In this whole-genome survey, 6 distinct PMCAs (Figure 1) and 7 distinct NCXs (Figure 2) were found in zebrafish. According to the phylogenetic analysis, these transporters were annotated as zPMCA1a, zPMCA1b, zPMCA2, zPMCA3a, zPMCA3b, zPMCA4 (Figure 3a, and in standard nomenclature ATP2B1a, ATP2B1b, ATP2B2, ATP2B3a, ATP2B3b, and ATP2B4, respectively) and zNCX1a, zNCX1b, zNCX2a, zNCX2b, zNCX3, zNCX4a, and zNCX4b (Figure 3b, in standard nomenclature SLC8a1a, SLC8a1b, SLC8a2a, SLC8a2b, SLC8a3, SLC8a4a, and SLC8a4b, respectively). zNCX1a and zNCX1b were previously described and named zNCX1h and zNCX1n, respectively 25, and zNCX2a, zNCX3, zNCX4a, and zNCX4b have also been published 1718. However, zNCX2b and the other zPMCAs are first described in this study.

Based on the topological structures of PMCA and NCX summarized in mammals 1226, zebrafish PMCAs and NCXs share similar patterns as mammals at the deduced amino acid sequences. According to the hydropathy analysis, 10 putative transmembrane domains were predicted for zPMCA (Figures 1, 4a), whereas NCX has 11 hydrophobic segments and 9 putative transmembrane domains (Figures 2, 4b) due to an intracellular helix front transmembrane domain 6 and a P loop-like structure of hydrophobic segment 8, which contained the amino acid sequence "GIG" 26. PMCA and sarco-endoplasmic reticulum calcium ATPase (SERCA), the closest kin to PMCA, share a similar topological structure: hydrophobic segments span the membrane 10 times, and 3 intracellular loops consist of Ca²⁺ binding and transporting, and calmodulin-regulating functions 1227; the zPMCAs were also found to have 3 cytoplasmic domains 1 each between TM2 and TM3, TM4 and TM5, and TM10 and the C-terminal (Figure 1). As for NCX, a large intracellular loop for ion translocation 2829 was located between TM5 and TM6 (Figure 2). There were 20 alternative splicing isoforms of zPMCAs found when cloning was conducted, while 9 alternative splicing isoforms were found in zNCXs (data not shown). The hot spots of alternative splicing regions in the zPMCAs are at the 5' untranslated regions (UTRs), the fragments around the 300^th amino acid including from 1 to 3 small exons (less than 100 bp each in those cases), and 3' amino acid tails that usually stretch to the 3' UTR and alter the location of stop codons. On the other hand, a region of alternative splicing sites was found in zNCX1b and zNCX3 around the 610^th amino acid, which is a part of the intracellular loop.

zPMCAs share 63%~86% identities with each other, while zNCXs exhibit higher variety among isoforms (i.e., 36%~76% identity). According to the phylogenetic analysis, zebrafish and 2 pufferfish species so far all have 6 PMCA and 7 NCX isoforms (Figure 3a,b). Compared with the isoforms annotated in tetrapods, duplication from the ancestor ray-finned fish occurred in PMCA1, PMCA3, NCX1, NCX2, and NCX4, according to the latest version of the genomic database utilized. Moreover, in the karyotype, zPMCAs and zNCXs are located on different chromosomes (see additional file 1), suggesting that these paralogous groups may have originated from genome duplication but not from tandem duplication. Relationships among these pairs of duplicates were evident not only in the phylogenetic tree, but also in their gene structures (Figure 4). All teleost PMCAs and NCXs examined in this study showed an outgroup topology to their orthologues of mammals (Figure 3a,b), and the paralogous teleosts formed an inner-group topology. The 2 prototypes of NCXs, NCX4a and NCX4b, found in teleosts were most closely related to NCX1 and NCX3 in both phylogenetic trees (Figure 3) and gene structures (Figure 4b). In gene structures, all of them had a relatively large first exon, even though some small introns were found in the split exons in zNCX4s (Figure 4b).

RT-PCR analysis revealed that zpmca2 and zpmca4 were ubiquitously expressed in various tissues compared with zpmca1a, zpmca1b, zpmca3a, and zpmca3b (Figure 5). All zpmcas were abundant in the brain and eye except zpmca1b, which was mainly expressed in the eye. These findings agree with the Ca²⁺ homeostasis function of PMCAs found in the mammalian neuron system 12. zncx1b and zncx4a were also ubiquitously expressed, while zncx2b and zncx3 were only detected in the brain and eyes. zpmca1a, zpmca1b, zpmca2, zpmca3a, and all the zncxs were expressed in ovary and 1-cell embryo (Figure 5), suggesting that these genes may regulate intracellular Ca²⁺ homeostasis at earlier stages. zpmca1a, zpmca2, zpmca4, zncx1, and zncx4a were expressed in gills (Figure 5).

According to the quantitative real time-PCR analysis, low-Ca²⁺ FW-acclimated zebrafish expressed significantly higher levels of zecac in gills than did high-Ca²⁺ FW-acclimated fish (Figure 6). However, no significant difference was found between the 2 groups in the expressions of any of the zpmcas and zncxs (Figure 6). Consistent with the results of the tissue scans shown in Figure 5, zpmca4, zpmca1a, and zpmca2 exhibited more than 20-fold higher expression levels than did the other zpmca isoforms in gills, whereas zncx1b was found to have an approximately 5-times stronger signal compared with the other zncxs (Figure 6).

Fluorescence in situ hybridization and immunocytochemistry were used to determine the isoforms of zpmca and zncx that were specifically expressed in gill ionocytes. Specific mRNA probes of all of the zpmca and zncx isoforms were used to conduct the in situ hybridizations for whole-mount embryos and adult gills. Among the isoforms, only zpmca2 and zncx1b mRNA signals were found in specific groups of cells in embryonic skin and adult gills (data not shown).

Subsequently, double-fluorescence in situ hybridizations were used for zecac and zpmca2 (Figure 7A–C), and zecac and zncx1b (Figure 7D–F), respectively. More gill cells expressed zpmca2 and zncx1b mRNAs than zecac mRNA (Figure 7A–F). Notably, zpmca2 and zncx1b were also abundantly expressed in gill lamellae, where fewer MR cells are usually found 24. zecac and zpmca2 mRNAs were colocalized only in a portion of gill cells, but were not colocalized all of the time. Around half of the zecac-positive cells co-expressed zpmca2 signals (51.7 +/- 3.50%, n = 3), while only 40.3 +/- 1.35% (n = 3) of zpmca2-positive cells showed zecac signals (Figures 7A–C, 8). Similar results were found in the case of double labeling of zecac and zncx1b (Figures 7D–F, 8). The proportion of zncx1b-positive cells that co-expressed zecac was 77.1 +/- 1.35% (n = 3), while zncx1b signals could be detected in all zecac-positive cells (Figures 7D–F, 8).

The subsequent triple-labeling experiments further demonstrated the colocalization of zecac mRNA/zpmca2 mRNA/Na⁺-K⁺-ATPase (Figure 8A–E), and zecac mRNA/zncx1b mRNA/Na⁺-K⁺-ATPase (Figure 8F–J), respectively, in specific groups of gill ionocytes. Taken all together, only a portion of gill ionocytes co-expressed the mRNA of zecac, zpmca2, and zncx1b, and Na⁺-K⁺-ATPase.

Zebrafish (Danio rerio) brood stocks at the Institute of Cellular and Organismic Biology, Academia Sinica were kept in fresh water (local tap water, FW) at 28.5°C under a 14-h:10-h light: dark photoperiod. Embryos were collected within 30 min after fertilization and incubated in a Petri dish until the desired developmental stages. For whole-mount in situ hybridization, PTU (1-phenyl 2-thiourea) at a final concentration of 0.003% was added to prevent melanogenesis. Fish were anesthetized with buffered MS222 before sampling following the guidelines of the Academia Sinica Institutional Animal Care and Utilization Committee (approval no.: RFiZOOHP2006086).

Following a previous process in our laboratory 49, high-Ca²⁺ (2.00 mM, control) and low-Ca²⁺ (0.02 mM) artificial fresh waters were prepared with double-deionized water (Milli-RO60, Millipore, Billerica, MA, USA) supplemented with adequate CaSO₄·2H₂O, MgSO₄·7H₂O, NaCl, K₂HPO₄, and KH₂PO₄. Nominal Ca²⁺ concentrations of the high- and low-Ca²⁺ media were 2.00 and 0.02 mM, respectively, but the other ion concentrations of the media were the same ([Na⁺], 0.5 mM; [Mg²⁺], 0.16 mM; and [K⁺], 0.3 mM) as those in the local tap water. Variations in the ion concentrations were maintained within 10% of the predicted values by examination with an atomic absorption spectrophotometer (Hitachi Z-8000, Tokyo, Japan). Zebrafish were transferred to high- and low-Ca²⁺ media, respectively, for 2 wk. At the end of acclimation, fish gills were sampled for quantitative real-time PCR analysis.

The peptide sequences from other species (teleost had a higher priority) were used to BLAST the genome databases (of NCBI and Ensembl) for zebrafish, pufferfish (Fugu rubripes), and tetradon (Tetraodon nigroviridis). The putative full-length or partial open reading frames of PMCAs and NCXs obtained were confirmed using the EST database, and/or were used to design primers for cloning and the RT-PCR analysis. PCR products thus obtained were subcloned into a pGEM-T Easy vector (Promega, Madison, WI, USA), and the nucleotide sequences were determined with an ABI 377 sequencer (Applied Biosystems, Warrington, UK). Sequence analysis was conducted with a BLASTx program (NCBI). The specific primers of 5' and 3'-rapid amplification of cDNA ends (RACE) were designed from the partial sequences obtained from the PCR with degenerate primers. The RACE PCR program followed a commercial protocol (Clontech, Mountain View, CA, USA), and the RACE PCR products were also subcloned into pGEM-T Easy vectors and sequenced. The full-length deduced amino-acid sequences were aligned with ClustalX and then phylogenetically analyzed with Mega3.1 51. For the proteomic structure analysis, the hydropathy plot and transmembranes were predicted with EMBOSS (the European Molecular Biology Open Software Suite, version 2.8.0 52 algorithm according to Persson and Argos 53.

An appropriate amount of zebrafish adult tissues was collected for the total RNA preparation with TRIZol reagent (Invitrogen, Carlsbad, CA, USA). The amount and quality of total RNA were determined by measuring the absorbances at 260 and 280 nm with a spectrophotometer (Hitachi U-2000) and RNA denatured gels. For cDNA synthesis, 5 μg of total RNA was reverse-transcribed in a final volume of 20 μL containing 0.5 mM dNTPs, 2.5 μM oligo (dT)₁₈, 5 mM dithiothreitol, and 200 units superscript reverse transcriptase III (Invitrogen) for 1.5 h at 42°C, followed by a 15-min incubation at 70°C. For the PCR amplification, 1 μL cDNA was used as a template in a 25-μL final reaction volume containing 0.25 mM dNTP, 1.25 units Gen-Taq polymerase (Genemark, Taipei, Taiwan), and 0.2 μM of each primer. The primer sets for the PCR of tissue scans were zPMCA1a (428-bp fragment) forward 5'-AGACAGGGTGGATAGAAGGT-3', reverse 5'-CCCCAATAATGTGAAGATGA-3'; zPMCA1b (556-bp fragment), forward 5'-AGCCCCTTATTTCCCGCAC-3', reverse 5'-GCCCCCTTCTCAGCTCCATT-3'; zPMCA2 (444-bp fragment) forward 5'-GGGATCGGGATGAGATGGTA-3', reverse 5'-GCGTGTGTTTGTCTGTGGGT-3'; zPMCA3a (457-bp fragment) forward 5'-TTGCTGGTACAGATGTGGC-3', reverse 5'-GGAGGAGAGTGAAGCGGAG-3'; zPMCA3b (614-bp fragment) forward 5'-GGTCTGCTGGGTTTCCTACT-3', reverse 5'-TCCTGCTCAATCTCTCCTTT-3'; zPMCA4 (388-bp fragment) forward 5'-GTCAAGGCTGTGATGTGGGC-3', reverse 5'-TGGTTGGGAGTGGAGAAGGG-3'; zNCX1b (358-bp fragment) forward 5'-AGAGACGAGGAGAAGGAGGT-3', reverse 5'-GCACGAAAGCAAAGAGAACT-3'; zNCX2a (276-bp fragment) forward 5'-TGATGAAGAAGGAGGTGAAC-3', reverse 5'-CTTGCGAATGTGTCTGGTAT-3'; zNCX2b (477-bp fragment) forward 5'-GCAGCCGCATTTCCTTCG-3', reverse 5'-CGCCAGAGTTTCGGACCAC-3'; zNCX3 (244-bp fragment) forward 5'-GAGGAAGCAAGAAGAATAGC-3', reverse 5'-AGTCAAAACAAGATGGCAGA-3'; zNCX4a (462-bp fragment) forward 5'-GGAAGAGAGCGGAGAAGAGC-3', reverse 5'- ATGGTGAAGAGGGTGACGGA-3'; and zNCX4b (350-bp fragment) forward 5'-ATGCAGCAGCAGAAGAGC-3', reverse 5'-ACATTCTCGCCAGGTTTG-3'. A 514-bp fragment of zebrafish beta-actin was used as an internal control to evaluate the relative amounts of complementary DNAs (cDNAs), and the design of the primer pairs followed Hsiao et al. 54. The amplicons were all sequenced to make sure that the PCR products were the desired gene fragments.

Fragments of the target genes obtained by PCR were inserted into pGEM-T Eeasy vectors. Digoxigenin- (Dig) (Roche, Penzberg, Germany) or dinitrophenol (DNP)-labeled (Perkin-Elmer, Boston, MA, USA) RNA probes were synthesized by in vitro transcription with T7 and SP6 RNA polymerase (Takara, Shiga, Japan). The qualities of the probes were examined using RNA gels, and the concentrations were determined by a dot-blot assay with standard DIG-labeled RNA (100 ng/μl) (Roche). Excised gills were fixed with 4% paraformaldehyde overnight at 4°C and then washed several times with PBS. For the cryosections, fixed samples were immersed in PBS containing 30% sucrose overnight, and embedded in Optimal Cutting Temperature (OCT) compound embedding medium (Sakura, Tokyo, Japan) at -20°C, and 10-μm frozen cross-sections were cut with a CM 1900 rapid sectioning cryostat (Leica, Heidelberg, Germany) and attached to poly-L-lysine coated slides (Electron Microscopy Sciences, Ft. Washington, PA, USA). Prepared samples from either cryosections or methanol-dehydrated whole gills were washed several times with phosphate-buffered saline with 0.1% tween-20 (PBST). After a brief washing with PBST, samples were incubated with hybridization buffer (HyB) containing 50% formamide, 5× SSC, and 0.1% Tween-20 for 5 min at 65°C. Prehybridization was performed for 2 h at 65°C with HyB+, which is the hybridization buffer supplemented with 500 ng/mL yeast tRNA and 50 μg/mL heparin. For hybridization, samples were incubated in 100 ng of the RNA probe in 200 μL HyB+ at 65°C overnight. Then, slides were washed at 65°C for 10 min in 75% HyB and 25% 2× SSC, 10 min in 50% HyB and 50% 2× SSC, 10 min in 25% HyB and 75% 2× SSC, 10 min in 2× SSC, and 30 min for 2 times in 0.2× SSC at 70°C. Further washes were performed at room temperature for 5 min in 75% 0.2× SSC and 25% PBST, 5 min in 50% 0.2× SSC and 50% PBST, 5 min in 25% 0.2× SSC and 75% PBST, and 5 min in PBST. Fluorescence staining was conducted with a commercial kit, TSA Plus Fluorescence Systems (Perkin-Elmer). The hybridization signals detected by the DIG-labeled RNA probes were amplified through fluorescein-tyramide signal amplification (TSA), while cyanine 3-TSA was used for the DNP-labeled probes. For triple labeling, samples were first subjected to in situ hybridizations of the 2 transporter genes and then to Na⁺/K⁺-ATPase immunocytochemistry (see below).

Zebrafish embryos and gills were fixed in 4% paraformaldehyde for 10 min at 4°C. After washing in PBS, fixed embryos were treated with 100% ethanol for 10 min at -20°C and subsequently subjected to blocking with 3% BSA at room temperature for 30 min. Embryos were then incubated with 1:200 PBS-diluted mouse anti-chicken Na⁺/K⁺-ATPase α subunit (cytosolic epitope for all isoforms) monoclonal immunoglobulin G (IgG) (Developmental Studies Hybridoma Bank, University of Iowa, Iowa City, IA, USA) at room temperature for 2 h. Samples were washed twice in PBS for 10 min each, and then incubated with 1:200 PBS-diluted goat anti-mouse IgG conjugated with Cy5 (Jackson Immunoresearch Laboratories, West Grove, PA, USA) for 2 h at room temperature. Images were acquired with a confocal laser scanning microscope (TCS-SP5, Leica Lasertechnik, Heidelberg, Germany).

Quantitative real-time PCR (qPCR) was carried out using a SYBR green dye (Applied Biosystems)-based assay with an ABI Prism 7000 Sequence Detection System (Applied Biosystems) according to the manufacturer's instructions. Primers targeting Ca²⁺ transporters (see additional file 1) and the endogenous control gene, beta-actin, were designed using the Primer Express 2.0 software (Applied Biosystems). In each assay, 25 ng cDNA was amplified in a 20-μL reaction containing 2x SYBR green master mix, 100 nM of forward and reverse primers, and nuclease-free water. Beta-actin, with a 151-bp amplicon (forward 5'-CACCTTCCAGCAGATGTGGA-3' and reverse 5'-AAAAGCCATGCCAATGTTGTC-3'), was used as an internal control to construct the standard curves.

Values are presented as the mean ± SE and were compared by the 2-sample t-test.