Arginylation regulates myofibrils to maintain heart function and prevent dilated cardiomyopathy

Abstract

Protein arginylation mediated by arginyltransferase (ATE1) is essential for heart formation during embryogenesis, however its cell-autonomous role in cardiomyocytes and the differentiated heart muscle has never been investigated. To address this question, we generated cardiac muscle-specific Ate1 knockout mice, in which Ate1 deletion was driven by α-myosin heavy chain promoter (αMHC-Ate1 mouse). These mice were initially viable, but developed severe cardiac contractility defects, dilated cardiomyopathy, and thrombosis over time, resulting in high rates of lethality after 6months of age. These symptoms were accompanied by severe ultrastructural defects in cardiac myofibrils, seen in the newborns and far preceding the onset of cardiomyopathy, suggesting that these defects were primary and likely underlay the development of the future heart defects. Several major sarcomeric proteins were arginylated in vivo. Moreover, Ate1 deletion in the hearts resulted in a significant reduction of active and passive myofibril forces, suggesting that arginylation is critical for both myofibril structural integrity and contractility. Thus, arginylation is essential for maintaining the heart function by regulation of the major myofibril proteins and myofibril forces, and its absence in the heart muscle leads to progressive heart failure through cardiomyocyte-specific defects.

Figure 1. Ate1 knockout in the heart results in dilated cardiomyopathy and postnatal lethality

(a) Survival curve of wild type (WT) and αMHC-Ate1 (CKO) mice. 61 WT and 131 CKO mice were used to derive the P0 100% values. (b) Heart/Body weight ratio of WT and CKO mice at the age of 10–12 months. Columns are significantly different from each other (p=0.001). 14 WT and 24 CKO were analyzed (error bars represent SEM). (c) Left panel: whole hearts fixed with formalin; Center panel: the hearts were cut in half to observe the inside; Right panel: heart sections stained with hematoxylin and eosin (H&E). 10 WT and 18 CKO mice were analyzed. (d) Left panels: echocardiograms of WT (top) and CKO (middle and bottom) mice at 3 (left) and 12 (middle) months of age. Right panel: electrocardiogram (ECG) of WT (top) and CKO (middle and bottom) mice at 12 month of the age (12 m). Double-headed arrows: PQR distance; arrowheads: S waves; arrows: T waves. 6 WT and 6 CKO mice at 3 months and 4 WT and 7 CKO at 12 months were analyzed.

Figure 2. Comparison of the ultrastructure of cardiac muscle between WT and CKO hearts by electron microscopy

Distribution of cardiomyocytes in adult (~12 months of age) and newborn (P0) wild type (WT) and αMHC-Ate1 (CKO) mouse hearts observed at different magnifications as indicated on the left-hand side for each group of images. 4 and 2 age-matched pairs of WT and CKO mice were analyzed in adult and newborn mice, respectively. Bars, 10μm in a– c, 500 nm in d–m.

Figure 3. Ate1 knockout in the heart affects some of the myofibril proteins

(a) Western blotting of heart muscle homogenates probed with antibodies to the proteins arginylated in cardiac myofibrils in WT and CKO newborn (n) and adult mice (2 mice in each group). Ate1: arginyltransferase; Actn2: α-actinin 2; Tnnt2: cardiac troponin T2; Tpm: Tropomyosin; Myl3: myosin light polypeptide 3. (b) Left: Coomassie blue gels of total heart homogenates (Ex) and isolated myofibrils (MF) from WT and CKO adult mice, showing that the levels of actin and myosin heavy chain (MHC) are overall similar in these preparations. Right: immunoblots of the wild type heart homogenate and myofibril preparations with anti-Ate1 shows that Ate1 is highly prominent in the heart and directly associates with cardiac myofibrils. (c) Localization of tropomyosin and α-actinin 2 in cardiac myofibrils. Left panel; localization of tropomyosin (Tpm); center panel: localization of α-actinin 2 (Actn2); right panel: merged pictures (Merged); top panel: Tpm appears as single bands in WT mice; middle panel: Tpm appears as double bands in WT mice; bottom panel: Tpm appears as single bands co-localized with Actn2 in CKO mice. (d) Ratio of myofibrils in which the Tpm and Actn2 distributed as single bands (as shown in (c), top panel), double bands (as shown in (c), middle panel) or co-localized (as shown in (c), bottom panel) in 2 WT and 3 CKO mice.

Fig. 4. Ate1 knockout in the heart results in reduction of active and passive myofibril forces

(a) Typical force measurements performed with two myofibril samples isolated from aged WT (black) and CKO (red) hearts. In both cases, myofibrils were activated at an initial SL of 2.2μm and after full activation the force was maintained stable until the myofibrils were relaxed. Note that the absolute force is significantly decreased in the mutant heart. (b) Consecutive stretches performed with non-activated myofibrils isolated from control (black) and mutant (red) hearts. Stretches were started at 1.8μm and were performed at magnitudes of 0.2μm at a speed of 10μm/1sec. Mean (± S.E.M.) values for (c) total active forces and (d) passive forces for myofibrils tested during these experiments. Multiple myofibrils from 4 age-matched control and αMHC-Ate1 mice at 415 and 485 days old (2 pairs at each age) were analyzed.

Figure 5. Model of the regulation of cardiac formation and function by protein arginylation

Lack of arginylation causes dilated cardiomyopathy associated with abnormal and supercontracted sarcomeres.