222 - Cord Blood Proteomic Profiles Reveal Novel Placenta-Lung Endotypes of Bronchopulmonary Dysplasia

Publication Number: 222.4415

Jiaqi Zhao, University of California, San Diego School of Medicine, San Diego, CA, United States; Janu Newar, University of California, San Diego School of Medicine, San Diego, CA, United States; Aliyah Abanes, University of California, San Diego School of Medicine, San Diego, CA, United States; Mana Parast, University of California, San Diego School of Medicine, La Jolla, CA, United States; Karen Mestan, UC San Diego/Rady Children's Hospital, La Jolla, CA, United States

Background: The placenta is an active mediator of the intrauterine environment and influences fetal lung development. Certain pathways of placental dysfunction may explain the development of chronic lung disease in neonates, the most common being bronchopulmonary dysplasia (BPD). Cord blood (CB) immune proteins circulating at birth may serve as proxies of placental dysfunction that inform BPD evolution.

Objective: Identify proteomic pathways of placental dysfunction that elucidate mechanisms of BPD endotypes.

Design/Methods: Large-scale proteomics (7,596 analytes) was conducted using aptamer-based SomaScan platform (SomaLogic, Boulder, CO) in archived CB plasma from 48 extremely preterm infants (≤28 completed weeks gestation) enrolled in a Chicago-based birth cohort. Proteomics output was linked to 4 domains of placental pathology as defined by the 2016 Amsterdam Workshop Criteria: acute inflammation (AI), chronic inflammation (CI), maternal vascular malperfusion (MVM), and fetal vascular malperfusion (FVM). BPD (N=19 infants) was defined based on NIH consensus criteria. Protein intensities were log-transformed and compared according to the presence of placental lesions among the 4 domains, using student’s t and ANOVA with multiple testing correction. KEGG pathway enrichment analysis was conducted with ClusterProfiler in R.

Results: Distinct proteomic profiles were observed across the 4 placental domains, with inflammatory domains (AI and CI) showing differential expression compared to controls (normal pathology, N=8), and fewer proteins identified with vascular dysfunction (MVM and FVM). 19 proteins were differentially expressed with AI and/or CI, with heatmaps demonstrating upregulation of oxidant stress and neutrophil defense pathway proteins (hemopexin, neutrophil defensins) (Fig. 1). When stratified by any AI vs. none, 42 proteins were differentially expressed, with distinct clusters in the AI group which varied according to BPD status (Fig. 2). Patterns for CI were largely reversed, suggesting that compensatory mechanisms of chronic inflammation play an active role in BPD pathogenesis. Using the KEGG database, 11 biological pathways were significantly associated with AI (Fig. 3; P-adjusted < 0.05). Of note, chemokine signaling and neutrophil extracellular trap formation pathways were suppressed with AI but activated with CI.

Conclusion(s): CB proteomic profiles vary across placental dysfunction among EPT infants with and without BPD. Immune profiling according to intrauterine pathology may elucidate mechanisms that define distinct placenta-lung endotypes, leading to a more personalized BPD management approach.

Figure 1
Heatmap displaying the protein levels [log2 relative fluorescent units (RFUs)] for selected proteins (rows) by subjects (columns) in inflammatory vs. healthy controls. Columns are annotated by BPD and placental pathologic domain status.

Figure 2
Volcano plots (2a, 2c) show the –log FDR-adjusted P values versus log2 fold change for protein expression comparisons between AI vs. No AI and CI vs. No CI patient samples. Heatmaps (2b, 2d) display the protein levels [log2 relative fluorescent units (RFUs)] for selected proteins (rows) by subjects (columns). Columns are annotated by BPD and path domain status.

Figure 3
Pathway and enrichment analysis of proteomics data. Biological pathways significantly (p adjusted < 0.05) associated with AI and CI domains, respectively, based on proteomics data.