XB-ART-811Dev Biol March 1, 2006; 291 (1): 96-109.
Retinoic acid signaling is essential for formation of the heart tube in Xenopus.
Retinoic acid is clearly important for the development of the heart. In this paper, we provide evidence that retinoic acid is essential for multiple aspects of cardiogenesis in Xenopus by examining embryos that have been exposed to retinoic acid receptor antagonists. Early in cardiogenesis, retinoic acid alters the expression of key genes in the lateral plate mesoderm including Nkx2.5 and HAND1, indicating that early patterning of the lateral plate mesoderm is, in part, controlled by retinoic acid. We found that, in Xenopus, the transition of the heart from a sheet of cells to a tube required retinoic acid signaling. The requirement for retinoic acid signaling was determined to take place during a narrow window of time between embryonic stages 14 and 18, well before heart tube closure. At the highest doses used, the lateral fields of myocardium fail to fuse, intermediate doses lead to a fusion of the two sides but failure to form a tube, and embryos exposed to lower concentrations of antagonist form a heart tube that failed to complete all the landmark changes that characterize looping. The myocardial phenotypes observed when exposed to the retinoic acid antagonist resemble the myocardium from earlier stages of cardiogenesis, although precocious expression of cardiac differentiation markers was not seen. The morphology of individual cells within the myocardium appeared immature, closely resembling the shape and size of cells at earlier stages of development. However, the failures in morphogenesis are not merely a slowing of development because, even when allowed to develop through stage 40, the heart tubes did not close when embryos were exposed to high levels of antagonist. Indeed, some aspects of left-right asymmetry also remained even in hearts that never formed a tube. These results demonstrate that components of the retinoic acid signaling pathway are necessary for the progression of cardiac morphogenesis in Xenopus.
PubMed ID: 16423341
PMC ID: PMC3539789
Article link: Dev Biol
Genes referenced: fubp1 gata4 hand1 myh4 myh6 nkx2-5 tnni3
Article Images: [+] show captions
|Fig. 1. Blocking retinoic acid signaling at stage 12/14 results in a failure to form a simple heart tube. Whole-mount in situ hybridization for cTnI was used to visualize the differentiated myocardium (blue staining). In stage 31 embryos, a normal linear heart tube that is just starting to loop (A). The tube and looping are best viewed from the ventral side of the embryo (B). When embryos are treated with 1 μM RA, the heart is severely disrupted (C), although when viewed from the ventral side, it can be seen that a tube is still formed (D). When treated with 1 μM AGN 194301, myocardial differentiation still occurs (E), but when viewed from the ventral side, the lack of tube formation is clear (F). Addition of equimolar (1 μM) concentrations of both RA and AGN 194301 resulted in an apparently normal heart tube indicating that the observed effects are not due to toxicity of the compounds used (G, H).|
|Fig. 2. Dose-dependent defects in cardiac morphogenesis caused by the retinoic acid antagonist. To better visualize the defects in cardiac morphogenesis due to addition of antagonist, αMHC expression was examined in serial sections along the anterior–posterior axis of stage 32/34 embryos subjected to varying amounts of AGN 194301. A complete tube with distinct left–right asymmetry was clearly evident throughout most of the control myocardial domain (first column, A, E, I, M, Q with anterior at the top and posterior at the bottom). Consistent with the whole-mount data, no evidence of linear heart tube formation was detected in 1 μM AGN 194301 treatments (second column, B, F, J, N, R), while a partial, relatively symmetrical heart tube could be detected within 100-nM-treated embryos (third column, C, G, K, O, S). Formation of a heart tube was observed in embryos subjected to 10 nM AGN 194301 (fourth column, D, H, L, P, T), although the fusion was not as extensive as seen in control embryos. In addition, at this stage, the left and right side appeared symmetric when compared to control embryos.|
|Fig. 3. Common morphologies of differentiated hearts from embryos treated with retinoic acid antagonists. When viewed from the ventral side, control embryos stained for αMHC expression by whole-mount in situ hybridization show a heart tube that is initiating the normal looping process (A). The most severe phenotype observed with AGN 194301 treatment was a lack of both tube formation and a ventral gap in the αMHC expression domain indicating a lack of fusion of the two heart primordia (B). Embryos that showed a fusion of the heart primordia at the ventral midline but still no tube formation (C) were seen with lower doses of antagonist. At even lower doses of antagonist, the most common heart phenotype observed was an apparent heart tube that appeared to not undergo looping at this stage (D).|
|Fig. 4. Blocking RA signaling is required in a narrow window of time for formation of the heart tube. The heart has been visualized by whole-mount in situ hybridization for cTnI and is viewed in whole cleared embryos in side view (top panel) and from the anterior end (bottom panel) in order to visualize the formation of a tube. Normal heart tube formation can be seen in stage 31/32 embryos that were treated with 2 μl/ml DMSO as a carrier control (A, G). Treatment with 1 μM RA at stage 14 resulted in a dramatic decrease of myocardial differentiation (B, H). Treatment with 1 μM AGN 194301 resulted in a block to heart tube formation remained as a sheet along the ventral midline (C, I). cTnI expression levels and spatial distribution were restored if RA and AGN 194301 were given in equimolar amounts (D, J). When embryos were subjected to an initial treatment of 1 μM AGN 194301 at stage 14 with subsequent addition of 1 μM RA at stage 16, cTnI expression (E, K) was also similar to control patterns. However, when the addition of RA was delayed until stage 18/20, cardiomyocyte differentiation was shown to be again restricted to posterior regions of the normal heart field and linear heart tube formation was not detected (F, L). Thus, blocking RA signaling in a tight window of time between stages 14 and 18 is sufficient to prevent formation of the heart tube.|
|Fig. 5. Treatment with retinoic acid antagonist alters the expression domains of Nkx2.5 and GATA-4. The expression domains of Nkx2.5 and GATA-4 are visualized by whole-mount in situ hybridization. In control embryos treated with 2 μl/ml DMSO (top panel, A–D), the normal expression domain of Nkx2.5 and GATA-4 can be seen in side view and viewed from the ventral side. Treatment with 1 μM RA caused a severe reduction in the expression pattern of Nkx2.5 (E, F) but did not cause any obvious change in the expression pattern of GATA-4 (G, H). Treatment with 1 μM AGN 194301 caused a marked change in the expression domain of Nkx2.5. As expected, the expression domain was restricted to a sheet of cells rather than a tube, and the expression was restricted to the posterior end of the region that normally expresses cTnI (I, J). In addition, there was a marked spur of Nkx2.5 expression that extended dorsally close to the level of the somites (red arrow). Treatment with AGN 194301 caused a reduction, but not elimination of the GATA-4 expression domain (K, L). If embryos were simultaneously exposed to 1 μM RA and 1 μM AGN 194301, the expression domains of Nkx2.5 and GATA-4 were similar to controls (M–P).|
|Fig. 6. Treatment of embryos with antagonist prior to gastrulation alters patterning of cardiac genes. At embryonic stage 20, Nkx2.5 is normally expressed in a broad sheet on either side of the cement gland (A). When embryos were treated with AGN 194301 prior to gastrulation (stage 6), this domain appeared expanded, particularly along the anterior–posterior axis (B). When compared to expression in control embryos (C), the expression of the lateral plate mesoderm marker, HAND1, was diminished by the same treatment (D) with shortening both from the posterior and anterior end of the embryo (anterior is left and posterior is right in both C and D). Despite the early changes in Nkx2.5 expression, the myocardium still failed to form a tube and the Nkx2.5 (E) and cTnI (F) expression domain was reduced with essentially the same phenotype as observed in embryos treated with RA antagonists post-gastrulation (see Fig. 1).|
|Fig. 7. Altering retinoic acid signaling has profound affects on the final morphology of the cardiovascular system. A normally developed was seen in stage 41 control embryos (A). Treatment with AGN 194301 resulted in a small contractile nub (B). Treatment with RA resulted in a smaller then normal contracting heart (C). Treatment with an agonist to RA results in a phenotypic heart similar to that produced by RA treatment (D). All hearts were outlined, in a relaxed phase, to better visualize heart size. (E) Altering RA signaling during early heart development results in a decreased heart rate in stage 41 embryos (n = 20, P ≤ 0.01).|
|Fig. 8. Cardiomyocytes retain an immature appearance after AGN 194301 treatment. Control embryos at stage 31 (A), stage 35 (B), stage 39 (C), and an AGN-194301-treated embryo (D) fixed at the same time as control sibling shown in C were immunolabeled and analyzed by confocal microscopy (10 RA antagonist embryos were analyzed; this embryo represents the trend seen). The black and white inset represents a compilation of all the optical sections of a z-series (60–70 4 μm sections) taken through the embryos as viewed with the muscle marker cardiac troponin I. The colored micrograph represents an optical thick section that is a subset of the inset (10 consecutive sections). Red = cardiac troponin I; green = F-22, flectin. The normal stage 31 embryo (A) shows the heart region after the left and right fields have fused but prior to tube closure. The troponin I staining reveals large columnar muscle cells characteristic of the stage. There is some flectin staining apparent on the basal side of the mesoderm as well as faint intracellular staining in the muscle cells of the left heart field in this particular subset of sections. At stage 35 (B), the heart has fused to form a tube and is beginning to undergo S-shaped looping. The muscle cells have lost their columnar appearance, and muscle striations are visible. Flectin is distributed in a subset of muscle cells as well as both the basal and apical surface of the endocardium. The normal stage 39 embryo heart (C) is viewed just prior to trabeculation; the striations more evident. The majority of the flectin staining is seen in the ECM. The 1 μM AGN-194301-treated embryos have unfused heart fields. They have not formed tubes but as they develop do not remain as flat sheets. The muscle cells have the columnar appearance of a less mature stage 31 embryo. The flectin staining has a temporal characteristic more similar to stage 39 with the majority of the localization in the ECM.|
|Fig. 9. Elements of the normal looping process are evident in the non-tubular hearts of RA-antagonist-treated embryos. Normal stage 33 embryos and RA-antagonist-treated embryos were fixed, immunolabeled with cardiac troponin I, imaged by confocal microscopy, and analyzed. Consistent characteristics among the groups were noted and are represented in the images of the individuals shown. Images A–D are a series of optical “thick sections”(ventral view) taken through a stage 33 normal Xenopus embryo (the inset in A is a compilation of the entire series). The series is presented in a ventral to dorsal progression as depicted in schematic I. At stage 33, the Xenopus heart field is just beginning tube closure (D), yet elements of looping have begun. (1) There is a ventral bulging of the heart field issue (evident by ventral–dorsal series A–D or I). (2) The asymmetric nature of the heart becomes more evident as it bends to form a “c-shaped” structure, curved to the embryos right (B, C). (3) The forming tube also rotates, positioning the original left heart field (pseudo-colored blue in series A–D, J) more ventrally and the original right heart field (pseudo colored yellow) more dorsally. These changes are illustrated in panels A–D with A, the most ventral sections, showing mainly left field (blue) progressing to D, the most dorsal sections, showing tissue only from the right field (yellow). The undeveloped future sino-atrial region remains evenly distributed across the right–left axis of the embryo, however, the right side is becoming more anterior than the left. In all species, the looping heart forms a series of complementary curves. For example, the outer portion of the “c shaped” curvature on the right side of the heart can be matched to the inner part of the “c-shaped” curve on the left side of the heart. The complementary curves are highlighted green and red on many of the panels and shown separately in panels K and L. Panels E–H (schematic J) are a ventral–dorsal series of the heart region of a 1 μM 194301-treated embryo fixed when untreated littermates reached stage 39. Inset E shows the entire z-series through the heart field compiled. Confocal imaging the myocardium of these embryos revealed that many characteristics of looping are present. (1) The heart fields of the RA-antagonist-treated embryos did not remain flat sheets. Although the most posterior (the would-be sino-atrial region of an earlier normal embryo) remained evenly distributed among the ventral–dorsal, the right–left, and the anterior–posterior axes (E, F), (2) anterior to this region, the left field (blue) assumed a ventral position (note presence in E, lack thereof in H) while the right field (yellow) assumed a more dorsal prominence (present in H and not E). (3) Additionally, the left and right fields make the appropriate “c-shaped” bends with respect to the embryo's midline (G, H). The most prominent left side curvature of F and the most prominent right side curvature of H are copy/pasted into cartoon L. By digitally swinging the bottom (anatomical posterior end) of the curved lines toward the anatomical midline, the complementary pair resembles the pattern of a stage 33 heart (L and K).|