- Email: [email protected]
Glucocorticoid-induced proteins in bovine endothelial cells
245
Molecular and Cellular Endocrinology, 32 (1983) 245-254 Elsevier Scientific Publishers Ireland. Ltd.
MCE 01054
GLUCOCORTICOID-INDUCED PROTEINS IN BOVINE ENDOTHELIAL CELLS N.R. NICHOLS ‘, C.J. LLOYD 2, F.A.O. MENDELSOHN J.W. FUNDER ‘,*
2, and
’ Medical Research Centre, Prince Henry’s Hospital, Melbourne, Victoria 3004, 2 Universiiy of Melbourne, Department of Medicine, Austin Hospital, Heidelberg, (Australia) Received
11 April
1983; accepted
Victoria
4 July 1983
High circulating levels of corticosteroids are associated with elevated blood pressure and an increased incidence of vascular damage. We have used high resolution two-dimensional gel electrophoresis and autoradiography to analyse direct steroid effects on the newly synthesized proteins in bovine aortic endothelial cells. At medium concentrations of lo-’ M, corticosteroids but not sex steroids increased the synthesis of two endothelial proteins: el (M.W. - 43K, p1 - 6.3) and e2 (M.W. - 28K, pI - 5.9). The responses were consistent for different endothelial cell cultures, different gel runs, and for natural and synthetic glucocorticoids, including the ‘pure’ synthetic glucocorticoid, RU 26988. Mineralocorticoids (aldosterone, deoxycorticosterone) at lo-’ M were less potent in their effect on the synthesis of e2, and had no discernible effect on the synthesis of el. These specific effects on cellular protein synthesis demonstrate that glucocorticoids can act directly on endothelial cells. These direct actions may thus mediate altered endothelial cell function in clinical and experimental cardiovascular conditions characterized by corticosteroid excess. Keywords:
R U 26988; endothelial protein synthesis; steroid domain; two-dimensional electrophoresis.
Factors influencing endothelial cell function may play a role in the development and maintenance of cardiovascular disease. In particular, high circulating levels of adrenal steroids have long been associated with an increased incidence of vascular damage and resulting atherosclerosis (Selye, 195 1; Cope, 1972; Brunner and Gavras, 1977). Endothelial cell injury by chemical as well as mechanical means is thought to be an initiating factor in the development of arterial disease (Ross, 1981). The molecular mechanisms underlying such injury, however, and the role of various agents, including steroids, are not well understood. * To whom correspondence 0303-7207/83/$03.00
should
be addressed.
0 1983 Elsevier Scientific
Publishers
Ireland,
Ltd.
246
N.R. Nichols et al.
Recently, Mendelsohn et al. (1982) have characterized the glucocorticoid induction of angiotensin-converting enzyme in cultured bovine aortic endothelial cells. These data are consistent with a direct effect of glucocorticoid-specific hormones on endothelial cells via receptor-initiated, DNA-dependent, mRNA-mediated protein synthesis. Changes in levels of vasoactive peptides by alterations in angiotensin-converting enzyme activity have been suggested to be responsible for endothelial apertures, increased vascular permeability and resultant vascular damage (Brunner and Gavras, 1977). In the present study we have used the techniques of high resolution two-dimensional (2-D) gel electrophoresis and autoradiography to examine the protein synthetic patterns of bovine endothelial cell cultures for additional steroid-responsive products which might alter endothelial cell function. Cells were cultured with a series of naturally occurring and synthetic steroids, representing the range of secretion products of the adrenal cortex. After 24 h incubation with glucocorticoid, mineralocorticoid, androgen or progesterone the effect of these various steroids on the protein synthetic profile was compared by autoradiography.
MATERIALS AND METHODS Chemicals and reagents [ 35S]Methionine in 0.02 M potassium acetate containing 0.1% /3mercaptoethanol (1200 Ci/rnmole) and [ “C]proteins of known molecu-
lar weight were obtained from the Radiochemical Centre (Amersham, U.K.). RU 26988 (1 1/3,17&dihydroxy- 17a-propynyl-androsta- 1,4,6-triene-3-one) was the gift of Roussel-UCLAF (Romainville, France), nonradioactive dexamethasone (DM) of Merck, Sharp and Dohme (Sydney, Australia) and 9cu-fluorocortisol (9cu-fF) of E.R. Squibb (Sydney, Australia); 5cu-dihydrotestosterone (Sa-DHT) was obtained from Ikapharm (Ramat-Gan, Israel), and other steroids from Steraloids (Wilton, N.H.). Steroids were stored as stock solutions (4 x 10e3 M) in absolute ethanol at 4°C. Acrylamide (electrophoresis pure grade) was from Bio-Rad Laboratories (Richmond, CA); N,N’-methylene bis-acrylamide and TEMED (tetramethyl ethylene diamide) were from Eastman Kodak Co. (Rochester, NY). &Mercaptoethanol, Tris-HCl, Trizma base and Nonidet P-40 (octyl phenoxypolyethoxy ethanol) were from Sigma Chemical Co. (St. Louis, MO); sodium dodecyl sulphate (SDS) of specially pure grade from BDH Chemicals Ltd. (Poole, U.K.) Agarose C, pharmalytes pH 3-10 and pharmalytes pH 5-8 were obtained from Pharmacia Fine Chemicals
Glucocorticoid-induced
proteins in endothelial ceils
241
(Uppsala, Sweden). All other laboratory chemicals were from Ajax Chemicals (Sydney, Australia). The following tissue culture media and media components were obtained from Flow Laboratories (Stanmore, N.S.W., Australia): Dulbecco modified Eagle’s medium (DMEM), foetal calf serum, non-essential amino acids, sodium bicarbonate solution (5.6%), glutamine (200 mM), penicillin (5000 U/ml) and streptomycin (5000 pg/ml), trypsin powder, Dulbecco’s phosphate buffer and Hepes buffer (1 M). Methionine-free DMEM was specially prepared by Commonweal~ Serum Laboratories (Melbourne, Australia). Bovine endothelial cell cultures Endothelial cells were cultured from fresh bovine aortae by collagenase dispersion (Booyse et al., 1975) and cultured in DMEM supplemented with non-essential amino acids, 0.37% sodium bicarbonate, 30% foetal calf serum, 2 mM glutamine, penicillin (100 U/ml) and streptomycin (100 pg/ml). Cells were cultured at 37°C in a humidified atmosphere of 95% air and 5% CO,. Three cell lines from different aortae were subcultured by treatment with 0.025% trypsin/0.02% EDTA in phosphate-buffered saline and used for expe~mentation at low (< 7) passage number when the cells still formed a confluent monolayer of small, flattened, cuboidal cells (Gospodarowicz and Tauber, 1980). Endothelial cell cultures produced angiotensin-converting enzyme (Mendelsohn et al., 1982), did not stain with antibodies specific for smooth muscle tropomyosin (Charnley-Campbell et al., 1977) and had similar between-line 2-D gel protein synthetic patterns, Steroid treatment and [35S]methionine incorporation Endothelial cells were grown to confluence in multi-well 35 mm tissue culture dishes. Cells were washed twice with serum-free DMEM, then incubated for 24 h at 37°C in 2 ml of the same medium containing either ethanol (0.01%) as a control, or lo-’ M of the following steroids: DM, RU 26988, 9cw-fF, corticosterone (B), aldosterone (Aldo), deoxycorticosterone (DOC), progesterone or SC+DHT. For [ 35S]methionine incorporation, methionine-free DMEM was supplemented with 0.1% sodium bicarbonate solution, 20 mM Hepes buffer and 2 mM glutamine. Cells were washed with methionine-free DMEM and incubated for 2 h at 37°C with 200 &X/ml [35S]methionine in 1 ml of the above medium containing ethanol (0.01%) or lo-’ M steroids. At the end of the pulse period the cells were chased with DMEM containing unlabelled methionine, and washed twice with Dulbecco’s phosphate buffer. Cells were then lysed in 150 ~1 buffer containing 9.5 M urea, 2% Nonidet P-40, 5% /&mercapto-
248
N.R. Nichols et al.
ethanol, 1.6% pharmalytes pH 5-8 and 0.4% pharmalytes pH 3-10. Aliquots (10 ~1) were analysed by liquid scintillation counting to determine the total incorporation of [35S]methionine into bovine endothelial cells (- 106cpm/5 X lo4 cells). Samples were frozen and stored at - 20°C routinely for 2 weeks, and on occasion up to 2 months. Two-dimensional gel electrophoresis For analysis of cellular proteins, samples labelled with [ 35S]methionine were subjected to two-dimensional gel electrophoresis by the method of O’Farrell(1975) except that gels were dried for 24 h between two layers of cellulose acetate stretched over a glass plate and clamped at the sides. The dried gels were exposed to Ilford (Ilflex 90) or Agfa-Gevaert (Mamoray RP3) X-ray films for l-2 weeks; films were processed using an automatic Kodak X-ray film developer. The pH gradient of the isoelectric focusing gels was measured by cutting 5 mm sections which were then shaken vigorously with 2 ml of 0.01 M KC1 for 30 min, and the pH measured using a Metrohm Herisan pH-meter type E512 (Switzerland). The SDS-electrophoresis gels were calibrated using standard [ “C]proteins of known molecular weights: lysozyme (14 300), carbonic anhydrase (30000), ovalbumin (46000), bovine serum albumin (69000) and phosphorylase b (92 500). Twelve gels were run concurrently, with each treatment represented by two separate sample preparations. Protein patterns of control and steroid-treated cells were compared visually using an X-ray viewing box by overlaying the gel autoradiographs. Spot intensities were compared on gels exposed to film for the same period of time, and differences within individual runs and between runs of different samples were recorded. On this basis, consistent variations between the protein patterns of control and steroid-treated cells are described.
RESULTS Effect of steroid on protein synthetic patterns analysed by 2-D gel electrophoresis Bovine endothelial cells were treated with 0.01% ethanol (control) or Fig. 1. Glucocorticoid-responsive proteins of cultured bovine endothelial cells. Contact prints of 2-D gel (12.5% acrylamide) autoradiographs of control and DM-treated bovine endothelial cells are shown. Consistent changes in protein patterns are shown by circles. The arrow points to the reference protein actin (M.W. - 43K, pI - 5.4). Direction of equilibrium isoelectric focusing (IEF) is from basic to acidic pH. Proteins are separated by electrophoresis in the second dimension according to their subunit molecular weight (M.W.).
Glucocorticoid- induced proteins in endothelial ceils
249
250
N. R. Nichols et al.
RUN
I
Fig. 2. Specificity of corticosteroid response in cultured bovine endothelial cells. Cells were treated with lo-’ M DM, 9&F, B, DOC and progesterone in run 1 and with lo-’ M DM, RU 26988, Aldo and Sa-DHT in run 2, and the patterns of [35S]methionine-labelled proteins compared with controls. One of duplicate gels is shown and circles enclose the responsive proteins. Actin (M.W. - 43K, pI - 5.4) is shown by an arrow for reference. el shows increased intensity with glucocorticoids and e2 shows increased intensity with all corticosteroids. Contact prints of autoradiographs were cut, then reproduced photographically.
Giucocorticoid - induced proteins in endothelial cells
251
lo-’ M glucocorticoids (DM, RU 26988, 9a-fF, B), mineralocorticoids (Aldo, DOC, 9a-fF) or, as additional controls, androgen (Sa-DHT) or progesterone. Protein profiles were altered by glucocorticoids, DOC and Aldo but not by Sa-DHT or progesterone. The consistent changes in newly synthesized proteins between runs are illustrated in Fig. 1, showing the changes in response to DM. A comparison of control and DM autoradiographs shows two proteins, el (M.W. - 43K, p1 - 6.3) and e2 (M.W. - 28K, p1 - 5.9) with increased intensities. These induced changes were consistent for 8 DM-treated samples in 4 separate runs. Glucocorticoid specificity of corticosteroid responses in bovine endothelial cells
The specificity of the dexamethasone effects on protein synthetic patterns shown in Fig. 1 was examined by 2-D gel electrophoresis of bovine endothelial cells treated with lo-’ M classical glucocorticoids (DM, 9a-fF, B), classical mineralocorticoids (Aldo, DOC, Ba-fF), sex steroids (5a-DHT, progesterone) and the ‘pure’ glucocorticoid, RU 26988, with high affinity for classical glucocorticoid receptors but negligible affinity for classical mineralocorticoid receptors (Teutsch et al., 1981). Cells were incubated in serum-free media to eliminate the effect of serum binding proteins on free steroid concentration. The steroid specificity data of two gel runs are shown in Fig. 2. In both runs, protein el is increased in intensity when cells were treated with lo-’ M glucocorticoids (DM, RU, 9a-fF, B). However, mineralocorticoid (Aldo, DOC) and sex steroids (Sa-DHT, progesterone) did not alter the rate of synthesis of protein el. Protein e2 exhibited a different specificity; it was increased in intensity by both glucocorticoids and mineralocorticoids, with the following hierarchy: DM, 9a-fF, RU > B > Aldo, DOC > Sa-DHT, Prog = control. Induction of proteins el and e2 thus comprises a glucocorticoid-specific response and not a mineralocorticoid response, based on two findings: the classical mineralocorticoids (Aldo, DOC) are less potent than DM, 9a-fF and B (e2) or show no effect on the protein synthetic rate (el), and the ‘pure’ glucocorticoid, Ru 26988, is a full agonist in terms of protein response.
DISCUSSION In the present studies a consistent domain of two glucocorticoid-induced proteins el and e2 (M.W. - 43K, p1 - 6.3 and M.W. - 28K, p1 - 5.9, respectively) has been demonstrated in confluent bovine aortic
252
N. R. Nichols et al.
endothelial cultures. The glucocorticoid specificity of the response was determined by comparing the intensities of el and e2 on 2-D gel autoradiographs from cells treated with a range of steroids. For both polypeptides, the potent synthetic glucocorticoids exhibited the greatest effect on spot intensities whereas Scu-DHT and progesterone have no effect. These data are consistent with the existence of type II [3H]DM binding classical glucocorticoid receptors in these cells ( Kd 4°C = 1.2 nM, capacity = 64 fmoles/mg protein) and our inability to find any specific binding of [3H]R1881, a potent synthetic androgen (Nichols and Funder, unpublished data). In terms of the regulation of protein e2, Aldo and DOC acted as agonists, albeit less potent ones, which presumably reflects their ability to occupy glucocorticoid receptors (Samuels and Tomkins, 1970; Rousseau and Schmit, 1977) and is similar to our findings for the action of these steroids in rat vascular smooth muscle cells (Nichols et al., 1983). In contrast, no effect of these steroids was seen on protein el. This may reflect an effect on spot intensity below the threshold of discrimination, since occupancy of glucocorticoid receptors by aldosterone or DOC has been shown, in other instances, to be followed by partial agonist/partial antagonist effects (Rousseau and Schmit, 1977). Another possibility is that aldosterone and DOC may be agonist for one facet of the glucocorticoid response and antagonist for another. Finally, the inability of progesterone to affect the rate of synthesis of either polypeptide is consistent with its reported full antagonist actions (Samuels and Ton&ins, 1970) by preventing the nuclear translocation of glucocorticoid receptors. Glucocorticoids have been shown to increase the surface area and rate of protein synthesis of human umbilical vein endothelial cells (Maca et al., 1978). Previously, Longnecker et al. (1981) reported that glucocorticoids induced two polypeptides (M.W. - 150K, p1 - 4.0) and increased the secretion of basement membrane proteins in a cloned line of bovine endothelial cells. On our 10 and 12.5% acrylamide gels we were not able to discern a responsive protein at 150K. However, angiotensinconverting enzyme, which has a.M.W. of 129-140K on SDS gels and a p1 of 4.5-4.8 (Soffer, 1976), is induced 6-7-fold by lo-’ M DM in bovine endothelial cell cultures over 2 days (Mendelsohn et al., 1982). A 41-43K polypeptide of similar isoelectric point to protein el has been identified as a glucocorticoid-specific induced protein in human skin fibroblasts (Khalid et al., 1983) and in rat thymus (Clements et al., 1982), vascular smooth muscle cells (Nichols et al., 1983) and cardiac cells (Nichols and Funder, 1982). However, the magnitude of the response to glucocorticoids varies, being greatest in the fibroblasts and least in the endothelial cells. This variation may reflect the different
Glucocorticoid- induced proteins in endothelial cells
253
culture conditions employed or simply be dependent on cell-specific factors. In the rat pituitary GH, tumour cells in vitro, apparently different effects of ~uc~rticoids are seen, depending on the metabolic state and differentiation of the cells (Johnson et al., 1979). Glucocorticoid-induced 40K (lipomodulin) and 30K species of phospholipase inhibitory proteins have been identified in rat macrophages and neutrophils and may be responsible for the anti-inflammatory actions of glucocorticoids (Hirata et al., 1982). A basement membrane protein, CSP-60, has a subunit M.W. of 30K, and appears as a new cell surface protein in confluent, highly contact-inhibited endothelial cell cultures (Vlodavsky et al., 1979). The functional identity of the 43K and 28K glucocorticoid-induced proteins in bovine endothelial cells and their relationship to the proteins described above is unknown.
ACKNOWLEDGEMENTS We thank Jenny Gannell and Chris Olsson (technical), Sue Smith (secretarial) and Debbie Frey (photographic) for assistance. This work was supported by the National Health and Medical Research Council of Australia.
REFERENCES Booyse, F.M., Sedlak, B.J. and Rafelson, M.E. (1975) Thrombosis and Haemostasis 34, 825-839. Bnmner, H.R. and Gavras, H. (1977) In: Hypertension: Mechanisms, Diagnosis and Management, Eds.: J.O. Davis, J.H. Laragh and A. Selwyn (HP Publishing Co., New York) pp. 169-180. Charnley-Campbell, J., Campbell, G.R., Groschel-Stewart, U. and Bumstock, G. (1977) Cell Tissue Res. 183, 153- 166. Clements, J.A., Ftmder, J.W., Khalid, B.A.K., Krozowski, S. and Lim, A.T. (1982) In: Endocrinology of Hypertension, Eds,: F. Mantero, E.G. Biglieri and C.R.W. Edwards (Academic Press, London) pp. 1- 10. Cope, C.L. (1972) In: Adrenal Steroids and Disease (J.B. Lippincott, Philadelphia) pp. 463-483. Gospodarowicz, D. and Tauber, J. (1980) Endocr. Rev. 1, 201-227. Hirata, F., Notsu, Y., Iwata, M., Parente, L., DiRosa, M. and Flower, R.J. (1982) B&hem. Biophys. Res. Commun. 109,223-230. Johnson, L.J., Lan, N.C. and Banter, J.D. (1979) J. Biol. Chem. 254,7785-7794. Kbalid, B.A.K., Gyorki, S., Wame, G.L. and Funder, J.W. (1983) Clin. Endocrinol. 18, 407-416. Longnecker, J.P., Leitman, D.E., Kilty, L.A. and Johnson, L.K. (1981) Proc. Endocr. Sot. (Cincinnati) 63, 12 (abstract 185).
254
N. R. Nichols et al.
Maca, R.D., Fry, G.L. and Hoak, J.C. (1978) Br. J. Haematol. 38, 501-509. Mendelsohn, F.A.O., Lloyd, C.J., Kachel, C. and Funder, J.W. (1982) J. Clin. Invest. 70, 684-692. Nichols, N.R. and Funder, J.W. (1982) Proc. 12th Int. Congr. B&hem., Perth, 276 (abstract). Nichols, N., Olsson, C. and Funder, J. (1983) Endocrinology (accepted for publication). O’Farrell, P.M. (1975) J. Biol. Chem. 250, 4007-4021. Ross, R. (198 1) Atherosclerosis 1, 293-3 11. Rousseau, G.G. and S&nit, J. (1977) J. Steroid B&hem. 8, 911-919. Samuels, H.H. and Ton&ins, G.M. (1970) J. Mol. Biol. 52, 57-74. Selye, H. (1951) In: Hypertension, Ed.: E.T. Bell (Minnesota Press, Minneapolis) pp. 119-132. Soffer, R.L. (1976) Ann. Rev. B&hem. 45, 73-94. Teutsch, G., Costerousse, G., Deraedt, R., Benz& J., Fortin, M. and Philibert, D. (1981) Steroids 38, 65 l-665. Vlodavsky, I., Johnson, L.K. and Gospodarowicz, D. (1979) Proc. Nat. Acad. Sci. (U.S.A.) 76, 2306-2310.
Molecular and Cellular Endocrinology, 32 (1983) 245-254 Elsevier Scientific Publishers Ireland. Ltd.
MCE 01054
GLUCOCORTICOID-INDUCED PROTEINS IN BOVINE ENDOTHELIAL CELLS N.R. NICHOLS ‘, C.J. LLOYD 2, F.A.O. MENDELSOHN J.W. FUNDER ‘,*
2, and
’ Medical Research Centre, Prince Henry’s Hospital, Melbourne, Victoria 3004, 2 Universiiy of Melbourne, Department of Medicine, Austin Hospital, Heidelberg, (Australia) Received
11 April
1983; accepted
Victoria
4 July 1983
High circulating levels of corticosteroids are associated with elevated blood pressure and an increased incidence of vascular damage. We have used high resolution two-dimensional gel electrophoresis and autoradiography to analyse direct steroid effects on the newly synthesized proteins in bovine aortic endothelial cells. At medium concentrations of lo-’ M, corticosteroids but not sex steroids increased the synthesis of two endothelial proteins: el (M.W. - 43K, p1 - 6.3) and e2 (M.W. - 28K, pI - 5.9). The responses were consistent for different endothelial cell cultures, different gel runs, and for natural and synthetic glucocorticoids, including the ‘pure’ synthetic glucocorticoid, RU 26988. Mineralocorticoids (aldosterone, deoxycorticosterone) at lo-’ M were less potent in their effect on the synthesis of e2, and had no discernible effect on the synthesis of el. These specific effects on cellular protein synthesis demonstrate that glucocorticoids can act directly on endothelial cells. These direct actions may thus mediate altered endothelial cell function in clinical and experimental cardiovascular conditions characterized by corticosteroid excess. Keywords:
R U 26988; endothelial protein synthesis; steroid domain; two-dimensional electrophoresis.
Factors influencing endothelial cell function may play a role in the development and maintenance of cardiovascular disease. In particular, high circulating levels of adrenal steroids have long been associated with an increased incidence of vascular damage and resulting atherosclerosis (Selye, 195 1; Cope, 1972; Brunner and Gavras, 1977). Endothelial cell injury by chemical as well as mechanical means is thought to be an initiating factor in the development of arterial disease (Ross, 1981). The molecular mechanisms underlying such injury, however, and the role of various agents, including steroids, are not well understood. * To whom correspondence 0303-7207/83/$03.00
should
be addressed.
0 1983 Elsevier Scientific
Publishers
Ireland,
Ltd.
246
N.R. Nichols et al.
Recently, Mendelsohn et al. (1982) have characterized the glucocorticoid induction of angiotensin-converting enzyme in cultured bovine aortic endothelial cells. These data are consistent with a direct effect of glucocorticoid-specific hormones on endothelial cells via receptor-initiated, DNA-dependent, mRNA-mediated protein synthesis. Changes in levels of vasoactive peptides by alterations in angiotensin-converting enzyme activity have been suggested to be responsible for endothelial apertures, increased vascular permeability and resultant vascular damage (Brunner and Gavras, 1977). In the present study we have used the techniques of high resolution two-dimensional (2-D) gel electrophoresis and autoradiography to examine the protein synthetic patterns of bovine endothelial cell cultures for additional steroid-responsive products which might alter endothelial cell function. Cells were cultured with a series of naturally occurring and synthetic steroids, representing the range of secretion products of the adrenal cortex. After 24 h incubation with glucocorticoid, mineralocorticoid, androgen or progesterone the effect of these various steroids on the protein synthetic profile was compared by autoradiography.
MATERIALS AND METHODS Chemicals and reagents [ 35S]Methionine in 0.02 M potassium acetate containing 0.1% /3mercaptoethanol (1200 Ci/rnmole) and [ “C]proteins of known molecu-
lar weight were obtained from the Radiochemical Centre (Amersham, U.K.). RU 26988 (1 1/3,17&dihydroxy- 17a-propynyl-androsta- 1,4,6-triene-3-one) was the gift of Roussel-UCLAF (Romainville, France), nonradioactive dexamethasone (DM) of Merck, Sharp and Dohme (Sydney, Australia) and 9cu-fluorocortisol (9cu-fF) of E.R. Squibb (Sydney, Australia); 5cu-dihydrotestosterone (Sa-DHT) was obtained from Ikapharm (Ramat-Gan, Israel), and other steroids from Steraloids (Wilton, N.H.). Steroids were stored as stock solutions (4 x 10e3 M) in absolute ethanol at 4°C. Acrylamide (electrophoresis pure grade) was from Bio-Rad Laboratories (Richmond, CA); N,N’-methylene bis-acrylamide and TEMED (tetramethyl ethylene diamide) were from Eastman Kodak Co. (Rochester, NY). &Mercaptoethanol, Tris-HCl, Trizma base and Nonidet P-40 (octyl phenoxypolyethoxy ethanol) were from Sigma Chemical Co. (St. Louis, MO); sodium dodecyl sulphate (SDS) of specially pure grade from BDH Chemicals Ltd. (Poole, U.K.) Agarose C, pharmalytes pH 3-10 and pharmalytes pH 5-8 were obtained from Pharmacia Fine Chemicals
Glucocorticoid-induced
proteins in endothelial ceils
241
(Uppsala, Sweden). All other laboratory chemicals were from Ajax Chemicals (Sydney, Australia). The following tissue culture media and media components were obtained from Flow Laboratories (Stanmore, N.S.W., Australia): Dulbecco modified Eagle’s medium (DMEM), foetal calf serum, non-essential amino acids, sodium bicarbonate solution (5.6%), glutamine (200 mM), penicillin (5000 U/ml) and streptomycin (5000 pg/ml), trypsin powder, Dulbecco’s phosphate buffer and Hepes buffer (1 M). Methionine-free DMEM was specially prepared by Commonweal~ Serum Laboratories (Melbourne, Australia). Bovine endothelial cell cultures Endothelial cells were cultured from fresh bovine aortae by collagenase dispersion (Booyse et al., 1975) and cultured in DMEM supplemented with non-essential amino acids, 0.37% sodium bicarbonate, 30% foetal calf serum, 2 mM glutamine, penicillin (100 U/ml) and streptomycin (100 pg/ml). Cells were cultured at 37°C in a humidified atmosphere of 95% air and 5% CO,. Three cell lines from different aortae were subcultured by treatment with 0.025% trypsin/0.02% EDTA in phosphate-buffered saline and used for expe~mentation at low (< 7) passage number when the cells still formed a confluent monolayer of small, flattened, cuboidal cells (Gospodarowicz and Tauber, 1980). Endothelial cell cultures produced angiotensin-converting enzyme (Mendelsohn et al., 1982), did not stain with antibodies specific for smooth muscle tropomyosin (Charnley-Campbell et al., 1977) and had similar between-line 2-D gel protein synthetic patterns, Steroid treatment and [35S]methionine incorporation Endothelial cells were grown to confluence in multi-well 35 mm tissue culture dishes. Cells were washed twice with serum-free DMEM, then incubated for 24 h at 37°C in 2 ml of the same medium containing either ethanol (0.01%) as a control, or lo-’ M of the following steroids: DM, RU 26988, 9cw-fF, corticosterone (B), aldosterone (Aldo), deoxycorticosterone (DOC), progesterone or SC+DHT. For [ 35S]methionine incorporation, methionine-free DMEM was supplemented with 0.1% sodium bicarbonate solution, 20 mM Hepes buffer and 2 mM glutamine. Cells were washed with methionine-free DMEM and incubated for 2 h at 37°C with 200 &X/ml [35S]methionine in 1 ml of the above medium containing ethanol (0.01%) or lo-’ M steroids. At the end of the pulse period the cells were chased with DMEM containing unlabelled methionine, and washed twice with Dulbecco’s phosphate buffer. Cells were then lysed in 150 ~1 buffer containing 9.5 M urea, 2% Nonidet P-40, 5% /&mercapto-
248
N.R. Nichols et al.
ethanol, 1.6% pharmalytes pH 5-8 and 0.4% pharmalytes pH 3-10. Aliquots (10 ~1) were analysed by liquid scintillation counting to determine the total incorporation of [35S]methionine into bovine endothelial cells (- 106cpm/5 X lo4 cells). Samples were frozen and stored at - 20°C routinely for 2 weeks, and on occasion up to 2 months. Two-dimensional gel electrophoresis For analysis of cellular proteins, samples labelled with [ 35S]methionine were subjected to two-dimensional gel electrophoresis by the method of O’Farrell(1975) except that gels were dried for 24 h between two layers of cellulose acetate stretched over a glass plate and clamped at the sides. The dried gels were exposed to Ilford (Ilflex 90) or Agfa-Gevaert (Mamoray RP3) X-ray films for l-2 weeks; films were processed using an automatic Kodak X-ray film developer. The pH gradient of the isoelectric focusing gels was measured by cutting 5 mm sections which were then shaken vigorously with 2 ml of 0.01 M KC1 for 30 min, and the pH measured using a Metrohm Herisan pH-meter type E512 (Switzerland). The SDS-electrophoresis gels were calibrated using standard [ “C]proteins of known molecular weights: lysozyme (14 300), carbonic anhydrase (30000), ovalbumin (46000), bovine serum albumin (69000) and phosphorylase b (92 500). Twelve gels were run concurrently, with each treatment represented by two separate sample preparations. Protein patterns of control and steroid-treated cells were compared visually using an X-ray viewing box by overlaying the gel autoradiographs. Spot intensities were compared on gels exposed to film for the same period of time, and differences within individual runs and between runs of different samples were recorded. On this basis, consistent variations between the protein patterns of control and steroid-treated cells are described.
RESULTS Effect of steroid on protein synthetic patterns analysed by 2-D gel electrophoresis Bovine endothelial cells were treated with 0.01% ethanol (control) or Fig. 1. Glucocorticoid-responsive proteins of cultured bovine endothelial cells. Contact prints of 2-D gel (12.5% acrylamide) autoradiographs of control and DM-treated bovine endothelial cells are shown. Consistent changes in protein patterns are shown by circles. The arrow points to the reference protein actin (M.W. - 43K, pI - 5.4). Direction of equilibrium isoelectric focusing (IEF) is from basic to acidic pH. Proteins are separated by electrophoresis in the second dimension according to their subunit molecular weight (M.W.).
Glucocorticoid- induced proteins in endothelial ceils
249
250
N. R. Nichols et al.
RUN
I
Fig. 2. Specificity of corticosteroid response in cultured bovine endothelial cells. Cells were treated with lo-’ M DM, 9&F, B, DOC and progesterone in run 1 and with lo-’ M DM, RU 26988, Aldo and Sa-DHT in run 2, and the patterns of [35S]methionine-labelled proteins compared with controls. One of duplicate gels is shown and circles enclose the responsive proteins. Actin (M.W. - 43K, pI - 5.4) is shown by an arrow for reference. el shows increased intensity with glucocorticoids and e2 shows increased intensity with all corticosteroids. Contact prints of autoradiographs were cut, then reproduced photographically.
Giucocorticoid - induced proteins in endothelial cells
251
lo-’ M glucocorticoids (DM, RU 26988, 9a-fF, B), mineralocorticoids (Aldo, DOC, 9a-fF) or, as additional controls, androgen (Sa-DHT) or progesterone. Protein profiles were altered by glucocorticoids, DOC and Aldo but not by Sa-DHT or progesterone. The consistent changes in newly synthesized proteins between runs are illustrated in Fig. 1, showing the changes in response to DM. A comparison of control and DM autoradiographs shows two proteins, el (M.W. - 43K, p1 - 6.3) and e2 (M.W. - 28K, p1 - 5.9) with increased intensities. These induced changes were consistent for 8 DM-treated samples in 4 separate runs. Glucocorticoid specificity of corticosteroid responses in bovine endothelial cells
The specificity of the dexamethasone effects on protein synthetic patterns shown in Fig. 1 was examined by 2-D gel electrophoresis of bovine endothelial cells treated with lo-’ M classical glucocorticoids (DM, 9a-fF, B), classical mineralocorticoids (Aldo, DOC, Ba-fF), sex steroids (5a-DHT, progesterone) and the ‘pure’ glucocorticoid, RU 26988, with high affinity for classical glucocorticoid receptors but negligible affinity for classical mineralocorticoid receptors (Teutsch et al., 1981). Cells were incubated in serum-free media to eliminate the effect of serum binding proteins on free steroid concentration. The steroid specificity data of two gel runs are shown in Fig. 2. In both runs, protein el is increased in intensity when cells were treated with lo-’ M glucocorticoids (DM, RU, 9a-fF, B). However, mineralocorticoid (Aldo, DOC) and sex steroids (Sa-DHT, progesterone) did not alter the rate of synthesis of protein el. Protein e2 exhibited a different specificity; it was increased in intensity by both glucocorticoids and mineralocorticoids, with the following hierarchy: DM, 9a-fF, RU > B > Aldo, DOC > Sa-DHT, Prog = control. Induction of proteins el and e2 thus comprises a glucocorticoid-specific response and not a mineralocorticoid response, based on two findings: the classical mineralocorticoids (Aldo, DOC) are less potent than DM, 9a-fF and B (e2) or show no effect on the protein synthetic rate (el), and the ‘pure’ glucocorticoid, Ru 26988, is a full agonist in terms of protein response.
DISCUSSION In the present studies a consistent domain of two glucocorticoid-induced proteins el and e2 (M.W. - 43K, p1 - 6.3 and M.W. - 28K, p1 - 5.9, respectively) has been demonstrated in confluent bovine aortic
252
N. R. Nichols et al.
endothelial cultures. The glucocorticoid specificity of the response was determined by comparing the intensities of el and e2 on 2-D gel autoradiographs from cells treated with a range of steroids. For both polypeptides, the potent synthetic glucocorticoids exhibited the greatest effect on spot intensities whereas Scu-DHT and progesterone have no effect. These data are consistent with the existence of type II [3H]DM binding classical glucocorticoid receptors in these cells ( Kd 4°C = 1.2 nM, capacity = 64 fmoles/mg protein) and our inability to find any specific binding of [3H]R1881, a potent synthetic androgen (Nichols and Funder, unpublished data). In terms of the regulation of protein e2, Aldo and DOC acted as agonists, albeit less potent ones, which presumably reflects their ability to occupy glucocorticoid receptors (Samuels and Tomkins, 1970; Rousseau and Schmit, 1977) and is similar to our findings for the action of these steroids in rat vascular smooth muscle cells (Nichols et al., 1983). In contrast, no effect of these steroids was seen on protein el. This may reflect an effect on spot intensity below the threshold of discrimination, since occupancy of glucocorticoid receptors by aldosterone or DOC has been shown, in other instances, to be followed by partial agonist/partial antagonist effects (Rousseau and Schmit, 1977). Another possibility is that aldosterone and DOC may be agonist for one facet of the glucocorticoid response and antagonist for another. Finally, the inability of progesterone to affect the rate of synthesis of either polypeptide is consistent with its reported full antagonist actions (Samuels and Ton&ins, 1970) by preventing the nuclear translocation of glucocorticoid receptors. Glucocorticoids have been shown to increase the surface area and rate of protein synthesis of human umbilical vein endothelial cells (Maca et al., 1978). Previously, Longnecker et al. (1981) reported that glucocorticoids induced two polypeptides (M.W. - 150K, p1 - 4.0) and increased the secretion of basement membrane proteins in a cloned line of bovine endothelial cells. On our 10 and 12.5% acrylamide gels we were not able to discern a responsive protein at 150K. However, angiotensinconverting enzyme, which has a.M.W. of 129-140K on SDS gels and a p1 of 4.5-4.8 (Soffer, 1976), is induced 6-7-fold by lo-’ M DM in bovine endothelial cell cultures over 2 days (Mendelsohn et al., 1982). A 41-43K polypeptide of similar isoelectric point to protein el has been identified as a glucocorticoid-specific induced protein in human skin fibroblasts (Khalid et al., 1983) and in rat thymus (Clements et al., 1982), vascular smooth muscle cells (Nichols et al., 1983) and cardiac cells (Nichols and Funder, 1982). However, the magnitude of the response to glucocorticoids varies, being greatest in the fibroblasts and least in the endothelial cells. This variation may reflect the different
Glucocorticoid- induced proteins in endothelial cells
253
culture conditions employed or simply be dependent on cell-specific factors. In the rat pituitary GH, tumour cells in vitro, apparently different effects of ~uc~rticoids are seen, depending on the metabolic state and differentiation of the cells (Johnson et al., 1979). Glucocorticoid-induced 40K (lipomodulin) and 30K species of phospholipase inhibitory proteins have been identified in rat macrophages and neutrophils and may be responsible for the anti-inflammatory actions of glucocorticoids (Hirata et al., 1982). A basement membrane protein, CSP-60, has a subunit M.W. of 30K, and appears as a new cell surface protein in confluent, highly contact-inhibited endothelial cell cultures (Vlodavsky et al., 1979). The functional identity of the 43K and 28K glucocorticoid-induced proteins in bovine endothelial cells and their relationship to the proteins described above is unknown.
ACKNOWLEDGEMENTS We thank Jenny Gannell and Chris Olsson (technical), Sue Smith (secretarial) and Debbie Frey (photographic) for assistance. This work was supported by the National Health and Medical Research Council of Australia.
REFERENCES Booyse, F.M., Sedlak, B.J. and Rafelson, M.E. (1975) Thrombosis and Haemostasis 34, 825-839. Bnmner, H.R. and Gavras, H. (1977) In: Hypertension: Mechanisms, Diagnosis and Management, Eds.: J.O. Davis, J.H. Laragh and A. Selwyn (HP Publishing Co., New York) pp. 169-180. Charnley-Campbell, J., Campbell, G.R., Groschel-Stewart, U. and Bumstock, G. (1977) Cell Tissue Res. 183, 153- 166. Clements, J.A., Ftmder, J.W., Khalid, B.A.K., Krozowski, S. and Lim, A.T. (1982) In: Endocrinology of Hypertension, Eds,: F. Mantero, E.G. Biglieri and C.R.W. Edwards (Academic Press, London) pp. 1- 10. Cope, C.L. (1972) In: Adrenal Steroids and Disease (J.B. Lippincott, Philadelphia) pp. 463-483. Gospodarowicz, D. and Tauber, J. (1980) Endocr. Rev. 1, 201-227. Hirata, F., Notsu, Y., Iwata, M., Parente, L., DiRosa, M. and Flower, R.J. (1982) B&hem. Biophys. Res. Commun. 109,223-230. Johnson, L.J., Lan, N.C. and Banter, J.D. (1979) J. Biol. Chem. 254,7785-7794. Kbalid, B.A.K., Gyorki, S., Wame, G.L. and Funder, J.W. (1983) Clin. Endocrinol. 18, 407-416. Longnecker, J.P., Leitman, D.E., Kilty, L.A. and Johnson, L.K. (1981) Proc. Endocr. Sot. (Cincinnati) 63, 12 (abstract 185).
254
N. R. Nichols et al.
Maca, R.D., Fry, G.L. and Hoak, J.C. (1978) Br. J. Haematol. 38, 501-509. Mendelsohn, F.A.O., Lloyd, C.J., Kachel, C. and Funder, J.W. (1982) J. Clin. Invest. 70, 684-692. Nichols, N.R. and Funder, J.W. (1982) Proc. 12th Int. Congr. B&hem., Perth, 276 (abstract). Nichols, N., Olsson, C. and Funder, J. (1983) Endocrinology (accepted for publication). O’Farrell, P.M. (1975) J. Biol. Chem. 250, 4007-4021. Ross, R. (198 1) Atherosclerosis 1, 293-3 11. Rousseau, G.G. and S&nit, J. (1977) J. Steroid B&hem. 8, 911-919. Samuels, H.H. and Ton&ins, G.M. (1970) J. Mol. Biol. 52, 57-74. Selye, H. (1951) In: Hypertension, Ed.: E.T. Bell (Minnesota Press, Minneapolis) pp. 119-132. Soffer, R.L. (1976) Ann. Rev. B&hem. 45, 73-94. Teutsch, G., Costerousse, G., Deraedt, R., Benz& J., Fortin, M. and Philibert, D. (1981) Steroids 38, 65 l-665. Vlodavsky, I., Johnson, L.K. and Gospodarowicz, D. (1979) Proc. Nat. Acad. Sci. (U.S.A.) 76, 2306-2310.