A 66-kda Protein of Bovine Hypophyseal Pars Tuberalis Induces Luteinizing Hormone Release from Rat Pars Distalis Introduction

In this study, evidence for a factor secreted by bovine hypophyseal pars tuberalis that stimulates luteinizing hormone (LH) release from rat pars distalis cells is shown. The secretion products of bovine pars tuberalis cells into the culture medium were assayed on dispersed rat pars distalis cells in 30 min incubations and superfusion experiments. The culture medium from pars tuberalis total cell populations, added at a dose of 6 μg per tube, induced the greater LH release from pars distalis cells, without effect on follicle stimulating hormone (FSH) release. After pars tuberalis cells separation on a discontinuos Percoll gradient, only the culture medium of cells from 50 and 60% strength Percoll were able to release LH from rat pars distalis cells. Therefore, cell fractions from 50 and 60% strenght Percoll were cultured together. To elicit maximal LH release (6 times the basal output), with the addition of 2 μg of pars tuberalis protein was required , suggesting that these cells produce the factor or factors which affect pars distalis gonadotrope cells. After applying the pars tuberalis culture medium on 12% SDS-PAGE, the band with biological activity was that of 66-kDal. Fifty ng protein of its eluate released almost 9 times the basal output of LH from pars distalis cells. Results suggest a modulating effect of a protein from the bovine pars tuberalis on rat cultured gonadotrope cells from the pars distalis.


Introduction
Pars tuberalis is a structurally distinct region of the adenohypophysis and it has been studied in rat, human, ovine and other species (Gross et al., 1984;Skinner et al., 1992).Pars tuberalis of most mammalian species contains specific cells which are structurally and functionally different from pars distalis cells (Bock et al., 2001).Pars tuberalis cell types include the follicular cells, pars distalis cells and the so-called pars tuberalisspecific cells.The latter are distinct endocrine cells dis-playing melatonin receptors (Wittkowski et al., 1999).
It is generally accepted that LH and FSH are stored and secreted from different secretory granules within pars distalis bihormonal cells (Thomas and Clarke, 1997;Farnworth, 1995).However, numerous gonadotropes have been detected in the anteroventral region of the pars tuberalis (Skinner and Robinson, 1996).It has been hypothesized that LH is released from pars tuberalis gonadotropes and diffuses into the medial eminence where it suppresses LHRH release from the terminals of the LHRH neurons, either by a direct action or by an action of other neuronal terminals (Nakazawa et al., 1991).Few LHRH fibers have been reported to be present in the adjacent pars tuberalis of the rat pituitary by fluorescence immunohistochemistry for LHRH and electron-microscopy.It has been postulated that the LHRH message might be transmitted not only by the established hypophyseal portal vein system but also via the folliculo-stellate cells in the pars tuberalis to aid in LH release modulation (Mabuchi et al., 2004).Due to the close spatial relationship between pars tuberalis cells and the hypotalamic-hypophyseal portal capillaries, it has also been suggested that the portal circulation might be the avenue for the transport of the putative secretory product(s) from pars tuberalis, thus reaching the pars distalis (Dellmann et al., 1974;Fitzgerald, 1979).On the basis of this anatomical arrangement, Dellmann et al. (1974) hypothesized that pars distalis is a target organ for pars tuberalis secretion(s).
A few studies have shown the possibility that factors released by pars tuberalis regulate the activity of lactotrophs in pars tuberalis.It has been reported that co-culture of ovine pars distalis cells with pars tuberalis cells causes prolactin secretion, with an increase of 50-100% above control levels over 24 h (Hazlerigg et al.,1996).We have demonstrated that conditioned medium of cultured bovine cells of the pars tuberalis for 48 h has a potent stimulatory effect on cell prolactin and GH secretions by pars distalis cells (Lafarque et al., 1998;2004).
Although the nature of the secretory product(s) of the pars tuberalis-specific cells has not yet been clarified, for practical reasons, previous authors have designated the unidentified pars tuberalis hormone(s) as tuberalin(s) (Morgan et al., 1996).Pars tuberalis-specific cells synthesize the common a subunit of the pars distalis glycoprotein hormones.It has been suggested that tuberalin II, a 21 kDa protein secreted by a predominant population of pars tuberalis-specific cells (type 3 cells) would correspond to the beta chain of a specific glycoprotein secreted by those pars tuberalis-specific cells (Guerra and Rodriguez, 2001).However, the endocrine function of pars tuberalis is currently unknown.
The aim of this study was to establish whether pars tuberalis cell secretions can affect LH and FSH release from pars distalis cells.With this objective, bovine pars tuberalis cells were isolated by Percoll gradient and cultured, and the culture medium was subsequently used to stimulate LH and FSH release from rat pars distalis cells.

Animals
Adult male Holtzman rats (250 g) bred in our animal facilities and maintained under conditions of con-trolled humidity and temperature (21-23 o C) and a 12-h light/dark cycle were used.Tap water and pelleted rat chow (Cargill SACI, Argentina) were available ad libitum.The bovine partes tuberalis were collected from young bulls killed at a local abattoir, between 8:00 and 9:00 a.m., throughout the year.

Chemicals
The following drugs were purchased from Sigma (St.Louis, MO, USA): 199 culture medium, crude trypsin, soybean trypsin inhibitor, collagenase type IV, DNAse, penicillin G, streptomycin, nistatine and bovine serum albumin fraction V (BSA).All other chemicals were of reagent grade and were provided by Merck Laboratory, Buenos Aires, Argentina.LH and FSH kits were kindly provided by NIADDK of NIH (Dr.Raiti, NIADDK Rat Pituitary Hormone Distribution Program).

Pars tuberalis cell culture
Primary cultures of bovine pars tuberalis were generated from 10-12 glands.The pars tuberalis tissues were dissociated in 199 medium containing 0.25% crude trypsin by stirring at 37ºC and assisted by repeatedly drawing tissue fragments into a siliconized pipette.After 30 min 0.2 mg/mLtrypsin inhibitor was added.The suspension was centrifuged at 100 x g for 10 min at 4ºC.The cell pellet was washed twice in 10 mL of 199 medium containing 2% penicillin G, 3% streptomycin and 0.04% nistatine and centrifuged.Then, the cell pellet was resuspended in the above indicated medium.Cell viability was determined to be 90% by trypan blue exclusion test.Cells (2 x 10 6 cells/mL) were cultured in 199 medium supplemented with antibiotics at 37ºC in 95% air-5% CO 2 for 24 h.The medium was removed, replaced by fresh medium with antibiotics and cells were cultured for additional 24 h.After that, the culture medium was replaced again, as described, and cells were cultured for 48 h in the same conditions.Finally, the culture medium was aspirated, centrifuged to remove cellular debris, supplemented with a mix of protease inhibitors and used to stimulate pars distalis cells.

Percoll gradient
Due to the existence of different cell populations in bovine pars tuberalis, a density gradient centrifugation of pars tuberalis cells was performed on a discontinuous Percoll gradient to identify the cells whose secretions could be responsible for LH and FSH release from pars distalis cells.Stock Percoll was diluted 1:10 with Hanks' balanced salt solution and then diluted again to different percentage (30, 40, 50, 60 and 80%) strengths of Percoll with Hanks' solution.Starting with 80% Percoll, 1.5 mL aliquots of decreasing strength were layered into a 12 mL conical centrifuge tube up to 30% Percoll.After dissociation and culture for 24 h twice as indicated above, pars tuberalis cells were layered above the 30% Percoll in 1 mL of Hanks'solution.The tube was then centrifuged at 400 x g for 15 min at 4ºC.Cells were harvested from each strength of Percoll, washed twice with 199 medium, tested by trypan blue test and cultured (2 x 10 6 viable cells/mL) for 48 h as indicated above.Cell viability was of 90-95% in all Percoll fractions.The collected culture media were used for subsequent experiments.

Dispersed pars distalis cells
Adult male rats were killed by decapitation.The brains and pituitaries were rapidly dissected out.The anterior lobes were separated from the neurointermediate lobes.Pars distalis cells were dispersed as previously described (Spinedi et al., 1984).Briefly, each pars distalis was cut into four pieces and placed into 1 mL 199 medium containing 0.2% collagenase.During shaking at 37ºC under 95% O 2 -5% CO 2 atmosphere, remaining fragments of tissue were repeatedly (30-40 times) passed through a siliconized Pasteur pipette.Three minutes before the procedure was finished, 0.1 mg/mL DNase was added.Then, the suspension was centrifuged at 100 x g for 10 min at 4ºC.The cell pellet was washed twice in 5 mL 199 medium containing FIGURE 1. LH (A) and FSH (B) release by dispersed cells from rat pars distalis that were incubated in vitro with culture medium of all pars tuberalis cell populations (dose-dependent effect).The μg of protein represents the amount of pars tuberalis protein in 500 μl of final volume of incubation.Bars represent mean ± SEM of duplicate measurements from four independent experiments with 3 tubes per point per experiment.Letters indicate values with significant differences vs basal value, a P < 0.01 and b P <0.001.0.25% BSA, 20 mg/L penicillin G and 30 mg/L streptomycin, centrifuged as described and finally resuspended in the same medium.After viability determination, 1 x 10 5 cells/tube were placed in 0.5 mL of 199 medium and stabilized for 6 h at 37ºC under 95% O 2 -5% CO 2 atmosphere.Then, the medium was replaced by 0.45 mL of fresh medium and cells were stimulated by addition of different doses (0.03-15 μg protein), dissolved in 50 μL of medium from cultured pars tuberalis cells.Basal secretions of LH and FSH were obtained by pars distalis cells incubation containing only fresh 199 medium with 0.25 BSA and antibiotics.Tubes were incubated in a shaking bath for 30 min at 37ºC gassed with 95% O 2 -5% CO 2 .After centrifugation, the supernatants were kept at -70ºC until LH and FSH determinations.Protein concentrations were determined by the method of Bradford et al. (1976).

Superfusion of pars distalis dispersed cells
Cells (2 x 10 6 viable cells) were resuspended in Biogel P2 (200 -400 mesh), previously hydrated with 199 culture medium at 4ºC for 12 hs.This cell preparation was packed into a plastic column and superfused with fresh 199 medium supplemented with 0.25% BSA, 20 mg/L ascorbic acid and antibiotics, with a flow of 0.75 mL/min for 10 min.Then, cells were stimulated with 1 mL 28 mM KCl in the medium and the superfusion was continued with medium alone.Fractions were collected every 2 min, for 20 min.After that, the column was washed with fresh 199 medium and cells were stimulated with 1 mL of culture medium from pars tuberalis cells of 50-60% strenght Percoll containing 3 and 0.3 μg protein per mL.Fractions were collected for 20 min, every 2 min.Collected fractions were stored at -70ºC until LH measurement.

SDS-PAGE
Slab gel electrophoresis was performed using the method of Laemmli (1970).Molecular weight markers were visualized by Coomassie blue staining, followed by destaining in glacial acetic acid/methanol/water (10:40:50, v/v/v) at room temperature.
After protein separation by 12% SDS-PAGE of medium from cultured pars tuberalis cells from 50+60% Percoll gradient, individual bands were cut and the proteins contained in each piece of gel were eluted by incubation in phosphate-buffered saline (PBS) pH 7.5 at 4ºC for 4 h, in a shaking bath.After concentrating the eluted protein by Microcom (OMEGA) filters, their bio-logical activity was tested by incubations on pars distalis cells by determining LH release.

LH and FSH radioimmunoassays
LH and FSH were measured in duplicate by double antibody radioimmunoassay (RIA) using a kit provided by the NIADDK of NIH, as recommended.Results were expressed in terms of rat LH or FSH RP-2 standard preparations.The assay sensitivity was less than 5 ng/ mL and the inter-and intra-assay coefficients of variation was less than 10.0.

Statistical analysis
Data are expressed as mean ± SEM.The statistical significance of the differences between means was assessed by one-way ANOVA followed by Tukey post-test.A P value of <0.05 was considered significant.

Results
The response of rat pars distalis cells to culture medium from total pars tuberalis cell populations (without Percoll separation) measured as the release of LH and FSH is shown in figure 1.The concentration of released LH was proportional to the amount of protein in the pars tuberalis culture medium (50 μL) until 6 μg of protein were present in the final volume of pars distalis cells incubation (500 μL), where the maximal amount of LH was detected.With higher protein concentrations the LH release was less than with 6 μg of protein (Fig. 1A).No effect on FSH release was observed with any of the used doses (Fig. 1B).
When the stimulations of pars distalis cells were done with 6 μg of protein obtained from cultures of each of the five Percoll bands of pars tuberalis cells, LH release was observed with culture media obtained from only two cell bands (50 and 60% Percoll; Fig. 2A).Again no effect on FSH release was observed (Fig. 2B).
Therefore, cells from 50 and 60% strength Percoll were cultured together.Figure 3A shows LH release from pars distalis cells that were incubated with different protein concentrations of culture medium from 50-60% strength Percoll cells fractions.The maximal response of pars distalis cells was observed when they were stimulated with 2 and 4 μg of pars tuberalis protein in the final volume of incubation (500 μl).This response was not mantained with higher protein concentrations.
Results from the superfusion experiments are shown in figure 3B.After unspecific stimulation with 28 mM KCL (black squares) a significant increase in the release of LH from pars distalis cells was induced over baseline values after 4 min.Also a significant response of pars distalis cells was observed when they were stimulated with culture medium of pars tuberalis cells from 50-60% Percoll fractions, containing either 3 or 0.3 μg of protein per mL (black dots and white triangles, respectively).
Figure 4A shows the results from SDS-PAGE studies.After protein separation by 12% SDS-PAGE of the culture medium obtained from 50-60% Percoll cell fractions, eight bands were visualized.The LH release response obtained with the protein recovered from each band is shown in Fig. 4B.LH release was only detected (p< 0.001) when pars distalis cells were stimulated with the eluate from the band of molecular size of 66 kDa.In this case a high LH response was obtained with 50 ng of pars tuberalis protein in the final volume of incubation (500 μl).

Discussion
Several types of secretory cells have been described in the pars tuberalis of the hypophysis.Previously we have shown that culture medium from total bovine cell populations of pars tuberalis stimulate the release of prolactin and GH from rat pars distalis cells (Lafarque et al., 1998;2004).In this study we report evidence for a factor of 66 kDa secreted by the pars tuberalis of bovine pituitary that shows an LH releasing action on rat pars distalis cells.
The pars tuberalis cellular populations were fractioned over a discotinuous Percoll gradients allowed to identify only two of the cell fractions (the 50 and 60% Percoll fractions) that were capable to induce LH release from pars distalis cells.As expected, the maximum LH release induced by secretions from the 50+60% Percoll fractions occurred at a three times lower protein concentration, suggesting that pars tuberalis cells in those fractions could be responsible for producing the factor(s) which affect pars distalis gonadotrope cells.
Studies in the rat indicate that the mode of hormone synthesis and/or release from LH cells of the pars tuberalis is different from those of the pars distalis (Gross et al., 1984;Watanabe, 1986).The secreting LH capacity of pars tuberalis cells appears to be smaller and inconsistent, with low LHRH sensitivity (Watanabe, ).In our case, the measured LH comes from pars distalis cells, and not from pars tuberalis cells, since LH was not detected in the culture medium of pars tuberalis cells using the specific LH antibody for rats.Also, it is not likely that the measured LH comes from new synthesis in pars distalis cells, since in superfusion experiments the LH response was observed at shorter times after stimulation with pars tuberalis cell secretions.
In the rat, numerous LHRH-positive fibers have been observed as tiny lines with several varicosities both in the primary vascular plexus and in the hypothalamus corresponding to the posterior half of the portal vein area.LHRH-positive fibers have also been noted around S-100-positive cells in the pars tuberalis (Mabuchi et al., 2004).In our experimental conditions, the stimulation of pars distalis cells cannot be attributed to the presence of hypothalamic factors in the culture medium of these cells, since they were cultured long enough to eliminate the nerve endings which could be present in the pars tuberalis dissection.Additionally, the partial purification of pars tuberalis cells on Percoll gradients and the successive cell washes to eliminate Percoll solution also contributes to the tuberalis cultures being exempt of hypothalamic factors.
After pars distalis cell stimulation with culture medium of pars tuberalis cells from 50+60% Percoll fractions containing 6 μg protein or more in the final volume of pars distalis cells incubation, LH release decreased.This fact could be explained by a possible direct inhibitory effect of prolactin on LH secretion.In the same experimental model, we have previously reported that secretions of pars tuberalis cells from 50+60% Percoll fractions containing 4 μg protein or more in 500 μl of final incubation volume increase prolactin release from pars distalis cells (Lafarque et al., 1998).Studies in the rat have shown that in vitro treatment with prolactin is able to suppress both basal and LHRH-stimulated LH release from cultured pituitary fragments (Cheung, 1983;Foreman and Porter, 1981) and reduce the percentage of pituitary cells secreting LH in vitro (Sortino and Wise, 1989).Similarly, the in vivo and in vitro LH response to exogenous LHRH administration was shown to be reduced in hyperprolactinemic rats, compared with controls (Smith, 1982).
Culture medium from pars tuberalis cells did not induce FSH release from pars distalis cells.The intracellular mechanism for the differential LH and FSH release after stimulation with pars tuberalis secretions might be explained by the fact that gonadotropins are stored and secreted from different secretory granules within bihormonal cells (Thomas and Clarke, 1997;Farnworth, 1995).
Our results suggest a possible modulating effect of pars tuberalis on pars distalis gonadotrope cells.Although the exact nature of the factor involved in LH release is not known yet, in this study we have found the biological activity on the basis of a pars tuberalis protein of molecular size of 66-kDa.
Future studies will be carried out to identify the structure of these pars tuberalis proteins and also to determine the signalling pathway involved in LH release from pars distalis cells.It has been shown that a factor(s) released from ovine pars tuberalis cells increases c-fos expression in some, but not all, lactotrophs in pars distalis cell cultures (Morgan et al., 1996).This study reveals the potential influence of the pars tuberalis on a second endocrine axis and also extends our knowledge of pituitary regulation.

FIGURE 2 .
FIGURE 2. LH (A) and FSH (B) release by dispersed cells from rat pars distalis incubated in vitro with 6 μg of protein from the culture medium of each of the fractions of pars tuberalis cells separated on discontinuous Percoll gradients in 500 μl of final volume of incubation.Bars represent mean ± SEM of duplicate measurements as indicated in Fig. 1.Letters indicate values with significant differences vs basal value, a P < 0.001.

FIGURE 4 .
FIGURE 4. A, SDS-PAGE (12%) gel of culture medium from pars tuberalis cells of 50-60% Percoll fractions, bands 1 to 8. MW: molecular weight markers.B, LH release from dispersed pars distalis cells stimulated with the protein (50 ng) eluted from each band from the SDS-PAGE (12%) gel.Bars represent mean ± SEM of duplicate measurements of 4 independent experiments.a P < 0.001 band 6 corresponding to 66 kDa vs basal value.