Sulfate (SO42–) is maintained at ∼1 mmol–1 l–1 in teleost fishes that are exposed to media of varying SO42– concentrations. We first measured plasma SO42– concentration in euryhaline fishes that adapt to both SO42–-poor freshwater (<0.5 mmol l) and SO42–-enriched seawater (30 mmol l–1). Unlike Mozambique tilapia and chum salmon, Japanese eels maintained higher plasma SO42– concentration in freshwater (6.2±2.3 mmol l–1) than in seawater (0.7±0.1 mmol l–1). We then analyzed the whole-body SO42– budget using 35SO42–. 35SO42– influx in seawater-adapted eels occurred by 84.5% via body surfaces and 15.5% via digestive tracts. The SO42– influx was higher in seawater eels (1.55 μmol kg–1 h–1) than in freshwater eels (0.09 μmol kg–1 min–1), but it was facilitated in freshwater eels when the difference in SO42– concentrations between plasma and environment was taken into account (freshwater eels, 6.2 vs 0.3 mmol l–1; seawater eels, 0.7 vs 30 mmol l–1). One hour after injection of 35SO42– into the blood of seawater eels, the kidney excreted ∼97% of the ionized form, whereas the radioactivity increased gradually in the medium and the rectal fluid more than 3 h after injection. As the radioactivity was poorly adsorbed by anion-exchange resin, 35SO42– in the blood may be incorporated into cells and excreted by the intestine, gills and skin, probably as mucus. These results show that freshwater eels take up SO42– actively from the environment, but seawater eels cope with the obligatory influx of SO42– through the gills by excreting excess SO42–via the kidney and in mucus.
Because seawater is a hyperosmotic environment where various ions are dissolved, marine teleost fish continuously cope with a threat of dehydration and excess ion invasion. To compensate for the osmotic loss of water, they drink copious seawater and absorb most of the ingested water via the intestine in parallel with monovalent ions such as Na+ and Cl– (Smith, 1930; Marshall and Grosell, 2006; Takei and Balment, 2009; Grosell, 2011). The excess Na+ and Cl– are excreted actively by mitochondria-rich cells in the gills (Evans, 2008). In contrast to the many studies on Na+ and Cl– regulation in fishes, studies on the regulation of divalent ions such as Mg2+, Ca2+ and SO42– are still limited. In terms of the concentration difference between plasma and environmental seawater, the ratio is 3–4 for Na+ and Cl– (Na+, 170 vs 450 mmol l–1; Cl–, 120 vs 500 mmol l–1), but it is 30–50 for Mg2+ and SO42– (Mg2+, 1 vs 50 mmol l–1; SO42–, <1 vs 30 mmol l–1). Therefore, divalent ions unavoidably enter the body via the gills and digestive tracts across the concentration gradient (Hickman, 1968). The excess divalent ions are excreted actively from the proximal tubules of the kidney (Beyenbach et al., 1986; Renfro and Pritchard, 1983).
Sulfate (SO42–) is the second most abundant anion in seawater following Cl–. However, because of the relative difficulty of its measurement, research on SO42– regulation, compared with Cl– regulation, has been largely neglected in marine teleost fish (Marshall and Grosell, 2006). Topologically, the organs that directly contact the external media and have extensive surface area such as the gills and digestive tracts (lumen is an external environment) are possible sites of SO42– influx. It is generally accepted that little SO42– is absorbed across the intestinal epithelium (Marshall and Grosell, 2006), but Hickman (Hickman, 1968) estimated that 11.3% of SO42– derived from ingested seawater is absorbed by the intestine of southern flounder (Paralichthys lethostigma). Concerning the gills, SO42– permeability has been examined in freshwater teleosts and variable data were reported: significant permeability was found in the guppy (Rosenthal, 1961), whereas little permeability was detected in the goldfish (Garcia and Maetz, 1964). However, detailed analyses of the SO42– budget between body fluids and media have not yet been reported in euryhaline fish that experience profound changes in SO42– concentration between freshwater and seawater.
To collect basic information about the whole-body SO42– budget in teleost fish, we used Japanese eels (Anguilla japonica Temminck and Schlegel 1846) as a model because they adapt readily to both freshwater and seawater, and various techniques for in vivo experiments have been established. Initially, we measured plasma SO42– and other ion concentrations in eels adapted to freshwater or seawater and compared the data with those of other euryhaline species, Mozambique tilapia [Oreochromis mossambicus (Peters 1852)] and chum salmon [Oncorhynchus keta (Walbaum 1792)]. We found that SO42– concentration in freshwater eels was much higher than those of tilapia and chum salmon, as reported previously by Nakada et al. (Nakada et al., 2005). However, the plasma SO42– concentration in seawater eels was even lower than in tilapia and salmon, showing suppressed influx and/or enhanced excretion of SO42– ions. We thus compared SO42– concentrations in the urine and rectal fluid between freshwater and seawater eels to estimate the site of SO42– excretion. Further, we examined the whole-body SO42– regulation in detail by measuring the uptake and excretion of SO42– at the regulatory sites (gills/skin, intestine and kidney) using 35SO42– as a tracer in conscious, catheterized eels. As radioactivity appeared immediately in the urine but slowly in the media and rectal fluid after tracer injection, we determined whether the material secreted into the media and the rectal fluid is ionized 35SO42– using anion-exchange resin.
MATERIALS AND METHODS
Cultured Japanese eels (192±5 g, N=39) were purchased from a local dealer. Fifteen eels were maintained in freshwater tanks and 24 eels were maintained in seawater tanks for more than 2 weeks before use. Chum salmon (109±4 g, N=12) and tilapia (125±3 g, N=12) were reared in either freshwater or seawater for more than 1 month. Temperatures were maintained at 18°C for eels, 12°C for chum salmon and 25°C for tilapia. Eels were not fed after purchase, and chum salmon and tilapia were starved for 5 days before experiments. The ionic compositions of freshwater and seawater were as follows: 1.0 mmol l–1 Na+, 0.5 mmol l–1 K+, 0.5 mmol l–1 Cl–, 0.25 mmol l–1 Ca2+, 0.5 mmol l–1 Mg2+ and 0.3 mmol l–1 SO42– for freshwater; and 450 mmol l–1 Na+, 12.5 mmol l–1 K+, 525 mmol l–1 Cl–, 10 mmol l–1 Ca2+, 50 mmol l–1 Mg2+ and 30 mmol l–1 SO42– for seawater. All conditions for fish maintenance were in accordance with the Guidelines for Animal Care and Maintenance at University of Tokyo and the experiments were approved by the Bioscience Committee of the University of Tokyo.
Comparison of plasma ions among euryhaline teleosts
Plasma ion concentrations (Na+, Cl–, SO42–, Ca2+ and Mg2+) and osmolality were measured in eels, tilapia and salmon and compared between freshwater- and seawater-acclimated fish. The ion concentrations were also measured in the urine and rectal fluid of freshwater- and seawater-acclimated eels (N=9 each). Fluid samples were collected from the caudal vein, urinary bladder and rectum after anesthesia in 0.1% (w/v) tricaine methanesulfonate (Sigma-Aldrich, St Louis, MO, USA). Blood samples without anticoagulant were centrifuged immediately after collection at 10,000 g for 5 min at 4°C to separate plasma. Ionized anions and cations were measured by ion chromatography (AV10, Shimadzu, Kyoto, Japan) using a cation-exchange column (IC-C3) and an anion-exchange column (IC-A3).
Measurement of SO42– influx
Freshwater and seawater eels (N=6 each) were anesthetized as above and the ventral aorta was catheterized as described previously (Tsuchida and Takei, 1998). A schematic drawing of the experimental set up is shown in supplementary material Fig. S1. After more than 18 h of recovery from surgery, eels were transferred to a bucket with 2 l of freshwater or seawater, and 7.4 MBq of 35SO42– (1.59 TBq mg–1, Muromachi Science, Tokyo, Japan) was added to the medium. Urine and 75 μl of blood were collected 0, 1, 3, 6, 12 and 24 h after the isotope administration, and 25 μl of plasma or urine was mixed with 5 ml of scintillation cocktail (AQUASOL-2, PerkinElmer, Shelton, CT, USA) for measurement in a scintillation counter (LS6000 SC, Beckman, Fullerton, CA, USA).
To evaluate the role of the digestive tract in SO42– uptake, six seawater eels were cannulated as above, and the esophagus was intubated with a polyethylene tube (0.61 mm o.d.). Seawater was infused through the tube into the esophagus at a constant rate (1 ml h–1) to maintain water balance in seawater (Takei, 2000). After more than 18 h of recovery, three eels were transferred to a bucket with 2 l of seawater containing 7.4 MBq of 35SO42– and infused seawater into the esophagus, and the other three eels were transferred to 35SO42–-free seawater and infused seawater containing 3.7 MBq l–135SO42–. Blood was collected as above for measurement of radioactivity.
Measurement of SO42– efflux
After anesthesia, the ventral aorta and urinary bladder of seawater eels (N=6) were catheterized as described previously (Renfro and Pritchard, 1983) using a vinyl tube (1.5 mm o.d.) inserted into the rectum through the anus. A schematic drawing of the experimental setup is shown in supplementary material Fig. S1. After more than 18 h after surgery, 1.7 MBq of 35SO42– was injected into the ventral aorta in a volume of 100 μl for 30 s followed by a flush with 100 μl of isotonic saline. The dead volume of the tube was ca. 30 μl. Then 500 μl of medium in the bucket, whole urine and rectal fluid, as well as 75 μl of blood were collected 0.5, 1, 3, 6, 12 and 24 h after 35SO42– administration for measurement of radioactivity.
To separate the ionic 35SO42– from 35S containing metabolites, collected fluids were incubated with activated anion-exchange resin (DEAE Sephadex™ G-25, GE Healthcare, Uppsala, Sweden) at 25°C for 3 h. The mixture was centrifuged at 10,000 g for 15 min at 4°C. The absorptive capacity of the resin was confirmed by cold SO42–. Supernatant (10 μl for plasma, urine and rectal fluid and 50 μl for the medium) was mixed with 5 ml of scintillation cocktail for measurement of radioactivity.
Calculation of SO42– fluxes
Extracellular fluid volume of the Japanese eel was estimated to be 15% of body mass based on the Cl– space and Na+ space of European eel (Kirsch, 1972a; Mayer and Nibelle, 1969). The time course of 35SO42– influx was highly linear (r2=0.98 for seawater eels and r2=0.96 for freshwater eels, P<0.01) for 24 h after administration of isotope to the medium. In contrast, the disappearance curve of 35SO42– from the circulation was fitted best to a dual exponential function (At=ae–αt+be–βt+c), where At is plasma radioactivity at time t, and a, b, c, α and β are constants (Takei and Hatakeyama, 1987). The efflux rate of 35SO42– was calculated based on the specific activity of 35SO42– in the plasma. Curve fitting was performed using statistical software (KyPlot 5.0, Kyens, Tokyo, Japan).
Ionic concentrations of body fluids and flux rates were compared between freshwater and seawater fish using Student’s t-tests and Tukey–Kramer tests, respectively. Influx rate was analyzed by linear regression. All analyses were performed using KyPlot 5.0 software. Significance was defined at P<0.05. All results are expressed as means ± s.e.m.
Plasma ion concentrations of euryhaline teleosts in freshwater and seawater
Plasma ionic concentrations and osmolality were generally higher in seawater fish than in freshwater fish irrespective of species (chum salmon, Mozambique tilapia and Japanese eel), except for SO42– concentration in freshwater eels (Fig. 1). The higher values for Na+, Cl–, Mg2+, Ca2+ and osmolality in seawater fish were more prominent in eels than in tilapia or salmon, particularly plasma Cl– concentration. Unlike salmon and tilapia, plasma SO42– concentration was suppressed to ∼1 mmol l–1 in seawater eels (Fig. 1A). Urine SO42– concentration was 40-fold higher in seawater eels than in freshwater eels, indicating active excretion of SO42– by the kidney (Fig. 2). SO42– concentration in the rectal fluid was 25% higher in seawater eels than in freshwater eels.
SO42– influx in freshwater and seawater eels
Plasma radioactivity increased linearly after administration of 35SO42– to the medium in both seawater and freshwater eels, and the influx rate was calculated to be 1.55±0.21 μmol kg–1 h–1 for seawater eels (N=6) and 0.09±0.03 μmol kg–1 h–1 for freshwater eels (N=6) after correction of specific activity of 35SO42– in each medium (Fig. 3A). To take into account the difference in the SO42– concentrations between plasma and the medium of freshwater (6.2 vs 0.3 mmol l–1, 20.6 times) and seawater eels (0.7 vs 30 mmol l–1, 0.02 times), we corrected the influx value at each time point after administration of 35SO42–. The rate was reversed and more facilitated in freshwater eels (1.76±0.21 μmol kg–1 h–1) than in seawater eels (0.03±0.01 μmol kg–1 h–1) after correction (Fig. 3B). When the influx was compared between intact eels and esophagus-ligatured eels, the rate decreased by ∼15% (Fig. 4). Furthermore, when esophagus-ligatured eels in 35SO42–-free seawater were infused with 35SO42–-containing seawater through the esophagus at the normal drinking rate, the increase in plasma radioactivity was ∼15% compared with intact eels in 35SO42–-containing seawater (Fig. 4). Based on these data, SO42– influx was calculated to be 84.5% from the body surfaces (gills and skin) and 15.5% from the digestive tracts (esophagus, stomach and intestine).
SO42– excretion in seawater eels
Plasma radioactivity gradually decreased after injection of 35SO42– into the blood of seawater eels (Fig. 5A). Urine radioactivity quickly increased within 1 h after injection, and a high level was maintained for 6 h followed by a gradual decrease (Fig. 5B). Radioactivity in the rectal fluid remained low for 3 h and increased linearly thereafter up to 24 h (Fig. 5C), indicating secretion of metabolic products containing 35S into the intestinal lumen. Radioactivity in the medium also remained low for almost 6 h and then increased linearly (Fig. 5D), again suggesting secretion of 35S-containing metabolites from the gills and skin. In fact, 85.6% of radioactivity in urine was identified as 35SO42–, as the radioactivity was adsorbed to anion-exchange resin (Table 1). However, only 32.3% in rectal fluid and 10.1% in medium was adsorbed by the resin. Therefore, using the data from 30 min to 1 h in Fig. 5, the excretion rate of ionic SO42– was calculated to be 1.56±0.18 μmol kg–1 h–1 by the kidney, 0.05±0.01 μmol kg–1 h–1 into the intestinal lumen and 0.0011±0.0008 μmol kg–1 h–1 into the medium by the gills and skin in seawater eels (N=6). This result shows that 96.8% of ionic SO42– was excreted via the kidney of seawater eels.
Marine teleost fish are continually faced with an excess of SO42– that permeates the body surfaces according to the concentration gradient imposed by the high SO42– concentration in environmental seawater. The major site of influx may be the gills because they have an extensive surface area and separate the blood from environmental seawater only by a thin monolayer of respiratory cells (Evans, 2008; Marshall and Grosell, 2006). The SO42– excess is further exaggerated by intestinal absorption due to copious drinking of environmental seawater in marine teleosts (Hickman, 1968; Takei and Balment, 2009). However, SO42– transport activity is generally very low in the intestine of marine teleosts (Marshall and Grosell, 2006). The secretion of excess SO42– is accomplished in part by its active secretion at the proximal tubules of the kidney, as evidenced by high SO42– concentration in the urine compared with other anions (Renfro and Pritchard, 1983; Dickman and Renfro, 1986; Cliff and Beyenbach, 1992). Based on these previous data, the following conclusions can be drawn by the present study: (1) freshwater eels maintain high plasma SO42– concentrations (6.2 mmol l–1) by facilitated SO42– uptake from the medium (17.7 μmol kg–1 h–1); (2) seawater eels offset obligatory SO42– influx via the gills by enhanced SO42– excretion by the kidney and maintain low plasma SO42– concentrations (0.7 mmol l–1); and (3) a part of SO42– fluxed into the body of seawater eels was also excreted by the gills, skin and digestive tracts as mucus and other metabolic products.
High plasma SO42– concentrations in freshwater eels
Nakada et al. (Nakada et al., 2005) were the first to report unusually high plasma SO42– concentrations in freshwater eels (∼20 mmol l–1). The high value may be exaggerated, as the plasma Cl– concentration of the eels is extremely low (∼70 mmol l–1). In this study, we measured a plasma SO42– concentration of freshwater eels of 6.2 mmol l–1, which was much higher than that of the other euryhaline teleosts examined (Mozambique tilapia and chum salmon). The SO42– concentration was lower in freshwater than in seawater tilapia and salmon, but the relationship was reversed in eels, where plasma SO42– concentration was suppressed to ∼1 mmol l–1. The SO42– concentrations in the urine and rectal fluid of seawater eels were also comparable to those of other marine teleost fishes (Berglund and Forster, 1958; Hickman, 1968; McDonald and Grosell, 2006). These results clearly show that, in eels, profound changes occur in SO42– regulation when they move from freshwater to seawater. Using the eel as a model, we recently found that in seawater, Cl–, not SO42–, is responsible for switching the SO42– regulation from a freshwater-retention type to a seawater-extrusion type (Watanabe and Takei, 2011a).
In the plasma of Mozambique tilapia and chum salmon, not only SO42– but also other ions (Na+ and Cl–) were higher in seawater than in freshwater. In eels, plasma Na+ and, particularly, Cl– concentrations were also higher in seawater than in freshwater. Because the sum of negative charges (Cl– + SO42–) appears to be similar among different species, SO42– may compensate for the lack of negative charge caused by low plasma Cl– concentration in freshwater eels. In fact, the total negative charge of Cl– and SO42– was almost the same in three euryhaline fishes in freshwater and seawater in the present study (data not shown), and plasma Cl– was inversely correlated with plasma SO42– concentration (Watanabe and Takei, 2011b). Thus, freshwater eels appear to have an unusually high tolerance to hypersulfatemia or a unique mechanism to neutralize the toxic effects of excess SO42–.
Tracer experiments using 35SO42– showed that the influx rate of SO42– was 1.55 μmol kg–1 h–1 in seawater eels, which is balanced by active excretion by the kidney at a rate of 1.56 μmol kg–1 h–1. Comparing the flux rates of ions, the rate for SO42– influx is much lower than that of Na+ (13.2 mmol kg–1 h–1) and Cl– (0.48 mmol kg–1 h–1) in seawater eels (Kirsch and Meister, 1982; Tsukada et al., 2005), as in other marine teleosts (Marshall and Grosell, 2006). Therefore, it seems that SO42– transport across the body surfaces is suppressed in seawater eels compared with Na+ and Cl–. In fact, the turnover rate of Na+ and Cl– is greatly accelerated in euryhaline teleosts when they are in seawater compared with freshwater (Kirsch, 1972b; Miyazaki et al., 1998), probably because their influxes counter osmotic water loss by the gills and their absorption increases water uptake by the intestine (Marshall and Grosell, 2006).
SO42– influx from the environment
The results obtained from the esophagus-ligatured seawater eels revealed that 84.5% of SO42– influx occurred via the body surfaces and only 14.5% entered via the intestine. The major site of influx in the body surface may be the gills, as the eel skin is covered by small scales and mucus and its surface area is 1/20 that of the total area of the gill surface (Motais and Isaia, 1972). It has been suggested that the intestine of marine teleosts is almost impermeable to SO42–, which allows changes in luminal SO42– concentration to be used as a marker for water absorption along the intestine (Marshall and Grosell, 2006). However, the present study showed that intestinal absorption of SO42– is significant in seawater eels.
The SO42– influx in the gills and intestine of seawater eels may be regulated by the transcellular and intercellular pathways through transporters located on the apical and basolateral membranes of epithelial cells. Concerning the transcellular pathway, little is known about SO42– transporters and channels in the gills. In the intestine, solute carrier (Slc) 26a6 family members, anion exchangers that potentially transport SO42–, have been identified in the eel (Watanabe and Takei, 2011b) and pufferfish (Kurita et al., 2008). However, Slc26a6 localized on the apical membrane of epithelia may secrete SO42– into the lumen in exchange for Cl–, as has been shown in the proximal tubule of seawater eel kidney (Watanabe and Takei, 2011a). It was also shown that Slc26a6 in the seawater fish intestine is involved in the secretion of HCO –3 into the lumen for CaCO3 precipitation to decrease luminal fluid osmolality (Grosell, 2011). Slc26a1, which also transports SO42–, was localized abundantly in the basolateral membrane of proximal tubule of eel kidney (Nakada et al., 2005; Watanabe and Takei, 2011b) and in the kidney of other teleost fish (Katoh et al., 2006; Kato et al., 2009), but its gene transcripts were detectable only in the rectum of seawater eels (Watanabe and Takei, 2011b). Judging from the fact that the gills and intestine of seawater eels take up SO42–, unknown SO42– transporters are likely to be present in these tissues. A possible candidate is Slc26a3, which is briskly expressed in the anterior intestine of seawater eels and is a major SO42– transporter in the intestine of mammals (Markovich, 2001); however, its physiological role is not known in fish.
As SO42– concentration in seawater is more than 40-fold greater than that of plasma, the paracellular pathway is a probable route of SO42– uptake. However, the uptake is influenced not only by the concentration gradient but also the transepithelial potential difference. It has been reported that the transepithelial potential is +23 mV (inside positive) in seawater killifish (Wood and Grosell, 2008). If this is the case also in eels, then both the concentration gradient and the transepithelial potential facilitate SO42– uptake through the intercellular route. In fact, the influx rate of SO42– is 1.55 μmol kg–1 h–1 in seawater eels.
For freshwater fish, the gills may also be a major site of SO42– uptake; for example, eels usually drink little in freshwater and thus intestinal absorption may be negligible (Takei, 2000). The influx rate of SO42– of freshwater eels from the medium into the body was as low as 0.09 μmol kg–1 h–1 in the present study. However, because the plasma SO42– concentration was more than 20-fold higher than that of environmental freshwater, SO42– was lost across the concentration gradient but uptake occurred in freshwater. Further, transepithelial potential was shown to decrease to –39 mV in killifish after transfer to freshwater, which also impedes SO42– uptake from the medium (Wood and Grosell, 2008). Thus, it is obvious that SO42– influx is much enhanced in freshwater eels, although the actual influx rate is much smaller than that in seawater eels.
SO42– efflux into the environment
In the present study, 35SO42– injected into the plasma was secreted immediately into the urine but not into the intestinal lumen or into the medium. Thus, ionized SO42– is secreted mostly (∼97%) by the kidney within 1 h after injection. Consistently, 85.1% of the radioactivity was adsorbed to the anion-exchange resin in the urine collected even after 24 h of tracer injection. Previous studies have shown that the secretion of SO42– is achieved at the proximal tubules of the kidney in several teleost species (Katoh et al., 2006; Kato et al., 2009; Watanabe and Takei, 2011a), in which apical Slc26a6 and basolateral Slc26a1 localized on the same cell play major roles in SO42– transport from blood into the tubular lumen. However, ionized SO42– was excreted into the intestinal lumen only by 32.5%. The gene transcripts of Slc26a6a, b and c and Slc26a1 were detected in the eel intestine but in a different segment (Watanabe and Takei, 2011a). Both Slc26a6a and Slc26a1 mRNA were detected only in the rectum of seawater eels (Watanabe and Takei, 2011a), where SO42– may be secreted into the lumen in an ionic form. The ionic SO42– secreted from the gills and skin was only 10% as judged by the adsorption to the anion-exchange resin.
Significant amounts of 35SO42– injected into the circulation was taken up by the tissues, processed, and secreted into the medium and rectal fluid more than 3 h after injection. This indicates that 35SO42– was incorporated into some metabolic products and secreted from the gills, skin and intestine. In mammals, mucus is made from various sulfated proteins and sugars (Amerongen et al., 1998). Marine teleosts have large amounts of sulfated glycoproteins in abundant goblet cells in the digestive tract and mucus cells in the gills and skin, as shown by Alcian Blue staining (Sarasquete et al., 2001). Thus most 35S may be incorporated into the mucus. Because 35S-radioactivity in the rectal fluid increased to a significant amount after 24 h (Fig. 5C), the role of intestine in the secretion of excess SO42– seems to be important for maintenance of low plasma SO42– concentration in seawater eels. In mammals, sulfotransferase is capable of converting glycoprotein into sulfated substances including mucin (Brockhausen, 2003), but such a pathway has not been demonstrated in fishes.
Fig. 6 summarizes the relative contribution of osmoregulatory sites to the whole-body SO42– budget in seawater eels as shown by this study. It is obvious that the vigorous influx of SO42–via the gills across the concentration gradient is nullified by active renal excretion. Recently, we localized several Slc transporters in the apical and basolateral membrane of epithelial cells in different segments of the renal proximal tubule in seawater eels (Watanabe and Takei, 2011a). Our next target may be to identify molecules and the mechanism of SO42– transport in the gills and intestine of seawater eels and those concerned in the active uptake of SO42– by the gills of freshwater eels. Because eels are unique in their SO42– regulation among euryhaline fishes, they provide an interesting model for identifying important regulatory mechanisms in teleost fish.
We thank Drs Susumu Hyodo, Makoto Kusakabe and Jillian Healy and Ms. Sanae Hasegawa of the Laboratory of Physiology, Department of Marine Bioscience, Atmosphere and Ocean Research Institute, University of Tokyo, and Dr John Donald of Deakin University for advice and comments on the manuscript.
This research was supported in part by a Grant-in-Aid for Basic Research (A) from the Japan Society for the Promotion of Science to Y.T. (13304063 and 16207004). T.W. was supported by the Global COE Program (Integrative Life Sciences Based on Study of Biosignaling Mechanisms), Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan.