRF (SRE; white box), CREB (CRE; white oval), SP1 (black oval), Ets-1 (EBS; black box), and AP-1 (◇) are pictured in approximate location within the EGR1 promoter (not to scale). Arrows indicate PCR primer locations, and the dotted lines show areas of promoter amplification. E: ChIP was performed with ATF3, c-Jun, and SRF (or IgG as control) at 6 h following treatments with PA or PA + TNFα. PCR was performed using immunoprecipitated DNA or input DNA for various regions of the EGR1 promoter. Band intensity was quantified by densitometry and normalized to input DNA. Values are expressed relative to the control group. Statistical differences were determined using Student's t-test. *Significance, P ≤ 0.05. IP, immunoprecipitation.

Fig. 5.

EGR1 regulation of inflammatory cytokine gene expression in response to PA or PA + TNFα. A: Western blot analysis and densitometry quantitation of EGR-1 nuclear protein levels following 24 h of PA or PA + TNFα treatment. Densitometry values were normalized to TATA-binding protein (n = 5). B: ChIP for EGR-1 (or IgG as control) was performed following 24-h treatments with PA or PA + TNFα. PCR was performed using immunoprecipitated or input DNA for regions containing the EGR-1 binding site on the IL-6, TNFα, and IL-8 promoters. Band intensity was quantified by densitometry and normalized to input DNA. Values are expressed as relative to the control group. C: EGR1 mRNA expression in BeWo cells transfected with either scrambled or EGR1 short hairpin RNAs (shRNAs) analyzed using real-time PCR. D: mRNA expression of inflammatory cytokines TNFα, IL-6, and IL-8 following 24-h exposure to PA or PA + TNFα in BeWo cells transfected with either scrambled or EGR1 shRNAs (n = 6). mRNA expression was normalized to cyclophilin mRNA and expressed as means ± SE relative to control group. *Significance, P ≤ 0.05.

Fig. 6.

Protein expression of MAPKs and EGR-1 in term human placenta from lean and obese subjects. A and B: immunoblots (A) and densitometric quantitation (B) for total and phosphorylated p38, JNK1/2, and ERK1/2 in lean and obese placenta collected at term (n = 11). Data are normalized to α-tubulin levels and expressed relative to the lean group (means ± SE). C: immunoblot analysis and densitometric quantitation of EGR-1 protein levels in lean or obese placenta (n = 11). Densitometry values were normalized to GAPDH and expressed as means ± SE. *Significance, P ≤ 0.05.