Vertebrate CASTOR is required for differentiation of cardiac precursor cells at the ventral midline

Abstract

The CASTOR (CST) transcription factor was initially identified for its role in maintaining stem cell competence in the Drosophila dorsal midline. Here we report that Xenopus CST affects cardiogenesis. In CST-depleted embryos, cardiomyocytes at the ventral midline arrest and are maintained as cardiac progenitors, while cells in more dorsal regions of the heart undergo their normal program of differentiation. Cardia bifida results from failed midline differentiation, even though cardiac cell migration and initial cell fate specification occur normally. Our fate mapping studies reveal that this ventral midline population of cardiomyocytes ultimately gives rise to the outer curvature of the heart; however, CST-depleted midline cells overproliferate and remain a coherent population of nonintegrated cells positioned on the outer wall of the ventricle. These midline-specific requirements for CST suggest the regulation of cardiomyocyte differentiation is regionalized along a dorsal-ventral axis and that this patterning occurs prior to heart tube formation.

Figure 1. CST Is Required for Vertebrate Heart Development

(A) Predicted schematic representation of CSTα and CSTβ proteins; nuclear localization signal (yellow), zinc finger repeats (red), serine-rich region (red). (B–F) Whole-mount in situ analysis of Stage 27 (early tailbud), Stage 32 (tailbud), and Stage 36 (early tadpole) embryos using a Cst-specific probe common to Cstα and Cstβ. ([B], lateral view with anterior to the left; [C and D] ventral and dorsal views, respectively, with anterior to the top). (D and F) Transverse sections of whole-mount in situ Stage 36 embryos through (D) the heart and (F) the hindbrain: hindbrain (hb), somites (s), heart primordium (hp), heart (h), myocardium (m), endocardium (en), commissural neurons (c). (G–J) Representative (G and I) control MO and (H and J) CstMO embryos. (G) Stage 32 control MO and (H) CstMO embryos are indistinguishable. (I) Stage 41 control MO and (J) CstMO embryos. CST-depleted embryos present with dorsal fin edema and no gross ventral region abnormalities. (K) RT-PCR analysis of Stage 42 tadpoles injected at the one-cell stage with the CstMO demonstrating inhibition of proper slicing of Cst pre-mRNA. Control MO (Con MO) and 5-mismatch MO (5-mis MO) are negative controls. (L) Whole-mount MHC antibody staining of tadpole Stage 37 CST-depleted embryos (lateral views with anterior to the left); inflow tract (i), ventricle (v), outflow tract (o). Scale bars: (B–C) = 0.5 mm, (G) = 1 mm, (D and L) = 100 μm.

Figure 2. CST Is Required for Cardiomyocyte Differentiation at the Ventral Midline

(A–L) Whole-mount in situ analysis with early cardiac markers Nkx2.5 (A, B, G, and H), Tbx5 (C, D, I, and J), and Tbx20 (E, F, K, and L) of tailbud Stage 26 and 29 control and CST-depleted embryos (ventral view with anterior to the top). Cardiac progenitors have properly migrated and completely fused across the ventral midline. (M–U) Whole-mount MHC antibody staining at Stage 29 (onset of cardiac differentiation), Stage 32 (completion of linear heart tube formation), and Stage 37 (chamber formation) (ventral view with anterior to the top). (M) Stage 29 control MO embryos and (N and O) CST-depleted embryos. (P) Stage 32 control MO embryos and (Q and R) CST-depleted embryos display varying degrees of cardia bifida of the linear heart tube upon CST depletion. (S) Stage 37 control MO embryos and (T and U) CST-depleted embryos display morphological consequences of CST depletion on chamber formation. (V–Y) Whole-mount Tmy antibody staining of Stage 29 and 32 (V and X) control MO and (W and Y) CST-depleted embryos demonstrates that lack of differentiation is not specific to MHC. (Z–B′) Simultaneous detection of cardiac progenitor cells and differentiated cardiac cells in a Stage 29 CST-depleted embryo. (Z and A′) Whole-mount double in situ analysis using a Nkx2.5-specific probe (pink) to mark cardiac progenitor cells and Cardiac troponin I-specific probe (blue) to mark differentiated cardiac cells in (Z) control MO and (A′) CST-depleted embryos. (B′) Magnified image of the cardiac region in the CST-depleted embryo in (A′). (C′ and D′) Transverse sections of Stage 29 (C′) control MO and (D′) CST-depleted embryos stained with MHC antibody and DAPI. Brackets highlight the lack of differentiation at the ventral midline. (E′) Quantification of differentiated cardiomyocytes determined by counting the total MHC-positive cells derived from serial sectioned embryos. Bars represent the average of at least six embryos per condition ± SEM; *p < 0.01. Representative images are derived from a single experiment, and all experiments were repeated at least twice with independent batches of embryos. Scale bars: (G) = 0.5 mm, (S and X) = 100 μm, (C′) = 200 μm.

Figure 3. CST Is Not Required for Formation or Patterning of Endodermal Tissue

(A) Schematic representation of endodermal tissue markers that demarcate pharyngeal endoderm (Sox2 and Endodermin), ventral midgut (Vito and Endodermin), and posterior endoderm (Endocut). (B) Relative expression levels of endodermal markers Sox2, Vito, Endodermin, and Endocut in Stage 29 CST-depleted embryos (n = 5) relative to control MO embryos (n = 5) using GAPDH as the housekeeping gene. Bars represent the relative expression levels ± SEM. (C–F) Whole-mount in situ analysis of endodermal markers (C) Sox2, (D) Vito, (E) Endodermin, and (F) Endocut in Stage 29 (top) control MO and (bottom) CST-depleted embryos (lateral views with anterior to the left). (G–H) In situ analysis of endodermal and cardiac markers on adjacent transverse sections through the cardiac region of (top) control MO and (bottom) CST-depleted Stage 29 embryos. (G) Sox2 and Nkx2.5 expression on adjacent sections demonstrating proper expression within pharyngeal tissue of CST-depleted embryos. (H) Endodermin and Tbx20 expression on adjacent sections demonstrating proper relative spatial expression within the cardiac tissue and endoderm of the embryo.

Figure 4. Fate Mapping Cardiac Ventral Midline Cells

(A) Bright-field image of a living cardiac actin-GFP transgenic embryo injected with MitoTracker at Stage 29 along the ventral midline 5.5 mm posterior to the cement gland (CG). (B) Fluorescent image of the same embryo demonstrating the location of incorporated MitoTracker into cells at the ventral midline (ventral views with anterior to the top). Fluorescence anterior to site of injection is reflection off the surface of the live embryo. (C and D) Tabulation of the location of MitoTracker-labeled cardiac cells of control MO and CST-depleted embryo at (C) midtailbud Stage 35 and (D) tadpole Stage 45. Images of CA-GFP transgenic control MO-injected and CST-depleted (E–H, J–M, and O–U) Stage 35 and (V–Y and A′–D′) Stage 45 dissected hearts. (F, K, P, T, U, and Z) Corresponding images of GFP expression. (G, L, Q, X, and C′) Corresponding images of fated MitoTracker-labeled cardiac ventral midline cells. (H, M, R, U, Y, and D′) Merged images of GFP and fated MitoTracker-labeled ventral midline cells. (T and U) Note the fated ventral midline cells in a pocket of undifferentiated (GFP-negative) cardiomyocytes. (I, N, S, Z, and E′) Schematics representing fate of the cardiac ventral midline cells to the outer curvature of the ventricle in (I) Stage 35 and (Z) Stage 45 control MO-injected hearts. CST-depleted fated ventral midline cells located in the (N) posterior midline or (S) in an undifferentiated cleft in the outer ventricular myocardium in Stage 35 CST-depleted hearts and (E′) in a condensed mass of cells on the outer ventricle in Stage 45 CST-depleted hearts.