The Calcineurin-FoxO-MuRF1 signaling pathway regulates myofibril integrity in cardiomyocytes

Essential revisions:

1) Prior studies in zebrafish (from the Xu and Yelon labs) have implicated cardiac contractility in the regulation of sarcomere integrity. While it is clear that calcium is overloaded in ncx1 mutants, cardiac contractility is also compromised. Is it possible that the latter change could contribute to the myofibril integrity defects? How might it be possible to distinguish the relative contributions of reduced contractility and calcium overload? Is there any scenario that would allow analysis of the calcineurin-FoxO-Murf1 pathway in the context of calcium overload with normal contractility (or is this impossible to achieve)? At a minimum, is it possible to demonstrate that the calcineurin-FoxO-Murf1 pathway is not triggered by reduction of contractility (perhaps in silent heart mutants, or in blebbistatin treated embryos)? Or, since murf1 morpholinos and MG132 can restore myofibril integrity in ncx1-deficient hearts, is it informative to examine cardiac rhythm in these rescued hearts?

To address the reviewer’s question, we provide a new video showing that ncx1h/murf1a/murf1b triple deficient hearts exhibit a chaotic fibrillation-like movement similar to ncx1h hearts (Video 3). This study demonstrates that knockdown of Murf1 preserves sarcomere integrity without restoring cardiac contraction in ncx1h deficient embryos, indicating that the regulation of contractility and sarcomere integrity can be uncoupled (subsection “Blocking MuRF1-induced protein degradation preserves myofibril integrity in ncx1h mutant hearts”).

2) Another important issue relates to Ca²⁺ compartmentalization. Distinct patterns of Ca²⁺ signals are generated in space and time to stimulate specific cellular responses. Is the Ca²⁺ that controls muscle contraction also involved in activating calcineurin?

We apologize for the confusion. In this revised manuscript, we have stated that both ncx1h deficiency (Introduction) and A23187 treatment (subsection “Ca²⁺ induces MuRF1a expression”) result in a global Ca²⁺ homeostasis defect in affected cells.

3) Figure 4: Comparing the ventricular morphology in Figure 1C/3A vs. Figure 4B, the ventricle in Tg:murf1 embryos was relatively normal in comparison to what is seen in a ncx1 mutant. Does Tg:murf1 fully recapitulate the phenotype of the ncx1 mutant? Quantification of the level of overexpression of the transgene is likely to help with the interpretation of these data. Along the same lines, the authors state that murf1 overexpression led to impaired cardiac function. However, Figure 4E only shows a reduced atrial fraction shortening, instead of ventricular fraction shortening. Can the authors provide an assessment of ventricular performance in Tg:murf1 embryos?

Murf1 overexpression recapitulates the sarcomeric disassembly phenotype present in tre/ncx1h hearts, but does not induce the severe morphological defects that are likely attributable to the complex cellular responses induced by aberrant Ca²⁺ handling and the failure of tre/ncx1h hearts to initiate a heartbeat (Langenbacher et al., 2005, Shimizu et al., 2015).

Overexpression of Murf1 affects both atrial and ventricular performance (new Videos 1 and 2). The quantitative data on ventricular fractional shortening (VFS) is now presented in the text (subsection “MuRF1 regulates myofibril integrity in cardiomyocytes”).

4) Figure 10: Figure 10A demonstrates that FoxO translocates into the nuclei of some ncx1 mutant cardiomyocytes, yet it seems that many of the ncx1 mutant cardiomyocytes exhibit a decreased level of FoxO or no FoxO. Please clarify.

FoxO is regulated by its subcellular localization. As current FoxO antibodies are not suitable for whole mount immunostaining in zebrafish, we injected a small amount of DNA encoding FLAG-tagged FoxO3a into zebrafish embryos and used the FLAG tag as a proxy to assess the subcellular localization of FoxO. This procedure creates a mosaic heart with a small number of cardiomyocytes expressing FLAG-tagged FoxO. We have revised the text to clarify the experimental design (subsection “Cn and FoxO regulate MuRF1 expression in the heart”).

5) Validation of reagents: A pair of morpholinos are employed in Figure 5 to knockdown murf1a and murf1b function. The sequences of the morpholinos are provided in Table 1, but no validation of these reagents is described. Considering that the morpholinos improve sarcomere integrity in the ncx1 mutants, there is no major concern that they are causing general toxicity, but it would be helpful to describe whatever controls the authors performed for these tools.

In addition to examining sarcomeric integrity in tre/ncx1h mutant hearts that received control morpholino, we assessed the amount of Murf1 protein in control embryos and morphants by Western blotting. We found that control MO cannot preserve the sarcomeric structure in ncx1h deficient hearts (New Figure 5A) and that the protein level of Murf1 is reduced by 27% in Murf1 morphants (New Figure 5C, and subsection “Blocking MuRF1-induced protein degradation preserves myofibril integrity in ncx1h mutant hearts”).

6) Validation of reagents: Figure 10A employs a FoxO antibody, but there is no description of this reagent in the Materials and methods. Where does this antibody come from, and how has it been validated?

As explained above, we used the myl7:FLAG-foxo3a-IRES-EGFP transgene as a proxy to assess the subcellular localization of FoxO. For this purpose, we used an anti-FLAG antibody for immunostaining (Sigma-Aldrich). This information is presented in the “Materials and methods” section.