A Novel Transgenic Mouse Model of Cardiac Hypertrophy and Atrial Fibrillation

Abstract

Cardiac hypertrophy is a major risk factor for the development of atrial fibrillation (AF). However, there are few animal models of AF associated with cardiac hypertrophy. In this study, we describe the in vivo electrophysiological characteristics and histopathology of a mouse model of cardiac hypertrophy that develops AF. Myostatin is a well-known negative regulator of skeletal muscle growth that was recently found to additionally regulate cardiac muscle growth. Using cardiac-specific expression of the inhibitory myostatin pro-peptide, we generated transgenic (TG) mice with dominant-negative regulation of MSTN (DN-MSTN). One line (DN-MSTN TG13) displayed ventricular hypertrophy, as well as spontaneous AF on the surface electrocardiogram (ECG), and was further evaluated. DN-MSTN TG13 had normal systolic function, but displayed atrial enlargement on cardiac MRI, as well as atrial fibrosis histologically. Baseline ECG revealed an increased P wave duration and QRS interval compared with wild-type littermate (WT) mice. Seven of 19 DN-MSTN TG13 mice had spontaneous or inducible AF, while none of the WT mice had atrial arrhythmias (p<0.05). Connexin40 (Cx40) was decreased in DN-MSTN TG13 mice, even in the absence of AF or significant atrial fibrosis, raising the possibility that MSTN signaling may play a role in Cx40 down-regulation and the development of AF in this mouse model. In conclusion, DN-MSTN TG13 mice represent a novel model of AF, in which molecular changes including an initial loss of Cx40 are noted prior to fibrosis and the development of atrial arrhythmias.

Figure 1.. Cardiac Enlargement and Fibrosis in DN-MSTN TG13 mice.

A. Masson’s Trichrome stain of wild-type (above) and DN-MSTN TG13 (below) hearts demonstrates ventricular hypertrophy with predominantly right atrial enlargement and fibrosis. B. In vivo magnetic resonance image of a DN-MSTN TG13 transgenic heart displays dramatic right atrial enlargement filling much of the right hemithorax (arrows). Top image: 4 chamber view. Spin refreshment due to flow is seen in the top half of the right atrium generating bright contrast. The inferior half of the right atrium (bordered by the right ventricle and the liver) is isointense with the myocardium, consistent with thrombosis. Bottom image: Short axis image at the level of the AV groove showing dramatic enlargement of the right atrium. C. AW/BW ratio (top) and VW/BW ratio (bottom) are both increased in DN-MSTN TG13 (p<0.05).

Supplemental Figure 1.. Microscopy of Ventricular Cellular Hypertrophy.

Transgenic mice (DN-MSTN TG13) display cellular hypertrophy compared with wild type littermates (WT). A. Hematoxylin and Eosin (H and E) staining. B. Wheat germ agglutinin (WGA) staining of cellular membranes shows increase cardiomyocyte area. (200x magnification)

Supplemental Figure 2.. Picosirius red staining for collagen.

Left and right atrial tissue from 16 week animals reveals a slight increase in staining for collagen in DN-MSTN TG13 mice compared with wild type in both atria. Representative 20x micrographs are display from the 3 DN-MSTN TG13 and 3 WT animals in whom these studies were performed.

Figure 2:. Atrial arrhythmias in DN-MSTN TG13 mice.

Representative ECG tracings from DN-MSTN TG13 mice and Wild Types. Abbreviations: S = surface electrocardiogram, A = atrial electrogram, V = ventricular electrogram, AP = atrial pacing. A. Representative ECG tracings with QRS complexes from WT (left) and DN-MSTN TG13 (right) animals. Note the enlarged P wave and broad QRS complex in DN-MSTN TG13 compared with WT. B. Representative surface tracing from a DN-MSTN TG13 mouse with spontaneous, sustained AF at 650 bpm. C. Sinus Rhythm in wild-type animal at 357 bpm. D. Spontaneous and sustained atrial tachycardia with ventricular rate 352 bpm, atrial rate 610 bpm. E. Atrial Tachycardia induced with extrastimuli atrial stimulation with S1/S2 of 100/80 ms.

Supplemental Figure 3:. Kv1.4, Cx40, and Cx43 protein expression in atria of DN-MSTN TG13, TG21, and WT littermates.

Shown are representative blots for protein expression of Kv1.4, Cx40, and Cx43 proteins, as well as endogenous control protein GAPDH, in atrial tissue from male mice from DN-MSTN TG13 and TG21 lines, with WT littermates as control. Below is quantification of protein expression after adjustment for endogenous control, GAPDH. Protein expression of Kv1.4, after adjustment for endogenous control, was significantly increased compared with wild type littermates (*p <0.05 compared with WT). There was no significant difference in Cx40 or Cx43 in TG21 mice (quantification not shown). Quantification and description of Cx40 and Cx43 in DN-MSTN TG13 is reported elsewhere. See Methods Section for details.

Figure 3.. Relationship of atrial size and age to arrhythmia in DN-MSTN TG13 mice.

A. AW/BW ratio is increased in DN-MSTN TG13 mice with arrhythmia compared to WT or DN-MSTN TG13 mice without arrhythmia, p<0.01 for both, p=NS for WT vs. DN-MSTN TG13 without arrhythmia. B. Increased probability of arrhythmia with increasing age in DN-MSTN TG13 mice. Shown is DN-MSTN TG13 mice divided in quartiles based on age (N=4, 5, 6, 4 for quartiles 1 – 4). On the left axis (bars) is the percentage of animals with arrhythmia in the quartile. The right axis (line) demonstrates the trend for probability of arrhythmia with increasing age based on a logistic regression model (OR 4.41, CI 1.10 – 17.8, p<0.01). The addition of an interaction term including AW/BW and age improved the model (Pseudo-R2 increased from 0.175 to 0.715, p<0.001), although AW/BW did not increase significantly with age for either group (R2=0.0191, p=NS for WT; R2=0.0314, p=NS for DN-MSTN TG13). Shown below the graph is the number of DN-MSTN TG13 animals in each quartile with arrhythmia over the total number of DN-MSTN TG13 animals in each quartile.

Supplemental Figure 4:. Gene expression of fibrotic genes.

QRT-PCR performed on atrial tissue from 6 DN-MSTN TG13 and 6 wild type animals ages 16 – 26 weeks comparing expression of various fibrosis and ECM-related genes reveals no significant increase in expression between DN-MSTN TG13 and WT hearts for overall (A), or for male mice specifically (B). Male comparison between 3 DN-MSTN TG13 and 3 wild type mice. Expression of all genes normalized to HPRT-1 as endogenous control prior to comparison between groups. None of the genes examined were significantly different between groups.

Figure 4.. Connexin and Ion Channel Expression in DN-MSTN TG13 hearts.

A. Western blot for connexins and flag protein demonstrates decreased Cx40 in TG (DN-MSTN TG13) animals compared with WT. DN-MSTN TG13 mice overall displayed decreased Cx40 expression (p<0.001) compared with WT (4B, upper panel), but not Cx43 (4B, lower panel) (p=NS). The difference between DN-MSTN TG13 mice with and without arrhythmia was not significant for Cx40 (p=NS) or Cx43 (p=NS). N=16 total animals (5 WT, 11 DN-MSTN TG13). C. QRT-PCR for RNA expression of various ion channels for DN-MSTN TG13 animals compared with WT (N=4 for each group). Only Kv1.4 displayed a statistically significant increase in expression compared with WT. ◊ Indicates p<0.05 vs WT.

Figure 5.. Supplemental Figure 5: Connexin40 distribution

IHC staining for connexin40 protein on right atrial tissue of 16 week old mice demonstrates overall decrease in connexin40 protein, but no obvious changes in distribution from cell borders (arrow) when viewed under 100x magnification. Shown: Representative images from immunostaining of 3 DN-MSTN TG13 and 3 WT animals.

Supplemental Figure 6.. Connexin40 and Connexin43 localization.

Shown is IF stain for Cx40 (red stain, Rhodamine conjugate), Cx43 (light blue, Cy5 conjugate), Pan-cadherin (yellow, FITC conjugate), and merged image of all 4 dyes for DN-MSTN TG13 and WT atria of 16 week old mice. Images on the left are at 63x magnificantion, and the right are zoom-enhanced (3x). For reference, DAPI nuclear stain is included in zoom-enhanced images on the right. Overall, Cx40 is decreased in DN-MSTN TG13 compared with WT, although the Cx40 that is present appears to be located along the cell-cell border, as shown by its colocalization with cadherin. Cx43 was not highly expressed, and was not significantly different between the groups.

Figure 7.. Connexin43 distribution

Connexin43 IHC stain of right atrial tissue from 16 week old mice. Overall connexin43 expression was lower than connexin40 in the atria we compared. However, no difference between DN-MSTN TG13 and WT was seen in distribution or expression levels. Representative images from studies of 3 DN-MSTN TG13 and 3 WT animals are shown.

Supplemental Figure 8.. Connexin expression in cardiomyocytes in vitro.

A. Representative Western blot demonstrates decreases in both Cx40 and Cx43 in MSTN- and DN-MSTN-infected cells, compared with uninfected control and GFP-infected cells. B. Quantification of Western blot data for Cx40 and Cx43. ◊ Indicates p<0.05 vs. uninfected control. N=3 for each group.