00l3-7227/90/1261-0512$02.00/0 Endocrinology Copyright © 1990 by The Endocrine Society
Vol. 126, No. 1 Printed in U.S.A.
Alterations in in Situ Prolactin Secretory Granule Morphology and Immunoactivity by Thiols and Divalent Cations* JANE GREENAN, MARY Y. LORENSON, MICHAEL V. CONCONI, AND AMEAE M. WALKER Division of Biomedical Sciences (J.G), University of California, Riverside, California 92521-0121; and the Endocrine Metabolism Unit (M.Y.L.), Department of Medicine, and the Clinical Research Center, University of Rochester School of Medicine and Dentistry, Rochester, New York 14642
ABSTRACT. The mechanisms involved in PRL storage in secretory granules are generally poorly understood. Recent studies with isolated granules, however, have suggested that granule storage forms may be relatively osmotically inactive due to oligomerization involving hormonal intermolecular disulfide bonds. Thus, expenditure of metabolic energy by the cell in order to maintain granule integrity would be reduced. When secretion is stimulated, oligomer depolymerization by thiol exchange mechanisms has been proposed to occur before or even concomitant with exocytosis. The present studies were designed to investigate the influence of metabolic inhibitors, thiols, and divalent cations on PRL storage in situ, rather than in isolated
granules. The results suggest that 1) PRL granules require little energy to maintain their structure, since a combination of azide (10 mM), fluoride (10 HIM), and cyanide (1 mM) had no effect on PRL granule morphology in normal anterior pituitary cells in primary culture; 2) disulfide linkages are involved in the osmotic activity of the PRL granule contents, since thiols induced granule swelling in lightly fixed cells; and 3) thiols and divalent cations are capable of altering the arrangement of stored hormone molecules, since PRL immunoactivity could be modified by these agents in glycol methacrylate-embedded exposed granule cores. {Endocrinology 126: 512-518,1990)
R
uct and large polyanionic compounds (1). A sulfated protein, present in low concentrations in PRL granules, may play such a role (6). However, in the pituitary mammotroph there is also evidence that osmotic inactivity is produced, at least in part, by hormone oligomerization via intermolecular disulfide bonding (7-11). In in vitro incubations of bovine pituitary granules, the presence of thiols increases the ratio of monomers to higher mol wt forms and increases the apparent intragranular osmolality (11), PRL, and GH release from these granule incubations is profoundly stimulated by thiols (9), and this can be blocked by increasing the medium osmolality (11). Divalent cations, on the other hand, reduce the proportion of monomers (7, 11,12) and inhibit hormone release (12). In the present study we examined the effects of metabolic inhibitors, thiols, and divalent cations on the ultrastructural and immunocytochemical properties of PRL secretory granules. We found that storage of PRL within cells does not require the expenditure of significant amounts of metabolic energy. Also, we observed that the endogenous concentrations of reduced glutathione (GSH) and zinc in pituitary secretory granules were higher than levels previously found to significantly alter
ELATIVELY little is known about the mechanisms involved in secretory product storage and the release of that product during exocytosis. Work by Jamieson and Palade in 1971 (1) showed that pancreatic zymogens were stored in a form that did not require the continual expenditure of metabolic energy; the secretory product was essentially osmotically inactive. In the insulin secretory granule, reduced intragranular osmolality may be obtained by the complexation of zinc with insulin (two zinc for every six insulin molecules) (2, 3). As in most granule systems, intragranular pH may also contribute to the stability of the stored insulin. Chromaffin granules and serotonin-containing platelet granules are stabilized by interaction of the secretory product with nucleotides, divalent cations, and intragranular proteins (4, 5). In some cell systems it has been suggested that osmotic inactivity of secretory product is brought about by strong ionic interactions between the secretory prodReceived July 24, 1989. Address requests for reprints to: Dr. Ameae M. Walker, Division of Biomedical Sciences, University of California, Riverside, California 92521-0121. * This work was supported by NIH Grant AM-28534, ACS Grant BC-562, an Academic Senate grant (to A.M.W.), and NIH Grants DK31326 and RR-00044 (to M.Y.L.).
512
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 02:09 For personal use only. No other uses without permission. . All rights reserved.
IN SITU PRL SECRETORY GRANULE STUDIES
hormone release or hormone monomer to oligomer ratios in granule incubations (9, 12). In addition, by lightly fixing cells to maintain some ultrastructural integrity, we were able to demonstrate in situ that thiol exposure resulted in granule swelling, and further, that thiols or zinc increased the number of immunohistochemically demonstrable PRL antigenic sites. Magnesium, another intragranular ion, negated the effects of zinc. These data further support the hypothesis that the conversion of hormone storage forms to releasable forms by thioldisulfide exchange mechanisms results in increased intragranular osmolality, which may be a critical feature in the regulation of pituitary hormone exocytosis. They also support a role for divalent cations in the physical arrangement of hormone molecules in the granule.
Materials and Methods Primary cultures Anterior pituitaries from female Sprague-Dawley-derived rats (Bantin and Kingman, Fremont, CA), weighing approximately 250 g, were dissociated by the method of Hymer et al. (13), as modified by Hopkins and Gregory (14), and plated in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with nonessential amino acids (NEAA) and 10% fetal bovine serum (all from Gibco, Grand Island, NY) at a density of 2 x 106 cells/35-mm petri dish. The cells were maintained for 48 h at 37 C under a water-saturated atmosphere of 95% 02-5% CO2 before experimental manipulation. Incubation in metabolic inhibitors Cells were washed three times in DMEM-NEAA containing 0.1% gelatin and then incubated in medium containing 10 mM NaN3, 1 mM NaCN, and 10 mM NaF in the same medium at 37 C for 1.5 h. The cells were washed free of gelatin and processed for electron microscopy. Gluteraldehyde fixation and processing for electron microscopy Cells were fixed in 50:50 DMEM-Karnovsky's fixative (15) for 2 h at room temperature, postfixed in Palade's osmium (16) for 1 h at 4 C, stained in Kellenberger's uranyl acetate (17) for 1 h at 4 C, pelleted by centrifugation at 10,000 x g for 15 min, and then dehydrated before embedding in Polybed 812 (Polysciences, Inc., Warrington, PA). Seventy-nanometer sections were cut on an LKB ultratome V (LKB, Rockville, MD), stained with uranyl acetate and lead citrate, and viewed on an Hitachi 600 electron microscope (Hialeah, FL), using an acceleration voltage of 75 keV. Mild fixation and incubation in thiols Cells were washed three times in DMEM and then fixed in periodatelysineparaformaldehyde (18) containing 6% sucrose for 1.5 h at room temperature. The fixed cells were washed three more times and then incubated in 0.5 or 5 mM /3-mercaptoethanol (MCE) or reduced GSH in DMEM-NEAA for 1 h at
513
room temperature. The cells were fixed again, this time in 50:50 DMEM-Karnovsky's fixative, and processed for electron microscopy as described above. Atomic absorption measurement of divalent cations Divalent cation concentrations in pituitary homogenates, fresh (rat and bovine) or frozen (bovine and porcine; Pelfreeze, Rogers, AK), and secretory granules, after removal of protein by 10% tricholoroacetic acid precipitation at 4 C, were determined using a Perkin-Elmer 290B atomic absorption spectrophotometer (Norwalk, CT). An air-acetylene mixture with a tank regulator pressure of 8 psi was used, with a slit width of 0.7 nm. The wavelengths for optimal sensitivity were 214 nm (zinc), 285 nm (magnesium), and 423 nm (calcium). Secretory granules from bovine pituitaries were isolated as previously described (19), using differential and sucrose density gradient centrifugation; a slightly modified procedure was used to prepare a partially purified rat secretory granule fraction (20). These two procedures produce a large granule fraction containing only GH and PRL granules. Protein concentrations were determined by the method of Lowry et al. (21), using BSA as standard. All glassware used was prewashed in dilute nitric acid to assure no divalent cation contamination. Fluorometric determination of GSH using o-phthalaldehyde GSH levels were measured fluorometrically by complexing GSH with o-phthaldehyde and measuring emission at 420 nm (excitation, 350 nm), using a JASCO FP550 spectrofluorometer, as previously described (22, 23). The sensitivity of the assay was in the 40- to 50-nM range. Glycol methacrylate embedding and immunohistochemistry Approximately 2 x 2-mm pieces of rat anterior pituitary were fixed for 2 h in the periodatelysineparaformaldehyde fixative (18) containing 6% sucrose, dehydrated in a graded series of alcohols to 95% ethanol, infiltrated overnight with JB4 embedding medium (Polysciences), and then embedded under vacuum for 4 h before sectioning. Five-micron sections were hydrated in antibody-binding buffer (AB buffer; 150 mM NaCl, 50 mM Tris, and 0.05% Nonidet P-40, pH 7.4) for 15 min at room temperature before immersion in AB buffer containing thiols or chloride salts of the divalent cations. When more than one agent was used, these were coincubated. The thiol and divalent cation solutions were removed by four changes of AB buffer before incubation in blocking serum (normal goat serum) for 15 min (1:75 in AB buffer) at room temperature. Incubations in specific antisera, nonimmune rabbit serum, or specific antisera supplemented with highly purified PRL (5 iig/m\) continued for 2 h at 37 C. Specific antisera and purified rat PRL (1-5, AFP4451B) were obtained through the National Pituitary Hormone Distribution Program. The rabbit antirat PRL serum (IC3, AFP) was diluted 1:100 in AB buffer before use. All binding, above that seen with nonimmune serum, was eliminated by the addition of 15 Mg/ml purified PRL. Unbound antibody was removed by washing three times with AB buffer and once with PBS (0.01 M phosphate). Bound antibody was visualized using a biotinylated second antibody
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 02:09 For personal use only. No other uses without permission. . All rights reserved.
IN SITU PRL SECRETORY GRANULE STUDIES
514
Endo • 1990 Voll26«Nol
and Vector ABC reagents (Vector Laboratories, Burlingame, CA). Briefly, sections were incubated in biotinylated goat antirabbit immunoglobulins, washed, and then incubated in the avidin-biotin complex reagent, which produces biotin-avidinbiotin-peroxidase complexes. The sections were washed again and then incubated in peroxidase substrate solution (0.01% H2O2-0.05% diaminobenzidene in 0.05 M Tris buffer, pH 7.2). After peroxidase development, the sections were washed with water and densitometrically scanned using a Hoefer GS-300 scanning densitometer (San Francisco, CA). All of the slides, each carrying four or five sections, were processed together and in the same order throughout the entire protocol, so that incubation times were exactly the same.
Results Incubation of living cells in metabolic inhibitors Incubation of anterior pituitary cells in a combination of NaN3, NaF, and NaCN had no effect on PRL granule morphology (compare Figs. 1, A and B). For these purposes, mammotrophs were identified on the basis of ultrastructural criteria (14). The only morphologically apparent effect of the metabolic inhibitors was a slight swelling of the Golgi complex and the occasional appearance of "blown" mitochondria. The concentrations of inhibitors chosen should have reduced cellular ATP levels by at least 90% (1, 24). Given adequate fixation, the morphological changes were consistent with depressed ATP levels. Incubation of lightly fixed cells in thiols Figure 2 shows the effects of thiol incubation on granule integrity. Thiol incubation resulted in major swelling of immature granules, an overall decreased density of some granules (Fig. 2, B-D, stars), and peripheral dissolution of others {arrows). In some instances, the swelling of immature granules was such that no granule core was visible in the section {arrowhead on Fig. 2B). All of these effects occurred with both thiols, although there was a tendency for GSH to cause more overall decreases in granule density and for MCE to cause more peripheral dissolution. For both thiols, 5 mM was more effective than 0.5 mM. At both concentrations, there were some granules that seemed unaffected by the treatment. Treatment of living cells with the same concentrations of GSH or MCE had no effect on granule morphology (data not presented). Measurement of intragranular GSH and divalent cations GSH in pituitary homogenates of both bovine and rat pituitaries was determined; values of 1.5 ± 0.2 yug/mg protein (rat) and 1.3 ± 0.3 /xg/mg protein (bovine) were obtained (n = 6 for each). Concurrently, the GSH level in rat liver homogenates was 1.8 ± 0.3 ixg/mg protein, in
B FIG. 1. Mammotrophs in primary culture after incubation in the absence (A) and presence (B) of metabolic inhibitors. N, Nucleus; G, Golgi complex; m, mitochondrion; sg, secretory granule. Magnification: A, X9300; B, X7000.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 02:09 For personal use only. No other uses without permission. . All rights reserved.
IN SITU PRL SECRETORY GRANULE STUDIES
515
FIG. 2. Mammotrophs prefixed with the periodate-lysine paraformaldehyde fixative and then exposed to 5 mm thiols. A, Fixation control; B and C, MCE; D, GSH. N, Nucleus; G, Golgi complex; ig, immature granules; stars, evenly decondensed granules; arrowhead, swollen immature granule where granule core is not visible; arrows, peripheral dissolution of granules. Magnifications: A, X6.690; B, X6.690; C, X19.440; D, X12.240.
good agreement with previously reported values (22, 23). When isolated bovine and rat pituitary secretory granules were tested, the concentration of GSH was 0.92 ± 0.17 Mg/mg protein (n = 6). Assuming an intragranular water space of 1 ^1/mg protein (25, 26), the GSH concentration in granules was equivalent to about 3 mM. The concentrations of total calcium, magnesium, and zinc in bovine, porcine, and rat pituitary homogenates are given in Table 1. These values were consistent whether the tissue was from a fresh or a frozen source. TABLE 1. Divalent cation concentrations in homogenates of bovine, porcine, and rat adenohypophyses Pituitary homogenates (Mg/g wet wt) Calcium Magnesium Zinc
Bovine
Porcine
Rat
226 ± 25 162 ± 8 19.7 ± 0.7
166 ± 25 211 ± 10 19.9 ± 1.3
290 ± 19 121 ± 9 11.1 ± 0.6
Determinations were by atomic absorption spectroscopy, as described in Materials and Methods. Values given are the mean ± SE (n = 6 different preparations) for all divalent cations.
i
Secretory granules also contained these divalent cations. Levels (micrograms per milligram protein) were 0.46 for calcium, 0.18 for magnesium, and 0.072 for zinc, which correspond to approximately 11.5 mM calcium, 7.7 mM magnesium, and 1.1 mM zinc. Immunodetectability of PRL after section pretreatment with thiols or divalent cations Preincubation of pituitary sections with thiols or divalent cations for 15 min had a marked effect on the subsequent immunohistochemical detection of PRL. For all studies, it was important to use freshly cut sections. Dehydration of the surface of the section significantly reduced our ability to produce any change in hormone immunodetectability. Table 2 summarizes the results obtained when pituitary sections were pretreated with GSH or MCE; binding was measured in the absence or presence of added purified PRL (termed optimal and reduced binding conditions, respectively). Incubation in either GSH or MCE increased the amount of specific antiserum binding, up to the maximum observed in this
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 02:09 For personal use only. No other uses without permission. . All rights reserved.
IN SITU PRL SECRETORY GRANULE STUDIES
516
TABLE 2. PRL immunolocalization after section pretreatment with thiols: relative absorbancies of peroxidase reaction product Specific binding Section pretreatment
Without added PRL
No additions 5 mM GSH 5 mM MCE
14 41"
With added PRL
100 290 260
36.5°
3.5
18.5" 11"
100 530 310
The mean scan values for three sections per slide are given both in relative absorbance units (A) and as a percentage of the control value (no additions). For each scan, the optimum absorption was recorded, and the section was microscopically inspected along the scan line to ensure that intracellular peroxidase and not nonspecific deposits gave rise to the value. Specific binding is binding with the specific antiserum (in the absence or presence of purified PRL) minus binding with nonimmune antiserum for the same section pretreatment. All values were significantly different from pretreatment without additions. This experiment has been repeated four times with the same result. "P
Vol. 126, No. 1 Printed in U.S.A.
Alterations in in Situ Prolactin Secretory Granule Morphology and Immunoactivity by Thiols and Divalent Cations* JANE GREENAN, MARY Y. LORENSON, MICHAEL V. CONCONI, AND AMEAE M. WALKER Division of Biomedical Sciences (J.G), University of California, Riverside, California 92521-0121; and the Endocrine Metabolism Unit (M.Y.L.), Department of Medicine, and the Clinical Research Center, University of Rochester School of Medicine and Dentistry, Rochester, New York 14642
ABSTRACT. The mechanisms involved in PRL storage in secretory granules are generally poorly understood. Recent studies with isolated granules, however, have suggested that granule storage forms may be relatively osmotically inactive due to oligomerization involving hormonal intermolecular disulfide bonds. Thus, expenditure of metabolic energy by the cell in order to maintain granule integrity would be reduced. When secretion is stimulated, oligomer depolymerization by thiol exchange mechanisms has been proposed to occur before or even concomitant with exocytosis. The present studies were designed to investigate the influence of metabolic inhibitors, thiols, and divalent cations on PRL storage in situ, rather than in isolated
granules. The results suggest that 1) PRL granules require little energy to maintain their structure, since a combination of azide (10 mM), fluoride (10 HIM), and cyanide (1 mM) had no effect on PRL granule morphology in normal anterior pituitary cells in primary culture; 2) disulfide linkages are involved in the osmotic activity of the PRL granule contents, since thiols induced granule swelling in lightly fixed cells; and 3) thiols and divalent cations are capable of altering the arrangement of stored hormone molecules, since PRL immunoactivity could be modified by these agents in glycol methacrylate-embedded exposed granule cores. {Endocrinology 126: 512-518,1990)
R
uct and large polyanionic compounds (1). A sulfated protein, present in low concentrations in PRL granules, may play such a role (6). However, in the pituitary mammotroph there is also evidence that osmotic inactivity is produced, at least in part, by hormone oligomerization via intermolecular disulfide bonding (7-11). In in vitro incubations of bovine pituitary granules, the presence of thiols increases the ratio of monomers to higher mol wt forms and increases the apparent intragranular osmolality (11), PRL, and GH release from these granule incubations is profoundly stimulated by thiols (9), and this can be blocked by increasing the medium osmolality (11). Divalent cations, on the other hand, reduce the proportion of monomers (7, 11,12) and inhibit hormone release (12). In the present study we examined the effects of metabolic inhibitors, thiols, and divalent cations on the ultrastructural and immunocytochemical properties of PRL secretory granules. We found that storage of PRL within cells does not require the expenditure of significant amounts of metabolic energy. Also, we observed that the endogenous concentrations of reduced glutathione (GSH) and zinc in pituitary secretory granules were higher than levels previously found to significantly alter
ELATIVELY little is known about the mechanisms involved in secretory product storage and the release of that product during exocytosis. Work by Jamieson and Palade in 1971 (1) showed that pancreatic zymogens were stored in a form that did not require the continual expenditure of metabolic energy; the secretory product was essentially osmotically inactive. In the insulin secretory granule, reduced intragranular osmolality may be obtained by the complexation of zinc with insulin (two zinc for every six insulin molecules) (2, 3). As in most granule systems, intragranular pH may also contribute to the stability of the stored insulin. Chromaffin granules and serotonin-containing platelet granules are stabilized by interaction of the secretory product with nucleotides, divalent cations, and intragranular proteins (4, 5). In some cell systems it has been suggested that osmotic inactivity of secretory product is brought about by strong ionic interactions between the secretory prodReceived July 24, 1989. Address requests for reprints to: Dr. Ameae M. Walker, Division of Biomedical Sciences, University of California, Riverside, California 92521-0121. * This work was supported by NIH Grant AM-28534, ACS Grant BC-562, an Academic Senate grant (to A.M.W.), and NIH Grants DK31326 and RR-00044 (to M.Y.L.).
512
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 02:09 For personal use only. No other uses without permission. . All rights reserved.
IN SITU PRL SECRETORY GRANULE STUDIES
hormone release or hormone monomer to oligomer ratios in granule incubations (9, 12). In addition, by lightly fixing cells to maintain some ultrastructural integrity, we were able to demonstrate in situ that thiol exposure resulted in granule swelling, and further, that thiols or zinc increased the number of immunohistochemically demonstrable PRL antigenic sites. Magnesium, another intragranular ion, negated the effects of zinc. These data further support the hypothesis that the conversion of hormone storage forms to releasable forms by thioldisulfide exchange mechanisms results in increased intragranular osmolality, which may be a critical feature in the regulation of pituitary hormone exocytosis. They also support a role for divalent cations in the physical arrangement of hormone molecules in the granule.
Materials and Methods Primary cultures Anterior pituitaries from female Sprague-Dawley-derived rats (Bantin and Kingman, Fremont, CA), weighing approximately 250 g, were dissociated by the method of Hymer et al. (13), as modified by Hopkins and Gregory (14), and plated in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with nonessential amino acids (NEAA) and 10% fetal bovine serum (all from Gibco, Grand Island, NY) at a density of 2 x 106 cells/35-mm petri dish. The cells were maintained for 48 h at 37 C under a water-saturated atmosphere of 95% 02-5% CO2 before experimental manipulation. Incubation in metabolic inhibitors Cells were washed three times in DMEM-NEAA containing 0.1% gelatin and then incubated in medium containing 10 mM NaN3, 1 mM NaCN, and 10 mM NaF in the same medium at 37 C for 1.5 h. The cells were washed free of gelatin and processed for electron microscopy. Gluteraldehyde fixation and processing for electron microscopy Cells were fixed in 50:50 DMEM-Karnovsky's fixative (15) for 2 h at room temperature, postfixed in Palade's osmium (16) for 1 h at 4 C, stained in Kellenberger's uranyl acetate (17) for 1 h at 4 C, pelleted by centrifugation at 10,000 x g for 15 min, and then dehydrated before embedding in Polybed 812 (Polysciences, Inc., Warrington, PA). Seventy-nanometer sections were cut on an LKB ultratome V (LKB, Rockville, MD), stained with uranyl acetate and lead citrate, and viewed on an Hitachi 600 electron microscope (Hialeah, FL), using an acceleration voltage of 75 keV. Mild fixation and incubation in thiols Cells were washed three times in DMEM and then fixed in periodatelysineparaformaldehyde (18) containing 6% sucrose for 1.5 h at room temperature. The fixed cells were washed three more times and then incubated in 0.5 or 5 mM /3-mercaptoethanol (MCE) or reduced GSH in DMEM-NEAA for 1 h at
513
room temperature. The cells were fixed again, this time in 50:50 DMEM-Karnovsky's fixative, and processed for electron microscopy as described above. Atomic absorption measurement of divalent cations Divalent cation concentrations in pituitary homogenates, fresh (rat and bovine) or frozen (bovine and porcine; Pelfreeze, Rogers, AK), and secretory granules, after removal of protein by 10% tricholoroacetic acid precipitation at 4 C, were determined using a Perkin-Elmer 290B atomic absorption spectrophotometer (Norwalk, CT). An air-acetylene mixture with a tank regulator pressure of 8 psi was used, with a slit width of 0.7 nm. The wavelengths for optimal sensitivity were 214 nm (zinc), 285 nm (magnesium), and 423 nm (calcium). Secretory granules from bovine pituitaries were isolated as previously described (19), using differential and sucrose density gradient centrifugation; a slightly modified procedure was used to prepare a partially purified rat secretory granule fraction (20). These two procedures produce a large granule fraction containing only GH and PRL granules. Protein concentrations were determined by the method of Lowry et al. (21), using BSA as standard. All glassware used was prewashed in dilute nitric acid to assure no divalent cation contamination. Fluorometric determination of GSH using o-phthalaldehyde GSH levels were measured fluorometrically by complexing GSH with o-phthaldehyde and measuring emission at 420 nm (excitation, 350 nm), using a JASCO FP550 spectrofluorometer, as previously described (22, 23). The sensitivity of the assay was in the 40- to 50-nM range. Glycol methacrylate embedding and immunohistochemistry Approximately 2 x 2-mm pieces of rat anterior pituitary were fixed for 2 h in the periodatelysineparaformaldehyde fixative (18) containing 6% sucrose, dehydrated in a graded series of alcohols to 95% ethanol, infiltrated overnight with JB4 embedding medium (Polysciences), and then embedded under vacuum for 4 h before sectioning. Five-micron sections were hydrated in antibody-binding buffer (AB buffer; 150 mM NaCl, 50 mM Tris, and 0.05% Nonidet P-40, pH 7.4) for 15 min at room temperature before immersion in AB buffer containing thiols or chloride salts of the divalent cations. When more than one agent was used, these were coincubated. The thiol and divalent cation solutions were removed by four changes of AB buffer before incubation in blocking serum (normal goat serum) for 15 min (1:75 in AB buffer) at room temperature. Incubations in specific antisera, nonimmune rabbit serum, or specific antisera supplemented with highly purified PRL (5 iig/m\) continued for 2 h at 37 C. Specific antisera and purified rat PRL (1-5, AFP4451B) were obtained through the National Pituitary Hormone Distribution Program. The rabbit antirat PRL serum (IC3, AFP) was diluted 1:100 in AB buffer before use. All binding, above that seen with nonimmune serum, was eliminated by the addition of 15 Mg/ml purified PRL. Unbound antibody was removed by washing three times with AB buffer and once with PBS (0.01 M phosphate). Bound antibody was visualized using a biotinylated second antibody
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 02:09 For personal use only. No other uses without permission. . All rights reserved.
IN SITU PRL SECRETORY GRANULE STUDIES
514
Endo • 1990 Voll26«Nol
and Vector ABC reagents (Vector Laboratories, Burlingame, CA). Briefly, sections were incubated in biotinylated goat antirabbit immunoglobulins, washed, and then incubated in the avidin-biotin complex reagent, which produces biotin-avidinbiotin-peroxidase complexes. The sections were washed again and then incubated in peroxidase substrate solution (0.01% H2O2-0.05% diaminobenzidene in 0.05 M Tris buffer, pH 7.2). After peroxidase development, the sections were washed with water and densitometrically scanned using a Hoefer GS-300 scanning densitometer (San Francisco, CA). All of the slides, each carrying four or five sections, were processed together and in the same order throughout the entire protocol, so that incubation times were exactly the same.
Results Incubation of living cells in metabolic inhibitors Incubation of anterior pituitary cells in a combination of NaN3, NaF, and NaCN had no effect on PRL granule morphology (compare Figs. 1, A and B). For these purposes, mammotrophs were identified on the basis of ultrastructural criteria (14). The only morphologically apparent effect of the metabolic inhibitors was a slight swelling of the Golgi complex and the occasional appearance of "blown" mitochondria. The concentrations of inhibitors chosen should have reduced cellular ATP levels by at least 90% (1, 24). Given adequate fixation, the morphological changes were consistent with depressed ATP levels. Incubation of lightly fixed cells in thiols Figure 2 shows the effects of thiol incubation on granule integrity. Thiol incubation resulted in major swelling of immature granules, an overall decreased density of some granules (Fig. 2, B-D, stars), and peripheral dissolution of others {arrows). In some instances, the swelling of immature granules was such that no granule core was visible in the section {arrowhead on Fig. 2B). All of these effects occurred with both thiols, although there was a tendency for GSH to cause more overall decreases in granule density and for MCE to cause more peripheral dissolution. For both thiols, 5 mM was more effective than 0.5 mM. At both concentrations, there were some granules that seemed unaffected by the treatment. Treatment of living cells with the same concentrations of GSH or MCE had no effect on granule morphology (data not presented). Measurement of intragranular GSH and divalent cations GSH in pituitary homogenates of both bovine and rat pituitaries was determined; values of 1.5 ± 0.2 yug/mg protein (rat) and 1.3 ± 0.3 /xg/mg protein (bovine) were obtained (n = 6 for each). Concurrently, the GSH level in rat liver homogenates was 1.8 ± 0.3 ixg/mg protein, in
B FIG. 1. Mammotrophs in primary culture after incubation in the absence (A) and presence (B) of metabolic inhibitors. N, Nucleus; G, Golgi complex; m, mitochondrion; sg, secretory granule. Magnification: A, X9300; B, X7000.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 02:09 For personal use only. No other uses without permission. . All rights reserved.
IN SITU PRL SECRETORY GRANULE STUDIES
515
FIG. 2. Mammotrophs prefixed with the periodate-lysine paraformaldehyde fixative and then exposed to 5 mm thiols. A, Fixation control; B and C, MCE; D, GSH. N, Nucleus; G, Golgi complex; ig, immature granules; stars, evenly decondensed granules; arrowhead, swollen immature granule where granule core is not visible; arrows, peripheral dissolution of granules. Magnifications: A, X6.690; B, X6.690; C, X19.440; D, X12.240.
good agreement with previously reported values (22, 23). When isolated bovine and rat pituitary secretory granules were tested, the concentration of GSH was 0.92 ± 0.17 Mg/mg protein (n = 6). Assuming an intragranular water space of 1 ^1/mg protein (25, 26), the GSH concentration in granules was equivalent to about 3 mM. The concentrations of total calcium, magnesium, and zinc in bovine, porcine, and rat pituitary homogenates are given in Table 1. These values were consistent whether the tissue was from a fresh or a frozen source. TABLE 1. Divalent cation concentrations in homogenates of bovine, porcine, and rat adenohypophyses Pituitary homogenates (Mg/g wet wt) Calcium Magnesium Zinc
Bovine
Porcine
Rat
226 ± 25 162 ± 8 19.7 ± 0.7
166 ± 25 211 ± 10 19.9 ± 1.3
290 ± 19 121 ± 9 11.1 ± 0.6
Determinations were by atomic absorption spectroscopy, as described in Materials and Methods. Values given are the mean ± SE (n = 6 different preparations) for all divalent cations.
i
Secretory granules also contained these divalent cations. Levels (micrograms per milligram protein) were 0.46 for calcium, 0.18 for magnesium, and 0.072 for zinc, which correspond to approximately 11.5 mM calcium, 7.7 mM magnesium, and 1.1 mM zinc. Immunodetectability of PRL after section pretreatment with thiols or divalent cations Preincubation of pituitary sections with thiols or divalent cations for 15 min had a marked effect on the subsequent immunohistochemical detection of PRL. For all studies, it was important to use freshly cut sections. Dehydration of the surface of the section significantly reduced our ability to produce any change in hormone immunodetectability. Table 2 summarizes the results obtained when pituitary sections were pretreated with GSH or MCE; binding was measured in the absence or presence of added purified PRL (termed optimal and reduced binding conditions, respectively). Incubation in either GSH or MCE increased the amount of specific antiserum binding, up to the maximum observed in this
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 19 November 2015. at 02:09 For personal use only. No other uses without permission. . All rights reserved.
IN SITU PRL SECRETORY GRANULE STUDIES
516
TABLE 2. PRL immunolocalization after section pretreatment with thiols: relative absorbancies of peroxidase reaction product Specific binding Section pretreatment
Without added PRL
No additions 5 mM GSH 5 mM MCE
14 41"
With added PRL
100 290 260
36.5°
3.5
18.5" 11"
100 530 310
The mean scan values for three sections per slide are given both in relative absorbance units (A) and as a percentage of the control value (no additions). For each scan, the optimum absorption was recorded, and the section was microscopically inspected along the scan line to ensure that intracellular peroxidase and not nonspecific deposits gave rise to the value. Specific binding is binding with the specific antiserum (in the absence or presence of purified PRL) minus binding with nonimmune antiserum for the same section pretreatment. All values were significantly different from pretreatment without additions. This experiment has been repeated four times with the same result. "P
