Identification of Zebrafish Calcium Toolkit Genes and Their Expression in the Brain

Zebrafish are well-suited for in vivo calcium imaging because of the transparency of their larvae and the ability to express calcium probes in various cell subtypes. This model organism has been used extensively to study brain development, neuronal function, and network activity. However, only a few studies have investigated calcium homeostasis and signaling in zebrafish neurons, and little is known about the proteins that are involved in these processes. Using bioinformatics analysis and available databases, the present study identified 491 genes of the zebrafish Calcium Toolkit (CaTK). Using RNA-sequencing, we then evaluated the expression of these genes in the adult zebrafish brain and found 380 hits that belonged to the CaTK. Based on quantitative real-time polymerase chain reaction arrays, we estimated the relative mRNA levels in the brain of CaTK genes at two developmental stages. In both 5 dpf larvae and adult zebrafish, the highest relative expression was observed for tmbim4, which encodes a Golgi membrane protein. The present data on CaTK genes will contribute to future applications of zebrafish as a model for in vivo and in vitro studies of Ca2+ signaling.


Introduction
Calcium ions (Ca 2+ ) are essential for the regulation of intracellular processes and communication between cells. As a ubiquitous secondary messenger, they are involved in multiple signaling cascades [1]. To control and respond to such Ca 2+ signals, cells require sophisticated machinery, namely the Calcium Toolkit (CaTK) that consists of Ca 2+ -binding and Ca 2+ -transporting proteins, including channels, pumps, and exchangers in plasma and organellar membranes [1,2]. Ca 2+ concentrations inside cells are kept at a very low level of approx. 100 nM [3]. Ca 2+ signals occur through an influx from the external environment or from intracellular storages, such as the endoplasmic reticulum (ER). The generation of such signals depletes Ca 2+ storages. To maintain cellular Ca 2+ homeostasis, these storages need to be refilled. The process that allows the maintenance of high Ca 2+ concentrations inside the ER while Ca 2+ flux and leakage occur is called store-operated calcium entry (SOCE) [4]. The proteins that are involved in this process include stromal interaction molecules (STIMs; i.e., sensors of Ca 2+ levels inside the ER) and ORAI or transient receptor potential (TRP) channels [5][6][7][8][9] that enable Ca 2+ influx into the cell. When ER Ca 2+ levels decrease, STIMs undergo conformational changes, oligomerize, and bind to ORAI/TRP to activate them. These complexes can be detected as puncta-like structures

Real-Time Polymerase Chain Reaction Arrays of CaTK
Adult, 1-year-old zebrafish were anesthetized with MS-222, and the brains were dissected. The material from six fish was mixed for one RNA sample and homogenized in Qiazol (Qiagen, catalog no. 79306). The RNA from zebrafish larvae was prepared using the same protocol with the exception that 50 heads were dissected from the larvae with needles at 5 days postfertilization (dpf) and pooled as one sample. The RNA quality was verified by measuring the absorbance at 260, 280, and 230 nm. Only samples with A260/280 nm and A230/280 nm > 1.8 were used for the analysis. The RNA templates (1000 ng) were used to synthesize first-strand cDNA using iScript reverse transcription supermix (Bio-Rad, catalog no. 1708840). Real-time polymerase chain reaction (RT-PCR) was performed in duplicate using SsoAdvanced Universal SYBR Green Supermix (Bio-Rad, catalog no. 1725274), and 96-well custom plates that contained primers for target and reference genes (Bio-Rad). cDNA (25 ng) was used for each reaction. Actin β 1 (actb1), glyceraldehyde-3-phosphate dehydrogenase (gapdh), tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein zeta polypeptide (yhwaz), and ribosomal protein S18 (rps18) were used as reference genes. The data were analyzed using Bio-Rad CFX Maestro 1.0 software. The stability of the reference genes was checked using the Reference Gene Selection Tool, and all of them were evaluated as at least acceptable. The expression levels were calculated using the ∆Cq method with an assumption of equal efficiencies for all reactions, and the average normalized expression was calculated for each gene according to the following formula: Normalized expression GOI = RQ GOI (RQ actb1 × RQ gapdh × RQ yhwaz × RQ rps18 ) 1 4 where GOI is the gene of interest and RQ is the relative quantity, calculated as RQ = 2 (−Cq GOI ) . The expression levels in two developmental stages were compared using an analysis of variance (ANOVA) and Tukey's Honestly Significant Difference (HSD) post hoc test. Changes in the expression are presented as a fold change ± SD using an expression in larvae as a reference value. The charts were prepared using GraphPad Prism 5.

Real-Time PCR of Components of SOCE Expression Analysis
Adult, 1-year-old zebrafish were anesthetized with MS-222 and dissected. Various organs (i.e., brain, eyes, skeletal muscles, gonads, liver, and intestines) were removed and washed with phosphate-buffered saline (PBS). The material from two fish was pooled together, and the RNA was isolated using Qiazol (Qiagen, catalog no. 79306) according to a published protocol [43]. Samples from the larval heads and trunks were obtained by separating these parts with needles and pooling tissue from fifty 5 dpf zebrafish together. The RNA templates (1000 ng for larvae and 500 ng for organs from adult fish) were used to synthesize first-strand cDNA using the SuperScript IV First-Strand Synthesis System (Invitrogen, catalog no. 18091050). RT-PCR was performed in duplicate using FastStart Essential DNA Green Master (Roche, catalog no. 06402712001) and 50 ng cDNA for larvae and 25 ng cDNA for adults. Eukaryotic translation elongation factor 1 α 1, like 1 (eef1a1l1) was used as a reference gene. The primer sequences are listed in Table S1 in the Supplementary Materials. The data were analyzed using LightCycler 96 SW 1.1 (Roche) and Microsoft Excel. The efficiency of the PCR reaction was estimated based on an analysis of the dilution curves and slope calculations (Table S1). The expression levels were calculated using the 2 (−∆Cq) method. The charts were prepared using GraphPad Prism5.

Members of the Zebrafish Brain CaTK
We identified 491 genes in zebrafish that were annotated in the Gene database of NCBI, which were members of the CaTK (Table S2 in Supplementary Materials). Approximately 4% of these genes were specific to zebrafish only. Next, we determined whether transcripts of these genes could be detected in the zebrafish brain. Using RNA-sequencing (RNA-seq) analysis, we detected transcripts for 22,046 genes in the samples that were obtained from adult zebrafish brains (Table S4). We found 444 CaTK hits of 491 predicted members among them ( Figure 1, Table S3). For 77 genes which the expression patterns of are not yet established in zebrafish, we assessed the mRNA levels in larval heads and adult brains semiquantitatively using RT-PCR arrays ( Figure 5).
Genes 2019, 10, x FOR PEER REVIEW 4 of 15 isolated using Qiazol (Qiagen, catalog no. 79306) according to a published protocol [43]. Samples from the larval heads and trunks were obtained by separating these parts with needles and pooling tissue from fifty 5 dpf zebrafish together. The RNA templates (1000 ng for larvae and 500 ng for organs from adult fish) were used to synthesize first-strand cDNA using the SuperScript IV First-Strand Synthesis System (Invitrogen, catalog no. 18091050). RT-PCR was performed in duplicate using FastStart Essential DNA Green Master (Roche, catalog no. 06402712001) and 50 ng cDNA for larvae and 25 ng cDNA for adults. Eukaryotic translation elongation factor 1 α 1, like 1 (eef1a1l1) was used as a reference gene. The primer sequences are listed in Table S1 in the Supplementary Materials. The data were analyzed using LightCycler 96 SW 1.1 (Roche) and Microsoft Excel. The efficiency of the PCR reaction was estimated based on an analysis of the dilution curves and slope calculations ( Table S1). The expression levels were calculated using the 2 (−ΔCq) method. The charts were prepared using GraphPad Prism5.

Members of the Zebrafish Brain CaTK
We identified 491 genes in zebrafish that were annotated in the Gene database of NCBI, which were members of the CaTK (Table S2 in Supplementary Materials). Approximately 4% of these genes were specific to zebrafish only. Next, we determined whether transcripts of these genes could be detected in the zebrafish brain. Using RNA-sequencing (RNA-seq) analysis, we detected transcripts for 22,046 genes in the samples that were obtained from adult zebrafish brains (Table S4). We found 444 CaTK hits of 491 predicted members among them ( Figure 1, Table S3). For 77 genes which the expression patterns of are not yet established in zebrafish, we assessed the mRNA levels in larval heads and adult brains semiquantitatively using RT-PCR arrays ( Figure 5). The expression of genes that were identified as members of the CaTK was examined by RNA-seq in the brain in adult zebrafish. Next, the mRNA levels of the selected genes were determined in zebrafish brain samples using RT-PCR arrays.
We next performed a GO annotation analysis of 444 brain CaTK genes using the PANTHER classification system and zebrafish genome as a reference gene list. A majority of the identified transcripts that are shown in Figure 2 encoded approx. 45% cytosolic proteins, approx. 25% membrane proteins, and nearly 20% organelle proteins. This classification roughly overlapped with the plausible annotated function: binding, transporter activity, and catalytic activity, respectively. The CaTK genes were classified into several signal transduction and metabolic pathways. Interestingly, approximately 10% of the genes were related to neurodegenerative diseases and belong The expression of genes that were identified as members of the CaTK was examined by RNA-seq in the brain in adult zebrafish. Next, the mRNA levels of the selected genes were determined in zebrafish brain samples using RT-PCR arrays.
We next performed a GO annotation analysis of 444 brain CaTK genes using the PANTHER classification system and zebrafish genome as a reference gene list. A majority of the identified transcripts that are shown in Figure 2 encoded approx. 45% cytosolic proteins, approx. 25% membrane proteins, and nearly 20% organelle proteins. This classification roughly overlapped with the plausible annotated function: binding, transporter activity, and catalytic activity, respectively. The CaTK genes were classified into several signal transduction and metabolic pathways. Interestingly, approximately 10% of the genes were related to neurodegenerative diseases and belong to pathways that are associated with the pathogenesis of Alzheimer's disease and Huntington's disease. Among the most enriched pathways, we also found signaling via important neurotransmitters (serotonin, glutamate, acetylcholine, and histamine) and hormones that are released from hypothalamic neurons (oxytocin, thyrotropin-releasing hormone, and gonadotropin-releasing hormone). disease. Among the most enriched pathways, we also found signaling via important neurotransmitters (serotonin, glutamate, acetylcholine, and histamine) and hormones that are released from hypothalamic neurons (oxytocin, thyrotropin-releasing hormone, and gonadotropinreleasing hormone). As observed in the GO annotation analysis, the highest number of genes encodes proteins with binding abilities. We found the mRNA of genes that encode well-described Ca 2+ -binding proteins, such as calbindin, calretinin, parvalbumin, calmodulin, calumenin, and S100 Ca 2+ -binding proteins. Most of these proteins are localized in the cytosol, but some are present in the PM (e.g., neuronal Ca 2+ sensors), ER (e.g., calreticulin and calnexin), and extracellular matrix (e.g., cell growth regulator with EF-hand domain and fibulin).
We also found the mRNA of genes that encode proteins that are responsible for Ca 2+ release and transport across the ER membrane, such as inositol 1,4,5-trisphosphate (IP3) receptors (itpr1a, itpr1b, and itpr2), ryanodine receptors (ryr1a, ryr1b, ryr2a, and ryr3), and Serca. The highest mRNA level was observed for ryr2a, but ryr1b and itpr1b also reached high expression levels. We detected several As observed in the GO annotation analysis, the highest number of genes encodes proteins with binding abilities. We found the mRNA of genes that encode well-described Ca 2+ -binding proteins, such as calbindin, calretinin, parvalbumin, calmodulin, calumenin, and S100 Ca 2+ -binding proteins. Most of these proteins are localized in the cytosol, but some are present in the PM (e.g., neuronal Ca 2+ sensors), ER (e.g., calreticulin and calnexin), and extracellular matrix (e.g., cell growth regulator with EF-hand domain and fibulin).
We also found the mRNA of genes that encode proteins that are responsible for Ca 2+ release and transport across the ER membrane, such as inositol 1,4,5-trisphosphate (IP 3 ) receptors (itpr1a, itpr1b, and itpr2), ryanodine receptors (ryr1a, ryr1b, ryr2a, and ryr3), and Serca. The highest mRNA level was observed for ryr2a, but ryr1b and itpr1b also reached high expression levels. We detected several mitochondrial Ca 2+ transporters, including members of the solute carrier family 25 (slc25a), leucine zipper and EF-hand containing transmembrane proteins (letm1 and letm2), components of the mitochondrial uniporter complex that consists of mitochondrial calcium uniporter (mcu) that forms a Ca 2+ permeable pore, and three regulatory subunits (micu1, micu2, micu3a, and micu3b). Micu1 and letm2 were enriched in the brain tissue.
A similar percentage of genes that encode transport proteins in the CaTK comprises those with catalytic activity. Most of them are responsible for protein phosphorylation, such as calcium binding protein 39 (cab39), protein kinase C (prkcs, with the highest level of prkcbb), and serum/glucocorticoid regulated kinase (sgks, with the highest level of sgk1).
A much smaller subgroup is involved in the regulation of transcription. This includes cyclic adenosine monophosphate response element binding proteins (crebbpa and crebbpb) and cyclic adenosine monophosphate response element binding protein-regulated transcription coactivators (crebs), cell cycle/division and apoptosis regulators (ccar1 and ccar2), and amyloid β precursor binding family B members (apbb).
In addition to the mRNAs of orai1a and orai2, the expression of other genes that encode proteins that are involved in SOCE was detected in the zebrafish brain, including the Ca 2+ -binding Stim proteins and TRP Ca 2+ channels (trpc3, trpc6) and SOCE-regulating proteins (calcium release activated channel regulator (cracr), SOCE-associated regulatory factor (saraf ), and sigma non-opioid intracellular receptor (sigmar1)).
However, the RNA-seq analysis did not detect mRNAs of some genes that were identified in the zebrafish CaTK. There were no transcripts for several annexins, parvalbumins, Ca 2+ voltage-gated channel, and TRP channel subunits. Voltage-dependent Ca 2+ channels are multi-subunit complexes composed of α-1, β, α-2/delta, and γ subunits. Here, we detected most of the transcripts for genes that encode pore-forming α subunits. However, the transcript for cacna1ea was not detected despite known Cacna1e abundant expression in the rodent brain. α-2/delta, β, and γ are auxiliary subunits, that are important for the assembly and membrane localization of the complex, can modulate calcium currents and channel activation as well inactivation kinetics. Apart from cacnb1 and cacng3a, all genes encoding those subunits were expressed in zebrafish brains. Among TRPC channels, neither transcripts for trpc5b nor trpc7a were found to be expressed in zebrafish brain and the transcript for trpm1a was also absent. Also, no transcripts for most of the calcium homeostasis modulators were detected by RNA-seq in the brain samples.

Expression of SOCE Components in Zebrafish
RNA-seq revealed low levels of mRNA encoding Stim proteins. To determine their expression pattern in the zebrafish, we applied another methodology and performed a precise analysis of a full set of zebrafish SOCE transcripts using RT-PCR. We first checked their mRNA levels in 5 dpf larvae. Surprisingly, we detected the expression of genes for all isoforms of stims and orais (Figure 3). The stim1a transcript was detectable at a very low level. The stim1b transcript was expressed predominately in the head, and the stim2a transcript was expressed predominately in the trunk. Stim2b expression was maintained at similar moderate levels in both the head and trunk of the larvae (Figure 3a). All three orai mRNA transcripts were present in the zebrafish head and trunk, with the highest relative expression of orai1a in both parts (Figure 3b). stim1a transcript was detectable at a very low level. The stim1b transcript was expressed predominately in the head, and the stim2a transcript was expressed predominately in the trunk. Stim2b expression was maintained at similar moderate levels in both the head and trunk of the larvae (Figure 3a). All three orai mRNA transcripts were present in the zebrafish head and trunk, with the highest relative expression of orai1a in both parts (Figure 3b). The mRNA levels of stims and orais were also estimated by RT-PCR in the brain in 8-month-old fish and in several other organs for comparison purposes (Figure 4). The level of mRNA that encode isoforms of these genes varied, depending on the tissue. In the brain, the highest relative expression was found for stim2b and stim1b. The relative expression of stim2a was low, and the stim1a relative expression was very low. That is inconsistent with the RNA-seq results, which showed the highest level of stim1a, rather low levels of stim2b and stim1b and a very low amount of stim2a mRNA in the brain samples (Table S3). Differences between the RNA-Seq and RT-PCR results for some percentage of genes were previously reported [44], especially in the case of lowly expressed genes. Noteworthily, cDNA samples used in these two experiments were obtained using slightly different methods that also could affect the obtained results. Consistent with the relatively high stim2a level in the larval trunk, the high enrichment of stim2a was found in skeletal muscles. Stim1a was relatively highly expressed in the eyes and gonads, and the stim1b relative expression was high in the liver. In the liver and intestine, stim isoforms were expressed at very low levels (Figure 4a). The mRNA levels of stims and orais were also estimated by RT-PCR in the brain in 8-month-old fish and in several other organs for comparison purposes (Figure 4). The level of mRNA that encode isoforms of these genes varied, depending on the tissue. In the brain, the highest relative expression was found for stim2b and stim1b. The relative expression of stim2a was low, and the stim1a relative expression was very low. That is inconsistent with the RNA-seq results, which showed the highest level of stim1a, rather low levels of stim2b and stim1b and a very low amount of stim2a mRNA in the brain samples (Table S3). Differences between the RNA-Seq and RT-PCR results for some percentage of genes were previously reported [44], especially in the case of lowly expressed genes. Noteworthily, cDNA samples used in these two experiments were obtained using slightly different methods that also could affect the obtained results. Consistent with the relatively high stim2a level in the larval trunk, the high enrichment of stim2a was found in skeletal muscles. Stim1a was relatively highly expressed in the eyes and gonads, and the stim1b relative expression was high in the liver. In the liver and intestine, stim isoforms were expressed at very low levels (Figure 4a). We detected a high relative level of orai2 mRNA in the zebrafish brain, where it reached the highest relative level of expression and dominated over orai1a and orai1b. A similar expression pattern (i.e., with the highest orai2 transcript level) was also detected in the eyes. Similar to stims, a low orai isoforms expression was observed in the liver and intestine (Figure 4b).

Analysis of Expression Patterns of Previously Uncharacterized Members of the CaTK in Zebrafish Larval Head and Adult Brain
We next analyzed the relative expression of selected genes of the CaTK using RT-PCR. Of the 380 genes, we chose 77 (63 had unknown expression patterns and 14 had known expression patterns to validate the efficiency of the assay). Table S3 shows the CaTK reference genes which the expressions of were found in the ZFIN database [41].
We assessed the relative mRNA levels of 77 genes in the heads of 5 dpf larvae and in the brains of adult zebrafish ( Figure 5). To compare the relative levels of different genes, we made an assumption that all of them were amplified with equal efficiency. At both stages, the highest relative We detected a high relative level of orai2 mRNA in the zebrafish brain, where it reached the highest relative level of expression and dominated over orai1a and orai1b. A similar expression pattern (i.e., with the highest orai2 transcript level) was also detected in the eyes. Similar to stims, a low orai isoforms expression was observed in the liver and intestine (Figure 4b).

Analysis of Expression Patterns of Previously Uncharacterized Members of the CaTK in Zebrafish Larval Head and Adult Brain
We next analyzed the relative expression of selected genes of the CaTK using RT-PCR. Of the 380 genes, we chose 77 (63 had unknown expression patterns and 14 had known expression patterns to validate the efficiency of the assay). Table S3 shows the CaTK reference genes which the expressions of were found in the ZFIN database [41].
We assessed the relative mRNA levels of 77 genes in the heads of 5 dpf larvae and in the brains of adult zebrafish ( Figure 5). To compare the relative levels of different genes, we made an assumption that all of them were amplified with equal efficiency. At both stages, the highest relative expression was observed for efhd1, slc25a28, apbb2b, cacna1ab, tmbim4, atox1, and cgref1. This observation is roughly consistent with the RNA-seq results, that showed a high expression of a majority of these genes (efhd1, apbb2b, cacna1ab, and tmbim4). The expression of all of these genes is observed in the human brain [45]. In 5 dpf zebrafish larvae, the highest relative expression was observed for tmbim4 (which encodes Golgi membrane protein) and slc25a28 (which encodes the mitochondrial transmembrane transporter). In the adult zebrafish brain, the highest relative expression was observed for efhd1 (which encodes mitochondrial Ca 2+ -binding protein) and grin1a (which encodes a subunit of the NMDA receptor NR1.1). The expression of the latter gene has been previously reported in the zebrafish brain [46]. Also, mRNA for baxa, slc24a4a, and tenm3 were detected by us (on relatively low levels), and those genes were shown to be expressed in parts of zebrafish brain [47][48][49], thus indicating that our assay enables the detection of transcripts present in the brain. observed for tmbim4 (which encodes Golgi membrane protein) and slc25a28 (which encodes the mitochondrial transmembrane transporter). In the adult zebrafish brain, the highest relative expression was observed for efhd1 (which encodes mitochondrial Ca 2+ -binding protein) and grin1a (which encodes a subunit of the NMDA receptor NR1.1). The expression of the latter gene has been previously reported in the zebrafish brain [46]. Also, mRNA for baxa, slc24a4a, and tenm3 were detected by us (on relatively low levels), and those genes were shown to be expressed in parts of zebrafish brain [47][48][49], thus indicating that our assay enables the detection of transcripts present in the brain. Figure 5. A heat map of the normalized expression levels of zebrafish brain CaTK genes, estimated by RT-PCR arrays: The expression levels of 77 CaTK genes in zebrafish brain tissue, including 14 genes for validation (baxa, cacna1ab, cacng2a, chp1, crebbpa, crtc3, grin1a, kcnn1b, micu1, slc24a4a, tenm3, and trpm3; Table 7), were normalized to the 4 reference genes. The expression levels were calculated using the ΔCq method with an assumption of equal efficiencies for all reactions. The relative mRNA levels are color-coded: dark blue (low) to red (high). Notice the different scales that are used for the results from larvae and adults. The RNA samples were obtained from larval heads (a) and from the brains of adult fish (b). n = 3. N, nucleus; EM, extracellular matrix; G, Golgi apparatus; Mt, mitochondria; PM, plasma membrane; and ER, endoplasmic reticulum.
Approximately 30% of the genes had a stable level of expression. The expression of a majority of the genes (70%) significantly increased in the mature fish. The highest fold change (>10-fold change in expression at the larval stage) was observed for several genes that encode VGCCs (cacna1aa, cacna1ab, and cacna1g), an NMDA channel subunit (grin1a), and transmembrane regulators of Ca 2+ channels (cacng2a, cacng3b, cacng7a, cacng7b, cacng8a, and cacng8b). The expression of a few K + channels (kcnma1b, kcnn1a, and kcnt1) significantly increased during development ( Figure 6). Moreover, we detected higher levels of transcripts for members of the amyloid  precursor proteinbinding families A and B (apba1a, apba1b, apba2b, apbb2b, and apbb3). Increases in the mRNA levels of ormdl3, pdzd8, efhd1, ghitm, faim2a, crtc1b, and ppef1 were also observed in the adult brain ( Figure 6).

Figure 5.
A heat map of the normalized expression levels of zebrafish brain CaTK genes, estimated by RT-PCR arrays: The expression levels of 77 CaTK genes in zebrafish brain tissue, including 14 genes for validation (baxa, cacna1ab, cacng2a, chp1, crebbpa, crtc3, grin1a, kcnn1b, micu1, slc24a4a, tenm3, and trpm3), were normalized to the 4 reference genes. The expression levels were calculated using the ∆Cq method with an assumption of equal efficiencies for all reactions. The relative mRNA levels are color-coded: dark blue (low) to red (high). Notice the different scales that are used for the results from larvae and adults. The RNA samples were obtained from larval heads (a) and from the brains of adult fish (b). n = 3. N, nucleus; EM, extracellular matrix; G, Golgi apparatus; Mt, mitochondria; PM, plasma membrane; and ER, endoplasmic reticulum.
Approximately 30% of the genes had a stable level of expression. The expression of a majority of the genes (70%) significantly increased in the mature fish. The highest fold change (>10-fold change in expression at the larval stage) was observed for several genes that encode VGCCs (cacna1aa, cacna1ab, and cacna1g), an NMDA channel subunit (grin1a), and transmembrane regulators of Ca 2+ channels (cacng2a, cacng3b, cacng7a, cacng7b, cacng8a, and cacng8b). The expression of a few K + channels (kcnma1b, kcnn1a, and kcnt1) significantly increased during development ( Figure 6). Moreover, we detected higher levels of transcripts for members of the amyloid β precursor protein-binding families A and B (apba1a, apba1b, apba2b, apbb2b, and apbb3). Increases in the mRNA levels of ormdl3, pdzd8, efhd1, ghitm, faim2a, crtc1b, and ppef1 were also observed in the adult brain ( Figure 6). Only two genes, cracr2ab and trpm4a, presented an opposite effect, in which their mRNA levels dropped in adult fish (14-and three-fold, respectively). Cracr2ab is a Ca 2+ -binding protein that regulates Ca 2+ transport by modulating SOCE. Trpm4a is a cation channel, the activity of which increases as intracellular Ca 2+ concentrations increase. Both of these genes were shown to be expressed in the human brain throughout development, although at very low levels [45].

Discussion
Ca 2+ -mediated signaling regulates numerous cellular processes and involves a vast number of proteins that form a complex network of interactions. To make this possible, Ca 2+ signals differ in speed, amplitude, and spatial patterning, with a broad repertoire of Ca 2+ -sensing, Ca 2+ -buffering, and Ca 2+ -transport proteins, comprising the CaTK [1]. This comprehensive machinery is especially broad in neuronal cells. Neurons possess numerous Ca 2+ channels in different cellular compartments where they may have different functions, such as NMDA channels [50]. There is ample crosstalk between various signaling pathways (e.g., Ca 2+ influx via SOCE refills the ER and may also regulate transcription [28]), cellular organelles (e.g., ER-mitochondria tethering [51]), and cellular systems (e.g., inflammation that affects central nervous system function [52]). To better understand these mechanisms and interactions of Ca 2+ -dependent processes in neurons, detailed analyses of the Ca 2+dependent players are needed.
The present study elucidated the zebrafish CaTK, further demonstrating that this animal model is useful for studying the molecular mechanisms of Ca 2+ signaling and homeostasis. We found that 96% of zebrafish CaTK genes have mammalian orthologues that encode protein machinery for the aforementioned processes, thus allowing analyses of the role of Ca 2+ -dependent processes in fish that are relevant to mammals and the development of translational models of human diseases. We identified the expression patterns of numerous CaTK genes in the zebrafish brain and changes in their mRNA expression between two developmental stages (i.e., in 5 dpf larvae and the adult brain).
The genes we assigned to the zebrafish brain CaTK belong to various signaling pathways. Among the most enriched genes are those that signal important central nervous system neurotransmitters and neurohormones [53][54][55]. Approximately 7% of the genes are involved in various glutamatergic signaling pathways, including both ionotropic and metabotropic glutamate receptor (mGluR) pathways I, II, and III. Such an enrichment could be expected because glutamate is one of the most abundant neurotransmitters in the zebrafish central nervous system [53]. This indicates that many of the secretion and signal transduction processes depend on Ca 2+ signaling [1,2], and zebrafish are a good animal model for studying them. The changes in the expression of selected CaTK genes between mature fish and 5dpf larvae, quantified by RT-PCR arrays: Only genes with a fold change >10 and p < 0.05 are shown (Tukey's Honestly Significant Difference test). The normalized expression in larvae was used as a reference value (= 1, marked as a red line). The data are presented as a fold change ± SD, normalized to the expression level in 5 dpf larvae (n = 3).
Only two genes, cracr2ab and trpm4a, presented an opposite effect, in which their mRNA levels dropped in adult fish (14-and three-fold, respectively). Cracr2ab is a Ca 2+ -binding protein that regulates Ca 2+ transport by modulating SOCE. Trpm4a is a cation channel, the activity of which increases as intracellular Ca 2+ concentrations increase. Both of these genes were shown to be expressed in the human brain throughout development, although at very low levels [45].

Discussion
Ca 2+ -mediated signaling regulates numerous cellular processes and involves a vast number of proteins that form a complex network of interactions. To make this possible, Ca 2+ signals differ in speed, amplitude, and spatial patterning, with a broad repertoire of Ca 2+ -sensing, Ca 2+ -buffering, and Ca 2+ -transport proteins, comprising the CaTK [1]. This comprehensive machinery is especially broad in neuronal cells. Neurons possess numerous Ca 2+ channels in different cellular compartments where they may have different functions, such as NMDA channels [50]. There is ample crosstalk between various signaling pathways (e.g., Ca 2+ influx via SOCE refills the ER and may also regulate transcription [28]), cellular organelles (e.g., ER-mitochondria tethering [51]), and cellular systems (e.g., inflammation that affects central nervous system function [52]). To better understand these mechanisms and interactions of Ca 2+ -dependent processes in neurons, detailed analyses of the Ca 2+ -dependent players are needed.
The present study elucidated the zebrafish CaTK, further demonstrating that this animal model is useful for studying the molecular mechanisms of Ca 2+ signaling and homeostasis. We found that 96% of zebrafish CaTK genes have mammalian orthologues that encode protein machinery for the aforementioned processes, thus allowing analyses of the role of Ca 2+ -dependent processes in fish that are relevant to mammals and the development of translational models of human diseases. We identified the expression patterns of numerous CaTK genes in the zebrafish brain and changes in their mRNA expression between two developmental stages (i.e., in 5 dpf larvae and the adult brain).
The genes we assigned to the zebrafish brain CaTK belong to various signaling pathways. Among the most enriched genes are those that signal important central nervous system neurotransmitters and neurohormones [53][54][55]. Approximately 7% of the genes are involved in various glutamatergic signaling pathways, including both ionotropic and metabotropic glutamate receptor (mGluR) pathways I, II, and III. Such an enrichment could be expected because glutamate is one of the most abundant neurotransmitters in the zebrafish central nervous system [53]. This indicates that many of the secretion and signal transduction processes depend on Ca 2+ signaling [1,2], and zebrafish are a good animal model for studying them.
A large number of active CaTK genes in the brain in adult zebrafish encodes Ca 2+ channels in the cell membrane. Genes that are involved in glutamatergic transmission were enriched, but we also identified genes that encode subunits of VGCCs and Trp and Orai channels. We also found genes that encode proteins that are responsible for Ca 2+ transport in the ER (e.g., Serca and IP 3 and ryanodine receptors), several mitochondrial Ca 2+ transporters (e.g., Scl25 and Letm, components of the mitochondrial uniporter complex), and Ca 2+ -binding proteins (e.g., Calbindin1, Calbindin2 (Calretinin), Calreticulin, and Calnexin) [12,[56][57][58]. The expression of all four isoforms of stims (stim1a, stim1b, stim2a, and stim2b) was detected in the brain of adult zebrafish. In mammals, different Stim1-to-Stim2 ratios are observed, depending on the specific tissue. In mice, Stim1 has a higher level of expression than Stim2, except in the brain where the Stim2 transcript levels dominate over Stim1 [24]. In humans, based on the RNA-seq data from the Expression Atlas [23], more uniform STIM1 and STIM2 expressions are observed, and the STIM1 expression predominates. In zebrafish, a high relative level of Stim1b and Stim2b expressions was observed in the brain, and both Stim2 isoforms predominated in muscles. This is in contrast to mammalian cells, in which the STIM1 splice variant STIM1L exists and is highly expressed in muscle tissue [59]. However, a similar sequence is not found in zebrafish, based on Ensembl (release 94) [60]. Based on Expression Atlas RNA-seq data, ORAI1 is the dominate ORAI channel in most human tissues. However, ORAI2 is highly enriched in the brain, the expression of which predominates over ORAI1 and ORAI3. A similar pattern was observed in mouse tissue [22] and in the zebrafish brain, as shown in the present work. The existence of nSOCE has been questioned [29], but further studies of nSOCE machinery in zebrafish might help resolve this issue.
RT-PCR arrays allowed us to validate the expression patterns for 63 CaTK genes, the expression patterns of which were not yet established in zebrafish. We compared their levels of expression in 5 dpf larval heads with the adult brain. Generally, approximately 70% of the genes had a higher expression in the adult brain than in the larvae. This could be related to the method of preparation, in which samples that were obtained from larval heads were only enriched in neuronal tissue and were contaminated with muscles and skin. However, this change may be the result of brain maturation, based on genes which the expression of increases in the adult brain.
The highest increase in expression was observed for genes that encode channels in the cell membrane and their regulators. These genes represented over half of the most differentially expressed genes. That indicates their importance in brain maturation and adult brain function. Among these, several cacng genes that encode transmembrane AMPA receptor regulatory proteins (TARPs) were found. TARPs modulate the gating properties of AMPA receptors and affect their trafficking and subcellular localization. Six classic TRAPs that are encoded by cacng genes are found in the mammalian central nervous system: γ2, γ3, γ4, γ5, γ7, and γ8 [61]. In the zebrafish brain, we detected transcripts for cacng1a, cacng1b, cacng2a, cacng3b, cacng4a, cacng4b, cacng5a, cacng5b, cacng5a, cacng5b, cacng7a, cacng7b, cacng8a, and cacng8b. In larval heads, the mRNA levels of most of them (except cacng4b and cacng5b) were at a very low level and significantly increased in the adult brain. A similar tendency was found in the human brain for CACNG3, whereas CACGN2, CACGN7, and CACGN8 presented rather stable levels during maturation [45]. Increases in the mRNA levels of genes that encode pore-forming subunits of VGCCs (cacna1aa, cacna1ab, cacna1ba, cacna1bb, cacna1c, cacna1da, cacna1db, cacna1eb, cacna1fa, cacna1fb, cacna1g, cacna1ha, cacna1hb, and cacna1i) were also noted. Cacna1g encodes T-type, low-voltage activated calcium channels, the expression of which also increases during mouse development [62], reaches the highest level during childhood in the human brain [45], and is activated by the Lef1/β-catenin complex in thalamic neurons [63]. We also found an increase in the expression of the NMDA receptor subunit encoding gene grin1a and APP-binding family A and B genes. Most of these genes that encode the AMPA, NMDA, and CACN channels are associated with neurological diseases, such as epilepsy [64] and Alzheimer's disease [65], thus making them a potential treatment target. Increases in the expression of members of the amyloid β precursor protein-binding families A and B were also found. These are neuronal adapter proteins that are involved in amyloid β production and play a role in the pathogenesis of Alzheimer's disease [66]. Although zebrafish have been extensively studied as a model organism for neurodegenerative diseases and brain aging [31], the link between these studies and Ca 2+ needs to be strengthened. The aforementioned genes are potential targets for further research on neurodegeneration with regards to the regulation of Ca 2+ .
The GO term analysis in the present study showed increases in the mRNA levels of proteins that are involved in important metabolic pathways in the mature zebrafish brain (e.g., the transcript for ormdl3 that encodes a protein that is involved in sphingolipid synthesis [67] and pdzd8). Pdzd8 participates in lipid transport, the formation of ER-mitochondrion contacts, and the regulation of Ca 2+ dynamics in mammalian neurons [68]. Additionally, the expression of ghitm1 (which encodes a mitochondrial protein) and faim2a (which encodes a plasma membrane protein) increased in the mature brain, similar to observations in human brain samples [45]. However, little is known about the roles of these two proteins in the central nervous system. The expression of the protein serine/threonine phosphatase ppef1 and transcriptional coactivator crtc1b, which present Ca 2+ -dependent activity, also increased in the adult zebrafish brain, but such increases were not observed for their orthologues in the human brain [45].
The transcripts of the following genes that encode proteins without well-defined functions were detected for the first time in the zebrafish brain: atox1, calhm2, cgref1, cherp, tusc2, and genes that encode proteins with EF-hand Ca 2+ -binding domains (e.g., efcab1, efcab7, efcab11, efcc1, efhc2, efhd1, efhd2, necab1, necab2, and ppef1). The efhd1 gene seems to be of high interest. Although it already had a relatively high level of expression in 5 dpf larvae, its expression increased approx. 15-fold in the adult brain. A similar pattern was observed in the human brain [45]. EFHD1 has been shown to be involved in mitoflash activation in HeLa cells [69].
Very high relative mRNA levels of tmbim4 and slc25a28 were found in the zebrafish brain, with no major changes in expression during maturation. TMBIM4 is located in the Golgi apparatus and ER membrane and was shown to be ubiquitously expressed in mouse tissues [70]. TMBIM4 overexpression reduces Ca 2+ release from internal stores [70,71], but its function in neuronal cells is unknown. SLC25a28 (mitoferrin-2) is a mitochondrial iron transporter. Disruptions of iron homeostasis may be involved in dopaminergic neuron degeneration in Parkinson's disease [72].
Our research supports the use of zebrafish for studying Ca 2+ homeostasis and signaling. The transparency of its larvae, the rapid development of the nervous system, and the ease of genetic manipulation make zebrafish useful for many types of research. Zebrafish embryos can be genetically manipulated relatively easily for transient or stable protein expression, knockouts, or other types of mutations [39]. Functional conservation among vertebrates (approx. 70% of human genes have at least one zebrafish orthologue) makes zebrafish an outstanding model for examining the molecular mechanisms of human diseases [73], assaying gene activity, and discovering novel compounds for the treatment of disease. CRISPR/Cas9 technology is well-established for zebrafish, and crispants can be created to introduce mutations of specific genes [30]. This creates the possibility of generating fish models of human diseases. Moreover, Ca 2+ probes can be introduced to target the cytosol, ER, mitochondria, and lysosomes, thus permitting analyses of Ca 2+ levels [32,74]. This provides many opportunities to implement various experimental approaches, including in vivo visualization and analyses of the dynamics of Ca 2+ signaling.

Conclusions and Future Directions
The present CaTK data will be useful for designing experiments to dissect the functions of specific genes and to understand the molecular mechanisms of Ca 2+ homeostasis and signaling in wildtype fish and models of neurological diseases. Ca 2+ homeostasis and signaling are dysregulated in many neurological diseases [12,75]. The fact that zebrafish are amenable to in vivo experiments and the high conservation of CaTK members (96% as found in the present study) allow broad investigations of Ca 2+ signaling at the systemic level. The genes that were identified in the present study will allow further studies of neurodevelopment, neuronal Ca 2+ signaling, and neurodegeneration in zebrafish as a model organism.