Distributional changes in branchial chloride cells were examined in Japanese sea bass (Lateolabrax japonicus) juveniles transferred from seawater (SW) to fresh water (FW) during their migration season toward low salinity habitat in nature. Chloride cells were identified by immunocytochemistry with a specific antiserum for Na ,K -ATPase. In fish reared in SW as controls, branchial chloride cells were localized exclusively in the filaments and absent in the lamellae. When sea bass were transferred from SW to FW, chloride cells emerged in gill lamellae, starting at the proximal part of the lamellae and thereafter spread over the lamellar epithelium. On 7th and 15th days after FW transfer, chloride cells were mostly found on the lamellae, whereas the number of filament chloride cells was decreased. These results suggest that, in Japanese sea bass juveniles, chloride cells in the gill lamellae are important in FW adaptation, and that lamellar chloride cells originated from the filaments and migrated to the lamellae during FW adaptation.
Teleost fish maintain ionic and osmotic gradients between the body fluid and external environments (Evans, 1984). In both freshwater (FW) and seawater (SW) fish, plasma osmolality is usually maintained at narrow physiological ranges, which are about one third of SW osmolality. The gills, kidney and intestine are important osmoregulatory organs, creating ionic and osmotic gradients. In particular, chloride cells mainly present in the gills are best known as the salt-secreting site in SW fish (Foskett and Scheffey, 1982; Zadunaiskey, 1984; Avella and Bornancin, 1990). Chloride cells are characterized by the presence of numerous mitochondria and an extensive tubular system. The tubular system is continuous with the basolateral membrane, resulting in the large surface area for the placement of ion-transporting proteins, such as Na+,K+-ATPase (Karnakey et al., 1976; Zadunaiskey, 1984; McCormick, 1995). In contrast to their salt-secreting function in SW, osmoregulatory functions of chloride cells in FW are still poorly understood. Chloride cells have been suggested to be in charge of ion uptake in FW to compensate for diffusional ion losses in hypoosmotic environments (McCormick et al., 1992; Flik et al., 1996), although direct evidence for this is still lacking.
In salmonids and eel, two distinct types of chloride cells are detected in filament and lamellar epithelia in the gills (Avella and Bornancin, 1990; Uchida et al., 1996; Sasai et al., 1998). In chum salmon fry, filament chloride cells are activated following transfer from FW to SW, whereas lamellar cells are frequently observed in FW, but disappear after SW transfer (Uchida et al., 1996). These findings suggest that filament and lamellar chloride cells are important in adaptation to SW and FW, respectively; filament chloride cells are considered to be the site for salt secretion in SW, and lamellar chloride cells are presumably responsible for ion uptake in FW.
Japanese sea bass Lateolabrax japonicus is an euryhaline marine fish inhabiting coastal areas and estuaries in Japan, and occasionally appears in FW (Ochiai and Tanaka, 1986). As is the case with other marine teleosts, eggs are spawned in SW; however, the advanced larvae are often found in brackish water zones such as estuaries (Matsumiya et al., 1982, 1985; Fujita et al., 1988). In fact, their amphydromous characteristics are pointed out in some population (Tanaka, 1997); they immigrate to FW at the larva-juvenile transformation phase or early juvenile stage. Therefore, in contrast to other marine teleosts so far examined, Japanese sea bass is expected to acquire the ability for FW adaptation during their early life stages.
Considering the unique FW adaptability, Japanese sea bass would be a good experimental model to explore the functions of chloride cells in FW. In this study, changes in distribution and density of branchial chloride cells were investigated following transfer from SW to FW in Japanese sea bass at their potential FW entry stage. Our observations indicated that chloride cells emerged in the gill lamellae after FW transfer, and that those lamellar chloride cells originated from the gill filaments and migrated to the lamellae during FW adaptation.
MATERIALS AND METHODS
Fish rearing and transfer experiments
Artificially fertilized eggs of sea bass were obtained from Chiba Prefectural Tokyo Bay Fish Farming Center, and transported to Fisheries Research Station, Kyoto University, on January 16, 1997. They were reared in a stock tank (500-l) with filtered SW at 15°C. After hatching on January 17, larvae were fed on rotifers Brachionus plicatilis, Artemia sp. nauplii and artificial diet according to their developmental stages until 103 days after hatching when the transfer experiment started. Three tanks with closed filtering systems were filled with FW (0 ppt salinity), diluted SW (11 ppt) or SW (33 ppt). Seventy juveniles (juvenile stage; total length, 40 mm; wet weight, 700 mg) were directly transferred to each tank, reared at 18–22°C under the natural photoperiod. Dead fish were counted daily and removed.
Ten fish from each tank were sampled on 0 day, 1 day, 3 days, 7 days and 15 days after transfer (days 0, 1, 3, 7 and 15). After anesthesia with MS-222, the gills were removed and fixed in 4% paraformaldehyde in 0.1M phosphate buffer (pH 7.4) for 24 hrs at 4°C, and preserved in 70% ethanol at 4°C.
Immunocytochemical detection of gill chloride cells
The second gill arch was removed, dehydrated in ethanol and embedded in paraffin. Serial sections (5 μm) were cut parallel to the long axis of the filament, and mounted on slides. Branchial chloride cells were detected immunocytochemically with an antiserum specific for Na+,K+-ATPase, which was raised against a synthetic peptide corresponding to part of the highly conserved region of the Na+,K+-ATPase α-subunit (Ura et al., 1996). The sections were immnocytochemically stained by the avidin-biotin-peroxidase complex (ABC) method (Hsu et al., 1981) using a commercial kit (DAKO sABC kit, Glostrup, Denmark). The deparaffined sections were incubated sequentially with: (1) 0.3% H2O2 for 60 min, (2) 5% normal goat serum for 60 min, (3) anti-Na+,K+-ATPase serum diluted 1:2000 for 20 hr at 4°C, (4) biotinylated anti-rabbit IgG for 30 min, (5) ABC for 30 min, and (6) 0.01% 3,3′-diaminobenzidine tetrahydrochloride containing 0.005% H2O2 for 8 min.
For Quantitative analyses, the density of chloride cells was measured. Immnoreactive cells were classified into lamellar chloride cells when more than 50% of the sectional area was on a lamellae, and into filament chloride cells in other cases. Our preliminary observations and a previous study (Laurent and Dunel, 1980) have indicated that chloride cells in the filament were more abundant on the afferent vascular side than on the efferent. To overcome the uneven distribution of chloride cells in the gills and to minimize the counting loss, the analysis was made on the central zone of the filament, which had the full length of the lamella. All immunoreactive cells in the filaments and lamellae were counted along 700–1000 μm length of a filament, and the density was expressed as cell number/mm gill filament. Since the length of lamella varied among individual fish, and moreover, no tendency among treatment groups was found, the locations of chloride cells were expressed as relative values. To represent the relative location of lamellar chloride cells to the filament, the distance from a chloride cell to the filament (D) and the total length of the lamella (L) were measured on the section, as shown in Fig. 1. The relative distance to the filament was calculated as D × 100/L (%).
Data for chloride cells density were shown as the mean ± S.E.M. Significance of differences was determined by the Student's t-test after variance analysis by F-test.
Fig. 2 shows changes in the survival rates after transfer to different salinities. More than 90% of fish transferred to SW (control) and 1/3 SW survived for 15 days. When transferred to FW, the number of surviving larvae decreased until day 10; however, no fish were dead thereafter, the survival rate being 77% on day 15.
Fig. 3 shows sagittal sections of the gills stained with antiNa+,K+-ATPase serum to identify chloride cells. Distribution of chloride cells was restricted to the filament epithelia in fish reared in SW (Fig. 3a, b) and those transferred to 1/3 SW (Fig. 3c), and no chloride cells were observed in the lamellae. In fish transferred to FW, however, chloride cells emerged at the proximal part of the lamellae on day 1 (Fig. 3d), and then expanded their distribution over the lamellae on days 3, 7 and 15 (Fig. 3e, f, g).
For quantitative analyses, the density of chloride cells in both filaments and lamellae was measured (Fig. 4). On days 0 and 15 in SW and on day 15 in 1/3 SW, chloride cells were exclusively detected in filaments, and no chloride cells were detectable in the lamellae. When transferred to FW, a few, but significant numbers of, chloride cells appeared in the lamellae as early as day 1. Thereafter, lamellar chloride cells increased significantly (p<0.01) and predominated over filament cells on days 3, 7 and 15. In sharp contrast, chloride cells in the filaments decreased significantly (p<0.01) following FW transfer.
Changes in chloride cell distribution in the lamellar epithelia following FW transfer are shown in Fig. 5. In SW (day 0), chloride cells were not present in the lamellae. Following FW transfer, lamellar chloride cells were found only in the region close to the filament on day 1, and then spread out to be distributed over the full length of the lamellae on days 7 and 15.
The most drastic change in the gills of sea bass juveniles transferred from SW to FW is the appearance of chloride cells in the lamellae. All branchial chloride cells were detected in the filaments in fish reared in SW and 1/3 SW, whereas lamellar chloride cells become evident in fish transferred to FW (Fig. 3). The presence of chloride cells in the lamellar epithelia was reported in chum salmon (Uchida et al., 1996) and Japanese eel (Sasai et al., 1998) adapted to FW. Those lamellar chloride cells disappeared when transferred to SW in chum salmon (Uchida et al., 1996). In the present study, after transfer to FW, lamellar chloride cells first appeared at the base close to the filament, and expanded from the proximal to distal part of the lamellae, probably suggesting an unidirectional migration of chloride cells. In chum salmon gills, undifferentiated stem cells and immature chloride cells are often observed in the filaments in close association with the central venus sinus, but not found in the gill lamellae (Uchida and Kaneko, 1996). Cell proliferation was only observed in filaments, and not in lamellae (Laurent et al., 1994). Taken together, lamellar chloride cells appearing after FW transfer are considered to originate from undifferentiated cells in the filament, and migrate toward the distal part of the lamellae to spread their distribution over the lamellar epithelia. Such a spacial shift of the chloride cell distribution might facilitate a possible ion-transporting function, allowing to expand their apical membrane in contact with external environments.
The juveniles transferred to SW or 1/3 SW showed a lower mortality during the first 7 days (Fig. 2), suggesting that the handling stress caused by transfer could be minor. On the contrary, the survival rate decreased to about 80% in fish transferred to FW. This mortality is probably due to the exposure to the hypoosmotic environment. It is noticeable that a higher mortality was observed only during the first 7 days. On the other hand, the chloride cell distribution reached the maximum expansion over the lamellae, and their density became saturated on day 7 (Figs. 4 and 5). This coincidence suggests that the observed changes in chloride cell distribution during the first 7 days after FW transfer are critical for full adaptation to FW in juvenile sea bass.
In previous studies, diverse functions of chloride cells have been considered in SW-and FW-adapted fish. The involvement of chloride cells in ion secretion has been demonstrated in SW-adapted fish (Foskett and Scheffey, 1982; Zadunaiskey, 1984; Avella and Bornancin, 1990), whereas chloride cells have been suggested to be involved in ion absorption in FW-adapted fish (Perry et al., 1992; Flik et al., 1996; Li et al., 1997), especially in calcium uptake (McCormick et al., 1992; Flik et al., 1995, 1996; Li et al., 1997). On the other hand, morphologically different types of chloride cells were reported by Pisam et al. (1987), who classified filament chloride cells into two types, based on their ultrastructural observations in guppy Lebistes reticulatus. The chloride cells located on the base of gill lamellae were referred to as α type and those on the interlamellar region of the filaments as ß type. While α cells exist in both FW and SW, ß cells are only present in FW. The occurrence of α and ß cells has also been confirmed in other several teleost species (Pisam et al., 1987, 1990, 1993; Pisam and Rambourg, 1991). Although they claimed that ß cells in the interlamellar region of the gill filaments were FW-type chloride cells, it is also possible that α-type chloride cells at the base of the lamellae migrate up to the lamellar epithelia in response to FW exposure. This may explain the absence of α cells in FW. Such explanation might also support our finding that FW-type lamellar chloride cells originate from gill filaments.
In the present study, appearance of chloride cells in the lamellae was demonstrated in juvenile sea bass transferred to FW. Our findings indicate the importance of lamellar chloride cells in FW adaptation. The function of lamellar chloride cells in sea bass is probably ion uptake from dilute environmental water, as is expected in salmonids and eel. Considering the unique FW adaptability during their juvenile stage, sea bass would provide a valuable experimental model in examining the activation and production mechanisms of chloride cells in gill lamellae.
The mechanisms of the alteration in chloride cell distribution and number, as well as of putative changes in their functions may be under hormonal control. FW-specific ß chloride cells (Pisam et al., 1993) were induced in salt wateradapted tilapia Oreochromis niloticus by prolactin, a hormone important for FW adaptation. In vitro treatment with cortisol, a SW-adapting hormone, increased the number and size of chloride cells in the opercular membrane (McCormick, 1990), and chloride cells in the yolk-sac membrane also increased their size in response to cortisol treatment both in vivo and in vitro (Ayson et al., 1995). Moreover, Uchida et al. (1998) demonstrated intense immunoreaction to cortisol receptor on lamellar chloride cells in chum salmon. On the contrary, the decrease in the chloride cell size following prolactin injections to SW-adapted tilapia was reported by Herndon et al. (1991). Although at present no information is available on hormonal control of gill chloride cells in sea bass, development of lamellar chloride cells in FW is likely to be regulated by endocrine systems.
We thank Mr. N. Makino, Chiba Prefectural Tokyo Bay Fish Farming Center, for supplying sea bass eggs. We are grateful to Dr. K. Soyano, Nagasaki University, for supplying anti-Na+,K+-ATPase. We would like to thank Mr. J. Hiroi, Kyoto University, for his technical advice. This study was supported in part by grants-in-aid from the Ministry of Education and from Cooperative Program (No. 27, 1998) of Ocean Research Institute, University of Tokyo.