A Potential Role of Esophageal Cancer Related Gene-4 for Atrial Fibrillation

Abstract

Epidemiological studies have shown a strong correlation between tumor and AF. However, the molecular link between tumor and AF remains unknown. ECRG4, a tumor suppressor gene that is expressed in the A-V node and in sporadic ventricular myocytes, inhibits tumorigenesis and monitors tissue homeostasis by functioning as a 'sentinel' molecule gauging inflammatory and cell proliferative responses. To explore the potential physiological function of Ecrg4 in heart, we evaluated its distribution in heart, analyzed its expression in patients with persistent AF and in a canine AF model, and dissected the molecular events downstream of Ecrg4. The results showed that the level of Ecrg4 expression is homogenously high in atria and the conduction systems and in sporadic ventricular myocytes. Importantly, the expression of Ecrg4 was significantly decreased in atrial appendages of AF patients than patients with SR. Moreover, in rapid pacing canine AF models, the expression of ECRG4 in atria was significantly decreased compared to that of the controls. Mechanistically, knockdown ECRG4 in atrial myocytes significantly shortened the APDs, inhibited the expression of Gja1, and activated pro-inflammatory cascades and genes involved in cardiac remodeling. These results suggest that Ecrg4 may play a critical role in the pathogenesis of AF.

Figure 1

Distribution of Ecrg4 in adult rat heart by immunohistochemistry. Specimens of Sinus node, A-V node, left atrium, and left ventricle were dissected from adult SD rat hearts, and processed for IHC as described. Representative images showing that Ecrg4 (brown staining) is expressed in Sinus node (A), A-V node (B), left atrium (C), and sporadically in ventricular myocytes of left ventricle (D). Scale bar, 200 µm.

Figure 2

Distribution of Ecrg4 in adult rat heart by real-time PCR. Hearts were harvested from 9 adult rats, specimens of Sinus node, left and right atria (L. and R. At.), and left and right ventricles (L. and R. Vt.) were dissected, which were processed for total RNA isolation and real-time PCR as described in materials and methods. SN and L. At. express the highest level of ECRG4, followed by R. At. and both ventricles. Atria express significantly higher level of ECRG4 than ventricles (n = 9, P < 0.05), and L. At. express higher level of ECRG4 than R. At. (n = 9, P < 0.05). Data were presented as “Mean ± SD”, and the experiments were performed in triplicate and repeated three times.

Figure 3

Ecrg4 is expressed in neonatal rat cardiomyocytes. Neonatal atrial and ventricular myocytes were prepared separately as described in materials and methods, which were seeded in a 6-well plate containing coverslips for overnight. Cells were fixed, permeabilized, and proceeded for immunofluorence using anti-Ecrg4 as primary antibody, goat anti-rabbit Alexa Fluor 488 as secondary antibody, and DAPI for nuclear staining. Representative images showing relative higher level of Ecrg4 (bright green) in perinuclear region and in endoplasmic reticulum in neonatal atrial myocytes (A, left panel) compared to relative lower level of Ecrg4 (faint green) diffused in the cytoplasm of ventricular myocytes (B, right panel). Scale bar = 50 µm.

Figure 4

Expression of Ecrg4 is down-regulated in appendages of AF patients. Specimens of atrial appendage from right atrium were harvested from 5 RHD patients with persistent AF (A, labeled AF1–5 on left panel) and 5 RHD patients with SR (B, labeled SR1–5 on right panel) respectively. The specimen were fixed, microtomed, and proceeded for IHC using anti-Ecrg4 as primary antibody and goat anti-rabbit-HRP as secondary antibody per instructions in materials and methods. Representative images (with 200x magnification on the left side and 400x magnification on the right side of A and B respectively) showing a light brown staining of cardiomyocytes (low level of Ecrg4) in all 5 specimens of atrial appendages from patients with AF (A, left panel) relative to a dark brown staining of cardiomyocytes (high level of Ecrg4) in all 5 specimens of atrial appendages from patients with SR (B, right panel).

Figure 5

Establishment of canine AF model and expression of Ecrg4 in atria. AF was induced in canines using rapid atria pacing as described in materials and methods. AF was confirmed using surface and intracardiac electrograms showing a normal regular P and atrial waves (A) in a control dog relative to fast and irregular atrial waves (B) in an AF model dog. When gene expression was analyzed by real-time PCR (C), ECRG4 was significantly down-regulated (n = 8, P < 0.05) in AF models (AF, black bar) compared to controls in SR (SR, Open Bar). Data were presented as “Mean ± SD”, and the experiments were performed in triplicate and repeated three times.

Figure 6

Knockdown ECRG4 in neonatal atrial myocytes shortens APD. Neonatal atrial CMs were transduced with lentivirus expressing Ecrg4-siRNA or luciferase-siRNA for 36 hours and the knock down of ECRG4 was confirmed by real-time PCR showing a 74% decreased ECRG4 expression in Ecrg4-siRNA (black bar) relative to Luc-siRNA transduction (open bar) (A, P < 0.05). Thirty-six hours after transduction, AP was recorded using patch-clamp on GFP positive CMs. Representative recordings showing APs in Luc-siRNA (C) and Ecrg4-siRNA (D) transduced CMs. Quantitative analysis of APD₅₀ (B, left panel) and APD₉₀ (B, right panel) showed 53% and 81% decrease, respectively, in Ecrg4-siRNA (black bars) relative to Luc-siRNA (open bar) transduced atrial myocytes (n = 8, P < 0.05 in both APD₅₀ and APD₉₀). Data were presented as “Mean ± SD”, and the knockdown experiment was performed in triplicate and repeated three times.

Figure 7

Knockdown ECRG4 affects the expression of genes implicated in inflammation and cardiac remodeling. Knockdown ECRG4 was performed as described in materials and methods and gene expression normalized to GAPDH was expressed as relative expression of the knockdown using Luc-siRNA, which was arbitrarily set as 1. Knockdown ECRG4 significantly increased the expression of IL1a, IL6, and MCP1 (A, P < 0.05 in all) but not that of NF-kb p50 (ns, non-significant) (A), inhibited the expression of Gja1 (B, P < 0.05) and activated the expression of MMP3, s100a1 and s100a8 (B, P < 0.05 in all). Data were presented as “Mean ± SD”, and the experiments were performed in triplicate and repeated three times.