Introduction
The calcium ion is a ubiquitous intracellular messenger that controls many processes. As steep gradients between the extracellular medium and intracellular stores and the cytosol are maintained, the regulation of calcium pathways is fundamental to control diverse cellular functions. During development, calcium participates in fertilization, cytokinesis, cell motility, axis formation and differentiation (Whitaker, Reference Whitaker2006, Slusarski & Pelegri, Reference Slusarski and Pelegri2007). Voltage-activated calcium channels (CaV) provide the principal pathway for calcium entry into electrically excitable cells, such as neurons, skeletal muscle, and cardiac muscle and endocrine cells. As cells in the early embryo are devoid of electrical excitability, information on the expression or function of CaV before differentiation of excitable cells is sparse. It is noteworthy that functional CaV have been recorded from the eggs or embryos of echinoderms and ascidians (Dale et al., Reference Dale, Talevi and DeFelice1991, Reference Dale, Yazaki and Tosti1997). Moreover, calcium entry through CaV regulates cytokinesis in the early sea urchin egg (Yazaki et al., Reference Yazaki, Tosti and Dale1995). In vertebrates, CaV are expressed in the mesoderm of Xenopus and mediate the acquisition of its dorsal fate (Palma et al., Reference Palma, Kukuljan and Mayor2001) and also appear to participate in neural induction in amphibians (Webb et al., Reference Webb, Moreau, Leclerc and Miller2005, Leclerc et al., Reference Leclerc, Néant, Webb, Miller and Moreau2006).
The zebrafish (Danio rerio) embryo combines the advantages of classical embryological approaches with the availability of genetic tools, including a fully sequenced genome. Studies in this model show long-range elevations of its intracellular concentration during early development (Webb & Miller, Reference Webb and Miller2006). Calcium mobilization from intracellular stores is the main mechanism involved in these signals. Furthermore, the InsP3 receptors are critically involved in the control of early development (Ashworth et al., Reference Ashworth, Devogelaere, Fabes, Tunwell, Koh, De Smedt and Patel2007). Calcium fluxes have been also proposed to mediate signals that define cell fates (Schneider et al., Reference Schneider, Houston, Rebagliati and Slusarski2008). At later stages, the development of the heart requires the expression of CaV (Rottbauer et al., Reference Rottbauer, Baker, Wo, Mohideen, Cantiello and Fishman2001). Considering the antecedents of CaV expression in the early embryo in other anamniotes, the role of calcium in the control of developmental processes and the advantages of the zebrafish embryo to study early vertebrate development, we explored the expression of CaV in this model.
Methods
We conducted BLAST searches of the zebrafish genome using identified zebrafish and mammalian CaV cDNA sequences. Putative transcriptional units were identified by the existence of ESTs. Wild type Danio rerio were maintained in aquaria under standard conditions and fertilizations were carried out using routine procedures. Embryos were maintained in medium E3 (5 mM NaCl, 0.17 mM KCl, 0.4 mM CaCl2, and 0.16 mM MgSO4, methylene blue 0.01%) at 28 °C.
Total RNA was isolated using Trizol (Invitrogen), and cDNA was synthesized using Superscript II (Invitrogen). We designed PCR primers for each of the identified genes. Primer sequences are shown in Table 1. PCR products were cloned into the pGEM-T easy vector system (Promega) and sequenced. In situ hybridization (ISH) probes were synthesized by in vitro transcription of cloned PCR products, incorporating digogixenin-11-UTP (Roche). ISH procedures followed standard protocols. Whole mount immunohistochemistry was conducted using the anti-pan-calcium channel α1 subunit (Alomone), anti-gemini (CaV1.3a or CACNA1Da, Sidi et al., Reference Sidi, Busch-Nentwich, Friedrich, Schoenberger and Nicolson2004) or anti-dihydropiridine binding complex (α1 subunit) monoclonal (Chemicon International) antibodies. The secondary antibody was anti-rabbit Fc coupled to horseradish peroxidase (Amersham Pharmacia Biotech). Peroxidase activity used diaminobenzidene as substrate according to standard protocols. Loss of function experiments were conducted exposing embryos to pharmacological blockers (all purchased from Alomone), dissolved from stocks in methanol or DMSO; embryos were maintained in the dark to avoid photoinactivation of the blockers. A morpholino-modified antisense oligonucleotide targeting CaV1.3.a was purchased from Gene-Tools; its sequence was GCGTCCACCTGCATGGGCCTGCCAG. The oligonucleotide or a nonsense control was injected into 2-cell embryos.
Table 1 Genes encoding CaV in the zebrafish
Genes are named according to their closest mammalian homologue; CACNA1Aa and CACNA1Ab, and CACNA1Ba and CACNA1Bb represent distinct genes, highly homologous to one mammalian counterpart each. GenBank accession numbers listed are for identified cDNAs or sequences annotated as putative transcripts. Only one number for each is listed, despite some cases of redundancy. Primers sequences of the oligonucleotides used for RT-PCR are shown.
Results and discussion
In identification of CaV expressed in the early zebrafish embryo, α-subunits are the main components of functional CaV; genes encoding these subunits are termed CACNA1. Our compilation of the known cDNA sequences that encode α subunits of CaV in the zebrafish and the results of the BLAST searches identified 12 genes belonging to the CACNA family in this species. Based on the closest matches to CaV genes in mammalians, we assigned the designations annotated in Table 1. Consistent with evidence of a wide-scale gene duplication in the teleost lineage, which has yielded duplicate versions of genes when compared with tetrapod genomes, such as the genes encoding CaV1.3a and b (Sidi et al., Reference Sidi, Busch-Nentwich, Friedrich, Schoenberger and Nicolson2004) and sodium channels (Novak et al., Reference Novak, Jost, Lu, Taylor, Zakon and Ribera2006), we found two genes that yield predicted proteins with a high similarity to CACNA1A and two genes that would originate CACNA1B-like proteins.
RT-PCR analysis of the expression of the 12 genes, using RNA isolated from 5 dpf (days post fertilization), fish yielded amplicons of the expected size in all cases (data not shown). RT-PCR analysis using RNA isolated from gastrulating embryos (75% of epiboly) as the templates, resulted in consistent amplification of fragments that corresponded to CACNA1Da, CACNA1S, CACNA1G, CACNA1Db, CACN1AI and CACNA1A (data not shown). Sequencing of these products confirmed their identity as predicted from the genome databases. We conducted ISH for all transcripts expressed at gastrulation to determine their spatial and temporal patterns during early development. We observed signals with only two probes, CACNA1Da and CACNA1S (Figure 1), in spite of extensive testing of different hybridization conditions for the other probes. In mammals, CACANA1Da encodes a CaV expressed in neuroendocrine cells and neurons, whereas CACNA1S encodes the skeletal muscle calcium channel. It is noteworthy that early Xenopus embryos also express transcripts homologous to these (Palma et al, Reference Palma, Kukuljan and Mayor2001). In the adult zebrafish, CACNA1A is expressed in the ear (Sidi et al., Reference Sidi, Busch-Nentwich, Friedrich, Schoenberger and Nicolson2004). The CACNA1Da and CACNA1S transcripts were detected as early as the 2-cell stage and their expression continues during blastula and gastrula stages; this early expression indicates the presence of maternally transcribed mRNAs. The distribution of the transcripts is widespread during early development, whereas, at later stages, labelling is restricted to the developing CNS and the somites respectively (Figure 1).
Figure 1 (a, b): ISH showing the expression of the CACNA1Da (CaV1.3a) transcript in the early zebrafish embryo. Expression is during early blastula (8-cell stage, a) and gastrulation (c, d). (e) shows the hybridization with the sense probe at midgastrulation. At later stages (24–48 hpf) labelling is restricted to the central nervous system (f, g). (h) and (i) show embryos at the same stages hybridized with the sense probe. j–l show the expression of the CACNA1S (CaV1.1) transcript at the 2-cell stage (j) and during gastrulation (k, l). At 48 hpf the expression is restricted to the somites (m). (n) shows embryos hybridized to the sense probe. hpf, hours post fertilization.
We used a ‘pan-CaV’ antibody, which recognizes an epitope conserved in all CaVs (see Methods), to test if CaV proteins are expressed at early stages of development. We observed distinct labelling at the 2-cell stage, which persists throughout gastrulation (Figure 2a). Embryos stained with an antibody targeting the rabbit skeletal muscle CaV (α1S subunit, CaV1.1 or the product of the CACNA1S gene), displayed a similar pattern (data not shown). We then used an antibody raised against the zebrafish CaV1.3a protein (Sidi et al., Reference Sidi, Busch-Nentwich, Friedrich, Schoenberger and Nicolson2004). As shown in Figure 2b, distinct labelling is observed very early in development, including 4-cell embryos, and is maintained during gastrulation.
Figure 2 Expression of CaV proteins: (a–f) whole mount immunohistochemical analysis using the anti-CaV1.3a (CACNA1Da) antibody. Expression of immunoreactivity is shown from the early blastula stage (a) through gastrulation and organogenesis. (g) and (h) show control embryos that were incubated only with the secondary antibody and subjected to the peroxidase reaction. (i–m) immunohistochemical analysis using the anti-‘pan-CaV’ antibody, from the 4-cell stage (i) to 36 hpf. (n) shows embryos subjected to the immunohistochemical procedure with the primary antibody or the primary antibody preincubated with the immunogenic peptide. hpf, hours post fertilization.
These data demonstrate that transcripts that encode CaV are translated; the use of three different antibodies, targeting different epitopes supports the specificity of this finding. This study did not allow us to distinguish in more detail which transcripts are actually translated, beyond the distinct characterization of CaV1.3a expression and the possible expression of CaV1.1.
We used pharmacological tools to study the function of the expressed CaV. Mammalian products of the CACNA1D and CACNA1S genes are sensitive to dihydropyridines; therefore we exposed embryos to the blockers nifedipine (20 μM) or nimodipine (10 μM), or the agonist (–)BK-8644 (5 μM) or to methanol (0.2%). Embryos were exposed to the blockers at 6 hpf and were maintained in that condition for 6 or 12 h. Three series of 50 embryos for each condition were studied. Examination of living embryos at 12, 24 or 48 h did not reveal any significant distinguishing features in treated embryos versus controls (Figure 3a). This result contrasts with the effect of InsP3 receptor blockers on early development of the zebrafish (Ashworth et al., Reference Ashworth, Devogelaere, Fabes, Tunwell, Koh, De Smedt and Patel2007). In order to detect possible abnormalities in gastrulation or early patterning, we subjected embryos to blockers to ISH using a probe for no tail, which is expressed in the developing mesoderm (Schulte-Merker et al., Reference Schulte-Merker, Hammerschmidt, Beuchle, Cho, De Robertis and Nüsslein-Volhard1994), and thus may be used to evaluate the progression of gastrulation. Exposure to blockers was not associate with perturbations in its expression pattern in two series of 30 embryos for each condition (Figure 3b).
Figure 3 Exposure to dihydropyridine blockers or to a dihydropyridine agonist of CaV does not alter early development of the zebrafish embryo, as evidenced by the appearance of living embryos at midgastrula (left panels) or the expression pattern of the no tail transcript (right panels), used here as a marker of mesoderm. Panels show lateral and dorsal views of embryos subjected to ISH at late gastrula stages.
Early exposure to blockers were associated with defects evident at late stages, particularly marked edema and abnormalities in heart function. In three series of 60 embryos, around 50% of the embryos showed defects evident at 5 dpf, such as those depicted in Figure 4. This result is compatible with the role of CaV in cardiac development (Rottbauer et al., Reference Rottbauer, Baker, Wo, Mohideen, Cantiello and Fishman2001) (Figure 4). We did not pursue analysis of this effect further, as the focus of our work was the description of CaV expression before the excitable cell lineages arise.
Figure 4 Early exposure of zebrafish embryos to dihydropyridine blockers, but not to a related agonist of CaV results in late developmental defects. Embryos representative of the defects observed in association to the exposure to blockers, in a dorsal and lateral view are shown in the right panels.
We attempted to decrease the expression of one of the identified CaV, Cav1.3a, using a morpholino antisense oligonucleotide, injected into the two blastomeres at the 2-cell stage. In three series of more than 100 embryos each, we did not observe significant differences in the injected embryos with respect to controls at 24 or 48 hpf. Defects attributable to the injection procedure were observed in about 20% of embryos injected with the specific CaV1.3a and the nonsense control oligonucleotide (data not shown). We speculate that the lack of any marked effect of blockers on development may be because CaV proteins play roles different to that of calcium conduction. Recent evidence shows that CaV β4 subunits are expressed during early development in the zebrafish, and play a role as scaffold proteins (Ebert et al., Reference Ebert, McAnelly, Srinivasan, Linker, Horne and Garrity2008). Alternatively, CaV function may be to solely contribute to replenish stores responsible for release of calcium mediated by intracellular messengers (Kukuljan et al., Reference Kukuljan, Rojas, Catt and Stojilkovic1994); thus the loss of this function may have mild consequences. Furthermore, and as we were not able to ascertain the translation of transcripts encoding putative dihydropyridine-insensitive CaV (i.e. CACNA1G), redundant functions shared by various subtypes of channels may have not been affected by the blockers we used. These results demonstrate that the expression of several CaV mRNAs and the translation of the messengers is a common phenomenon in vertebrate embryos. Furthermore, these data complement functional data obtained from studies in the eggs and embryos of invertebrates (Dale et al., Reference Dale, Talevi and DeFelice1991, Reference Dale, Yazaki and Tosti1997), suggesting that these channels may serve conserved roles in development throughout the animal kingdom. The finding of multiple transcripts is compatible with redundant functions, which renders an immediate genetic or pharmacological approach to their study more difficult. The results presented here set the basis for further dissection of the physiological role of CaV during development.
Acknowledgements
Supported by FONDECYT (Chile) 1040829. We are indebt to Dr Miguel Concha's lab for providing embryos and technical advice.
