0013-7227/90/1263-1630$02.00/0 Endocrinology Copyright© 1990 by The Endocrine Society
Vol. 126, No. 3 Printed in U.S.A.
M. FARRINGTONf AND W. C. HYMER Department of Molecular and Cell Biology, Pennsylvania State University, University Park, Pennsylvania 16802
ABSTRACT. Although it has been known for some time that
rats, but were in glands from the thyroidectomized rats injected
GH aggregates are contained within the rat anterior pituitary
with T4; 3) that GH aggregates were uniquely associated with a
gland, the role that they might play in pituitary function is unknown. The present study examines this issue using the technique of Western blotting, which permitted visualization of 11 GH variants with apparent mol wt ranging from 14-88K. Electroelution of the higher mol wt variants from gels followed by their chemical reduction with /3-mercaptoethanol increased GH immunoassayability by about 5-fold. With the blot procedure we found 1) that GH aggregates greater than 44K were associated with a 40,000 X g sedimentable fraction; 2) that GH aggregates were not present in glands from thyroidectomized
heavily granulated somatotroph subpopulation isolated by density gradient centrifugation; and 4) that high mol wt GH forms were released from the dense somatotrophs in culture, since treatment of the culture medium with /3-mercaptoethanol increased GH immunoassayability by about 5-fold. Taken together, the results show that high mol wt GH aggregates are contained in secretory granules of certain somatotrophs and are also released in aggregate form from these cells in vitro. (Endocrinology 126: 1630-1638, 1990)
M
thermore, packaging of GH polymers within the somatotroph secretory granule may lead to changes in intragranular osmotic pressure, which, in turn, facilitate exocytosis (23). GH secretory granules are known to be affected by the thyroid hormone status of the host (24-26), zinc ions (27), and somatotroph subtype (28, 29). We have evaluated the extent of GH aggregation under each of these three conditions by Western blotting.
ULTIPLE forms of GH are contained in the pituitary gland, blood, and urine (1-5). A family of GH variants, all structurally related to the 22K form of the hormone, include 1) the 20K form, which results from alternate mRNA splicing (6-9); 2) a 23K glycosylated variant (10, 11); 3) a two-chain 25K form resulting from proteolytic cleavage in the large disulfide loop (1214); 4) phosphorylated GH (15, 16); and 5) disulfidelinked aggregated forms (17, 18). The physiological significance of many of these GH variants is poorly defined. This is especially true for the aggregate forms. Dimerization of GH probably results from antiparallel alignment of the cystines in two GH monomers (19). The dimer shows relatively poor immunoreactivity (10-40% of the monomer) (20) and has only 15% of its activity in the tibial line assay (19). Nevertheless, there is evidence to support the hypothesis that GH aggregates have physiological significance. For example, GH dimers are present in the circulation; they have a longer half-life than the monomeric form (21, 22). Fur-
Materials and Methods Animals Sprague-Dawley male rats (Harlan, Indianapolis, IN), weighing 300-450 g, served as pituitary donors. In some experiments animals were chemically throidectomized by incorporating 6propyl-2-thiouracil (PTU; 0.5 mg/ml; Sigma, St. Louis, MO) and glucose (50 mg/ml) in their drinking water for 42 days. Controls received only glucose in their water. At death, the mean body weight of control animals was 491 ± 19 g; PTUtreated animals weighed 356 ± 12 g.
Received July 26, 1989. Address all correspondence and requests for reprints to: Dr. W. C. Hymer, 401 Althouse Laboratory, University Park, Pennsylvania 16802. * Presented in part at the 70th Annual Meeting of The Endocrine Society, New Orleans, LA, June 1988. This work was supported by NASA Grants NCC 2-370, 9-17416, and NAGW 1196. t Submitted in partial fulfillment for requirements of the Ph.D. degree in Biochemistry at the Pennsylvania State University.
GH extraction Individual pituitary glands were homogenized in 1 ml 0.01 N NaHCO3 at 4 C using 10 strokes of a ground glass McShan type homogenizer (Bellco Glass, Inc., Vineland, NJ). These homogenates were kept on ice for up to 24 h before centrifugation (40,000 x g; 20 min; 4 C). The supernatant fraction was flash frozen and stored at —70 C. Under these conditions GH
1630
Downloaded from https://academic.oup.com/endo/article-abstract/126/3/1630/2532990 by University of Newcastle user on 03 February 2019
Growth Hormone Aggregates in the Rat Adenohypophysis*
GH AGGREGATES IN RAT ADENOHYPOPHYSIS yield was 120 ± 10 ng GH/mg protein (Bradford method) (30). Storage of the homogenates for 0.5 or 2 h at 4 C yielded 50% Or 70% of that obtained at 24 h. Reextraction of the 40,000 x g pellet prepared from the 24-h samples yielded less than 10% of the initial GH extracted. Therefore, a single 24-h extraction procedure was used in this study.
Samples to be run on sodium dodecyl sulfate-polyacrylamide gels (SDS-PAGE) were prepared in 0.025 M Tris (pH 6.8), 2% SDS, 10% glycerol, and 0.0001% bromphenol blue. If reduced, they also contained 5% mercaptoethanol. The samples were heated in a boiling water bath for 5 min and cooled to room temperature before loading onto the gel. SDS-PAGE was performed in a discontinuous Tris-glycine buffer system at 25 C (31). Linear (5-15%) gradient separating gels (0.75 mm thick) were poured using a Hoeffer 30-ml gradient marker (Hoeffer Scientific, San Francisco, CA); these were overlayered with a stacking gel (5%) prepared in 0.25 M Tris, pH 6.8. The running buffer was 0.025 M Tris, 0.1% SDS, and 0.192 M glycine at pH 8.3. Protein mobility was monitored using solutions of 0.02% (wt/vol) bromophenol blue in the cathodic buffer reservoir. Western blot Proteins in polyacrylamide gels were electrophoretically transferred onto nitrocellulose paper according to a modification of the method of Towbin et al. (32). Prewetted pure nitrocellulose paper (Bio-Rad, Richmond, CA; 0.45 nm) was rolled onto the gel surface to promote a seal between the paper and gel, thereby facilitating transfer. Transfer was performed at constant amperage (300 mamp/gel) for 6 h in 0.10 M Tris, pH 9.3, at temperatures below 25 C. Immediately after transfer the paper was dried either in air or in a vacuum oven at 60 C for 10 min. This step decreased diffusion of protein bands. All subsequent steps were carried out in Seal-a-meal bags that were purged of air bubbles. The paper was first incubated in 0.01 M PBS, pH 7.4, containing 1% BSA at room temperature for 4 h or overnight at 4 C to block nonspecific binding sites. Immunoperoxidase staining of GH on the paper was accomplished by 1) incubation with monkey antirat GH antiserum (kindly provided by R. Grindeland, NASA Ames Research Center; 1:40,000 final dilution) (33) for 3 days at 4 C or 2 days at room temperature; 2) incubation with horseradish peroxidase-conjugated goat antimonkey immunoglobulin G (IgG; Cappel Laboratories, Cochranville, PA; 1:4,000 final dilution) for 6 h at room temperature; 3) overnight incubation at 4 C with rabbit antigoat IgG (Cappel) at 1:2,000 final dilution; 4) incubation for 4 h at room temperature in goat peroxidase antiperoxidase (Cappel; 1:4,000 final dilution); 5) incubation with enzyme substrate [0.2 mg 3,3'-tetrahydrochlorodiamino-benzidine; Sigma; 4 /ul 3% (vol/vol) hydrogen peroxide/ml PBS] for no longer than 1 h at room temperature. If 0.01 M sodium citrate, pH 5.2, was used in place of PBS in step 5, steps 3 and 4 could be eliminated. The nitrocellulose paper was washed four times with 0.15 M NaCl between each of the steps above. At the last step the blots were washed in distilled H2O, air dried, and analyzed by reflectance densitometry within 2 weeks.
Prestained mol wt markers (Bethesda Research Laboratories, Gaithersburg, MD) were used to estimate mol wt of transferred proteins. The coefficients of variation for the Rf values of different markers, based on analysis of 17 gels, were: /3lactoglobin, 0.9%; a-chymotrypsinogen, 1.6%; ovalbumin, 1.7%; phosphorylase-B, 2.2%; and myosin H chain, 5.9%. The relative reflective optical densities of GH Western blots were determined using a Bio-Rad 620 videodensitometer. The optical density profiles of GH provided a convenient technique for comparison of individual GH forms. Quantitation of these profiles was not attempted for two reasons. First, the time required for transfer of high mol wt forms sometimes permitted smaller GH forms to pass through to the other side. Second, several variants have different (6) or unknown antigenicities. Background staining was subtracted from each sample trace by a computer program incorporated into the densitometer. In every case this was determined on an empty lane that was adjacent to the sample lane in the same electorphoresis-blot trial. Protein elution Protein elution from the SDS-gel was performed by cutting a 2-cm wide slice and subjecting that slice to Western blotting (see above). The remaining portion of the gel was wrapped in plastic film, stored at -70 C, thawed in elution buffer (0.1 M Tris, pH 8.6), and finally cut into regions of GH staining after alignment with the template. The gel pieces were each homogenized in elution buffer, and proteins eluted electrophoretically (34). Briefly, this procedure involved 1) electrophoresis at constant voltage (500 V; 24 h) of protein through a glass tube (plugged at the end with 5% stacking gel) into dialysis tubing; 2) current reversal for 2 min to minimize protein attachment to the tubing; 3) dialysis against water overnight; and 4) lyophilization. Recovery of GH using this procedure was 111 ± 8% (three experiments). Differential centrifugation Cell-free homogenates were prepared in 0.25 M sucrose with or without 0.2 mM ZnCl2 (two pituitaries per ml) at 4 C using a Teflon homogenizor (Belco Glass, Vineland, NJ). After centrifugation at 275 X g for 10 min to remove nuclei, the supernatant fraction was centrifuged at 40,000 X g for 20 min. GH contained in the homogenate, the 40,000 X g pellet, and final supernatant fraction was obtained by extraction with NaHCO3 (0.01 M final concentration). Pituitary cell separation and culture Cells were prepared from pituitary glands by trypsinization (35). Yields were 2-3 X 106/gland, with viability greater than 95%. Somatotrophs were fractionated into two subpopulations by gradient centrifugation (28). Total cell recovery and GH recovery after this procedure were 90 ± 2% and 86 ± 3%, respectively. Cell culture was performed in Limbro plates (Flow Laboratories, McLean, VA) at a density of 1.5 x 105 cells/ml for 3 days at 37 C in humidified air plus 5% CO2. The culture medium was Minimum Essential Medium containing 5% horse serum and antibiotics (36).
Downloaded from https://academic.oup.com/endo/article-abstract/126/3/1630/2532990 by University of Newcastle user on 03 February 2019
Electrophoresis
1631
GH AGGREGATES IN RAT ADENOHYPOPHYSIS
1632 GH immunoassay
Data analysis Data are reported as the mean ± SEM. Duncan's multiple range test was used to measure differences.
Results Western blot methodologies GH molecules having apparent mol wt of 14-88K were routinely observed in pituitary extracts that were electrophoresed under nonreducing conditions, transferred to nitrocellulose paper, and then stained immunochemically for the presence of GH (Fig. 1A). Reflectance profiles of such immunoblots also showed heterogenity in GH forms on the paper (Fig. 1, middle). Immunoreactive GH forms greater than 25K disappeared upon chemical reduction of the extracts before electrophoresis (Fig. IB). The results of three tests helped to establish the validity of the GH immunostaining. First, use of the NIDDK antirat GH serum (1:8000 final dilution) in place of the flbsorbance
B
Grindeland antiserum yielded GH profiles identical to those in Fig. 1A. Second, replacement of the Grindeland antiserum with either normal monkey serum (1:8000) or BSA (2%) abolished GH staining. Third, preincubation of the Grindeland antiserum with 20, 40, or 80 ng GH (RP-1) for 24 h at 4 C before use resulted in severe dimunition (20 and 40 /zg) to elimination (80 fig) in staining of all GH bands (data not shown). When a glycine-methanol buffer (32) was used for blotting, transfer of high mol wt GH forms was severely inhibited, even though transfer time was increased by 10 h. Inhibition of transfer of high mol wt proteins was documented by silver staining of the gel after blotting. Results of other experiments, not shown here, indicated that GH profiles 1) were identical between fresh and frozen extracts, 2) were not affected by inclusion of aprotinin (2 U/ml) in the extract, and 3) were identical between extracts prepared in the presence or absence of iodoacetamide (1 x 10~3 M). Isoelectric focusing (IEF) The possibility that charge heterogeneity could in part account for the GH variants seen in Fig. 1 was addressed by focusing pituitary extracts in thin (0.5-mm) polyacrylamide-based gels containing ampholytes (pH 3.5-10), followed by pressure immunoblotting. The results revealed the presence of five or six immunoactive GH forms that had isoelectric points ranging from 4.8-7.1 regardless of whether the extracts were prepared in acetic acid (pH 3.3; Fig. 2A), Na2CO3 buffer (pH 10; Fig. 2B), or NaHCO3 (pH 8; Fig. 2C). In another experiment, unstained IEF gels were divided into five regions that contained GH of varying pi, viz. 4.8, 5.4, 5.9, 6.1-6.9, and 7.1. Proteins were then eluted from these IEF gels, lyophilized, electrophoresed by SDS-PAGE under nonreducing conditions, and immunoblotted. Although the mol wt patterns of GH variants were similar between IEF regions (Fig. 3), there were minor differences. For example, the 20K GH and its multimers tended to be found in the more basic regions, while the 25K variant tended to be found in the more acidic regions. /3ME treatment
25K22K20K" 14K-
FlG. 1. Representative Western blots of rat pituitary GH contained in extracts electrophoresed under nonreducing (A) or reducing (B) conditions (see Materials and Methods). The reflectance optical density tracing of blot in A is shown in the middle panel.
Since treatment of alkaline extracts with /3ME enhances the immunoreactivity of GH (37), and since this treatment also eliminates the high mol wt GH variants seen in nonreducing gels (Fig. IB), it was of interest to combine the two approaches to further study high mol wt GH variants. Shown in Table 1 are data which confirm previous findings that chemical reduction enhances GH immunoassayability and further show the repeatability of that effect. Maximal enhancement was achieved within 1 h (data not shown).
Downloaded from https://academic.oup.com/endo/article-abstract/126/3/1630/2532990 by University of Newcastle user on 03 February 2019
GH levels were measured by a specific enzyme immunoassay using three dilutions in duplicate (37). In some cases samples were treated with an equal volume of 0.4 M 2-mercaptoethanol (2-ME) for 1 h at 4 C. Before GH assay it was necessary to remove the 2-ME by centrifugation through Sephadex G-10 columns (38). Control samples were treated with water instead of 2-ME.
Endo• 1990 Voll26«No3
GH AGGREGATES IN RAT ADENOHYPOPHYSIS
1633
0.60-r
0.30--
Region 4 (pi 6.6, 6.9)
Region 3 (pi 5.9)
Region 2 (pi 5.4)
Region I (pl 4.8) 15
30
45
60
75
90
120
105
135 150
(•)
Migration in Gel (mm)
FIG. 3. Reflectance tracings of GH blots after two-dimensional gel electrophoresis. GH forms separated by IEF were eluted, dialyzed, lyophilized, reconstituted, and reelectrophoresed on SDS-PAGE gels (n = 2 experiments).
B
TABLE 1. Effect of /3ME treatment of rat pituitary extracts on GH immunoassayability
FIG. 2. Representative GH Western blots of IEF gels electrophoresed under different conditions: A, 0.01 N HOAc, pH 3.3; B, 0.05 N Na2CO3, pH 10; and C, 0.01 N Na2CO3, pH 8.0 (n = 3 experiments), aw, Application wick.
Electrocution of GH from slices of SDS-polyacrylamide gels followed by chemical reduction and GH immunoassay (see Materials and Methods) clearly showed enhancement of hormone activity in the higher mol wt regions (Table 2). Physiological studies Subcellular distribution. Alkaline extracts of a crude 40,000 x g mitochondrial-granule pellet (39) prepared from 0.25 M sucrose homogenates contained high mol wt GH variants (Fig. 4B). Extracts of the 40,000 X g supernatant fraction revealed virtually identical profiles (Fig. 4C). The distribution of recovered GH in this and other experiments, measured by immunoassay, was 32 ± 5% in the 40,000 X g pellet and 66 ± 4% in the 40,000 X g supernatant fraction. However, when homogenates were prepared in 0.25 M sucrose containing 0.2 mM ZnCl2, the distribution patterns of high mol wt GH were markedly
0ME
GH fold increase 1 1.39 ± 0.09 3.38 ± 0.73
molarity 0.00 0.05 0.20
Data represent the increase in G H immunoactivity (37) after /3ME treatment (n = 7 experiments). See Materials and Methods for additional details.
T A B L E 2. Effect of /3ME treatment of rat pituitary extracts on immunoassayability of different G H forms separated by SDS-PAGE
GH (Mg) Region
mol wt (kD)
Fold increase -j3ME
land 2 3 4 5 6 7 and 8
44K) GH forms in extracts of pituitary tissue has been recognized previously (5,18). However, their visualization on blots is not routine, probably because of technical issues relating to the Western blotting procedure itself. For example, 15% gels appear to have been used most often, and these have a fractionation range of 12-45K; in such gels 22K GH has an Rf of 0.42, and its dimer has an Rf of 0.14 (40). As a result, higher mol wt variants would probably be excluded from the separating gel. On the other hand, the fractionation range of 5-15% gels is about 14-250K. The most commonly used blotting buffers contain glycine and methanol (32). Methanol 1) counteracts swelling of the gel during blotting, 2) increases hydrophobic sites on nitrocellulose, 3) removes SDS from SDSprotein complexes (41), and 4) reduces pore size, thus
0 -i0
15
30
45
60
75
90
105 120
135
150
Migration in Gel (mm)
FIG. 7. Representative reflectance tracings of GH Western blots of tissue extracts prepared from A, pituitary glands of rats receiving sucrose in their drinking water for 42 days and a single injection of saline 24 h before death; B, pituitary glands of rats receiving sucrose in their drinking water for 42 days and a single injection of T4 24 h before death; C, pituitary glands of rats receiving sucrose and PTU in their drinking water for 42 days and a single injection of saline 24 h before death; and D, pituitary glands of rats receiving sucrose and PTU in their drinking water for 42 days and a single injection of T4 24 h before death (n = 2 experiments).
restricting transfer of larger molecules and causing protein precipitation in the gel. Elimination of methanol and glycine in the blotting buffer increased transfer efficiency and often GH staining intensity as well. Taken together, these modifications allowed us to study GH aggregates in a new way. Since the antisera used in our study was directed against 22K GH, it is not surprising that 22K and 25K were in the most heavily stained forms. The 25K form probably represents clipped GH (13, 14), while the 23K form may be the glycosylated variant (10, 11). It is probable that the higher mol wt forms are not attributable 1) to aggregations of clipped GH, since inclusion of aprotinin did not affect GH profiles; 2) to artificial disulfide-linked aggregation generated during extraction, since inclusion of iodoacetamide did not change GH profiles; 3) to deamidation, since extraction in buffers of different pH did not affect the GH profiles; or 4) to unique charge variants, since most of the mol wt variants were found in each of five variants isolated by IEF. It seems probable that the higher mol wt GH variants
Downloaded from https://academic.oup.com/endo/article-abstract/126/3/1630/2532990 by University of Newcastle user on 03 February 2019
Average
0.80 T
1635
1636
GH AGGREGATES IN RAT ADENOHYPOPHYSIS
Some years ago studies by Ellis et al. (33) showed that high mol wt forms of rat GH in tissue or plasma were rich in tibial line activity, but poor in immunological activity. Since culture medium from heavily granulated somatotrophs is potent in GH tibial line assay activity (46), and since implantation of heavily granulated (but not lightly granulated) somatotrophs into the cerebral ventricles of hypophysectomized rats restores animal growth (47), we suggest that the disulfide-linked GH aggregates identified in our study could represent a more bioactive form of the hormone molecule that its monomer. In the third experimental series we took advantage of the well documented effects of thyroidectomy on somatotroph degranulation (25, 26) as well as somatotroph regranulation after T4 injection (24). Since GH aggregates disappeared after thyroidectomy and reappeared after T4 administration, their association with secretory granules was again suggested. In summary, the results from these three experimental test systems offer strong support for the hypothesis that GH aggregates are localized within the GH secretion granule and that they are released from heavily granulated GH cells. Of course they do not eliminate the possibility that aggregates are also localized in other subcellular compartments. It is worthwhile to note that the GH aggregates appear to account for a major percentage of the total measurable hormone; the data in Tables 2 and 3 suggest that more than 70% of the total measurable extracted or secreted GH measured after chemical reduction was derived from the aggregate forms. How might the GH aggregates be packed within the secretory granule? Lewis et al. (19) showed that tryptic digests of the 22K GH dimer generated no new disulfide peptides other than those that were contained in the monomer, a result that suggested an antiparallel arrangement of cystines. Simple extension of this configuration could account for higher multiples of the monomer. Since 20K GH is also contained in rat pituitary (43), heteropolyomers of 20K or 22K GH could be present in tissue extracts. This issue was studied in 20 extracts by generating a frequency histogram of multimers that could be detected on the immunoblots. The result shows that most of the aggregates are multiples of either 20K or 22K monomers (Fig. 8). Although the significance of this interesting result is unknown, it is tempting to speculate that different granules and/or somatotrophs could contain unique aggregates of GH molecules. In the past, heterogeneity in the rat pituitary GH system has been demonstrated 1) by cell separation/ culture studies (28), 2) by measurements of hormone activities (4, 37, 48), and 3) by biochemical studies of the GH molecules themselves (48). Our study shows the importance of GH aggregates in this system and ties
Downloaded from https://academic.oup.com/endo/article-abstract/126/3/1630/2532990 by University of Newcastle user on 03 February 2019
represent disulfide-linked aggregates. For example, the higher mol wt forms corresponded to the expected mol wt of dimers, trimers, etc., of monomeric 20K and 22K GH. The disappearance of GH forms greater than 25K on reducing gels is consistent with the idea that these forms are disulfide linked. High mol wt disulfide-linked forms of placental lactogens have also been reported recently (3). The results of Bell et al. (42) showed that a majority of the GH contained in purified bovine GH granules consisted of dimers. Furthermore, the studies of Lorenson et al. (43) inferred that GH aggregates were contained in secretory granules. We, therefore, tested the idea that GH aggregates were localized in granules. To do this, three experimental test systems were chosen, each because it caused a perturbation within the GH secretory granule compartment. Our strategy was to see if (and how) that perturbation affected the state of GH aggregation. In the first experimental series we reasoned that GH aggregates should be recovered in a 40,000 X g pellet prepared from 0.25 M sucrose homogenates, since this speed is sufficient to quantitatively sediment GH granules (39). It was, therefore, surprising to find virtually identical distribution patterns of aggregate forms between the 40,000 x g pellet and supernatant fractions. However, since Zn2+ inhibits GH release from isolated secretory granules in vitro (43), we repeated the experiment in the presence of ZnCl2. The redistribution of a majority of the GH aggregates to the 40,000 X g pellet implies that Zn2+ somehow stabilizes hormone within the particle, perhaps via chelation to intermolecular disulfide bonds. Other findings document the importance of Zn2+ in GH cell structure and function. It is known, for example, that zinc ion is localized in GH secretory granules (27). Furthermore, zinc-deficient diets are associated with slow growth (44, 45). In the second experimental series, our method to separate somatotroph subpopulations by density gradient centrifugation (28) offered a convenient way to study GH forms contained in lightly granulated us. heavily granulated somatotrophs. Ultrastructural studies indicate that the less dense GH cells contain only 1-5% of the secretory granule complement that the more dense cells have. The blot profiles of high mol wt hormone extracted from the dense cells obviously correlate with their high secretory granule content. Since culture medium from well granulated GH cells contains hormone whose immunoreactivity is enhanced 5-fold upon chemical reduction, it seems clear that GH aggregates are also secreted from a subpopulation of GH cells in vitro. Aggregates are also found in plasma; it has been suggested that these have unique roles in either metabolic maintenance or receptor-mediated events (21, 22).
Endo • 1990 Vol 126-No 3
GH AGGREGATES IN RAT ADENOHYPOPHYSIS
1637
25 i 20 S
22
15 -\
CD
£
10 5 -
U, 30
Molecular Weight 1000
together previous results obtained solely with GH cells of GH molecules.
References 1. Benveniste R, Stachura ME, Szabo M, Frohman LA 1975 Big growth hormone (GH): conversion to small GH without peptide bond cleavage. J clin Endocrinol Metab 41:422 2. Bauman G, Abramson EC 1983 Urinary growth hormone in man: evidence for multiple molecular forms. J Clin Endocrinol Metab 56:305 3. Silverlight JJ, Prysor-Jones RA, Jenkins JS 1985 Growth hormone in normal female rat plasma appears as a large molecular weight form. Life Sci 36:1927 4. Sinha YN 1980 Molecular size variants of prolactin and growth hormone in mouse serum: strain differences and alterations of concentrations by physiological and pharmacological stimuli. Endocrinology 107:1959 5. Stolar MW, Baumann G 1986 Big growth hormone forms in human plasma: immunochemical evidence for their pituitary origin. Metabolism 35:75 6. Sinha R, Seavy B, Lewis U 1974 Heterogeneity of human growth hormone. Endocr Res Commun 1:449 7. Lewis U, Dunn J, Bonewald L, Seavey B, VanderLaan W 1978 A naturally occurring structural variant of human growth hormone. J Biol Chem 253:2679 8. Lewis U, Bonewald L, Lewis L 1980 The 20,000 dalton variant of human growth hormone: locations of the amino acid deletions. Biochem Biophys Res Commun 92:511 9. Wallis M 1980 Growth hormone: deletions in the protein and introns in the gene. Nature 284:512 10. Sinha R, Lewis U 1986 A lectin-binding assay indicates a possible glycosylated growth hormone in the human pituitary gland. Biochem Biophys Res Commun 140:491 11. Sinha R, Jacobsen BP 1987 Glycosylated growth hormone: detection in murine pituitary gland and evidence of physiological fluctuations. 145:1368 12. Ellis S, Nuenke JM, Grindeland RE 1968 Identity between the growth hormone degrading activity of the pituitary gland and plasmin. Endocrinology 83:1029 13. Macaig T, Forand R, Kelley P, Canalis 1979 The identification and characterization of a growth activator. In Vitro 15:188 (Abstract) 14. Lewis UJ 1984 Variants of growth hormone and prolactin and their postranslational modifications. Annu Rev Physiol 46:33 15. Liberti JP, Joshi GS 1986 Synthesis and secretion of phosphorylated growth hormone by rat pituitary glands in vitro. Biochem Biophys Res Commun 137:806 16. Liberti JP, Antoni BA, Chiebowski JF 1985 Naturally-occurring pituitary growth hormone is phosphorylated. Biochem Biophys
90
60
(n=20
qe|$) a
Res Commun 137:713 17. Wright DR, Goodman AD, Trimble KD 1975 Studies on "big" growth hormone from human plasma and pituitary. J Clin Invest 54:1064 18. Stolar MW, Amburn K, Baumann G 1984 Plasma "big" and "bigbig" growth hormone (GH) in man: an oligomeric series composed of structurally diverse GH monomers. J Clin Endocrinol Metab 59:212 19. Lewis U, Peterson SM, Bonewald LF, Seavy BK, VanderLaan WP 1977 An interchain disulfide dimer of growth hormone. J Biol Chem 252:3697 20. Sigel MB, VanderLaan WP, Kobrin MS, VandeLaan EF and Thorpe NA 1982 The biological half life of human growth hormone and a biologically active 20,000 dalton variant in mouse blood. Endocr Res Commun 9:67 21. Baumann G, Stolar MW, Buchanan TA 1985 Slow metabolic clearance rate of the 20,000-dalton variant of human growth hormone: implications for biological activity. Endocrinology 117:1309 22. Hendricks CM, Eastman RC, Takeda S, Asakawa K, Gorden P 1983 Plasma clearance of intravenously administered pituitary human growth hormone: gel filtration studies of heterogeneous components. J Clin Endocrinol Metab 60:864 23. Lorenson MY, Miska SP, Jacobs LS 1984 Molecular mechanisms of prolactin release from pituitary secretory granules. In: Mena F, Valverde CM (eds) Prolactin Secretion: A Multidisciplinary Approach. Academic Press, New York, p 141 24. Hervas F, Morreale de Escobar G, Escobar del Rey F 1975 Rapid effects of single doses of L-thyroxine and triiodo-L-thyronine on growth hormone, as studied in the rat by radioimmunoassay. Endocrinology 97:91 25. Augustine EC, Hymer WC 1978 Thyroid hormone effects on protein and RNA metabolism in the anterior pituitary. Mol Cell Endocrinol 10:225 26. Surks MI, DeFesi CR 1977 Determination of the cell number of each cell type in the anterior pituitary of euthyroid and hypothyroid rats. Endocrinology 101:946 27. Thorlacius-Ussing O, Flyvbjerg A, Esmann J 1987 Evidence that selenium induces growth retardation through reduced growth hormone and somatomedin C production. Endocrinology 120:659 28. Snyder G, Hymer WC, Snyder J 1977 Functional heterogeneity in somatotrophs isolated from the rat anterior pituitary. Endocrinology 101:788 29. Hopkins CR, Farquhar MG 1975 Hormone secretion by cells dissociated from rat anterior pituitaries. J Cell Biol 59:276 30. Bradford MM 1976 Protein assay utilizing Coomassie brilliant blue G-250. Anal Biochem 72:248 31. Laemmli UK 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680 32. Towbin H, Staehelin T, Gordon J 1979 Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: proce-
Downloaded from https://academic.oup.com/endo/article-abstract/126/3/1630/2532990 by University of Newcastle user on 03 February 2019
FiG. 8. Frequency distribution histogram of GH aggregate sizes contained in 20 different pituitary extracts electrophoresed on separate gels.
1638
GH AGGREGATES IN RAT ADENOHYPOPHYSIS
immunobiochemical detection. Biotechniques 3:276 42. Bell JA, Moffat K, Vonderhaar BK, Golde DW 1985 Crystallization and preliminary X-ray characterization of bovine growth hormone. J Biol Chem 260:8520 43. Lorenson MY, Robson DL, Jacobs LS 1983 Detectability of pituitary PRL and GH by immunoassay is increased by thiols and suppressed by divalent cations. Endocrinology 112:1880 44. Coble YD, Bardin CW, Ross GT, Darby WT 1971 Studies of endocrine function in boys with retarded growth, delayed sexual maturation and zinc deficiency. J Clin Endocrinol Metab 32:361 45. Oner G, Bhaumick B, Bala RM 1984 Effect of zinc deficiency on serum somatomedin levels and skeletal growth in young rats. Endocrinology 114:1860 46. Tietjen GH, Gindeland RE, Hymer WC, Vasques M, Holley DC, Somatostatin (SRIF) and growth hormone releasing factor (rGRF43) differentially regulate immunoassayable and bioassayable growth hormone secretion from two populations of rat somatotrophs. 71st Annual Meeting of the Endocrine Society, Seattle WA, 1989 (Abstract), p 219 47. Hymer WC, Wilbur DL 1980 Dispersed pituitary cells: their use in vitro and in vivo. In: Mahesh V, Muldoon T, Saxena B, Sadler W (eds) Functional Correlates of Hormone Receptors in Reproduction. Elsevier/North-Holland, p 13-44 48. Sigel MV, VanderLaan WP, VandeLaan EF, Lewis UJ 1980 Measurement of multiple forms of hGH: cross reactivities in conventional and two-chain radioimmunoassays. Endocrinology 106:92
Downloaded from https://academic.oup.com/endo/article-abstract/126/3/1630/2532990 by University of Newcastle user on 03 February 2019
dure and some applications. Proc Natl Acad Sci USA 76:4350 33. Ellis S, Grindeland RE, Nuenke JM, Callahan PX 1986a Isolation and properties of rat and rabbit growth hormones. Ann NY Acad Sci 148:328 34. Walker JE, Auffret AD, Came A, Gurnett A, Haniseh P, Saraste M 1982 Solid phase sequence analysis of polypeptides eluted from polyacrylamide gels. Eur J Biochem 123:2543 35. Wilfinger WW, Larsen WJ, Downs TR, Wilbur DJ 1982 An in vitro model for studies of intracellular communication in cultured rat anterior pituitary cells. Tissue Cell 16:483 36. Wilfinger W, Davis J, Augustine E, Hymer WC 1979 The effects of culture conditions on PRL and GH production by rat anterior pituitary cells. Endocrinology 105:530 37. Farrington MA, Hymer WC 1987 An enzyme immunoassay for rat growth hormone: applications to the study of GH variants. Life Sci 40:2479 38. Helmerhorst E, Stokes GB 1980 Microcentrifuge desalting: rapid quantitative method for desalting small amounts of protein. Anal Biochem 104:130 39. Lorenson MY, Jacobs LS 1982 Thiol regulation of protein, growth hormone, and prolactin released from isolated adenohypophysial secretory granules. Endocrinology 110:1164 40. Yokoya S, Friesen HG 1986 Human growth hormone (GH)-releasing factor stimulates and somatostatin inhibits the release of rat GH variants. Endocrinology 119:2097 41. Bers G, Garfin D 1985 Protein and nucleic acid blotting and
Endo • 1990 Vol 126 • No 3
Vol. 126, No. 3 Printed in U.S.A.
M. FARRINGTONf AND W. C. HYMER Department of Molecular and Cell Biology, Pennsylvania State University, University Park, Pennsylvania 16802
ABSTRACT. Although it has been known for some time that
rats, but were in glands from the thyroidectomized rats injected
GH aggregates are contained within the rat anterior pituitary
with T4; 3) that GH aggregates were uniquely associated with a
gland, the role that they might play in pituitary function is unknown. The present study examines this issue using the technique of Western blotting, which permitted visualization of 11 GH variants with apparent mol wt ranging from 14-88K. Electroelution of the higher mol wt variants from gels followed by their chemical reduction with /3-mercaptoethanol increased GH immunoassayability by about 5-fold. With the blot procedure we found 1) that GH aggregates greater than 44K were associated with a 40,000 X g sedimentable fraction; 2) that GH aggregates were not present in glands from thyroidectomized
heavily granulated somatotroph subpopulation isolated by density gradient centrifugation; and 4) that high mol wt GH forms were released from the dense somatotrophs in culture, since treatment of the culture medium with /3-mercaptoethanol increased GH immunoassayability by about 5-fold. Taken together, the results show that high mol wt GH aggregates are contained in secretory granules of certain somatotrophs and are also released in aggregate form from these cells in vitro. (Endocrinology 126: 1630-1638, 1990)
M
thermore, packaging of GH polymers within the somatotroph secretory granule may lead to changes in intragranular osmotic pressure, which, in turn, facilitate exocytosis (23). GH secretory granules are known to be affected by the thyroid hormone status of the host (24-26), zinc ions (27), and somatotroph subtype (28, 29). We have evaluated the extent of GH aggregation under each of these three conditions by Western blotting.
ULTIPLE forms of GH are contained in the pituitary gland, blood, and urine (1-5). A family of GH variants, all structurally related to the 22K form of the hormone, include 1) the 20K form, which results from alternate mRNA splicing (6-9); 2) a 23K glycosylated variant (10, 11); 3) a two-chain 25K form resulting from proteolytic cleavage in the large disulfide loop (1214); 4) phosphorylated GH (15, 16); and 5) disulfidelinked aggregated forms (17, 18). The physiological significance of many of these GH variants is poorly defined. This is especially true for the aggregate forms. Dimerization of GH probably results from antiparallel alignment of the cystines in two GH monomers (19). The dimer shows relatively poor immunoreactivity (10-40% of the monomer) (20) and has only 15% of its activity in the tibial line assay (19). Nevertheless, there is evidence to support the hypothesis that GH aggregates have physiological significance. For example, GH dimers are present in the circulation; they have a longer half-life than the monomeric form (21, 22). Fur-
Materials and Methods Animals Sprague-Dawley male rats (Harlan, Indianapolis, IN), weighing 300-450 g, served as pituitary donors. In some experiments animals were chemically throidectomized by incorporating 6propyl-2-thiouracil (PTU; 0.5 mg/ml; Sigma, St. Louis, MO) and glucose (50 mg/ml) in their drinking water for 42 days. Controls received only glucose in their water. At death, the mean body weight of control animals was 491 ± 19 g; PTUtreated animals weighed 356 ± 12 g.
Received July 26, 1989. Address all correspondence and requests for reprints to: Dr. W. C. Hymer, 401 Althouse Laboratory, University Park, Pennsylvania 16802. * Presented in part at the 70th Annual Meeting of The Endocrine Society, New Orleans, LA, June 1988. This work was supported by NASA Grants NCC 2-370, 9-17416, and NAGW 1196. t Submitted in partial fulfillment for requirements of the Ph.D. degree in Biochemistry at the Pennsylvania State University.
GH extraction Individual pituitary glands were homogenized in 1 ml 0.01 N NaHCO3 at 4 C using 10 strokes of a ground glass McShan type homogenizer (Bellco Glass, Inc., Vineland, NJ). These homogenates were kept on ice for up to 24 h before centrifugation (40,000 x g; 20 min; 4 C). The supernatant fraction was flash frozen and stored at —70 C. Under these conditions GH
1630
Downloaded from https://academic.oup.com/endo/article-abstract/126/3/1630/2532990 by University of Newcastle user on 03 February 2019
Growth Hormone Aggregates in the Rat Adenohypophysis*
GH AGGREGATES IN RAT ADENOHYPOPHYSIS yield was 120 ± 10 ng GH/mg protein (Bradford method) (30). Storage of the homogenates for 0.5 or 2 h at 4 C yielded 50% Or 70% of that obtained at 24 h. Reextraction of the 40,000 x g pellet prepared from the 24-h samples yielded less than 10% of the initial GH extracted. Therefore, a single 24-h extraction procedure was used in this study.
Samples to be run on sodium dodecyl sulfate-polyacrylamide gels (SDS-PAGE) were prepared in 0.025 M Tris (pH 6.8), 2% SDS, 10% glycerol, and 0.0001% bromphenol blue. If reduced, they also contained 5% mercaptoethanol. The samples were heated in a boiling water bath for 5 min and cooled to room temperature before loading onto the gel. SDS-PAGE was performed in a discontinuous Tris-glycine buffer system at 25 C (31). Linear (5-15%) gradient separating gels (0.75 mm thick) were poured using a Hoeffer 30-ml gradient marker (Hoeffer Scientific, San Francisco, CA); these were overlayered with a stacking gel (5%) prepared in 0.25 M Tris, pH 6.8. The running buffer was 0.025 M Tris, 0.1% SDS, and 0.192 M glycine at pH 8.3. Protein mobility was monitored using solutions of 0.02% (wt/vol) bromophenol blue in the cathodic buffer reservoir. Western blot Proteins in polyacrylamide gels were electrophoretically transferred onto nitrocellulose paper according to a modification of the method of Towbin et al. (32). Prewetted pure nitrocellulose paper (Bio-Rad, Richmond, CA; 0.45 nm) was rolled onto the gel surface to promote a seal between the paper and gel, thereby facilitating transfer. Transfer was performed at constant amperage (300 mamp/gel) for 6 h in 0.10 M Tris, pH 9.3, at temperatures below 25 C. Immediately after transfer the paper was dried either in air or in a vacuum oven at 60 C for 10 min. This step decreased diffusion of protein bands. All subsequent steps were carried out in Seal-a-meal bags that were purged of air bubbles. The paper was first incubated in 0.01 M PBS, pH 7.4, containing 1% BSA at room temperature for 4 h or overnight at 4 C to block nonspecific binding sites. Immunoperoxidase staining of GH on the paper was accomplished by 1) incubation with monkey antirat GH antiserum (kindly provided by R. Grindeland, NASA Ames Research Center; 1:40,000 final dilution) (33) for 3 days at 4 C or 2 days at room temperature; 2) incubation with horseradish peroxidase-conjugated goat antimonkey immunoglobulin G (IgG; Cappel Laboratories, Cochranville, PA; 1:4,000 final dilution) for 6 h at room temperature; 3) overnight incubation at 4 C with rabbit antigoat IgG (Cappel) at 1:2,000 final dilution; 4) incubation for 4 h at room temperature in goat peroxidase antiperoxidase (Cappel; 1:4,000 final dilution); 5) incubation with enzyme substrate [0.2 mg 3,3'-tetrahydrochlorodiamino-benzidine; Sigma; 4 /ul 3% (vol/vol) hydrogen peroxide/ml PBS] for no longer than 1 h at room temperature. If 0.01 M sodium citrate, pH 5.2, was used in place of PBS in step 5, steps 3 and 4 could be eliminated. The nitrocellulose paper was washed four times with 0.15 M NaCl between each of the steps above. At the last step the blots were washed in distilled H2O, air dried, and analyzed by reflectance densitometry within 2 weeks.
Prestained mol wt markers (Bethesda Research Laboratories, Gaithersburg, MD) were used to estimate mol wt of transferred proteins. The coefficients of variation for the Rf values of different markers, based on analysis of 17 gels, were: /3lactoglobin, 0.9%; a-chymotrypsinogen, 1.6%; ovalbumin, 1.7%; phosphorylase-B, 2.2%; and myosin H chain, 5.9%. The relative reflective optical densities of GH Western blots were determined using a Bio-Rad 620 videodensitometer. The optical density profiles of GH provided a convenient technique for comparison of individual GH forms. Quantitation of these profiles was not attempted for two reasons. First, the time required for transfer of high mol wt forms sometimes permitted smaller GH forms to pass through to the other side. Second, several variants have different (6) or unknown antigenicities. Background staining was subtracted from each sample trace by a computer program incorporated into the densitometer. In every case this was determined on an empty lane that was adjacent to the sample lane in the same electorphoresis-blot trial. Protein elution Protein elution from the SDS-gel was performed by cutting a 2-cm wide slice and subjecting that slice to Western blotting (see above). The remaining portion of the gel was wrapped in plastic film, stored at -70 C, thawed in elution buffer (0.1 M Tris, pH 8.6), and finally cut into regions of GH staining after alignment with the template. The gel pieces were each homogenized in elution buffer, and proteins eluted electrophoretically (34). Briefly, this procedure involved 1) electrophoresis at constant voltage (500 V; 24 h) of protein through a glass tube (plugged at the end with 5% stacking gel) into dialysis tubing; 2) current reversal for 2 min to minimize protein attachment to the tubing; 3) dialysis against water overnight; and 4) lyophilization. Recovery of GH using this procedure was 111 ± 8% (three experiments). Differential centrifugation Cell-free homogenates were prepared in 0.25 M sucrose with or without 0.2 mM ZnCl2 (two pituitaries per ml) at 4 C using a Teflon homogenizor (Belco Glass, Vineland, NJ). After centrifugation at 275 X g for 10 min to remove nuclei, the supernatant fraction was centrifuged at 40,000 X g for 20 min. GH contained in the homogenate, the 40,000 X g pellet, and final supernatant fraction was obtained by extraction with NaHCO3 (0.01 M final concentration). Pituitary cell separation and culture Cells were prepared from pituitary glands by trypsinization (35). Yields were 2-3 X 106/gland, with viability greater than 95%. Somatotrophs were fractionated into two subpopulations by gradient centrifugation (28). Total cell recovery and GH recovery after this procedure were 90 ± 2% and 86 ± 3%, respectively. Cell culture was performed in Limbro plates (Flow Laboratories, McLean, VA) at a density of 1.5 x 105 cells/ml for 3 days at 37 C in humidified air plus 5% CO2. The culture medium was Minimum Essential Medium containing 5% horse serum and antibiotics (36).
Downloaded from https://academic.oup.com/endo/article-abstract/126/3/1630/2532990 by University of Newcastle user on 03 February 2019
Electrophoresis
1631
GH AGGREGATES IN RAT ADENOHYPOPHYSIS
1632 GH immunoassay
Data analysis Data are reported as the mean ± SEM. Duncan's multiple range test was used to measure differences.
Results Western blot methodologies GH molecules having apparent mol wt of 14-88K were routinely observed in pituitary extracts that were electrophoresed under nonreducing conditions, transferred to nitrocellulose paper, and then stained immunochemically for the presence of GH (Fig. 1A). Reflectance profiles of such immunoblots also showed heterogenity in GH forms on the paper (Fig. 1, middle). Immunoreactive GH forms greater than 25K disappeared upon chemical reduction of the extracts before electrophoresis (Fig. IB). The results of three tests helped to establish the validity of the GH immunostaining. First, use of the NIDDK antirat GH serum (1:8000 final dilution) in place of the flbsorbance
B
Grindeland antiserum yielded GH profiles identical to those in Fig. 1A. Second, replacement of the Grindeland antiserum with either normal monkey serum (1:8000) or BSA (2%) abolished GH staining. Third, preincubation of the Grindeland antiserum with 20, 40, or 80 ng GH (RP-1) for 24 h at 4 C before use resulted in severe dimunition (20 and 40 /zg) to elimination (80 fig) in staining of all GH bands (data not shown). When a glycine-methanol buffer (32) was used for blotting, transfer of high mol wt GH forms was severely inhibited, even though transfer time was increased by 10 h. Inhibition of transfer of high mol wt proteins was documented by silver staining of the gel after blotting. Results of other experiments, not shown here, indicated that GH profiles 1) were identical between fresh and frozen extracts, 2) were not affected by inclusion of aprotinin (2 U/ml) in the extract, and 3) were identical between extracts prepared in the presence or absence of iodoacetamide (1 x 10~3 M). Isoelectric focusing (IEF) The possibility that charge heterogeneity could in part account for the GH variants seen in Fig. 1 was addressed by focusing pituitary extracts in thin (0.5-mm) polyacrylamide-based gels containing ampholytes (pH 3.5-10), followed by pressure immunoblotting. The results revealed the presence of five or six immunoactive GH forms that had isoelectric points ranging from 4.8-7.1 regardless of whether the extracts were prepared in acetic acid (pH 3.3; Fig. 2A), Na2CO3 buffer (pH 10; Fig. 2B), or NaHCO3 (pH 8; Fig. 2C). In another experiment, unstained IEF gels were divided into five regions that contained GH of varying pi, viz. 4.8, 5.4, 5.9, 6.1-6.9, and 7.1. Proteins were then eluted from these IEF gels, lyophilized, electrophoresed by SDS-PAGE under nonreducing conditions, and immunoblotted. Although the mol wt patterns of GH variants were similar between IEF regions (Fig. 3), there were minor differences. For example, the 20K GH and its multimers tended to be found in the more basic regions, while the 25K variant tended to be found in the more acidic regions. /3ME treatment
25K22K20K" 14K-
FlG. 1. Representative Western blots of rat pituitary GH contained in extracts electrophoresed under nonreducing (A) or reducing (B) conditions (see Materials and Methods). The reflectance optical density tracing of blot in A is shown in the middle panel.
Since treatment of alkaline extracts with /3ME enhances the immunoreactivity of GH (37), and since this treatment also eliminates the high mol wt GH variants seen in nonreducing gels (Fig. IB), it was of interest to combine the two approaches to further study high mol wt GH variants. Shown in Table 1 are data which confirm previous findings that chemical reduction enhances GH immunoassayability and further show the repeatability of that effect. Maximal enhancement was achieved within 1 h (data not shown).
Downloaded from https://academic.oup.com/endo/article-abstract/126/3/1630/2532990 by University of Newcastle user on 03 February 2019
GH levels were measured by a specific enzyme immunoassay using three dilutions in duplicate (37). In some cases samples were treated with an equal volume of 0.4 M 2-mercaptoethanol (2-ME) for 1 h at 4 C. Before GH assay it was necessary to remove the 2-ME by centrifugation through Sephadex G-10 columns (38). Control samples were treated with water instead of 2-ME.
Endo• 1990 Voll26«No3
GH AGGREGATES IN RAT ADENOHYPOPHYSIS
1633
0.60-r
0.30--
Region 4 (pi 6.6, 6.9)
Region 3 (pi 5.9)
Region 2 (pi 5.4)
Region I (pl 4.8) 15
30
45
60
75
90
120
105
135 150
(•)
Migration in Gel (mm)
FIG. 3. Reflectance tracings of GH blots after two-dimensional gel electrophoresis. GH forms separated by IEF were eluted, dialyzed, lyophilized, reconstituted, and reelectrophoresed on SDS-PAGE gels (n = 2 experiments).
B
TABLE 1. Effect of /3ME treatment of rat pituitary extracts on GH immunoassayability
FIG. 2. Representative GH Western blots of IEF gels electrophoresed under different conditions: A, 0.01 N HOAc, pH 3.3; B, 0.05 N Na2CO3, pH 10; and C, 0.01 N Na2CO3, pH 8.0 (n = 3 experiments), aw, Application wick.
Electrocution of GH from slices of SDS-polyacrylamide gels followed by chemical reduction and GH immunoassay (see Materials and Methods) clearly showed enhancement of hormone activity in the higher mol wt regions (Table 2). Physiological studies Subcellular distribution. Alkaline extracts of a crude 40,000 x g mitochondrial-granule pellet (39) prepared from 0.25 M sucrose homogenates contained high mol wt GH variants (Fig. 4B). Extracts of the 40,000 X g supernatant fraction revealed virtually identical profiles (Fig. 4C). The distribution of recovered GH in this and other experiments, measured by immunoassay, was 32 ± 5% in the 40,000 X g pellet and 66 ± 4% in the 40,000 X g supernatant fraction. However, when homogenates were prepared in 0.25 M sucrose containing 0.2 mM ZnCl2, the distribution patterns of high mol wt GH were markedly
0ME
GH fold increase 1 1.39 ± 0.09 3.38 ± 0.73
molarity 0.00 0.05 0.20
Data represent the increase in G H immunoactivity (37) after /3ME treatment (n = 7 experiments). See Materials and Methods for additional details.
T A B L E 2. Effect of /3ME treatment of rat pituitary extracts on immunoassayability of different G H forms separated by SDS-PAGE
GH (Mg) Region
mol wt (kD)
Fold increase -j3ME
land 2 3 4 5 6 7 and 8
44K) GH forms in extracts of pituitary tissue has been recognized previously (5,18). However, their visualization on blots is not routine, probably because of technical issues relating to the Western blotting procedure itself. For example, 15% gels appear to have been used most often, and these have a fractionation range of 12-45K; in such gels 22K GH has an Rf of 0.42, and its dimer has an Rf of 0.14 (40). As a result, higher mol wt variants would probably be excluded from the separating gel. On the other hand, the fractionation range of 5-15% gels is about 14-250K. The most commonly used blotting buffers contain glycine and methanol (32). Methanol 1) counteracts swelling of the gel during blotting, 2) increases hydrophobic sites on nitrocellulose, 3) removes SDS from SDSprotein complexes (41), and 4) reduces pore size, thus
0 -i0
15
30
45
60
75
90
105 120
135
150
Migration in Gel (mm)
FIG. 7. Representative reflectance tracings of GH Western blots of tissue extracts prepared from A, pituitary glands of rats receiving sucrose in their drinking water for 42 days and a single injection of saline 24 h before death; B, pituitary glands of rats receiving sucrose in their drinking water for 42 days and a single injection of T4 24 h before death; C, pituitary glands of rats receiving sucrose and PTU in their drinking water for 42 days and a single injection of saline 24 h before death; and D, pituitary glands of rats receiving sucrose and PTU in their drinking water for 42 days and a single injection of T4 24 h before death (n = 2 experiments).
restricting transfer of larger molecules and causing protein precipitation in the gel. Elimination of methanol and glycine in the blotting buffer increased transfer efficiency and often GH staining intensity as well. Taken together, these modifications allowed us to study GH aggregates in a new way. Since the antisera used in our study was directed against 22K GH, it is not surprising that 22K and 25K were in the most heavily stained forms. The 25K form probably represents clipped GH (13, 14), while the 23K form may be the glycosylated variant (10, 11). It is probable that the higher mol wt forms are not attributable 1) to aggregations of clipped GH, since inclusion of aprotinin did not affect GH profiles; 2) to artificial disulfide-linked aggregation generated during extraction, since inclusion of iodoacetamide did not change GH profiles; 3) to deamidation, since extraction in buffers of different pH did not affect the GH profiles; or 4) to unique charge variants, since most of the mol wt variants were found in each of five variants isolated by IEF. It seems probable that the higher mol wt GH variants
Downloaded from https://academic.oup.com/endo/article-abstract/126/3/1630/2532990 by University of Newcastle user on 03 February 2019
Average
0.80 T
1635
1636
GH AGGREGATES IN RAT ADENOHYPOPHYSIS
Some years ago studies by Ellis et al. (33) showed that high mol wt forms of rat GH in tissue or plasma were rich in tibial line activity, but poor in immunological activity. Since culture medium from heavily granulated somatotrophs is potent in GH tibial line assay activity (46), and since implantation of heavily granulated (but not lightly granulated) somatotrophs into the cerebral ventricles of hypophysectomized rats restores animal growth (47), we suggest that the disulfide-linked GH aggregates identified in our study could represent a more bioactive form of the hormone molecule that its monomer. In the third experimental series we took advantage of the well documented effects of thyroidectomy on somatotroph degranulation (25, 26) as well as somatotroph regranulation after T4 injection (24). Since GH aggregates disappeared after thyroidectomy and reappeared after T4 administration, their association with secretory granules was again suggested. In summary, the results from these three experimental test systems offer strong support for the hypothesis that GH aggregates are localized within the GH secretion granule and that they are released from heavily granulated GH cells. Of course they do not eliminate the possibility that aggregates are also localized in other subcellular compartments. It is worthwhile to note that the GH aggregates appear to account for a major percentage of the total measurable hormone; the data in Tables 2 and 3 suggest that more than 70% of the total measurable extracted or secreted GH measured after chemical reduction was derived from the aggregate forms. How might the GH aggregates be packed within the secretory granule? Lewis et al. (19) showed that tryptic digests of the 22K GH dimer generated no new disulfide peptides other than those that were contained in the monomer, a result that suggested an antiparallel arrangement of cystines. Simple extension of this configuration could account for higher multiples of the monomer. Since 20K GH is also contained in rat pituitary (43), heteropolyomers of 20K or 22K GH could be present in tissue extracts. This issue was studied in 20 extracts by generating a frequency histogram of multimers that could be detected on the immunoblots. The result shows that most of the aggregates are multiples of either 20K or 22K monomers (Fig. 8). Although the significance of this interesting result is unknown, it is tempting to speculate that different granules and/or somatotrophs could contain unique aggregates of GH molecules. In the past, heterogeneity in the rat pituitary GH system has been demonstrated 1) by cell separation/ culture studies (28), 2) by measurements of hormone activities (4, 37, 48), and 3) by biochemical studies of the GH molecules themselves (48). Our study shows the importance of GH aggregates in this system and ties
Downloaded from https://academic.oup.com/endo/article-abstract/126/3/1630/2532990 by University of Newcastle user on 03 February 2019
represent disulfide-linked aggregates. For example, the higher mol wt forms corresponded to the expected mol wt of dimers, trimers, etc., of monomeric 20K and 22K GH. The disappearance of GH forms greater than 25K on reducing gels is consistent with the idea that these forms are disulfide linked. High mol wt disulfide-linked forms of placental lactogens have also been reported recently (3). The results of Bell et al. (42) showed that a majority of the GH contained in purified bovine GH granules consisted of dimers. Furthermore, the studies of Lorenson et al. (43) inferred that GH aggregates were contained in secretory granules. We, therefore, tested the idea that GH aggregates were localized in granules. To do this, three experimental test systems were chosen, each because it caused a perturbation within the GH secretory granule compartment. Our strategy was to see if (and how) that perturbation affected the state of GH aggregation. In the first experimental series we reasoned that GH aggregates should be recovered in a 40,000 X g pellet prepared from 0.25 M sucrose homogenates, since this speed is sufficient to quantitatively sediment GH granules (39). It was, therefore, surprising to find virtually identical distribution patterns of aggregate forms between the 40,000 x g pellet and supernatant fractions. However, since Zn2+ inhibits GH release from isolated secretory granules in vitro (43), we repeated the experiment in the presence of ZnCl2. The redistribution of a majority of the GH aggregates to the 40,000 X g pellet implies that Zn2+ somehow stabilizes hormone within the particle, perhaps via chelation to intermolecular disulfide bonds. Other findings document the importance of Zn2+ in GH cell structure and function. It is known, for example, that zinc ion is localized in GH secretory granules (27). Furthermore, zinc-deficient diets are associated with slow growth (44, 45). In the second experimental series, our method to separate somatotroph subpopulations by density gradient centrifugation (28) offered a convenient way to study GH forms contained in lightly granulated us. heavily granulated somatotrophs. Ultrastructural studies indicate that the less dense GH cells contain only 1-5% of the secretory granule complement that the more dense cells have. The blot profiles of high mol wt hormone extracted from the dense cells obviously correlate with their high secretory granule content. Since culture medium from well granulated GH cells contains hormone whose immunoreactivity is enhanced 5-fold upon chemical reduction, it seems clear that GH aggregates are also secreted from a subpopulation of GH cells in vitro. Aggregates are also found in plasma; it has been suggested that these have unique roles in either metabolic maintenance or receptor-mediated events (21, 22).
Endo • 1990 Vol 126-No 3
GH AGGREGATES IN RAT ADENOHYPOPHYSIS
1637
25 i 20 S
22
15 -\
CD
£
10 5 -
U, 30
Molecular Weight 1000
together previous results obtained solely with GH cells of GH molecules.
References 1. Benveniste R, Stachura ME, Szabo M, Frohman LA 1975 Big growth hormone (GH): conversion to small GH without peptide bond cleavage. J clin Endocrinol Metab 41:422 2. Bauman G, Abramson EC 1983 Urinary growth hormone in man: evidence for multiple molecular forms. J Clin Endocrinol Metab 56:305 3. Silverlight JJ, Prysor-Jones RA, Jenkins JS 1985 Growth hormone in normal female rat plasma appears as a large molecular weight form. Life Sci 36:1927 4. Sinha YN 1980 Molecular size variants of prolactin and growth hormone in mouse serum: strain differences and alterations of concentrations by physiological and pharmacological stimuli. Endocrinology 107:1959 5. Stolar MW, Baumann G 1986 Big growth hormone forms in human plasma: immunochemical evidence for their pituitary origin. Metabolism 35:75 6. Sinha R, Seavy B, Lewis U 1974 Heterogeneity of human growth hormone. Endocr Res Commun 1:449 7. Lewis U, Dunn J, Bonewald L, Seavey B, VanderLaan W 1978 A naturally occurring structural variant of human growth hormone. J Biol Chem 253:2679 8. Lewis U, Bonewald L, Lewis L 1980 The 20,000 dalton variant of human growth hormone: locations of the amino acid deletions. Biochem Biophys Res Commun 92:511 9. Wallis M 1980 Growth hormone: deletions in the protein and introns in the gene. Nature 284:512 10. Sinha R, Lewis U 1986 A lectin-binding assay indicates a possible glycosylated growth hormone in the human pituitary gland. Biochem Biophys Res Commun 140:491 11. Sinha R, Jacobsen BP 1987 Glycosylated growth hormone: detection in murine pituitary gland and evidence of physiological fluctuations. 145:1368 12. Ellis S, Nuenke JM, Grindeland RE 1968 Identity between the growth hormone degrading activity of the pituitary gland and plasmin. Endocrinology 83:1029 13. Macaig T, Forand R, Kelley P, Canalis 1979 The identification and characterization of a growth activator. In Vitro 15:188 (Abstract) 14. Lewis UJ 1984 Variants of growth hormone and prolactin and their postranslational modifications. Annu Rev Physiol 46:33 15. Liberti JP, Joshi GS 1986 Synthesis and secretion of phosphorylated growth hormone by rat pituitary glands in vitro. Biochem Biophys Res Commun 137:806 16. Liberti JP, Antoni BA, Chiebowski JF 1985 Naturally-occurring pituitary growth hormone is phosphorylated. Biochem Biophys
90
60
(n=20
qe|$) a
Res Commun 137:713 17. Wright DR, Goodman AD, Trimble KD 1975 Studies on "big" growth hormone from human plasma and pituitary. J Clin Invest 54:1064 18. Stolar MW, Amburn K, Baumann G 1984 Plasma "big" and "bigbig" growth hormone (GH) in man: an oligomeric series composed of structurally diverse GH monomers. J Clin Endocrinol Metab 59:212 19. Lewis U, Peterson SM, Bonewald LF, Seavy BK, VanderLaan WP 1977 An interchain disulfide dimer of growth hormone. J Biol Chem 252:3697 20. Sigel MB, VanderLaan WP, Kobrin MS, VandeLaan EF and Thorpe NA 1982 The biological half life of human growth hormone and a biologically active 20,000 dalton variant in mouse blood. Endocr Res Commun 9:67 21. Baumann G, Stolar MW, Buchanan TA 1985 Slow metabolic clearance rate of the 20,000-dalton variant of human growth hormone: implications for biological activity. Endocrinology 117:1309 22. Hendricks CM, Eastman RC, Takeda S, Asakawa K, Gorden P 1983 Plasma clearance of intravenously administered pituitary human growth hormone: gel filtration studies of heterogeneous components. J Clin Endocrinol Metab 60:864 23. Lorenson MY, Miska SP, Jacobs LS 1984 Molecular mechanisms of prolactin release from pituitary secretory granules. In: Mena F, Valverde CM (eds) Prolactin Secretion: A Multidisciplinary Approach. Academic Press, New York, p 141 24. Hervas F, Morreale de Escobar G, Escobar del Rey F 1975 Rapid effects of single doses of L-thyroxine and triiodo-L-thyronine on growth hormone, as studied in the rat by radioimmunoassay. Endocrinology 97:91 25. Augustine EC, Hymer WC 1978 Thyroid hormone effects on protein and RNA metabolism in the anterior pituitary. Mol Cell Endocrinol 10:225 26. Surks MI, DeFesi CR 1977 Determination of the cell number of each cell type in the anterior pituitary of euthyroid and hypothyroid rats. Endocrinology 101:946 27. Thorlacius-Ussing O, Flyvbjerg A, Esmann J 1987 Evidence that selenium induces growth retardation through reduced growth hormone and somatomedin C production. Endocrinology 120:659 28. Snyder G, Hymer WC, Snyder J 1977 Functional heterogeneity in somatotrophs isolated from the rat anterior pituitary. Endocrinology 101:788 29. Hopkins CR, Farquhar MG 1975 Hormone secretion by cells dissociated from rat anterior pituitaries. J Cell Biol 59:276 30. Bradford MM 1976 Protein assay utilizing Coomassie brilliant blue G-250. Anal Biochem 72:248 31. Laemmli UK 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680 32. Towbin H, Staehelin T, Gordon J 1979 Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: proce-
Downloaded from https://academic.oup.com/endo/article-abstract/126/3/1630/2532990 by University of Newcastle user on 03 February 2019
FiG. 8. Frequency distribution histogram of GH aggregate sizes contained in 20 different pituitary extracts electrophoresed on separate gels.
1638
GH AGGREGATES IN RAT ADENOHYPOPHYSIS
immunobiochemical detection. Biotechniques 3:276 42. Bell JA, Moffat K, Vonderhaar BK, Golde DW 1985 Crystallization and preliminary X-ray characterization of bovine growth hormone. J Biol Chem 260:8520 43. Lorenson MY, Robson DL, Jacobs LS 1983 Detectability of pituitary PRL and GH by immunoassay is increased by thiols and suppressed by divalent cations. Endocrinology 112:1880 44. Coble YD, Bardin CW, Ross GT, Darby WT 1971 Studies of endocrine function in boys with retarded growth, delayed sexual maturation and zinc deficiency. J Clin Endocrinol Metab 32:361 45. Oner G, Bhaumick B, Bala RM 1984 Effect of zinc deficiency on serum somatomedin levels and skeletal growth in young rats. Endocrinology 114:1860 46. Tietjen GH, Gindeland RE, Hymer WC, Vasques M, Holley DC, Somatostatin (SRIF) and growth hormone releasing factor (rGRF43) differentially regulate immunoassayable and bioassayable growth hormone secretion from two populations of rat somatotrophs. 71st Annual Meeting of the Endocrine Society, Seattle WA, 1989 (Abstract), p 219 47. Hymer WC, Wilbur DL 1980 Dispersed pituitary cells: their use in vitro and in vivo. In: Mahesh V, Muldoon T, Saxena B, Sadler W (eds) Functional Correlates of Hormone Receptors in Reproduction. Elsevier/North-Holland, p 13-44 48. Sigel MV, VanderLaan WP, VandeLaan EF, Lewis UJ 1980 Measurement of multiple forms of hGH: cross reactivities in conventional and two-chain radioimmunoassays. Endocrinology 106:92
Downloaded from https://academic.oup.com/endo/article-abstract/126/3/1630/2532990 by University of Newcastle user on 03 February 2019
dure and some applications. Proc Natl Acad Sci USA 76:4350 33. Ellis S, Grindeland RE, Nuenke JM, Callahan PX 1986a Isolation and properties of rat and rabbit growth hormones. Ann NY Acad Sci 148:328 34. Walker JE, Auffret AD, Came A, Gurnett A, Haniseh P, Saraste M 1982 Solid phase sequence analysis of polypeptides eluted from polyacrylamide gels. Eur J Biochem 123:2543 35. Wilfinger WW, Larsen WJ, Downs TR, Wilbur DJ 1982 An in vitro model for studies of intracellular communication in cultured rat anterior pituitary cells. Tissue Cell 16:483 36. Wilfinger W, Davis J, Augustine E, Hymer WC 1979 The effects of culture conditions on PRL and GH production by rat anterior pituitary cells. Endocrinology 105:530 37. Farrington MA, Hymer WC 1987 An enzyme immunoassay for rat growth hormone: applications to the study of GH variants. Life Sci 40:2479 38. Helmerhorst E, Stokes GB 1980 Microcentrifuge desalting: rapid quantitative method for desalting small amounts of protein. Anal Biochem 104:130 39. Lorenson MY, Jacobs LS 1982 Thiol regulation of protein, growth hormone, and prolactin released from isolated adenohypophysial secretory granules. Endocrinology 110:1164 40. Yokoya S, Friesen HG 1986 Human growth hormone (GH)-releasing factor stimulates and somatostatin inhibits the release of rat GH variants. Endocrinology 119:2097 41. Bers G, Garfin D 1985 Protein and nucleic acid blotting and
Endo • 1990 Vol 126 • No 3
