Biochimica et Biophysica Acta, 1033 (1990) 41-48
41
Elsevier BBAGEN 23236
Purification and characterization of the activated mineralocorticoid receptor from rat myocardium G e o r g e L a z a r 1, M a u r i c e P a g a n o 2 a n d M a n j u l K . A g a r w a l 3 1 Institute of Pathophysiology, Szeged Medical School, Szeged (Hungary), Departments of Physiology and 2Biochemistry and 3Hormone Laboratory, UFR Broussais, Centre Unioersitaire des Cordeliers, Paris (France)
(Received 5 July 1989)
Key words: Mineralocorticoidreceptor; Corticosteroid;RU 26752; R 5020; Hormone receptor; (Rat rnyocardium)
Cardiac cytosol from adrenalectomized rats was radiolabelled with 10 nM tritiated RU 26752, R 5020 or aldosterone, to saturate the mineralocorticoid receptor (MCR) in the presence of 1 /tM RU 38486 to block the glucocorticoid and progestin receptors. Free steroids were removed by charcoal treatment and the radiolabelled cytosol was passed through a phosphoceHulose column. The MCR peak in the phosphoceilulose eluate was activated at 25 °C for 45 min, adsorbed onto the DNA-cellulose and finally extracted once each with buffers containing 1 M potassium chloride or 25 mM magnesium chloride. The pooled DNA-cellulose extracts, reequilibrated with 10 nM [3HIRU 26752, were resolved as a single, homogeneous band of 78 kDa upon polyacrylamide gel electrophoresis, lon-exchange analysis of the purified MCR on DEAE-cellulose-52 revealed a single peak in the 0.017 M sodium phosphate region with both RU 26752 and R 5020, but aldosterone dissociated during this procedure. Molecular filtration on Ultrogel AcA-44 columns revealed a major 145 kDa peak, with some smaller components of 40 and 80 kDa. These hydrodynamic properties of the purified MCR are at variance with those of the native receptor in crude myocardial cytosoi, and suggest that some post-translational modifications in vivo may be required for the expression of MCR-mediated responses.
Introduction It is generally accepted that steroid hormones bind to intracellular receptors with high affinity. The amplification of target-specific genes by the steroid-receptor complex remains an important tool in the elucidation of the organization and expression of the complex mammalian genome (for reviews see Refs. 1-3). Mineralocorticoid hormones regulate the ionic balance in the mammal, whose dysfunction leads to hypertension and related syndromes [4]. Rat myocardium apparently contains mineralocorticoid-specific receptors (MCR) whose function remains unknown [5] and whose saturation kinetics and stability appear somewhat different from rat renal MCR [6-8]. The problem is further complicated by the apparent existence of MCR components that preferentially bind antagonists such as R 5020 and RU 26752 and are thus seemingly distinct from aldosterone-specific sites in crude cytosol [7-11].
Abbreviations: MCIL mineraloeorticoid receptor; G-CR, glueocorticoid receptor; PR, progesteronereceptor. Correspondence: M.K. Agarwal, 15 rue de l'Ecole de Mrdecine, 75270 Paris Cedex 06, France.
It is obvious from the above that, as a prelude to elucidating MCR-specific genetic amplification, the purification of the M C R is of paramount importance. Although receptors for all classes of steroid hormone have been biochemically purified to homogeneity, the MCR has hitherto escaped this sort of analysis due primarily to its low concentration in the cell and its instability. We present here a rapid and simple procedure that yields a homogeneous preparation of MCR. The purified M C R from rat kidney and from heart were compared to assess organ-specific differences seen in crude preparations [7,8]. Additionally, the purified material was analyzed to gain understanding of the nature of the steroid-dependent receptor heterogeneity described earlier [10-12].
Materials and Methods A n i m a l s a n d tissues. Male, Wistar rats (Iffa-Credo, France), aged 5 - 6 weeks (150-200 g), were bilaterally adrenalectomized under diethyl ether anesthesia~ and maintained in a climate-controlled room with access to pellet food and water ad hbitum. Animals were killed under diethyl ether anesthesia, blood was collected by
0304-4165/90/$03.50 © 1990 ElsevierSciencePublishers B.V. (BiomedicalDivision)
42 aortic cannulation, and the organs were subsequently perfused with an excess of ice-cold buffer A. The organs were weighed, minced and homogenized in 2 vol. of ice-cold buffer A (w/v) using a teflon pestle. A lipid-free supernate was finally obtained by centrifugation at 105000 × g (50 min, 4°C). Blood was allowed to clot at room temperature for 1 h, followed by 1 h at 4 ° C, and finally centrifuged at 10000 × g (15 min, 4°C) to get a clear serum. Buffers. Buffer A contained 0.02 M sodium phosphate (pH 7.5), 1 mM EDTA, 12 mM monothioglycerol and 10% glycerol and was used for the preparation of organ cytosol, as well as the DNA-cellulose resin. The receptor was eluted from DNA-cellulose by buffers B and C containing 20 M Tris-HC1 (pH 7.5), 1 mM EDTA, 12 mM monothioglycerol, 10% glycerol and 25 mM magnesium (B) or 1 M potassium (C) chloride, respectively. Buffer D for Ultrogel chromatography contained 0.02 M sodium phosphate (pH 7.5), 12 mM monothioglycerol, and I mM EDTA. DEAE-cellulose-52 chromatography was performed in buffer E containing 0.002 M sodium phosphate (pH 7.5), 12 mM monothioglycerol and 1 mM EDTA; gradient elution was carried out with this buffer E plus that containing 0.2 M sodium phosphate (pH 7.5). Purification Protocol. In order to circumvent the problem of receptor denaturation observed with the natural hormone aldosterone, we standardized our procedure with the synthetic antagonist RU 26752 that stabilizes MCR [7-9] for periods longer than the total purification procedure here. The organ cytosol was equilibrated for 60 min at 4 ° C with 10 nM [3H]RU 26752 (or another steroid), as described in our detailed kinetic studies [7-11], in the presence of 1 /LM RU 38486 which has high affinity for the glucocorticoid (GCR) and progesterone (PR) receptors [13,14]. Free steroids were subsequently removed by further incubation (10 min 4°C) in the presence of 50 mg activated charcoal (Sigma C-5260) per ml organ cytosol that was finally removed by centrifugation (3000 × g, 10 min, 4 ° C). Aliquots of the cytosol were quantitated for protein [15] and radioactivity [7-11]. Labelled cytosol was quickly passed through glass wool to remove traces of charcoal. The resulting filtrate was loaded onto a 10 ml bed of phosphocellulose P-11 (Whatman), prepared and packed in a 10 ml polyethylene syringe according to the manufacturer's instructions. Fractions of 2 ml were collected at a flow rate of 50-60 m l / h at 4 ° C to elute MCR from the phosphocellulose resin with the aid of buffer A. A gradient elution with 30 ml each of 0.02 M and 0.2 M phosphate buffer A was thereafter attempted to detect further retention of radioactivity. Aliquots of 100 #1 were counted in 10 ml Picofluor to localize the receptor peak which was immediately pooled and analyzed for protein-bound radioactivity. This procedure was finally
established after several trials under various conditions of elution. The phosphocellulose pool (generally 12-16 ml) was allowed to stand at 25 °C for 45 min to activate the MCR and then incubated with 1 g of DNA-ceUulose (Sigma) for 45 min in an ice bath with occasional agitation. The DNA-cellulose resin was subsequently washed three times with an excess of ice-cold buffer A by centrifugation for 5 rain at 5000 × g at 4°C. The receptor adsorbed by the resin was finally extracted for 30 min at 4 ° C by elution once each with 2 ml of buffers B and C. Fresh DNA-ceUulose was used each time. The two DNA-cellulose extracts were pooled and incubated with 10 nM tritiated steroid (60 min 4°C). Unbound radioactivity was removed by additional incubation (10 min 4°C) in the presence of an equal volume of activated charcoal (2.5%) -dextran (0.25%) that was centrifuged out at 3000 x g, 10 min, 4°C. A flow chart of the purification procedure is given below. cytosolradiolabeUedwith10 nM tritiatedsteroid+1/~MRU 38486
1 t l 1 1
phosphocellulosechromatography
activationof the PC pool
DNA-cellulosechromatography
requilibrationof the eluatewith10 nM radioligand polyacrylamideget electrophoresis ion-exchangechromatography
molecularfiltration
Gel electrophoresis. SDS-polyacrylamide (15% w/v) slab gels were prepared by the standard procedure [16] used routinely in our laboratory [17]. Samples were first concentrated by precipitation with an equal volume of 60% (w/v) trichloroacetic acid together with 10 /~1 of 2% (w/v) sodium deoxycholate as carrier. The precipitate, collected by centrifugation (5000 × g, 15 min, 4 ° C) was washed with 1 ml acetone and finally dissolved in 30 t~l of 0.07 M Tris-HC1 (pH 6.8) containing 1% SDS (w/v), 0.001% (w/v) Bromophenol blue, 100 mM dithiothreitol and 10% (w/v) glycerol, as described previously [18]. After 5 min at 100°C, 30 #1 sample was applied to the gel slab which, following migration, was fixed with 12% trichloroacetic acid overnight, and finally stained with Coomassie violet (Serva). Pharmacia LM r markers kit was used to calibrate the gel with materials of known molecular weight. Molecular filtration. Rat blood serum was incubated (60 min, 4°C) with 0.2 ~Ci/ml [14C]cortisol. 1.5 ml
43 serum and 2 ml DNA-cellulose extract were mixed, charcoal-treated (50 mg/ml), passed through glass wool and loaded onto an Ultrogel column (1 x 30 cm) prepared and packed in buffer D according to our original methods [6-11]. Fractions of 2 ml were collected at 15 m l / h at 4 ° C and analyzed for protein and radioactivity. The column was calibrated with markers of known molecular weight (Sigma kit MWCF-200). Ion-exchange chromatography. 2 ml DNA-cellulose extract and 1.5 ml serum, double labelled as above for Ultrogel columns, were desalted by passage through a 10 ml bed of Bio-Gel P-10 resin (Bio-Rad No. 81103) in a 10 ml syringe. Fractions of 2 ml were collected with the low ionic buffer E (60-70 ml/h, 4°C) and 0.1 ml samples were counted in 10 ml Picofluor to localize the radioactive peak which was immediately pooled. The P-10 pool (8-12 ml) was loaded onto a DEAEcellulose-52 column (0.5 x 25 cm). After a prewash with 20 ml of low ionic buffer E, a gradient elution was begun (at the arrow in the figures) with 30 ml each of 0.002 M and 0.2 M sodium phosphate buffer E, at 20-30 m l / h at 4°C. Fractions of 2 ml were collected and analyzed for protein and radioactivity [6-11]. Radioactive and protein determinations. After column chromatography, aliquots of 1 ml were mixed with 10 ml Picofluor 40 (Packard) in 20 ml polyethylene vials and counted in a Tricarb scintillation counter (Packard), taking into account background and spectral overlap according to a quench curve established for this purpose [6-11]. Absorbance at 280 nm was determined in a double-beam spectrophotometer, with automatic base line comparison against the elution buffer. Specific protein determinations were made by the Bradford technique [15]. Reagents and chemicals. Ultrogel AcA-44 (Batch No. 4030) was a product of LKB, Sweden; Microgranular DEAE-cellulose-52 (Batch No. 6152024) and cellulose phosphate P-11 (Batch No. 6611013) were obtained from Whatman, U.K. Bio-Gel P-10 (control No. 569) was supplied by Bio-Rad Laboratories, CA, U.S.A. The activated charcoal (Lot 270-0022), Dextran D-3759 (Lot 125 F-C254), and DNA-cellulose (Lot 117 F-8135) were all purchased from Sigma, MO, U.S.A. The kit (Sigma MW-GF-200) for the calibration of Ultrogel columns contained the following molecular mass markers: Blue dextran, 2 MDa; fl-amylase, 200 kDa; alcohol dehydrogenase, 150 kDa; carbonic anhydrase, 30 kDa; and cytochrome c, 12 kDa. Polyacrylamide gels were calibrated by the LMr markers kit from Pharmacia containing: phosphorylase b, 94 kDa; bovine serum albumin, 68 kDa; ovalbumin, 43 kDa; carbonic anhydrase, 30 kDa; soybean trypsin inhibitor 20 kDa; and lactalbumin, 14 kDa. Steroids. Tritiated RU 26752 (50 Ci/mM, reference X3025 A) and the corresponding cold steroid (Lot 7) were provided free of charge by Roussel-Uclaf, Ro-
mainville, France. [1,2-3H]Aldosterone (45 Ci/mM, batch 34) and [4-14C]cortisol 55 mCi/mM, batch 57) were purchased from Amersham, U.K. [3H]R 5020 (87 Ci/mM, lot 1009-149) was a product of New England Nuclear. The radiochemcial purity exceeded 97~ in thin-layer chromatography in all cases. Trivial names of the steroids used in this study are as follows: aldosterone, 11fl,21-dihydroxy-3,20-dioxopregn-4-ene-18-al; cortisol, 11fl,17a,21-trihydroxypregn-4-ene-3,20-dione; RU 26752, 7a-propyl-3-oxo17ot-pregn-4-ene-21,17-carbalactone; R 5020, promegestone 17a-methyl. Results
Given the unstable nature of MCR, and since the natural hormone aldosterone binds to several types of receptor with varying degrees of affinity [2,3,10], we chose to standardize our purification procedure with RU 26752 that binds specifically to MCR [7-9,19,20], stabilizes MCR for more than 8 h [7,8], and permits the activation process [7,8] required for the subsequent DNA-cellulose step. Its affinity for rat renal MCR is approx. 50% that of aldosterone ( K d = 15 nm at 4°C), but it does not bind to other groups of receptors [20]. MCR stability was additionally assured by the extensive use of phosphate buffers that are known to inhibit a number of degradative processes such as endogenous proteinases [21-22]. An excess of radioinert RU 38486 was systematically used to saturate GCR and PR [13,14]. Data in fig. 1 show that the MCR-[3H]RU 26752 complex eluted in the prewash from the phosphocellulose column; no further radioactivity could be detected with a phosphate gradient. This peak was absent if 1 #M radioinert RU 26752 was added at the time of
_3 E X
E o. u
1. o o'
J
o
lO 20 FRACTION NO. Fig. 1. Partialenrichmentof rat cardiac M C R on phosphocellulose. 15 ml myocardialcytosol,labelledwith 10 nM tritiated RU 26752 in the presence of 1 /~M R U 38486, was loaded onto the resin. After a prewash with 40 ml buffer A, a gradient elution was begun (not shown).The radioactivepeak(o o) betweenarrowswas pooled for subsequentanalyses.
44 TABLE I
Two-step purification scheme of the activated mineralocorticoid receptor from rat myocardium Myocardial cytosol from 15 adrenalectomized rats was incubated with 10 n M [3H]RU 26752, in the presence of 1 p M R U 38486, passed through a phosphocellulose column, activated at 25 ° C for 45 min, and then extracted from DNA-cellulose, as described in the Materials and Methods. Protein (Bradford) and charcoal-resistant radioactivity were determined at each step. Source Cytosol Phosphocellulose pool DNA-cellulose extract
cpm/mg protein
Fold enrichment
336
-
640
1.9
534 558
1591
Total cpm
% recovery
114240
-
59136
52
26 727
23
at times observed, as with other receptors [18], but the component of more than 75 kDa was by far the most consistent and abundant. Several attempts were made to further purify this preparation of MCR. Rechromatography of the final product on DNA-cellulose led to inactivation of steroid binding. Ion-exchange chromatography, prior to the DNA-ceUulose step, also resulted in the loss of steroidbinding capacity during desalting and concentration of the pooled eluate. An affinity resin, containing deoxycorticosterone as the active ligand, appeared to be of only limited value for partial enrichment of MCR although more than 1000-fold purification by a similar 1
2
3
Kd
a
94 Phosphorylase B incubation of cardiac cytosol with 10 nM [3H]RU 26752, confirming specificity (not shown). This step leads to approx. 2-fold enrichment of MCR (see above). Similar results were obtained with aldosterone or R5020 (not shown). After several trials with various buffers and elution protocols, this phosphocellulose passage was systematically used as the first step in all subsequent studies. Since the activated RU 26752-MCR complex specifically binds to the DNA-cellulose [7,8], it was considered logical to exploit this property for MCR purification. RU 26752-MCR complex, ehited from the phosphocellulose column, was activated for 45 rain at 25 °C [7-8], and then incubated with DNA-ceUulose for 45 min at 4°C. The MCR retained by this resin was finally extracted once each with buffers containing 1 M potassium chloride or 25 mM magnesium chloride, and then requilibrated with 10 nM [3H]RU 26752 to asses MCR purity and recovery. A typical experiment shown in Table I reveals a purification of almost 1600-fold by this simple procedure, with 23% recovery. Actual yield in both parameters is certainly higher, since dextran-charcoal, used to remove the free steroid after final equilibration of the DNA-cellulose extract with 10 nM radiolabelled RU 26752, invariably led to nonspecific adsorption of MCR-bound material due to a low protein content in the sample; interexperimental variation was therefore difficult to assess with certainty. Inclusion of a carrier protein actually improved the percent purification, though inconsistently. These are very comparable to the results obtained with rat renal, in place of cardiac, cytosol (data not shown). Analysis of the activated cardiac MCR, purified as above, on polyacrylamide gels revealed a single, homogeneous band of 78 kDa (Fig. 2a) with an R F of 0.14 (Fig. 2b), again very comparable to the 75 kDa band in the kidney (not shown). A more than 90 kDa band was
68 BSA
43
Ovalbumin
30 Carbonic Anhydrase
20
Soybean
Trypsin Inhibitor 10]_ ._~hosphor ylase B
/
b
BSA
i
bumin
=/ 0
ic Anhydrase
21
!
| 14 0
I a
"~.~sin , 0.5
Inhibitor
Lactalbumin~'~O
, Rf 1.0
Fig. 2. Activated rat cardiac M C R migrates as a homogeneous band on polyacrylamide gels. 50 /~g of whole cytosol (lane 1, 2a) or less than 1 p g extract from D N A cellulose (lane 2, 2a) were concentrated and electrophoresed as described in the Materials and Methods. Markers of known molecular weight (lane 3, 2a) and R F (b) were run concurrently for gel calibration.
45 a • 3H (10 -3)
+I'
M PO~
L E .5"
"
2
X
O.
:E
o
n
o :r..xr:t~" X.X~
2o FRACTION
0
3;
"
o
I0
NO
3b
20 FRACTION
~o
NO
(1o-')I' C ,3 x (io +3)
2
'/.,
C
(IO-qlSlc 1H iIo-Z)
•
,,o,
~ 311 d 0.5
1.
O..
o
! IO
•' 20 FRACTION
°
30
0-'
40
~
0
NO
:
"
o
10
20 FRACTION
3'0
z,b
NO
Fig. 3. Analysis of homogeneous rat myocardial MCR by ion-exchange chromatography on DEAE-cellulos¢. Rat blood serum labelled with [:4C]¢ortisol, and cardiac MCR extracted from DNA-ccUulosein the presence of tritiated RU 26752 (a and b), R 5020 (c) or aldosterone (d) were coehromatographed on DE-52 as described in the Materials and Methods section; serum was omitted for b. x x, absorbance at 280 nm; O
O, 3H; O-
procedure has been described by some authors [23]. Finally, contrary to a report by one group [24], M C R did not bind to the commercial preparation of Heparin-Sepharose in o u r laboratory. AU thee data are not included here. In view of steroid- and tissue-dependent heterogeneity of native M C R in crude rat cardiac [6,8], or renal [7,11,12] cytosol, we used the procedure described above to compare the nature of homogeneous, activated M C R purified in the presence of various steroids. The M C R was analyzed on two different chromatographic supports by our original and unique approach of doublelabelled elution, where blood serum tagged with [t4C]cortisol is used as a position marker to assess reproducibility between runs. Additionally, serum also serves as a carrier to stabilize the minimal amounts of tritium-labelled M C R protein in the purified preparation. Data in Fig. 3 show that M C R purified in the presence of either tritiated RU 26752 (Fig. 3a) or R 5020 (Fig. 3c) was resolved as a sharp peak in the 0.017 M phosphate region from the DEAE-cellulose-52 column. Rat serum [t4C]cortisol-transcortin was consistently eluted as a sharp peak in the 0.06 M sodium
O, :4C.
phosphate region, confirming excellent reproducibility. Purified MCR-bound [3H]aldosterone appeared unstable under these conditions such that the radioactivity eluted as a broad band divided between the late prowash and the beginning of the phosphate gradient (Fig. 3d), reminiscent of dissociated [3H]RU 26752 eluted similarly if the carrier serum was omitted from the procedure (Fig. 3b). This elution Profile with the purified M C R is quite distinct from that observed in whol~ cardiac cytosol where aldosterone, R U 26752 and +R 5020 are eluted as M C R 2 [2+,3] M c R a and M C R 4 positions [6-12] in 0.006 M, 0.02 M and 0.06 M phosphate, respectively, as in the kidney cytosol (not shown)~ Data in Fig. 4a show a linear calibration curve of molecular mass markers on the Ultrogel AcA 44 column. Rat blood serum transcortin was consistently resolved in the 95 kDa region in all these double-labelled elution profiles, R U 26752 elmed:as, single peak of about 80 kDa (Fig. 4b) but aldostcrone was bound to a major component of 145 k D a along with two minor species (Fig. 4c). The R 5020 labelling p a u e r n was confined to the 145-80 kDa regions followed by another of approx. 40 kDa (Fig. 4d). The basic 40 kDa unit may thus dimerize into the 80 kDa species, followed by
46
200'
~*.YLASS
~
~ALCOHOL
7 100 o so.
~
b (IO-I)I4c 3H (IO "3)
GENASE
2
lo
40 .5
t
E
A280
C).
1.o
1:s
Ve/Vo
o
21o
lO FRACTION NQ
2o
(10 -II) IIIc.~. 3H (10 -3)
2'
d
c
• )
E
( 1 0 4 ) 1 1 C ~ 3 H (10 -3) T
2-.8
T
~1 -.4
,R A280
X~
~'280
2
O
i
o
i
lO FRACTION NO
,
"1
2O
O,i
~
O
o
10 F R A C T I O N NO
io
Fig. 4. Filtration of the purified rat mycr.ardial MCR on ultrogel AcA-44 columns. The ultrogel column was calibrated with markers or known molecular weight (a). Rat blood serum and myocardial MCR labelled with tritiated RU 26752 (b), aldosterone (c), or R 5020 (d), were coeluted as described in the Materials and Methods. The elution position of various MCR peaks was calculated relative to transcortin (T) shown on the molecular mass scale (a). x x, absorbance at 280 nm; o . o, 3H; • • , 14C.
another dimerization into a 145 kDa entity that, in the kidney, undergoes further aggregation into 290 kDa complexes (not shown), thereby leading to organ-specific differences. Discussion
The availability, in recent years, of a number of synthetic derivatives of spirolactone has provided potent new tools to dissect MCR structure and function [7-8,20]. Since the affinity of the natural hormone aldosterone for several receptors is only 2- or 3-fold lower than that for MCR [2,10,26], and since the resulting complex is labile [3,10,23,24], we standardized our purification procedure with R U 26752 that exhibits little or no affinity for receptors other than MCR [20], and that stabilized MCR for much longer than the 8 h
[7,8] purification protocol established here. Indeed, the purified aldosterone-MCR complex was rather unstable, obviating the nature of products studied in two earlier attempts at partial purification of MCR [23,24]. The use of phosphate buffers assured further stability and also protected against endogenous proteolysis [21,22]. We finally chose to omit exogenous proteinase inhibitors altogether, since they variously alter the steroid [22] and the DNA [27] binding sites. Addition of molybdate in buffers subsequent to the activation step did not alter the chromatographic elution profile in any way. The two-step procedure detailed here yields a homogeneous preparation of rat cardiac MCR within 8 h. Nonspecific adsorption of radioactivity bound to the minute quantity of protein makes it difficult to assess
47 the exact degree of purity and interexperimental variation by the charcoal technique, but the polyacrylamide gel pattern would suggest at least 60~ homogeneity with about 20~ yield, depending upon the steroid and the run. Further analysis of the purified product was not possible for several reasons. Scatchard analysis could not be attempted due to the instability of the purified MCR protein in the presence of charcoal required for the removal of free steroids. An exchange assay for MCR is currently wanting. Affinity labelling of MCR, too, has never been attempted hitherto, since no specific iigand for this purpose is actually available. We are attempting this with the aid of R 5020 despite its lack of specificity for MCR. The purified product possesses an intact steroid binding site on a 78 kDa domain. This agrees well with purified preparations for GCR where components range from above 90 to only 40 kDa [18]. We did at times observe a 94 kDa unit, but the predominance of the smaller species may represent a mechanism to distinguish MCR from GCR, and possibly PR. Attempts to enrich rat cardiac MCR by chromatography on Heparin-Sepharose [24], or an affinity gel [23], were uniformly negative in our hands. It is conceivable that these authors were not looking at MCR at all, since other cellular vectors such as GCR or PR [28], and perhaps the Heparin-binding protein [29], could he nonspecifically labelled by aldosterone. Differences in the steroid, the animal species, the resin source and elution conditions, could perhaps explain discrepancies. Analysis of purified MCR by two other methods was made possible by our original, two decade old technique of double-labelled chromatography, where serum transcortin is used as a position marker, and monitors reproducibility between column runs [3,7,8,30]. Additionally, serum proteins form a carrier to stabilize the minimal amounts of purified MCR. Whereas 50-110 kDa complexes are observed with MCR in crude heart cytosol [8], the purified cardiac MCR revealed aggregates ranging from 40 to 145 kDa, depending upon the steroid, suggesting possible polymerization into dimers and tetramers [31], perhaps via partial, endogenous proteolysis [21,22]. Purified renal MCR was consistently eluted as a 290 kDa peak, under similar conditions, suggesting organ-specific differences. Cytoplasmic components such as the heat shock protein [32], a metallic factor [33] and transfer RNA [34], are all believed to stabilize the native GCR, although glucocorticoids do not in any way alter tRNA structure and function [35]. Such cellular constituents apparently are not required for the action of cloned MCR in vitro [36]. Their importance seemingly lies in maintaining receptor stereospecifcity in vivo, and their removal during MCR purification can alter the assessment of the molecular size.
The purified, activated cardiac MCR singularly eluted in 0.017 M sodium phosphate region from DE-52 columns, irrespective of the occupying steroid, and quite distinct from blood serum transcortin in 0.06 M phosphate in these double-labelled studies. On the other hand, the unactivated receptor in crude cardiac cytosol eluted from the DEAE-cellulose-52 column with aldosterone, RU 26752 and R 5020 in 0.006 M, 0.024 M and 0.06 M phosphate regions, or the MR 2, MR 3, MR 4 positions, respectively [6-11]. This suggests that the receptor heterogeneity described more than a decade ago [12] may be a consequence of post-translational modifications of intracellular MCR, contrary to suggestions that native MCR appeared similar in crude cytosol and after partial purification [23]. A high degree of structural homology exists between GCR and MCR [36], and a hydrophobic pocket is common to both GCR and PR [13]. From the wellknown cross-reactivity between GCR and MCR [10,12,26], and the antagonism between mineralocorticoids and progestins [10,37,38], the same ligand can most obviously induce stereospecific configurations in several receptor types leading to an integrated hormone response in the whole organism. Two or more binding sites may actually be present on one and the same receptor, leading to an erroneous impression of heterogeneity [39,40]. It has recently been suggested that the enzyme 11flhydroxydehydrogenase may confer mineralocorticoid specificity in the appropriate target tissue [26]. However, liver is rich in this enzyme but neither contains MCR [30] nor expresses mineralocorticoid-specific functions. Additionally, synthetic spirolactones such as RU 26752 and ZK 91587 are potent mineralocorticoid antagonists [7,8,20], but apparently not substrates for the above enzyme. Thus, MCR-mediated genetic modulation remains to be elucidated. In conclusion, the stability of RU 26752-MCR complexes in phosphate buffers permitted the purification of cardiac MCR to homogeneity. Although also valid for the purification of R 5020-MCR complexes, some difficulty was encountered in the procedure when aldosterone was used and this confirms MCR instability observed in earlier studies. The heterogeneity described in biochemical studies of unactivated MCR in crude cytosol may be due to cytoplasmic modifcations of the steroid-binding domain, or of some associated component(s) that appear(s) to confer stability to the intracellular receptor. Comparison of the purified product with the cloned receptor should further the understanding of MCR-mediated processes.
Acknowledgements Thanks are due to Dr. D. Philibert for RU 26752 samples. This work was aided by grants from UFR
48 Broussais, and NATO (0782/87). Parts of this study were presented at the Annual Meeting of the Endocrine Society, Seattle, June 1989, and have therefore appeared as an Abstract. References 1 Jensen, E.V. and DeSombre, E.R. (1972) Annu. Rev. Biochem. 41, 203-230. 2 Agarwal, M.K. (1978) FEBS Lett. 85, 1-8. 3 Agarwal, M.K. (1983) in Principles of Recepterology (Agarwal, M.K., ed.), pp. 1-68, Walter de Gruyter, Berlin. 4 Grunfeld, J.P., Eloy, L., Moura, A.M., Ganeval, D., Frendo, B.R. and Worcel, M. (1985) Hypertension 7, 292-299. 5 Liew, C.C., Liu, D.K. and Gornall, A.G. (1972) Endocrinology 90, 488-495. 6 Agarwal, M.K. and Philippe, M. (1979) J. Mol. Cell. Cardiol. 11, 115-126. 7 Agarwal, M.K. and Kalimi, M. (1988) Biochim. Biophys. Acta 964, 105-112. 8 Agarwal, M.K. and Kalimi, M. (1989) Biochem. Med. Metab. Biol. 41, 36-45. 9 Lazar, G. and Agarwal, M.K. (1986) Biochem. Biophys. Res. Commun. 134, 261-265. 10 Agarwal, M.K. (1982) in Hormone Antagonists (Agarwal, M.K., ed.), pp. 307-334, Walter de Gruyter, Berlin. 11 Agarwal, M.K. and Paillard, J. (1979) Biochem. Biophys. Res. Commun. 89, 77-84. 12 Agarwal, M.K. (1977) in Multiple Molecular Forms of Steroid Hormone Receptors (Agarwal, M.K., ed.), pp. 93-112, Elsevier/ North Holland, Amsterdam. 13 Philibert, D. (1984) in Adrenal Steroid Antagonism (Agarwal, M.K., ed.), pp. 77-102, Walter de Gruyter, Berlin. 14 Agarwal, M.K., Hainque, B., Moustaid, N. and Lazar, G. (1987) FEBS Lett. 217, 221-226. 15 Bradford, D. (1976) Anal. Biochem. 72, 248-254. 16 Laemmli, U.K. (1970) Nature 227, 680-685. 17 Keppler, D., Fondan~he, M.C., Fumeron, V.D., Pagano, M. and Burtin, P. (1988) Can. Res. 48, 6855-6862. 18 Scheidereit, C., Geisse, S., Westphal, H.M. and Beato, M. (1983) Nature 304, 749-752.
19 Agarwal, M.K. and Kalimi, M. (1987) Biochem. Biophys. Res. Commun. 143, 398-402. 20 Nedelec, R., Philibert, D. and Torelli, V. (1986) in Proceedings of the 3rd SCI-RSC Medicinal Chemistry Symposium (Lambert, R.W., ed.), pp. 322-344, Cambridge. 21 Agarwal, M.K. and Philippe, M. (1979) in Proteases and Hormones (Agarwal, M.K., ed.), pp. 93-118, Elsevier, Amsterdam. 22 Agarwal, M.K. and Philippe, M. (1981) Biochem. Med. 26, 265-276. 23 Lombes, M., Claire, M., Lustenberger, P., Miehand, A. and Oblin, M.E.R. (1987) J. Biol. Chem. 262; 8121-8127. 24 Weisz, A., Cicatiello, L. and Bresciani, F. (1986) J. Steroid Binchem. 24, 461-467. 25 Agarwal, M.K. (1976) Biochem. J. 154, 567-575. 26 Funder, J.W., Pearce, P.T., Smith, R. and Smith, A.l. (1988) Science 242, 583-585. 27 Kasayama, S., Noma, K., Sato, B., Nakao, M., Nishizawa, Y., Matusumoto, K. and Nishimoto, S. (1987) J. Steroid Biochem. 28, 273-277. 28 Li, S. and Li, J.J. (1978) Endocrinology 103, 2119-2127. 29 Lankes, W., Greismacher, A., Grunwald, J., Albiez, R.S. and Keller, R. (1988) Biochem. J. 251, 831-842. 30 Agarwal, M.K. (1975) Nature 254, 623-625. 31 Rexin, M., Busch, W., Segnitz, B. and Gehring, U. (1988) FEBS Lett. 241,234-238. 32 Mendel, D.B. and Orti, E. (1988) J. Biol. Chem. 263, 6695-6702. 33 Meshinchi, S.,Grippo, J.F., Sanchez, E.R., Bresnick, E.H. and Pratt, W.B. (1988) J. Biol. Chem. 263, 16809-16817. 34 Aft, M. and Vedeckis, W. (1987) Science 235, 467-470. 35 Agarwal, M.K., Hanoune, J., Yu, F.L., Weinstein, I.B. and Feigeison, P. (1969) Biochemistry 8, 4806-4812. 36 Arriza, J.L., Weinberger, C., Cerelli, (3., Glaser, T.M., Handelin, B.L., Housman, D. and Evans, R.M. (1987) Science 237, 268-275. 37 Wambach, G. (1987) in Receptor Mediated Antisteroid Action (Agarwal, M.K., ed.), pp. 169-196, Walter de Gruyter, Berlin. 38 Landau, R.L. (1979) in Antthormones (Agarwal, M.K., ed.), pp. 153-166, Elsevier/North Holland, Amsterdam. 39 Agarwal, M.K. and Raynand, J.P. (1989) FEBS Lett. 243, 1-3. 40 Kalimi, M. and Agarwal, M.K. (1988) Biochem. Biophys. Res. Commun. 131, 265-371.
41
Elsevier BBAGEN 23236
Purification and characterization of the activated mineralocorticoid receptor from rat myocardium G e o r g e L a z a r 1, M a u r i c e P a g a n o 2 a n d M a n j u l K . A g a r w a l 3 1 Institute of Pathophysiology, Szeged Medical School, Szeged (Hungary), Departments of Physiology and 2Biochemistry and 3Hormone Laboratory, UFR Broussais, Centre Unioersitaire des Cordeliers, Paris (France)
(Received 5 July 1989)
Key words: Mineralocorticoidreceptor; Corticosteroid;RU 26752; R 5020; Hormone receptor; (Rat rnyocardium)
Cardiac cytosol from adrenalectomized rats was radiolabelled with 10 nM tritiated RU 26752, R 5020 or aldosterone, to saturate the mineralocorticoid receptor (MCR) in the presence of 1 /tM RU 38486 to block the glucocorticoid and progestin receptors. Free steroids were removed by charcoal treatment and the radiolabelled cytosol was passed through a phosphoceHulose column. The MCR peak in the phosphoceilulose eluate was activated at 25 °C for 45 min, adsorbed onto the DNA-cellulose and finally extracted once each with buffers containing 1 M potassium chloride or 25 mM magnesium chloride. The pooled DNA-cellulose extracts, reequilibrated with 10 nM [3HIRU 26752, were resolved as a single, homogeneous band of 78 kDa upon polyacrylamide gel electrophoresis, lon-exchange analysis of the purified MCR on DEAE-cellulose-52 revealed a single peak in the 0.017 M sodium phosphate region with both RU 26752 and R 5020, but aldosterone dissociated during this procedure. Molecular filtration on Ultrogel AcA-44 columns revealed a major 145 kDa peak, with some smaller components of 40 and 80 kDa. These hydrodynamic properties of the purified MCR are at variance with those of the native receptor in crude myocardial cytosoi, and suggest that some post-translational modifications in vivo may be required for the expression of MCR-mediated responses.
Introduction It is generally accepted that steroid hormones bind to intracellular receptors with high affinity. The amplification of target-specific genes by the steroid-receptor complex remains an important tool in the elucidation of the organization and expression of the complex mammalian genome (for reviews see Refs. 1-3). Mineralocorticoid hormones regulate the ionic balance in the mammal, whose dysfunction leads to hypertension and related syndromes [4]. Rat myocardium apparently contains mineralocorticoid-specific receptors (MCR) whose function remains unknown [5] and whose saturation kinetics and stability appear somewhat different from rat renal MCR [6-8]. The problem is further complicated by the apparent existence of MCR components that preferentially bind antagonists such as R 5020 and RU 26752 and are thus seemingly distinct from aldosterone-specific sites in crude cytosol [7-11].
Abbreviations: MCIL mineraloeorticoid receptor; G-CR, glueocorticoid receptor; PR, progesteronereceptor. Correspondence: M.K. Agarwal, 15 rue de l'Ecole de Mrdecine, 75270 Paris Cedex 06, France.
It is obvious from the above that, as a prelude to elucidating MCR-specific genetic amplification, the purification of the M C R is of paramount importance. Although receptors for all classes of steroid hormone have been biochemically purified to homogeneity, the MCR has hitherto escaped this sort of analysis due primarily to its low concentration in the cell and its instability. We present here a rapid and simple procedure that yields a homogeneous preparation of MCR. The purified M C R from rat kidney and from heart were compared to assess organ-specific differences seen in crude preparations [7,8]. Additionally, the purified material was analyzed to gain understanding of the nature of the steroid-dependent receptor heterogeneity described earlier [10-12].
Materials and Methods A n i m a l s a n d tissues. Male, Wistar rats (Iffa-Credo, France), aged 5 - 6 weeks (150-200 g), were bilaterally adrenalectomized under diethyl ether anesthesia~ and maintained in a climate-controlled room with access to pellet food and water ad hbitum. Animals were killed under diethyl ether anesthesia, blood was collected by
0304-4165/90/$03.50 © 1990 ElsevierSciencePublishers B.V. (BiomedicalDivision)
42 aortic cannulation, and the organs were subsequently perfused with an excess of ice-cold buffer A. The organs were weighed, minced and homogenized in 2 vol. of ice-cold buffer A (w/v) using a teflon pestle. A lipid-free supernate was finally obtained by centrifugation at 105000 × g (50 min, 4°C). Blood was allowed to clot at room temperature for 1 h, followed by 1 h at 4 ° C, and finally centrifuged at 10000 × g (15 min, 4°C) to get a clear serum. Buffers. Buffer A contained 0.02 M sodium phosphate (pH 7.5), 1 mM EDTA, 12 mM monothioglycerol and 10% glycerol and was used for the preparation of organ cytosol, as well as the DNA-cellulose resin. The receptor was eluted from DNA-cellulose by buffers B and C containing 20 M Tris-HC1 (pH 7.5), 1 mM EDTA, 12 mM monothioglycerol, 10% glycerol and 25 mM magnesium (B) or 1 M potassium (C) chloride, respectively. Buffer D for Ultrogel chromatography contained 0.02 M sodium phosphate (pH 7.5), 12 mM monothioglycerol, and I mM EDTA. DEAE-cellulose-52 chromatography was performed in buffer E containing 0.002 M sodium phosphate (pH 7.5), 12 mM monothioglycerol and 1 mM EDTA; gradient elution was carried out with this buffer E plus that containing 0.2 M sodium phosphate (pH 7.5). Purification Protocol. In order to circumvent the problem of receptor denaturation observed with the natural hormone aldosterone, we standardized our procedure with the synthetic antagonist RU 26752 that stabilizes MCR [7-9] for periods longer than the total purification procedure here. The organ cytosol was equilibrated for 60 min at 4 ° C with 10 nM [3H]RU 26752 (or another steroid), as described in our detailed kinetic studies [7-11], in the presence of 1 /LM RU 38486 which has high affinity for the glucocorticoid (GCR) and progesterone (PR) receptors [13,14]. Free steroids were subsequently removed by further incubation (10 min 4°C) in the presence of 50 mg activated charcoal (Sigma C-5260) per ml organ cytosol that was finally removed by centrifugation (3000 × g, 10 min, 4 ° C). Aliquots of the cytosol were quantitated for protein [15] and radioactivity [7-11]. Labelled cytosol was quickly passed through glass wool to remove traces of charcoal. The resulting filtrate was loaded onto a 10 ml bed of phosphocellulose P-11 (Whatman), prepared and packed in a 10 ml polyethylene syringe according to the manufacturer's instructions. Fractions of 2 ml were collected at a flow rate of 50-60 m l / h at 4 ° C to elute MCR from the phosphocellulose resin with the aid of buffer A. A gradient elution with 30 ml each of 0.02 M and 0.2 M phosphate buffer A was thereafter attempted to detect further retention of radioactivity. Aliquots of 100 #1 were counted in 10 ml Picofluor to localize the receptor peak which was immediately pooled and analyzed for protein-bound radioactivity. This procedure was finally
established after several trials under various conditions of elution. The phosphocellulose pool (generally 12-16 ml) was allowed to stand at 25 °C for 45 min to activate the MCR and then incubated with 1 g of DNA-ceUulose (Sigma) for 45 min in an ice bath with occasional agitation. The DNA-cellulose resin was subsequently washed three times with an excess of ice-cold buffer A by centrifugation for 5 rain at 5000 × g at 4°C. The receptor adsorbed by the resin was finally extracted for 30 min at 4 ° C by elution once each with 2 ml of buffers B and C. Fresh DNA-ceUulose was used each time. The two DNA-cellulose extracts were pooled and incubated with 10 nM tritiated steroid (60 min 4°C). Unbound radioactivity was removed by additional incubation (10 min 4°C) in the presence of an equal volume of activated charcoal (2.5%) -dextran (0.25%) that was centrifuged out at 3000 x g, 10 min, 4°C. A flow chart of the purification procedure is given below. cytosolradiolabeUedwith10 nM tritiatedsteroid+1/~MRU 38486
1 t l 1 1
phosphocellulosechromatography
activationof the PC pool
DNA-cellulosechromatography
requilibrationof the eluatewith10 nM radioligand polyacrylamideget electrophoresis ion-exchangechromatography
molecularfiltration
Gel electrophoresis. SDS-polyacrylamide (15% w/v) slab gels were prepared by the standard procedure [16] used routinely in our laboratory [17]. Samples were first concentrated by precipitation with an equal volume of 60% (w/v) trichloroacetic acid together with 10 /~1 of 2% (w/v) sodium deoxycholate as carrier. The precipitate, collected by centrifugation (5000 × g, 15 min, 4 ° C) was washed with 1 ml acetone and finally dissolved in 30 t~l of 0.07 M Tris-HC1 (pH 6.8) containing 1% SDS (w/v), 0.001% (w/v) Bromophenol blue, 100 mM dithiothreitol and 10% (w/v) glycerol, as described previously [18]. After 5 min at 100°C, 30 #1 sample was applied to the gel slab which, following migration, was fixed with 12% trichloroacetic acid overnight, and finally stained with Coomassie violet (Serva). Pharmacia LM r markers kit was used to calibrate the gel with materials of known molecular weight. Molecular filtration. Rat blood serum was incubated (60 min, 4°C) with 0.2 ~Ci/ml [14C]cortisol. 1.5 ml
43 serum and 2 ml DNA-cellulose extract were mixed, charcoal-treated (50 mg/ml), passed through glass wool and loaded onto an Ultrogel column (1 x 30 cm) prepared and packed in buffer D according to our original methods [6-11]. Fractions of 2 ml were collected at 15 m l / h at 4 ° C and analyzed for protein and radioactivity. The column was calibrated with markers of known molecular weight (Sigma kit MWCF-200). Ion-exchange chromatography. 2 ml DNA-cellulose extract and 1.5 ml serum, double labelled as above for Ultrogel columns, were desalted by passage through a 10 ml bed of Bio-Gel P-10 resin (Bio-Rad No. 81103) in a 10 ml syringe. Fractions of 2 ml were collected with the low ionic buffer E (60-70 ml/h, 4°C) and 0.1 ml samples were counted in 10 ml Picofluor to localize the radioactive peak which was immediately pooled. The P-10 pool (8-12 ml) was loaded onto a DEAEcellulose-52 column (0.5 x 25 cm). After a prewash with 20 ml of low ionic buffer E, a gradient elution was begun (at the arrow in the figures) with 30 ml each of 0.002 M and 0.2 M sodium phosphate buffer E, at 20-30 m l / h at 4°C. Fractions of 2 ml were collected and analyzed for protein and radioactivity [6-11]. Radioactive and protein determinations. After column chromatography, aliquots of 1 ml were mixed with 10 ml Picofluor 40 (Packard) in 20 ml polyethylene vials and counted in a Tricarb scintillation counter (Packard), taking into account background and spectral overlap according to a quench curve established for this purpose [6-11]. Absorbance at 280 nm was determined in a double-beam spectrophotometer, with automatic base line comparison against the elution buffer. Specific protein determinations were made by the Bradford technique [15]. Reagents and chemicals. Ultrogel AcA-44 (Batch No. 4030) was a product of LKB, Sweden; Microgranular DEAE-cellulose-52 (Batch No. 6152024) and cellulose phosphate P-11 (Batch No. 6611013) were obtained from Whatman, U.K. Bio-Gel P-10 (control No. 569) was supplied by Bio-Rad Laboratories, CA, U.S.A. The activated charcoal (Lot 270-0022), Dextran D-3759 (Lot 125 F-C254), and DNA-cellulose (Lot 117 F-8135) were all purchased from Sigma, MO, U.S.A. The kit (Sigma MW-GF-200) for the calibration of Ultrogel columns contained the following molecular mass markers: Blue dextran, 2 MDa; fl-amylase, 200 kDa; alcohol dehydrogenase, 150 kDa; carbonic anhydrase, 30 kDa; and cytochrome c, 12 kDa. Polyacrylamide gels were calibrated by the LMr markers kit from Pharmacia containing: phosphorylase b, 94 kDa; bovine serum albumin, 68 kDa; ovalbumin, 43 kDa; carbonic anhydrase, 30 kDa; soybean trypsin inhibitor 20 kDa; and lactalbumin, 14 kDa. Steroids. Tritiated RU 26752 (50 Ci/mM, reference X3025 A) and the corresponding cold steroid (Lot 7) were provided free of charge by Roussel-Uclaf, Ro-
mainville, France. [1,2-3H]Aldosterone (45 Ci/mM, batch 34) and [4-14C]cortisol 55 mCi/mM, batch 57) were purchased from Amersham, U.K. [3H]R 5020 (87 Ci/mM, lot 1009-149) was a product of New England Nuclear. The radiochemcial purity exceeded 97~ in thin-layer chromatography in all cases. Trivial names of the steroids used in this study are as follows: aldosterone, 11fl,21-dihydroxy-3,20-dioxopregn-4-ene-18-al; cortisol, 11fl,17a,21-trihydroxypregn-4-ene-3,20-dione; RU 26752, 7a-propyl-3-oxo17ot-pregn-4-ene-21,17-carbalactone; R 5020, promegestone 17a-methyl. Results
Given the unstable nature of MCR, and since the natural hormone aldosterone binds to several types of receptor with varying degrees of affinity [2,3,10], we chose to standardize our purification procedure with RU 26752 that binds specifically to MCR [7-9,19,20], stabilizes MCR for more than 8 h [7,8], and permits the activation process [7,8] required for the subsequent DNA-cellulose step. Its affinity for rat renal MCR is approx. 50% that of aldosterone ( K d = 15 nm at 4°C), but it does not bind to other groups of receptors [20]. MCR stability was additionally assured by the extensive use of phosphate buffers that are known to inhibit a number of degradative processes such as endogenous proteinases [21-22]. An excess of radioinert RU 38486 was systematically used to saturate GCR and PR [13,14]. Data in fig. 1 show that the MCR-[3H]RU 26752 complex eluted in the prewash from the phosphocellulose column; no further radioactivity could be detected with a phosphate gradient. This peak was absent if 1 #M radioinert RU 26752 was added at the time of
_3 E X
E o. u
1. o o'
J
o
lO 20 FRACTION NO. Fig. 1. Partialenrichmentof rat cardiac M C R on phosphocellulose. 15 ml myocardialcytosol,labelledwith 10 nM tritiated RU 26752 in the presence of 1 /~M R U 38486, was loaded onto the resin. After a prewash with 40 ml buffer A, a gradient elution was begun (not shown).The radioactivepeak(o o) betweenarrowswas pooled for subsequentanalyses.
44 TABLE I
Two-step purification scheme of the activated mineralocorticoid receptor from rat myocardium Myocardial cytosol from 15 adrenalectomized rats was incubated with 10 n M [3H]RU 26752, in the presence of 1 p M R U 38486, passed through a phosphocellulose column, activated at 25 ° C for 45 min, and then extracted from DNA-cellulose, as described in the Materials and Methods. Protein (Bradford) and charcoal-resistant radioactivity were determined at each step. Source Cytosol Phosphocellulose pool DNA-cellulose extract
cpm/mg protein
Fold enrichment
336
-
640
1.9
534 558
1591
Total cpm
% recovery
114240
-
59136
52
26 727
23
at times observed, as with other receptors [18], but the component of more than 75 kDa was by far the most consistent and abundant. Several attempts were made to further purify this preparation of MCR. Rechromatography of the final product on DNA-cellulose led to inactivation of steroid binding. Ion-exchange chromatography, prior to the DNA-ceUulose step, also resulted in the loss of steroidbinding capacity during desalting and concentration of the pooled eluate. An affinity resin, containing deoxycorticosterone as the active ligand, appeared to be of only limited value for partial enrichment of MCR although more than 1000-fold purification by a similar 1
2
3
Kd
a
94 Phosphorylase B incubation of cardiac cytosol with 10 nM [3H]RU 26752, confirming specificity (not shown). This step leads to approx. 2-fold enrichment of MCR (see above). Similar results were obtained with aldosterone or R5020 (not shown). After several trials with various buffers and elution protocols, this phosphocellulose passage was systematically used as the first step in all subsequent studies. Since the activated RU 26752-MCR complex specifically binds to the DNA-cellulose [7,8], it was considered logical to exploit this property for MCR purification. RU 26752-MCR complex, ehited from the phosphocellulose column, was activated for 45 rain at 25 °C [7-8], and then incubated with DNA-ceUulose for 45 min at 4°C. The MCR retained by this resin was finally extracted once each with buffers containing 1 M potassium chloride or 25 mM magnesium chloride, and then requilibrated with 10 nM [3H]RU 26752 to asses MCR purity and recovery. A typical experiment shown in Table I reveals a purification of almost 1600-fold by this simple procedure, with 23% recovery. Actual yield in both parameters is certainly higher, since dextran-charcoal, used to remove the free steroid after final equilibration of the DNA-cellulose extract with 10 nM radiolabelled RU 26752, invariably led to nonspecific adsorption of MCR-bound material due to a low protein content in the sample; interexperimental variation was therefore difficult to assess with certainty. Inclusion of a carrier protein actually improved the percent purification, though inconsistently. These are very comparable to the results obtained with rat renal, in place of cardiac, cytosol (data not shown). Analysis of the activated cardiac MCR, purified as above, on polyacrylamide gels revealed a single, homogeneous band of 78 kDa (Fig. 2a) with an R F of 0.14 (Fig. 2b), again very comparable to the 75 kDa band in the kidney (not shown). A more than 90 kDa band was
68 BSA
43
Ovalbumin
30 Carbonic Anhydrase
20
Soybean
Trypsin Inhibitor 10]_ ._~hosphor ylase B
/
b
BSA
i
bumin
=/ 0
ic Anhydrase
21
!
| 14 0
I a
"~.~sin , 0.5
Inhibitor
Lactalbumin~'~O
, Rf 1.0
Fig. 2. Activated rat cardiac M C R migrates as a homogeneous band on polyacrylamide gels. 50 /~g of whole cytosol (lane 1, 2a) or less than 1 p g extract from D N A cellulose (lane 2, 2a) were concentrated and electrophoresed as described in the Materials and Methods. Markers of known molecular weight (lane 3, 2a) and R F (b) were run concurrently for gel calibration.
45 a • 3H (10 -3)
+I'
M PO~
L E .5"
"
2
X
O.
:E
o
n
o :r..xr:t~" X.X~
2o FRACTION
0
3;
"
o
I0
NO
3b
20 FRACTION
~o
NO
(1o-')I' C ,3 x (io +3)
2
'/.,
C
(IO-qlSlc 1H iIo-Z)
•
,,o,
~ 311 d 0.5
1.
O..
o
! IO
•' 20 FRACTION
°
30
0-'
40
~
0
NO
:
"
o
10
20 FRACTION
3'0
z,b
NO
Fig. 3. Analysis of homogeneous rat myocardial MCR by ion-exchange chromatography on DEAE-cellulos¢. Rat blood serum labelled with [:4C]¢ortisol, and cardiac MCR extracted from DNA-ccUulosein the presence of tritiated RU 26752 (a and b), R 5020 (c) or aldosterone (d) were coehromatographed on DE-52 as described in the Materials and Methods section; serum was omitted for b. x x, absorbance at 280 nm; O
O, 3H; O-
procedure has been described by some authors [23]. Finally, contrary to a report by one group [24], M C R did not bind to the commercial preparation of Heparin-Sepharose in o u r laboratory. AU thee data are not included here. In view of steroid- and tissue-dependent heterogeneity of native M C R in crude rat cardiac [6,8], or renal [7,11,12] cytosol, we used the procedure described above to compare the nature of homogeneous, activated M C R purified in the presence of various steroids. The M C R was analyzed on two different chromatographic supports by our original and unique approach of doublelabelled elution, where blood serum tagged with [t4C]cortisol is used as a position marker to assess reproducibility between runs. Additionally, serum also serves as a carrier to stabilize the minimal amounts of tritium-labelled M C R protein in the purified preparation. Data in Fig. 3 show that M C R purified in the presence of either tritiated RU 26752 (Fig. 3a) or R 5020 (Fig. 3c) was resolved as a sharp peak in the 0.017 M phosphate region from the DEAE-cellulose-52 column. Rat serum [t4C]cortisol-transcortin was consistently eluted as a sharp peak in the 0.06 M sodium
O, :4C.
phosphate region, confirming excellent reproducibility. Purified MCR-bound [3H]aldosterone appeared unstable under these conditions such that the radioactivity eluted as a broad band divided between the late prowash and the beginning of the phosphate gradient (Fig. 3d), reminiscent of dissociated [3H]RU 26752 eluted similarly if the carrier serum was omitted from the procedure (Fig. 3b). This elution Profile with the purified M C R is quite distinct from that observed in whol~ cardiac cytosol where aldosterone, R U 26752 and +R 5020 are eluted as M C R 2 [2+,3] M c R a and M C R 4 positions [6-12] in 0.006 M, 0.02 M and 0.06 M phosphate, respectively, as in the kidney cytosol (not shown)~ Data in Fig. 4a show a linear calibration curve of molecular mass markers on the Ultrogel AcA 44 column. Rat blood serum transcortin was consistently resolved in the 95 kDa region in all these double-labelled elution profiles, R U 26752 elmed:as, single peak of about 80 kDa (Fig. 4b) but aldostcrone was bound to a major component of 145 k D a along with two minor species (Fig. 4c). The R 5020 labelling p a u e r n was confined to the 145-80 kDa regions followed by another of approx. 40 kDa (Fig. 4d). The basic 40 kDa unit may thus dimerize into the 80 kDa species, followed by
46
200'
~*.YLASS
~
~ALCOHOL
7 100 o so.
~
b (IO-I)I4c 3H (IO "3)
GENASE
2
lo
40 .5
t
E
A280
C).
1.o
1:s
Ve/Vo
o
21o
lO FRACTION NQ
2o
(10 -II) IIIc.~. 3H (10 -3)
2'
d
c
• )
E
( 1 0 4 ) 1 1 C ~ 3 H (10 -3) T
2-.8
T
~1 -.4
,R A280
X~
~'280
2
O
i
o
i
lO FRACTION NO
,
"1
2O
O,i
~
O
o
10 F R A C T I O N NO
io
Fig. 4. Filtration of the purified rat mycr.ardial MCR on ultrogel AcA-44 columns. The ultrogel column was calibrated with markers or known molecular weight (a). Rat blood serum and myocardial MCR labelled with tritiated RU 26752 (b), aldosterone (c), or R 5020 (d), were coeluted as described in the Materials and Methods. The elution position of various MCR peaks was calculated relative to transcortin (T) shown on the molecular mass scale (a). x x, absorbance at 280 nm; o . o, 3H; • • , 14C.
another dimerization into a 145 kDa entity that, in the kidney, undergoes further aggregation into 290 kDa complexes (not shown), thereby leading to organ-specific differences. Discussion
The availability, in recent years, of a number of synthetic derivatives of spirolactone has provided potent new tools to dissect MCR structure and function [7-8,20]. Since the affinity of the natural hormone aldosterone for several receptors is only 2- or 3-fold lower than that for MCR [2,10,26], and since the resulting complex is labile [3,10,23,24], we standardized our purification procedure with R U 26752 that exhibits little or no affinity for receptors other than MCR [20], and that stabilized MCR for much longer than the 8 h
[7,8] purification protocol established here. Indeed, the purified aldosterone-MCR complex was rather unstable, obviating the nature of products studied in two earlier attempts at partial purification of MCR [23,24]. The use of phosphate buffers assured further stability and also protected against endogenous proteolysis [21,22]. We finally chose to omit exogenous proteinase inhibitors altogether, since they variously alter the steroid [22] and the DNA [27] binding sites. Addition of molybdate in buffers subsequent to the activation step did not alter the chromatographic elution profile in any way. The two-step procedure detailed here yields a homogeneous preparation of rat cardiac MCR within 8 h. Nonspecific adsorption of radioactivity bound to the minute quantity of protein makes it difficult to assess
47 the exact degree of purity and interexperimental variation by the charcoal technique, but the polyacrylamide gel pattern would suggest at least 60~ homogeneity with about 20~ yield, depending upon the steroid and the run. Further analysis of the purified product was not possible for several reasons. Scatchard analysis could not be attempted due to the instability of the purified MCR protein in the presence of charcoal required for the removal of free steroids. An exchange assay for MCR is currently wanting. Affinity labelling of MCR, too, has never been attempted hitherto, since no specific iigand for this purpose is actually available. We are attempting this with the aid of R 5020 despite its lack of specificity for MCR. The purified product possesses an intact steroid binding site on a 78 kDa domain. This agrees well with purified preparations for GCR where components range from above 90 to only 40 kDa [18]. We did at times observe a 94 kDa unit, but the predominance of the smaller species may represent a mechanism to distinguish MCR from GCR, and possibly PR. Attempts to enrich rat cardiac MCR by chromatography on Heparin-Sepharose [24], or an affinity gel [23], were uniformly negative in our hands. It is conceivable that these authors were not looking at MCR at all, since other cellular vectors such as GCR or PR [28], and perhaps the Heparin-binding protein [29], could he nonspecifically labelled by aldosterone. Differences in the steroid, the animal species, the resin source and elution conditions, could perhaps explain discrepancies. Analysis of purified MCR by two other methods was made possible by our original, two decade old technique of double-labelled chromatography, where serum transcortin is used as a position marker, and monitors reproducibility between column runs [3,7,8,30]. Additionally, serum proteins form a carrier to stabilize the minimal amounts of purified MCR. Whereas 50-110 kDa complexes are observed with MCR in crude heart cytosol [8], the purified cardiac MCR revealed aggregates ranging from 40 to 145 kDa, depending upon the steroid, suggesting possible polymerization into dimers and tetramers [31], perhaps via partial, endogenous proteolysis [21,22]. Purified renal MCR was consistently eluted as a 290 kDa peak, under similar conditions, suggesting organ-specific differences. Cytoplasmic components such as the heat shock protein [32], a metallic factor [33] and transfer RNA [34], are all believed to stabilize the native GCR, although glucocorticoids do not in any way alter tRNA structure and function [35]. Such cellular constituents apparently are not required for the action of cloned MCR in vitro [36]. Their importance seemingly lies in maintaining receptor stereospecifcity in vivo, and their removal during MCR purification can alter the assessment of the molecular size.
The purified, activated cardiac MCR singularly eluted in 0.017 M sodium phosphate region from DE-52 columns, irrespective of the occupying steroid, and quite distinct from blood serum transcortin in 0.06 M phosphate in these double-labelled studies. On the other hand, the unactivated receptor in crude cardiac cytosol eluted from the DEAE-cellulose-52 column with aldosterone, RU 26752 and R 5020 in 0.006 M, 0.024 M and 0.06 M phosphate regions, or the MR 2, MR 3, MR 4 positions, respectively [6-11]. This suggests that the receptor heterogeneity described more than a decade ago [12] may be a consequence of post-translational modifications of intracellular MCR, contrary to suggestions that native MCR appeared similar in crude cytosol and after partial purification [23]. A high degree of structural homology exists between GCR and MCR [36], and a hydrophobic pocket is common to both GCR and PR [13]. From the wellknown cross-reactivity between GCR and MCR [10,12,26], and the antagonism between mineralocorticoids and progestins [10,37,38], the same ligand can most obviously induce stereospecific configurations in several receptor types leading to an integrated hormone response in the whole organism. Two or more binding sites may actually be present on one and the same receptor, leading to an erroneous impression of heterogeneity [39,40]. It has recently been suggested that the enzyme 11flhydroxydehydrogenase may confer mineralocorticoid specificity in the appropriate target tissue [26]. However, liver is rich in this enzyme but neither contains MCR [30] nor expresses mineralocorticoid-specific functions. Additionally, synthetic spirolactones such as RU 26752 and ZK 91587 are potent mineralocorticoid antagonists [7,8,20], but apparently not substrates for the above enzyme. Thus, MCR-mediated genetic modulation remains to be elucidated. In conclusion, the stability of RU 26752-MCR complexes in phosphate buffers permitted the purification of cardiac MCR to homogeneity. Although also valid for the purification of R 5020-MCR complexes, some difficulty was encountered in the procedure when aldosterone was used and this confirms MCR instability observed in earlier studies. The heterogeneity described in biochemical studies of unactivated MCR in crude cytosol may be due to cytoplasmic modifcations of the steroid-binding domain, or of some associated component(s) that appear(s) to confer stability to the intracellular receptor. Comparison of the purified product with the cloned receptor should further the understanding of MCR-mediated processes.
Acknowledgements Thanks are due to Dr. D. Philibert for RU 26752 samples. This work was aided by grants from UFR
48 Broussais, and NATO (0782/87). Parts of this study were presented at the Annual Meeting of the Endocrine Society, Seattle, June 1989, and have therefore appeared as an Abstract. References 1 Jensen, E.V. and DeSombre, E.R. (1972) Annu. Rev. Biochem. 41, 203-230. 2 Agarwal, M.K. (1978) FEBS Lett. 85, 1-8. 3 Agarwal, M.K. (1983) in Principles of Recepterology (Agarwal, M.K., ed.), pp. 1-68, Walter de Gruyter, Berlin. 4 Grunfeld, J.P., Eloy, L., Moura, A.M., Ganeval, D., Frendo, B.R. and Worcel, M. (1985) Hypertension 7, 292-299. 5 Liew, C.C., Liu, D.K. and Gornall, A.G. (1972) Endocrinology 90, 488-495. 6 Agarwal, M.K. and Philippe, M. (1979) J. Mol. Cell. Cardiol. 11, 115-126. 7 Agarwal, M.K. and Kalimi, M. (1988) Biochim. Biophys. Acta 964, 105-112. 8 Agarwal, M.K. and Kalimi, M. (1989) Biochem. Med. Metab. Biol. 41, 36-45. 9 Lazar, G. and Agarwal, M.K. (1986) Biochem. Biophys. Res. Commun. 134, 261-265. 10 Agarwal, M.K. (1982) in Hormone Antagonists (Agarwal, M.K., ed.), pp. 307-334, Walter de Gruyter, Berlin. 11 Agarwal, M.K. and Paillard, J. (1979) Biochem. Biophys. Res. Commun. 89, 77-84. 12 Agarwal, M.K. (1977) in Multiple Molecular Forms of Steroid Hormone Receptors (Agarwal, M.K., ed.), pp. 93-112, Elsevier/ North Holland, Amsterdam. 13 Philibert, D. (1984) in Adrenal Steroid Antagonism (Agarwal, M.K., ed.), pp. 77-102, Walter de Gruyter, Berlin. 14 Agarwal, M.K., Hainque, B., Moustaid, N. and Lazar, G. (1987) FEBS Lett. 217, 221-226. 15 Bradford, D. (1976) Anal. Biochem. 72, 248-254. 16 Laemmli, U.K. (1970) Nature 227, 680-685. 17 Keppler, D., Fondan~he, M.C., Fumeron, V.D., Pagano, M. and Burtin, P. (1988) Can. Res. 48, 6855-6862. 18 Scheidereit, C., Geisse, S., Westphal, H.M. and Beato, M. (1983) Nature 304, 749-752.
19 Agarwal, M.K. and Kalimi, M. (1987) Biochem. Biophys. Res. Commun. 143, 398-402. 20 Nedelec, R., Philibert, D. and Torelli, V. (1986) in Proceedings of the 3rd SCI-RSC Medicinal Chemistry Symposium (Lambert, R.W., ed.), pp. 322-344, Cambridge. 21 Agarwal, M.K. and Philippe, M. (1979) in Proteases and Hormones (Agarwal, M.K., ed.), pp. 93-118, Elsevier, Amsterdam. 22 Agarwal, M.K. and Philippe, M. (1981) Biochem. Med. 26, 265-276. 23 Lombes, M., Claire, M., Lustenberger, P., Miehand, A. and Oblin, M.E.R. (1987) J. Biol. Chem. 262; 8121-8127. 24 Weisz, A., Cicatiello, L. and Bresciani, F. (1986) J. Steroid Binchem. 24, 461-467. 25 Agarwal, M.K. (1976) Biochem. J. 154, 567-575. 26 Funder, J.W., Pearce, P.T., Smith, R. and Smith, A.l. (1988) Science 242, 583-585. 27 Kasayama, S., Noma, K., Sato, B., Nakao, M., Nishizawa, Y., Matusumoto, K. and Nishimoto, S. (1987) J. Steroid Biochem. 28, 273-277. 28 Li, S. and Li, J.J. (1978) Endocrinology 103, 2119-2127. 29 Lankes, W., Greismacher, A., Grunwald, J., Albiez, R.S. and Keller, R. (1988) Biochem. J. 251, 831-842. 30 Agarwal, M.K. (1975) Nature 254, 623-625. 31 Rexin, M., Busch, W., Segnitz, B. and Gehring, U. (1988) FEBS Lett. 241,234-238. 32 Mendel, D.B. and Orti, E. (1988) J. Biol. Chem. 263, 6695-6702. 33 Meshinchi, S.,Grippo, J.F., Sanchez, E.R., Bresnick, E.H. and Pratt, W.B. (1988) J. Biol. Chem. 263, 16809-16817. 34 Aft, M. and Vedeckis, W. (1987) Science 235, 467-470. 35 Agarwal, M.K., Hanoune, J., Yu, F.L., Weinstein, I.B. and Feigeison, P. (1969) Biochemistry 8, 4806-4812. 36 Arriza, J.L., Weinberger, C., Cerelli, (3., Glaser, T.M., Handelin, B.L., Housman, D. and Evans, R.M. (1987) Science 237, 268-275. 37 Wambach, G. (1987) in Receptor Mediated Antisteroid Action (Agarwal, M.K., ed.), pp. 169-196, Walter de Gruyter, Berlin. 38 Landau, R.L. (1979) in Antthormones (Agarwal, M.K., ed.), pp. 153-166, Elsevier/North Holland, Amsterdam. 39 Agarwal, M.K. and Raynand, J.P. (1989) FEBS Lett. 243, 1-3. 40 Kalimi, M. and Agarwal, M.K. (1988) Biochem. Biophys. Res. Commun. 131, 265-371.
