Calcineurin signaling in avian cardiovascular development

Authors

  • Christine M. Liberatore,

    1. Division of Molecular Cardiovascular Biology, Cincinnati Children's Medical Center, ML7020, Cincinnati, Ohio
    Search for more papers by this author
  • Katherine E. Yutzey

    Corresponding author
    1. Division of Molecular Cardiovascular Biology, Cincinnati Children's Medical Center, ML7020, Cincinnati, Ohio
    • Division of Molecular Cardiovascular Biology, Cincinnati Children's Medical Center, ML7020, 3333 Burnet Avenue, Cincinnati, OH 45229
    Search for more papers by this author

Abstract

Experiments were initiated in avian embryos to determine the embryonic expression of calcineurin protein phosphatase isoforms as well as to identify developmental processes affected by inhibition of calcineurin signal transduction. Chicken calcineurin A alpha (CnAα) and calcineurin A beta (CnAβ) are differentially expressed in the developing cardiovascular system, including primitive heart tube and valve primordia. Inhibition of calcineurin signaling by cyclosporin A (CsA) treatment in ovo resulted in distinct cardiovascular malformations, depending on the timing and localization of treatment. Initial formation of the heart tube was apparently normal in embryos treated with CsA from embryonic day (E)1 to E2, but hallmarks of heart failure were apparent with treatment from E2 to E3. Vascular defects were apparent in whole embryos treated on either day, but local administration of CsA directly to the forming vessels on E2 did not inhibit blood vessel formation. This observation supports an indirect effect of calcineurin inhibition on angiogenic remodeling as a result of compromised heart development. Together these studies are consistent with multiple roles for calcineurin signaling in the developing cardiovascular system. Developmental Dynamics 229:300–311, 2004.© 2003 Wiley-Liss, Inc.

INTRODUCTION

The calcineurin signal transduction pathway has been implicated in the embryonic development and adult pathology of several organ systems and tissue types, including cardiac and skeletal muscle, cartilage, vasculature, nervous system, and skin (Graef et al., 2001b; Crabtree and Olson, 2002). The calcineurin phosphatase is activated by sustained elevation of intracellular calcium levels in response to a variety of extracellular stimuli (Rusnak and Mertz, 2000). The calcineurin holoenzyme is associated with calmodulin and consists of highly conserved calcineurin A (CnA) catalytic and calcineurin B (CnB) regulatory subunits. Among the targets of the calcineurin phosphatase are members of the nuclear factor of activated T cells (NFAT) family of transcription factors that are translocated to the nucleus in response to dephosphorylation by calcineurin. The five vertebrate NFATs are differentially expressed during development and have diverse functions in a variety of tissues during embryogenesis and in adults (Rao et al., 1997; Graef et al., 2001b). NFATs have been described as integrators of multiple signaling pathways, because they act in conjunction with other transcription factors such as AP-1 and GATA4 (Crabtree, 1999). Until recently, the importance of calcineurin and NFATs during development has been underappreciated, but it is likely that additional functions for these proteins will be revealed as more is known about the signaling pathways that control organogenesis.

Gene targeting of calcineurin and NFAT family members has established the importance of this signaling pathway in the development of the cardiovascular system in mice (Graef et al., 2001b). Targeted mutagenesis of the single embryonically expressed calcineurin B1 (CnB) regulatory subunit gene results in embryonic lethal cardiovascular defects (Graef et al., 2001a). Embryonic lethality was not observed with loss of either CnAα or CnAβ isoforms, but mice lacking both are not viable (Zhang et al., 1996; Bueno et al., 2002a; Parsons, et al., 2003). Targeted mutagenesis of specific NFAT family members revealed additional requirements for calcineurin/NFAT signaling in the developing heart. Loss of NFATc1 results in embryonic lethality due to defects in cardiac valvuloseptal development (de la Pompa et al., 1998; Ranger et al., 1998). Embryos lacking both NFATc3 and NFATc4 have embryonic lethal cardiac metabolic defects and exhibit abnormalities in vascular maturation (Graef et al., 2001a; Bushdid et al., 2003). Together these studies demonstrate that calcineurin signaling and NFAT activation are required for cardiovascular development and embryonic viability. However, the upstream effectors, downstream targets, and temporal requirements of this signaling pathway in the developing heart and blood vessels are not completely characterized.

Developmental functions for calcineurin signaling also have been examined pharmacologically using cyclosporin A (CsA), a specific inhibitor of calcineurin phosphatase activity (Liu et al., 1991). In mice, maternal administration of CsA during E7.5–E8.5 recapitulated the defective vascular development and embryonic lethality observed with gene targeting of CnB (Graef et al., 2001a). Of interest, maternal administration of CsA after E8.5 did not affect vascular development, although embryonic lethality occurs at these stages with genetic manipulation of calcineurin and NFAT expression. In addition, cultured mouse embryos treated with CsA beginning at E8.5 exhibited a variety of malformations and embryonic growth retardation, supporting the requirement for calcineurin signaling at these later stages (Uhing et al., 1993). In cell culture experiments, CsA treatment inhibits the differentiation of skeletal and vascular smooth muscle cells and also inhibits contractile activity and calcium handling in cardiomyocytes (Abbott et al., 1998; Janssen et al., 2000; Robida et al., 2000; Ohkawa et al., 2003). Further studies are necessary to identify specific populations of cells in the embryo affected by calcineurin inhibition in vivo and to determine the timing of sensitivity of different cell lineages during development.

Relatively little is known about calcineurin signaling and NFAT activation in the development of vertebrate organisms other than mice. While important roles for calcineurin and NFATs have been established through murine genetic studies, the precise temporal and spatial requirements of this signaling pathway are often difficult to ascertain in embryos with genomic alterations. More targeted approaches are possible in the chicken system where delivery of signaling agents can be controlled in time and space. Here, we report the isolation and developmentally regulated expression of CnA isoforms in the chick embryo. Temporally and spatially controlled administration of cyclosporin was used to examine the effects of calcineurin inhibition on specific events in cardiovascular development, including cardiomyocyte differentiation and diversification, blood vessel formation, and myocardial maturation. Through these analyses, specific temporal and spatial requirements for calcineurin signaling were identified in the developing avian cardiovascular system.

RESULTS

Isolation of Chicken Calcineurin A α and β Isoforms

Degenerate oligonucleotides directed against highly conserved regions of calcineurin sequence were used to amplify CnA isoform sequences from embryonic day 2 (E2) avian embryos. Oligonucleotide sequences were directed against amino acid coding sequence corresponding to CnA phosphatase and CnB binding domains, which are completely conserved between CnA α, β, and γ isoforms of rat, human, frog, and cow. Amplified DNA of appropriate size was subcloned and the sequences of multiple CnA fragments were determined and classified based on BLAST searches of the GenBank database. The isolated chicken CnA sequences correspond to either α or β but not γ isoforms (Fig. 1), and sequences encoding other CnA isoforms were not identified in the degenerate polymerase chain reaction (PCR) amplification screen. The derived amino acid sequence for chicken CnAα is 100% identical to the corresponding region of human CnAα, and the CnAβ sequence is 97% identical to human CnAβ sequence (Kincaid et al., 1990). The chicken CnAβ sequence has less identity with human CnAα (94%), and amino acids that are diagnostic for CnAβ rather than α in other vertebrate species are conserved. A more complete cDNA sequence for CnAα was obtained after screening a stage 8–13 chicken heart cDNA library. The derived amino acid sequence of the CnAα isoform isolated is 98% identical to full-length human CnAα and corresponds to the human CnAα2 splice variant, which is missing 10 amino acids that are present in the CnAα1 C-terminal domain (Kincaid et al., 1990; McPartlin et al., 1991). PCR primers were designed that are specific for either the CnAα or CnAβ isoform. RT-PCR analysis of RNA isolated from E1 (Hamburger and Hamilton, 1951; stages 5–8) whole embryos and E2 (Stages 11–14) head, heart, and tail demonstrated that both CnAα and CnAβ isoforms are expressed in avian embryos at these stages (Fig. 2).

Figure 1.

Sequence alignment of chicken calcineurin A (CnA) α and CnAβ. Chicken CnA sequences were isolated with degenerate oligonucleotides corresponding to conserved amino acids in the calcineurin A phosphatase domain (single underline) and in the calcineurin B binding region (double underline) followed by cDNA library screening for CnAα. The CnA autoinhibitory domain also is indicated (broken underline). CnA-derived amino acid sequences were aligned by using BLASTp and the GenBank database. CnA isoforms were identified based on amino acid identity with human CnAα (CnAα2 splice form, accession no. BC025714) and CnAβ (accession no. M29551). Sequence identities for human and chicken CnA isoforms are indicated by dashed lines and boldface amino acids in the CnAβ sequence are characteristic of β rather than α isoforms.

Figure 2.

Rerverse transcriptase-polymerase chain reaction (RT-PCR) analysis of calcineurin A alpha (CnAα) isoform expression during early avian development. CnAα and CnAβ sequences were amplified using isoform-specific oligonucleotides from reverse transcribed RNA isolated from embryonic day (E) 1 chicken embryos (stages 5–8) or dissected heads (Hd), hearts (Ht), and tails (T) from E2 embryos (stages 11–14). Amplification of GAPDH was included as a loading control, and reactions in which reverse transcriptase was omitted (noRT) served as negative controls for nonspecific PCR amplification.

Expression of Calcineurin A Isoforms During Avian Cardiovascular Development

The tissue distribution of CnAα and CnAβ during early embryogenesis and heart formation was examined in whole-mount chicken embryos. At stage 10, CnAα expression is apparent in the primitive heart tube and more posteriorly in the lateral plate (Fig. 3A). Both CnAα and CnAβ are expressed in the developing neural tube at this stage (Fig. 3A,E). Differential distribution of CnAα and CnAβ was observed during heart looping and chamber formation. At stage 15, CnAα expression is restricted to the atrioventricular (AV) canal, but CnAβ is more widely expressed in throughout the looped heart (Fig. 3B,F). Both CnAα and CnAβ are expressed in the somites and CnAβ expression is apparent in the limb buds (Fig. 3C,F,G). Gene expression of CnA isoforms was examined in bisected hearts during valve formation (Fig. 3D,H). Strong expression of CnAα was observed in AV and semilunar valve primordia at E7 and expression of CnAβ also was detected in the heart valves at E10. Neither CnAα nor CnAβ are expressed strongly in the myocardium after heart chamber formation is complete. Together these expression analyses show that CnAα and CnAβ are temporally and spatially regulated in several developing embryonic structures, including the heart.

Figure 3.

Calcineurin A alpha (CnAα) and CnAβ are expressed in the developing cardiovascular system of the chick. A–H: Whole-mount chicken embryos or bisected hearts were hybridized with CnAα (A–D) or CnAβ (E–H) digoxigenin-labeled antisense RNA probes. A,E: At stage 10, expression of both CnA isoforms is detected in the neural tube (asterisks) and CnAα is present in the caudal regions of the heart primordia (arrow) and posterior lateral plate (arrowheads). B,F,G: CnAα is predominant in the atrioventricular (AV) canal (arrowhead) of a stage 15 heart, whereas CnAβ is more widely expressed throughout the heart (F,G, arrowheads). C: Expression of both CnA isoforms also was observed in the somites(s) with CnAβ expression evident in the forelimb (fl) and hindlimb (hl). The posterior intestinal portal (pip) also is indicated. D: CnAα is expressed in the semilunar (arrowhead) and AV valve primordia (arrows) at embryonic day 7. E: CnAβ is present in the aortic semilunar valve (arrowhead) and mitral valve (arrow) at embryonic day 10.

Effects of Calcineurin Inhibition on Cardiac and Skeletal Muscle Differentiation and Fiber Type Diversification

In skeletal muscle cells, calcineurin inhibition prevents myogenic differentiation and promotes fast vs. slow skeletal muscle fiber types (Abbott et al., 1998; Delling et al., 2000; Friday et al., 2000). The possibility that calcineurin has a similar function in the developing heart was investigated by treatment of avian embryos in ovo with cyclosporin A (CsA). A teratologic dose of CsA for chicken eggs was determined empirically to be 100 μg CsA in 100 μl (1.6 mg/kg) per egg, which is less than the teratologic dose reported for mice (12.5–25 mg/kg). Higher doses of CsA (500 μg/egg) are cytotoxic and caused arrested development and death of treated embryos. Chicken embryos were treated with a single teratologic dose of CsA at stages 5–8 (E1), and embryos were allowed to develop until stages 11–12 (E2). These stages correspond to the initial formation and differentiation of the primitive heart tube as well as the establishment of diversified myogenic lineages evident in positionally restricted myosin heavy chain isoform expression (Yutzey et al., 1994). The early formation and looping of the primitive heart tube were grossly normal in CsA-treated embryos, and heart beating was observed in both treated and control embryos (Fig. 4). Cardiomyogenic differentiation occurred in CsA-treated embryos as demonstrated by myosin heavy chain expression in the primitive hearts (Fig. 4A).

Figure 4.

Cardiac and skeletal myogenesis in chicken embryos treated with cyclosporin A (CsA) in ovo. A–F: Fertilized chicken eggs were injected with 100 μg of CsA (A–C) or phosphate buffered saline (D,E) on embryonic day (E) 1/stages 5–8 (A,B,D,E) or E2/stages 12–13 (C,F). Injected eggs were incubated an additional day, and embryos were hybridized with VMHC1 (A,C,D,F) or AMHC1 (B,E) antisense RNA probes. Embryos treated from E1 to E2 exhibited normal expression of VMHC1 throughout the primitive heart tube (A,D) and posterior restriction (asterisks in B,E) of AMHC1 (B,E) at stages 12–13. C,F: Embryos treated from E2 to E3 expressed VMHC1 in the ventricles (V) and not atria (A), although heart morphogenesis was abnormal in the presence of CsA. In the somites at stage 18, VMHC1 is predominant in the somitic core in control embryos (F, large arrow) but is redistributed to the rostral and caudal edges of the somites in CsA-treated embryos (C, small arrows).

Diversification of anterior and posterior cardiomyogenic lineages also was examined in CsA-treated embryos. At stages 11–12, the myosin heavy chain isoform VMHC1 is expressed throughout the primitive heart tube, but AMHC1 expression is restricted to the posterior atriogenic region of the heart (Bisaha and Bader, 1991; Yutzey et al., 1994). These myosin isoforms were used as indicators of cardiomyogenic differentiation and fiber type diversification in CsA-treated and control embryos. VMHC1 expression, indicative of cardiomyogenic differentiation, was apparent throughout the primitive heart tube of both experimental groups (Fig. 4A,D). Restricted expression of AMHC1 in the posterior heart tube also occurred in both treated and control embryos, indicative of normal fiber type specification in the presence of CsA (Fig. 4B,E). In older embryos treated from E2 to E3 with CsA, ventricular restriction of VMHC1 expression in the heart occurred normally, although abnormal heart morphogenesis was noted (Fig. 4C,F; Bisaha and Bader, 1991; Yutzey et al., 1994). Together these observations demonstrate that cardiomyogenic differentiation and diversified lineage specification occur when calcineurin signaling is inhibited.

The normal differentiation and diversification of cardiac myocytes was confirmed in heart forming region explants placed in culture and treated with CsA in minimal culture medium (Fig. 5). This assay provides a direct assessment of the effects of CsA on cardiomyogenic differentiation and diversification in a controlled culture environment. Anterior or posterior stage 5/6 heart forming region explants were placed directly in culture medium containing 0.5 μM CsA or control medium and cultured for 2 days, during which they differentiate and begin to beat. CsA-treated cardiomyocytes appeared to differentiate normally as indicated by strong VMHC1 expression and beating in both anterior and posterior explants (Fig. 5A). Cardiac myocyte diversification was assessed by posterior localization of AMHC1 expression in cultured explants (Yutzey et al., 1995). CsA treatment did not appear to affect this process, because the normal pattern of AMHC1 expression in posterior but not anterior explants was apparent in both treated and control groups (Fig. 5B,D). Further evidence for the inability of CsA to affect cardiomyocyte lineage diversification is that CsA treatment did not prevent the retinoic acid-induced anterior activation of AMHC1 expression (Yutzey et al., 1995) in explants treated with both drugs (data not shown). These explant culture experiments support the in vivo observations that early differentiation and diversification of cardiomyogenic lineages are unaffected by calcineurin inhibition.

Figure 5.

Cardiac differentiation and diversification in cultured heart forming region explants occur normally in the presence of cyclosporin A (CsA). A–D: Anterior and posterior heart forming regions were explanted from stage 5/6 embryos and placed in culture in the presence (A,B) or absence (C,D) of 0.5 μM CsA. Explants were cultured for 2 days and then hybridized with VMHC1 (A,C) or AMHC1 (B,D) antisense RNA probes. Myogenic differentiation was apparent in treated anterior and posterior explants (n = 5 of 5) and in control cultures (n = 6 of 6). Restriction of AMHC1 to the posterior heart forming region in treated explants also was apparently normal (treated n = 13 of 17; untreated n = 9 of 11). Note anterior explants in B and D are present but do not express AMHC1.

The efficacy of CsA treatments was confirmed and effects on skeletal myogenesis were investigated in embryos treated in ovo from E2 to E3. For these experiments, eggs were injected with CsA at stages 11–12 and embryos were isolated at stage 18. During this time period, somitic muscle is in the process of differentiating and diversified slow and fast myogenic lineages are established (Buffinger and Stockdale, 1994; Stern and Hauschka, 1995). Myogenic differentiation of somitic muscle evident in VMHC1 gene expression was observed in both treated and control embryos (Fig. 4C,F). Explanted somitic muscle precursors treated in culture also differentiated in the presence of CsA (data not shown). However, differential localization of myosin heavy chain isoform gene expression within the somites was affected with CsA treatment. In untreated embryos, VHMC1 expression is predominant in the somitic core region enriched for slow muscle fibers (Sacks et al., 2003) and lower expression is evident at the rostral and caudal edges of the somites (Fig. 4F). In CsA-treated somites VMHC1 expression is predominant at the rostral and caudal somitic borders but was notably weaker in the somite core (Fig. 4C). The differential localization of VHMC1 expression in the somites at this stage is consistent with diversified muscle fiber types in the newly differentiated somitic muscle (Sacks et al., 2003). Therefore, the shift in VMHC1 localization from the somite core, where slow fibers predominate, to the rostral–caudal somite edges in CsA-treated embryos likely reflects altered skeletal muscle fiber-type distribution as a result of calcineurin inhibition in vivo.

CsA Treatment and the Formation of a Functional Cardiovascular System In Ovo

Studies in mouse embryos and cultured vascular endothelial cells have implicated calcineurin signaling and NFAT activation in angiogenic remodeling events (Graef et al., 2001a). The effect of calcineurin inhibition on development of the entire cardiovascular system was further examined in CsA-treated avian embryos, where the timing and dosage of drug delivery can be controlled. Embryos were treated with CsA on E1 (stages 5–8) or E2 (stages 11–12) and assessed on E3 (stages 15–18) for heart morphogenesis and blood vessel formation. Obvious abnormalities in heart formation and vascular development were observed in embryos treated beginning on E1 or E2 (Table 1). The extent of developmental abnormalities was dependent on the dose of CsA administered, with more severe defects and increased mortality in embryos treated with 500 μg CsA and less severe, but significant, cardiovascular defects apparent with 50 μg treatments (data not shown). Similar defects in heart formation and vascular development were observed after treatment with 50, 10, or 1 μg of FK506, an alternative calcineurin inhibitor (data not shown; Liu et al., 1991). Abnormal heart and vessel development was observed in approximately 80% of embryos treated with CsA on E1, and defects in both heart and vasculature also were observed in the majority of embryos treated on E2.

Table 1. Cardiovascular Defects in Cyclosporin A–treated Embryos in Ovoa
 E1 to E3E2 to E3
+CsAControl+CsAControl
  • a

    Embryos were injected with 100 μg cyclosporin A (CsA) on embryonic day 1 (E1; stages 5–8) or E2 (stages 11–12), and all were assessed on E3 (stages 15–18). Myocardial thinning was determined by translucence or distension of the primitive heart tube. Decreased vascular remodeling was defined as persistence of vascular plexus or reduced formation of major blood vessels.

Myocardial thinning80%6%63%4%
 28/352/3515/241/24
Decreased vascular remodeling82%0%55%4%
 28/340/3212/221/24

The major developmental defects observed with calcineurin inhibition in ovo were in the cardiovascular system. Blood pooling and lack of circulation, indicative of cardiovascular failure, were evident in the majority of CsA-treated embryos evaluated on E3. However, heartbeats observed in CsA-treated embryos were indicative of embryonic viability and the initiation of cardiac function (Fig. 6A,B). Additional evidence for embryonic viability was the apparently normal development of eyes, pharyngeal arches, and otic vesicles in treated embryos. The most common cardiac anomalies observed in CsA-treated embryos were obvious thinning and distension of the ventricular myocardium relative to control embryos (Fig. 6C,G). In addition, heart looping was abnormal with atria positioned caudally relative to atria in control embryos. In histologic sections, formation of ventricular trabeculae and the initiation of endocardial cushion formation in the AV canal also were compromised in CsA-treated embryos (Fig. 6D,H). Impaired vascular development was apparent in the reduction of major vessel formation and persistent vascular plexuses in the vascular beds of the chorioallantoic membrane (Fig. 6B,F). In many treated embryos, complete lack of venous return to the embryo was evident in the absence of blood flow in anterior and posterior vitelline veins (Fig. 6A). Together these embryonic malformations are indicative of cardiovascular failure; however, several embryonic structures including the head, eye, pharyngeal arches, and somites continued to advance developmentally in the CsA-treated embryos (Fig. 6C,G). By E4, the majority of embryos treated with CsA on E1 or E2 were obviously necrotic. The timing of death at E4, but not earlier, in the absence of blood flow supports previous observations that a fully functional cardiovascular system is not necessary for avian development before that stage (Burggren et al., 2000).

Figure 6.

Development of a functional cardiovascular system is inhibited by cyclosporin A (CsA) treatment in ovo. Fertilized chicken eggs were injected with 100 μg of CsA (A–D) or phosphate buffered saline (E–H) on embryonic day (E) 2/stages 11–12 and incubated an additional day to E3/stage 15. Embryos were photographed in ovo to visualize blood circulation in the forming cardiovascular system (A,B,E,F). Regions within the rectangles (A,E) were enlarged (B,F) to show inhibition of angiogenic remodeling with CsA treatment. Arrows in A and E indicate blood in the anterior vitelline vein that is missing in the CsA-treated embryo. The posterior vitelline vein is indicated by an arrowhead in F. The embryos in A and E were removed from the egg and photographed in whole-mount (C,G). D,H: In histologic sections, initial formation of ventricular trabeculae (arrowheads) and atrioventricular endocardial cushions (asterisks) were disrupted in the CsA-treated embryo (D) relative to controls (H). A, atrium; V, ventricle; AVC, atrioventricular canal are indicated.

The administration of CsA by in ovo injections exposes the entire embryo to the drug, making it difficult to determine whether cardiovascular anomalies are the result of primary defects in vessel formation or are secondary to compromised cardiac function. Local administration of CsA to the developing extraembryonic vasculature was used to analyze the primary effects of the drug on angiogenic remodeling of vascular precursors beginning at E1 or E2 of development (Fig. 7). CsA was administered by agar squares saturated with the drug and placed directly on the developing embryo on E1 (stages 7–8) or E2 (stages 11–12) and embryos were allowed to develop until E3 (stages 15–17). Vascular development in close proximity to the CsA agar was evaluated visually and documented photographically.

Figure 7.

Local administration of cyclosporin in agar inhibits angiogenic remodeling when initiated on embryonic day (E) 1 but not on E2. Agar squares infiltrated with cyclosporin A (CsA) (A,B) or phosphate buffered saline (C,D) were placed directly on extraembryonic vascular precursors at E1/stage 5 (A,C) or E2/stage 11 (B,D). Windowed eggs were incubated until E3/stages 15–18, and agar squares (boxes) were photographed in ovo. A: Remodeling of the vascular plexus (asterisk) into major vessels was inhibited near the agar square and in embryos treated on E1 (n = 9 of 10). B–D: Major vessel formation in close proximity to the agar (indicated by arrowheads) was observed in E2 CsA-treated embryos (n = 9 of 11) and in control embryos treated on E1 or E2.

Severe defects in vascular patterning and angiogenic remodeling were observed in direct contact and in proximity to CsA agar treatments initiated on E1 (Fig. 7A; n = 9 of 10). However, teratologic effects also were observed in other embryonic structures, including the heart, with this drug treatment regimen. In contrast, large remodeled vessels were observed in direct contact with CsA agar when placed on the developing extraembryonic vasculature on E2 (Fig. 7B; n = 9 of 11). In parallel experiments, CsA agar placed near the heart at stages 11–12 resulted in cardiac anomalies, including distended ventricles (data not shown). Vascular defects and blood pooling also were observed in these embryos with cardiac anomalies. The observation that angiogenic remodeling is relatively unaffected by direct administration of CsA to the forming blood vessels (stages 11–12) supports the hypothesis that the vascular anomalies observed with whole embryo treatments are probably secondary to compromised heart function.

DISCUSSION

In the avian embryo, CnAα and CnAβ isoforms are expressed in embryonic structures affected with CsA treatment and calcineurin inhibition. In the cardiovascular system, CnAα and CnAβ are differentially expressed in the myocardium and developing valves of the heart as well as posterior lateral plate, which contains vascular precursors. CnAα and CnAβ also are expressed in the somites, limb buds, and neural tube. Calcineurin inhibition with CsA treatment led to developmental defects in structures that express CnAα and CnAβ, including the heart and somites. The initial differentiation and diversification of cardiac myocytes and formation of the primitive heart tube occurred in the presence of CsA in embryos treated from E1 to E2. However, heart morphologic abnormalities and vascular defects were observed with CsA treatment from E2 to E3. Cardiac defects observed in these embryos included abnormal looping, myocardial thinning, and disruption of endocardial cushion formation. Cardiovascular function also was impaired evident in pericardial effusion and reduced blood flow. Targeted delivery of CsA directly to forming vessels from E2 to E3 demonstrated that angiogenic remodeling occurs when calcineurin signaling is inhibited locally. Therefore, the defects in heart morphogenesis observed with treatments from E2 to E3 likely contribute to defects in blood vessel formation in CsA-treated embryos. Skeletal muscle fiber-type distribution in the somites also was affected with the E2 to E3 CsA treatments. These analyses of calcineurin expression and sensitivity periods for calcineurin inhibition in avian embryos further define the temporal and spatial requirements for this signaling pathway in cardiovascular and skeletal muscle development.

Temporally and spatially distinct patterns of expression of CnAα and CnAβ were observed during cardiovascular development in avian embryos. Differential expression of these isoforms also has been reported in other cell types where calcineurin signal transduction is important, including lymphocytes, cardiomyocytes, neurons, and skeletal muscle (Kuno et al., 1992; Jiang et al., 1997; Taigen et al., 2000; Parsons et al., 2003). The CnAα sequence expressed early in avian development corresponds to an alternatively spliced form lacking a specific 10 amino acids first identified in murine and human brain, but the functional or regulatory implications of this alternative form of CnAα have not been determined (Kincaid et al., 1990). The increased expression of calcineurin isoforms in embryonic structures where the pathway is activated is suggestive of a feedforward mechanism for calcineurin signaling whereby activation of the pathway leads to increased calcineurin transcription. In support of such a mechanism, CnAβ mRNA expression is elevated in response to agonist-stimulated hypertrophy of cardiomyoctes (Taigen et al., 2000). Specific signaling transduction activities for CnAα and CnAβ have not been demonstrated, but gene targeting in mice reveals different phenotypes with loss of either CnAα or CnAβ. Postnatal defects were detected in the heart for CnAβ nulls and in the nervous system for CnAα nulls, while immune system and skeletal muscle abnormalities were observed with loss of either CnA isoform (Zhang et al., 1996; Zhuo et al., 1999; Bueno et al., 2002a, c; Parsons et al., 2003). Severe embryonic malformations or lethality were not reported for either CnAα or CnAβ nulls, but embryos lacking both CnAα and CnAβ are not viable (Parsons et al., 2003). Because the CnB1 null embryos do not survive beyond E10.0, there may be redundancy or compensatory mechanisms for the loss of specific CnA isoforms during early development. However, multiple organ systems are affected later in life by the reduced capacity for calcineurin signal transduction that results from targeted mutagenesis of individual CnA genes.

The initial specification, differentiation, and diversification of cardiomyocyte cell lineages are apparently unaffected by inhibition or loss of calcineurin signal transduction. In CsA-treated avian embryos, cardiac-specific gene expression is activated and the heart primordia fuse to form the primitive heart tube. The initial diversification of anterior/ventriculogenic and posterior/atriogenic cardiomyogenic lineages in ovo or in cultured cardiac explants also appears normal in the presence of CsA. A similar result was obtained in mice, where the initial formation and differentiation of a functional heart tube was observed with targeted mutagenesis of CnB1 (Graef et al., 2001a). Together, these studies demonstrate that the early differentiation and diversification of cardiac myocytes is not dependent on calcineurin signal transduction. This finding is in contrast to the effects of CsA treatment or loss of CnA isoforms on the differentiation of cultured skeletal muscle myocytes and on the diversification of slow vs. fast skeletal muscle lineages (Abbott et al., 1998; Chin et al., 1998; Delling et al., 2000; Parsons et al., 2003). In general, calcineurin inhibition leads to loss of slow muscle fibers in cultured skeletal myocytes or genetically altered mice. Therefore, the observed reduction of VMHC1 expression in the somite core of CsA-treated avian embryos is consistent with a loss of slow muscle fibers during somitogenesis.

There is increasing evidence that cardiomyocyte maturation and heart function require calcineurin signaling and NFAT activation during development. Cyclosporin treatment of avian embryos during heart looping results in thinning of the ventricular myocardium and heart failure evident in blood pooling and pericardial effusion. A similar phenotype was observed in mouse embryos lacking both NFATc3 and NFATc4 that die at midgestation as a result of deficiencies in cardiomyocyte proliferation and metabolism (Bushdid et al., 2003). Together, these studies support a critical role for calcineurin signaling and NFAT activation in the maturation of cardiomyoctes at E3–E4 in avian embryos and E10.5–E11.5 in mice. These are comparable stages in terms of development of the heart and have been identified as critical time points for the initiation of effective blood circulation in both species (Burggren et al., 2000; Conway et al., 2003).

Calcineurin signaling and NFAT activation also have been implicated in vascular development. Avian embryos treated with CsA from stage 5 to stage 12 (E1 to E2) did not exhibit defects in endothelial specification or differentiation (data not shown), but vascular malformations and angiogenic remodeling defects were observed at stage 14–15 (E3) when treatment was initiated at stage 5 (E1) or at stage 12 (E2). The early sensitivity period for vascular development before any blood circulation also was observed in mice with maternal CsA administration at E7.5 to 8.5 or loss of CnB1 (Graef et al., 2001a). Although these studies are suggestive of a primary role for calcineurin signaling early in angiogenesis, it is also possible that the observed vascular remodeling defects are secondary to compromised heart development within the context of the whole embryo. A potential function for calcineurin in vascular development is as a mediator of vascular endothelial growth factor (VEGF) signal transduction. In cultured endothelial cells, VEGF induces calcineurin signaling and NFAT activation and VEGF-induced angiogenesis can be inhibited with CsA treatment (Armesilla et al., 1999; Hernandez et al., 2001). During early embryogenesis, VEGF has a critical role in the remodeling of endothelial cells to form the first blood vessels (Drake and Little, 1999; Argraves et al., 2002). Therefore, the early sensitivity period for calcineurin inhibition in vascular development could be related to VEGF control of early vasculogenesis and angiogenic remodeling.

Inhibition of angiogenic remodeling of the peripheral vascular plexus was observed in whole avian embryos treated with CsA from E2 to E3; however, local delivery of the drug directly to the remodeling vascular precursors did not affect major vessel formation. Thus, the inhibitory effect of CsA at this stage does not appear to be a primary defect in blood vessel development but rather is a likely secondary effect of compromised cardiac function. Avian embryos treated with CsA from E2 to E3 exhibit thinning of the myocardium and reduced trabeculae similar to mouse embryos lacking NFATc3 and NFATc4 (Bushdid et al., 2003). These mice were originally described as having embryonic lethal vascular patterning defects, but a subsequent study demonstrated prolonged embryonic viability with cardiomyocyte-specific restoration of NFAT function (Graef et al., 2001a; Bushdid et al., 2003). The dependence of vascular development on cardiac function also has been demonstrated in mice deficient in N-cadherin or the sodium/calcium exchanger Ncx1 (Wakimoto et al., 2000; Luo et al., 2001). The requirement for cardiac function for vascular plexus remodeling was observed as early as 1918, when W.B. Chapman reported that chick yolk sac capillary beds failed to develop into major vessels after extirpation of the heart at the 12-somite stage (Hamburger and Hamilton stage 11 or E2) (Chapman, 1918). Further studies are necessary to determine whether hemodynamic forces or loss of circulating factors are the basis for the dependence of the developing vasculature on cardiac function.

The calcineurin signal transduction pathway is active during critical stages of development in several organ systems. In the developing cardiovascular system, calcineurin signaling is required for cardiomyocyte maturation and function as well as early vascular development. During heart chamber formation, calcineurin signaling and NFATc1 function are required for valvuloseptal development, and calcineurin signaling also has been implicated in compaction of the ventricular myocardium (de la Pompa et al., 1998; Ranger et al., 1998; Shou et al., 1998; Guo et al., 2002). In the adult heart, calcineurin signaling has a central role in cardiac hypertrophy (Bueno et al., 2002b). In addition to the heart, calcineurin signal transduction is active during development and adult adaptive responses of skeletal muscle, neurons, and the immune system, and there is emerging evidence for the importance of this signaling pathway in cartilage, skin, and smooth muscle (Baksh and Burakoff, 2000; Olson and Williams, 2000; Ranger et al., 2000; Graef et al., 2001b, 2003; Crabtree and Olson, 2002). These initial studies have all been performed using mammalian systems and tissue culture. Future analyses in other vertebrate model systems, including avian embryos, will facilitate the examination of specific molecular pathways and developmental processes that require calcineurin signaling during embryogenesis.

EXPERIMENTAL PROCEDURES

Isolation of Chicken CnAα and CnAβ

Chicken CnAα (802 bp) and CnAβ (799 bp) sequences were amplified from reverse transcription reactions, including embryonic day 2 head RNA generated as previously described (Searcy and Yutzey, 1998). Degenerate oligonucleotides for CnAα isolation were 5′-CYGTYCCHTTTCCNCCAA-3′ (upper) and 5′-GVACRTTTACMARCATYTC-3′ (lower). CnAβ was isolated by using 5′ -TTTGAAGTNGGRGGATC-3′ and the same lower primer as for CnAα. PCR reactions included 200 pmol of oligonucleotide and were conducted for 30 cycles of 94° 1 min, 50° 1.5 min and 72° 3 min. The resulting DNA fragments (∼800 bp) were subcloned into pBluescript T-vector and subjected to sequence analysis. CnA isoform identity was determined after BLAST sequence alignment for several independent clones and only CnAα (802 bp) and CnAβ (799 bp) were identified. A more complete CnAα cDNA (2042 bp) was isolated after screening a stage 4–6 chicken embryo heart forming region λScreen library (Novagen) by using the CnAα (802-bp) fragment by previously described methods (Searcy and Yutzey, 1998). The full-length sequence was determined, and its identity as CnAα was confirmed by multiple sequence alignment. GenBank accession numbers are AY324834 for CnAα and AY324833 for CnAβ.

Isoform-specific primers for CnAα were 5′-TTCAACTGCTCCCCTCATCCTTAC-3′ and 5′-ATAGCGCTTTGCAGTGTTTGTTTT-3′ and for CnAβ were 5′-CTTCCTCGGTGACTATGTAGACAG-3′ and 5′-AATGCTGGAGGTTCTTTGAATC-3′. PCR amplifications were performed as above with RT reactions from whole E1 embryos (stages 5–8) or dissected E2 (stages 11–14) embryos, with the exception that the annealing temperature was 55°C. Isoform-specificity of amplification reactions was confirmed by PCR reactions with CnAα or CnAβ plasmid DNA templates. GAPDH amplification was included as a loading control for each RT sample as previously described (Searcy and Yutzey, 1998). PCR reactions were electrophoresed in the presence of ethidium bromide, and the ultraviolet-irradiated gel was documented using the Bio-Rad Gel Doc 2000 system.

In Situ Hybridizations

Fertilized white Leghorn chicken eggs (obtained from Charles River, North Franklin, CT) were incubated at 38°C in high humidity for 1–10 days. Embryos were isolated and in situ hybridizations performed as described previously (Searcy and Yutzey, 1998; Ehrman and Yutzey, 1999). E7–E10 whole hearts were bisected after dehydration and proteinase K treatments were performed for 12–15 min. CnAα antisense probes were generated with T3 polymerase from XhoI linearized plasmid for 802-bp riboprobe or with T7 from BamHI linearized plasmid for 2100-bp riboprobe. CnAβ antisense riboprobe was generated from XhoI linearized plasmid with T3 polymerase. VMHC1 and AMHC1 probes were generated and used for hybridizations of whole embryos and cultured explants as described previously (Yutzey et al., 1994, 1995; Liberatore et al., 2000).

Cyclosporin Treatments

Cyclosporin A (Sandimmune, Novartis) diluted in phosphate buffered saline (PBS) to 1 mg/ml was injected directly into chicken eggs toward the blunt end airspace by using a 28G1/2 insulin syringe. A teratologic dose of CsA 100 μg in 100 μl per egg (∼1.6 mg/kg) was determined empirically, but milder developmental defects also were observed with 50 μg of CsA per egg. Treatments with 500 μg of CsA were cytotoxic, and administration of 10 μg of CsA per egg had no obvious effect on development or embryonic viability. For analysis of the effects of calcineurin inhibition during development, chicken eggs were injected with 100 μg of CsA in 100 μl on E1 (stages 5–8) or E2 (stages 11–13) and incubated an additional 1 or 2 days. Control embryos were injected with PBS in parallel experiments. For local CsA treatments, 1-mm agar squares (2% agar in PBS) were soaked in 10 mg/ml CsA for at least 1 hr and placed directly on E1 (stages 7–8) or E2 (stages 11–12) embryos. Eggs were windowed by using standard procedures, resealed with tape, and incubated 1–2 days further (Selleck, 1996). After treatment, morphology and development of the cardiovascular system was assessed visually and documented photographically. Embryos were subsequently fixed in 4% paraformaldehyde for in situ hybridization or histology. Morphology of CsA-treated and control embryos was assessed in 5-μm paraffin sections stained with hematoxylin and eosin. For treatments of cardiogenic tissue in culture, anterior and posterior heart forming region explants were isolated at stage 5 as previously described (Yutzey et al., 1995) and cultured in the presence or absence of 0.5 μM CsA in M199 medium. Explants were incubated for 2 days and fixed for in situ hybridization with AMHC1 or VMHC1 probes (Yutzey et al., 1995).

Acknowledgements

We thank Paul Bushdid, Robin Searcy-Schrick, Christina Alfieri, Ron Waclaw, Jeff Molkentin, Orlando Bueno, Alex Lange, and members of the Division of Molecular Cardiovascular Biology for technical support and scientific advice. K.E.Y. received an Established Investigator Award from the American Heart Association.

Ancillary