INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

IN MANY EURYHALINE FISH, including the tilapia (Oreochromis mossambicus), prolactin (PRL) plays a central role in freshwater osmoregulation. By acting on osmoregulatory surfaces, PRL stimulates ion retention and decreases water influx (3, 6, 13, 22). Consistent with its osmoregulatory activity, PRL release from the tilapia pituitary increases as extracellular osmolality is reduced both in vivo and in vitro (8, 25, 33, 34 , 42). Blood osmolality in freshwater-acclimated tilapia (~310 mosmol/kgH₂O) is somewhat lower than that in seawater-acclimated fish (350 mosmol/kgH₂O) (33, 42). The release of PRL increases in vitro in direct relation to reductions in osmotic pressure that fall within the physiological range observed in vivo (7, 8, 27). Thus PRL release is governed by extracellular osmolality, the factor that PRL regulates at the organismic level.

These osmoreceptive properties of tilapia PRL cells facilitate the investigation of the mechanisms by which osmotic signals are transduced into osmoregulatory responses. Furthermore, of particular advantage, PRL cells are arranged into a nearly homogeneous tissue, comprising nearly 100% of the rostral pars distalis (RPD), allowing PRL cells to be easily isolated for in vitro studies (10, 26).

Under hyposmotic conditions, most cells swell and then undergo what has come to be called a regulatory volume decrease (RVD) (19). In perfusion studies, the rise in PRL release from intact RPDs reaches a peak within 30 min after the onset of hyposmotic stimulation before subsiding to an elevated plateau (7). Furthermore, this elevation in PRL release from whole pituitaries incubated in hyposmotic medium is maintained for up to 12 h compared with those maintained in hyperosmotic medium (33). These and other studies have employed either whole pituitaries or intact RPDs to investigate peak and sustained PRL release in response to hyposmotic medium. The time course followed by hyposmotically induced PRL release from dispersed PRL cells has not been described, although it is known that after overnight incubation, both intact RPDs and dispersed PRL cells show similar responses to hyposmotic medium (4). It is also not known whether the decrease in PRL release from peak levels after hyposmotic stimulation is due to a decrease in cell volume as a consequence of RVD.

Several studies using mammalian renal cells and cell lines have suggested that a rise in intracellular Ca²⁺ concentration ([Ca²⁺]_i) is important for RVD and that this volume adjustment is dependent on extracellular Ca²⁺ (21, 37, 41). Increases in PRL release can be induced experimentally by increasing [Ca²⁺]_i with Ca²⁺ ionophores and, over a long term (18 h), are dependent on extracellular [Ca²⁺] (8, 9, 11). Because hyposmotic medium alone is sufficient stimulus for a rise in [Ca²⁺]_i (10), it is hypothesized that extracellular Ca²⁺ entry is an important step in the activation of hyposmotically induced PRL release.

In the present experiments, a perfusion system utilizing dispersed PRL cells permitted switches to media of different osmotic concentrations and ionic compositions with minimal mixing. By video imaging the cells during perfusion, the time course of the hyposmotically induced increase in cell volume could be observed together with the increase in PRL release into the perfusate. A similar chamber, containing PRL cells exposed to the same conditions, allowed the determination of the changes in [Ca²⁺]_i from fura 2-AM-loaded cells. Thus we are able to show for the first time that a cell volume increase, occurring rapidly in response to hyposmotic medium, precedes the rise in PRL release in a model in which osmotically sensitive cells secrete a hormone responsible for maintaining osmotic homeostasis at the organismic level. Furthermore, the involvement of extracellular Ca²⁺ in changes in [Ca²⁺]_i, cell volume, and PRL release after hyposmotic stimulation has been characterized.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Fish. Mature tilapia (Oreochromis mossambicus) of both sexes, weighing 200-600 g, were obtained from a population maintained at the Hawaii Institute of Marine Biology. They were kept outdoors in 5,000-liter tanks in fresh water and fed twice daily with Purina Trout Chow (~2% of body wt, twice a day). Water temperature was 22-26°C. All experiments were conducted in accordance with the principles and procedures approved by the Institutional Animal Care and Use Committee, University of Hawaii.

Cell dispersion and perfusion. Pituitaries were removed after decapitation. The RPD, composed of nearly 100% PRL cells, was dissected and placed in groups of 3 in a 24-well culture plate with 500 µl of hyperosmotic (355 mosmol/kgH₂O) Krebs bicarbonate-Ringer solution, containing glucose (500 mg/l), glutamine (290 mg/l), and Eagle's minimal essential medium as described by Wigham et al. (40). Medium osmolality was adjusted by varying the concentration of NaCl and checked using a vapor pressure osmometer (Wescor 5100C; Logan, UT). Tissues were preincubated overnight (18-20 h) at 26-28°C on a gyratory platform (80 rpm) under a humidified atmosphere composed of 95% O₂-5% CO₂. After preincubation, PRL cells were dispersed in 0.125% trypsin (Sigma, St. Louis, MO) dissolved in phosphate-buffered saline (PBS, 355 mosmol/kgH₂O), washed twice with PBS to remove trypsin from the cells (4), and plated on a poly-L-lysine (Sigma; 0.1 mg/ml)-coated chamber. The chamber consisted of two rectangular coverslips (22 × 40 mm) held together with 100% silicone at the extremities, with cut hypodermic needles (23 guage) forming the inlet and outlet. The chamber volume ranged from 250 to 300 µl and accommodated an average of 537,000 ± 53,000 PRL cells (n = 12 different chamber preparations). The inlet of the chamber was connected to syringes (60 ml) containing media. The chamber was mounted on the stage of a microscope equipped with a video camera. Media of various compositions were gravity fed through the syringes and switched through a manifold valve set upstream of the chamber. The flow rate ranged from 60 to 80 µl/min. The time lapse between reaching the chamber and completely replacing the previous chamber solution ranged between 1.5 and 2.5 min. The dead time between changing the valve and the outflow of perfusate was 4-5 min. Thus the time difference between cell volume and PRL release measurements was defined by the temporal difference between the average time between reaching and clearing the chamber and the time taken to reach the point where the perfusate was collected. The mean time difference was 2 min and 9 ± 6 s (n = 4) and has been incorporated in Figs. 1, 2, and 5 to correctly express the time in which PRL cell volume and PRL release responded to a change in medium. The perfusate was collected manually every 5 min in 0.5-ml centrifuge tubes that had been previously weighed. Tubes containing perfusate were then reweighed and the volume in each sample determined. These volumes were used to determine the flow rate and to calculate the amount of PRL in each sample after radioimmunoassay. Each perfusion experiment for cell volume and PRL release determinations was replicated with cells from different RPD dissociations at least four times.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Effects of sequential 15% reductions in medium osmolality (from 355 to 300 mosmol/kgH2O) on the release of prolactin (PRL)177 and PRL188. PRL release is expressed as percent change from baseline. Release of both PRLs from the same preparations of dispersed PRL cells increased in a similar way after medium osmolality was reduced. Release of both PRLs exhibited near-identical responses to all other treatments in this study, and for clarity, only assays of PRL188 are shown in Figs. 2 and 5. Values are means ± SE (n = 4 replicate runs).

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Effects of 15 (A) and 30% (B) reduction in medium osmolality (from 355 to 300 mosmol/kgH2O and from 355 to 248 mosmol/kgH2O, respectively) on cell volume change and PRL release measured from the same preparation of dispersed PRL cells. PRL release and cell volume are expressed as percent change from the baseline. Horizontal bars represent the estimated time in which the solution reached the cells and are corrected for the temporal lag between cell volume and PRL release measurements. A: after exposure to 300 mosmol/kgH2O medium, cell volume increased 15% above baseline (), closely followed by a 10- to 11-fold increase in PRL release (). B: exposure to 248 mosmol/kgH2O medium increased cell volume 30% above baseline (), also followed by a 10-fold increase in PRL release (). Values are means ± SE (n = 4 replicate experiments for PRL release and 16 cells, taken from all experiments, for volume measurements). C: the slope of cell volume decline after the peak induced by a 30% reduction in osmolality (248 mosmol/kgH2O) is significantly (P < 0.001) steeper than that of a 15% reduction (300 mosmol/kgH2O). The slopes were compared by multiple regression analysis and are shown as solid lines.

Cell size. Cell images were captured every 5 min with a video camera and stored in a computer (Macintosh IIcx). The microscope was equipped with a ×100 oil-immersion objective lens (Nikon, Japan), and the total magnification of the image, as seen on the screen, was ×1,200. The cross-sectional areas of cells were estimated by tracing each cell from digitally captured images. Images were processed with the NSF Scion Image software. Areas (A) were obtained in pixels and then transformed into square micrometers as determined by viewing a stage micrometer. Cell volume (V) was estimated from the area as follows
Cell volume was expressed as a percent change from the baseline (taken as 100%). The baseline value was taken as the mean of the volume calculated for the first five time points in pretreatment medium.

Intracellular Ca²+ concentration. The procedure for fura 2 loading in dispersed PRL cells has been described previously (Hyde GN, Seale AP, Grau EG, and Borksi RJ, unpublished observations). Briefly, PRL cells were dispersed as described in Cell dispersion and perfusion but were placed on round coverslips (22-mm diameter) previously coated with poly-L-lysine (0.1 mg/ml) and preincubated overnight in 355 mosmol/kgH₂O medium. Cells were then loaded with 5 µM fura 2-AM (Molecular Probes, Eugene, OR), freshly diluted from a stock solution of 5 mM in anhydrous DMSO, for 90 min at 28°C. After loading, the coverslips with cells were rinsed and placed in baseline medium (355 mosmol/kgH₂O) for 30 min prior to data recording. Individual coverslips were mounted in a metal chamber that allows the cells to be perfused with media of different compositions (4). The chamber was mounted on the stage of an inverted microscope (Nikon).

8

6

Radioimmunoassays. The tilapia pituitary secretes two distinct PRL molecules, PRL₁₈₈ and PRL₁₇₇, that are encoded by separate genes. The quantity of both PRLs in the perfusate was measured by homologous radioimmunoassays (1, 33, 42). In the present experiment, PRL₁₈₈ and PRL₁₇₇ release showed highly similar patterns in response to a reduction in osmolality (Fig. 1). Thus, for clarity, only the PRL₁₈₈ response is presented. Values obtained as nanograms per milliliter were converted to a percent change from the baseline set at 100%, which was determined from the average of the first five time points under pretreatment conditions (355 mosmol/kgH₂O).

Statistical analysis. Slope comparisons of cell volume changes were carried out by multiple regression and simple regression analyses. For the comparison of PRL release in Ca²⁺-deleted medium and normal hyposmotic medium, five values for each treatment period were added to generate a net response over 25 min. The net response over 25 min during the pretreatment was subtracted from the net response over 25 min during treatment for each replicate and then averaged. A pairwise t-test was employed to compare the average percent pretreatment change between cells in hyposmotic medium and Ca²⁺-deleted hyposmotic medium. For comparisons of cell volume in Ca²⁺-deleted medium and normal hyposmotic medium, the net change in cell volume under both conditions was computed by subtracting the average cell volume change in pretreatment medium from that in treatment medium. Comparisons between treatments were performed using a pairwise t-test. Calculations were performed using the Minitab statistical software package (State College, PA).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Effects of a reduction in medium osmolality on cell volume and PRL release. The effects of a hyposmotic medium on PRL release and cell volume were examined over a time course of 80 min (Fig. 2A). The increase in cell volume was proportional to the degree that medium osmolality was reduced. As the osmolality was reduced by 15%, the peak volume increase amounted to ~15% above baseline. Cell volume was consistently increased at the first measurement (5 min) after hyposmotic medium reached the cells and was maximum at the second measurement (10 min). Volume then decreased gradually, following a trajectory that could be well fitted by a linear regression (see below). The second PRL determination (but not the first) following change to the hyposmotic medium consistently showed a severalfold increase in PRL, on average 700% for a 15% decrease in medium osmolality. A peak in PRL release occurred at the third determination (15 min), reaching as much as 11-fold above baseline release. The rate of PRL release then declined sharply to reach a plateau of two- to threefold above the baseline after 25 min. This plateau was sustained throughout the exposure to hyposmotic medium. The baseline rate of PRL release ranged between 0.8 and 1.6 ng/min, whereas the peak release following hyposmotic stimulation reached 12.5 ng/min. After the return to hyperosmotic (355 mosmol/kgH₂O) medium, cell volume declined to the original size within a single measurement, whereas PRL release declined more slowly.

Intracellular Ca ²+ oscillations. The baseline [Ca²⁺]_i of PRL cells perfused with 355 mosmol/kgH₂O medium was expressed as 340/380 ratios and classified into three main patterns (Fig. 3). Quiescent or silent cells, exhibiting low spontaneous variation from baseline (Fig. 3A), accounted for 60% of the cells studied (111 of 186 cells). The remainder showed spontaneous activity that could be further classified into two patterns: spontaneous high-frequency oscillators (Fig. 3B) and low-frequency-oscillators (Fig. 3C). The cells exhibiting these two patterns accounted for 32 and 8% respectively, of the total cells studied. Sporadic and transient peaks that show a range of amplitudes characterize the high-frequency spontaneous oscillators. Low-frequency oscillators are characterized by periods of slow increase and recovery to resting [Ca²⁺]_i.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Patterns of basal intracellular Ca2+ concentration ([Ca2+]i) in tilapia PRL cells. Traces represent the 340/380 ratio of single cells under control (355 mosmol/kgH2O) medium and directly reflect [Ca2+]i. Measurements under control medium were taken every 15 s for 15 min. A: trace of a single quiescent PRL cell displaying stable basal [Ca2+]i (observed in 111 of 186 cells). B: typical pattern of [Ca2+]i in a spontaneously active high-frequency oscillator (observed in 60 of 186 cells). C: a spontaneously active low-frequency oscillator (observed in 15 of 186 cells). These and all subsequent 340/380 ratios shown are from fura 2-AM-loaded, dispersed PRL cells. Traces shown are representative of those obtained in 186 PRL cells studied in 26 independent experiments.

Effects of reduction in medium osmolality on [Ca²+]_i. When medium osmolality was reduced from 355 to 300 mosmol/kgH₂O, [Ca²⁺]_i quickly increased (within 1-2 min) as indicated by the increase in 340/380 ratios (Fig. 4). Both oscillating and silent cells responded to the hyposmotic stimulus, and the traces presented in Fig. 4, A and B, are representative of 23 and 35 cells, respectively. The magnitude of [Ca²⁺]_i responses to hyposmotic medium varied among cells, reaching ratios of up to 6 and as low as 0.5 (Fig. 4, C and D, respectively). Of the 69 cells analyzed for [Ca²⁺]_i responses to hyposmotic medium, 55% responded with ratios up to 1, 25% responded with ratios up to 2, and the remaining 20% had peaks with ratios higher than 2. Most cells responded with a peak in [Ca²⁺]_i that typically lasted from 5 to 10 min and subsided either to an elevated plateau (45% of cells) or baseline levels (39% of cells). The remaining 16% of cells did not increase the ratio above baseline after exposure to hyposmotic medium but did respond to high-[K⁺] medium at the end of the experiment by increasing [Ca²⁺]_i.

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Effects of hyposmotic medium (300 mosmol/kgH2O) on [Ca2+]i in single PRL cells. Arrows indicate the estimated time in which the medium reached the cells. A: patterns of [Ca2+]i from a single spontaneously active PRL cell in response to hyposmotic medium. B: patterns of [Ca2+]i from a single quiescent PRL cell in response to hyposmotic medium. The magnitude of [Ca2+]i changes in response to hyposmotic medium varied among single PRL cells, from very strong (C) to slight increases in 340/380 ratio (D). E: mean ratio of cells from a single representative experiment in which medium osmolality was changed from 355 to 300 mosmol/kgH2O, switched back to 355 mosmol/kgH2O, and reexposed to 300 mosmol/kgH2O, followed by high-[K+] hyperosmotic medium (355 mosmol/kgH2O). Values are means ± SE (n = 10 distinct cells from a single experiment) at 1-min intervals. An increase in 340/380 ratio was observed after both hyposmotic stimulations, followed by a strong and rapid increase in ratios after exposure to high-[K+] hyperosmotic medium. Responsiveness to high-[K+] or hyposmotic medium at the end of all experimental runs was used as an indicator of cell viability.

Effects of Ca²+ deletion on PRL release and cell volume. To examine the involvement of extracellular Ca²⁺ in hyposmotically induced PRL release, we exposed PRL cells to hyposmotic medium devoid of CaCl₂ (Ca²⁺ deleted). Cells were initially perfused with normal hyperosmotic medium (355 mosmol/kgH₂O, with 2 mM CaCl₂ ), and after the switch to normal hyposmotic medium, PRL release was increased over threefold above the baseline. A significant (P <0.05) decrease in net PRL release over a period of 25 min was observed in hyposmotic Ca²⁺-deleted medium compared with the normal hyposmotic control (Fig. 5A). Experiments were also conducted with EGTA (2 mM) in the incubation medium to ensure a low extracellular Ca²⁺ concentration. These experiments produced results similar to those when Ca²⁺ was deleted from the incubation medium (data not shown). Prolactin cell volumes increased in proportion to the reduction in osmolality regardless of the presence or absence of Ca²⁺ (Fig. 5B). In both normal hyposmotic (i.e., containing Ca²⁺) and Ca²⁺-deleted hyposmotic media, RVD was not observed during the 30 min following a 15% drop in osmolality, as confirmed by regression analysis. As mentioned previously, RVD becomes measurable after 35 min in hyposmotic medium. In all treatments, cell volume increased ~10-15% above the baseline established in hyperosmotic medium, thus clearly indicating that deletion of Ca²⁺ from the incubation medium reduced hyposmotically induced PRL release but not cell swelling.

View larger version (25K): [in this window] [in a new window]   Fig. 5.   Effects of Ca2+-deleted hyposmotic medium (300 mosmol/kgH2O) on PRL release and cell volume measured from the same preparation of dispersed PRL cells. Horizontal bars indicate the estimated time in which the medium reached the cells. Ca2+-deleted hyposmotic medium was prepared by omitting 2 mM CaCl2 present in normal hyposmotic medium. A: the release of PRL increased in response to hyposmotic medium () but was significantly reduced in Ca2+-deleted hyposmotic medium (). In bar graph, values are means ± SE and indicate net PRL release change over 25 min (n = 5 and 7 for control and Ca2+-deleted, respectively). * Significantly different from control (300 mosmol/kgH2O) at P < 0.05. B: the change in PRL cell volume in response to hyposmotic medium () was not significantly different from that of Ca2+-deleted hyposmotic medium (). In bar graph, values are means ± SE and indicate net cell volume change over 30 min (values were computed from individual cells taken from at least 3 replicate experiments; n = 10 and 15 for control and Ca2+-deleted hyposmotic medium, respectively).

Effects of Ca²+ deletion on [Ca²+]_i. The changes in [Ca²⁺]_i in response to Ca²⁺-deleted hyposmotic medium clearly indicate the role of extracellular Ca²⁺ triggering hyposmotically induced PRL release. Exposure to Ca²⁺-deleted hyposmotic medium (300 mosmol/kgH₂O) immediately after normal hyperosmotic medium (355 mosmol/kgH₂O, with 2 mM CaCl₂) did not alter the 340/380 ratio in 9 of 11 cells. This pattern, followed by an increase in [Ca²⁺]_i in normal (i.e., Ca²⁺-containing) hyposmotic medium, is exemplified by the single cell recording in Fig. 6A. The other 2 of 11 cells responded to Ca²⁺-deleted hyposmotic medium with a transient increase in [Ca²⁺]_i. A switch from normal hyperosmotic medium to Ca²⁺-deleted hyperosmotic medium also produced either a transient increase (23%) or no change in [Ca²⁺]_i (77% of 52 cells). The subsequent exposure of these same cells to Ca²⁺-deleted hyposmotic medium did not produce a transient increase in [Ca²⁺]_i in 98% of cases. An example of a cell responding to Ca²⁺-deleted hyperosmotic medium, followed by no response to Ca²⁺-deleted hyposmotic medium, is shown in Fig. 6B. This cell also responded strongly to high [K⁺] in normal hyperosmotic media at the end of the experiment.

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Effects of Ca2+-deleted hyperosmotic and hyposmotic medium (355 and 300 mosmol/kgH2O, respectively) on [Ca2+]i in single PRL cells. Arrows indicate the estimated time in which the medium reached the cells and the treatment being perfused. A: [Ca2+]i trace of a single PRL cell that did not respond to Ca2+-deleted hyposmotic medium and recovered by increasing the 340/380 ratio after exposure to normal (Ca2+ containing) hyposmotic medium. This pattern is representative of 9 of 11 cells. B: [Ca2+]i trace of a single PRL cell that transiently and strongly increased [Ca2+]i after having been switched from normal hyperosmotic medium to Ca2+-deleted hyperosmotic medium but that did not change [Ca2+]i following exposure to Ca2+-deleted hyposmotic medium. This response was seen in 12 of 52 cells. After recovery, responsiveness to normal hyposmotic medium or high [K+] was observed in 75 and 87% of cells, respectively, although to different magnitudes.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

This is the first report to correlate changes in cell size and PRL release in an osmoreceptive model system, based on observations on the same population of PRL cells. The use of the tilapia PRL cell offers distinct advantages that are unavailable to researchers addressing similar questions in other osmoregulatory systems, such as the hypothalamo-neurohypophysial magnocellular systems that secrete vasopressin and oxytocin in mammals (5). Specifically, the tilapia PRL cell represents a model system in which the osmoregulatory output (PRL release) can be measured simultaneously with other parameters, such as cell size, that are involved in the osmoreceptive process.

The present findings indicate the importance of extracellular Ca²⁺ in hyposmotically-induced PRL release. In the absence of extracellular Ca²⁺, a reduction in osmolality and the subsequent increase in cell volume do not significantly increase PRL release. In the presence of extracellular Ca²⁺, however, an increase in cell volume seems to be the trigger for a rise in [Ca²⁺]_i and subsequent PRL release. Thus, in the activation of the osmoreceptive transduction pathway, the extent to which an increase in cell volume triggers a rise in [Ca²⁺]_i and subsequent PRL release relies on the availability of extracellular Ca²⁺. Therefore, extracellular Ca²⁺ entry would appear to be essential for the full response of PRL secretion in hyposmotic environments, although the participation of intracellular Ca²⁺ stores in this process remains to be examined.

In the present perfusion incubations, dispersed PRL cells responded within minutes to decreases in medium osmolality. Prolactin cells were preincubated in hyperosmotic medium (355 mosmol/kgH₂O) to show the effects of a physiologically relevant reduction in medium osmolality. The movement of a freshwater-acclimated tilapia to seawater produces a rapid increase in blood osmolality. Conversely, the transfer of a seawater-acclimated fish to freshwater elicits a rapid drop in blood osmolality. The degree of these changes depends to a considerable extent on the past experience of the animal with different salinities, and deviations between 290 and 450 mosmol/kgH₂O in the blood osmolality are commonly observed in tilapia subjected to transfer between freshwater and seawater (10, 33). Thus the changes of hyperosmotic medium (355 mosmol/kgH₂O) to hyposmotic medium (300 mosmol/kgH₂O) employed in this study are well within the range of blood osmolalities observed in vivo.

In the present study, a 15% reduction in osmolality elicited an increase in cell volume of roughly 15%, which returned gradually toward the baseline over the next 80 min. However, this cell volume decrease was slow, and it took 35 min of hyposmotic stimulation before a significant volume reduction could be observed. In many cells placed in hyposmotic media, the initial swelling due to water influx is followed rapidly by losses of ions and/or organic solutes that lead to RVD (19, 20). In many cell types, RVD is observed within 20 min of hyposmotic stimulation (2, 20, 21, 32, 37, 39). The absence of a rapid RVD in the PRL cell may reflect the smaller shift in osmolality used in this study (15% reduction in osmolality) compared with those typically used by others, i.e., 25-50% (2, 20, 21, 31, 32, 39). Therefore, in addition to testing a physiologically relevant 15% change, we subjected the cells to a 30% decrease in osmolality to examine PRL cell volume changes and PRL release. Indeed, after a 30% increase in cell volume, the RVD following a 30% reduction in osmolality was more accentuated than that observed after a 15% reduction, indicating that the degree of osmotic deviation may dictate the extent of cellular RVD.

No single mechanism for RVD is known to operate in all cells. A wide array of different ions, organic solutes, and pathways has been implicated in different cell types, and different transduction mechanisms linking the initial swelling and subsequent activation of transport pathways have been postulated (19). Many of these studies, however, are based on in vitro experiments that often extrapolate the conditions observed in vivo. In normal and tumor-derived GH4C1 and MMQ rat pituitary cells, for example, RVD was observed within 10 min after exposure to 27% hyposmotic medium (36). On the other hand, pancreatic -cells may exhibit RVD after a reduction in osmolality as low as 10% (23). In the latter case, however, the physiological significance of RVD is related to the regulation of insulin by blood glucose, rather than to osmoreception. If PRL cells are responding as osmoreceptors to physiological decreases in extracellular osmolality, changes in cell volume would need to be compatible with the time course of hyposmotically induced PRL response. The present study indicates that regulatory volume adjustments in the PRL cells occur only after extended periods (after 35 min when exposed to a 15% reduction in osmolality) or an extreme deviation (30%) in osmolality. The delayed or slow volume decrease in tilapia PRL cells exposed to small, physiological changes (15%) in extracellular osmolality may be of adaptive significance to the animal. Continued stimulation (hyposmotic extracellular environment) of the PRL cell maintains elevated rates of PRL secretion for up to 12 h (33). The absence of rapid volume regulation within the physiological range of osmolality indicates that volume regulatory changes in intracellular osmolyte pools do not occur during the time frame necessary for the peak in PRL release. Thus we believe that these findings provide significant evidence that PRL release in response to changes in osmolality is a physiologically important process and not a nonspecific outcome of cell volume regulatory mechanisms.

Evidence from our laboratory suggests that the effect of reduced osmolality on PRL release in the tilapia is mediated through a signal transduction system linked to changes in [Ca²⁺]_i. By measuring the uptake and loss of ⁴⁵Ca²⁺ in the RPD, Richman et al. (29, 30) have shown that intracellular Ca²⁺ metabolism is modified in response to extracellular osmolality and is directly linked to the stimulation of PRL release. Through the use of the Ca²⁺-sensitive fluorescent dye fura 2, we observed that [Ca²⁺]_i rises rapidly after a reduction in medium osmolality from 355 to 300 mosmol/kgH₂O. It was previously reported that 25% of tilapia PRL cells show spontaneous oscillatory activity, whereas the remainder are silent (10). The baseline of the oscillatory cells was elevated, whereas the amplitude of oscillations was decreased at the onset of hyposmotic stimulation. This report is consistent with our current findings, in which the majority of cells in hyperosmotic medium were silent (60%). Regardless of the oscillation pattern, PRL cells responded to hyposmotic medium by elevating [Ca²⁺]_i. This rise in [Ca²⁺]_i varied in magnitude among experiments. Although this variation may represent natural fluctuation in PRL cell responsiveness, it may be a reflection of variations in cell preparation. As an indication of cell responsiveness, high-[K⁺] medium was introduced at the end of the experiments, and >90% of the cells studied responded with a sharp increase in [Ca²⁺]_i. In addition to silent and spontaneously active cells, low-frequency oscillation patterns are described for the first time in the tilapia PRL cell.

Intracellular Ca²⁺ oscillations have been described in many cell types, including mammalian PRL cells (35, 38) and goldfish GH cells (43). Oscillations have been implicated in modulating specific signal transduction pathways (15, 28, 35), although there is little evidence describing how this modulation occurs. In goldfish somatotropes, 88% of cells were quiescent, and low-frequency oscillations were not reported (43). In mammalian PRL cells, four [Ca²⁺]_i oscillation patterns have been described (38). Some cells (~25%) showed no spontaneous [Ca²⁺]_i oscillations, whereas the remainder were divided into low-amplitude, high-frequency oscillators (42%), high-amplitude, high-frequency oscillators (25%), and low-frequency oscillators (8%).

This study also examined whether the short-term rise in [Ca²⁺]_i is dependent on an increase in cell volume. Alternatively, an increase in cell volume may directly lead to an increase in PRL release, independent of extracellular [Ca²⁺]. Indeed, the osmotically induced release of hormones from rat anterior pituitary cells seems to occur independently of extracellular Ca²⁺ availability (36). In such cases, however, large deviations in osmolality are required to induce a short-lived burst of the hormone release, and cell swelling may induce a universal secretion of exocytotic material that may represent a pathological response (36). The entry of extracellular Ca²⁺ following cell swelling has been described in several cases (21, 37, 41). Increases in [Ca²⁺]_i are often attributed to activation of a RVD mechanism that does not appear to be operating in the tilapia PRL cell. For example, in Xenopus renal A6 cell lines, removal of extracellular Ca²⁺ from the medium prevented a full RVD (37). In the present study, cell volume was maintained at an elevated level during hyposmotic stimulation in both Ca²⁺-deleted and normal media, suggesting that extracellular Ca²⁺ does not affect cell volume.

The deletion of Ca²⁺ from hyposmotic medium significantly reduced the hyposmotically induced PRL release, indicating that extracellular Ca²⁺ is a crucial component for the initiation of this signal transduction. The near absence of release in response to hyposmotic Ca²⁺-deleted medium has been previously described from intact RPDs perfused under similar conditions (29). The small (<2-fold) response in PRL release observed in Ca²⁺-deleted hyposmotic media suggests the participation of other second messenger systems and intracellular Ca²⁺ stores in the transduction pathway. Alternatively, this transient response may be a reflection of compensatory release of Ca²⁺ from intracellular stores, because it occurred after extracellular Ca²⁺ was depleted with either EGTA or with nominally Ca²⁺-free media. In some cells, a transient peak in [Ca²⁺]_i was observed after exposure to hyperosmotic or hyposmotic Ca²⁺-deleted and Ca²⁺-depleted media, followed by a decline or return to the baseline ratio. Because these transient responses occurred in both hyper- and hyposmotic media deprived of Ca²⁺, they may represent compensatory Ca²⁺ release from intracellular stores. This compensatory release has also been suggested in goldfish somatotropes (16), but the mechanism underlying such a response remains to be clarified.

The notion that stretch-sensitive ion channels may act in the transduction of changes in transmembrane osmotic pressure is particularly attractive in light of the fact that a 1 mosmol/kgH₂O decrease in medium osmotic pressure increases membrane tension to a degree that is well within the range to which stretch-sensitive ion channels are responsive (24). Changes in cell volume in response to an osmotic stimulus could account for the osmosensitivity of certain endocrine and neuroendocrine pathways. Changes in cell volume would lead to the activation or inactivation of stretch-sensitive ion channels that are linked to secretory mechanisms. In the case of tilapia PRL cells, this transduction process would link a reduction in extracellular osmolality to an increase in PRL, consistent with the osmoregulatory role of this hormone. Although the importance of cell volume and extracelluar Ca²⁺ to this process has been identified, further study of the nature and control of this transduction pathway is necessary. We are currently testing the hypothesis that the increase in cell volume following hyposmotic stimulation will increase the open probability of putative stretch-activated channels that allow extracellular Ca²⁺ entry and a subsequent increase in PRL release (Ref. 32a; see p. C1290 in this issue).