SUMMARYThe zebrafish heart has the capacity to regenerate after ventricular resection. Although this regeneration model has proved useful for the elucidation of certain regeneration mechanisms, it is based on the removal of heart tissue rather than its damage. Here, we characterize the cellular response and regenerative capacity of the zebrafish heart after cryoinjury, an alternative procedure that more closely models the pathophysiological process undergone by the human heart after myocardial infarction (MI). Localized damage was induced in 25% of the ventricle by cryocauterization (CC). During the first 24 hours post-injury, CC leads to cardiomyocyte death within the injured area and the near coronary vasculature. Cell death is followed by a rapid proliferative response in endocardium, epicardium and myocardium. During the first 3 weeks post-injury cell debris was cleared and the injured area replaced by a massive scar. The fibrotic tissue was subsequently degraded and replaced by cardiac tissue. Although animals survived CC, their hearts showed nonhomogeneous ventricular contraction and had a thickened ventricular wall, suggesting that regeneration is associated with processes resembling mammalian ventricular remodeling after acute MI. Our results provide the first evidence that, like mammalian hearts, teleost hearts undergo massive fibrosis after cardiac damage. Unlike mammals, however, the fish heart can progressively eliminate the scar and regenerate the lost myocardium, indicating that scar formation is compatible with myocardial regeneration and the existence of endogenous mechanisms of scar regression. This finding suggests that CC-induced damage in zebrafish could provide a valuable model for the study of the mechanisms of scar removal post-MI.

No abstract

Vertebrate limbs grow out from the flanks of embryos, with their main axis extending proximodistally from the trunk. Distinct limb domains, each with specific traits, are generated in a proximal-to-distal sequence during development. Diffusible factors expressed from signalling centres promote the outgrowth of limbs and specify their dorsoventral and anteroposterior axes. However, the molecular mechanism by which limb cells acquire their proximodistal (P-D) identity is unknown. Here we describe the role of the homeobox genes Meis1/2 and Pbx1 in the development of mouse, chicken and Drosophila limbs. We find that Meis1/2 expression is restricted to a proximal domain, coincident with the previously reported domain in which Pbx1 is localized to the nucleus, and resembling the distribution of the Drosophila homologues homothorax (hth) and extradenticle (exd); that Meis1 regulates Pbx1 activity by promoting nuclear import of the Pbx1 protein; and that ectopic expression of Meis1 in chicken and hth in Drosophila disrupts distal limb development and induces distal-to-proximal transformations. We suggest that restriction of Meis1/Hth to proximal regions of the vertebrate and insect limb is essential to specify cell fates and differentiation patterns along the P-D axis of the limb.

Respiratory chain complexes can super-assemble into quaternary structures called supercomplexes that optimize cellular metabolism. The interaction between complexes III (CIII) and IV (CIV) is modulated by supercomplex assembly factor 1 (SCAF1, also known as COX7A2L). The discovery of SCAF1 represented strong genetic evidence that supercomplexes exist in vivo. SCAF1 is present as a long isoform (113 amino acids) or a short isoform (111 amino acids) in different mouse strains. Only the long isoform can induce the super-assembly of CIII and CIV, but it is not clear whether SCAF1 is required for the formation of the respirasome (a supercomplex of CI, CIII and CIV). Here we show, by combining deep proteomics and immunodetection analysis, that SCAF1 is always required for the interaction between CIII and CIV and that the respirasome is absent from most tissues of animals containing the short isoform of SCAF1, with the exception of heart and skeletal muscle. We used directed mutagenesis to characterize SCAF1 regions that interact with CIII and CIV and discovered that this interaction requires the correct orientation of a histidine residue at position 73 that is altered in the short isoform of SCAF1, explaining its inability to interact with CIV. Furthermore, we find that the CIV subunit COX7A2 is replaced by SCAF1 in supercomplexes containing CIII and CIV and by COX7A1 in CIV dimers, and that dimers seem to be more stable when they include COX6A2 rather than the COX6A1 isoform.