Platyfish possess a heart lacking vascularized compact myocardium.

(A) Simplified phylogenetic lineages of the fish species examined in this study, based on ref. 14. Teleosts diversified into two main branches approximately 250 million years ago (mya). The Otophysi clade encompasses more than one-third of all fish species, including zebrafish and tetras 43. Percomorphaceae is a hyperdiverse clade described as “the bush at the top” of the fish phylogenetic tree 116, comprising approximately 55% of extant teleost diversity, including perciforms, cichlids, and poeciliids117. (B) AFOG staining of longitudinal sections of platyfish and zebrafish hearts. Collagen (Col.) is stained in blue. (C-H) Fluorescence staining of hearts shown in (B). The myocardium is labeled by F-actin staining (fluorescent phalloidin). Note the presence of the compact myocardium in zebrafish (D), and its absence in the platyfish (G). The N2.261 antibody recognizes a specific myosin isoform that is nearly absent in the uninjured zebrafish heart (C-E), with the exception of a few fibers at the outflow tract (oft). In platyfish, N2.261 immunolabels the atrium, suggesting evolutionary divergence of myosin isoform composition between species. Fibronectin, an extracellular matrix protein, is detected in the bulbus arteriosus. (I-L) Alkaline phosphatase activity staining reveals a dense coronary vasculature in zebrafish hearts, which is absent in platyfish. Frames depict the magnified areas shown on adjacent panels, labeled with the corresponding letter. Abbreviations: avc, atrioventricular canal; at, atrium; ba, bulbus arteriosus; oft, outflow tract; v, ventricle. These abbreviations and labeling conventions apply throughout all figures.

The outer layer of Xiphophorus ventricle consists of a thin, collagen-rich matrix.

Picrosirius Red (PSR) staining of longitudinal heart sections from zebrafish (A–H) and platyfish (I–P). Sections were imaged under bright-field (A–D, I–L) and polarized light (E–H, M–P). For each species, higher magnification views of the outer layer of the bulbus arteriosus (B, F, J, N), valves (C, G, K, O), and the outer layer of the myocardium (D, H, L, P) are shown. Under polarized light, collagen fibers display a range of color from green to red depending on fiber thickness and extracellular matrix composition 118. In contrast to zebrafish (H), platyfish (P) exhibit a distinct fibrillar collagenous coat surrounding the ventricle.

Cryoinjured ventricles in Xiphophorus fish display transient wound bulging and permanent scarring.

(A) Schematics of heart cryoinjury procedure in platyfish. (B) Representative images of hearts from uninjured fish and at 7 days post-cryoinjury (dpci), stained with fluorescent phalloidin (green). The damaged area is identified by a weak fluorescence signal reflecting the loss of contractile cardiomyocytes. Arrowheads indicate the border between the intact myocardium and the wound. Abbreviations: at, atrium; ba, bulbus arteriosus; v, ventricle. (C) AFOG staining of transverse ventricular sections from zebrafish, platyfish, and swordtail, collected at indicated time points after cryoinjury. Intact myocardium (orange); fibrin and other protein deposits (red); collagen (blue). Arrowheads indicate the wound edge; double-headed arrows mark the myocardial (myo) bridge; dashed lines encircle wound tissue that has expanded beyond the normal ventricular circumference, indicative of the wound bulging phenotype.

Partial restoration of the heart in platyfish and swordtail after cryoinjury.

(A) Classification of injury phenotypes based on AFOG staining, representative examples of which are shown in Figure 3. Stacked bar plots display the percentage of each category at indicated time points after cryoinjury in platyfish and swordtail. Numbers at the base of each bar indicate the number of biological replicates (fish). Pearson’s chi-squared test with Holm’s post hoc correction: ns, not significant; *, p < 0.05. (B) Fluorescence staining of transverse platyfish heart sections at indicated time points after cryoinjury. The fibronectin-positive wound area (red) contrasts with the intact myocardium labeled by F-actin (green). Sham-operated ventricles at 30 days post-thoracotomy are shown as controls. (C-D) Quantification of the wound size and fibronectin deposition, as represented in (B). Statistical comparisons were performed using the Kruskal-Wallis test followed by Dunn’s test with Holm’s post hoc correction. Adjusted p-value: * < 0.05, ** < 0.01, *** < 0.001, **** < 0.0001. Sample sizes: n = 24 (sham), 15 (7 dpci), 20 (14 dpci), 20 (30 dpci), 13 (60 dpci), and 12 (90 dpci).

Bulk RNA sequencing reveals divergent responses to cryoinjury between zebrafish and platyfish.

(A-B) Volcano plots of transcriptomes from cryoinjured zebrafish (A) and platyfish (B) ventricles at 7 dpci, compared to uninjured controls. The log₂ fold change (log₂FC) represents the ratio of transcript abundance in cryoinjured versus uninjured conditions. Genes with a significant change in transcript levels (adjusted p-value, padj < 0.05) are highlighted in blue (decreased) or orange (increased). (C) Scatter plot comparing log₂FC values between the two species. Genes with a higher log₂FC in platyfish than in zebrafish are shown in green; those with a higher log₂FC in zebrafish are shown in purple. Highlighted genes show at least a twofold difference in transcript abundance after cryoinjury between species. (D) Comparison of gene set enrichment analyses (GSEA) between species. Color intensity reflects the statistical significance of gene set reduction (blue) or enrichment (orange) at 7 dpci relative to uninjured controls. Dot size corresponds to the normalized enrichment score (NES) for each gene set. Selected gene sets are organized into broader functional categories. (E) Transcript abundance of orthologous genes across conditions (uninjured, dark shade; cryoinjured, light shade) and species (zebrafish, ZF, purple; platyfish, PF, green). Each point represents the DESeq2-based TPM-like normalized read count (see Methods) for one biological replicate. The log₂FC and −log₁₀(padj) values correspond to those shown in (A) and (B). Genes are grouped into broader functional categories.

Delayed recruitment of immune cells during platyfish heart repair.

(A) Transcript abundance of immune cell activation-related genes across conditions and species. For details, see the legend to Figure 5 and Methods. (B-C) Immunofluorescence staining for Mpx-positive neutrophils (B) and L-plastin-positive leukocytes (C) at 7, 14, and 30 dpci. Sections were counterstained with DAPI and phalloidin-488. Control sections were obtained from uninjured ventricles of sham-operated fish at 30 days post-thoracotomy (dpt). (D-E) Quantification of Mpx-positive area (D) and L-plastin-positive area (E), corresponding to representative images shown in (B) and (C). In the cryoinjured ventricles (CI), quantifications were assessed separately in the intact myocardium (CI: Intact) and the wounded area (CI: Wound). Statistical comparisons of Sham vs. CI: Intact were performed using unpaired Wilcoxon tests, as these groups comprised different animals, whereas comparisons between the intact myocardium and wound area within the same cryoinjured heart were performed using paired Wilcoxon tests. Holm’s post-hoc correction was applied for multiple comparisons. Adjusted p-value: * P< 0.05, ** P< 0.01. (F-G) Recruitment kinetics of neutrophils (grey, left Y-axis) and L-plastin-positive leukocytes (purple, right Y-axis) in the intact myocardium (F) and wound area (G) at 1, 3, 7, 14, and 30 dpci. Sham values pooled across all time points, represent the baseline level of each marker in uninjured myocardium. Kruskal-Wallis tests were used to compare marker levels across time points.

Ventricular cryoinjury triggers transient cardiomyocyte activation.

(A) Schematic illustration of the three analyzed areas: the tropomyosin-negative wounded area, the border zone (bz) myocardium within 100 µm from the injury border, and the remote myocardium distant from the border zone. (B) Transcript abundance of orthologous genes defined as markers of cardiomyocyte fate change associated with cardiac regeneration. For further details, see the legend to Figure 5 and Methods. (C) PCNA immunolocalization in sham-operated (7 dpt and 14 dpt) and cryoinjured (7 dpci and 14 dpci) platyfish ventricles. Orange frames in the top panels indicate the regions magnified in the middle panels. Bottom panels show the corresponding PCNA channel alone. Red arrowheads indicate PCNA-positive nuclei within the myocardium. bz, border zone. (D) Quantification of proliferating cells in the Tropomyosin-positive myocardium across time points and cardiac regions. Sham versus CI: Remote myocardium comparisons were performed using unpaired Wilcoxon tests. CI: Remote myocardium versus CI: Border zone comparisons were performed using paired Wilcoxon tests, as measurements were obtained from the same hearts. Holm’s post hoc correction was applied for multiple comparisons. Adjusted p-value: * < 0.05, ** < 0.01, *** < 0.001, **** < 0.0001. (E-F) BrdU incorporation in zebrafish (E) and platyfish (F) following continuous labeling from 7 to 30 dpci (23 days) and 3 to 30 dpci (27 days), respectively. Sections were counterstained with DAPI and anti-tropomyosin antibody. Orange frames in the top panels indicate the regions magnified in the middle panels. Bottom panels display the corresponding BrdU channel alone. In zebrafish, the initial wound area is delineated based on myocardial morphology and the high density of BrdU-positive nuclei. The myocardial bridge (m.b., double-headed arrow) is observed in zebrafish but not in platyfish. For panels (C, E, F), dashed red lines encircle the wound area and dashed yellow lines demarcate the border zone.