Drs. O'Brien and Gourdie are co-senior authors.
Differentiation of cardiac Purkinje fibers requires precise spatiotemporal regulation of Nkx2-5 expression
Article first published online: 21 OCT 2005
Copyright © 2005 Wiley-Liss, Inc.
Volume 235, Issue 1, pages 38–49, January 2006
How to Cite
Harris, B. S., Spruill, L., Edmonson, A. M., Rackley, M. S., Benson, D. W., O'Brien, T. X. and Gourdie, R. G. (2006), Differentiation of cardiac Purkinje fibers requires precise spatiotemporal regulation of Nkx2-5 expression. Dev. Dyn., 235: 38–49. doi: 10.1002/dvdy.20580
- Issue published online: 7 DEC 2005
- Article first published online: 21 OCT 2005
- Manuscript Accepted: 4 AUG 2005
- Manuscript Revised: 18 JUL 2005
- Manuscript Received: 21 JUN 2005
- NIH. Grant Numbers: HL56728, HL36059, HD39946
- Research and Development Service of the Department of Veteran Affairs
- National Center for Research Resources. Grant Number: P20RR016434
- American Heart Association. Grant Number: 0425546U
- heart conduction system;
- Purkinje fibers;
- atrioventricular block
Nkx2-5 gene mutations cause cardiac abnormalities, including deficits of function in the atrioventricular conduction system (AVCS). In the chick, Nkx2-5 is elevated in Purkinje fiber AVCS cells relative to working cardiomyocytes. Here, we show that Nkx2-5 expression rises to a peak as Purkinje fibers progressively differentiate. To disrupt this pattern, we overexpressed Nkx2-5 from embryonic day 10, as Purkinje fibers are recruited within developing chick hearts. Overexpression of Nkx2-5 caused inhibition of slow tonic myosin heavy chain protein (sMHC), a late Purkinje fiber marker but did not affect Cx40 levels. Working cardiomyocytes overexpressing Nkx2-5 in these hearts ectopically up-regulated Cx40 but not sMHC. Isolated embryonic cardiomyocytes overexpressing Nkx2-5 also displayed increased Cx40 and suppressed sMHC. By contrast, overexpression of a human NKX2-5 mutant did not effect these markers in vivo or in vitro, suggesting one possible mechanism for clinical phenotypes. We conclude that a prerequisite for normal Purkinje fiber maturation is precise regulation of Nkx2-5 levels. Developmental Dynamics 235:38–49, 2006. © 2005 Wiley-Liss, Inc.
Nkx2-5 (Csx/Tinman) is an NK2 class homeodomain transcription factor expressed in diverse species (Evans, 1999). In vertebrate embryos, Nkx2-5 is one of the earliest markers of heart-forming potential. Functional studies have shown that postgastrulation expression of Nkx2-5 is required for specification of cardiomyogenic cell fate. Absence of Nkx2-5 expression leads to abnormal heart formation in Drosophila, Xenopus, and mouse (Grow and Krieg, 1998). Nkx2-5 knockout mice die in utero shortly after looping morphogenesis and display disrupted expression of cardiac genes (Lyons et al., 1995). Whereas these findings establish a function for Nkx2-5 in early cardiogenesis, emerging data suggests this gene has additional developmental roles. Clinical interest in this gene arises from the identification of heterozygous mutations in human NKX2-5, including the Gln170ter NKX2-5 truncation mutant, that cause heart disease (Schott et al., 1998). Although several structural and functional cardiac abnormalities are associated with such mutations, one common and salient phenotype is atrioventricular conduction block (Schott et al., 1998). Of interest, the atrioventricular (AV) block phenotype emerges during postnatal development and is believed to result from progressive disease of the atrioventricular conduction system (AVCS). This prospect is supported by our recent work in Nkx2-5 haploinsufficent mice, where we reported that loss of a single Nkx2-5 allele is sufficient to cause a severe hyoplasia of the AVCS in the postnatal heart associated with a greater than 50% loss of Cx40-expressing Purkinje fibers in the ventricle (Jay et al., 2004).
The finding that Nkx2-5 expression is elevated in the forming AVCS of higher vertebrates, has provided evidence of a role for this transcription factor in conduction tissue development (Takebayashi-Suzuki et al., 2001; Thomas et al., 2001). Purkinje fiber cells of the AVCS coordinate rapid spread of action potential (AP) in the ventricular myocardium (Pennisi et al., 2002; Gourdie et al., 2003; Moorman and Christoffels, 2003). To fulfill this role of fast AP conduction, Purkinje fibers express a characteristic set of genes. Such genes may be either uniquely or differentially expressed relative to working cardiomyocytes and encode proteins including ion channels, contractile proteins, and signaling molecules. In a process best characterized in the avian embryo (Harris et al., 2002), the Purkinje fiber-specific pattern of gene expression emerges progressively after septation of the ventricles. Genes up-regulated early during this maturational differentiation include the gap junction proteins Cx40 and Cx45 (Gourdie et al., 1993; Delorme et al., 1995). A well-characterized marker of late phases of Purkinje fiber differentiation in the chick is the slow tonic myosin heavy chain protein (sMHC), which initiates expression just before hatching. At present, the molecular cues regulating the progressive differentiation of Purkinje fiber phenotype are unknown. Previously, we have shown that Nkx2-5 is first up-regulated in prospective Purkinje fibers during inductive recruitment of these cells into the AVCS. In this study, we demonstrate that levels of nuclear-localized Nkx2-5 in Purkinje fiber cells continue to increase throughout development and that disruption of this pattern by prematurely up-regulating Nkx2-5 levels has differential effects on genes marking early and late stages of differentiation of conduction cells. Our results suggest that there is a requirement for precise regulation of Nkx2-5 level during differentiation of Purkinje fiber cells.
Elevated Nkx2-5 Protein Is Expressed Throughout Purkinje Fiber Cell Differentiation
Previously, we have demonstrated that Nkx2-5 is elevated in the developing AVCS of the chick heart relative to working myocardium (Thomas et al., 2001). In the present study, we addressed the mechanistic relevance of this expression pattern by focusing on the role of Nkx2-5 in the differentiation of the peripheral network of subendocardial Purkinje fibers (SPFs) and periarterial Purkinje fibers (PPFs). We first characterized the localization of Nkx2-5 protein expression by multilabel immunohistochemistry using three specific markers that show a progressive sequence of up-regulated expression over the development of Purkinje fiber cells. At embryonic day 12 (E12), the gap junction protein connexin40 (Cx40) provides a marker of the initial stages of Purkinje fiber differentiation (Gourdie et al., 1993). At E12, Nkx2-5 localized to the nuclei of Cx40-positive Purkinje fibers at higher levels relative to working cardiomyocytes (Fig. 1A). At E15, subsequent differentiation of Purkinje fibers is characterized by the expression of the intermediate filament protein transitin/EAP-300 (McCabe et al., 1995). We also noted that higher levels of Nkx2-5 labeling were maintained within transitin-positive Purkinje fibers at this stage (Fig. 1B). The later stages of conduction cell differentiation in the chick embryo are marked by the onset of expression of the sMHC at E17 (Gonzalez-Sanchez and Bader, 1985). At E19, increased levels of nuclear Nkx2-5 immunolabeling were associated with sMHC-positive PPFs (Fig. 1C). Elevated Nkx2-5 levels also could be detected in sMHC-positive Purkinje fibers during the posthatching period (Fig. 1D). In summary, these results confirmed that Nkx2-5 is elevated in the nuclei of bona-fide conduction cells and also demonstrated for the first time that elevation of this Nkx2-5 signal was maintained at early, intermediate, and late stages of Purkinje fiber differentiation.
Nuclear-Localized Nkx2-5 Increases During Purkinje Fiber Maturation
Figure 1 shows that Nkx2-5 protein is elevated in the nuclei of differentiating conduction cells. However, it was our impression that nuclear-localized Nkx2-5 in Purkinje fiber cells continued to increase relative to working cardiomyocytes as development progressed. To quantify this increase, sections from E8, E12, E19, and postnatal day 7 (P7) hearts were multilabeled with Cx40 and Nkx2-5 antibodies and the nuclear label To-Pro3 and then imaged by confocal optical sectioning (arrows in Fig. 2A–C). In Figure 2A–C, note the preservation of nuclei within the thin endocardial layer (arrowheads) over SPFs, indicative of the excellent histological integrity of these heart sections. At each stage analyzed, nuclear To-Pro3 and Cx40 immunolabeling were used to confirm that Nkx2-5 was specifically increased in SPF nuclei. For clarity, these images are repeated below without the nuclear To-Pro3 counter stain (Fig. 2D–F). By staining samples at the same time using identical immunolabeling protocols and maintaining constant settings on the confocal microscope, we were able to compare relative levels of Nkx2-5 signal between developmental stages. Using this rigorously controlled approach, it was determined that the mean intensity of Nkx2-5 was at least twofold higher in SPF nuclei relative to nuclei of working cardiomyocytes throughout development (Fig. 2G). Moreover, consistent with our qualitative impression, nuclear-localized Nkx2-5 increased significantly (P < 0.05) from a twofold elevated level at E12 to a fourfold increase over that of working cardiomyocyte nuclei at E19, before dropping back to twofold increased levels 1 week after hatching (P7).
Adenoviral Targeting of Developing Chick Heart In Vivo Using a Shell-Less Culture Model
To explore the function of the progressive increase in nuclear-localized Nkx2-5, we devised a strategy to overexpress Nkx2-5 in conduction cells from embryonic day (E) 10, a time corresponding to the earliest stages of Purkinje fiber differentiation in advance of the appearance of markers, including Cx40, transitin, and sMHC (Fig. 1 and Harris et al., 2002). To facilitate this strategy, the shell-less culture protocol of Dunn (1974) was used in which chick embryos isolated from the egg shell at E2 can be cultured up until E19 (Fig. 3A,B). Heart development of shell-less embryos was histologically normal (Fig. 3C,D). Of particular utility was that the shell-less protocol facilitated direct visualization of the heart at the E10 time point, enabling high precision microinjection of adenoviral expression vectors at the stage of interest, i.e., the initiation of Purkinje fiber development. The adenoviral vector used, AdNkxHA, has been characterized in earlier studies (Muller et al., 2002). This construct is bicistronic, expressing a HA-tagged Nkx2-5 and green fluorescent protein (GFP) under independent CMV promoters (Fig. 3E). Western blotting of cultures of chick cardiomyocytes isolated from E3.5 embryos infected by AdNkx2-5 confirmed the ability of the virus to elevate Nkx2-5 levels (Fig. 3F). After microinjection at E10, shell-less embryos were incubated for a further 5–9 days. Survival was good, with 70% of embryos being viable 7 days after injection. Whole-mount heart preparations were visualized by microscopy, and sectors of GFP fluorescence were distinguished (Fig. 3G). The pattern of viral infection varied, depending on the region injected. Typically, sectors were extensive, encompassing the right ventricular myocardium and regions of the left ventricles. HA immunolabeling of histological sections confirmed that large numbers of cardiomyocytes in such GFP-positive sectors expressed the exogenous HA-tagged Nkx2-5 within nuclei (Fig. 3H). Cardiomyocytes overexpressing the exogenous Nkx2-5 were distributed throughout the myocardium, from epicardium to endocardium (Fig. 3I). HA immunolabeling of nuclei was absent in hearts microinjected with AdGFP or vehicle control solution (data not shown).
Constitutive Expression of Nkx2-5 Disrupts Progressive Differentiation of Purkinje Fibers In Vivo
Immunolocalization studies of AVCS markers were carried out on vehicle control and AdGFP- and AdNkxHA-infected E19 hearts. GFP-positive Purkinje fibers in hearts infected with the control AdGFP virus showed normal patterns of AVCS development and Nkx2-5 expression (Fig. 4A). Of interest, in hearts targeted with the AdNkxHA virus, Cx40 immunolabeling patterns of HA-positive sectors of the subendocardium were not altered (Fig. 4B). Thus, maintenance of overexpressed Nkx2-5 at higher than endogenous levels for a protracted period starting at E10 did not appear to affect Cx40 expression by Purkinje fibers. However, analysis of AdNkxHA-positive domains within regions of the working myocardium in the same hearts revealed a different response. The working ventricular myocardial tissues of the developing chick normally express low or negligible levels of Cx40 (Gourdie et al., 1993). Consistent with this finding, working ventricular myocardium infected with control AdGFP adenovirus revealed no evidence of Cx40 expression. However, similar regions of ventricle infected with the AdNkxHA adenovirus showed a striking up-regulation of Cx40 (compare Fig. 4C,D with 4E,F). These ectopic foci of bright punctate Cx40-immunopositive gap junctions were colocalized within cells expressing GFP, mediated by AdNkxHA infection. These data were reproducible in multiple experimental animals. Thus, Nkx2-5 overexpression was sufficient to up-regulate Cx40 ectopically in working cardiomyocytes in vivo.
More significantly with respect to specialized myocardial differentiation was the effect of Nkx2-5 overexpression on the late-expressed marker of Purkinje fiber phenotype, sMHC. Figure 4G shows a section through the right ventricular free-wall of an E19 heart that had been targeted with AdNkxHA. Prominent HA immunolabeling can be seen subendocardially within this region, as well as scattered nuclei within the myocardium, indicating the presence of cardiomyocytes overexpressing exogenous Nkx2-5 (Fig. 4G). In the subendocardium, normally highly immunopositive for sMHC at E19, there is a complete absence of signal in the sister section of that shown in Figure 4G (Fig. 4H). No such loss of sMHC-immunopositive cells was observed in uninfected control hearts incubated normally and in control hearts targeted ex ovo with the AdGFP virus (Fig. 4I). This change in the expression of sMHC was consistent with an inhibition of the normal progression of Purkinje fiber differentiation in hearts injected with the AdNkxHA virus. Of interest, although working myocardial cells infected by AdNkxHA showed increased levels of Cx40 immunolabeling (Fig. 4E), these same cells showed no detectable up-regulation of sMHC (data not shown).
Overexpression of Nkx2-5 Within Embryonic Cardiomyocytes In Vitro Differentially Affects AVCS Markers
At embryonic day 3.5 (E3.5), neither Cx40 nor sMHC are expressed at detectable levels in the ventricular myocardium of the chick embryo. However, after 3 days in culture, a proportion (∼5%) of cardiomyocytes derived from E3.5 hearts immunolabel for sMHC (Gourdie et al., 1998). Consistent with these findings, we confirmed that, when embryonic chick cardiomyocytes from E3.5 hearts were cultured for a period of 5 days, a subpopulation of MF20-positive cells immunolabeled for sMHC (Fig. 5A). The frequency of sMHC-immunolabeled cells in these cultures was increased by treatment with endothelin-1, which was also consistent with our previous results (data not shown). Bu using quantitative real-time reverse transcriptase-polymerase chain reaction (RT-PCR), Western blotting, and immunocytochemistry, we sought to investigate the effect of exogenous Nkx2-5 overexpression on levels of Cx40, sMHC, and endogenous Nkx2-5 in this in vitro model of conduction cell differentiation (Fig. 5).
Whereas exogenous Nkx2-5 expression had no effect on levels of endogenous chick Nkx2-5 mRNA, the Cx40 mRNA level showed proportional increases at multiplicities of AdNkx2-5HA infection (i.e., MOIs) of 2 and 10 (Fig. 5B). Western blotting of protein from the same samples confirmed that Cx40 protein was increased in response to overexpression of Nkx2-5 (Fig. 5C). In contrast to the effect of Nkx2-5 overexpression on Cx40, sMHC mRNA was reduced in cultures of E3.5 cardiomyocytes at an AdNkx2-5HA MOI of 10 (Fig. 5B). To explore this finding further, we infected cultures with increasing MOIs of 2, 5, 10, and 20 and used real-time RT-PCR to assay Cx40 and sMHC mRNAs (Fig. 5D). Cx40 mRNA levels showed a linear increase in relation to AdNkxHA MOI. By contrast, sMHC mRNA levels showed a steady downward trend as viral MOI increased.
To further characterize of the effect of Nkx2-5 overexpression in vitro, we examined sMHC and Cx40 protein expression by immunocytochemically after AdNkx2-5HA infection. Consistent with PCR and Western blot data, cultures infected with AdNkxHA consistently showed increases in Cx40 immunolabeling and suppression of sMHC relative to AdGFP and noninfected controls (Fig. 5E). The inhibition of sMHC immunolabeling by Nkx2-5 was particularly striking an observation that was reproduced in at least 10 independent experiments. In summary, the results derived from this in vitro model of Purkinje fiber cell differentiation were similar to those obtained in vivo. On the one hand, overexpression of Nkx2-5 caused up-regulated Cx40 expression in these cultures. In contrast, the same treatment inhibited sMHC, quantifiably suppressing the expression of this late marker of Purkinje fiber development.
Mutant NKX2-5 Associated With AV Block Has No Effect on Cx40 In Vivo or In Vitro
In the final part of this study, we used our in vivo and in vitro models of Nkx2-5 function to probe the Gln170ter NKX2-5 mutant, which was among the first NKX2-5 mutations to be described in humans (Schott et al., 1998). This mutation occurs at nucleotide position 618, within the DNA binding homeodomain and is predicted to result in a truncated NKX2-5. We cloned this mutant into an adenoviral vector, termed AdM170, and using an identical strategy as above, we targeted cardiomyocytes in vivo and in vitro. Immunoanalysis of the working myocardium within AdM170 targeted sectors of shell-less cultured chick hearts revealed that Cx40 levels remained unaltered (Fig. 6A,B). In fact, the low levels of Cx40 immunostaining remained identical to controls (see Fig. 3E,F). This finding was in striking contrast to similar regions of ventricle infected with the AdNkxHA virus, which as we outline above, displayed ectopic foci of Cx40 up-regulation (Fig. 6C,D, see also Fig. 3C,D). We also investigated what affect this truncated NKX2-5 mutant had on Purkinje fibers. As detailed above, overexpression of wild-type Nkx2-5 did not alter Cx40 immunolabeling patterns within Purkinje fibers. Infection with AdM170 also failed to effect the expression of Cx40 within any of the Purkinje fibers analyzed (Fig. 6E,F). Thus, overexpression of the truncated NKX2-5 mutant in developing conduction cells did not appear to disrupt normal levels of Cx40, suggesting that this mutant may not exert a dominant negative affect. We next investigated how the function of this Nkx2-5 mutant differed from wild-type Nkx2-5 using the in vitro model. Western blotting of these embryonic cardiomyocyte cultures confirmed the expression of Nkx2-5 protein after infection with constructs AdNkx2-5HA (HA-tag immunoblot) and AdM170 (FLAG-tag immunoblot) (Fig. 6H). Consistent with earlier results, AdNkx2-5HA induced up-regulation of Cx40 in comparison to control cultures in this experiment. By contrast, cells infected with AdM170 did not elicit a comparable effect, Cx40 remained low and similar to control levels. These data indicated that, unlike wild-type Nkx2-5, the NKX2-5 M170 mutant was unable to increase Cx40 ectopically in cardiomyocytes in vivo or in vitro.
In this study, we show that cardiac Purkinje fibers express at least twofold more Nkx2-5 in cell nuclei relative to adjacent working cardiomyocytes throughout differentiation. Moreover, using genes marking initial, intermediate, and later stages of Purkinje fiber development, it is demonstrated that this nuclear-localized Nkx2-5 signal increases progressively up until hatching. Constitutive overexpression of Nkx2-5 at the initiation of Purkinje fiber development disrupts this maturational sequence of gene expression, in particular inhibiting sMHC, a protein up-regulated at later phases of conduction cell differentiation. Finally, we show that a mutant Nkx2-5 form, M170, known to cause cardiac disease in humans, does not reciprocate the effects of overexpression of wild-type Nkx2-5, indicating that M170 may disrupt AVCS development by acting as a nonfunctional allele, reducing gene dosage by means of haploinsufficiency. Based on these data, we propose that elevated Nkx2-5 defines a population of cardiomyocytes actively differentiating into Purkinje fibers. Furthermore, we suggest that precise temporal regulation of Nkx2-5 level is necessary for normal differentiation of Purkinje fiber cells.
We and others have reported that Nkx2-5 is expressed at elevated levels in conduction tissues of the bird (Takebayashi-Suzuki et al., 2001; Thomas et al., 2001). Our earlier study also noted that the timing of increased Nkx2-5 in central and peripheral conduction cells corresponded with the sequence of cellular recruitment to these different AVCS tissues established by retroviral clonal analyses and birthdating studies (Gourdie et al., 1995; Cheng et al., 1999; Sedmera et al., 2003). Thus, in the present study, an important advance is the ability to correlate temporal variation in Nkx2-5 immunolabeling to a sequence of protein markers defining stages of Purkinje fiber maturation. As we show here, by precociously up-regulating Nkx2-5 levels, we were able to modify this sequence of gene expression, thereby implicating the importance of strict temporal regulation of Nkx2-5 dosage to the progressive maturation of Purkinje fiber phenotype.
The initial induction of Purkinje fibers has been shown to be associated with factors secreted by endothelial cells, including endothelin and neuregulin, and may involve input by epigenetic mechanical factors such as shear stress and tissue strain (Gourdie et al., 1998; Rentschler et al., 2002; Sedmera et al., 2003; Patel and Kos, 2005). For peripheral Purkinje fibers in vivo, the timing of this inductive recruitment corresponds to the initiation of Cx40 expression at E12. We show here that the onset of Cx40 expression in Purkinje fibers is marked by the coincident up-regulated expression of Nkx2-5 (quantified as a twofold induction). Given that Purkinje fibers subsequently undergo a progressive sequence of differentiation, apparently determined at least to some degree by increasing levels of nuclear-localized Nkx2-5, an interesting question is how is this process orchestrated? On the one hand, regulation of Nkx2-5 levels may be determined genetically. Certainly, the Cx40 promoter contains binding sites for Nkx2-5, Tbx5, and members of the GATA transcription factor family: the latter two being known binding partners of Nkx2-5 (Lee et al., 1998; Sepulveda et al., 1998; Hiroi et al., 2001; Sedmera et al., 2003). Nkx2-5 may transactivate Cx40 either alone or in association with Tbx5, another transcription critical to AVCS development (Hatcher et al., 2003; Moskowitz et al., 2004). However, what of the later-expressed markers of Purkinje fibers such as sMHC? As we demonstrate in our in vivo and in vitro models, sMHC is particularly sensitive to increase in exogenous Nkx2-5, and the maintenance of expression of sMHC in vivo appears to be associated with decreases in levels of nuclear-localized Nkx2-5. Ongoing work is necessary to determine whether Nkx2-5 down-regulation and sMHC up-regulation in vivo occurs as a consequence of a predetermined genetic program or whether these changes occur in response to paracrine cues, as is probably the case in the initial induction of Purkinje fibers. Most likely, a mechanism involving interplay between both genetically determined and inductive signaling processes orchestrate this differentiation.
While sMHC was sensitive to Nkx2-5 dose, Cx40 levels in Purkinje fibers were not noticeably perturbed by overexpression of Nkx2-5 in vivo. The results, furthermore, indicate that, although Nkx2-5 is sufficient to increase Cx40 in vitro in E3.5 cardiomyocytes, further elevation of Nkx2-5 after E10 in vivo will not drive Cx40 expression yet higher. These observations parallel results from our recent study in the Nkx2-5 heterozygous null mouse. Whereas a reduction in Nkx2-5 dosage by means of haploinsufficiency resulted in a greater than 50% loss of Purkinje fibers in this mouse model, Cx40 within the residual conduction cells remained at quantifiably normal levels (Jay et al., 2004). Together, the data in chick and mouse suggest that, once selection of Purkinje fiber fate has been initiated, incipient conduction cells become refractory to subsequent alterations in Nkx2-5 dose, at least with respect to Cx40 expression. We would emphasize that the same paradigm does not appear to hold for Cx40-negative cardiomyocytes, which can be induced to up-regulate Cx40 by Nkx2-5 overexpression. Whether the up-regulation of Cx40 represents an initial step down the Purkinje fiber differentiation pathway or a transactivation of a single gene with no particular implications for the potential of cardiomyocytes to undergo subsequent differentiation into a conduction cell remains to be determined.
It is noteworthy that our observation of up-regulated Cx40 in working cardiomyocytes in response to exogenous Nkx2-5 overexpression differs from the results of Kasahara and colleagues (2001). These workers reported that overexpression of wild-type Nkx2-5 in adult mouse cardiomyocytes caused decreased expression of Cx43, and no detectable change in Cx40. This apparent disparity may reflect interspecies variation and/or changes in the competence of different-staged cardiomyocytes to respond to Nkx2-5. It is of pertinence here that Nkx2-5 has been shown to be rapidly up-regulated in working myocardial tissues in a mammalian model of hypertrophy (Thompson et al., 1998). Elevated levels of Nkx2-5 have also been detected in the ventricle after stimulation of hypertrophy by adrenergic agonists (Saadane et al., 1999). As such, the ectopic up-regulation of Nkx2-5 in the working myocardium that we model in the embryonic chick may be an aspect of certain pathologic conditions of the mature heart. Of interest, Cx40 up-regulation has also been observed within the myocardium of humans with congestive heart failure (Dupont et al., 2001b), and increased Cx40 levels are associated with predisposition to atrial fibrillation (Dupont et al., 2001a). Although these studies did not assay Nkx2-5, our data showing that overexpression of this transcription factor is sufficient to up-regulate Cx40 expression in working cardiomyocytes in vivo, suggests the possibility of relationships between these two proteins in the diseased myocardium.
A second aspect of this study with implications for cardiac disease processes relates to the use of our in vitro and in vivo models to assay the function of known NKX2-5 mutants. Over 30 NKX2-5 mutations have been identified in humans that result in functional and structural cardiac abnormalities including AV block (Schott et al., 1998; Benson et al., 1999; Goldmuntz et al., 2001; Gutierrez-Roelens et al., 2002; Ikeda et al., 2002; Watanabe et al., 2002; McElhinney et al., 2003). The presence of AV block is especially associated with mutations involving the homeodomain of Nkx2-5 (Benson et al., 1999; Gutierrez-Roelens et al., 2002). One such mutant is the M170 variant in which the mutated NKX2-5 locus expresses a carboxy-terminal truncation at the level of homeodomain. Unlike wild-type Nkx2-5, overexpression of the M170 truncation mutant in our in vitro and in vivo models of Purkinje fiber differentiation did not appear to disrupt Cx40 or sMHC expression patterns. In both models, endogenous wild-type Nkx2-5 alleles were present, indicating that the M170 mutant is not exerting a dominant negative affect. As such, we would conclude that the Nkx2-5 M170 protein is probably a null mutation and results in loss of Nkx2-5 function by means of haploinsufficiency. Other recent studies in mice have pointed to the importance of Nkx2-5 gene dosage in AVCS development (Jay et al., 2004; Pashmforoush et al., 2004). Here, further novel insight is provided with the demonstration that elevation and subsequent controlled increases in Nkx2-5 level are required for normal Purkinje fiber differentiation. This requirement for precise spatiotemporal regulation of Nkx2-5 is likely to have implications in the development of therapeutic strategies designed to ameliorate the effects of Nkx2-5 mutation on AVCS function.
Adenovirus Vector Construction and Infection of Isolated Cardiomyocytes
The pAd-Track shuttle using the method of Vogelstein and the pAd Easy-1 adenoviral vector system (Stratagene, La Jolla, CA) were used to clone human wild-type Nkx2-5 and a known Gln170ter Nkx2-5 truncation mutant cDNA, termed AdM170, as previously described (He et al., 1998). Nkx2-5 M170 was cloned in a similar manner (Schott et al., 1998). AdGFP, an adenovirus expressing green fluorescent protein (GFP) driven by the CMV promoter, was a kind gift of Larry Rothblum's laboratory.
Isolated embryonic cardiomyocytes received AdNkxHA, AdM170, or AdGFP at an MOI of 10. GFP was the marker of infection, and expression was confirmed using immunocytochemistry and Western blotting (HA tag and FLAG). Data shown are representative of multiple independent experimental preparations of primary cardiomyocytes.
Shell-Less Chick Culture Preparations and Microinjections
Chick eggs were incubated to E2 and then cracked into polystyrene dishes (Fisher Scientific, Rockville, MD; Dunn, 1974). These were placed into Petri dishes containing sterile water to maintain humidity. Chicks were maintained at 38.5°C until E10. The hearts of these embryos were visualized and targeted by direct trans-thoracic microinjection using a Leica dissection microscope and Pico Pulser apparatus inside a glove box. Injection volumes were 5 and 10 μl of vehicle control (PBS) or high titer adenoviral vectors (1 × 106 PFU). After microinjection, shell-less embryos received an eggshell supplement to the albumin to sustain developmental calcium requirements. Shell-less chick preparations were further cultured up to and including E19. Finally, the embryos were killed by decapitation. The chest cavities were opened, and the chicks were placed into 4% paraformaldehyde for 2 hr. After 3 × 30-min PBS washes, the chicks were examined with a fluorescent dissection microscope for GFP expression in the heart, and whole-mount images were captured. Adenoviral infected hearts were then cryoprotected and processed for frozen sectioning.
Embryonic Chick Cardiomyocyte Cultures
Cardiomyocytes were isolated from the ventricles of E3 chicken embryos as detailed previously (Gourdie et al., 1998). Cardiomyocytes were dissociated using a low concentration of trypsin (0.1%) for 10 min before plating on slides or Petri dishes. Cultures were maintained in medium M199 (Gibco, Grand Island, NY) supplemented with penicillin and streptomycin, 1% chick serum, and 1:100 of insulin, transferrin, and selenium. Cells were allowed to attach for 12 hr, before the addition of medium containing adenoviral constructs. After a further 12 hr, this medium was exchanged for fresh medium. Cultures were incubated for up to 5 days at 37°C with 5% CO2 and were terminated by fixing with 4% paraformaldehyde (immunocytochemistry) or by scraping for Western blotting.
Protein Extraction and Western Blotting
Isolated cardiomyocytes were homogenized ultrasonically in protein extraction buffer as previously described (Muller et al., 2002). Normally, triplicate samples were separated using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) under denaturing conditions. Separated proteins were transferred to nitrocellulose membrane, Hybond-N, in gel transfer tanks (Bio-Rad, Richmond, CA). Nonspecific binding sites were blocked for 1 hr using 10% fat-free milk powder and 1% bovine serum albumin in TBST (10 mM Tris, 100 mM NaCl, 0.1% Tween 20) at 4°C. Antibodies included anti-HA, anti-FLAG (Sigma, St. Louis, MO), and anti–glyceraldehyde-3-phosphate dehydrogenase (Research Diagnostics, Inc., Minneapolis, MN). Antibodies were incubated overnight at 4°C in 1% fat-free milk powder TBST, at appropriate concentrations. Blots were washed in TBST, 6 × 5 min at room temperature. Secondary antibodies, directed against primary immunoglobulins, conjugated to horseradish peroxidase were used at 1:2,000 in TBST for 1 hr at room temperature. Blots were finally washed in TBST 6 × 5 min at room temperature before being developed for 1 min using the substrate Western Lightning Chemiluminescence Reagent Plus (NEN Life Science Products, Austin, TX) per manufacturer's protocol. Blots were covered in cling film and visualized by exposure to Kodak BioMax Mr-1 film developed from 45 sec to 10 min later in an X-Ograph Compact X2 developer.
Quantitative Real-Time RT-PCR
Total RNA was extracted from embryonic cardiomyocyte cultures by using Ultraspec RNA total RNA isolation agent. After RNA isolation, the RNA was resuspended in a 1-mM concentration of MgCl2. RQ1 DNase was added to the samples, which were incubated at 37°C for 20 to 30 min to remove DNA from the reactions. The DNase was then heat inactivated for 5 min at 90°C. RNA concentration was determined by spectroscopy and analyzed for integrity of the RNA by agarose gel electrophoresis under denaturing conditions. Equal amounts of RNA, equivalent to 250 ng of total RNA from the various samples were used for RT-PCR, together with a water control with oligonucleotide primer pairs. PCR was performed in 25-μl reaction volumes containing 20 mM Tris-HCl, pH 8.4, 50 mM KCl, 1.5 mM MgCl2, 0.2 mM dNTPs, 25 pmol of each primer, and 2.5 units Taq DNA-polymerase. The reaction conditions for each primer pair were optimized with respect to MgCl2 concentration, annealing temperature, extension time, and cycle number. The primers used were as follows: Cx40 forward GGAGGAGAAGAGAAAGATGAAG, reverse TCCCAGCAAGACAACTCAG; sMHC forward GCGGGAAGAGCAGGCAGAG, reverse TCACACGAGGATGGAGCAAGC; and 18S forward TATGGTTCCTTTGGTCGCTC, reverse GGTTGGTTTTGATCTGATCTGATAAAT.
To quantify mRNA levels, PCR reactions were performed in a Bio-Rad Icycler in combination with a Quantitect SYBR Green RT-PCR detection kit (Qiagen, Chatsworth, CA). PCR conditions were as follows: (1) 1 RT-PCR step of 45 min, 1 polymerase activation step of 15 min, followed by 35 cycles of a 15 sec at 94°C (denaturation); (2) 30 sec at 60°C (annealing); and (3) 30 sec at 72°C (elongation). A melt-curve analysis was used to determine specific target amplification. Upon completion, gel loading buffer (10×) was added to each sample and 10μl of each reaction was electrophoresed on 2% agarose gels, including ethidium bromide in 1× Tris acetate ethylenediaminetetraacetic acid (EDTA; TAE) buffer in a Horizon 58 gel apparatus (BRL-Life Technologies, Gaithersburg, MD). Loading buffer consisted of 0.5 M EDTA, pH 7.5, 10% SDS, 50% glycerol, and 0.25% bromophenol blue. Electrophoresed bands were visualized on a ultraviolet photometry dual intensity transilluminator and documented using Polaroid photography. Data were first normalized to 18S levels and then the DDCT method used to analyze samples, where control values were set to 1. All experiments were repeated in triplicate and carried out on multiple experiments as detailed in figure legends.
Multilabel Immunoconfocal Microscopy
Frozen sections of 8-micron thickness were blocked for 1 hr at room temperature and incubated with antibodies overnight at 4°C at concentrations determined to give the best signal to noise ratio. Primary antibodies used were anti-mouse Cx40 (Chemicon, Temecula, CA), anti-human Nkx2-5 and anti-GATA4 (Santa-Cruz Biotechnology, Inc., CA), anti-sMHC (Ald-58; Gonzalez-Sanchez and Bader, 1985), anti-MHC (MF20; Bader et al., 1982), anti-transitin (EAP300; McCabe et al., 1995, Gourdie et al., 1995), and anti-HA. After 3 × 5-min washes in PBS, specific fluorochrome-conjugated secondary antibodies were applied. These included anti-rabbit Alexa 488, Alexa 555 (Molecular Probes, Eugene, OR), anti-rabbit Cy5, anti-mouse fluorescein isothiocyanate, anti-mouse tetrarhodamine isothiocyanate (TRITC; Chemicon), and anti-goat TRITC (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA). Nuclei counterstains were Hoechst 33258 (Sigma), DRAQ5 (Biostatus Ltd., UK), and To-Pro-3 (Molecular Probes).
Immunolabeled slides were viewed by using either a Zeiss Axioskop epifluorescence light microscope or a TCS Laser-Scanning Confocal microscope (Leica, Germany). Images were imported into NIH image and Photoshop 7.0 using an Apple Macintosh G4. For image quantitations, all settings were standardized on the confocal microscope and in NIH image.
We thank Dr. David Sedmera, Dr. Tom Trusk, and Mr. T. Gallien for their insight and assistance. R.G.G., T.X.O., and D.W.B. were funded by the NIH; T.X.O. received Merit and Reap awards from the Research and Development Service of the Department of Veteran Affairs; and B.S.H. was funded by the National Center for Research Resources and by the American Heart Association.
- 1982. Immunochemical analysis of myosin heavy chain during avian myogenesis in vivo and in vitro. J Cell Biol 95: 763–770. , , .
- 1999. Mutations in the cardiac transcription factor NKX2.5 affect diverse cardiac developmental pathways. J Clin Invest 104: 1567–1573. , , , , , , , , , , , , .
- 1999. Development of the cardiac conduction system involves recruitment within a multipotent cardiomyogenic lineage. Development 126: 5041–5049. , , , , , .
- 1995. Developmental regulation of connexin 40 gene expression in mouse heart correlates with the differentiation of the conduction system. Dev Dyn 204: 358–371. , , , , , , , .
- 1974. Technique of shell-less culture of the 72-hour avian embryo. Poult Sci 53: 409–412. .
- 2001a. The gap-junctional protein connexin40 is elevated in patients susceptible to postoperative atrial fibrillation. Circulation 103: 842–849. , , , , , , .
- 2001b. Altered connexin expression in human congestive heart failure. J Mol Cell Cardiol 33: 359–371. , , , , , , , , .
- 1999. Vertebrate tinman homologues and cardiac differentiation. Semin Cell Dev Biol 10: 73–83. .
- 2001. NKX2.5 mutations in patients with tetralogy of fallot. Circulation 104: 2565–2568. , , .
- 1985. Characterization of a myosin heavy chain in the conductive system of the adult and developing chicken heart. J Cell Biol 100: 270–275. , .
- 1993. Evidence for a distinct gap-junctional phenotype in ventricular conduction tissues of the developing and mature avian heart. Circ Res 72: 278–289. , , , , .
- 1995. Terminal diversification of the myocyte lineage generates Purkinje fibers of the cardiac conduction system. Development 121: 1423–1431. , , , .
- 1998. Endothelin-induced conversion of embryonic heart muscle cells into impulse-conducting Purkinje fibers. Proc Natl Acad Sci U S A 95: 6815–6818. , , , , .
- 2003. Development of the cardiac pacemaking and conduction system. Birth Defects Res Part C Embryo Today 69: 46–57. , , , , , , , .
- 1998. Tinman function is essential for vertebrate heart development: elimination of cardiac differentiation by dominant inhibitory mutants of the tinman-related genes, XNkx2–3 and XNkx2–5. Dev Biol 204: 187–196. , .
- 2002. Progressive AV-block and anomalous venous return among cardiac anomalies associated with two novel missense mutations in the CSX/NKX2–5 gene. Hum Mutat 20: 75–76. , , , , .
- 2002. Coronary arteriogenesis and differentiation of periarterial Purkinje fibers in the chick heart: is there a link? Tex Heart Inst J 29: 262–270. , , .
- 2003. Transcription factor cascades in congenital heart malformation. Trends Mol Med 9: 512–515. , , , .
- 1998. A simplified system for generating recombinant adenoviruses. Proc Natl Acad Sci U S A 95: 2509–2514. , , , , , .
- 2001. Tbx5 associates with Nkx2–5 and synergistically promotes cardiomyocyte differentiation. Nat Genet 28: 276–280. , , , , , , .
- 2002. Novel point mutation in the cardiac transcription factor CSX/NKX2.5 associated with congenital heart disease. Circ J 66: 561–563. , , , , , , , , , , .
- 2004. Nkx2–5 mutation causes anatomic hypoplasia of the cardiac conduction system. J Clin Invest 113: 1130–1137. , , , , , , , , , , , , .
- 2001. Progressive atrioventricular conduction defects and heart failure in mice expressing a mutant Csx/Nkx2.5 homeoprotein. J Clin Invest 108: 189–201. , , , , , , , , , , , .
- 1998. The cardiac tissue-restricted homeobox protein Csx/Nkx2.5 physically associates with the zinc finger protein GATA4 and cooperatively activates atrial natriuretic factor gene expression. Mol Cell Biol 18: 3120–3129. , , , , , , .
- 1995. Myogenic and morphogenetic defects in the heart tubes of murine embryos lacking the homeo box gene Nkx2–5. Genes Dev 9: 1654–1666. , , , , , , .
- 1995. Developmentally regulated neural protein EAP-300 is expressed by myocardium and cardiac neural crest during chick embryogenesis. Dev Dyn 203: 51–60. , , , .
- 2003. NKX2.5 mutations in patients with congenital heart disease. J Am Coll Cardiol 42: 1650–1655. , , , , .
- 2003. Development of the cardiac conduction system: a matter of chamber development. Novartis Found Symp 250: 25–34; discussion 34–43, 276–279. , .
- 2004. The T-Box transcription factor Tbx5 is required for the patterning and maturation of the murine cardiac conduction system. Development 131: 4107–4116. , , , , , , , , , .
- 2002. Differential regulation of the cardiac sodium calcium exchanger promoter in adult and neonatal cardiomyocytes by Nkx2.5 and serum response factor. J Mol Cell Cardiol 34: 807–821. , , , , , , , , .
- 2004. Nkx2–5 pathways and congenital heart disease; loss of ventricular myocyte lineage specification leads to progressive cardiomyopathy and complete heart block. Cell 117: 373–386. , , , , , , , , , , , , , , .
- 2005. Endothelin-1 and Neuregulin-1 convert embryonic cardiomyocytes into cells of the conduction system in the mouse. Dev Dyn 233: 20–28. , .
- 2002. Induction and patterning of the cardiac conduction system. Int J Dev Biol 46: 765–775. , , , , .
- 2002. Neuregulin-1 promotes formation of the murine cardiac conduction system. Proc Natl Acad Sci U S A 99: 10464–10469. , , , , , , , , .
- 1999. Expression of immediate early genes, GATA-4, and Nkx-2.5 in adrenergic-induced cardiac hypertrophy and during regression in adult mice. Br J Pharmacol 127: 1165–1176. , , .
- 1998. Congenital heart disease caused by mutations in the transcription factor NKX2–5. Science 281: 108–111. , , , , , , , , .
- 2003. Spatiotemporal pattern of commitment to slowed proliferation in the embryonic mouse heart indicates progressive differentiation of the cardiac conduction system. Anat Rec 274A: 773–777. , , , , , , .
- 1998. GATA-4 and Nkx-2.5 coactivate Nkx-2 DNA binding targets: role for regulating early cardiac gene expression. Mol Cell Biol 18: 3405–3415. , , , , , .
- 2001. Purkinje fibers of the avian heart express a myogenic transcription factor program distinct from cardiac and skeletal muscle. Dev Biol 234: 390–401. , , , .
- 2001. Elevated expression of Nkx-2.5 in developing myocardial conduction cells. Anat Rec 263: 307–313. , , , , , , .
- 1998. Upregulation of the cardiac homeobox gene Nkx2–5 (CSX) in feline right ventricular pressure overload. Am J Physiol 274: H1569–H1573. , , .
- 2002. Two novel frameshift mutations in NKX2.5 result in novel features including visceral inversus and sinus venosus type ASD. J Med Genet 39: 807–811. , , , , , .