Placental transcriptomic signatures of spontaneous preterm birth

Abstract

Background: Spontaneous preterm birth accounts for most preterm births and leads to significant morbidity in the newborn and childhood period. This subtype of preterm birth represents an increasing proportion of all preterm births when compared with medically indicated preterm birth, yet it is understudied in omics analyses. The placenta is a key regulator of fetal and newborn health, and the placental transcriptome can provide insight into pathologic changes that lead to spontaneous preterm birth.

Objective: This analysis aimed to identify genes for which placental expression was associated with spontaneous preterm birth (including early preterm and late preterm birth).

Study design: The ECHO PATHWAYS consortium extracted RNA from placental samples collected from the Conditions Affecting Neurocognitive Development and Learning in Early Childhood and the Global Alliance to Prevent Prematurity and Stillbirth studies. Placental transcriptomic data were obtained by RNA sequencing. Linear models were fit to estimate differences in placental gene expression between term birth and spontaneous preterm birth (including gestational age subgroups defined by the American College of Obstetricians and Gynecologists). Models were adjusted for numerous confounding variables, including labor status, cohort, and RNA sequencing batch. This analysis excluded patients with induced labor, chorioamnionitis, multifetal gestations, or medical indications for preterm birth. Our combined cohort contained gene expression data for 14,023 genes in 48 preterm and 540 term samples. Genes and pathways were considered statistically significantly different at false discovery rate-adjusted P value of <.05.

Results: In total, we identified 1728 genes for which placental expression was associated with spontaneous preterm birth with more differences in expression in early preterm samples than late preterm samples when compared with full-term samples. Of those, 9 genes were significantly decreased in both early and late spontaneous preterm birth, and the strongest associations involved placental expression of IL1B, ALPL, and CRLF1. In early and late preterm samples, we observed decreased expression of genes involved in immune signaling, signal transduction, and endocrine function.

Conclusion: This study provides a comprehensive assessment of the differences in the placental transcriptome associated with spontaneous preterm birth with robust adjustment for confounding. Results of this study are in alignment with the known etiology of spontaneous preterm birth, because we identified multiple genes and pathways for which the placental and chorioamniotic membrane expression was previously associated with prematurity, including IL1B. We identified decreased expression in key signaling pathways that are essential for placental growth and function, which may be related to the etiology of spontaneous preterm birth. We identified increased expression of genes within metabolic pathways associated exclusively with early preterm birth. These signaling and metabolic pathways may provide clinically targetable pathways and biomarkers. The findings presented here can be used to understand underlying pathologic changes in premature placentas, which can inform and improve clinical obstetrics practice.

Keywords: ALPL; GABRP; IL1B; chemokine signaling; placenta; placental metabolism; signal transduction; spontaneous preterm birth; transcriptomics.

Figure 1:

(A) Summary of ECHO PATHWAYS Consortium Cohort and Study Specific Design. (B) Directed Acyclic Diagram of covariates used in Analysis.

Figure 2:

Volcano plots depicting significance (-log10 p values, Y axis) and strength (LogFC, x axis) in our analyses of (A) Preterm vs.Term infants and (B) ACOG Gestational Length subgroups vs. full term infants. Purple Xs represent placental maturation signatures (from Supplemental table 1) not included in final analysis. The top 5 genes based on log fold change are labelled

Figure 3:

Summary of Differentially Expressed Genes whose placental expression was associated with preterm birth and/or early and/or late preterm birth. (A) Venn Diagram and UpsetR plot depicting shared and distinct changes in placental gene expression. The horizontal bars indicate the total number of positive or negative DEGs in each subgroup, and the vertical bars represent the number of unique or overlapping DEG (B)The Log fold change of top 5 differentially expressed genes (based on combined FDR adjusted p values) for genes significantly associated with more than one outcome from the subgroups noted in the UpsetR plot, with subgroups highlighted by color in upsetR Plot.

Figure 4:

Rotational gene set testing revealed shared and distinct pathways whose placental expression was associated with prematurity. (A) Venn diagram of shared and distinct pathways between analysis groups. (B) dotplot depicting results for 65 pathways associated with all three analyses. The size of the dots are scaled to the log adjusted p value, and ordered on the X axis based on the proportion of genes up or downregulated. Pathways are grouped into KEGG subgroups.