Introduction

Freshwater (FW) teleosts maintain their body fluids hypertonic to the surrounding medium. Ionic homeostasis is achieved through constant absorption of ions from the environment via specialized epithelial cells termed mitochondrion-rich cells (MRCs) or ionocytes, predominantly distributed on the adult gill or larval yolk skin (for recent reviews see [12, 15, 16, 18, 19, 26]). Over the past 80 years, numerous regulatory agents, including cortisol [24, 29, 34], prolactin [6, 7, 41], isotocin [11], angiotensin [22], vitamin D [33], and catecholamines [27], have been implicated in stimulating ionic uptake in several species of FW fish. However, to maintain ionic homeostasis, teleosts must also possess physiological mechanisms to reduce rates of ion uptake. For Ca2+ uptake, both stanniocalcin [46] and calcitonin [28] were shown to act as potent hypocalcemic hormones in developing zebrafish, and another study [27] hinted at the possibility that catecholamines, when acting through α-adrenergic receptors, could inhibit Na+ uptake in zebrafish. While these previous studies indicate the presence of mechanisms to reduce ion uptake in FW fish, it is likely that there are other inhibitory mechanisms that have not yet been discovered.

Several gaseous molecules, including nitric oxide and carbon monoxide, are recognized as important mediators of physiological signals and have been termed gasotransmitters or gasometers [48]. Following the discovery in vertebrate tissues of two key hydrogen sulfide (H2S) synthesizing enzymes, cystathionine β-synthase (CBS) and cystathionine γ-lyase (CSE) [8, 14, 45], H2S has become recognized as a third gasotransmitter [36, 48]. Although H2S is known to play an important role in cardiorespiratory control/oxygen sensing in vertebrates including fish [37–39], it has been recently implicated in the regulation of transepithelial ion transport. For example, treatment with NaSH (a H2S donor) of cell cultures derived from human bronchial epithelial cells (H441) and native trachea derived from pig and mouse inhibited Na+ uptake by those cells [1, 2]. With this background, the aim of the present study was, through pharmacological treatments, gene knockdown, and immunohistochemistry, to demonstrate an inhibitory role for H2S on Na+ uptake in developing zebrafish. The results of the current study introduce a novel mechanism for Na+ homeostasis in FW fish.

Materials and methods

Experimental animals and water preparation

Zebrafish (Danio rerio Hamilton-Buchanan 1822) were purchased from Noah’s Pet Ark (Vancouver) or Big Al’s Aquarium Services (Ottawa) and kept in dechlorinated city of Vancouver tap water (FWV) or in dechloraminated city of Ottawa tap water (FWO). See Table 1 for the ionic composition of Vancouver and Ottawa tap water. At both facilities, water temperature was kept at 28 °C. Fish were subjected to constant 14-h L/10-h D photoperiod and fed daily until satiation. Embryos were collected and reared until 4 days post-fertilization (dpf) in 50-ml Petri dish kept in the same room as adult fish and raised in FWV or FWO waters supplemented with 0.05 % methylene blue. All protocols were approved by the Animal Care Committee at University of British Columbia and University of Ottawa (Protocol BL-226).

Table 1 The ionic composition of Vancouver and Ottawa tap water
Full size table

To determine the potential role of H2S on Na+ uptake in developing zebrafish, the following series of experiments were performed. Unless stated otherwise, all chemicals used for the experiments were purchased from Sigma.

Pharmacological agents

Two chemicals known to generate H2S, Na2S, and GYY4137 (VIVA Bioscience, VB3200, UK) were used to test the effects of exogenous H2S on Na+ uptake. Endogenous production of H2S was inhibited by treating the fish with either aminooxyacetic acid (AOA, an inhibitor of CBS but also see [3] and discussion below) or propargylglycine (PPG, an inhibitor of CSE).

Series 1. Effects of H2S on Na+ uptake

This experiment was performed in FWV. To test whether H2S could regulate uptake of Na+, 4 dpf larvae were exposed to 5 or 10 μM Na2S (pH of these solutions ranged between 7.69 and 7.87 where control water pH was 7.0). Na+ uptake was measured using 22Na as described previously [27]. The 2-h Na+ uptake flux was started shortly (<5 min) after the addition of Na2S. Unless stated otherwise, the uptake measurement was performed in the continued presence of chemical treatment.

Although Na2S has been used previously in physiological studies to produce equimolar quantities of H2S rapidly, the response is markedly transient with peak levels being achieved within minutes and then typically falling back to control levels within 1 h [30]. Thus, we used a second donor of H2S, GYY-4137, which is known to yield lower but constant levels of H2S for up to 7 days [30]. Briefly, larvae were raised in ion-depleted FWO (tenfold dilution of Ottawa tap water), and at 4 dpf, larvae were exposed to 100, 200, and 500 μM GYY-4137 for 2 h before starting the 2-h Na+ uptake measurement as described above, in the continued presence of the inhibitor. Because the addition of up to 500 μM GYY-4137 did not significantly alter water pH (pH remained between 7.1 and 7.2), no pH-matched control experiment was performed.

Series 2. Effect of H2S synthesis inhibitors on Na+ uptake

Pharmacological inhibition

Because FWV is extremely dilute (the [Na+] is even lower than the affinity constant (KM) for Na+ uptake in adult zebrafish [5]), it was predicted that Na+ uptake is under tonic stimulation even in “controlˮ Vancouver conditions. Thus, to test for the physiological role of H2S as an inhibitor of Na+ uptake, larvae were exposed to high-[Na+] media (1 mM [Na+] by adding NaCl to FWV; hereafter referred to as FWV + Na). Specifically, Na+ uptake was measured in the following three groups: (1) fish raised in FWV until 4 dpf (controls), (2) fish raised in FWV until 3 dpf and then exposed to FWV + Na for 24 h, (3) fish raised as in group 2 but exposed to FWV + Na in the presence of either 100 μM AOA (an inhibitor of CBS) but also see [3] or 100 μM PPG (an inhibitor of CSE). At 4 dpf, larvae from groups 2 and 3 were transferred back to the control FWV (without inhibitors), and their Na+ uptake was measured immediately after the transfer as described above. While the addition of 100 μM PPG did not significantly affect water pH (7.5–7.6), the addition of AOA to a final concentration of 100 μM caused a mild acidification of the water (pH 6.7–6.9). To control for the potential effect of the lowered pH on Na+ uptake, we also performed an experiment where Na+ uptake was measured following 24-h exposure to water adjusted to a pH of 6.2 (pH was lowered by adding H2SO4).

Gene knockdown

This experiment was performed in Ottawa. Because the ionic composition of FWO is significantly higher than FWV (see Table 1; [Na+] is almost five times higher than the KM for Na+ uptake by adult zebrafish in FWO), the fish were not subjected to a “high-salt” pretreatment; all larvae were raised in FWO until 4 dpf. To implicate H2S in Na+ homeostasis, CSE was knocked down by microinjecting a fluorescein isothiocyanate (FITC)-conjugated antisense morpholino against zebrafish CSE (5′-CGCACAAGAGTGAACAGCTCTCTGT-3′) (Gene Tools, LCC). Briefly, 1–2 cell stage embryos were injected with the morpholino (MO) at a dose of 4 ng/embryo. Another group of embryos (“sham”) were injected with control MO (5′-CCTCTTACCTCAGTTACAATTTATA-3′; also tagged with FITC and prepared as CSE MO). At 1 dpf, injected larvae were screened for the widespread presence of fluorescence, and only FITC-positive embryos were raised in FWO until 4 dpf. At 4 dpf, rates of Na+ uptake were determined in the CSE-morphants and sham-injected larvae in FWO as described above. No developmental abnormality or higher mortality rate was observed in the CSE-morphants.

The effectiveness of the CSE knockdown was confirmed by western blotting using a commercial polyclonal antibody raised in rabbit against human CSE (ARP46067_P050; Aviva Systems Biology, San Diego, USA) whose epitope shared 100 % identity with zebrafish CSE (accession number AAH67624.1). Total protein was extracted from MO- and sham-injected larvae using Tris buffer (10 mM Tris-HCl with 2 % Triton X-100; pH adjusted to 7.4) supplemented with protease inhibitor tablet (Complete Mini, Roche). Ten larvae were pooled to prepare one (N = 1) sample. Extracted samples were loaded onto a 10 % SDS-PAGE, size-fractionated at 200 V, and transferred onto a polyvinylidene difluoride (PVDF) membrane (Bio-Rad). After transfer, membranes were blocked with 5 % bovine serum albumin (BSA) in 0.2 % Tween 20 in Tris-buffered saline (TBST) for 2 h at room temperature. Membranes were incubated overnight with gentle shaking with anti-CSE antibody (1:750 in 2 % BSA in TBST) at 4 °C. Subsequently, membranes were washed (3 × 5 min) with TBST and incubated with horseradish peroxidase-conjugated secondary antibody against rabbit IgG (Invitrogen; 1:15,000 in 2 % BSA in TBST) for 2 h at room temperature. The membranes were then washed (4 × 10 min), and the immunoreactive bands were detected using enhanced chemiluminescence (Millipore) with a ChemiDoc system (Bio-Rad). Subsequently, the membrane was reprobed with β-actin antibody (1:4,000; Sigma) after stripping with Re-Blot Plus solution (Millipore).

Series 3. Elucidation of potential Na+ transporters regulated by H2S

Effect of H2S treatment on acid secretion

To help identify the transporter(s) inhibited by H2S treatment, whole-body net acid excretion (JNETH+) following Na2S treatment was assessed using 4-dpf zebrafish larvae chronically raised in FWV. Briefly, 15 4-dpf larvae (to generate N = 1) were placed in 4.5 ml of FWV in a 15-ml tube. Water samples (1.5 ml) were collected at the beginning and end of a 1.5-h flux period and stored at 4 °C until analysis (all samples were analyzed within 24 h of collection). To determine titratable alkalinity flux (JnetTA), 1.2 ml of the initial and final water samples from each flux period was titrated to pH 4.00 using 0.001 M HCl; the difference in titrant volume added between the samples was noted. Acid was delivered with a Gilmont microburette (GS-1200 GE 2.0 ml; Fisher), and pH was recorded (pHC3005-8 electrode and PHM201 pH meter; Radiometer Analytical) throughout the titration. Samples were bubbled with N2 continuously to ensure mixing. Ammonia excretion (JNETAMM) was determined by measuring water total ammonia levels at the start and end of the flux period. Water ammonia concentrations were determined colorimetrically [47], with slight modifications for microplate reading. JNETH+ was calculated as the sum of JnetTA and JnetAMM, signs considered [35]. After the first measurement, fluid volume was readjusted to 4.5 ml. Na2S was added to a final concentration of 10 μM, and another whole-body acid-flux measurement was started immediately, in the presence of Na2S in water.

To confirm that zebrafish larvae reared in FWV absorb Na+ via H+-secretion coupled mechanism, 4 dpf zebrafish larvae were treated with 50 and 100 μM 5-(N-ethyl-N-isopropyl amiloride (EIPA), which effectively inhibits zebrafish Na+-H+-exchanger 3b (NHE3b) expressed in oocytes [20]. A stock solution of EIPA was prepared using DMSO as vehicle, and the control group was exposed to DMSO alone (the concentration did not exceed 0.1 %). Na+ uptake measurements were started immediately after exposure to EIPA to the medium as described above, in the continued presence of the inhibitor.

Effect of H2S treatment on acid-exposed zebrafish larvae

We recently showed that exposing zebrafish larvae to acidified (pH 4.0) FWO for 24 h increase Na+ uptake and that this stimulation is largely attributable to activation of NHE3 [25, 24]. To assess the potential regulation of Na+ uptake via NHE3 by H2S, zebrafish larvae were raised in control Ottawa tap water until 3 dpf. Subsequently, a subset of larvae was exposed to acidified FWO (pH 4.0; prepared by adding H2SO4 to FWO). After acid exposure, larvae were transferred back to FWO and exposed to 10 μM Na2S. Immediately after the addition of Na2S, the measurement of Na+ uptake was started as described above.

Effect of H2S treatment on glial cell missing 2 (gcm2) morphants

The transcription factor, gcm2, plays a critical role in the differentiation of H+-ATPase-rich cells (HRCs), and its knockdown prevents or severely restricts HRC formation [9, 10, 43]. The loss of HRCs is accompanied by a compensatory increase in the numbers of Na+-Cl− co-transporter (NCC; zslc12a10.2)-expressing cells [9, 10, 43]. To test whether NCC is regulated by H2S, larvae were injected with a MO targeting gcm2 or a control MO [24]. The larvae were raised in FWO until 4 dpf, and Na+ uptake was measured in response to treatment with Na2S (final concentration 5 or 10 μM) as described above. The effectiveness of gcm2 knockdown was confirmed by staining sham and gcm2 MO larvae with concanavalin A (conA; 5 μg/ml), a specific vital dye for HR cells, following a standard protocol [27]. Images were acquired using a confocal laser scanning microscope (A1R+; Nikon Instruments, USA). For each individual larva, three regions of interest (10,000 μm2 each) were randomly selected, and the number of conA positive cells was determined. The average of those three counts (regarded as N = 1) was used in the statistical analysis.

To confirm that the loss of gcm2 indeed increased expression of NCC, we analyzed the messenger RNA (mRNA) expression level of NCC in 4-dpf sham and gcm2 MO larvae (chronically raised in FWO). Total RNA was extracted using Trizol (Invitrogen) according to the manufacturer’s instructions. cDNA was synthesized by treating 1 μg of extracted RNA with DNase (Invitrogen) and RevertAid M-MNuLV reverse transcriptase (Fermentas, Burlington, ON, Canada) according to the manufacturer’s instructions. Real-time PCR was performed using a Bio-Rad CFX96 qPCR system with Brilliant III SYBR Green Master Mix (Agilent Technologies, La Jolla, CA, USA). PCR conditions were as follows: 95 °C for 3 min, 40 cycles of 95 °C for 20 s, and 58 °C for 20 s, with final extension for 5 min at 72 °C. Data were normalized to the expression of 18S. For the primer pairs, see Table 2.

Table 2 Primer sequence for NCC and 18S
Full size table

Series 4. Distribution of CSE in zebrafish larvae

To determine whether CSE is expressed in ionocytes, 4 dpf larvae were vitally stained as above with Alexa-633 conjugated conA for 30 min at room temperature (RT). Subsequently, larvae were killed by overdose with MS-222 and fixed using a 4 % paraformaldehyde solution prepared in PBS for 2 h at RT. After rinsing with PBS, larvae were subjected to a heat-induced antigen retrieval protocol (15 min heating at 70 °C in 150 mM Tris solution, pH 9.0) and blocked with 10 % natural donkey goat serum in 5 % PBS-Triton-X (PBST) for 2 h at RT. Larvae were stained with the same anti-CSE antibody used for the western blotting and co-stained with an anti-Na+-K+-ATPase antibody (α5, Developmental Studies Hybridoma Bank) in 5 % PBST (1:200 dilution for both antibodies overnight at RT). Subsequently, larvae were rinsed with PBS and then incubated with secondary antibodies (Alexa-488 conjugated anti-rabbit IgG and Alexa-596 conjugated anti-mouse IgG; Invitrogen; both used at 1:400) in PBS for 2 h at RT. After rinsing in PBS, samples were mounted onto microscope slides and observed with an Olympus FV1000 BX61 confocal microscope with Olympus FluoView software.

The specificity of antibody staining was tested in two ways: (1) preabsorption of antibody with peptide and (2) immunostaining of CSE MOs. For preabsorption, larvae were immunostained as described above (but without any vital staining). The antibody was mixed with blocking peptide (catalog # AAP46067, Aviva Systems Biology) for 2 h at RT at 1:10 antibody/peptide ratio. Representative images were selected from four larvae (both for control and preabsorption groups). CSE morphants were prepared as described in Series 2.2 and raised until 4 dpf in FWO media and immunostained with CSE antibody as described above. Representative images were selected from 12 sham and morphants (in two separate trials, each consisting of six larvae).

Analytical methods and calculations

To determine Na+ uptake, all collected water samples were supplemented with 5 ml of scintillation cocktail (Biosafe-II, RPI corp., Mt. Prospect, IL, USA), and their radioactivity was measured with a liquid scintillation counter (model LS-6500 Beckman Coulter, Co. Mississauga ON, Canada). After being rinsed in an isotope-free medium, larvae were digested in a tissue solubilizer (SolvableTM, PerkinElmer) overnight at 65 °C. After complete digestion, samples were supplemented with 5 ml of the same scintillation cocktail. Samples were then neutralized by adding 400 μl of glacial acetic acid before measuring their radioactivity. The concentration of total Na+ in the water was measured using flame emission spectrophotometry (model AA260, Varian, Palo Alto, CA, USA). Owing to the limited volume of water, [Na+] was measured, and hence external specific activity was determined, only at the end of the flux period. It was assumed that, given the typical [Na+] of FWO (700–1000 μM), a low Na+ influx rate (on the scale of 1 nmol/fish/h) and an even smaller expected net flux of Na+ (difference between influx and efflux), changes in total [Na+] during the flux period would be negligible. The rate of Na+ uptake (J Na in , pmol/fish/h) was calculated as follows:

$$ {J}_{in}^{Na}=\frac{F}{ SA\cdot n\cdot t} $$

where F is the total incorporated radioactivity (DPM, disintegration per minute), SA is the specific activity of the medium (DPM/pmol), n is the number of larvae per digest, and t is the duration of the incubation (h). DPM was calculated by the liquid scintillation counting program after taking quenching and counting efficiency into consideration.

Statistical analysis

All statistical analyses were performed, and data were graphed with SigmaPlot (v. 11, Systat Inc. Chicago, IL, USA). Data presented in Figs. 1, 2, 4a, and 5e were analyzed with one-way ANOVA followed by Tukey’s post hoc test. Other data were analyzed with Student’s t test. When normality or equal variance assumptions were violated, data were transformed by calculating log or square root. In all statistical analysis, the significance level was set at p < 0.05.

Fig. 1
figure 1

Effect of Na2S treatment on Na+ uptake (Series 1). Acute (up to 4 h) treatment of zebrafish larvae at 4 days post-fertilization (dpf) with Na2S (a) and GYY4173 (b) led to a dose-dependent inhibition of Na+ uptake (N = 5–8; one-way ANOVA). Data are presented as means ± SEM; bars not sharing the same letter are significantly different from one another

Full size image
Fig. 2
figure 2

Effect of CBS/CSE inhibition during exposure to FWV + Na (Series 2.1). Preexposing zebrafish larvae to FWV + Na (1 mM NaCl solution added to Vancouver tap water) for 24 h between 3 and 4 dpf significantly reduced Na+ uptake capacity when returned to FWV where [Na+] = 80 μM (N = 14, one-way ANOVA; the measurement was carried out using 4 dpf larvae). This inhibition in Na+ uptake was prevented when larvae were exposed to high saltwater with an inhibitor for CBS (AOA; N = 6–8; one-way ANOVA) or CSE (PPG, N = 6–8; one-way ANOVA). Data are presented as means ± SEM; bars not sharing the same letter are significantly different from one another

Full size image

Results

Series 1. Effect of waterborne H2S on Na+ uptake

Treatment with two different H2S donors, Na2S (Fig. 1a; N = 5–8; one-way ANOVA) and GYY-4137 (Fig. 1b; N = 6; one-way ANOVA), reduced Na+ uptake by zebrafish larvae in an apparent concentration-dependent manner. At concentrations of 5 and 10 μM Na2S, Na+ uptake was reduced by 25 and 75 %, respectively (Fig. 1a), whereas treatment with 500 μM GYY-4137 reduced Na+ by about 40 % (Fig. 1b).

Series 2.1. Effect of CSE/CBS inhibition during high-salt exposure

Exposure to FWV + Na between 72 and 96 hpf reduced the capacity of larvae to absorb Na+ when they were transferred back to FWV. The reduction in Na+ uptake in fish preexposed to FWV + Na was prevented in the presence of commonly used CBS and CSE inhibitors, AOA and PPG, respectively (Fig. 2; N = 6–8; one-way ANOVA). Indeed, the capacity to absorb Na+ in fish treated with AOA was even significantly higher than in the unhandled control group. When pH-control experiments (3 dpf larvae were exposed to pH 6.2 medium for 24 h) were performed, Na+ uptake was slightly reduced (251 ± 30 vs 178 ± 13 pmol/fish/h, N = 6), indicating that the observed increase in Na+ uptake was not caused by changes in water pH.

Series 2.2. Effect of CSE knockdown on Na+ uptake

Western blotting with a commercial anti-CSE antibody yielded multiple bands (Fig. 3a). Following the knockdown of CSE, intensity of the band that corresponded to the predicted molecular weight of CSE (~45 kDa) was greatly diminished (Fig. 3a shows a representative western blot image; N = 4 for Fig. 3b). In agreement with the proposed inhibitory role of H2S on Na+ uptake as well as the previously observed effects of pharmacological inhibition of H2S-synthesizing enzymes (Figs. 1 and 2), Na+ uptake was significantly elevated in CSE morphants (Fig. 3c; N = 6; Student’s t test).

Fig. 3
figure 3

Effect of CSE knockdown on Na+ uptake (Series 2.2). Microinjection of CSE morpholino successfully reduced the protein expression level by ~98 % (b; representative blots are shown in a). Whole-body protein samples for Western blotting were extracted from 4 dpf larvae. Na+ uptake (measured using 4 dpf larvae) was significantly higher in CSE morphants compared to sham-injected larvae (c; N = 6; Student’s t test, performed in FWO). Data are presented as means ± SEM, and an asterisk denotes significant difference between shams and morphants

Full size image

Series 3. Elucidation of the Na+ uptake mechanism regulated by H2S

Treating 4 dpf zebrafish larvae raised in FWV with 50 or 100 μM EIPA significantly reduced Na+ uptake by about 50 % (Fig. 4a; N = 5–6; one-way ANOVA). Treatment of 4 dpf larvae (raised in FWV) with 10 μM Na2S significantly reduced whole-body net acid secretion (Fig. 4b; N = 7; paired t test), although two components of acid secretion, namely, titratable alkalinity flux (Fig. 4c) and ammonia excretion (Fig. 4d), were not significantly affected when analyzed individually. When zebrafish larvae preexposed to acidified FWO between 3 and 4 dpf and then treated with waterborne Na2S (in acidified FWO), Na+ uptake was strongly inhibited (Fig. 4e, N = 6; one-way ANOVA).

Fig. 4
figure 4

Analysis of the Na+ uptake mechanism regulated by H2S (Series 3.1 and 3.2). Treating 4 days post-fertilization (dpf) zebrafish larvae in FWV with 50 or 100 μM EIPA significantly reduced Na+ uptake (a; N = 5–6; one-way ANOVA). Treating zebrafish larvae at 4 dpf with Na2S significantly reduced whole-body acid secretion (b; N = 7; paired t test) but did not affect whole-body titratable alkalinity (TA) flux (c; N = 7; paired t test), or ammonia secretion (d; N = 7; paired t test. Na2S treatment was highly effective in inhibiting Na+ uptake in acid-exposed larvae (e; N = 6; one-way ANOVA; performed in FWO). Data are presented as means ± SEM, and an asterisk in b indicates significant difference between control and Na2S-treated groups. Bars labeled with different letters in a and e indicate significant differences

Full size image

As reported previously, knockdown of gcm2 led to near-complete loss of conA positive HR cells on the yolk sack (Fig. 5a; N = 5; Student’s t test). Representative images of sham and gcm2 MO used for quantifications are shown in Fig. 5b, c, respectively. On the other hand, the mRNA expression level of NCC was significantly elevated in gcm2 MO (Fig. 5d; N = 5; Student’s t test). Exogenous treatment of 4 dpf gcm2 morphants (chronically raised in FWO) with Na2S had no effect on Na+ uptake, suggesting that NCC is not regulated by H2S (Fig. 5e; N = 5–6; one-way ANOVA).

Fig. 5
figure 5

The effect of H2S treatment on gcm2 morphants (Series 3.3). The loss of gcm2 led to significant reduction in conA positive HR cell density on yolk of 4 dpf zebrafish larvae (a; N = 5; Student’s t test). Representative images for sham and MO are shown in b and c. In agreement with previous publications, knockdown of gcm2 led to significant increase in NCC mRNA expression (d; N = 5; Student’s t test). Despite the elevated Na+ uptake rate seen in gcm2 morphants raised in FWO, treatment with up to 10 μM Na2S did not have any effect on Na+ uptake by morphants (e; N = 5–6; one-way ANOVA). Data are presented as means ± SEM, and an asterisk in a and d indicates significant difference between control and Na2S-treated groups. Bars labeled with different letters in e indicate significant difference from one another

Full size image

Series 4. Localization of CSE in developing zebrafish larvae

Preabsorption of CSE antibody with blocking peptide or knockdown of CSE greatly reduced immunostaining of CSE on the yolk sack epithelium (Fig. 6). Whole-mount immunohistochemistry revealed numerous CSE-positive cells on the yolk sack of larvae. CSE was co-localized with conA (a vital dye used to identify HRCs) as well as with Na+/K+-ATPase (a marker for NaR cells), suggesting that CSE is highly expressed in at least two types of ionocytes (Fig. 7).

Fig. 6
figure 6

Confirmation of CSE immunostaining in zebrafish larvae (Series 4). Immuohistochemistry of 4 dpf zebrafish larvae with a commercially available CSE antibody revealed numerous positively stained cells on the yolk epithelia (a). Preabsorption of CSE antibody with blocking peptide eliminated the staining (b). Similarly, the number of stained cells was greatly diminished in CSE MO (d), in comparison to sham-injected larvae (c; immunostaining was performed without antibody preabsorption). Positive CSE stains are indicated with asterisks. Scale bars = 100 μm

Full size image
Fig. 7
figure 7

Distribution of CSE in developing zebrafish larvae (Series 4). Larvae at 4 days post-fertilization (4 dpf) were stained with anti-CSE antibody (a, e), conA (an HRC marker, b, f), and anti-NaK antibody (a NaRC marker, c, g). A merged image showed high expression of CSE in both HRCs and NaRCs (d, h). Scale bars indicate 50 μm for a–d; 10 μm for e–h

Full size image

Discussion

While previous studies have focused on the physiological mechanisms that contribute to stimulating salt uptake (see “Introductionˮ), ionic homeostasis presumably must also rely on mechanisms that reduce ion uptake capacity. Indeed, for Ca2+, stanniocalcin and calcitonin are recognized as potent hypocalcemic hormones in fish [28, 46]. The present study was designed to test the hypothesis that Na+ uptake by zebrafish larvae is inhibited by endogenously synthesized H2S. As a potential inhibitor of Na+ uptake, we reasoned that the involvement of endogenous H2S would likely be lowest when Na+ transport capacity was elevated and presumably highest when Na+ transport capacity was diminished. Thus, the experimental design involved raising zebrafish embryos either in Na+-poor Vancouver tap water (FWV) where Na+ transporters were expected to be in an activated state or in Na+-enriched water (either Vancouver tap water supplemented with 1 mM NaCl; FWV + Na or regular Ottawa tap water; FWO) where Na+ transporters were expected to be less active. We reasoned that it would be most effective to assess the inhibitory effects of exogenous H2S on Na+ uptake in FWV but that the involvement of endogenous H2S would be best evaluated in Na+-enriched water. Indeed, the results revealed a significant role for endogenous H2S as a tonic inhibitor of Na+ uptake in fish exposed to FWV + Na or to FWO.

To first establish the potential for H2S to act as an inhibitor of Na+ uptake, zebrafish larvae were treated with two chemicals capable of generating H2S: Na2S and GYY-4137. Na2S is known to raise H2S levels to peak values within about 15 min after which the concentration may decline rapidly [30]. GYY-4137, however, continuously releases H2S for up to 7 days [30] leading to stable values. Treatment with either chemical caused significant and similar reductions in Na+ uptake despite the widely different levels used (up to 10 μM Na2S versus 500 μM GYY-4137). However, based on [30], 500 μM GYY4137 is expected to produce stable levels of H2S of about 20–25 μM. Thus, the actual concentration of waterborne H2S that zebrafish larvae were exposed to was likely similar regardless of whether Na2S or GYY-4137 was used. The observation that both Na2S and GYY-4137 treatment reduced Na+ uptake in zebrafish larvae strengthens the putative role played by H2S in osmoregulation. Furthermore, when 4 dpf zebrafish larvae raised in ion-poor water were treated with 200 and 500 μM pyridoxalphosphate (PLP; a substrate for H2S synthesis by CSE), such treatment significantly reduced Na+ uptake by 30 and 80 %, respectively (data not shown). On the other hand, when larvae were treated with a similar dose of L-cysteine (up to 500 μM), Na+ uptake was unaffected. The lack of an effect of L-cysteine may reflect its low rate of uptake via skin (e.g., [13]). Unfortunately, owing to the small size of zebrafish larvae, we were unable to measure the whole body content of H2S following treatment with Na2S, GYY-4137, or PLP. While this technical limitation somewhat confounds the interpretation of Fig. 1, the conclusion that H2S is a physiologically relevant molecule regulating Na+ balance is strengthened by the observation that inhibition of either of two H2S-synthesizing enzymes, CBS or CSE, caused an increase in Na+ uptake in fish exposed to FWO or FWV + Na.

Owing to the difference in water chemistry between FWV and FWO, two different protocols were used to investigate the physiological consequences of CBS/CSE inhibition on Na+ uptake (see above). The observation that despite the difference in experimental protocols, as well as the mode of inhibition (pharmacological inhibition vs gene knockdown), and that all experiments yielded the same result (increase in Na+ uptake) strengthens the conclusion that H2S is an endogenous inhibitor of Na+ uptake. The similarity of the results obtained from CSE/CBS knockdown and their pharmacological inhibition experiment is particularly important, given that the specificity of AOA and PPG as selective blockers has not been confirmed in zebrafish and that larvae were exposed for 24 h to relatively high levels of inhibitors (100 μM). It is also important to note that, although AOA has been used as a selective inhibitor for CBS, a recent study by Asimakopoulou et al. [3] demonstrated that, AOA could effectively inhibit both human CBS and CSE. While it remains to be determined whether AOA is able to inhibit both CBS and CSE in zebrafish, the potential lack of specificity might explain the greater stimulation of Na+ uptake in zebrafish larvae treated with AOA during their acclimation to FWV − Na (Fig. 2).

It is also worth noting that, because H2S can be synthesized from cysteine by four enzymes, namely, CSE, CBS, cysteine aminotransferase-3-mercaptopyruvate, and cysteine lyase [31], the stimulatory effect from CSE and/or CBS inhibition is likely to be lower than otherwise expected owing to the presence of these additional sources of H2S. Indeed, we also performed experiments to knockdown CBSb and observed a similar increase in Na+ uptake as for CSE knockdown (Supplemental Fig. 2). We were unable to confirm this knockdown at the protein level, because of the presence of another isoform (CBSa) with a similar protein sequence (all commercially available antibodies against CBS recognize both isoforms). However, mRNA expression of CBSb was significantly reduced in CBSb MO (Supplement Fig. 1). The CBSb MO was designed to splice out exon 3 (385–491 bp; accession number: NM_001014345.2), and in some instances, injection of splice-variant inducing MO can lead to degradation of target mRNA (e.g., [51]). When we analyzed the tissue distribution of the two CBS isoforms in adult zebrafish, only CBSb was highly expressed in the gill (data not shown), suggesting that this isoform might be playing a role in physiological processes occurring in adult gill (and potentially equivalent functions in larval skin). The idea that multiple H2S-synthesizing enzymes are compensating for each other in ion homeostasis would benefit from further investigation.

Current models for Na+ uptake in zebrafish larvae implicate at least two mechanisms: one involving a NCC (zslc12a10.2) and another involving the Na+/H+ exchanger NHE3b (zslc9a3b) [17, 26]. To determine which Na+ uptake mechanism was being targeted by H2S, we measured the effect of H2S on acid secretion (to test for the involvement of the net acid exporter NHE3b) and its effects in gcm2 morphants in which the differentiation of the NHE3b-containing HR cells is prevented.

When larvae were treated with Na2S, whole-body acid secretion was significantly reduced (Fig. 4b–d), suggesting that Na2S inhibits Na+ uptake operating via a mechanism linked to acid excretion such as NHE3b. Indeed, treatment with 50 and 100 μM 5-(N-Ethyl-5-isopropyl)amiloride, a selective inhibitor of NHE, significantly inhibited Na+ uptake by zebrafish larvae raised in FWV by almost 40 %, suggesting that in FWV, NHE3b is playing a predominant role in Na+ uptake in zebrafish larvae (Fig. 4a). It is worth noting that although whole-body acid secretion was significantly reduced in H2S-treated larvae (Fig. 4b), there was no significant effect on titratable alkalinity (TA) flux following H2S treatment (Fig. 4c). The absence of a significant effect may reflect the experimental design because a reduction in TA flux could reflect both (1) a reduction in acid secretion (potentially caused by NHE3b inhibition) and (2) a stimulation of base secretion, likely to be mediated by one or more Cl−/HCO3 − exchangers thought to be involved with osmoregulation in zebrafish [4, 40]. Furthermore, treatment with Na2S markedly reduced Na+ uptake in larvae raised in acidified (pH 4.0) FWO (Fig. 4e), where NHE3b is the predominant route for Na+ uptake [24, 25]. These two observations, when combined, strongly suggest that Na+ uptake via NHE3b is targeted by H2S. To directly demonstrate that zebrafish NHE3b is under the control of H2S would probably require pHi measurements using an oocyte expression (or similar) system [20]. However, because Na+ uptake via NHE3b is coupled with other “transporters”/enzymes in ionocytes, most notably apical ammonia-conducting channels Rhcg1 [25, 42], cytosolic carbonic anhydrase [21], and basolateral Na+-K+-ATPase (subunit zatp1a1a.5) [32], it is possible that H2S is affecting the function of these transporters/enzymes, rather than NHE3b itself. Indeed, the inhibitory effect of H2S treatment on Na+ uptake by H441 cells (a human bronchiolar epithelial cell line) is caused by the inhibition of Na+-K+-ATPase, rather than direct inhibition of epithelial Na+ channels [2]. Thus, additional experiments (e.g., testing to see if H2S treatment inhibits the stimulation in ammonia secretion following acid exposure) would be required to address whether Na+ uptake via NHE3b is under the direct control of H2S or controlled indirectly via interaction with these other transporters/enzymes.

The knockdown of the transcription factor gcm2 is known to prevent the formation of HRCs and induce the formation of NCC-expressing cells [9, 43]. The present study confirmed that loss of gcm2 greatly reduced the density of conA-positive HRCs (Fig. 5a, b, c shows representative images) and concurrently increased NCC mRNA expression (Fig. 5d). However, it is also worth noting that when gcm2 MO was treated with metolazone, an inhibitor of NCC previously reported to decrease Na+ and Cl− uptake in zebrafish larvae [49], no inhibition of Na+ uptake was observed. As reported previously [10, 24], knockdown of gcm2 alone significantly increased Na+ uptake (Fig. 5e). It is noteworthy that waterborne treatment with H2S did not affect Na+ uptake in gcm2 morphants, despite their high Na+ uptake rate which should have made it easier to detect any inhibitory effects of H2S, if present. This observation strongly suggests that Na+ uptake via NCC is not under the control of H2S. The immunohistochemistry data agree with the idea that H2S does not regulate Na+ uptake via NCC but rather via NHE3b, because CSE was expressed in conA-positive HRCs, as well as in an additional population of ionocytes enriched with Na+-K+-ATPase (Fig. 7). Conclusively excluding the presence of CSE in NCC-expressing ionocytes is not possible owing to the lack of suitable NCC antibodies. Furthermore, demonstrating whether other H2S-synthesizing enzymes are present or absent in NCC-expressing cells should be undertaken before a physiological role for H2S in modulating NCC can be ruled out. However, the absence of an effect of exogenous H2S treatment on gcm2 morphants (enriched with NCC cells) strongly suggests that H2S does not contribute to regulating NCC function. An important, but technically challenging, issue is to assess whether the concentration of H2S is regulated within HRCs when zebrafish larvae are osmotically challenged. Given the proposed role of H2S as an inhibitor of Na+ uptake, it would be expected that [H2S] would rise in HRCs when zebrafish are exposed to Na+-rich water.

It is likely that H2S is acting as a signaling molecule within HRCs to modulate the function of Na+ transporters. The exact nature of the intracellular signaling induced by H2S is currently being explored in several tissues/cell lines, but it appears that H2S may at least induce signaling via protein kinase C (PKC), extracellular signal-regulated kinase (ERK1/2), and mitogen-activated protein kinase (MAPK) (for review, see [31]). An interesting observation is that H2S treatment (via NaHS) significantly attenuated the activation of several isoforms of adenylyl cyclase and cAMP synthesis following morphine treatment in mice striatum [50]. We recently showed that Na+ uptake is stimulated via a cAMP-PKA pathway in HR- and NCC-expressing cells of zebrafish larvae [23]. Although the idea currently is speculative, at least for HRCs, H2S might be interacting with the cAMP-PKA intracellular signaling pathway. Elucidating the intracellular signaling mechanism for H2S and determining the exact Na+ transporter targeted by H2S are the key priorities for future research to better understand the role of H2S as an endogenous osmoregulatory mechanism in FW fish.

Conclusions

Despite the long history of research in the field of FW fish ionic regulation, little attention has been paid to potential mechanisms inhibiting ion uptake capacity. Using a combination of approaches including pharmacological treatment, gene knockdown, and immunohistochemistry, the present study introduces the gasotransmitter H2S as an endogenous inhibitor of Na+ uptake in developing zebrafish. Future studies should focus on several outstanding questions, such as (1) how does intracellular [H2S] change in ionocytes when zebrafish are exposed to osmotic challenge, (2) what are the intracellular signaling pathways regulated by H2S to ultimately affect Na+ uptake, and (3) is the uptake of other ions, including Ca2+ and Cl−, also under the control of H2S? Addressing these questions will provide insight into the physiological role of gasotransmitters in osmoregulation by aquatic vertebrates.