Dosing and formulation of antenatal corticosteroids for fetal lung maturation and gene expression in rhesus macaques

Abstract

Antenatal corticosteroids (ANS) are the major intervention to decrease respiratory distress syndrome and mortality from premature birth and are standard of care. The use of ANS is expanding to include new indications and gestational ages, although the recommended dosing was never optimized. The most widely used treatment is two intramuscular doses of a 1:1 mixture of betamethasone-phosphate (Beta-P) and betamethasone-acetate (Beta-Ac) - the clinical drug. We tested in a primate model the efficacy of the slow release Beta-Ac alone for enhancing fetal lung maturation and to reduce fetal corticosteroid exposure and potential toxic effects. Pregnant rhesus macaques at 127 days of gestation (80% of term) were treated with either the clinical drug (0.25 mg/kg) or Beta-Ac (0.125 mg/kg). Beta-Ac alone increased lung compliance and surfactant concentration in the fetal lung equivalently to the clinical drug. By transcriptome analyses the early suppression of genes associated with immune responses and developmental pathways were less affected by Beta-Ac than the clinical drug. Promoter and regulatory analysis prediction identified differentially expressed genes targeted by the glucocorticoid receptor in the lung. At 5 days the clinical drug suppressed genes associated with neuronal development and differentiation in the fetal hippocampus compared to control, while low dose Beta-Ac alone did not. A low dose ANS treatment with Beta-Ac should be assessed for efficacy in human trials.

Figure 1

(A) Pressure-volume curves showing comparable improved static compliance after treatment with Beta-Ac (0.125 mg/kg) and the clinical drug. (B) Lung gas volumes were increased significantly for Beta-Ac 0.125 mg/kg and the clinical drug. (C) Saturated phosphatidylcholine (SatPC) concentration in the bronchoalveolar lavage fluid (BALF) increased with Beta-Ac 0.125 mg/kg and the clinical drug compared to control. *p-value < 0.05 vs. control.

Figure 2

Transcriptomic analysis was performed on whole lung RNA 4 hours and 5 days after administration of the clinical drug, and 6 hours and 5 days after Beta-Ac 0.125 mg/kg and from saline-treated animals (n = 3 animals per group). 3-dimensional principal component analysis (PCA) was generated using log-transformed read counts. (A) At the early timepoint PCA separated the animals treated with the clinical drug at 4 h, Beta-Ac at 6 h, and controls. (B) At 5 days there is separation of control animals and overlap of animals treated with the clinical drug and Beta-Ac (C) Sample correlation heatmap displaying the similarity in gene expression profile. Red color indicates increasing sample correlation and yellow color indicates decreasing sample correlation. Dendrogram clustering indicates the overall similarity of the samples. Both analyses showed separation of the clinical drug at 4 h and overlap of animals in the other treatment groups.

Figure 3

Top common differentially expressed genes in each group relative to control were determined using thresholds of p-value < 0.05, q-value < 0.1 and fold-change > 1.5. (A) Top 20 up and down regulated genes relative to control and predicted regulation by the glucocorticoid receptor (GR) either by the presence of a GR motif or predicted interaction by ingenuity pathway analysis. Genes reported in the literature to be associated with “lung”, “respiratory disease”, or “lung cell line” are bolded. Scatterplots of log fold-changes (logFC) of differentially expressed genes from. (B) Beta-Ac 6 h and the clinical drug 4 h relative to control and (C) Beta-Ac 5d and the clinical drug 5d relative to control. Most commonly differentially expressed genes had a similar direction and magnitude of changes. There was strong correlation between logFC values (r = 0.87 for 6 h and 4 h; r = 0.96 for 5d).

Figure 4

Gene set enrichment analysis comparing the clinical drug and Beta-Ac to control in the lung at time of peak Beta plasma concentration in the fetus. Selected gene ontology terms are displayed with the bar chart representing log p-values. Positive p-values denote induced genes and negative p-values denote suppressed genes. Genes suppressed by the clinical drug were more strongly associated with morphogenesis and developmental processes than genes suppressed by the Beta-Ac.

Figure 5

Transcriptomic analysis was performed for fetal hippocampus from saline-treated animals and 5 days after administration of ANS (n = 5 animals per group). (A) 3-dimensional principal component analysis (PCA) was generated using log-transformed read counts. (B) Sample correlation heatmap displaying the similarity in gene expression profile. Red color indicates increasing sample correlation and yellow color indicates decreasing sample correlation. Dendrogram clustering indicates the overall similarity of the samples. Both analyses showed separation of the control from animals treated with the clinical drug, while animals treated with Beta-Ac are interspersed between the controls and clinical treatment animals. (C) Top differentially induced and suppressed genes by the clinical drug in the hippocampus at 5 days. Genes are ordered by magnitude of gold change; p values are adjusted by the Benjamin-Hochberg method.

Figure 6

Gene set enrichment analysis comparing differentially expressed genes for the clinical drug compared to control in the fetal hippocampus. There was no differential expression between Beta-Ac and control. Selected gene ontology terms are displayed with the bar chart representing log p-values. Positive p-values denote induced genes and negative p-values denote suppressed genes.

Figure 7

Network of genes suppressed by the clinical drug in the fetal hippocampus at 5 days associated with the biological processes of neurogenesis and neuron differentiation.