Integumental and branchial chloride cells of tilapia larvae (Oreochromis mossambicus) were studied at the light-microscopical and ultrastructural level. Total numbers and distribution of chloride cells were quantified after immunostaining of cross sections of the entire larvae with an antibody against the alpha-subunit of Na+/K+-ATPase. The majority (66%) of Na+/K+-ATPase-immunoreactive (ir) cells, i.e. chloride cells, of freshwater tilapia larvae were located extrabranchially up to 48 h after hatching. Five days after hatching, the majority (80%) of chloride cells were found in the buccal cavity. Transfer of 24-h-old larvae to 20% sea water speeded up this process; 24 h after transfer (i.e. 48 h after hatching), the majority (59%) of chloride cells were located in the buccal cavity. The branchial chloride cell population of 24-h- and 120-h-old larvae consisted of immature, mature, apoptotic and necrotic chloride cells. However, relatively more immature chloride cells were observed in freshwater larvae (42-63%) than in (previously studied) freshwater adults (21%), illustrating the developmental state of the gills. After transfer to sea water, the incidence of degenerative chloride cells did not change. Furthermore, the incidence of immature cells had decreased and a new subtype of chloride cells, the "mitochondria-poor" cells, appeared more frequently. These mitochondria-poor chloride cells were characterised by an abundant tubular system and relatively few mitochondria, which were aligned at the border or concentrated in one part of the cytoplasm. Most of these cells did not contact the water. The function of their enhanced appearance after seawater transfer is unknown.

No abstract

Summary. Branchial chloride cells, which actively take up ions in the gills of freshwater fish, were studied in tilapia {Oreochromis mossamhicus) exposed to sublethally acidified freshwater. Structural damage of cells, result ing in cell death by necrosis, only occurred transiently, when the reduction of water pH was acute rather than gradual. The most prominent effects of water acidifica tion were the rapid increase in the num ber of chloride cells and the changes in frequency of the different stages of the chloride cell cycle. In the opercular inner epitheli um, a twofold increase in cells occurred 48 h after gradu al acidification. Cell density stabilized after 4 weeks at a level 5 times that of control fish. Four transitory stages were distinguished in the chloride cell cycle: accessory or replacement cells, immature, mature, and degenerat ing (apoptotic) cells. In control fish, m ature chloride cells dom inated (over 50%) with immature and apoptot ic cells totalling about 40%. After 4 weeks in acid water, only 13% of the cells were mature. Immature and apop totic cells dominated, each representing about 40% of the total num ber of chloride cells. M ature cells apparent ly age rapidly under these conditions. Thus, chloride cells turn over quickly in acid water, with a minor in crease in ion transport capacity of the gills. This conclu sion is supported by the observation that opercular and branchial Na + / K + ATPase activities in treated fish are only 4 0 % -5 0 % higher than in controls.

In the presence of low ambient calcium levels, prolactin cell activity is directly related to the height of the osmotic gradient between blood plasma and external medium, and not to ambient osmolarity. Prolactin cell activity is minimal in fish adapted to iso-osmotic saline. The osmotic water permeability of the gills is inversely related to the height of the osmotic gradient and to prolactin cell activity. In gills of fish from iso-osmotic saline the osmotic water permeability is maximal. This high permeability is reduced after injection of ovine prolactin. It is concluded that the rate of prolactin secretion is related directly to the rate of the osmotic water fluxes-and. possibly, passive ion fluxes-the fish are facing, irrespective of the direction of these fluxes. In the presence of high calcium levels, however, prolactin cell activity as well as osmotic water permeability of the gills were low and independent of ambient osmolarity. Prolactin injections did not influence the osmotic water permeability of gills from high-calcium-adapted fish. High prolactin secretion in freshwater-adapted fish is likely due to the presence of low environmental calcium levels and a high osmotic gradient between blood and environment. Low prolactin secretion in seawater fish-fish that are facing an even higher osmotic gradient-is probably caused by the high ambient calcium and magnesium levels, which may make prolactin superfluous for the control of the osmotic water permeability.

Although exposure to acid water (pH 3.5) induces severe and prolonged reduction in plasma osmolarity and total plasma calcium concentration in tilapia (Oreochromis mossambicus) and goldfish (Carassius auratus), the responses of the hypophyseal cells are clearly different. In tilapia, the size of the rostral pars distalis of the pituitary gland is enlarged as a result of the increase in size and number of prolactin cells. The pars intermedia PAS-positive (PIPAS) cells are not noticeably changed. Conversely, in goldfish, prolactin cells are unaffected, whereas the number of enlarged PIPAS cells increases markedly. Stimulation of prolactin secretion may be responsible for the partial restoration of plasma osmolarity and calcium levels observed in tilapia after two weeks exposure to acid water. Prolactin cells apparently play a role in the adaptation to acid stress by counteracting osmoregulatory disturbances. Goldfish show no restoration of plasma osmolarity during the course of the experiment. Plasma calcium levels tend to increase. Although prolactin may have an osmoregulatory function in goldfish under steady state conditions, goldfish prolactin cells do not seem to participate in the physiological adaptation to environmental changes that disturb water and ion homeostasis. The function of PIPAS cells in tilapia remains unclear and is apparently unconnected with ion regulation. The observations on these cells in goldfish are consistent with the hypercalcemic activity suggested for them.