Dual roles of Sema6D in cardiac morphogenesis through region-specific association of its receptor, Plexin-A1, with off-track and vascular endothelial growth factor receptor type 2

Abstract

Semaphorins, originally identified as axon guidance facto s in the nervous system, play integral roles in organogenesis. Here, we demonstrate a critical involvement of Sema6D in cardiac morphogenesis. Ectopic expression of Sema6D o RNA interference against Sema6D induces expansion or narrowing of the ventricular chamber, respectively, during chick embryonic development. Sema6D also exerts region-specific activities on cardiac explants, a migration-promoting activity on outgrowing cells from the conotruncal segment, and a migration-inhibitory activity on those from the ventricle. Plexin-A1 mediates these activities as the major Sema6D-binding receptor. Plexin-A1 forms a receptor complex with vascular endothelial growth factor receptor type 2 in the conotruncal segment or with Off-track in the ventricle segment; these complexes are responsible for the effects of Sema6D on the respective regions. Thus, the differential association of Plexin-A1 with additional receptor components entitles Sema6D to exert distinct biological activities at adjacent regions. This is crucial for complex cardiac morphogenesis.

Figure 1.

Sema6D mRNA is expressed in the developing neural and cardiac tubes. (A) Lateral view of E10.5 mouse embryo. (B–E) Cross-sections through the E10.5 mouse embryos shown in A. Sema6D mRNA is expressed in the dorsal side of the neural tube (B–D) and the cardiac tube, including the CT segment (C, arrow), atrioventricular segment (D, arrow), and ventricle (C,D, arrowheads). The magnified view of the ventricle (E) demonstrates Sema6D mRNA expression in the myocardial cells in compact zone and trabeculae (arrowhead), more than in the endocardial cells (arrow). Bars: A–D, 100 μm; E, 20 μm. (F–H) Expression of cSema6D mRNA in HH stage 7 (F,G) and HH stage 12 (H) chicken embryos. Ventral view (F) and cross-sections (G,H) display cSema6D mRNA expression in the dorsal side of the neural fold (G, arrow) and the cardiac ventricular wall (H, arrow). The expression pattern of cSema6D mRNA in the chick embryo is similar to that seen in the mouse embryo. Bar, 100 μm.

Figure 2.

Ectopic expression of Sema6D and RNAi against cSema6D result in abnormal cardiac tube formation. (A–F) HH stage 12 chick embryos implanted with control cells (A–C) or Sema6D-expressing cells (D–F). HH stage 4 embryos were removed from eggs, placed in the culture dishes, and implanted with cells (10⁶ cells/aggregate) on the edge of the lateral mesoderm. Embryos were cultured until HH stage 12 and either immunostained with anti-vMHC antibody (A,C,D,F) or visualized with hemotoxylene and eosin staining (B,E). Cross-sections through each embryo are indicated in the respective ventral view. Ventral view (D) and cross-section (E) of embryos implanted with Sema6D-expressing cells display defects in the neural tube closure and expansion of the cardiac cavity in comparison with embryos implanted with control cells (A,B). Magnified view of ventricular wall of Sema6D-treated embryos (F) also exhibit thin myocardial layer. Bars: A,B,D,E, 100 μm; C,F, 20 μm. (G–I) HH stage 16 chick embryos implanted with control (G), Sema6D-expressing (H), or Sema3A-expressing cells (I). HH stage 4 embryos were treated as described above. Embryos were cultured until HH stage 16 and then immunostained with anti-vMHC antibody. Embryos implanted with Sema6D-expressing cells (H) display a phenotype similar to that of HH stage 12 (D), whereas embryos implanted with Sema3A-expressing cells (I) display no apparent changes in either neural or cardiac morphology, in comparison with control embryos (G). Bar, 100 μm. (J,K) HH stage 36 chick embryos treated with control (J) or Sema6D-expressing cells (K). Sema6D-expessing cells or control cells (10⁶ cells) were injected into the space between the vitelline membrane and chick embryo at HH stage 12. Embryos were allowed to develop in ovo up to HH stage 36. Cross-sections of embryos were visualized by hematoxylene and eosin staining. In Sema6D-treated embryos, cardiac ventricle exhibits expanded cavity and poor myocardial trabeculae (K, arrowhead) and the endocardial cushion of the CT segment is extended (K, asterisk), compared with the respective regions in control embryo (J). Bar, 100 μm. (L–R) HH stage 12 chick embryos treated with control siRNA (L,O,P) or siRNA specific for cSema6D in the absence (M,Q,R) or presence (N) of Sema6D-Fc. The cardiac tubes of HH stage 6 chick embryos were electroporated with either control or Sema6D-specific siRNA (40 pmole/μL). Embryos were cultured with or without Sema6D-Fc (100 ng/mL) until HH stage 12. Embryos were then immunostained with anti-vMHC antibody (L–N) or in situ hybridized with riboprobes for VMHC1 (O,Q) or AMHC1 (P,R). Ventral view of embryos treated with siRNA specific for cSema6D display a narrowing of the ventricle (M, arrowhead) in comparison with the right-side bended cardiac tube in embryos treated with a control siRNA (L). Note that siRNA specific for cSema6D distorted exclusively the ventricular portion (Q, arrowhead) but not the atrial portion (R), compared with control siRNA (O,P). In the presence of Sema6D-Fc, embryos treated with siRNA specific for cSema6D display nearly normal or expanded cardiac tube (N). Bar, 100 μm.

Figure 4.

Plexin-A1 binds to Sema6D and its mRNA is expressed in the developing neural and cardiac tubes and mediates the effect of Sema6D on chick embryo. (A) Micrographs of binding assays that test the binding of truncated Sema6D on HEK293 cells transfected with Plexin-A1, Plexin-A2, Plexin-A4, Plexin-B1, or NP1. An alkaline phosphatase fusion with the truncated Sema6D-Fc domain was incubated with the transfectants; the final detection of binding was performed using alkaline phosphatase substrates. Significant staining is observed in Plexin-A1-expressing cells and cells coexpressing Plexin-A1 and NP1. Bar, 50 μm. (B) Expression of Plexin-A1 mRNA in E10.5 mouse embryo. (Panel a) Lateral view of E10.5 mouse embryo. Bar, 100 μm. (Panels b,c) Cross-sections through the mouse embryo shown in panel a. Plexin-A1 mRNA is expressed in the lateral side of the neural tube (panel b, arrowhead), the atrioventricular cushion (panel c, arrow), and the myocardial layer of cardiac ventricular wall (panel c, arrowhead). Bar, 100 μm. (Panels d,e) The magnified views of the cardiac tube. Plexin-A1 mRNA is expressed in the endocardial cells of the cardiac cushion (panel d, arrow) and the ventricular wall (panel e, arrows). The intense expression of Plexin-A1 mRNA is observed in the myocardial cells on the epicardial side of the myocardium (panel e, arrowhead) at higher levels than in trabeculae. Bar, 20 μm. (Panels f–h) Expression of cPlexin-A1 mRNA in HH stage7 (panels f,g) and HH stage 12 (panel h) chick embryo. Ventral view (panel f) and cross-sections (panels g,h) display cPlexA1 mRNA expression in the lateral side of the neural fold (panel g, arrow) and the cardiac ventricular wall (panel h, arrow). The expression pattern of cPlexin-A1 mRNA in the chick embryo is similar to that observed in mouse embryos. Bar, 100 μm. (C, panels a–h) HH stage 12 chick embryos treated with in the combination of PlexA1-myc (panels a,c–f,h) and soluble Sema6D-Fc (panels b,c,g,h). HH stage 4 chick embryos were treated with various combinations of soluble Sema6D-Fc (100 ng/mL) and PlexA1-myc (10 μg/mL). Embryos were then immunostained with anti-vMHC antibody (panels a–c) or in situ hybridized with riboprobes for VMHC1 (panel d) or AMHC1 (panel e). The neural folds of cross-sectioned embryos (panels f–h) were visualized with hemotoxylene and eosin staining. PlexA1-myc alone induces a narrowing of the cardiac cavity instead of the normal ballooning (panel a, arrowhead). Note that cPlexA1-myc distorted exclusively the ventricular portion (panel d, arrowhead) but not the atrial portion (panel e). In the presence of PlexA1-myc, embryos treated with Sema6D-Fc display neural tube closure (panel h) and an almost normal cardiac cavity (panel c). Bar, 100 μm. (D, panels a–e) HH stage 12 chick embryos treated with control (panel a) or cPlexin-A1 specific siRNA (panels b–e) in the presence of soluble Sema6D-Fc (panel c). HH stage 6 chick embryos were electroporated with either control (panel a) or cPlexin-A1 siRNAs (40 pmole/μL; panels b–e), then incubated with soluble Sema6D-Fc (100 ng/mL; panel c). Embryos were cultured until HH stage 12. Embryos were then immunostained with anti-vMHC antibody (panels a–c) or in situ hybridized with riboprobes for VMHC1 (panel d) or AMHC1 (panel e). Embryos treated with siRNA specific for cPlexin-A1 display poorly looped cardiac tube (panel b, arrowhead). Note that siRNA specific for cPlexin-A1 distorted exclusively the ventricular portion (panel d, arrowhead) but not the atrial portion (panel e). This phenotype is similar to embryos treated with siRNA specific for cSema6D (Fig. 2M,Q,R). In the presence of Sema6D-Fc, nearly normal cardiac tube formation is restored in embryos treated with siRNA specific for cPlexin-A1 (panel c). Bar, 100 μm.

Figure 3.

Sema6D induces distinct effects on the outgrowth from explants of different regions of cardiac tube. (A, panels a–h) Collagen gel coculture of explants with control (panels b,f), Sema6D-expressing (panels c,g), or Sema3A-expressing cells (panels d,h). Tissues were removed from chick embryos, placed on collagen gels, and cultured with medium. Explants were positioned on the right, whereas the cell aggregates (transfect.) were positioned on the left. Cardiac explants of the CT segment (panels a–d) and ventricle (panels e–h) were visualized by phase contrast microscopy. CT segment explants (panel c) demonstrate increased outgrowth toward the Sema6D-expressing cells, whereas ventricular explants (panel g) exhibit decreased outgrowth. Bar, 100 μm. (Panels i,j) Quantification of the relative degree of outgrowth as a percentage of each control value (panel i) and total numbers of phospho-histone H3-positive nuclei within outgrowth (panel j) of each explant. Data are means ± S.E.M. (*) p < 0.05, vs. explants with control cells. (B, panels a–h) Immunohistochemistry of cardiac explant outgrowths. Cardiac explants were immunostained for factorVIII (panels a,c,e,g) or fibrillin-2 (panels b,d,f,h). CT segment explants exhibit increased numbers of factorVIII-positive cells (panel e), whereas ventricle explants demonstrate decreased numbers of factorVIII-positive cells in response to Sema6D (panel g). Bar, 50 μm.

Figure 5.

Truncated Plexin-A1 and RNAi against cPlexin-A1 block the effect of Sema6D on cardiac explants. (A, panels a–i) Collagen gel coculture of cardiac explants with control (panels a,e) or Sema6D-expressing cells (panels b–d,f–h). Cardiac explants were positioned on the right, whereas cell aggregates (transfect.) were positioned on the left. Cardiac explants were visualized by phase contrast microscopy. Some cultures (panels c,g) contained PlexA1-myc, whereas in other cultures (panels d,h), explants were infected with retroviruses encoding the PlexA1-DN and placed on collagen gels. Incubation with PlexA1-myc or infection with PlexA1-DN block Sema6D-induced changes in outgrowth from both of CT segments and ventricle explants. Bar, 100 μm. (Panel i) Quantification of the relative degree of outgrowth from cardiac explants as a percentage of control value. Data are means ± S.E.M. (*) p < 0.05, vs. explants with control cells without reagents. (B, panels a–e) Collagen gel cocultures of cardiac explants treated with control (panels a,c) or cPlexin-A1-specific (panels b,d) siRNA with Sema6D-expressing cells (panels a–d). Cardiac tissue explants were transfected with either control or cPlexin-A1 siRNA and then placed on the collagen gels. Explants were positioned on the right, whereas cell aggregates were placed on the left. Cardiac explants were visualized by phase contrast microscopy. (Panels b,d) RNAi against cPlexin-A1 blocks Sema6D-induced changes in outgrowth from both CT segments and ventricle explants. Bar, 100 μm. (Panel e) Quantification of the relative degree of outgrowth from cardiac explants as a percentage of control value. Data are means ± S.E.M. (*) p < 0.05, vs. explants with control cells without reagents.

Figure 6.

Plexin-A1 interacts with OTK and VEGFR2. (A) RT–PCR analysis of mRNA extracted from outgrowing cells from cardiac explants. Cardiac tissues were removed from chick embryos, placed on the collagen gel, and cultured with medium for 2 d. After removing the core region of the explant, outgrowing cells were subjected to mRNA extraction. The outgrowing cells from ventricle explants exhibit NP1 and OTK mRNA expression, whereas cells from CT segments exhibit VEGFR1 and 2 mRNA expression. (B) Association of endogenous Plexin-A1 and VEGFR2 in outgrowing cells from CT segments. Outgrowing cells were prepared as described in A. Lysates prepared from outgrowing cells were immunoprecipitated (IP) with anti-NP1 antibodies (NP), or anti-VEGFR2 antibodies (R2). Immunoprecipitates were blotted (Blot) with the indicated antibodies. In outgrowing cells from the CT segment, Plexin-A1 is detected in immunoprecipitates using anti-VEGFR2 antibodies, but not in immuoprecipitates using anti-NP1 antibodies in outgrowing cells from the ventricle. Molecular weight markers are indicated at left. (C) Association of Plexin-A1 with OTK, following expression in HEK293 cells. Lysates prepared from transfected cells were immunoprecipitated with anti-Flag antibodies (FLAG) or anti-V5 antibodies (V5). Immunoprecipitates were blotted with the indicated antibodies. OTK is detected in lysates from cells cotransfected with Plexin-A1. Molecular weight markers are indicated at left. (D) Association of Plexin-A1 with VEGFR2, following expression in HEK293 cells. Lysates prepared from transfected cells were immunoprecipitated with Flag or V5. Immunoprecipitates were blotted with the indicated antibodies. VEGFR2, but not VEGFR1, is detected in lysates from cells cotransfected with Plexin-A1. Molecular weight markers are indicated at left. (E) Tyrosine phosphorylation of VEGFR2 in outgrowing cells from CT segments. Outgrowing cells from CT segments were prepared as described in Figure 6A. Outgrowing cells were stimulated with VEGF (50 ng/mL) and/or Sema6D-Fc (100 ng/mL) for 30 min before harvest. Lysates of outgrowing cells were immunoprecipitated with anti-VEGFR2 antibody. Immunoprecipitates were blotted with anti-VEGFR2 antibodies and anti-phosphotyrosine antibody (p-Tyr).

Figure 7.

Differential association of Plexin-A1 with OTK and VEGFR2 conveys the opposite effects of Sema6D on ventricle and CT segment. (A) Quantification of the relative degree of outgrowth from cardiac explants as a percentage of control value. Collagen gel coculture was performed using cardiac explants treated with control or cOTK-specific siRNA in the presence of control or Sema6D-expressing cells. RNAi against cOTK blocks Sema6D-induced changes in outgrowth from ventricle explants but not from CT segments. Data are means ± S.E.M. (*) p < 0.05, vs. explants with control cells without reagents. (B) Quantification of the relative degree of outgrowth from cardiac explants as a percentage of control value. Collagen gel coculture was performed using cardiac explants with control or Sema6D-expressing cells. Some culture medium contained VEGF (50 ng/mL) or truncated VEGFR2-myc (10 μg/mL). In other cultures, explants were treated with siRNA specific for cVEGFR2. CT segments exhibit increased outgrowth by VEGF alone, but decrease Sema6D-induced outgrowth by incubation with truncated VEGFR2-myc or treatment with siRNA specific for VEGFR2. Sema6D-induced outgrowth is augmented by VEGF. Ventricle explants do not demonstrate significant changes in Sema6D-induced outgrowth as well as control outgrowth by the treatment of VEGF and VEGFR2 mutants. Data are means ± S.E.M. (*) p < 0.05, vs. explants with control cells without reagents. (C) HH stage 12 chick embryos treated with control (panel a), cOTK (panels b,d,e), or cVEGFR2 specific siRNA (panel c). HH stage 6 chick embryos were electroporated with either control (panel a), OTK siRNAs (40 pmole/μL; panels b,d,e), or cVEGFR2 siRNAs (40 pmole/μL; panel c). Embryos were cultured until HH stage 12. Embryos were then immunostained with anti-vMHC antibody (panels a–c) or in situ hybridized with riboprobes for VMHC1 (panel d) or AMHC1 (panel e). Embryos treated with siRNA specific for OTK display poorly looped cardiac tube (panel b, arrowhead). Note that siRNA specific for cOTK distorted exclusively the ventricular portion (panel d, arrowhead) but not the atrial portion (panel e). (Panel c) In contrast, embryos treated with cVEGFR2 specific siRNA display nearly normal cardiac tube. Bar, 100 μm.