Chamber-specific transcriptional responses in atrial fibrillation
Catherine E. Lipovsky, … , Bo Zhang, Stacey L. Rentschler
Catherine E. Lipovsky, … , Bo Zhang, Stacey L. Rentschler
Published August 25, 2020
Citation Information: JCI Insight. 2020;5(18):e135319. https://doi.org/10.1172/jci.insight.135319.
View: Text | PDF
Research Article Cardiology Development

Chamber-specific transcriptional responses in atrial fibrillation

Abstract

Atrial fibrillation (AF) is the most common cardiac arrhythmia, yet the molecular signature of the vulnerable atrial substrate is not well understood. Here, we delineated a distinct transcriptional signature in right versus left atrial cardiomyocytes (CMs) at baseline and identified chamber-specific gene expression changes in patients with a history of AF in the setting of end-stage heart failure (AF+HF) that are not present in heart failure alone (HF). We observed that human left atrial (LA) CMs exhibited Notch pathway activation and increased ploidy in AF+HF but not in HF alone. Transient activation of Notch signaling within adult CMs in a murine genetic model is sufficient to increase ploidy in both atrial chambers. Notch activation within LA CMs generated a transcriptomic fingerprint resembling AF, with dysregulation of transcription factor and ion channel genes, including Pitx2, Tbx5, Kcnh2, Kcnq1, and Kcnip2. Notch activation also produced distinct cellular electrophysiologic responses in LA versus right atrial CMs, prolonging the action potential duration (APD) without altering the upstroke velocity in the left atrium and reducing the maximal upstroke velocity without altering the APD in the right atrium. Our results support a shared human/murine model of increased Notch pathway activity predisposing to AF.

Authors

Catherine E. Lipovsky, Jesus Jimenez, Qiusha Guo, Gang Li, Tiankai Yin, Stephanie C. Hicks, Somya Bhatnagar, Kentaro Takahashi, David M. Zhang, Brittany D. Brumback, Uri Goldsztejn, Rangarajan D. Nadadur, Carlos Perez-Cervantes, Ivan P. Moskowitz, Shaopeng Liu, Bo Zhang, Stacey L. Rentschler

×

Figure 1

Chamber-specific transcriptomic differences in human atrial fibrillation cardiomyocyte nuclei.

Options: View larger image (or click on image) Download as PowerPoint
Chamber-specific transcriptomic differences in human atrial fibrillation...
RNA-sequencing was performed on isolated cardiomyocyte nuclei (CMN) from human right atrium (RA) and left atrium (LA) from nonfailing (NF) donor hearts rejected for transplantation with no history of atrial fibrillation (AF), explanted hearts from individuals with end-stage heart failure (HF) who received a heart transplant without a history of AF, and end-stage heart failure with a documented history of AF (AF+HF). (A) Principal components analysis (PCA) plot showing the distribution of variance among all human samples (NF RA, light teal; AF+HF RA, dark blue; HF RA, light blue; NF LA, brown; AF+HF LA, bright orange; HF LA, orange). Venn diagrams representing the number of differentially upregulated (top) and downregulated (bottom) transcripts in the human RA or LA CMN with (B) HF or (C) AF+HF when compared with NF controls. (D) Heatmap with color scale showing the relative expression levels (Z score; red, up; blue, down) of transcription factors (TFs) differentially expressed in AF+HF RA when compared with NF RA. Relative expression levels of the same TFs are shown for each of the other group comparisons, including AH+HF LA compared with NF LA, HF LA compared with NF LA, HF RA compared with NF RA, as well as a comparison of NF LA to NF RA. (E) Similar heatmap analysis as shown in D, except the TFs shown are those differentially expressed in AF+HF LA when compared with NF LA. RNA-sequencing performed sequentially on 2 batches: batch 1: symbols with black outline on PCA, n = 9 individual RA (n = 6 NF, n = 3 HF) and n = 9 individual LA (n = 6 NF, n = 3 HF). Batch 2: symbols without black outline on PCA, n = 8 individual RA (n = 3 NF, n = 5 AF+HF) and n = 10 individual LA (n = 5 NF, n = 5 AF+HF). Transcripts with a fold change threshold of no more than 0.667 or at least 1.5 and false discovery rate (FDR) less than 0.05 were considered statistically significant.