Reverse Effect of Mammalian Hypocalcemic Cortisol in Fish: Cortisol Stimulates Ca2+ Uptake via Glucocorticoid Receptor-Mediated Vitamin D3 Metabolism

Cortisol was reported to downregulate body-fluid Ca2+ levels in mammals but was proposed to show hypercalcemic effects in teleostean fish. Fish, unlike terrestrial vertebrates, obtain Ca2+ from the environment mainly via the gills and skin rather than by dietary means, and have to regulate the Ca2+ uptake functions to cope with fluctuating Ca2+ levels in aquatic environments. Cortisol was previously found to regulate Ca2+ uptake in fish; however, the molecular mechanism behind this is largely unclear. Zebrafish were used as a model to explore this issue. Acclimation to low-Ca2+ fresh water stimulated Ca2+ influx and expression of epithelial calcium channel (ecac), 11β-hydroxylase and the glucocorticoid receptor (gr). Exogenous cortisol increased Ca2+ influx and the expressions of ecac and hydroxysteroid 11-beta dehydrogenase 2 (hsd11b2), but downregulated 11β-hydroxylase and the gr with no effects on other Ca2+ transporters or the mineralocorticoid receptor (mr). Morpholino knockdown of the GR, but not the MR, was found to impair zebrafish Ca2+ uptake function by inhibiting the ecac expression. To further explore the regulatory mechanism of cortisol in Ca2+ uptake, the involvement of vitamin D3 was analyzed. Cortisol stimulated expressions of vitamin D-25hydroxylase (cyp27a1), cyp27a1 like (cyp27a1l), 1α-OHase (cyp27b1) at 3 dpf through GR, the first time to demonstrate the relationship between cortisol and vitamin D3 in fish. In conclusion, cortisol stimulates ecac expression to enhance Ca2+ uptake functions, and this control pathway is suggested to be mediated by the GR. Lastly, cortisol also could mediate vitamin D3 signaling to stimulate Ca2+ uptake in zebrafish.


Introduction
Corticosteroids (CSs) are primarily synthesized from cholesterol through a series of reactions. CSs, which consist of glucocorticoids (GCs) and mineralocorticoids (MCs), are vital hormones for mammals, and are involved in regulating the osmolality and ion levels, body fluids, energy metabolism, respiration, and immune reactions [1,2]. GCs are efficient treatment for asthma, rheumatoid arthritis, and atopic dermatitis because they can reduce immune responses; however, several systemic side-effects including osteoporosis are induced. An imbalance of Ca 2+ handling is an important factor causing osteoporosis [3]. Regulation of Ca 2+ absorption and emission is closely associated with the bone structure, and GC was reported to cause malabsorption and malemission of Ca 2+ in the intestines and kidneys [2][3][4]. In mammals, GC was proposed to downregulate Ca 2+ levels of body fluids through modulating the renal and duodenal expressions of TRPV6 and calbindin-D 9 K [5,6].
CS is synthesized in the adrenal cortex of mammals, but in the interrenal tissue of the head kidneys in teleosts. Physiological functions of CSs in teleosts are similar to those in mammals, and CS signaling is also mediated by the GC receptor (GR) and MC receptor (MR), which are ligand-activated transcription factors [7,8]. Both the GR and MR can bind the GC-responsive element (GRE) of the gene promoter and form GR-GR, MR-MR, and MR-GR dimers [9]. In CS synthesis, 11b-hydroxylase (CYP11B1) and aldosterone synthase (CYP11B2) are enzymes in the final step of the synthesis of cortisol and aldosterone, respectively. Teleosts may lack aldosterone synthase, and therefore cortisol is the main CS hormone in teleosts [10,11]. Some in vitro studies demonstrated that cortisol stimulated the transcriptional activity in mammalian cell lines transiently transfected with an expression construct containing a fish GR or MR and a reporter plasmid containing multiple GREs, implying that both teleostean CS receptors can be bound by cortisol with different affinities [12][13][14][15]. Based on those results, cortisol was suggested to have both GC and MC functions through different CS receptors, GR or MR, in teleosts; however, very few studies have investigated if the GR, MR, or both are involved in specific physiological processes in teleosts. In a recent study on Atlantic salmon, the GR and MR were found to differentially mediate the stimulation of various ion transporters in the gills during acclimation to salinity changes [16].
Vitamin D 3 was a vital calcitrophic endocrine to regulate Ca 2+ homeostasis in vertebrates. Liver vitamin D-25hydroxylase (CYP27A1) converts vitamin D 3 precursor to 25-hydroxyvitamin D 3 (25(OH)D 3 ), which is then converted to 1,25-dihydroxytamin D 3 (1,25 (OH) 2 D 3 ), the active form of the vitamin D 3 , by renal 1a-OHase (CYP27B1) [17]. Vitamin D 3 spreads its function through its receptor, vitamin D 3 receptor (VDR). VDR is a ligand-active transcription factor, and duodenal trpv6 (ecac) is the target of the vitamin D 3 -VDR complex in mammals [18]. Stimulation of the duodenal trpv6 expression by vitamin D 3 is one of important pathways to enhance Ca 2+ uptake in mammals [19]. In fish, vitamin D 3 had been demonstrated to elevate serum Ca 2+ level [20,21] .Vitamin D 3 was also proposed to be associated with Ca 2+ transport in the gills based on the vitamin D 3 deficiency-induced impairment of growth and mineralization in the fish skeleton [22,23]. Indeed, changes in the plasma levels of 1,25 (OH) 2 D 3 or mRNA expression of gill/kidney VDR have been observed in salmon undergoing smoltification and migrating from freshwater (low Ca 2+ concentrations) to seawater (high Ca 2+ concentrations), suggesting that synthesis of the sterol and its receptor might be regulated depending upon ambient Ca 2+ concentrations [23]. In mammals, glucocorticoid has been well documented to affect vitamin D 3 metabolism although the actions varied depending on species [24][25][26][27][28][29]. This suggests a possible association between glucocorticoid, vitamin D 3 and Ca 2+ homeostasis in mammals; however, it is unknown whether this connection is also developed in fish.
In teleosts, cortisol is well demonstrated to regulate the mechanisms of ionic and osmotic balance, but only few studies investigated the role of cortisol in controlling transepithelial Ca 2+ transport [30,31]. Teleostean fish, unlike terrestrial vertebrates, obtain Ca 2+ mainly via absorption from the environment rather than by dietary means. In adult fish, the predominant route of Ca 2+ entry from the environment is across the gill epithelium while in larvae, the body skin is the major route of Ca 2+ uptake before full development of the gills occurs [32,33]. The Ca 2+ uptake function is well regulated for maintaining the internal Ca 2+ homeostasis during acclimation to aquatic environments with fluctuating Ca 2+ levels (,0.01 mM in soft fresh water to .10 mM in seawater) [30,34]. According to the current model in mammals and teleosts, active transcellular Ca 2+ transport is carried out through the operation of apical epithelial Ca 2+ channels (ECaC, TRPV5, and/or TRPV6), and basolateral plasma membrane Ca 2+ -ATPase (PMCA) and the Na + /Ca 2+ exchanger (NCX) [34,35]. Exposure of rainbow trout to a reduced ambient calcium level induces a rapid increase in systemic cortisol levels [36], and exogenous cortisol can stimulate branchial ECaC mRNA and protein expressions [37]. In an experiment with cultured gills, cortisol was also found to enhance transepithelial Ca 2+ transport [38]. Those studies suggested a role of cortisol in control of Ca 2+ uptake in teleosts; however, it is still unclear if cortisol controls only ECaCs or other Ca 2+ transporters (NCX and PMCA), and if this control pathway is mediated by the GR, MR, or both. Elucidation of these issues would enhance our understanding of the molecular physiological mechanisms of cortisol's control of epithelial Ca 2+ transport, an essential component associated with bone structure and formation in vertebrates.
Recently, zebrafish have become a model for research on ion regulation and related endocrine controls due to the well-developed genetic database and applicability of various molecular physiological approaches [32,39]. In zebrafish gills and skin (in embryonic stages), a specific ionocyte type that expresses ECaC, PMCA2, and NCX1b was identified to be responsible for transepithelial Ca 2+ uptake function [32,33,[39][40][41], providing a suitable model to further explore cortisol's control of Ca 2+ uptake mechanisms. The present study attempted to address 3 specific questions: (I) Does cortisol control zebrafish Ca 2+ uptake function by regulating the ecac, ncx1b, and/or pmca2? (II) Does cortisol control zebrafish Ca 2+ uptake function through mediation by the GR, MR, or both receptors? (III) Does cortisol control zebrafish Ca 2+ uptake function by regulating vital calcitrophic endocrine suchlike vitamin D 3 ? The effects of environmental Ca 2+ levels and exogenous cortisol on Ca 2+ contents and influx, the mRNA expressions of Ca 2+ transporters (ecac, ncx1b, and pmca2), the steroidogenesis enzymes 11b-hydroxylase and hydroxysteroid 11-beta dehydrogenase 2 (hsd11b2), CS receptors (gr and mr) and vitamin D 3 -related genes (vdra, vdrb, cyp27a1, cyp27a1l and cyp27b1) were investigated. Moreover, effects of knockdown of the GR or MR on Ca 2+ contents and influx, and the expression of Ca 2+ transporters and vitamin D 3 -related genes in zebrafish embryos were examined.

Experiment animals
Zebrafish (Danio rerio) were kept in local tap water ([Ca 2+ ] = 0.2 mM) at 28.5uC under a 14:10-h light-dark photoperiod at the Institute of Cellular and Organismic Biology, Academia Sinica, Taipei, Taiwan. Experimental protocols were approved by the Academia Sinica Institutional Animal Care and Utilization Committee (approval no.: RFIZOOHP220782).

Acclimation experiments
Artificial fresh waters with high-(2 mM) and low-Ca 2+ (0.02 mM) levels were prepared with double-deionized water (model Milli-RO60; Millipore, Billerica, MA, USA) supplemented with adequate CaSO 4 ?2H 2 O, MgSO 4 ?7H 2 O, NaCl, K 2 HPO 4 , and KH 2 PO 4 . Ca 2+ concentrations (total Ca 2+ levels measured by absorption spectrophotometry) of the high-and low-Ca 2+ media were 2 and 0.02 mM, respectively, but the other ion concentrations of the 3 media were the same ([Na + ], 0.5 mM; [Mg 2+ ], 0.16 mM; and [K + ], 0.3 mM) as those in local tap water. Variations in ion concentrations were maintained within 10% of the predicted values. Fertilized zebrafish eggs were transferred to high-and low-Ca 2+ media, respectively, and incubated thereafter until sampling at 3 d post-fertilization (dpf). The sampling time in this study was based previously [42].

Cortisol incubation experiments
For cortisol incubation experiments, we based cortisol dosage from previous study [43]. Cortisol (hydrocortisone, H4881, Sigma Chemical Co., St Louis, MO, USA) was dissolved in local tap water at 0 (control), 20, and 40 mg/l. Zebrafish embryos were incubated in the cortisol media immediately after fertilization, and were sampled at 1 or 3 dpf for the subsequent analysis. The incubation media were changed with new cortisol solution every day to maintain constant levels of cortisol. During incubation, neither significant mortality nor abnormal behavior was found.

Whole-body Ca 2+ content
Zebrafish embryos were anesthetized with MS-222 (Sigma) and then briefly rinsed in deionized water. 30 individuals were pooled as 1 sample. HNO 3 (13.1 N) was added to samples for digestion at 60uC overnight. Digested solutions were diluted with doubledeionized water, and the total Ca 2+ content was measured with a Z-8000 atomic absorption spectrophotometer (Hitachi, Tokyo, Japan). Standard solutions (Merck, Darmstadt, Germany) were used to make the standard curves.

Whole-body Ca 2+ influx
By following previously described methods [44] with some modifications, zebrafish embryos were dechorionated, rinsed briefly in deionized water, and then transferred to 2 ml of 45 Ca 2+ (Amersham, Piscataway, NJ; with a final working specific activity of 1,2 mCi/mmol)-containing medium for a subsequent 4-h incubation. After incubation, embryos were washed several times in isotope-free water medium. Six embryos were pooled into 1 vial, anesthetized with MS-222, and digested with tissue solubilizer (Solvable; Packard, Meriden, CT, USA) at 60uC for 8 h. The digested solutions were supplemented with counting solution (Ultima Gold; Packard), and the radioactivities of the solutions were counted with a liquid scintillation beta counter (LS6500; Beckman, Fullerton, CA, USA). The Ca 2+ influx was calculated using the following formula: J in = Q embryo X out 21 t 21 W 21 ; where J in is the influx (pmol?mg 21 ?h 21 ), Q embryo is the radioactivity of the embryo (cpm per individual) at the end of incubation, X out is the specific activity of the incubation medium (cpm/pmol), t is the incubation time (h), and W is the average body wet weight of different-stage embryos (mg).

RNA extraction
After anesthetized with 0.03% MS222, appropriate amounts of zebrafish tissues or embryos were collected and homogenized in 1 ml Trizol reagent (Invitrogen, Carlsbad, CA, USA), then mixed with 0.2 ml chloroform, and thoroughly shaken. After centrifugation at 4uC and 12,0006g for 30 min, the supernatants were obtained. The samples were then mixed with an equal volume of isopropanol. Pellets were precipitated by centrifugation at 4uC and 12,0006g for 30 min, washed with 70% alcohol, and stored at 220uC until use.

Reverse-transcription polymerase chain reaction (RT-PCR) analysis
For complementary (c)DNA synthesis, 1,5 mg of total RNA was reverse-transcribed in a final volume 20 ml containing 0.5 mM dNTPs, 2.5 mM oligo (dT) 20 , 250 ng random primers, 5 mM dithiothreitol, 40 units RNase inhibitor, and 200 units Superscript RT (Invitrogen) for 1 h at 50uC followed a 70uC incubation for 15 min. For PCR amplification, 2 ml cDNA was used as template in a 50-mL final reaction volume containing 0.25 mM dNTPs, 2.5 units Taq polymerase (Takara, Shiga, Japan), and 0.2 mM of each primer (Table S1). 30 cycles were performed for each reaction. All amplicons were sequenced to ensure that the PCR products were the desired gene fragments.

Quantitative real-time PCR (qPCR)
qPCR was performed with a LightCycler real-time PCR system (Roche, Penzberg, Germany) in a final volume of 10 ml, containing 5 ml 26 SYBR Green I Master (Roche Applied System), 300 nM of the primer pairs, and 20,30 ng cDNA. The standard curve for each gene was checked in a linear range with b-actin as an internal control. The primer sets for the qPCR are shown in Table S2.

In situ hybridization
Zebrafish ecac or gr Fragments were obtained by PCR and inserted into the pGEM-T easy vector (Promega, Madison, WI, USA). The inserted fragments were amplified with the T7 and SP6 primers by PCR, and the products as templates were used for the in vitro transcription with T7 and SP6 RNA polymerase (Roche) in the presence of digoxigenin (DIG)-UTP (Roche) to, respectively, synthesize sense and anti-sense probes. Zebrafish embryos were anesthetized on ice and fixed with 4% paraformaldehyde (PFA) in a phosphate-buffered saline (PBS; 1.4 mM NaCl, 0.2 mM KCl, 0.1 mM Na 2 HPO 4 , and 0.002 mM KH 2 PO 4 ; pH 7.4) solution at 4uC overnight. To do in situ hybrizization, we followed previously [40]. For the quantification of density, eight areas (85680 mm 2 each) on the yolk sac surface of an embryo were chosen for counting.

Organ culture
Adult fish were anesthetized with 0.03% MS222 and then gills were dissected and directly transferred to the pre-incubation DMEM medium (Invitrogen) containing 50 mg/ml of penicillin (Invitrogen), 50 mg/ml of streptomycin (Invitrogen) and 20% Fetal Bovine Serum (Invitrogen). Individual gill arches were carefully separated from the whole gill structure. Each gill arch was cut lengthwise, and the cut filament was designated as one sample to be incubated in a well (96-well). The cut gill filaments were incubated with the freshly prepared pre-incubation medium (control group) and the DMEM with supplementary 20 mg/l cortisol, respectively. The media were freshly prepared and replaced twice per day. Organ culture was carried out at 28uC for 1 d in 96-well culture plates in a humidified chamber supplied with 95% O 2 and 5% CO 2 .  (Table S3) and then cloned into the pCS2+GFP XLT vector, and the expression constructs were linearized to synthesize capped mRNA (cRNA) using an SP6 message RNA polymerase kit (Ambion, Austin, TX, USA). To confirm safety and efficiency of MOs and cRNA, we tested to inject various dosages of MOs and cRNA. Finally, we choose 2 ng/embryo (for MO) and 300 pg/ embryo (for cRNA) to inject. Under these dosages, neither significant mortality nor abnormal behavior was found. The MO (2 ng/embryo) and/or cRNA (300 pg/embryo) were injected into embryos at the 1,2 cell stage using an IM-300 microinjector system (Narishige Scientific Instrument Laboratory, Tokyo, Japan). Green fluorescent protein (GFP) signals in 1-dpf embryos were observed by fluorescence microscopy (Axioplan 2 Imaging; Carl Zeiss, Oberkochen, Germany). MO-injected embryos at 1 or 3 dpf were sampled for subsequent analyses.

Morpholino oligonucleotide (MO) knockdown and rescue
To rescue the defects caused by the MO, a full-length GR was PCR-amplified with a specific primer (Table S3) and cloned into the pCS2+ vector, and the construct was used to synthesize cRNA. Another GR MO, GR-SB MO, was designed at an intron-exon boundary by following a previous study [45], and this MO could only block endogenous GR translation. The full-length GR cRNA(300 pg/embryo) and GR-SB MO (59-CTGCTTCATG-TATTTTAGG-39; 2 ng/embryo) were injected into embryos at the 1,2 cell stage, and embryos were sampled at 3 dpf.

Western blot analysis
Thirty embryos were pooled as one sample and homogenized. Protein of 50 mg/well was loaded to a 10% SDS-PAGE at 100 V for 2 h. After separation, proteins were transferred onto polyvinylidene difluoride membrane (Millipore) at 100 V for 2 h. After being blocked for 1.5 h in 5% nonfat milk, the blots were incubated with GR (Santa Cruz Biotechnology) or MR polyclonal antibody (Abcam) overnight 4uC, diluted 1:500 and with an alkaline-phos-phatase-conjugated goat anti-rabbit IgG (diluted 1:2500, room tem-perature; Jackson Laboratories) for another 2 h. The blots were developed with 5-bromo-4-chloro-3indolylphosphate/nitro-blue tetrazolium.

Cryosectioning
Fresh zebrafish gills were fixed with 4% PFA at 4uC for 3 h and then immersed serially in PBS containing 5, 10, and 20% sucrose for 15 min at room temperature. Finally, gills were soaked in a mixed PBS solution (OCT compound: 20% sucrose at 1:2) overnight and then embedded with OCT compound embedding medium (Sakura, Tokyo, Japan) at 20uC. Cryosections at 6 mm were made with a cryostat (CM 1900; Leica, Heidelberg, Germany), and these were placed onto poly-L-lysine-coated slides (EMS, Hatfield, PA).

Immunocytochemistry
Prepared slides were rinsed in PBS and blocked with 3% BSA for 30 min. Afterward, the slides were first incubated with an a5 monoclonal antibody against the a-subunit of the avian Na,K-ATPase (NKA) (Hybridoma Bank, University of Iowa, Ames, IA; 1:600 dilution) overnight at 4uC. The slides were washed twice with PBS and incubated with an Alexa Fluor 568 goat anti-mouse IgG antibody (Molecular Probes, Carlsbad, CA; 1:200 diluted with PBS) for 2 h at room temperature. Images were acquired with a Leica TCS-NT confocal laser scanning microscope (Leica) or an Axioplan 2 imaging microscope.

Potential regulatory elements upstream of the zebrafish ecac gene
The zebrafish ecac genomic sequence was obtained from a zebrafish genome database (http://www.ncbi.nlm.nih.gov/projects/genome/guide/zebrafish/). Potential regulatory elements upstream of the ecac gene were predicted by Genomatix MatInspector (http://www.genomatix.de).

Statistical analysis
Data are presented as the mean6SD and were analyzed by oneway ANOVA and Student's t-test.

mRNA expressions of mr and gr
Both mr and gr mRNAs were universally expressed in all tissues studied (Fig. 1A). In developing embryos, mRNA expression of the gr was first detected at 1 h post-fertilization (hpf) and throughout development; however, mr mRNA only began to be expressed at 12 hpf (Fig. 1B).

Ca 2+ incubation on Ca 2+ influx and Ca 2+ -related genes
After acclimation to artificial fresh water containing different levels of Ca 2+ for 3 d, zebrafish Ca 2+ influx was significantly stimulated by low-Ca 2+ water ( Fig. 2A). Similarly, ecac, 11bhydroxylase and gr mRNA expressions were also significantly stimulated by low-Ca 2+ water (Fig. 2B). On the contrary, pmca2, ncx1b, hsd11b2, and mr mRNA expressions were not affected by environmental Ca 2+ levels (Fig. 2B).

Exogenous cortisol on Ca 2+ influx/contents and mRNA expressions of Ca 2+ -related target genes
To test the hypothesis of whether cortisol can affect Ca 2+ uptake, zebrafish embryos at the 1,2-cell stage were treated with exogenous cortisol for 3 d. Incubation with exogenous cortisol caused dose-dependent effects on both Ca 2+ contents and influx in 3-dpf zebrafish embryos. Compared to the control group (0 mg/l), cortisol-treated groups (20 and 40 mg/l) showed significant increases in Ca 2+ content and influx (Fig. 3A, B). The qPCR revealed differential effects of exogenous cortisol on the mRNA expressions of Ca 2+ transporters. mRNA expressions of zebrafish ncx1b and pmca2 were not affected by cortisol treatment (Fig. 3C); however, that of ecac was significantly upregulated by exogenous cortisol in a dose-dependent pattern (Fig. 3C). Furthermore, exogenous cortisol was also used to treat cultured gills. Similarly, ecac mRNA expression was affected by exogenous cortisol in gills, but ncx1b and pmca2 were not affected (Fig. S1). To support the data (Fig. 3C) of qPCR analyses for Ca 2+ transporter expressions, in situ hybridization of ecac was conducted in the embryos treated with cortisol. As shown in Fig. 4A and B, exogenous cortisol significantly stimulated the density of ecac -expressing cells in 3-dpf zebrafish embryos. Exogenous cortisol also caused differential effects on mRNA expressions of 11b-hydroxylase, hsd11b2, gr, and mr in zebrafish embryos. According to the qPCR in 3-dpf embryos, exogenous cortisol significantly inhibited mRNA expressions of 11b-hydroxylase (in a dose-dependent manner) and gr and stimulated hsd11b2, but did not affect that of the mr (Fig. 3C).

Loss-of-function effects on Ca 2+ contents/influx/ transporters and density of ecac-expressing cells in zebrafish embryos
To block the endogenous cortisol signaling pathway, MR and GR MOs were used to respectively inhibit translation of zebrafish GR and MR. The specificity and effectiveness of the MR and GR MOs were respectively confirmed by co-injection with zebrafish MR or GR cRNAs. Zebrafish embryos injected with only cRNAs (with GFP fusion) revealed signals of GFP translation (Fig. 5A, B), confirming the translation of MR and GR cRNAs. On the other hand, embryos co-injected with the MR (or GR) MO and MR (or GR) cRNA with GFP showed no green fluorescence (Fig. 5C, D), indicating that the MO specifically and effectively knocked-down the translation of MR (or GR) mRNA. In addition, Western blot was also used to further demonstrate MO specificity. As a result, GR or MR MO was found to specifically downregulate GR or MR protein level in 3-dpf zebrafish embryos (Fig. 5E).
After specificity tests, respective MOs were injected into 1,2cell embryos. Compared to the control MO, the GR MO caused   significant increases in Ca 2+ content and influx in 3-dpf zebrafish embryos, but the MR MO had no effects (Fig. 6A, B). The qPCR assay of the mRNA expressions of Ca 2+ transporters showed that the GR MO significantly reduced expression of the ecac, but did not affect ncx1b and pmca2 mRNA expressions in 3-dpf zebrafish embryos (Fig. 6C).
To further support these data, mRNA density of ecac-expressing cells in the skin of zebrafish morphants were analysed. The ecac-expressing cell was also significantly decreased upon GR MO injection (Fig. 7A,  B). In contrast with the GR MO, the MR MO did not affect the expressions of ncx1b, pmca2, and ecac genes as well as the density of ecacexpressing cells in 3-dpf zebrafish embryos (Fig. 6C, 7A, B).

Effects of MR/GR MO on Ca 2+ influx and ecac mRNA expression in zebrafish embryos with exogenous cortisol or low Ca 2+ media
To precisely ascertain the different roles of the zebrafish GR and MR, the zebrafish were incubated with or without cortisol (20 mg/l) after injections with the MOs. Compared to the control group (control MO injection without cortisol), both groups of the control MO with cortisol and the MR MO with cortisol exhibited a significantly higher Ca 2+ influx at 3-dpf, but the GR MO-injected embryos with exogenous cortisol did not (Fig. 8A). Similarly, ecac mRNA expression in the control MO with cortisol and MR MO with cortisol was significantly stimulated, while that of the GR MOinjected embryos was not affected by exogenous cortisol (Fig. 8B).
To further support the role of the GR in the Ca 2+ uptake mechanism, the zebrafish GR-SB MO and/or GR cRNA were coinjected into zebrafish 1,2-cell-stage embryos. Compared to the control MO-injected group, GR-SB MO, similar to GR MO, also caused significant decreases in both Ca 2+ influx and ecac mRNA expression in 3-dpf zebrafish embryos (Fig. 9A, B); however, coinjection with zebrafish GR cRNA and the GR-SB MO rescued the defective Ca 2+ influx and ecac mRNA expression caused by the injection of the GR-SB MO (Fig. 9A, B).
Low Ca 2+ medium was known to stimulate ecac expression in zebrafish [40,41] (Fig. 10).Whether this ecac expression upregulation by low Ca 2+ medium is mediated by GR or MR was further clarified in the following experiments. One,two-cell-stage embryos were injected with the control MO, MR MO, and GR MO, respectively, and then were incubated in 2.0 mM (high) or 0.02 mM (low) Ca 2+ medium. Compared to the control MO in low Ca 2+ medium, the GR morphants in low Ca 2+ medium were significantly lower in the ECaC mRNA expression at 3 dpf, but the MR morphants in low Ca 2+ medium were similar to the control group (Fig. 10).

Potential regulatory elements in the zebrafish ecac gene
The 59 flanking region of the zebrafish ecac gene was putatively identified and analyzed to search for possible regulatory elements. Several hormone-responsive elements, including the GRE, were identified within the 1653 bp analyzed in the putative upstream promoter region of the ecac gene (Fig. S2).

Exogenous cortisol on mRNA expressions of ecac and vitamin D 3 -related genes in zebrafish embryos
To investigate the effect of cortisol on the vitamin D 3 -related genes, exogenous cortisol (20 mg/l) was used to treat zebrafish  embryos. To trace the effect of exogenous cortisol, 1-and 3-dpf embryos were sampled. Cortisol caused differential effects on mRNA expressions of ecac and the vitamin D 3 -related genes in 1-dpf or 3-dpf zebrafish embryos. According to the qPCR, cortisol significantly stimulated ecac and cyp27b1 mRNA expression in 1and 3-dpf zebrafish embryos; however, cortisol only significantly stimulated the mRNA expressions of vitamin D 3 receptor a (vdra), cyp27a1l and cyp27a1in 1-or 3-dpf embryos (Fig. 11A, B). On the other hand, vdrb was not affected by cortisol in 1-and 3-dpf zebrafish embryos (Fig. 11A, B).

Effects of MR/GR MOs on mRNA expressions of the vitamin D 3 -related genes in zebrafish embryos
To further support the data of Fig. 7, 1,2-cell-stage embryos were injected with the control MO, MR MO, and GR MO, respectively. MR MO did not cause any effects on mRNA expressions of the vitamin D 3 -related genes in 1-and 3-dpf zebrafish embryos (Fig. 12A, B). On the contrary, GR MO caused differential effects on those vitamin D 3 -related genes (Fig. 12A, B). Only cyp27b1 mRNA expression was downregulated by GR MO at 1 dpf (Fig. 12A), but the mRNA expressions of vdra, cyp27a1l and cyp27a1 were all decreased by GR MO at 3 dpf (Fig. 12B).

Discussion
Glucocorticoid (cortisol) showed different impact in Ca 2+ handling between mammal and fish; however, the understanding is little of cortisol control in fish. For this purpose, we explored cortisol effect in zebrafish Ca 2+ handling. In the present study, the major findings were the following: (I) expression of zebrafish 11bhydroxylase was stimulated by low-Ca 2+ environment; (II) exogenous cortisol increased zebrafish Ca 2+ influx and content through upregulating the expression of ecac but not those of ncx1b or pmca2; (III) exogenous cortisol probably through a feedback pathway modulated the mRNA expression of 11b-hydroxylase, hsd11b2 and gr but not that of mr in zebrafish; and (IV) translational knockdown of  GR but not MR caused defects in the binding of cortisol or mRNA expression of vitamin D3-related genes, and thus impaired the ecac expression and Ca 2+ uptake function of zebrafish ECaC via the GR, a regulatory pathway that could be mediated by vitamin D 3 .
Fish mainly obtain Ca 2+ from the aquatic environment with fluctuating Ca 2+ levels, and therefore the internal Ca 2+ homeostasis is impacted by environmental Ca 2+ levels. Fish (at least teleost) bone is acellular, and thus the bone does not provide a pool for Ca 2+ as it does in terrestrial vertebrates [30]. For internal Ca 2+ homeostasis and bone formation (particularly in embryonic and larval stages), fish have to regulate the function of Ca 2+ uptake to cope with a fluctuating environment. Like other teleosts, zebrafish can enhance Ca 2+ uptake function by stimulating ECaC expression during acclimation to low-Ca 2+ fresh water [40,41,44], and this functional regulation may be associated with cortisol. Flik and Perry [36] reported that acclimation to low-Ca 2+ fresh water caused an increase in serum cortisol levels in rainbow trout. The present study further explored the mechanism behind this phenomenon. Low-Ca 2+ fresh water stimulated the mRNA expression of 11b-hydroxylase, the enzyme in the final step of cortisol synthesis, which suggests that environmental conditions affect steroidogenesis and thus cortisol levels in zebrafish. All these results imply a possible role of cortisol in the control of Ca 2+ uptake. To test this hypothesis, we treated zebrafish with exogenous cortisol. Similar to trout and eel [37,46], in zebrafish, exogenous cortisol stimulated the mRNA expression of ecac and Ca 2+ influx, and these functional enhancements resulted in increased Ca 2+ contents in the whole body (the present study). The present comprehensive data from molecular to the physiological level demonstrated the calciotropic effects of cortisol, and these effects showed a dose-dependent pattern and were of physiological significance.
The ECaC, NCX1b, and PMCA2 are coexpressed in a specific type of ionocyte [41], which achieves the epithelial Ca 2+ uptake function through the operations of the 3 transporters in zebrafish [32,33,39]. Previously, exogenous cortisol was found to stimulate branchial ECaC mRNA and protein expressions in trout [37], but  no attempt was made to examine the effects on the other Ca 2+ transporters (NCX and PMCA). The present study first reports that exogenous cortisol affected ecac only but not those of ncx1b or pmca2; similar result found in cultured gills with exogenous cortisol (Fig. S1). On the other hand, we further directly demonstrated ecac-expressing cell density in skin of zebrafish embryos was stimulated by exogenous cortisol. These results supported the previous notion that the ECaC is the rate-limiting step and the gatekeeper channel for active Ca 2+ transport in fish [39] as in mammals [47]. On the other hand, translational knockdown of stanniocalcin, a hypocalcemic hormone, was reported to stimulate ecac expression and Ca 2+ influx, but not affect the expressions of ncx1b or pmca2 in zebrafish [42]. Similarly in the present study, knockdown of the GR downregulated the expression of ecac, but showed no effect on ncx1b or pmca2. The ecac appears to be the major regulatory target gene in response to environmental Ca 2+ levels and also the upstream control of hormones in zebrafish.
In the present study, exogenous cortisol suppressed the mRNA expression of 11b-hydroxylase (decreasing the cortisol level) and simultaneously stimulated that of hsd11b2 (increasing the cortisone level), reflecting a feedback mechanism in controlling levels of corticoids. In mammals, HSD11B2 converts cortisol to cortisone in MR-specific tissues, and this prevents cortisol binding to the MR in these tissues and thus allows aldosterone to bind to the MR [48]. In fish, HSD11B2 can also convert cortisol to cortisone [49,50]. Thus, stimulation of hsd11b2 by exogenous cortisol treatment may be a feedback to control cortisol levels in zebrafish. This notion was further supported by the data of 11b-hydroxylase and the gr; gene expressions of zebrafish 11b-hydroxylase and the gr were inhibited by exogenous cortisol treatment. In mammals, dexamethasone treatment suppressed the secretion of the adrenocorticotrophic hormone, resulting in a decline in serum corticosterone levels [51,52], and the duodenal and renal gr expressions were significantly downregulated by dexamethasone treatment [6]. Similarly in trout and salmon, cortisol treatment by infusion, feeding, or soaking also caused downregulation of CS receptors or the gr in gills and liver [53][54][55][56][57]. Taken together, cortisol may regulate the function of the Ca 2+ mechanism through sophisticated feedback pathways, in which the expressions of 11bhydroxylase, hsd11b2, and CRs are differentially modulated.
Physiological functions of cortisol signaling are mediated by the GR and MR, which are ligand-activated transcription factors. In many previous studies, exogenous cortisol treatment was found to stimulate Ca 2+ uptake in teleosts [36,38,46]; however, it was unknown until the present study that cortisol controls the Ca 2+ uptake function through the GR but not the MR. In addition to experiments of exogenous cortisol treatment, we used a gene-specific MO to abolish endogenous cortisol signaling in zebrafish and directly explored the effect of cortisolsignaling defects on the Ca 2+ uptake function. Translational knockdown of the MR did not affect Ca 2+ uptake in zebrafish, but GR knockdown evidently impaired the Ca 2+ uptake mechanism by decreasing both Ca 2+ influx and content. Moreover, these defects in the Ca 2+ uptake mechanism were due to suppression of ecac expression but not the expressions of ncx1b or pmca2. GR defect was also demonstrated to downregulate ecac -expressing cell density in skin of zebrafish. These results indicated that the target of the GR is the ecac. Interestingly, the Ca 2+ influx, ecac mRNA and ecac -expressing cell density in MR MOinjected zebrafish morphants could still be stimulated (compared to the control) by exogenous cortisol, suggesting that the effects of exogenous cortisol on Ca 2+ uptake function is not through the MR. On the other hand, exogenous cortisol did not cause further stimulation (compared to the control) in Ca 2+ influx or ecac mRNA in GR morphants, but successfully rescued the Ca 2+ uptake functional defects caused by the GR MO. Similar effect was also found in GR MO-injected zebrafish with low Ca 2+ freshwater. GR MO morphant can abolish stimulation of low Ca 2+ freshwater on ecac mRNA expression, but MR MO morphant can not. Reinforcing these results, overexpression of the GR by injection with GR cRNA rescued the Ca 2+ uptake mechanism that was impaired in GR MO morphants. This evidence strongly suggests that endogenous cortisol stimulates Ca 2+ uptake through the GR, but not the MR, in zebrafish.
Some previous in vitro studies indicated that the trout MR and GR could bind cortisol and stimulate transcriptional activity in the mammalian cell lines transfected with a reporter plasmid [12][13][14][15]. In a recent study on Atlantic salmon, Kiilerich et al. [16] used GR and MR antagonists to discover that the 2 CS receptors were involved in regulating various ion transporters (NKA, NKCC, and CFTR) during acclimation to salinity changes. Differences between salmon (involvement of both the GR and MR) and zebrafish (only the GR) may be because of differences in the ion transport functions (Na + /Cl 2 vs. Ca 2+ ) and species (euryhaline vs. stenohaline), and clarification of this point requires further studies. On the other hand, it was noted that zebrafish MR expression was not regulated by exogenous cortisol in the present study. Teleosts might not synthesize aldosterone, and DOC was proposed to play a similar role as aldosterone [15]. Moreover, DOC was suggested to be a potent agonist of the fish MR because DOC can induce greater transcription activity than cortisol through the trout MR expressed in a mammalian cell line co-transfected with a reporter plasmid [15]. It will be challenging to see if a lack of an effect of exogenous cortisol on MR expression is due to different ligand affinities between the 2 CS receptors in zebrafish.
Cortisol shows hypercalcemic effects in zebrafish and other fish species as described above. On the contrary, cortisol was reported to inhibit the intestinal Ca 2+ absorption in chickens [58], and GC drugs, dexamethasone and prednisolone, were demonstrated to inhibit duodenal Ca 2+ uptake and trpv6 (ecac) gene expression in mice [5,6]. Recently, Kim et al. [6] found only one GRE in the promoter region of mice trpv6, and suggested that dexamethasone might not directly regulate trpv6 transcription, but instead, downregulates the Ca 2+ uptake function through other hormones. Based on a bioinformatics analysis, there are some putative GREs in the promoter region of the zebrafish ecac (Fig. S2). Previous in situ hybridization study on developing zebrafish embryos [59], gr transcript was detected over the skin where ionocytes appear. As shown in Fig. S3, GR was colocalized in NaR cells, which express ECaC [40]. These imply the possibility that cortisol can stimulate mRNA expression of the ecac through interaction with the GREs; however, further studies are needed to support this notion.
In addition to direct regulation of ecac transcript, cortisol may modulate the other endocrine to affect Ca 2+ uptake in zebrafish. Vitamin D 3 is a well-known calcitrophic endocrine to regulate Ca 2+ homeostasis in vertebrates [19][20][21], and several mammal studies demonstrated that glucocorticoid could affect vitamin D 3 metabolism [24][25][26][27][28][29]. Vitamin D 3 can bind VDR, forming a vitamin D3-VDR complex. This complex directly stimulates mammalian intestinal ecac transcript by binding vitamin D 3 responsive element (VDRE) in the promoter region of ecac [18]. In fish, vitamin D 3 was also reported to elevate the serum Ca 2+ level [20,21], and putative VDRE was also identified in ecac promoter region [60] (Fig. S2). Recent in vitro study in medaka, VDR was found to simulate transcript level of VDRE-containing construct with 1,25(OH) 2 D 3 treatment [61]. Our unpublished data also indicated stimulation of ecac mRNA expression by1,25(OH) 2 D 3 in zebrafish embryos. Taken together, vitamin D3-VDR control of ecac expression and function may also exist in fish. However, the associations between cortisol, vitamin D 3 and Ca 2+ handling in fish were not clear until the present study. CYP27A1 is an enzyme to synthesize 25(OH)D 3 , the vitamin D 3 precusor, and subsequently 25(OH)D 3 is converted to 1,25(OH) 2 D 3 (ctive vitamin D 3 ) by CYP27B1 [17]. In zebrafish embryos, exogenous cortisol could upregulate the expressions of cyp27b1, cyp27a1 and/or cyp27a1 like (cyp27a1l) at 1-and 3 dpf, and also affected vdra mRNA expression at 1 dpf. These implied that cortisol starts affecting the vitamin D 3 -VDR signaling from early stage. Cortisol could not only stimulate synthesis of vitamin D 3 precursors but also accelerate the synthesis of active vitamin D 3 . Moreover, our subsequent knockdown experiment further reinforced this notion. Knockdown of GR, but not MR, was found to suppress the expressions of cyp27b1, cyp27a1, cyp27a1l and vdra at 1and/or 3 dpf. According to these results, we suggested that cortisol probably regulates Ca 2+ handling through vitamin D 3 -VDR system, besides directly regulating ecac expression.
In this study, cortisol was suggested to directly or indirectly stimulate Ca 2+ uptake in zebrafish. Cortisol appears to cause different effects on Ca 2+ uptake between teleosts and higher vertebrates. Sources of Ca 2+ fundamentally differ among verte- brates. For terrestrial vertebrates, food is the major source of Ca 2+ , and the intestines serve as a primary site for Ca 2+ uptake. However, fish are constantly facing aquatic environments with variable Ca 2+ concentrations (of as low as ,0.01 mM in soft fresh water), and gills or the skin serve as the primary site for Ca 2+ uptake [30,32,33]. These differences may be associated with different modes of cortisol control of Ca 2+ uptake, which is a challenging issue to be explored from an evolutionary point of view.