Membrane permeability to K and the control of aldosterone synthesis: effects of valinomycin and cromakalim in bovine adrenocortical cells R. M.
Shepherd,
R. Fraser and C.
J. Kenyon
MRC Blood Pressure Unit, Western Infirmary, Glasgow Gil 6NT, U.K. (Requests for offprints should be addressed to C. J. Kenyon) received
24 March 1992
ABSTRACT
Stimulation of aldosterone synthesis by angiotensin II (All) is associated with depolarization of the cell membrane. Since the potential difference of adrenocortical cells is dependent on membrane permeability to potassium ions, the effects of agents which hyperpolarize the cell (by increasing permeability to K+) on the control of aldosterone synthesis were investigated further. Basal and All-stimulated aldosterone synthesis was increased by 20\p=n-\70% in cells incubated with 1 or 10 nM of the potassium ionophore valinomycin; higher concentrations markedly inhibited All-stimulated synthesis. Cromakalim, a potential antihypertensive drug which facilitates the opening of K+ channels in smooth muscle cells, stimulated basal aldosterone synthesis at 2 \g=m\M but had no effect at 40 \g=m\M. All-stimulated aldosterone synthesis was not affected by cromakalim except at 40 \g=m\M, which was inhibitory. The inhibitory effects of cromakalim, unlike those of valinomycin, were not reversible. Aldosterone synthesis from added hydroxycholesterol and pregnenolone (but not from
INTRODUCTION
deoxycorticosterone and corticosterone) was significantly inhibited by 40 \g=m\M cromakalim. Potassium efflux from cells preloaded with 43K was unaffected by low concentrations of valinomycin, but was markedly increased by concentrations which inhibited All-stimulated aldosterone production. Small decreases and increases in K efflux, caused by 1 and 40 \g=m\M cromakalim respectively, corresponded with increases and decreases in basal aldosterone production; cromakalim did not affect
43K efflux from All-stimulated cells. We suggest that increasing adrenocortical cell membrane permeability to K+ reduces steroidogenesis, but that valinomycin and cromakalim have other actions which complicate the relationship between 43K efflux and aldosterone production. Cromakalim appears to inhibit 21-hydroxylase activity in the biosynthetic pathway and may also affect
3\g=b\-hydroxysteroiddehydrogenase activity.
Journal of Molecular Endocrinology (1992) 9,
165-173
membrane potential (Natke & Kabela, 1979; Quinn al. 19876), we have argued that decreased and increased potassium efflux represent depolarization and repolarization respectively. It follows that agents which are known to increase potassium transport across the cell membrane should also affect aldoster¬ one synthesis, particularly in cells stimulated with All. Two such agents, valinomycin, an ionophore which forms a membrane soluble complex with K (see Pressman, 1976), and cromakalim (( 4- )6 et
In a previous study (Shepherd et al. 1991), we noted that potassium efflux from bovine adrenocortical cells could be separated into two components. Angiotensin II (All) stimulation of aldosterone pro¬ duction was associated with differential effects on these efflux components; All retarded the faster efflux component and quickened the slower compo¬ nent. Similar biphasic effects of All on the perme¬ ability of adrenocortical cells to K have been described by Lobo & Marusic (1988). Since the resting membrane potential of adrenocortical cells is dependent on permeability to K (Matthews, 1967; Quinn et al. 1987ö), and All has been shown to alter
cyano-3,4-dihydro-2,2-dimethyl-trans-4(2-oxo-lpyrrolidyl)-2H-benzo[b]pyran-3-ol; BRL 34915), an example of a class of vasorelaxant drugs which is thought to act by opening membrane potassium channels (reviewed by Hamilton & Weston, 1989),
have been used in the present study of the mecha¬ nism of All-stimulated aldosterone synthesis. Both agents should hyperpolarize cells by allowing K efflux. If these agents affect adrenocortical cell potas¬ sium metabolism, as has been demonstrated for other cell types, All-stimulated aldosterone synthe¬ sis should be inhibited. The effect may be partial, since the actions of All are also mediated by increases in intracellular Ca + which in turn are due to the release of stored calcium as well as uptake through voltage-gated calcium channels (Capponi et al. 1984; Braley et al. 1986). To check that any steroidogenic effects of valinomycin and cromakalim match their supposed pharmacological actions on potassium permeability, we have measured the efflux of radioactivity in cells preloaded with K. ,
MATERIALS AND METHODS
Cell isolation and culture
glomerulosa cells were isolated by colladigestion of minced slices of the outermost region of six to seven bovine adrenal glands (Kenyon et al. 1985; Shepherd et al. 1991). The collagenase digest was first filtered through a nylon gauze (100 |xm mesh; Henry Simon, Stockport, Cheshire, U.K.) to remove undigested tissue. After centrifugation at 500 £ for 10 min, the cells were resuspended in Medium 199 (modified so that the final K con¬ centration was 3-8 mM; Flow Laboratories, Irvine, Strathclyde, U.K.) and centrifuged once more. Broken cells and red blood cells were removed by separation on a discontinuous gradient of isotonic Percoll (Sigma Chemical Co., Poole, Dorset, U.K.) at 1750£ for 10 min. Cells with a density greater than 30% but less than 40% Percoll were washed by centrifugation at 500 £ for 10 min and resuspended Bovine
zona
genase
in Medium 199. Tissue for cell culture was pro¬ cessed in the same way except that after separation on Percoll, cells were centrifuged and resuspended in Ham's F12 medium (Flow Laboratories) modified according to Kramer (1988) to contain fungizone,
gentamicin, penicillin, butylated hydroxyanisole (50 ".m), a-tocopherol (1-2 um), sodium ascorbate (100 um), disodium selenium oxide (005 |ím), glutathione (03 |!m), insulin (10 ng/ml), transferrin (10|ig/ml), metyrapone (5 (ím), valine (01 nM), sodium bicarbonate (4 mM) and Hepes (25 mM). Calcium chloride was also added to give a final Ca concentration of 25 mM. After washing three times with Ham's F12, cells were seeded at a density of 20 000 cells/well in microtitre plates and maintained in culture in an atmosphere of 5% C02 at 37 °C for 5 days before measuring aldosterone synthesis.
43K efflux
measurements
Cells (2-3 x 10f)/ml) were preincubated in Medium 199 containing 1-3 MBq 43K/ml (MRC Cyclotron Unit, Hammersmith Hospital, London, U.K.) for 2 h at 37°C, then centrifuged at 500 £ for 10 min and resuspended in Medium 199 (2xl06 cells/ml). At timed intervals, 50 or 100 n.1 aliquots of this suspen¬ sion were added to a larger volume of Medium 199, containing various concentrations of All (Cam¬ bridge Research Biochemicals, Cambridge, Cambs, U.K.) with or without 20 [IM ouabain (Sigma Chemical Co.) and/or various doses of valinomycin (Sigma Chemical Co.) or cromakalim (Beecham Research, Welwyn Garden City, Herts, U.K.) and incubated at 37 °C. Samples (200 |il) of these incuba¬ tions were withdrawn and cells separated from medium by immediate centrifugation at 12 000 £ for 30 s through a layer of oil with a specific gravity of 1015 (a mixture of dibutyl phthalate and dinonyl phthalate; Sigma Chemical Co. and Fluka, Glossop, Derbyshire, U.K.). Radioactivity in the cells and medium were counted separately (Kenyon et al.
1985).
The efflux rates of K in static incubations were calculated from log values of intracellular radioactiv¬ ity expressed as percentages of the total radioactivity (intracellular plus extracellular) as measured at different times. Where indicated, standard curvestripping techniques were used to resolve two efflux components (Atkins, 1969; Shepherd et al. 1991).
Aldosterone assays Cells in suspension were incubated in sextuplícate in the presence and absence of All in Medium 199 containing various concentrations of either valino¬ mycin or cromakalim in an atmosphere of 95% 02 and 5% C02. After a 1-h incubation at 37 °C, cells were centrifuged (2000 £ for 10 min) and the super¬ natant was stored at —20 °C. Aldosterone was meas¬ ured by direct radioimmunoasay using antiserum given by Professor Dr Th. J. Benraad, Department of Internal Medicine, University of Nijmegen, The Netherlands. Aldosterone-3-(carboxymethyl)oximino-(2-[ Ijiodohistamine) was obtained from Amersham International pic, Amersham, Bucks, U.K. Aldosterone synthesis from added precursors was measured in the presence and absence of cromakalim in cultured cells. Culture medium containing anti¬ biotics (see above) was replaced (with two washes) at with Ham's F12 medium containing Ca 2-5 mM, 0-2 g bovine serum albumin/1 and 10 mM glucose but without antibiotics or antioxidants. Cells were incubated for 1 h at 37 °C in an atmosphere of
5% C02, in medium containing 25-hydroxycholesterol, pregnenolone, progesterone, 11-deoxycorticosterone or corticosterone (all at 5 )IM and all from Sigma Chemical Co.). The medium was removed and stored at —20 °C for later radioimmunoassay of aldosterone. Cells were removed from the plate by brief exposure to trypsin and then immediately counted. Aldosterone synthesis was calculated as pmol/10 cells per h after corrections for crossreaction with added steroid had been made by sub¬ tracting values obtained in parallel incubations with precursors but
no
cells.
Statistics The effects of AI I
on
K efflux
rates
and aldosterone
synthesis were compared by analysis of variance using Neuman-Keul's multiple range test where appropriate. Paired i-tests were used where indicated.
RESULTS
Effect of valinomycin and cromakalim basal and All-stimulated aldosterone
aldosterone release. Low concentrations increased aldosterone release in all conditions by 20—70% (P
Shepherd,
R. Fraser and C.
J. Kenyon
MRC Blood Pressure Unit, Western Infirmary, Glasgow Gil 6NT, U.K. (Requests for offprints should be addressed to C. J. Kenyon) received
24 March 1992
ABSTRACT
Stimulation of aldosterone synthesis by angiotensin II (All) is associated with depolarization of the cell membrane. Since the potential difference of adrenocortical cells is dependent on membrane permeability to potassium ions, the effects of agents which hyperpolarize the cell (by increasing permeability to K+) on the control of aldosterone synthesis were investigated further. Basal and All-stimulated aldosterone synthesis was increased by 20\p=n-\70% in cells incubated with 1 or 10 nM of the potassium ionophore valinomycin; higher concentrations markedly inhibited All-stimulated synthesis. Cromakalim, a potential antihypertensive drug which facilitates the opening of K+ channels in smooth muscle cells, stimulated basal aldosterone synthesis at 2 \g=m\M but had no effect at 40 \g=m\M. All-stimulated aldosterone synthesis was not affected by cromakalim except at 40 \g=m\M, which was inhibitory. The inhibitory effects of cromakalim, unlike those of valinomycin, were not reversible. Aldosterone synthesis from added hydroxycholesterol and pregnenolone (but not from
INTRODUCTION
deoxycorticosterone and corticosterone) was significantly inhibited by 40 \g=m\M cromakalim. Potassium efflux from cells preloaded with 43K was unaffected by low concentrations of valinomycin, but was markedly increased by concentrations which inhibited All-stimulated aldosterone production. Small decreases and increases in K efflux, caused by 1 and 40 \g=m\M cromakalim respectively, corresponded with increases and decreases in basal aldosterone production; cromakalim did not affect
43K efflux from All-stimulated cells. We suggest that increasing adrenocortical cell membrane permeability to K+ reduces steroidogenesis, but that valinomycin and cromakalim have other actions which complicate the relationship between 43K efflux and aldosterone production. Cromakalim appears to inhibit 21-hydroxylase activity in the biosynthetic pathway and may also affect
3\g=b\-hydroxysteroiddehydrogenase activity.
Journal of Molecular Endocrinology (1992) 9,
165-173
membrane potential (Natke & Kabela, 1979; Quinn al. 19876), we have argued that decreased and increased potassium efflux represent depolarization and repolarization respectively. It follows that agents which are known to increase potassium transport across the cell membrane should also affect aldoster¬ one synthesis, particularly in cells stimulated with All. Two such agents, valinomycin, an ionophore which forms a membrane soluble complex with K (see Pressman, 1976), and cromakalim (( 4- )6 et
In a previous study (Shepherd et al. 1991), we noted that potassium efflux from bovine adrenocortical cells could be separated into two components. Angiotensin II (All) stimulation of aldosterone pro¬ duction was associated with differential effects on these efflux components; All retarded the faster efflux component and quickened the slower compo¬ nent. Similar biphasic effects of All on the perme¬ ability of adrenocortical cells to K have been described by Lobo & Marusic (1988). Since the resting membrane potential of adrenocortical cells is dependent on permeability to K (Matthews, 1967; Quinn et al. 1987ö), and All has been shown to alter
cyano-3,4-dihydro-2,2-dimethyl-trans-4(2-oxo-lpyrrolidyl)-2H-benzo[b]pyran-3-ol; BRL 34915), an example of a class of vasorelaxant drugs which is thought to act by opening membrane potassium channels (reviewed by Hamilton & Weston, 1989),
have been used in the present study of the mecha¬ nism of All-stimulated aldosterone synthesis. Both agents should hyperpolarize cells by allowing K efflux. If these agents affect adrenocortical cell potas¬ sium metabolism, as has been demonstrated for other cell types, All-stimulated aldosterone synthe¬ sis should be inhibited. The effect may be partial, since the actions of All are also mediated by increases in intracellular Ca + which in turn are due to the release of stored calcium as well as uptake through voltage-gated calcium channels (Capponi et al. 1984; Braley et al. 1986). To check that any steroidogenic effects of valinomycin and cromakalim match their supposed pharmacological actions on potassium permeability, we have measured the efflux of radioactivity in cells preloaded with K. ,
MATERIALS AND METHODS
Cell isolation and culture
glomerulosa cells were isolated by colladigestion of minced slices of the outermost region of six to seven bovine adrenal glands (Kenyon et al. 1985; Shepherd et al. 1991). The collagenase digest was first filtered through a nylon gauze (100 |xm mesh; Henry Simon, Stockport, Cheshire, U.K.) to remove undigested tissue. After centrifugation at 500 £ for 10 min, the cells were resuspended in Medium 199 (modified so that the final K con¬ centration was 3-8 mM; Flow Laboratories, Irvine, Strathclyde, U.K.) and centrifuged once more. Broken cells and red blood cells were removed by separation on a discontinuous gradient of isotonic Percoll (Sigma Chemical Co., Poole, Dorset, U.K.) at 1750£ for 10 min. Cells with a density greater than 30% but less than 40% Percoll were washed by centrifugation at 500 £ for 10 min and resuspended Bovine
zona
genase
in Medium 199. Tissue for cell culture was pro¬ cessed in the same way except that after separation on Percoll, cells were centrifuged and resuspended in Ham's F12 medium (Flow Laboratories) modified according to Kramer (1988) to contain fungizone,
gentamicin, penicillin, butylated hydroxyanisole (50 ".m), a-tocopherol (1-2 um), sodium ascorbate (100 um), disodium selenium oxide (005 |ím), glutathione (03 |!m), insulin (10 ng/ml), transferrin (10|ig/ml), metyrapone (5 (ím), valine (01 nM), sodium bicarbonate (4 mM) and Hepes (25 mM). Calcium chloride was also added to give a final Ca concentration of 25 mM. After washing three times with Ham's F12, cells were seeded at a density of 20 000 cells/well in microtitre plates and maintained in culture in an atmosphere of 5% C02 at 37 °C for 5 days before measuring aldosterone synthesis.
43K efflux
measurements
Cells (2-3 x 10f)/ml) were preincubated in Medium 199 containing 1-3 MBq 43K/ml (MRC Cyclotron Unit, Hammersmith Hospital, London, U.K.) for 2 h at 37°C, then centrifuged at 500 £ for 10 min and resuspended in Medium 199 (2xl06 cells/ml). At timed intervals, 50 or 100 n.1 aliquots of this suspen¬ sion were added to a larger volume of Medium 199, containing various concentrations of All (Cam¬ bridge Research Biochemicals, Cambridge, Cambs, U.K.) with or without 20 [IM ouabain (Sigma Chemical Co.) and/or various doses of valinomycin (Sigma Chemical Co.) or cromakalim (Beecham Research, Welwyn Garden City, Herts, U.K.) and incubated at 37 °C. Samples (200 |il) of these incuba¬ tions were withdrawn and cells separated from medium by immediate centrifugation at 12 000 £ for 30 s through a layer of oil with a specific gravity of 1015 (a mixture of dibutyl phthalate and dinonyl phthalate; Sigma Chemical Co. and Fluka, Glossop, Derbyshire, U.K.). Radioactivity in the cells and medium were counted separately (Kenyon et al.
1985).
The efflux rates of K in static incubations were calculated from log values of intracellular radioactiv¬ ity expressed as percentages of the total radioactivity (intracellular plus extracellular) as measured at different times. Where indicated, standard curvestripping techniques were used to resolve two efflux components (Atkins, 1969; Shepherd et al. 1991).
Aldosterone assays Cells in suspension were incubated in sextuplícate in the presence and absence of All in Medium 199 containing various concentrations of either valino¬ mycin or cromakalim in an atmosphere of 95% 02 and 5% C02. After a 1-h incubation at 37 °C, cells were centrifuged (2000 £ for 10 min) and the super¬ natant was stored at —20 °C. Aldosterone was meas¬ ured by direct radioimmunoasay using antiserum given by Professor Dr Th. J. Benraad, Department of Internal Medicine, University of Nijmegen, The Netherlands. Aldosterone-3-(carboxymethyl)oximino-(2-[ Ijiodohistamine) was obtained from Amersham International pic, Amersham, Bucks, U.K. Aldosterone synthesis from added precursors was measured in the presence and absence of cromakalim in cultured cells. Culture medium containing anti¬ biotics (see above) was replaced (with two washes) at with Ham's F12 medium containing Ca 2-5 mM, 0-2 g bovine serum albumin/1 and 10 mM glucose but without antibiotics or antioxidants. Cells were incubated for 1 h at 37 °C in an atmosphere of
5% C02, in medium containing 25-hydroxycholesterol, pregnenolone, progesterone, 11-deoxycorticosterone or corticosterone (all at 5 )IM and all from Sigma Chemical Co.). The medium was removed and stored at —20 °C for later radioimmunoassay of aldosterone. Cells were removed from the plate by brief exposure to trypsin and then immediately counted. Aldosterone synthesis was calculated as pmol/10 cells per h after corrections for crossreaction with added steroid had been made by sub¬ tracting values obtained in parallel incubations with precursors but
no
cells.
Statistics The effects of AI I
on
K efflux
rates
and aldosterone
synthesis were compared by analysis of variance using Neuman-Keul's multiple range test where appropriate. Paired i-tests were used where indicated.
RESULTS
Effect of valinomycin and cromakalim basal and All-stimulated aldosterone
aldosterone release. Low concentrations increased aldosterone release in all conditions by 20—70% (P
