0013-7227/91/1281-0280$02.00/0 Endocrinology Copyright © 1991 by The Endocrine Society
Vol. 128, No. 1 Printed in U.S.A.
Autoradiographic Evidence of Nuclear Binding of Spironolactone in Rabbit Cortical Collecting Tubule* J. P. BONVALET, M. BLOT-CHABAUD, AND N. FARMAN WITH THE TECHNICAL ASSISTANCE OF F. WANSTOK INSERM U 246, SBCe, Departement de Biologie, CEN Saclay, 91191 Gif/Yvette Cedex, France
show the presence of specific nuclear labeling for both [3H] aldosterone and [3H]SC9420. Specific cytoplasmic labeling was
ABSTRACT. Previous biochemical studies indicated that the spirolactone-mineralocorticoid receptor complexes are unable to
very low for both [ 3 H]aldosterone and [ 3 H]SC9420. The nuclear
translocate into the nucleus. The present study was designed to
reinvestigate the intracellular distribution of spirolactone-binding sites, using autoradiography. For this purpose, rabbit kidney pyramids were incubated at 30 C with tritiated SC9420 or aldosterone. Thereafter, aldosterone-sensitive cortical collecting tubules were microdissected and processed for dry film autoradiography. The concentration was 2 nM for both steroids. Nonspecific labeling was determined by incubations with tritiated steroids plus a 100-fold excess of unlabeled steroids. Results
S
PIROLACTONES are a class of synthetic steroids that antagonize the effects of aldosterone (A) (13). They compete for binding of A to mineralocorticoid receptors (2-11) and inhibit competitively the Na+-retaining action of A in vivo (1, 3, 6, 9) and in isolated epithelia (5, 7). Previous studies, using biochemical methods, indicated that tritiated spirolactones specifically bind to cytosolic receptors, but no or very few spirolactone-receptor complexes were detected in the nuclei (3, 6, 7). On the basis of these observations, the spirolactone-receptor complex has been supposed to be unable to translocate into the nucleus or to have little or no affinity for the chromatin acceptor. In the present study we reexamined the question of the intracellular distribution of spirolactone-receptor complex. For this purpose, we performed an autoradiographic study of the binding of [3H] A and the tritiated spironolactone ([3H]SC9420). Experiments were performed on isolated cortical collecting tubules (CCT) of the rabbit kidney, a tubular segment highly sensitive to A (12-14, according to a method previously used to Received July 5,1990. Address all correspondence and requests for reprints to: J. P. Bonvalet, U 246 INSERM, Faculte de Medecine X. Bichat, 16 rue H. Huchard, 75018 Paris, France. *This work was supported by INSERM and a grant from the Fondation Searle pour la Recherche sur l'Hypertension Arterielle.
labeling by [3H]SC9420 was equally and almost completely displaced by a 100-fold excess of unlabeled aldosterone or SC9420 (91% and 87%, respectively). We conclude that spironolactone-receptor complexes migrate into the nucleus. The difference between these results and those of previous studies with biochemical techniques, which failed to detect specific nuclear binding of spirolactone, may be due to methodological reasons. (Endocrinology 128: 280-284, 1991)
examine the binding of a variety of steroids along the nephron (13, 15). Materials and Methods Experiments were performed on four normal female NewZealand White rabbits (1.5-2.0 kg BW). The animals were fed a standard laboratory diet and had free access to a 0.9% sodium solution to drink for 48 h before experiments in order to lower their endogenous A secretion. They were killed by a blow behind the neck, and the left kidney was treated as previously described (13): perfusion of the kidney with an ice-cold solution containing (in millimolar concentrations) NaCl, 137; KC1, 5; MgSO4, 0.8; Na2HPO4, 0.33; KH2PO4, 0.44; MgCl2) 1; CaCl2, 1; Dglucose, 5; and Tris-HCl, 10; pH 7.4, via the renal artery, followed by perfusion of a collagenase solution [similar solution to which 0.1% collagenase (Serva, Heidelberg, Germany: 0.71 U/mg) was added] under pressure up to rupture of the renal capsule. Thin pyramid pieces were cut and incubated for 30 min with 0.1% collagenase in the presence of radioactive steroids, with or without a 100-fold excess of unlabeled competitors. Since we have previously demonstrated (16) that nuclear binding of A is not temperature dependent, all incubations were performed at 30 C. Microssection was performed at 4 C in 4 ml microdissection solution [similar to the incubation solution except for the absence of collagenase, 0.25 mM CaCl2 instead of 1 mM, and the addition of 0.1% BSA (Sigma, St. Louis, MO)] without collagenase or steroids. The duration of the microdissection period was about 1 h. CCT were dissected in their light portion. In each experiment, about 100-120 CCT were collected.
280
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 17:57 For personal use only. No other uses without permission. . All rights reserved.
NUCLEAR BINDING OF SPIRONOLACTONE Autoradiographic procedure After their dissection, CCT were immediately transferred on coverslips maintained at 4 C (two to four CCT per coverslip). Forty coverslips were treated for autoradiography in each experiment. A dry film technique was used, as previously described (13). In brief, the coverslip was transferred in the dark room, and the tubular fragments were applied on a precoated slide (AR 10 stripping film). The coverslip was removed, and the slides were kept at —20 C for 4 months. Then, the slides were fixed (ethanol-acetic acid, 9:1), the films were developed (developer Ilford LS, Mobberley Cheshire, GB), and the nephron segments were stained with methyl green pyronin. The method used for analysis of silver grain counts has been described in detail previously (13). Briefly, each tubular segment was photographed, and the counting of silver grains was performed on enlargements of the films, over nuclei and cytoplasm, using circles of known surface area superimposed over the structure considered. The background of the film was determined in the same way. For each tubular segment, counting was performed on about 20 nuclear and cytoplasmic areas, and the mean value was calculated. Total nuclear labeling was calculated by subtracting from nuclear silver grain counts the cytoplasmic silver grain counts of the same photograph, considered as background for the nuclei (13). Cytoplasmic labeling is given after subtraction of the background of the film. For each tubular segment, the specific labeling is the difference between the total labeling of the considered segment (incubation with the tritiated steroid alone) and the mean value of nonspecific labeling (tritiated steroid plus competitor) in the same experiment. Steroids The steroids used were: tritiated A ([3H]A; Amersham: 41.25 Ci/mmol), tritiated spironolactone ([3H]SC9420; Searle, Des Plaines, IL; 54.7 Ci/mmol), unlabeled A (Sigma), and unlabeled SC9420 (Searle). Five different incubation conditions were used: 1) 2 nM [3H]A, 2) 2 nM [3H]A plus a 100-fold excess of unlabeled A, 3) 2 nM [3H]SC9420, 4) 2 nM [3H]SC9420 plus a 100-fold excess of SC9420, and 5) 2 nM [3H]SC9420 plus a 100fold excess of A.
Results Figure 1 consists of photographs of the labeling on the CCT in the different conditions. In the presence of [3H] A or [3H]SC9420 alone (Fig. 1, A and B), a clear nuclear labeling was observed. It has to be noted that the labeling by [3H]A was slightly higher than that by [3H]SC9420. The nuclear labeling by [3H]A and [3H]SC9420 was almost completely displaced by unlabeled A or SC9420, respectively (Fig. 1, C and D). The nuclear labeling by [3H]SC9420 was displaced by unlabeled A (Fig. IE) in a manner equivalent to that by unlabeled SC9420 (Fig. ID). In all conditions, the cytoplasmic labeling was very low and did not differ from the background (mean values of background: A, 3.51 ± 0.43 silver grains/100 /im2 surface area; B, 3.30 ± 0.43; C, 2.14 ± 0.20; D, 2.47 ±
281
0.31; E, 3.27 ± 0.21). Figure 2 gives the mean value of nuclear labeling. To take into account the difference in the specific activities of [3H]A and [3H]SC9420, we have corrected the silver grain counts of [3H]A labeling by the ratio of specific activity between the two steroids (54.7:41.25). The total nuclear labeling was more than 10 silver grains/100 /*m2 for [3H]A. It was somewhat lower for [3H]SC9420 (6.0/ 100 Aim2). The nonspecific nuclear labeling was low (0.52.3). The displacement of [3H]SC9420 labeling over nuclei by unlabeled A and that by spironolactone were roughly equivalent (91% and 87%, respectively). Table 1 gives the values of specific nuclear and cytoplasmic labeling by [3H]A and [3H]SC9420, calculated by subtracting the nonspecific labeling (in the presence of an excess of unlabeled A or SC9420) from the total labeling. This specific nuclear labeling was 35% lower for spirolactone than for A. No significant specific cytoplasmic labeling was observed with either steroids.
Discussion Previous studies have demonstrated that spirolactones specifically bind to mineralocorticoid-sensitive tissues. A specific cytosolic binding of a tritiated spirolactone (SC26304) was found by Marver et al. (3) in rat kidney. On the other hand, Claire et al. (6) and Rossier et al. (7) reported specific binding of tritiated prorenone to rat kidney and toad urinary bladder, respectively. In these three studies, the biological activity of spirolactones, consisting of inhibition of A-induced sodium transport, was associated with specific binding. In the present study we compared the binding of tritiated spironolactone (SC9420) and tritiated A in CCT microdissected from the same rabbit kidney, using an autoradiographic technique on intact cells (13). SC9420 has been shown to have a high A antagonist activity and no agonist activity (5). The CCT has repeatedly been demonstrated to be the most sensitive nephron segment for A (12-15). A physiological concentration of A (2 nM) was used, which has been shown to result in maximal specific binding in the rabbit CCT (13). The same concentration of [3H]SC9420 was used in the present study. The results with A were very similar to those of our previous reports (13, 15): high specific nuclear labeling, with very low nonspecific nuclear labeling, and very low specific cytoplasmic labeling. Similar results were obtained in the present study with [3H]SC9420. The specific binding was 65% that of A, suggesting a slightly lower affinity of SC9420 for the receptor than A. If this were the case, one cannot be assured that the observed labeling represents the maximal specific binding of [3H] SC9420, which could be higher than that found in the present study. The labeling by [3H]SC9420 was almost
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 17:57 For personal use only. No other uses without permission. . All rights reserved.
NUCLEAR BINDING OF SPIRONOLACTONE
282
Endo • 1991 Voll28«Nol ,
FIG. 1. Photographs (x530) of the autoradiographic labeling of CCT by [3H]A and [3H]SC9420. Incubation was performed for 30 min at 30 C with [3H]A (2 nM; A), [3H]SC9420 (2 nM; B), [3H]A (2 nM) plus a 100-fold excess of unlabeled A (C), [3H]SC9420 (2 nM) plus a 100fold excess of unlabeled SC9420 (D), and [3H]SC9420 (2 nM) plus a 100-fold excess of unlabeled A (E). Silver grains are concentrated over nuclear shadows and are almost completely displaced by the
• -i
.#1
excess of unlabeled steroids. Very few
silver grains are visible over cytoplasmic areas.
•#
• / •• TABLE 1. Specific nuclear and cytoplasmic labeling by [3H]aldosterone and [3H]SC9420
3H-A alone
Specific labeling (silver grains/100 /im2 surface ares)
cT 10 H
_J UJ
CD ^>
QC
Vol. 128, No. 1 Printed in U.S.A.
Autoradiographic Evidence of Nuclear Binding of Spironolactone in Rabbit Cortical Collecting Tubule* J. P. BONVALET, M. BLOT-CHABAUD, AND N. FARMAN WITH THE TECHNICAL ASSISTANCE OF F. WANSTOK INSERM U 246, SBCe, Departement de Biologie, CEN Saclay, 91191 Gif/Yvette Cedex, France
show the presence of specific nuclear labeling for both [3H] aldosterone and [3H]SC9420. Specific cytoplasmic labeling was
ABSTRACT. Previous biochemical studies indicated that the spirolactone-mineralocorticoid receptor complexes are unable to
very low for both [ 3 H]aldosterone and [ 3 H]SC9420. The nuclear
translocate into the nucleus. The present study was designed to
reinvestigate the intracellular distribution of spirolactone-binding sites, using autoradiography. For this purpose, rabbit kidney pyramids were incubated at 30 C with tritiated SC9420 or aldosterone. Thereafter, aldosterone-sensitive cortical collecting tubules were microdissected and processed for dry film autoradiography. The concentration was 2 nM for both steroids. Nonspecific labeling was determined by incubations with tritiated steroids plus a 100-fold excess of unlabeled steroids. Results
S
PIROLACTONES are a class of synthetic steroids that antagonize the effects of aldosterone (A) (13). They compete for binding of A to mineralocorticoid receptors (2-11) and inhibit competitively the Na+-retaining action of A in vivo (1, 3, 6, 9) and in isolated epithelia (5, 7). Previous studies, using biochemical methods, indicated that tritiated spirolactones specifically bind to cytosolic receptors, but no or very few spirolactone-receptor complexes were detected in the nuclei (3, 6, 7). On the basis of these observations, the spirolactone-receptor complex has been supposed to be unable to translocate into the nucleus or to have little or no affinity for the chromatin acceptor. In the present study we reexamined the question of the intracellular distribution of spirolactone-receptor complex. For this purpose, we performed an autoradiographic study of the binding of [3H] A and the tritiated spironolactone ([3H]SC9420). Experiments were performed on isolated cortical collecting tubules (CCT) of the rabbit kidney, a tubular segment highly sensitive to A (12-14, according to a method previously used to Received July 5,1990. Address all correspondence and requests for reprints to: J. P. Bonvalet, U 246 INSERM, Faculte de Medecine X. Bichat, 16 rue H. Huchard, 75018 Paris, France. *This work was supported by INSERM and a grant from the Fondation Searle pour la Recherche sur l'Hypertension Arterielle.
labeling by [3H]SC9420 was equally and almost completely displaced by a 100-fold excess of unlabeled aldosterone or SC9420 (91% and 87%, respectively). We conclude that spironolactone-receptor complexes migrate into the nucleus. The difference between these results and those of previous studies with biochemical techniques, which failed to detect specific nuclear binding of spirolactone, may be due to methodological reasons. (Endocrinology 128: 280-284, 1991)
examine the binding of a variety of steroids along the nephron (13, 15). Materials and Methods Experiments were performed on four normal female NewZealand White rabbits (1.5-2.0 kg BW). The animals were fed a standard laboratory diet and had free access to a 0.9% sodium solution to drink for 48 h before experiments in order to lower their endogenous A secretion. They were killed by a blow behind the neck, and the left kidney was treated as previously described (13): perfusion of the kidney with an ice-cold solution containing (in millimolar concentrations) NaCl, 137; KC1, 5; MgSO4, 0.8; Na2HPO4, 0.33; KH2PO4, 0.44; MgCl2) 1; CaCl2, 1; Dglucose, 5; and Tris-HCl, 10; pH 7.4, via the renal artery, followed by perfusion of a collagenase solution [similar solution to which 0.1% collagenase (Serva, Heidelberg, Germany: 0.71 U/mg) was added] under pressure up to rupture of the renal capsule. Thin pyramid pieces were cut and incubated for 30 min with 0.1% collagenase in the presence of radioactive steroids, with or without a 100-fold excess of unlabeled competitors. Since we have previously demonstrated (16) that nuclear binding of A is not temperature dependent, all incubations were performed at 30 C. Microssection was performed at 4 C in 4 ml microdissection solution [similar to the incubation solution except for the absence of collagenase, 0.25 mM CaCl2 instead of 1 mM, and the addition of 0.1% BSA (Sigma, St. Louis, MO)] without collagenase or steroids. The duration of the microdissection period was about 1 h. CCT were dissected in their light portion. In each experiment, about 100-120 CCT were collected.
280
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 17:57 For personal use only. No other uses without permission. . All rights reserved.
NUCLEAR BINDING OF SPIRONOLACTONE Autoradiographic procedure After their dissection, CCT were immediately transferred on coverslips maintained at 4 C (two to four CCT per coverslip). Forty coverslips were treated for autoradiography in each experiment. A dry film technique was used, as previously described (13). In brief, the coverslip was transferred in the dark room, and the tubular fragments were applied on a precoated slide (AR 10 stripping film). The coverslip was removed, and the slides were kept at —20 C for 4 months. Then, the slides were fixed (ethanol-acetic acid, 9:1), the films were developed (developer Ilford LS, Mobberley Cheshire, GB), and the nephron segments were stained with methyl green pyronin. The method used for analysis of silver grain counts has been described in detail previously (13). Briefly, each tubular segment was photographed, and the counting of silver grains was performed on enlargements of the films, over nuclei and cytoplasm, using circles of known surface area superimposed over the structure considered. The background of the film was determined in the same way. For each tubular segment, counting was performed on about 20 nuclear and cytoplasmic areas, and the mean value was calculated. Total nuclear labeling was calculated by subtracting from nuclear silver grain counts the cytoplasmic silver grain counts of the same photograph, considered as background for the nuclei (13). Cytoplasmic labeling is given after subtraction of the background of the film. For each tubular segment, the specific labeling is the difference between the total labeling of the considered segment (incubation with the tritiated steroid alone) and the mean value of nonspecific labeling (tritiated steroid plus competitor) in the same experiment. Steroids The steroids used were: tritiated A ([3H]A; Amersham: 41.25 Ci/mmol), tritiated spironolactone ([3H]SC9420; Searle, Des Plaines, IL; 54.7 Ci/mmol), unlabeled A (Sigma), and unlabeled SC9420 (Searle). Five different incubation conditions were used: 1) 2 nM [3H]A, 2) 2 nM [3H]A plus a 100-fold excess of unlabeled A, 3) 2 nM [3H]SC9420, 4) 2 nM [3H]SC9420 plus a 100-fold excess of SC9420, and 5) 2 nM [3H]SC9420 plus a 100fold excess of A.
Results Figure 1 consists of photographs of the labeling on the CCT in the different conditions. In the presence of [3H] A or [3H]SC9420 alone (Fig. 1, A and B), a clear nuclear labeling was observed. It has to be noted that the labeling by [3H]A was slightly higher than that by [3H]SC9420. The nuclear labeling by [3H]A and [3H]SC9420 was almost completely displaced by unlabeled A or SC9420, respectively (Fig. 1, C and D). The nuclear labeling by [3H]SC9420 was displaced by unlabeled A (Fig. IE) in a manner equivalent to that by unlabeled SC9420 (Fig. ID). In all conditions, the cytoplasmic labeling was very low and did not differ from the background (mean values of background: A, 3.51 ± 0.43 silver grains/100 /im2 surface area; B, 3.30 ± 0.43; C, 2.14 ± 0.20; D, 2.47 ±
281
0.31; E, 3.27 ± 0.21). Figure 2 gives the mean value of nuclear labeling. To take into account the difference in the specific activities of [3H]A and [3H]SC9420, we have corrected the silver grain counts of [3H]A labeling by the ratio of specific activity between the two steroids (54.7:41.25). The total nuclear labeling was more than 10 silver grains/100 /*m2 for [3H]A. It was somewhat lower for [3H]SC9420 (6.0/ 100 Aim2). The nonspecific nuclear labeling was low (0.52.3). The displacement of [3H]SC9420 labeling over nuclei by unlabeled A and that by spironolactone were roughly equivalent (91% and 87%, respectively). Table 1 gives the values of specific nuclear and cytoplasmic labeling by [3H]A and [3H]SC9420, calculated by subtracting the nonspecific labeling (in the presence of an excess of unlabeled A or SC9420) from the total labeling. This specific nuclear labeling was 35% lower for spirolactone than for A. No significant specific cytoplasmic labeling was observed with either steroids.
Discussion Previous studies have demonstrated that spirolactones specifically bind to mineralocorticoid-sensitive tissues. A specific cytosolic binding of a tritiated spirolactone (SC26304) was found by Marver et al. (3) in rat kidney. On the other hand, Claire et al. (6) and Rossier et al. (7) reported specific binding of tritiated prorenone to rat kidney and toad urinary bladder, respectively. In these three studies, the biological activity of spirolactones, consisting of inhibition of A-induced sodium transport, was associated with specific binding. In the present study we compared the binding of tritiated spironolactone (SC9420) and tritiated A in CCT microdissected from the same rabbit kidney, using an autoradiographic technique on intact cells (13). SC9420 has been shown to have a high A antagonist activity and no agonist activity (5). The CCT has repeatedly been demonstrated to be the most sensitive nephron segment for A (12-15). A physiological concentration of A (2 nM) was used, which has been shown to result in maximal specific binding in the rabbit CCT (13). The same concentration of [3H]SC9420 was used in the present study. The results with A were very similar to those of our previous reports (13, 15): high specific nuclear labeling, with very low nonspecific nuclear labeling, and very low specific cytoplasmic labeling. Similar results were obtained in the present study with [3H]SC9420. The specific binding was 65% that of A, suggesting a slightly lower affinity of SC9420 for the receptor than A. If this were the case, one cannot be assured that the observed labeling represents the maximal specific binding of [3H] SC9420, which could be higher than that found in the present study. The labeling by [3H]SC9420 was almost
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 17:57 For personal use only. No other uses without permission. . All rights reserved.
NUCLEAR BINDING OF SPIRONOLACTONE
282
Endo • 1991 Voll28«Nol ,
FIG. 1. Photographs (x530) of the autoradiographic labeling of CCT by [3H]A and [3H]SC9420. Incubation was performed for 30 min at 30 C with [3H]A (2 nM; A), [3H]SC9420 (2 nM; B), [3H]A (2 nM) plus a 100-fold excess of unlabeled A (C), [3H]SC9420 (2 nM) plus a 100fold excess of unlabeled SC9420 (D), and [3H]SC9420 (2 nM) plus a 100-fold excess of unlabeled A (E). Silver grains are concentrated over nuclear shadows and are almost completely displaced by the
• -i
.#1
excess of unlabeled steroids. Very few
silver grains are visible over cytoplasmic areas.
•#
• / •• TABLE 1. Specific nuclear and cytoplasmic labeling by [3H]aldosterone and [3H]SC9420
3H-A alone
Specific labeling (silver grains/100 /im2 surface ares)
cT 10 H
_J UJ
CD ^>
QC
