Maternal NO2 exposure and fetal growth restriction: Hypoxia transmission and lncRNAs-proinflammation-mediated abnormal hematopoiesis

Abstract

Epidemiological studies show a strong correlation between air pollution and fetal growth restriction (FGR), but existing results are controversial due to inherent limitations, such as causality of specific pollutants, developmental origin, and maternal-fetal transmission. To address this controversy, we first conducted a retrospective analysis of 28,796 newborns and revealed that maternal nitrogen dioxide (NO₂) exposure during the second trimester was positively associated with FGR, with an adjusted odds ratio of 1.075 (95% confidence interval: 1.020-1.133) per 10 μg/m³ NO₂ increase for small for gestational age. Then, by establishing an animal model of prenatal NO₂ exposure, we confirmed its adverse effects on embryonic growth and hematopoiesis in the yolk sac and fetal liver, primarily affecting the differentiation of hematopoietic stem and progenitor cells and erythroid maturation. By applying internal exposure analyses coupled with ¹⁵N isotope tracing, we found that maternal NO₂ inhalation induced acquired methemoglobinemia through its byproducts and placental hypoxia in pregnant mice. Importantly, by combining transcriptional profiling, bioinformatics analysis, and RNA binding protein immunoprecipitation (RIP)/chromatin immunoprecipitation (CHIP), we clarified that placental-fetal hypoxia transmission activated hypoxia-inducible factors, disturbed hematopoiesis through the hypoxia-inducible factor 1β-long noncoding RNAs-CCAAT/enhancer binding protein alpha-proinflammatory signaling pathway, ultimately contributing to FGR progression. These findings provide insights for risk prevention and clinical intervention to promote child well-being in NO₂-polluted areas.

Keywords: air NO2 exposure; fetal growth restriction; fetal hematopoietic dysfunction; lncRNAs-proinflammation modulation; maternal–fetal hypoxia transmission.

Fig. 1.

Effects of maternal NO₂ exposure on embryonic growth in mice offspring. (A) Weight (n = 16 to 23) and crown-rump length (n = 10 to 12) of embryos at E10.5. (B) Weight (n = 35 to 39) and crown-rump length (n = 38 to 39) of embryos at E14.5. Representative images of uteri (triangles indicate reabsorbed embryos), fetal YS (white asterisk), AGM (black line circle), and liver (red line circle) at E10.5 (C) and E14.5 (D). (Scale bar, 1 mm.) *P < 0.05, ***P < 0.001, indicates significance between different groups.

Fig. 2.

Effects of maternal NO₂ exposure on fetal hematopoietic development. (A) Schematic diagram of hematopoiesis in mouse embryos. Primitive EryP are first generated in the yolk sac (E7.5) and give rise to primitive nucleated erythrocytes (P-Ery), and they were still present in low frequencies at birth. A second wave of progenitors emerges in the yolk sac (E8.25-E11) as EMPs differentiate into multiple types of blood cells, including megakaryocytes (Mk), granulocytes (Gr), erythrocytes (Ery), and monocyte-derived macrophages (MΦ). After establishing circulation at E10.5, EMPs and HSCs migrate to the FL, and the FL gradually becomes the major hematopoietic organ where EMPs and HSCs continue to proliferate and differentiate until they peak at E14.5. (B) Representative images of o-dianisidine staining of the YS at E10.5. (Scale bar, 1 mm.) (C) CFU assay of YS in the control and NO₂-exposed groups at E10.5 (n = 5 to 6). (D) Representative images of the FL at E14.5 (Left) and the ratio of FL weight to body weight (Right) (n = 32). (Scale bar, 1 mm.) (E) Representative IHC images of the FL with anti-c-kit for HSCs and EMPs at E14.5 (Left), and the relative IOD of c-kit (Right) (n = 4). (Scale bar, 50 μm.) (F) CFU assay of FL in the control and NO₂-exposed groups at E14.5 (n = 8 to 11). (G) Representative images of Wright-Giemsa staining for blood cells collected from peripheral blood at E14.5 (Left), and the ratio of enucleated to nucleated erythrocytes (Right) (n = 17 to 21). (Scale bar, 50 μm.) (H) mRNA expression of α-globin and β-globin in the FL at E14.5 (n = 5 to 6). *P < 0.05, **P < 0.01, indicates significance between different groups.

Fig. 3.

mRNA expression profiles and biological processes in the FL at E14.5 following maternal NO₂ exposure. (A) Flowchart illustrating the steps for identifying and validating the differentially expressed mRNA and lncRNA in the FL at E14.5. (B) Top 5 GSEA results of all genes. The significantly enriched terms were screened by q < 0.05 and |NES| > 1.5. (C) GO terms enriched for the DEGs using DAVID database analysis. The significantly enriched terms were screened by FDR < 0.05. (D) KEGG pathways enriched by DEGs using DAVID database analysis. The significantly enriched terms were screened by FDR < 0.05. (E) Sankey plot depicting the relationship between DEGs and two proinflammatory signaling pathways. (F) Volcano map showing the significant downregulation of the genes involved in proinflammatory signaling. (G) Validation of the expression of 12 DEGs involved in proinflammatory signaling by qRT-PCR. (n = 4 to 8). *P < 0.05, indicates significance between different groups.

Fig. 4.

lncRNA regulation of gene expression in response to maternal NO₂ exposure in the FL at E14.5. (A) Regulatory mechanisms for the three proinflammatory signaling pathways. (B) GSEA showing the core genes in TNF and IL-17 signaling pathways. (C) Coexpression network of the 9 proinflammation-related DEGs and the top 4 LncRNAs with the highest coexpression number. (D) Validation of the expression of lncRNAs (n = 4 to 8). (E) Prediction of the TFs for 9 DEGs in ChIP-Atlas. (F) Hub genes determined by PPI network analysis, with Il1b, Ccl2, and Cxcl10 being considered as hub genes. (G) Prediction of the binding sites between C/EBPα and hub genes via JASPAR. (H) Validation of the predicted binding sites between C/EBPα and hub genes by ChIP analysis (n = 3). (I) Prediction of the interaction between lncRNAs (NONMMUT044528.2 and NONMMUT053442.2) and C/EBPα by RPISeq. (J) Validation of the interaction between lncRNA (NONMMUT044528.2 and NONMMUT053442.2) and C/EBPα by RIP analysis (n = 3). *P < 0.05, **P < 0.01, ***P < 0.001, indicates significance between different groups.

Fig. 5.

Maternal internal exposure and maternal–fetal hypoxia transmission following NO₂ exposure. (A) The levels of NO₃⁻ and NO₂⁻ in maternal blood at E14.5 (n = 5 to 8). (B) Percentage of ¹⁵NO₃⁻ in total NO₃⁻ in maternal blood after 2, 5, and 8 h of continuous ¹⁵NO₂ inhalation at E14.5 (n = 3). (C and D) Methemoglobin levels (n = 6) and blood cell counts (n = 7 to 10) in maternal blood at E14.5. (E) Representative images of hypoxyprobe staining (Left) and the percentage of positive area (Right) in the different regions of the placenta (n = 3). (Scale bar, 200 μm.) (F) Representative images of hypoxyprobe staining (Left) and the percentage of positive area (Right) in the FL at E14.5 (Right) (n = 3). (Scale bar, 200 μm.) (G) The mRNA expression of Hif1α and Hif1β in the FL at E14.5. (n = 8 to 10). (H) Prediction of the binding sites between HIF-1β and lncRNAs NONMMUT044528.2 and NONMMUT053442.2 via the JASPAR. (I) Validation of the binding sites between HIF-1β and lncRNAs NONMMUT044528.2 and NONMMUT053442.2 by ChIP analysis (n = 3). *P < 0.05, ***P < 0.001, indicates significance between different groups.

Fig. 6.

Epidemiological correlation and toxicological mechanisms for FGR progression induced by maternal NO₂ exposure.