0013.7227/92/1311~0374503.00/0 Endocrinology Copyright 0 1992 by The Endocrine

Vol. 131, No. 1 Prrnted in U.S.A.

Society

Characterization of Nuclear Angiotensin-II-Binding Sites in Rat Liver and Comparison with Plasma Membrane Receptors* SHIOW-SHIH

TANG,

HARALD

ROGGt,

ROBERT

SCHUMACHER+,

AND

VICTOR

J. DZAUS

Molecular and Cellular Vascular Research Laboratory, Division of Vascular Medicine and Atherosclerosis, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts 02115 ABSTRACT

753). a selective blocker of the Am-11 recentor subtvne I. fullv inhibits nuclear Ang-II binding with affinity similar to that in ‘plasma membrane. The pH optimum for [““IlAng-II binding to nuclei was 7.0, while binding to plasma membrane was optimal at pH 8.0. Low concentrations (0.05-0.1 mM) of dithiothreitol increased J’““IlAng-II binding to nuclei, but not to plasma membrane. In the absence>of detergent, Ang-II-binding sites appear to consist of soluble protein releasable from nuclei by freezing and thawing, hence distinct in physicochemical properties from the membrane-bound receptor. Size-exclusion HPLC estimated the mol wt of the soluble Ang-II-binding sites to be 66 kilodaltons. These nuclear Ang-II-binding sites have some similarities to but also show notable physicochemical differences from plasma membrane Ang-II receptors, and they may play a role in mediating the intracellular action of Ang-II. (Endocrinology 131: 374-380, 1992)

Although the action of angiotensin-II (Ang-II) is believed to be mediated by a transmembrane signal transduction mechanism, accumulating evidence suggests that Ang-II may also have a direct nuclear action. We have characterized the nuclear Ang-II-binding site in purified nuclei preparation from rat liver and compared it to plasma membrane Ang-II receptors. [““IlAng-II binding to isolated nuclei reached equilibration in 30 min at 25 C, slower than binding to plasma membrane, which reached equilibration within 10 min. Scatchard analysis of [““IlAng-II binding to isolated nuclei revealed a single class of binding sites (Kd = 1.4 nM; binding capacity = 10 fmol/mg protein or 460 sites/nucleus). In the nuclear preparation, Ang-II and its fragments competed for binding a potency order of Ang-III = Ang-II > Ang-II-(l-7) > Ang-II-(l-6) > Ang-II-(l-5). Losartan potassium (DuP

T

of binding of the hormone-receptor complex to specific DNA sequences, leading to changes in the expression of target genes. This mechanism of action has been documented for thyroid and steroid hormones (6, 7). Circumstantial evidence suggests that Ang-11 may also have a direct nuclear action. In the first place, Ang-II is taken up by cells (8, 9) and can be localized in the nucleus (10). Second, in a preliminary report the occurrence of Ang-IIbinding sites in the nuclei of hepatocytes and spleen cells has been described (11). Third, Ang-II increases RNA synthesis and changes chromatin conformation in isolated rat liver nuclei (12). These data, thus, suggest that a functional Ang receptor(s) may exist in the nucleus. More recently, a cytosolic Ang-II-binding protein from the rabbit liver was identified and purified (13). Taken together, all of these findings suggest that Ang-II may have an intracellular target of action. It is not clear whether the Ang-II-binding sites in the plasma membrane and nuclei are identical or different. In this paper the characteristics of the nuclear Ang-II-binding sites in rat liver are described. It will be shown that this entity has a high affinity for the highly selective, nonpeptidic Ang-II antogonist Losartan (DuP 753) and, hence, has properties in common with the recently discovered Ang receptor subtype AT1 (14), but it will be demonstrated that they are distinct in several physicochemical properties from the plasma membrane Ang-II receptors.

HE PEPTIDE hormone angiotensin-II (Ang-II) plays a central role in cardiorenal regulation. This octapeptide has multiple sites of action and induces diverse tissue responses Thus far, the mechanism by way of which Ang-II can regulate such diverse responses has remained elusive, nor is it clear how the temporal control of these responses is mediated. Some of them occur immediately (1, 2), but others are considerably delayed (3,4). Evidence points to the plasma membrane as a primary site of action for rapid effects of Ang-II (1, 2), presumably mediated through signal transduction. However, these signals are transient, and it is still not clear whether the delayed responses are mediated by the same mechanisms. One other possibility is that Ang-II or its homolog plays a direct intracellular role. Major effects of insulin are thought to be linked to its action on cell surface receptors. However, direct microinjection of insulin into the nucleus of Xenopus laevis oocytes also increases the synthesis of RNA, protein, and glycogen (5), suggesting that the nucleus may be an important intracellular site of insulin action. These nuclear effects may be the result Received January 6, 1992. Address all correspondence and requests for reprints to: Shiow-Shih Tang, Ph.D., Pediatric Nephrology Unit, Massachusetts General Hospital, BHX-4, Fruit Street, Boston, Massachusetts 02114. *This work was supported by NIH Grants HL-35610, HL-35792, HL-19259, HL-35252, HL-43131, and HL-42663; University of California Tobacco-Related Disease Program lRT215; an unrestricted gift from Bristol Myers Squibb for Cardiovascular Research; and the William F. Milton Fund. t Present address: Department of Research, Pharmaceuticals Division, Ciba-Geigy Ltd., CH-4002 Basel, Switzerland. $ Present address: Division of Cardiovascular Medicine, Stanford University School of Medicine, Stanford, California 94305.

Materials

and Methods

Materials [‘251]Ang-II land Nuclear

(2200 FCi/nmol) was purchased (Boston, MA). Ang-II, Ang-III,

from DuPont-New Engsaralasin ([Sar’,Ala’]Ang-

374

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 February 2015. at 08:30 For personal use only. No other uses without permission. . All rights reserved.

NUCLEAR

ANG-II-BINDING

II), dithiothreitol (DTT), phenylmethylsulfonylfluoride (PMSF), 8-hydroxylquinonine, DNA, and fraction V BSA were obtained from Sigma (St. Louis, MO). The nonpeptitidic Ang-II receptor antagonist Losartan (DuP 753) (15), Ang-II-(l-7), Ang-II-(l-6), and Ang-II-(l-5) were provided by Ciba-Geigy (Basel, Switzerland). Losartan potassium is the adopted name for DuP 753 in the U.S. (Dr. R. Smith, DuPont Merck, Wilmington, DE). Sephadex G-50 (fine) and Ultropac TSK-G2000SW were purchased from Pharmacia (I’iscataway, NJ). Methods Preparation of rat liver nuclei. Isolation of rat liver nuclei was performed using a modification of the method of Surks et al. (16). Male SpragueDawley rats (120-180 g) were killed by cervical dislocation, and livers were removed rapidly, freed of connective tissue, and rinsed with icecold 0.32 M sucrose, 3 rnM MgC12, and 20 rnM Tris, pH 7.4. The following steps were performed at 4 C. The livers were homogenized in 3 vol of the above buffer using a glass homogenizer and a motor-driven Teflon pestle, filtered through a metal sieve (USA Standard Sieve, American Scientific Products, McGraw Park, IL; opening size, 53 Frn) and centrifuged for 10 min at 1000 X g. The pellet was resuspended by gentle homogenization with a Dounce homogenizer (Kontes Co., Vineland, NJ) in 2.2 M sucrose, 1 rnM MgC12, and 10 rnM Tris, pH 7.4, and centrifuged at 64,000 x g for 1 h. The nuclear pellet at bottom of the centrifuging tube was washed twice by resuspending in 0.32 M sucrose, 3 mM MgCls, and 20 rnM Tris, pH 7.4, and centrifuged at 1,000 x g for 10 min. Preparation of rat liver plasma membrane. The supernatant of 1,000 x centrifugation of the rat liver homogenate was centrifuged at 100,000 g for 1 h. The pellet was washed twice by resuspending in 0.32 sucrose, 3 rnM MgCl*, and 20 rnM Tris, pH 7.4, and centrifuging 100,000 x g for 1 h. The plasma membrane suspension was aliquoted, frozen in liquid nitrogen, and stored at -70 C.

g X M at

Preparation of rat liver cytoplasm.

The cytoplasm was collected from the supernatant fraction of the 1,000 X g supernatant of rat liver homogenate, which was centrifuged at 100,000 x g for 1 h. The cytoplasm was aliquoted, frozen in liquid nitrogen, and stored at -70 C.

Determination of protein and DNA.

Protein was assayed by the method of Bradford (17), using BSA as the standard. The measurement of DNA was performed according to the method of Burton (18).

Determination of 5’-nucleotidase. 5’.Nucleotidase,

a plasma membrane marker, was assayed according to the method of Goldfine et al. (19). Sample membrane or nuclei were incubated with 50 mM Tris (pH 8.5), 10 rnM MgC12, and 5 rnM AMP in a total volume of 1.2 ml at 37 C. After 15, 30, and 60 min, the reaction was terminated by the addition of an equal volume of ice-cold 10% trichloroacetic acid, and the mixture was centrifuged to remove the denatured proteins. The inorganic phosphate in the supernatant was then measured according to the methods of Ames (20). Supernatant (0.3 ml) was incubated with 0.6 ml 0.4% (wt/ vol) ammonium molybdate in 1 N H2S04 and 0.1 ml 10% (wt/vol) ascorbic acid at 37 C for 1 h. The optical density of the samples was assayed at 820 nm with a spectrophotometer.

Determination of rz51]Ang-ZI binding to rat liver nuclei. For saturation binding assay, nuclei (100 pg protein) were incubated at 25 C for 30 min with varying concentrations of ‘ZSI-Ang-II (1 to 10 nM) in a O.l-ml final volume of an assay buffer containing 50 mM Tris (pH 7.4), 100 mM NaCl, 10 mM MgCl,, 0.1% BSA (heat denatured), 1 mM PMSF, and 1 rnM 8-hydroxyquinoline, To determine nonspecific binding (-15-20% at the Kd), parallel incubations were made in the presence of excess unlabeled Ang-II (lo-* M). After incubation, bound and free radioactivity were separated by filtration through glass-fiber filters (Whatman GF/C, Clifton, NJ) presoaked with assay buffer. Each filter was rinsed three times with 5 ml ice-cold 0.9% NaCl, and radioactivity was counted in an LKB y-counter (Rockville, MD) with a counting efficiency of 73%. For competition binding assay, nuclei were incubated with a fixed concentration of 1 nM [‘251]Ang-II and varying concentrations of Ang-II, Ang-III, saralasin, Losartan (DuP 753), Ang-I-(1-7), Ang-I-(1-6), or Ang-I-(1-5).

pH optimum of [‘251]Ang-ll binding. Nuclei incubated

with

[‘251]Ang-II,

with

or without

or plasma cold Ang-II,

membranes were in a total volume

SITES

375

of 100 ~1 50 mM Tris buffer, titrated to various final pH values (6, 6.5, 7, 7.5, 8, and 9) at 25 C for 30 min. After incubation, bound and free radioactivity were separated by filtration through GF/C glass-fiber filters and determined as stated above. The pH was measured in the complete assay mixture and rechecked at the end of the incubation.

Solubilization

and Ang-II-binding

sites from rat liver nuclei

Suspensions of nuclei in 0.32 M sucrose, 3 mM MgCl,, and 20 mM Tris, pH 7.4, without the presence of detergent were frozen, thawed and vortexed six times, and centrifuged at 20,000 x g for 10 min at 4 C. The supernatant was collected as the nuclear extract. The nuclear extract was incubated with 2 nM [ ‘251]Ang-II in the presence or absence of 10m4 M unlabeled Ang-II in a total volume of 0.1 ml of a buffer containing 50 mM Tris (pH 7.4), 100 mM NaCl, 10 mM MgC12, 0.1% BSA, 1 mM PMSF, and 1 mM 8-hydroxyquinoline (binding assay buffer). After incubation at 25 C for 30 min, bound and free [‘251]Ang-II were separated on a lml bed volume Sephadex G-50 column, which was equilibrated and eluted with the binding assay buffer. Eluate (0.45 ml) was collected stepwise, and radioactivity was determined as described above.

Size-exclusion HPLC. The nuclear

extract was incubated with 2 nM [‘?I Ang-II in the absence or presence of 10e4 M unlabeled Ang-II in a total volume of 0.1 ml containing the binding assay buffer at 25 C for 30 min. The reaction mixtures were injected into a Ultropac TSK G2000SW (Tosoh, Nanyo, Japan; 7.5 x 300 mm) column and eluted with a buffer containing 0.4 M NaCl and 10 mM potassium phosphate, pH 7.4, at a flow rate of 1 ml/min. One-milliliter fractions were collected, and the radioactivity in each fraction was determined in a y-counter.

Results Purity

of

rat liver nuclei

Using a standard nuclei isolation technique, the nuclei settled at the bottom of the centrifugation tube containing 2.2 M sucrose. DNA was enriched in the nuclei preparation by 20-fold. The protein/DNA ratio of the nuclear preparation was 2.2 (data not shown), consistent with the range previously reported (11, 19). Phase contrast microscopic examination (X 100) revealed intact nuclei. No intact cells, vesicles, or other particulate materials were seen (Fig. 1). 5’-Nucleotidase activity, a plasma membrane marker of the nuclear preparation, was lessthan 1% of the liver whole homogenate (Table l), indicating that the preparations were free of significant membrane contamination. Time dependence

studies

Time dependence studies revealed that Ang-II binding to isolated rat liver nuclei was rapid. In Fig. 2 (top panel), specific binding of [‘251]Ang-II to the nuclei (100 PLgprotein) reached a maximum at 30 min and diminished gradually thereafter. However, specific binding of [‘251]Ang-II to plasma membrane was more rapid and reached a maximum at 10 min, thereafter diminishing gradually (Fig. 2, bottom panel). The difference in binding kinetics between nuclei and plasma membrane provides additional evidence that [‘251]Ang-II binds to nuclei rather than to a contamination deriving from the plasma membrane. Saturability

of

pz51]Ang-II

binding

To demonstrate that the binding of [‘*‘IlAng-II was of high affinity and limited capacity, rat liver nuclei were incubated with varying concentrations of [‘251]Ang-II. The data pre-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 February 2015. at 08:30 For personal use only. No other uses without permission. . All rights reserved.

NUCLEAR ANG-II-BINDING

376

SITES

Endo. Vol131.

1992 No 1

FIG. 1. Phase contrast micrograph. Rat liver nuclei were isolated, as described in Materials and Methods, and examined without fixation. Magnification, X100. This is representative of three separate experiments.

TABLE 1. Estimation of contamination of nuclei preparation by plasma membrane (based on activity of 5’-nucleotidase/pg DNA) 5’-Nucleotidase activity (nmol Pi/min)

DNA hd

NUCLEAR

I

SA (nmol Pi/min. pg DNA)

Nuclei 0.66 3.4 0.19 Homogenate 2.04 0.071 28.7 % Contamination = 0.19/28.7 x 100% = 0.7%. Data shown are the means of two separate experiments, performed in duplicate. Pi, Inorganic phosphate.

sented in Fig. 3 showed that these nuclei bound [1251]Ang-II in a saturable fashion, with a high affinity (Kd, 1.4 nM), and with limited binding capacity (10 fmol/mg protein). The binding capacity represented 460 binding sites/nucleus. The presenceof a single classof specific Ang-II-binding sites was suggestedby the linearity of the Scatchard plot. In contrast, 2 classesof Ang-II binding sites are present on the rat liver membrane preparation with Kd values ranging from 0.2-0.4 nM (high affinity) and 2.9-3.0 nM (low affinity), respectively (11, 21). Thus, the affinity for nuclear Ang-II-binding sites appeared to be different from the affinity for plasma membrane Ang-II receptors. Liand

specificity

Several fragments and analogs of Ang and a nonpeptidic Ang antagonist were tested for their ability to compete with the binding of [‘251]Ang-II to rat liver nuclei. These agents competed with [‘251]Ang-II for the binding siteswith a relative potency of Ang-III = Ang-II > Losartan (DuP 753) > saralasin > Ang-II-(l-7) > Ang-II-(l-6) > Ang-II-(l-5), as shown in Fig. 4. When tested for its ability to inhibit the binding of [‘251]Ang-II to the nuclei, the heptapeptide AngIII [Ang-11-(2-B)] was found to have the same potency as that of unlabeled Ang-II. However, another heptapeptide, Ang-II-(l-7), was a significantly lesspotent competitor and displayed inhibitory effects on binding only at high concentrations, suggesting that phenylalanine at position 8 of Ang-

0

10

30

lncubatlon

60

Time (min)

FIG. 2. Kinetics of [‘251]Ang-II binding to isolated liver nuclei and liver plasma membranes. Rat liver nuclei (A) or plasma membranes (B) were incubated with 1 nM [Y]Ang-II at 25 C. Incubations were started at time zero. Specific binding (0) was measured. Nonspecific binding (A) refers to binding in the presence of 10m4M unlabeled Ang-II. Each value was the mean of closely agreeing duplicate determinations, and the result shown is representative of three separate experiments.

II is important for binding to nuclear binding sites.Fragments of Ang-II-(l-7) could not effectively displace the [‘251]Ang-II binding. This progressive decrease in competitive binding activity with decreasing chain length of the amino-terminal Ang-II fragments has also been reported for membranebound Ang-II receptor in smooth muscle cells (23). Among the compounds tested for inhibitory effects on binding were

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 February 2015. at 08:30 For personal use only. No other uses without permission. . All rights reserved.

NUCLEAR

ANG-II-BINDING

SITES

377

000

500

-2 a 0 F 7: Lii & a, 2 Z 0

0

2

4

6

8

10

B(fmoles) I

I

I

I

I

I

I

I

I

I,

0

1

2

3

4

5

6

7

8

9

lANGIt

5

7

6

10

8

9

10

PH

(nM)

5. Effect of pH on [‘2”I]Ang-II binding to rat liver nuclei and plasma membranes. Rat liver nuclei (0) or plasma membranes (0) were incubated with 1 nM [‘*“I]Ang-II at 25 C in Tris buffers of various pH. The pH was measured in the complete assay mixture and rechecked at the end of the incubation. Specific binding was determined. Data shown are the means of three separate experiments, performed in duplicate. FIG.

FIG. 3. Saturation binding of [‘“51]Ang-II to rat liver nuclei. Rat liver nuclei were incubated at 25 C for 30 min with various concentrations of [iz51]Ang-II in the absence or presence of 10m4 M Ang-II to determine specific Ang-II binding. A Scatchard analysis of the data is shown in inset. Data shown are representative of two separate experiments, performed in duplicate. B/F, Bound/free ratio.

o L z

1004

s O\”

50-

0'

-12

-11

-10

-9

Competitor

-8

-7

-6

-5

(Log M)

two potent antagonists of Ang-II, namely, Losartan (DuP 753) and saralasin. Ang-II was at least 5 times more potent than Losartan (DuP 753), which was twice as potent as saralasin. Losartan (DuP 753) is highly selective for the AT1 subtype of Ang-II receptor that shows variations in its tissue distribution (24). Losartan (DuP 753) fully displaced [lz51] Ang-II from its nuclear binding sites. The same finding has been reported by Bauer et al. (25), who investigated rat liver plasma membrane and detected only the receptor subtype that highly selectively binds Losartan (DuP 753). binding

of

0

Membrone

I 0.1

r’“I]Ang-II

[‘251]Ang-II binding to rat liver nuclei displayed a fairly sharp pH optimum at pH 7.0, as shown in Fig. 5. With increasing alkalinity or acidity, the binding decreased. An equal amount of protein from the rat liver plasma membrane

I 1.0

mM

-4

FIG. 4. Competition-binding studies of [‘251]Ang-II binding to rat liver nuclei. Rat liver nuclei were incubated with 2 nM [‘*“I]Ang-II for 30 min at 25 C in the presence of varying concentrations of fragments and analogs of Ang-II as well as Losartan (DuP 753). [iz51]Ang-II binding in the presence of lo-’ M unlabeled Ang-II was subtracted from each value, and results were expressed as percentages of specific [iz51] Ang-II bound in the absence of any unlabeled competition: Ang-II (O), Ang-III (A), saralasin (O), Losartan (DuP 753; 0), Ang-W-7) (W), Ang-11-(1-B) (A), or Ang-B-(1-5) (x). Data shown are the means of two to five separate experiments, performed in duplicate.

Effect of pH on optimal

0 Plosmo

I 10.0

1c

DTT

6. Effects of DTT on [iz51]Ang-II binding to rat liver nuclei and plasma membranes. Rat liver nuclei (0) or plasma membranes (0) were incubated with 1 nM [9]Ang-II and varying concentrations of DTT in 50 mM Tris (pH 7.5), 0.15 M NaCl, 5 mM MgCl,, and 1 mg/ml BSA at 25 C. Specific binding was determined. Each point represents the mean of five separate experiments, performed in duplicate; variability is given as the SE bar. Statistical comparisons were made by paired t test (*, P < 0.05; **, P < 0.01). FIG.

preparation was employed to compare the pH optimum for binding of [‘251]Ang-II. Liver plasma membrane Ang-II binding exhibited an optimum at pH 8.0. The differential pH optimum of binding is consistent with the observation of Goldfine et al. (26) that the pH optimum for [‘251]insulin binding to nuclear binding sites was between 7.0-7.5, while the optimum for [‘251]insulinbinding to plasma membrane was 8.0. Effects of DTT

on pz51]Ang-II

binding

Several lines of evidence suggest that the inactivation of Ang-II receptors by DTT is due to the reduction of intramolecular disulfide bonds (23). At pH 7.4, DTT inhibited the binding of Ang-II to liver plasma membrane in a dosedependent fashion, with an ICsOof 10 mM (Fig. 6). Paradoxically, DTT at low concentrations (0.05-0.1 mM) increased binding to the rat liver nuclear preparation, while at higher concentrations (lo-100 mM) it reduced binding in a fashion similar to its effect on binding to plasma membrane. Similar results were seen at pH 7.0 and 8.0.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 February 2015. at 08:30 For personal use only. No other uses without permission. . All rights reserved.

378

NUCLEAR

Solubilization

of

nuclear Ang-II-binding

ANG-II-BINDING

sites

Fresh rat liver nuclei that were resuspended in the buffer containing 0.32 M sucrose,3 mM MgC&, and 20 mM Tris, pH 7.4, were frozen, thawed, and vortexed five times and centrifuged at 20,000 X g for 10 min at 4 C. Soluble Ang-11 receptors were assayedas described in Materials and Methods. In the absenceof detergent in the extraction and the receptorbinding buffer, specific [‘251]Ang-II-binding sites were recovered mostly (60%) in the supernatant (Table 2). The extracted pellet was subsequently extracted with 0.2% Triton X-100, 0.14 M NaCl, and 0.4 M NaCl in sequence. NaCl (0.4 M) extracted most proteins, but failed to obtain more nuclear Ang-II-binding sites from the extracted pellet. Nuclei and plasma membrane preparations were subjected to freezing and thawing five times to release soluble binding sites. A direct comparison of soluble Ang-II-binding sites in these subcellular organelles was performed by incubating the respective supernatant with [‘251]Ang-II, with or without cold Ang-II, in the absence of detergent, followed by separation on a Sephadex column. Basedon the assay and by Sephadex column separation in the absenceof detergent, more soluble Ang-II-binding sites (92.5 fmol/mg protein) were produced from the nuclei preparation than from the plasma membrane (10.6 fmol/mg protein) after freeze-thaw treatment (Table 3). Nuclei or plasma membranes were subjected to 0.2% Triton X-100 treatment. No significant soluble Ang-II-binding sitesfrom nuclei were extracted with 0.2% Triton, while 21.5 fmol/mg binding protein were released from plasma membrane. These results suggestthat nuclear Ang-II receptors are mainly of a soluble nature. In this property they are distinctly different from the plasma membrane receptor. In Table 3, few or no specific Ang-II-binding sites could be detected from rat liver cytoplasm, even when 387 pg cytosolic protein were assayed. This is at variance with the results of a recent study in which abundant cytosolic Ang-II-binding siteswere detected in rabbit liver (13). Mol wt of nuclear Ang-II-binding

sites

To estimate the mol wt of the nuclear Ang-II receptor, a nuclear extract obtained after freezing and thawing was incubated with 2 nM [‘251]Ang-II, with or without 10P4M unlabeled Ang-II, at 25 C for 30 min in the binding buffer. After incubation, the reaction mixtures were injected onto an Ultropac TSK G2000SW column and eluted at a flow rate of TABLE

2. Localization

Nuclei preparation Supernatant of freezethaw, 5X 0.2% Triton X-100 extract 0.14-M NaCl extract 0.4-M NaCl extract Residual pellet suspension

of Ang-II-binding

sites in rat liver

SITES

Endo. Voll31.

1992 No 1

1 ml/min. In Fig. 5, [‘251]Ang-II-binding site complexes appeared to be eluted as a major peak with a retention time of 8 min at the position corresponding to a mol wt of about 66 kilodaltons (kDa). The binding of [1251]Ang-IIwas blocked in the presence of 10m4M Ang-II. Thus, the mol wt of nuclear Ang-II receptor was estimated to be about 66 kDa. Discussion

Although it is believed that the action of Ang-II is mediated by transmembrane signal transduction mechanisms, evidence is accumulating which suggeststhat Ang-II may also have a direct nuclear effect. This nuclear action may account for some of the delayed, long term effects. Our results confirm the existence of specific Ang-II-binding sites in the nuclei of rat liver, described in a preliminary report (11). We demonstrated further that the specific Ang-II-binding sites in the nuclei have features distinct from those in the plasma membrane. Thus, these results establish the basis for future studies of the mechanismsof nuclear effect of Ang-II at the molecular level. In this study we adduced multiple lines of evidence for the existence of specific Ang-II-binding sites. 1) On the basis of 5’-nucleotidase activity, protein/DNA ratio, and phase contrast microscopy, our nuclei preparations were free of significant plasma membrane contamination. 2) Specific [‘25I] Ang-II binding to nuclei was shown to take place in a timedependent manner. Moreover, different binding kinetics of [‘*‘IlAng-II to nuclei and plasma membrane were observed. Binding to the plasma membrane reached equilibrium more rapidly than binding to the nuclei. If the nuclei preparation had been contaminated with plasma membrane, the kinetics of binding to nuclei should have been similar to plasma membrane binding kinetics. 3) The specific binding of [125I] Ang-II to nuclei was found to be a saturable phenomenon. Scatchard analysis yielded a single classof binding sites with an apparent affinity constant (Kd) of 1.4 nM and a limited binding capacity of 10 fmol/mg protein or 460 sites/nucleus. By contrast, there are 2 classesof Ang-II-binding siteson the rat liver membrane preparation: high affinity binding sites (Kd, 0.3 nM) and low affinity binding sites (Kd, 3.0 nM) (21, 22). 4) There are distinct pH optima for the binding of [‘25I] Ang-II to nuclei and plasma membranes. Interestingly, the pH optimum of [‘251]Ang-II binding is 7, which is comparable to that of [‘2”I]insulin binding to nuclei, whereas the optimal

nuclei

Protein (w/ml)

Vol (ml)

Total Protein (ma)

Specific binding (fmol/mg protein)

Total binding (fmol)

% Total binding

7.0 0.3

5.0 4.5

35.0 1.4

12.0 177.0

420 248

100 59

0.9 0.8 12.0 1.9

2.0 2.0 2.0 2.0

1.8 1.6 24.0 3.8

54.0 3.1 2.3 1.3

Nuclei preparation and extracts were prepared as described in Materials and Methods. Specific was determined by the filtration technique through glass-fiber filters, and those of the solubilized 50 column (l-ml bed volume). Each value is the mean of closely agreeing duplicate determinations, separate experiments.

97 5 55 5

23 1 13 1

Ang-II-binding activity of particulate fractions fractions by chromatography on a Sephadex Gand the result shown is representative of three

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 February 2015. at 08:30 For personal use only. No other uses without permission. . All rights reserved.

NUCLEAR

ANG-II-BINDING

TABLE 3. Comparison of Ang-II-binding sites in nuclei and plasma membranes, using freezing and thawing or Triton X-100 extraction, and detection of soluble Ang-II-binding sites in rat liver cytoplasm Soluble Ang-II-binding activity (fmol/mg protein) Supernatant freezing-thawing,

of 5X

Nuclei

Plasma membrane

92.5

10.6

Supernatant of 0.2% Triton X-100 extract

Below limit detection

CvtoDlasm ” ’

Plasma membrane

Nuclei of

21.5

0.4

Fresh nuclei, plasma membrane, and cytoplasm were prepared as described in Materials and Methods. Nuclei and plasma membrane preparations were frozen and thawed five times or extracted with 0.2% Triton X-100. Soluble Ang-II-binding sites were determined by chromatography on a Sephadex G-50 (l-ml bed volume), and binding experiments were performed as described in Materials and Methods. Data shown are the means of two separate experiments, performed in duplicate.

500000

I

0

5

10

15

20

25

30

35

40

Retention Time (mln) FIG. 7. Size-exclusion HPLC of soluble nuclear binding sites. Rat liver nuclear extract obtained after freezing and thawing was incubated with 2 nM [‘251]Ang-II in the absence or presence of lo-” M unlabeled AngII at 25 C for 30 min. At the end of incubation, the reaction mixtures were immediately injected into an Ultropac TSK G2000SW column and eluted at a flow rate of 1 ml/min. One-milliliter fractions were collected, and the radioactivity in each fraction was determined in a ycounter. Each value is the mean of closely agreeing duplicate determinations, and the result shown is representative of three separate experiments.

binding of [‘251]Ang-II to plasma membrane occurs at pH 8, which is comparable to the optimal pH for insulin binding to plasma membrane. 5) DTT, an agent capable of reducing disulfide bonds at a low concentration (0.05-0.1 mM) enhanced [‘251]Ang-II binding to nuclei, but had no significant effect on [‘251]Ang-II binding to plasma membrane. Different responses of Ang binding to DTT may reflect different structures of the Ang-binding sites or differences in their environments. Recently, 2 subtypes of Ang receptors were demonstrated in studies using specific ligands (24). Subtypes AT, and AT2 are not eqully affected by DTT (24). DTT inhibited the binding of Ang to rat aorta smooth muscle cells (AT,), whereas it increased the affinity of Ang-II binding to human uterus (AT,). However, the DTT concentrations needed to affect the two receptor subtypes mentioned are lower than that needed to influence the nuclear binding site, indicating structural differences in these entities or their microinvironments. 6) Nuclear Ang-II-binding sites are predominantly

SITES

379

soluble in nature, whereas plasma Ang-II-binding sites are membrane-bound proteins, and to a large extent, their solubilization requires detergent. The mol wt and ligand specificities of nuclear and plasma membrane binding sites are similar. Recently, the mol wt of liver plasma membrane Ang-II-binding protein has been determined by photoaffinity labeling, cross-linking, or affinity chromatography and found to be 66 kDa (27, 28). Using size-exclusion HPLC, we have estimated that the [12’I]AngII nuclear binding site complex has a mass of about 66 kDa, which is very similar to that of the plasma membrane (Fig. 7). The relative potencies of various ligands to compete with [‘251]Ang-II binding to the nuclei and plasma membrane are very similar: Ang-III = Ang-II > Losartan (DuP 753) > saralasin > Ang-II-( l-7) > Ang-II-( 1-6) > Ang-II-(l-5). This indicates that the Ang-binding sites in the nuclei preferentially bind Ang-II and Ang-III, but not the amino-terminal Ang fragments, such as Ang-II-(l-7), Ang-II-(l-6), and AngII-(l-5). Thus, the nuclear Ang-l-binding site is distinct from the Ang receptors that bind Ang-II-(l-7) located in the rat hypothalamo-neurohypophyseal system (29), since these researchers showed an activity of this heptapeptide in the submicromolar range, whereas its affinity for the nuclear sites observed in this investigation is at least 1 order of magnitude lower. However, in its affinity to Losartan (DuP 753), the nuclear Ang-I-binding site behaves like the receptor sites of rat liver plasma membranes (25) and, hence, cannot be distinguished from the latter using this pharmacological tool. The possibility that the nuclear binding site is a degradative enzyme for Ang-II is unlikely. The affinity of Ang-11 for the nuclear binding site (Kd = 1.4 nM) is 3 orders of magnitude greater than the corresponding affinity for degradative enzymes of Ang-II (Michaelis-Menten constant = -10 PM) (30). In addition, the specificity for the nuclear Ang-II-binding sites is much stricter than the substrate specificity for enzymes that metabolize Ang-II (31). Finally, we found that Ang-II was not metabolized by nuclei in the assay buffer containing PMSF and 8-hydroxyquinoline (data not shown). In spite of the fact that Ang-II has been shown to be taken up by cells (8, 9) and localized in the nucleus (lo), the origin of intracellular Ang that binds nuclear sites and the pathway(s) by which this Ang-II might reach the nucleus site in intact cells are unresolved. This raises the question of whether the saturable Ang-II-binding sites detected in the rat liver nuclei are of cytosolic origin. Recently, Rosenberg et al. (32) reported a fairly abundant Ang-II-binding protein in rat liver cytosol that requires p-chloromercurisulfonic acid as an essential component for its activity. To avoid contamination with the cytosolic binding protein, we omitted the use of p-chloromercurisulfonic acid and, accordingly, did not detect it under our experimental conditions. Moreover, the possibility that the nuclear binding sites may derive from the cytosolic binding protein described as being due to contamination by the latter can be ruled out, since the cytosolic Ang-II-binding sites have no affinity for Losartan (DuP 753) (14), which is in contrast to the high affinity of nuclear AngII-binding sites for Losartan (DuP 753). In sum, we have demonstrated soluble nuclear Ang-IIbinding sites that are distinct in their physicochemical prop-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 February 2015. at 08:30 For personal use only. No other uses without permission. . All rights reserved.

380

NUCLEAR

ANG-II-BINDING

erties from the membrane-bound receptor. However, at the moment the possibility cannot be excluded that these dissimilarities are due to differences in the environments of the binding sites, rather than to differences in the proteins themselves. Further work toward the elucidation of the structure of these sites is needed to clarify these two posibilities. However, nuclear Ang-II-binding protein may be important in mediating the intracellular action of Ang-II. The exact mechanism of the interaction between the Ang-11 nuclear binding site and these regulatory elements of Ang-responsive genes is not known. Future research will focus on the interaction between the Ang-binding site complex and specific DNA sequences. These studies may provide novel insights into the molecular action of Ang-II on gene expression.

to the chemists Drs. B. Kamber and F. Ostermayer help, and to Ms. Elaine Bubrzycki for typing the

13.

Griendling

14.

H+ exchange regulates cumulation in vascular 10624 2.

RW 1988 Evidence

that Na+/ angiotensin II-stimulated diacylglycerol acsmooth muscle cells. J Biol Chem 263:10620-

Baukal AJ, Balla T, Hunyady L, Hausdorff W, Gruillemette G, Catt KJ 1988 Angiotensin II and guanine nucleotides stimulate formation meabilized

of inositol 1,4,5-triphosphate and its metabolites in peradrenal glomerulosa cells. J Biol Chem 263:6087-6092 II induces 3. Geisterfer AAT, Peach MJ, Owens GK 1988 Angiotensin hypertrophy, not hyperplasia, of cultured rat aortic smooth muscle cells. Circ Res 62:749-756 4. Naftilan AJ, Pratt RE, Dzau VJ 1989 Induction of platelet-derived growth factor A-chain and c-myc gene expressions by angiotensin II in cultured rat vascular smooth muscle cells. J Clin Invest 83:14191424 5. Miller DS 1988 Stimulation of RNA and protein synthesis by intracullular insulin. Science 240:506-509 6. Koenig RJ, Brent GA, Warne RL, Larsen PR, Moore DD 1987 Thyroid hormone receptor binds to a site in the rat growth hormone promoter required for induction by thyroid hormone. Proc Nat1 Acad Sci USA 84:5670-5674 7. Peale FV, Ludwig LB, Zain S, Hilf R, Bambara RA 1988 Properties of a high-affinity DNA binding site for estrogen receptor. Proc Nat1 Acad Sci USA 85:1038-1042 8. Bianchi C, Gutkowska J, De Lean A, Ballak M, Anand-Srivastava MB, Genest J, Cantin M 1986 Fate of (1251) angiotensin II in adrenal zona glomerulosa cells. Endocrinology 118:2605-2607 9. Crozat A, Penhoat A, Saez JM 198k Processing of angiotensin II (A-II) and (Sar’,Ala’)A-II bv cultured bovine adrenocortical cells. Enddcrinoldgy 118:2312-23i8 10 Robertson AL, Khairallah PA 1971 Angiotensin II: rapid localization in nuclei of smooth and cardiac muscle. Science 172: 1138-l 139 11 Re RN, MacPhee AA, Fallon JT 1981 Specific nuclear binding of angiotensin II by rat liver and spleen nuclei. Clin Sci [Suppl 71 61:245s-247s

II-binding

J, Bryan SE 1984 Angiotensin II fragments generated by micrococcal muRes Commun 119:220-227 1989 Purification and properties of a soluble protein from rabbit liver. J Biol Chem

Bumpus FM, Catt KJ, Chiu AT, de Gasparo M, Goodfriend T, Husain A, Peach MJ, Taylor DG, Timmermans PBMWM 1991 Nomenclature ture committee Hypertension

for angiotensin receptors. A report of the nomenclaof the council for high blood pressure research. 17:720-721

15.

Chiu AT, Herblin WF, McCall DE, Ardecky RJ, Carini DJ, Duncia IV, Pease LJ, Wong PC, Wexler RR, Johnson AL, Timmermans PBMWM 1989 Identification of angiotensin II receptor subtypes.

16.

Surks MI, Koerner D, Dillman

Biochem

18. 19.

KK, Berk BC, Alexander

DC, Brown

Kiron MAR, Soffer RL angiotensin 264:4138-4142

References 1.

Re RN, Vizard

receptors in chromatin clease. Biochem Biophys

17.

Acknowledgments We are indebted for their generous manuscript.

12.

Endo. 1992 Voll31. No 1

SITES

20. 21.

22. 23.

Biophys

Res Commun

165:196-203

W, Oppenheimer JH 1973 Limited capacity binding sites for L-triiodothyronine in rat liver nuclei. Localization to the chromatin and partial characterization of the Ltriiodothyronine-chromatin complex. J Biol Chem 248:7066-7072 Bradford MM 1976 A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248-254 Burton K 1966 Isolation of certain oligonucleotides obtained by degradation of deoxyribonucleic acid. Biochem J 98:68-69 Goldfine ID, Smith GI, Wong KY, Jones AL 1977 Cellular uptake and nuclear binding of insulin in human cultured lymphocytes: evidence for potential intracellular sites of insulin action. Proc Nat1 Acad Sci USA 74:1368-1372 Ames BN 1966 Assay of inorganic phosphate, total phosphate and phosphatases. Methods Enzymol 8:115-l 18 Campanile Cl’, Crane JK, Peach MJ, Garrison JC 1982 The hepatic angiotensin II receptor. I. Characterization of the membrane-binding site and correlation with physiological response in hepatocytes. J Biol Chem 257:4951-4958 Gunther S 1984 Characterization of angiotensin 11 receptor subtypes in rat liver. J Biol Chem 259:7622-7629 Gunther

S, Alexander

RW, Atkinson

Functional angiotensin II receptors muscle cells. J Cell Biol 92:289-298 24.

25. 26.

27. 28.

29.

30. 31. 32.

WJ, Gimbrone

in cultured

vascular

MA 1982 smooth

Whitebread

S, Mele M, Kamber B, de Gasparo M 1989 Preliminary biochemical characterization of two angiotensin II receptor subtypes. Biochem Biophys Res Commun 163:284-291 Bauer PH, Chiu AT, Garrison JC 1991 DuP 753 can antagonize the effects of angiotensin II in rat liver. Mol Pharmacol 39:579-585 Goldfine ID, Purrello F, Vigneri R, Clawson GA 1985 Insulin and the regulation of isolated nuclei and nuclear subfractions: potential relationship to mRNA metabolism. Diabetes Metab Rev 1:119-137 Sen I, Bull HG, Soffer RL 1984 Isolation of an angiotensin IIbinding protein from liver. Pro Nat1 Acad Sci USA 81:1679-1683 Laribi C, Allard M, Vincent JD, Simonnet G 1987 Solubilization and characterization of covalently labeled angiotensin II receptors in cultured mouse spinal cord cells. Neuropeptides 9:345-356 Schiavone MT, Santos RAS, Brosnihan KB, Khosla MC, Ferrario CM 1988 Release of vasopressin from the rat hypothalamo-neurohvuouhvsial svstem bv anriotensin-(l-7) heutaueptide. Proc Nat1 i:ad’SA USA’85:409