Cyp26 Enzymes Facilitate Second Heart Field Progenitor Addition and Maintenance of Ventricular Integrity

Abstract

Although retinoic acid (RA) teratogenicity has been investigated for decades, the mechanisms underlying RA-induced outflow tract (OFT) malformations are not understood. Here, we show zebrafish embryos deficient for Cyp26a1 and Cyp26c1 enzymes, which promote RA degradation, have OFT defects resulting from two mechanisms: first, a failure of second heart field (SHF) progenitors to join the OFT, instead contributing to the pharyngeal arch arteries (PAAs), and second, a loss of first heart field (FHF) ventricular cardiomyocytes due to disrupted cell polarity and extrusion from the heart tube. Molecularly, excess RA signaling negatively regulates fibroblast growth factor 8a (fgf8a) expression and positively regulates matrix metalloproteinase 9 (mmp9) expression. Although restoring Fibroblast growth factor (FGF) signaling can partially rescue SHF addition in Cyp26 deficient embryos, attenuating matrix metalloproteinase (MMP) function can rescue both ventricular SHF addition and FHF integrity. These novel findings indicate a primary effect of RA-induced OFT defects is disruption of the extracellular environment, which compromises both SHF recruitment and FHF ventricular integrity.

Fig 1. Cyp26-deficient embryos fail to add ventricular cardiomyocytes.

(A–F) Control (Ctrl) and Cyp26-deficient Tg(myl7:DsRed2-NLS) embryo hearts at 36 (A,D), 48 (B,E), and 72 hpf (C,F). Ventricles are red only (brackets) and atria are green (AMHC immunostaining). (G) Graph indicating cardiomyocyte counts at 36 (n = 10 for control and n = 8 for Cyp26 deficient), 48 (n = 15 per group) and 72 hpf (n = 15 per group). (H) Graph indicating fold difference of mRNA relative to β-actin assayed with RT-qPCR for the cardiac differentiation markers myl7, vmhc, and amhc. (I,J) Confocal images of hearts from Tg(myl7:Kaede) embryos at 48 hpf that were photoconverted at 36 hpf. Brackets indicate ventricular addition (green only cells) to the hearts. Arrows denote cardiomyocytes outside of the heart tube. (K) Graph depicting quantification of the amount of ventricular addition to the OFT (n = 8 per group). Frontal views, anterior up (A–F); n > 20 embryos per group (A-F,I,J). Error bars are standard error of the mean (SEM), asterisks denote p < 0.05 compared to controls by Student’s t test. Scale bar: 50 μm.

Fig 2. SHF progenitors do not add to the heart tube in Cyp26-deficient embryos.

(A–H) ISH of control and Cyp26-deficient Tg(nkx2.5:ZsYellow) embryos at 24 (A,E), 30 (B,F), 36 (C,G), and 48 (D,H) hpf. Nkx2.5:zsyellow (red) and myl7 (purple). (A–C,E–G) Brackets indicate nkx2.5:ZsYellow+ cells in the first nkx2.5+ pharyngeal arch. (F,G) Arrows indicate accumulation of nkx2.5+ cells adjacent to the arterial pole of the heart. (I,I’,K,K’) Tg(nkx2.5:Kaede) before and after photoconversion of the anterior lateral population of nkx2.5+ cells (bracket) in control and Cyp26-deficient embryos at 24 hpf. (J–J”,L–L”) Position of photoconverted cells (red, arrows) relative to the heart tube (green) in control and Cyp26-deficient embryos at 48 hpf. Images in I–J” and K–L”, respectively, are of the same control and Cyp26-deficient embryos. (M) Schematic depicting the three regions where photoconverted nkx2.5:Kaede+ cells contributed at 48 hpf. (N) Graph depicting percentage of contribution to the third and fourth PAAs, non-ventricular OFT, and ventricular cardiomyocytes (control-lateral n = 39, control-adjacent n = 8, Cyp26-deficient–lateral n = 52, Cyp26-deficient–adjacent n = 5). (O,P) Confocal images of arch arteries in Tg(nkx2.5:Kaede; kdrl:nlsEGFP) control and Cyp26-deficient embryos. (Q) Graph depicting quantification of the endothelial cell number in the third and fourth PAAs in control (n = 18) and Cyp26-deficient (n = 24) embryos. Ventral view, anterior up (A–H); dorsal view, anterior up (I,I’,K,K’); lateral view, anterior up (J,J’,J”,L,L’,L”); lateral view, anterior right (O,P). n > 20 per group for (A–H). Asterisk denote p < 0.05 by Chi Squared test. H, heart. Scale bar: 50 μm.

Fig 3. Ventricular cardiomyocytes can exit the heart tube in Cyp26-deficient embryos.

(A,B) 30-min interval frames from confocal time-lapse movies of control and Cyp26-deficient Tg(myl7:EGFP) hearts. Images of hearts were depth-coded with the spectrum ranging from pink at 120 μm to blue at 0 μm. (C,D) Hematoxylin–eosin (HE) stained frontal sections of the hearts from control and Cyp26-deficient embryos. Endocardium (white arrowheads) and myocardium (black arrowheads). Control hearts n = 4. Cyp26-deficient hearts n = 5. (E,F) Control and Cyp26-deficient Tg(myl7:EGFP);Tg(kdrl:mCherry) embryos at 48 hpf. Endocardium (red) and myocardium (green). Arrowheads indicate the inner border of endocardial and outer border of myocardial cells. (G) Graph showing the distance between the endocardial and myocardial layers (control n = 5, Cyp26 deficient n = 7). (H–I”‘) Confocal images of control and Cyp26-deficient Tg(myl7:EGFP) hearts stained for zonula occludens 1 (ZO1) (red) and green fluorescent protein (GFP) (green), with schematized outlines of cell boundaries and ZO1 staining. Arrow denotes cardiomyocyte protruding into the pericardial space. (J) Graph depicting the percentage of ZO1 expression along the height of cardiomyocytes (control n = 15, Cyp26 deficient n = 15). (K) Graph depicting circularity measurement of ventricular cells (control n = 15, Cyp26 deficient n = 30). (L–M”‘) Confocal images of control and Cyp26-deficient Tg(myl7:EGFP) hearts stained for β-catenin (red) and GFP (green), with schematized outlines of cell boundaries and β-catenin staining. Error bars are SEM, asterisk denotes p < 0.05 by Student’s t test. Frontal views, anterior up (A–F,H,I,L,M); n > 20 embryos per group (E,F,H,I,L,M). A, apical; B, basal. Scale bar for C–M’: 50 μm. Scale bar for H”–M”‘: 25 μm.

Fig 4. Restoring FGF signaling rescues SHF addition but not ventricular integrity.

(A,B) Graphs indicating fold difference of mRNA relative to β-actin assayed with RT-qPCR of fgf8a expression in whole embryos and isolated hearts at 48 hpf. (C) Graph depicting quantification of ventricular addition to the OFT (control n = 8, CA-Fgfr n = 6, Cyp26 deficient n = 10, Cyp26 deficient+CA-Fgfr n = 6). (D) Graph of ventricular cardiomyocyte counts at 48 and 72 hpf (n = 10 per group). (E–H) Confocal images of optical slices from hearts of Tg(myl7:Kaede) and Tg(myl7:Kaede);Tg(hsp70:ca-fgfr1) embryos at 48 hpf after heat-shock at 24 hpf and photoconversion at 36 hpf. Brackets indicate added ventricular cells (green only). (I–L) IHC for hearts from Tg(myl7:DsRed-NLS) and Tg(myl7:DsRed-NLS);Tg(hsp70:ca-fgfr1) at 48 hpf after heat-shock at 24 hpf. Arrows indicate ectopic cardiomyocytes. Error bars are SEM, asterisk denotes p < 0.05 by Student’s t test. Frontal views with anterior up (E–L); n > 20 embryos per group (E–L). Scale bar: 50 μm.

Fig 5. Attenuating MMP function restores SHF addition and ventricular integrity.

(A,B) Graphs indicating fold difference of mRNA relative to β-actin assayed with RT-qPCR of mmp9 expression in whole embryos and isolated hearts at 48 hpf. (C) Graph showing quantification of ventricular addition to the OFT (control n = 11, GM6001 treated n = 11, Cyp26 deficient n = 12, Cyp26 deficient + GM6001 n = 12). (D) Graph depicting ventricular cardiomyocyte counts at 48 and 72 hpf (n = 10 per group). (E–H) Confocal images of optical slices from hearts of Tg(myl7:Kaede) embryos at 48 hpf after photoconversion at 36 hpf. Brackets indicate ventricular addition (green only). (I–L) IHC of hearts from control and Cyp26-deficient Tg(myl7:DsRed-NLS) embryos after DMSO or GM6001 treatment. Arrows indicate ectopic cardiomyocytes. (M) Graph depicting the percentage of embryos with ectopic cardiomyocytes in Cyp26 deficient treated with DMSO or GM6001 at 48 hpf. (N) Graph depicting average number of ectopic cardiomyocytes (per embryo with ectopic cardiomyocytes) at 48 hpf. (O,P) Confocal images of optical slices through control (lineage tracer alone) or activated MMP9-injected Tg(myl7:EGFP);Tg(kdrl:mCherry) hearts at 48 hpf. Arrowheads denote the inner border of endocardial and outer border of myocardial cells. (Q) Graph depicting the quantification of the distance between the endocardium and myocardium (control n = 12, MMP9 injected n = 10). (R) Graph depicting the percentage of embryos with linearized, dysmoprhic hearts and cells outside the heart (n > 50 per group). (S) Graph indicating fold difference of mRNA relative to β-actin assayed with RT-qPCR of fgf8a expression at 48 hpf in control and Cyp26-deficient embryos treated with DMSO or GM6001. (T) Graph indicating fold difference of mRNA relative to β-actin assayed with RT-qPCR of mmp9 expression at 48 hpf after heat-shock induction of CA-Fgfr1. (U) Model of Cyp26 enzyme function in the ventricular OFT development. Red indicates reagents used to inhibit function. Blue indicates reagents used to activate function. Controls in C–E, I, and S were DMSO treated. Controls in O, Q, R indicate Cascade blue-dextran injected alone. Error bars are SEM, asterisks denote p < 0.05 compared to controls by Student’s t test (A–D,N,Q–T), asterisk denotes p<0.05 by Chi Squared test (M). Frontal views, anterior up (E–L,M,N); n > 20 embryos per group (E–L,M,N). CMs, cardiomyocytes. Scale bar: 50 μm.