Use of pituitary cells in primary culture to study the regulation of gonadotropin Hormone (GtH) secretion in rainbow trout: setting up and validationng thesystem as assessed by its responsiveness to mammalian and salmo gonatropin releasing hormone

the by its responsiveness to To study the regulation of gonadotropin secretion in rainbow trout in vitro, a method for preparing primary cultures of dispersed pituitary cells is described. Cells were dispersed by collagenase 0.1% in Hank’s saline solution for 20 hr at 12” and a high yield of viable cells was obtained. Attempts to improve cell functioning were made by varying culture condi- tions (density of cells initially plated, age of the culture). Cell functioning was assessed by their ability to respond to increasing doses of mammalian and salmon GnRH. Pituitaries were collected from spermiating males whose pituitaries are known to be sensitive to mammalian GnRH in viva. Using 96well plates, optimal conditions for good biological activity, are initial plating with 6.2 x 104 cells, incubation with GnRH for 24 hr on the third day after plating. In these conditions mammalian analog and salmon GnRH induced an increase in GtH release for doses ranging from 10-g to 10-C M. The GtH released during the GnRH incubation period does not decrease the sensitivity of the system since addition of 20 ng of GtH at the beginning of incubation does not modify the response profile.

To study the regulation of gonadotropin secretion in rainbow trout in vitro, a method for preparing primary cultures of dispersed pituitary cells is described. Cells were dispersed by collagenase 0.1% in Hank's saline solution for 20 hr at 12" and a high yield of viable cells was obtained. Attempts to improve cell functioning were made by varying culture conditions (density of cells initially plated, age of the culture). Cell functioning was assessed by their ability to respond to increasing doses of mammalian and salmon GnRH. Pituitaries were collected from spermiating males whose pituitaries are known to be sensitive to mammalian GnRH in viva. Using 96well plates, optimal conditions for good biological activity, are initial plating with 6.2 x 104 cells, incubation with GnRH for 24 hr on the third day after plating. In these conditions mammalian analog and salmon GnRH induced an increase in GtH release for doses ranging from 10-g to 10-C M. The GtH released during the GnRH incubation period does not decrease the sensitivity of the system since addition of 20 ng of GtH at the beginning of incubation does not modify the response profile. o 1986 Academic In previous work we have studied the regulation of GtH secretion by mammalian gonadotropin releasing hormone (mGnRH) in vivo in carp and trout (Weil et al., 1975(Weil et al., , 1978. The use of in vitro systems for studying the regulation of gonadotropin hormone (GtH) secretion provides a number of advantages over in viva experimentation. In such a system the pituitary is isolated from neural influences and the action of test substances can be easily studied.
The use of dispersed pituitary cells is more suitable than the use of whole organs or half organs because it requires less donor animals and provides homogeneous r The results were presented in part as a poster at the 7th Conference of the European Society for Comparative Physiology and Biochemistry, Barcelona, Spain. August 26-28, 1985(Weil, 1985. replicates since the pituitaries are pooled. These dispersed cells can be used in monolayer culture (static model) or anchored to beads and used in a perifusion system (dynamic model). In the former case each well is used for a test substance whereas in the latter, cells are usually used for several tests after a recovery period. For this reason we preferred to use a primary culture in a static model. However, dispersion and culture conditions can affect cell functioning and therefore the optimization of these parameters can greatly increase the usefulness of this technique. In the present work, we decided to study a few such parameters. Cell functioning was assessed by the ability of gonadotrophs present in the primary cell culture to respond to mammalian GnRH analog (D-Ala6 LHRH) and salmon GnRH (sGnRH) whose structure has been recently determined (Sherwood et al., 1983). Trials were conducted on adult spermiating trout which are known to have high pituitary GtH levels and to be responsive to mGnRH in viva (Weil et al., 1978).

Animals.
All the studies were conducted on 2-yearold spermiating males or on those just beginning regression of a spring-spawning strain reared at the fish farm of Gournay (Oise).
Cell preparation.
Pituitary glands were removed from 10 to 20 fish and placed in ice-cold Hank's balanced salt solution (GIBCO Laboratories), buffered with 20 mM Hepes (P-L Biochemicals, Inc.) and 9 mM sodium bicarbonate as indicated by Wolf (1982) and supplemented with penicillin (100 U/ml), streptomycin (100 pg/ml), and fungizone (0.25 l&ml) (GIBCO Laboratories). The buffered saline solution had an osmotic pressure of 300 mOsm/kg and a pH of 7.5, characteristic values of rainbow trout blood plasma. Pooled glands were rinsed with the buffered saline solution and chopped into small pieces with a surgical blade. The tissue fragments were transferred for dispersion to a siliconized glass-stoppered flask filled with 20-30 ml of Hank's buffered solution containing 0.1% (wt/ vol) collagenase (Boehringer Mannheim) and 3% (wti vol) BSA (RIA grade--Sigma).
The mixture was incubated at 12" for 20 hr and during this period dissociation was mec,hanically aided by aspirating the fragments three times in a siliconized Pasteur pipette. Cells were harvested by centrifugation at 200g for 10 min at 15" and washed twice with the culture medium. The culture medium consisted of RPM1 (GIBCO Laboratories) containing 20 mM Hepes, 9 mM NaHCo, and had a pH of 7.5 and an osmotic pressure of 300 mOs'm/kg. This culture medium was supplemented with 2% of a serum substitute, ultroser G or SF (IBF) and with the same dosage of antibiotic-antimycotic as previously described. Cells were resuspended in 1 ml of culture medium. Cell counting and cell viability were determined by mixing 40 ~1 of the cell suspension with 10 )*l of 0.1% trypan blue (BDH chemicals) in 0.01 &&phosphate buffer (pH 7.3) and examining the cells in a Thoma hemocytometer. After counting, the cells were diluted to 2.5 X lo5 cells or 1.25 x lo5 ceils/ml of medium and 250 pl of suspension (i.e., 6.2 X IO4 cells or 3.1 x IQ4 cells) were plated in each well of 96-well plates (NUNC). The dishes were maintained at 18".
Microscopical observation of the cells was performed each day after plating, After 2, 3, or 5 days of culture (preincubation), the cells were washed twice with 250 yl of ultroser and antibiotic-antimycotic free culture medium and subsequently reincubated in 250 pl of this medium with or without increasing concen-trations of synthetic GnRH. A synthetic analog of LHRH, des-Glyia o-Ala6 LHRH ethylamide (Beckmann), as well as a synthetic salmon GnRH (Breton er al., 1984) synthesized according to the sequence described by Sherwood et al. (1983) were used at different doses (10-n to 10-G M). At the end of the incubation period (at 1K' for 6 to 24 hr) the media were collected, centrifuged, and stored at -20" for subsequent determination of GtH. The intracellular concentration of GtH was determined in some experiments. After the removal of the incubation media, culture dishes were washed twice with buffered saline solution and 200 ~1 of radioimmunoassay buffer containing 0.1% (voVvo1) of Triton X-100 (Serva) was added to each well for 3 to 4 hr at 18" until complete breaking up of the cells was achieved as assessed by microscopic observation. The cell lysate was stored at -20 until GtH measurement. Results were expressed as nanograms of GtH released or present in the cell extracts per number of cells initially plated for the specified time intervals.
The overload of the medium by GtH was made by adding 20 rig/well of a pure salmon GtH (Breton et al., 1978).

GtH measurement.
A radioimmunoassay (RIA) established for assaying plasma and pituitary GtH content was used as described by Breton et al. (197ib), This RIA used a pure salmon GtH as a standard against guinea pig anti-trout GtH. Assessment OF this RIA for measuring GtH contained in the cells and in the culture medium was performed by comparing the parallelism of the usual standard curve and that of the standard curve containing 0.1% Triton or culture medium in a logit versus log dose representation. Statktical analysis.
The data were analyzed using an analysis of variance and comparisons between means were made using the Newman-Kerns test.

Dissociation and Culture Procedures
The exposure of pituitaries to collagenase at 12" for 20 hr produced a high yield of single cells but some clusters were also observed. Cell viability tested immediately after dissociation was high and the mean yield of live cells was 88.98 -t 1.49% for the three trials reported in the present work.
Microscopically, freshly dispersed cells had spherical shapes. After 24 hr in cuhure at 18" they still had this shape, and had begun division. By Day 2, the cells had flattened, spread, and begun to be at-tached. The entire area of the dish was not covered with cells when initial plating is done with 6.2 x 104 cells. On the other hand by Day 3 cells were firmly attached and the entire area was covered with cells and by Days 4 and 5 the wells had the same appearance. By Day 3, when wells were initially plated with half the quantity of cells (3.1 x 104 cells) the density of spread cells was equivalent to that observed on Day 2 after an initial plating of 6.2 x 104 cells per dish.

Assessment of the Biological Activity of
Pituitary Gonadotrophs Present in the Primary Culture Cells (6.2 x 104) were initially plated in control and experimental dishes, The biological activity of the pituitary gonadotrophs was tested by their ability to respond to GnRH stimulation after 3 days of preincubation.
At that time cells were firmly attached, had spread and had covered the entire dish area.
(I) Time course of GtH release by control and IO-8 M D-Ala6 LHRH stimulated plates (Fig. I). GtH release was studied 6, 12, and 24 hr during the incubation period in control and experimental quadruplicates. GtH levels rose slowly in the media collected from control dishes whereas a marked increase was noticed in GnRH stimulated ones. However, GtH levels were highly different from control ones only at 24 hr (P < 0.005). As a consequence a period of 24 hr was used in subsequent incubations.
(2) GtH release with various concentrations of D-Ala6 LHRH or sGnRH by cultured pituitary cells (Fig. 2). Quadruplicates of monolayer cultures were submitted for 24 hr to increasing doses (0 to 10-G M) of D-Ala6 LHRH and sGnRH, respectively. Doses ranging from lo-12 to 10 -10 M of both types of GnRH did not significantly change GtH release in comparison with controls (dose 0) and mean GtH levels were 7.62 t 0.39 ng/6.2 x 104 cells. On the other hand, doses ranging from IO-9 to IO-6 M induced a significant increase compared with controls and mean GtH levels were, respectively, 16.34 -t 0.96 ng/ well with o-Ala6 LHRH and 16.69 & 0.56 rig/well with sGnRH. tested on preadruplicate wells either 2, 3, and 5 days after an initial plating with 4.2 x 104 cells or 3 days after an initial plating with 3.1 x 104 cells. In each experimental condition the biological activity of the gonadotrophs was assessed by studying basal GtH release (GtH present in the dose 0 wells at the end of the incubation period), cell GtH content (measured at the term of the incubation period), and the responsiveness to various doses of sGnRH (GtH present in the GnRH stimulated wells at the end of the incubation period).

Injluence of the Preincubation
(I) Basal GtH release and cell GtH content (Table I). As the different doses of sGnRH treatment did not influence cell GtH content (P > 0.05), this parameter is expressed by the mean of all the wells.
Similar basal GtH levels and cell GtH contents were observed 2 days after an initial plating with 6.2 x 104 cells and 3 days after an initial plating with 3.1 x 104 cells. These values were significantly lower than those noticed in wells initially plated with 6.2 x 104 cells and preincubated for 3 and 5 days. These observations are in agreement with the microscopic observation of cell development in the dishes and related above.
(2) Responsiveness to GnRH (Fig. 3). Following an initial plating with 6.2 x 104 cells, only the highest tested dose of sGnRH (lo-" M) induced a significant Gt release compared to basal level when incubation is performed after 2 and 5 days of preincubation.
On the other hand, a significant GtH release was observed for doses ranging from 10-s to 10-G M when incubation is performed after 3 days of preincubation. As mentioned above, increase in GtI-I release was gradual and began at a dose of 10-9 M. On Day 3 basal and sGnRW induced GtH release presented good homogeneity among replicates and the global CQefficient of variation was 18%. On the other hand, 5 days after plating these two parameters presented heterogeneous values. The global coefficient of variation was ~63% and it is the reason why, a significant .increase in GtH release was noticed only for the highest tested dose (10-b M).
Following an initial plating with a low b Corresponds to the response to the 0 dose of sGnRH (see the results). c Values are the mean of control and sGnRH treated wells since the different doses of sGnlW did not influence cell GtH content (see the results).
*,** Values significantly different from wells initially plated with 6.2 x lo4 cells and studied after 3 and 5 days of preincubation. *P < 0.02, **P < 0.001. density of cells (3.1 X lo4 cells) a response to GnRH was observed on Day 3 only for the highest tested dose (10-e M).
Influence of an Overload of GtH in the Media on Gonadotrophs Responsiveness to Varying Doses of sGnRH Three days after plating, quadruplicate dishes containing 6.2 x 104 cells initially were submitted for a 24-hr incubation period to 20 ng of GtH plus increasing doses of sGnRH (O-lo-6 A4) or to sGnRH alone (same culture as previously reported Table l- Fig. 3). Basal GtH levels (dose 0) are 27.21 t 0.89 ng and 7.72 +-0.64 ngldish, respectively. Figure 4 shows sGnRH induced GtH release expressed as increase over basal levels. An overload of GtH did not modify sGnRH induced GtH release. GtH patterns are similar in control and experimental dishes: a significant increase in GtH release was obtained for sGnRH doses ranging from lo-8 to lo-GM and the values obtained are not significantly different.

DlSCUSSlON
The use of different in vitro systems for studying the regulation of GtH secretion is now commonplace in mammals. In teleost fish some attempts have been made. Authors used either hemipituitaries in shortterm incubation (carp: Breton et al., (1971aBreton et al., ( , 1972 or whole organs in short term incubation (rainbow trout: Crim and Evans, 1980) and in organ culture (rainbow trout: Fghraeus-van Ree et al., 1983). Recently pituitary fragments in a perifusion system have been described in goldfish . The use df dispersed pituitary cells of the whole pituitary has been also reported either in short-term incubation (rainbow trout: Fihraeus-van Ree et al., 1982a,b) or in primary culture in a static model (rainbow trout: Fkhraeusvan Ree et al., 1982b;carp: Ribeiro and Ahne, 1982;Ribeiro et al., 1983) and in a perifusion system (goldfish: Chang et al., 1984). Recently a primary culture of purified pituitary gonadotrophs have been set up for African catfish (Leeuw et al., 1984). The present paper describes a technique of dispersion and culture of pituitary cells in rainbow trout. The enzymatic collagenase procedure that we used produces a high yield of living cells which are firmly attached on Day 3 of culture and are able to be responsive to GnRH. Using another enzymatic dispersion technique, F$hraeusvan Ree et al. (1982b) have also obtained dispersed pituitary cells of rainbow trout able to be kept in primary culture and to be responsive to GnRH. In our study the response to GnRH is a function of incubation time, as already described for short-term  Influence of an overload of 20 ng of GtH on cells responsiveness to various doses of sGnRH during the 24hr period of an incubation performed 3 days after plating. Wells were initially plated with 6.2 x lo4 cells. The results for aGnRH stimulated GtH release without a GtH overload come from the same cultures as shown in Fig. 3. Values are 2 t SEM (n = 4) and are expressed as GtH increase compared to basal level (GtH measured in doses 0 wells), incubation of whole pituitary in trout (Crim and Evans, 1980) or for carp pituitaries in monolayer culture (Ribeiro el al., 1983). We found that a 24-hr incubation period is necessary to observe a GtH release highly different from control wells. Furthermore, we demonstrated for the first time that GnRH stimulates GtH release in a dose-related fashion from isolated pituitary cells of rainbow trout. We observed a response to GnRH for doses ranging from 10-9 to 10-C h4 of both tested peptides, the salmon GnRH and the mammalian LHRH analog (n-Ala6 LHRH).
The average increase in comparison with basa1 release (dose 0) was approximately two and a half times higher. As the two tested peptides are active in the same dose range we think that they are equipotent in our in vitro system. To our knowledge our work is the second one comparing, in vifvo the action of mammalian and salmon GnRH on GtH secretion. The first trials were made in goldfish with superfused pituitary fragments colIected from postspawning males by Mackenzie et al. (1984). Because both peptides were active in the same dose range, elicited response profiles which were similar in time course and stimulated maximal responses of equal magnitude, these authors concluded also that the peptides are approximately equipotent in their system. However, in viva in goldfish, Peter et al. (1985) have recently demonstrated that D-Alas LHRH is more active than sGnRH in terms of magnitude and duration of the GtH response. These authors suggest that in vivo the decreased degradation of LHRHa may be the primary factor determining its potency. As, in our static in vitro system, we did not study the degradation of both peptides as welt as their pituitary receptor binding affinity during the 24-hr incubation period, we cannot determine the factors contributing to their equal efficiency.
As the gonadotrophs present in our primary cultures were sensitive to GnRH, we attempted to optimize the conditions of culture to increase the sensitivity of our system and to decrease the numbers of fish required. First the'time was varie&between plating and incubation with GnRH. Second, wells were initially plated with half the quantity of cells. It appears that the responsiveness of our system is dependent on the number of cells present in the dish at the time of GnRH Iincubation. In fact, when the den&y of cells was weak after either a low initial plating (3.1 x 104 cells) or when cells are not fully grown and Ifirmly attached on Day 2 after an initial plating with 6.2 x 104 cells, only the highest dose of sGnRH (IO-6 M) induced, at the term of the 24-hr period of incubation, an increase in GtH release, in both cases,'the'smalI number of cells was assessed' by microscopic observation and corroborated by the low cells GtH content and basal Gtsln release in compar&n~with those tibserved in wells initially plated with 4.2 x 104 cells and stimulated on Day 3. A longer duration of incubation could have been sufficient to increase the sensitivity of this in vitro system and to allow us to use a smaller amount of cells. However, by increasing the duration of incubation, we may have encountered some problems related to the age of the culture. Indeed, we noticed heterogeneous GtH release in control wells (dose 0) as well as in sGnRH treated wells when incubation with GnRH is performed 5 days after plating. As a consequence, this heterogeneity caused a decrease in the sensitivity of our system since only the highest dose of GnRH, i.e., 10-e M, induced a significant increase in GtH release. However, by microscopic observation no changes in cell morphology were observed and the plates were not invaded by fibroblast proliferation as already described in mammals (O'Conner et al., 1980). In mammals, it is well known that in monolayer culture, basal as well as LHRH stimulated LH and FSH release decrease with culture time (Vale et al., 1972;Tixier-Vidal et al., 1973). However, this decrease appears more than 1 week after plating. In our work, on Day 5 basal GtH release during the incubation period, although lower, is not significantly different from the one seen on Day 3. The fact that spontaneous GtH release as well as sGnRH induced GtH release are heterogeneous between replicates, leads us to think that this heterogeneity is not due to the damage of GnRH receptor. This heterogeneity could be due to a decrease in the synthesis capacity of the cells with time. Additional studies on the kinetics of the synthesis capacity of gonadotrophs present in the primary culture are necessary to confirm or refute this hypothesis.
Therefore, optimal conditions for studying the regulation of GtH secretion by GnRH in mature male rainbow trout in 96well plates are initial plating with 6.2 x 104 cells, incubation with GnRH 3 days after plating and a duration of incubation of 24 hr. However, the presence of released GtH in the medium for 24 hr could have interfered with the sensitivity of the system. As the pattern of sGnRH induced GtH release as well as the sensitivity of the system were not modified when we added GtH in the wells, we conclude that this static in vitro system is convenient for a certain type of studies on regulation of GtH synthesis. Preliminary work in progress, unreported here and conducted on males and females at different stages of the sexual cycle corroborate this fact since the gonadotrophs respond to GnRH in a characteristic pattern as predicted from in vivo experiments (Weil et al., 1978). Further investigations are in progress to clarify this variable pituitary responsiveness to GnRH throughout the sexual cycle and to study the possible role of different steroids on its variation.