- Email: [email protected]
Comparison of Binding Proteins of Glucocorticoids in Mammary Tissue and in Blood Sera from Lactating Cows1
Comparison of Binding Proteins of Glucocorticoids in Mammary Tissue and in Blood Sera from Lactating Cows I R O N A L D C. G O R E W I T 2 and H. A L L E N T U C K E R Animal Reproduction Laboratory Department of Dairy Science Michigan State University East Lansing 48824
INTRODUCTION
ABSTRACT
Binding of glucocorticoids has been described for thymus (7), liver (1), brain (6), and blood (10). In addition, investigators partially have characterized glucocorticoid binding to bovine (4, 14), mouse (11), vole (15), and rat (2, 3) mammary tissue and rat (3) and mouse (12) mammary tumors. Results of these studies suggested that tissue binders of glucocorticoids were proteins since they were destroyed by proteolytic enzymes, mercurials, and heating at 100 C for 10 rain. Furthermore, sedimentation coefficients of these binding proteins ranged from 4 to 8 S on sucrose gradients (13). The molecular weight of tissue binding proteins ranged from 50,000 to 250,000 depending upon experimental conditions. The following study was undertaken to determine if glucocortieoid receptor protein(s) in mammary tissue from lactating cattle (4) was unique and not a result of contamination from corticosteroid binding globulin (CBG) of blood origin (10).
Supernatant fractions (700 x g) isolated from homogenates of mammary tissue slices from three lactating Holstein cows (slices previously incubated with [hydrogen-3lcortisol for 1 h at 37 C) were electrophoresed on 7% polyacrylamide gels. The majority of radioactivity was in a protein(s) 2.5 to 3 cm from the origin whereas bovine serum incubated with [hydrogen-3]cortisol showed the majority of radioactivity in a protein 5 to 6 cm from the origin. N,N-diethylaminoethyl cellulose chromatography of 700 x g supernatant fluids from three lactating cows revealed that [hydrogen-3]cortisol was associated with a protein component that eluted with .3 M potassium phosphate, pH 8.0. In contrast, [hydrogen-3] cortisol bound to bovine sera eluted as two protein components with .05 and .1 M potassium phosphate, pH 8.0. Approximate molecular weights determined from gel filtration (four lactating cows) and sucrose density studies (two lactating cows) were estimated to be 2.5 x l 0 s to 3 × 106 for the 700 x g mammary receptor of cortisol and 6 x 104 to 8 × 104 for the primary protein which binds cortisol in serum. We conclude that lactating bovine mammary tissue contains a protein(s) capable of binding tritiated cortisol which is unique from the corticosteroid binding proteins of blood.
MATERIALS AND METHODS Routine Procedures
Received January 15, 1976. Published with the approval of the Director of the Michigan Agricultural Experiment Station as Journal Article No. 7542. This research was supported in part by USPHS Grant HD-05750. 2Department of Animal Science, Cornell University, Ithaca, NY 14853.
Tissue samples (approximately 50 g each) were removed from eight random sites in the mammary glands of 12 lactating Holstein cows within 15 to 30 rain after slaughter. Samples were transferred immediately to ice-cold .01 M Tris-EDTA buffer [.01 M Trizma base and .01 M ethylenediamine tetra-acetic acid (Sigma Chemical) adjusted to pH 7.4 with 6 N HC1]. Within 1 to 2 h, each of the eight tissue samples was cut further into approximately .2 to .4 mm 3 slices. Single slices from each of the original eight tissue samples were placed in tubes for use in individual experiments as described below. One cow was used on a given day. Whole blood from the same 12 cows as above was collected at slaughter, and sera were
1247
1248
GOREWIT AND TUCKER
harvested for later use. Unless otherwise specified, mammary tissue and blood sera were incubated for 1 h at 37 C in 3 ml of .01 M Tris-EDTA buffer (pH 7.4) containing 1.3 × 10-7 M [ 1 , 2 ) H ] cortisol (44 Ci/mmol, New England Nuclear). After incubation, mammary tissue slices were washed five times with 5 ml of Tris-EDTA buffer at 4 C and homogenized in 2 ml of Tris-EDTA buffer as previously described (4). Homogenates were centrifuged at 700 × g for 10 min at 4 C. Radioactivity was quantified in a NuclearChicago Mark I liquid scintillation spectrometer. Protein concentrations were monitored as a function of optical density at 280 nm. Gel Filtration
Approximate molecular weights of glucocorticoid binding components within 700 × g supernatant fractions of mammary tissue slices and blood sera were compared by gel filtration. Eight .2 to .4 mm 3 mammary slices from each of four lactating cows were incubated with [3 HI cortisol, washed, homogenized, and centrifuged as described in Routine Procedures. The lipid layer of the 700 × g supernatant fluid was aspirated and discarded. A .6-ml sample of lipid-free supernatant fluid was applied to a 1.5 × 25 cm column of Sephadex G-200 and eluted at 37 C with Tris-EDTA buffer at a flow rate of .25 ml/min with a hydrostatic pressure head of 6 to 10 cm. Three milliliter fractions were collected; protein concentrations and radioactivity were determined in each fraction. The elution volume for the protein component associated with [3 H] cortisol in 700 × g supernatant fluids was recorded. Three milliliters of sera were incubated with [3 H] cortisol under the same experimental conditions as for mammary tissue slices. After incubation, a .6-ml sample of serum was eluted on Sephadex G-200 as described above. Elution volumes of protein fractions of 700 × g mammary supernatants and bovine sera which bound [3 H] cortisol were compared with a series of standard protein solutions which had been chromatographed in the same manner. The standard protein solutions were Dextran 2000 (2 x 106 molecular weight); 2.0% ovalbumin (4.3 × 104 molecular weight); 2.0%/3-1actoglobulin (1.8 × 104 molecular weight); and 2.0% ribonuclease A (1.5 × 104 molecular Journal of Dairy Science Vol. 59, No. 7
weight). The 700 × g mammary supernatant and serum binding proteins were fitted on a standard curve based upon their respective elution volumes. Sucrose Density Gradient Analysis
Eight representative mammary tissue slices from each of two lactating cows were homogenized and centrifuged at 700 x g as described in Routine Procedures. Lipid material was aspirated and discarded, and the remaining supernatant fractions were centrifuged at 45,000 × g for 10 rain to remove mitochondria. Supernatant fluids then were centrifuged at 100,000 × g for 2 h to isolate cytosol fractions. We chose 2 h centrifugations at 100,000 x g to assure that supernatant fluids were particle-free. Three milliliters of the 100,000 x g cytosol fractions were incubated with 1.3 x 10 -7 M (final concentration) [3 H] cortisol for 1 h at 37 C. Five milliliters bovine sera from each of three lactating cows were centrifuged at 100,000 x g for 2 h. After centrifugation, 3 ml of sera supernatant were .incubated with [3HI cortisol as described for mammary cytosol. Mammary cytosol and sera (100,000 x g supernatant fluids) were adjusted to contain 25 mg protein/ml. Samples of .3 ml of 100,000 x g mammary cytosol or bovine sera, previously incubated with [3 H] cortisol, were layered on 4.5 ml gradients of 10 to 30% sucrose in Tris-EDTA buffer and were centrifuged at 4 C for 18 h at 100,000 × g in a Beckman Model L-2 ultracentrifuge. Bottoms of tubes were pierced after centrifugation, 32 fractions of .1-ml each were collected directly into scintillation vials, and radioactivity were determined. Approximate sedimentation coefficients were determined by centrifuging standards (Sigma Chemical) of yeast alcohol dehydrogenase (8.0 S), liver alcohol dehydrogenase (5.0 S), and catalase (11.0 S) under the same conditions as the 100,000 × g mammary cytosols and sera. Migration of standards was estimated from optical density determinations at 280 nm. Polyacrylamide Disc-Gel Electrophoresis
Eight mammary slices from each of three lactating cows were incubated with [3H]cor-
GLUCOCORTICOID
tisol, washed, homogenized, and centrifuged as described in Routine Procedures. Samples (50 /~1) of lipid-free 700 x g supernatants were electrophoresed on 7% polyacrylamide gels containing 10% glycerol (running gel pH 9.3; electrode buffer pH 8.6). Electrophoresis was at 2 mA/disc at 4 C for 4 or 5 h. Gels were stained 10 min for protein with 1.0% amido Schwartz and destained with several washes of 7.0% glacial acetic acid for 24 h. Gels were scanned at 460 nm for protein in a Gilford Spectrophotometer with linear transporter. Gels were frozen on dry ice, cut into .5 cm slices, dissolved in 1 ml of 30% H202 for 12 h at 56 C, and counted for [3 H] cortisol. Whole bovine sera from each of three lactating cows were incubated with [3H] cortisol, diluted 1:5 with .85% NaC1, and electrophoresed under the same experimental conditions as mammary 700 × g supernatant fluids.
1249
BINDING PROTEINS
Binding of Tritated Cortisol, Dexamethasone, and Progesterone to Bovine Sera
To determine relative magnitudes and orders of binding of tritiated steroids to bovine sera at 37 C, the following was done. Three tubes containing 3 ml each of bovine sera from each of three lactating cows were incubated with 10/Jl of either 2.7 × 10 -9 M [3H] cortisol (44 Ci/m mol), [3H] dexamethasone (28 Ci/m mol), or [3 H] progesterone (48 Ci/m mol) for 1 h at 37 C. After incubation, .6 ml of serum containing [3H] steroid were added to .4 ml of 40% sucrose, and this mixture was applied to Sephadex G-200 columns and eluted at 37 C with Tris-EDTA buffer. Three-milliliter fractions were collected, counted for radioactivity, and assayed for protein at 280 nm. Data were expressed as /~mole steroid bound per. mg protein. R ESU LTS Gel Filtration
N,N-diethylaminoethyl Cellulose (DEAE) Chromatography
Approximately 1 g of mammary tissue slices from each of three lactating cows was incubated for 1 h at 37 C in 2.0 ml of .01 M potassium phosphate buffer, pH 8.0, containing 1.3 × 10- 7 M [ 3 H] cortisol. After incubation, tissues were washed five times with 5 ml .01 M potassium phosphate buffer at 4 C, homogenized in 2 ml of this buffer, and centrifuged at 700 × g for 30 min. The 700 × g lipid-free supernatant fractions were eluted (desalted) with .01 M potassium phosphate buffer through a Sephadex G-25 column. Three milliliters of the eluate containing protein-bound [3HI cortisol were applied to DEAE cellulose columns (Whatman, DE-52, Whatman Co.) which had been pre-equilibrated with .01 M potassium phosphate buffer pH 8.0. Mammary cytoplasmic proteins were eluted by a stepwise .01 M to 2.0 M potassium phosphate gradient, pH 8.0. Ten milliliters of bovine sera from each of three cows were dialyzed against 100 ml of .01 M potassium phosphate buffer for 18 h then centrifuged at 2000 × g for 10 min. The supernatant fluid (15 ml in swelled dialysis tubing) were incubated with [3 H] cortisol; and 3 ml of this solution were applied to DEAE cellulose columns and eluted, as described above for mammary cytoplasmic proteins (16).
Approximate molecular weights of the major [3 H] cortisol binding proteins in 700 x g mammary supernatants and bovine sera averaged (+ SE) 3 + .86 × 106 and 6 + .94 × 10 a, respectively (Fig. 1). There was a second protein component in sera which bound [3 H] cortisol and eluted with the void volume. This component had a molecular weight of approximately 1 + .53 × 106. Radioactive cortisol associated with this component was
~ 700xg e • F ~',,,~Mammary Cortisol Binding Protein mw~3xlO L Dextron 2 0 0 0 ~ S e c o n d o r y Serum Cortisol Rmding Protein 6.0[- ~ row- I xlO Q 70
"~
Primary
J
Ovalbumm " " ~ - lact*globulin
; 5.
'%lO....r
Rib
....
....
....
¢o ....
es ....
....
;0 ....
z'"
Volumeof Effluent (rot) FIG. t. Gel filtration (Sephadex G-200) elution
volumes of 700 X g mammary proteins binding cortisol, serum proteins binding cortisol, and various standard proteins of known molecular weight. J o u r n a l o f D a i r y S c i e n c e Vol. 59, N o . 7
1250
GOREWIT AND TUCKER
'°~ogl: ,'~o° ~
,ooo[
/ ,r / /l
/
I 100~
jj
x x x'lr
-
.c
6
...=.oJl,
~. ,Ao(~os)\Z,~ ,-,r'k i .oo[ / cArA..s~.,o= ¢ // /] II
,ooL
1",, "~ \
'°°V/ i .... 2
4
6
Bottom
8
,/"
x
I
.%.:' .t ....
I
I
"t
Polyacrylamide Disc-Gel Electrophoresis
Electrophoretic properties of glucocorticoid-binding proteins of mammary tissue and blood sera were compared by polyacrylamide disc-gel electrophoresis. The majority of radioactivity in the 700 × g mammary supernatants after electrophoresis was associated with protein(s) which migrated 2.5 to 3 cm from the gel origin (Fig. 3A). In contrast [aH] cortisol was
7Q
A
I0 12 14 16 18 20 22 24 26 2 8 ~ 3 2 FRACTION NUMBER
Top
FIG. 2. Sucrose density gradient patterns of 100,000 × g mammary proteins binding cortisol, serum proteins binding cortisol, and various standard proteins of known molecular weights. MCBC = [aHlcortisol bound to mammary cortisol binding component. SBC = [3H]cordsol bound to serum binding component. Milligrams of catalase, yeast alcohol dehydrogenase (YAD), and liver alcohol dehydrogenase (LAD) standard proteins in each fraction were plotted for comparison. The MCBC and SBC curves were the average of two and three cows, respectively. Catalase, YAD, and LAD curves were the average of two experiments. Counting efficiency was 50%.
700 X G MAMMARY SUPERNATANT
Z 25
5£
cPM ~
0
4( ~0
0
ZO
.?5
"', / , :
I0
\
,I
', I0
210
]iO
,oo ---,
410 5~0 60 Get LenQlh (cm)
B
]I 70
,,,,., 80
l-l'"
3 O0
I
9o
only 10% of that associated with the 6 × 104 molecular weight serum binding component.
8o
i
BOVINESERUM I 2 25
Sucrose Density Gradient Analysis
To determine more precisely molecular weights of glucocorticoid binding proteins, sucrose density gradient analysis of 100,000 × g supernatant fluids from mammary tissue and proteins of sera was used. The mammary cytosol binding component (MCBC) which bound [aH] cortisol sedimented in the 11.0 S region of sucrose gradients (Fig. 2). This suggested that the molecular weight of the mammary receptor for glucocorticoids in 100,000 x g supernatant fluids was 250,000 or greater since it sedimented in the region of catalase which has a molecular weight of 250,000. In contrast, the serum binding component (SBC) which bound [3HI cortisol sedimented in the 5.0 S region of the sucrose gradient. This suggested that the molecular weight of the SBC was approximately 80,000 since it sedimented in the region of liver alcohol dehydrogenase (LAD) which has a molecular weight of 83,000. Journal of Dairy Science Vol. 59, No. 7
70
CPM
,0
;".,
I
30
'~%
20
! I
~
8
* I
/
I
/
",-/
~C
I '
/!
0 D ot 4 6 0 nM I I0
i ~'0
i
~10 , ; 0
I ~i0 ' 610 '%%70--~r'~~17
IO
Gel Length (cm) FIG. 3. A. Polyacrylamide disc-gel electrophoretic profile of [3H] cortisol bound to protein in 700 X g supernatant fractions of mammary tissue slices from lactating cows. B. Polyactylamide disc-gel electrophoretic profile of [ 3 l-.l cortisol bound to proteins of bovine sera. Counts per min (CPM) profiles in Fig. 3A and 3B each represent the average o f three cows, whereas optical density (OD) in both figures is representative of one experiment each. Counting efficiency was 50%.
GLUCOCORTICOID BINDING PROTEINS
12 51
.30 A
700 x g MAMMARY SUPERNATANT OEAE CELLULOSE
400
300
c~
.ISO
200 .075
I00 3
5 7
9 II 13 IS 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 FRACTION NUMBER
,01M
I
.05 M
|
'.1M
I
.3 M
I
20 M
I
M PO 4
8000 7000 6000 5000 4000 3000
.30 ~
~
BOVINE SERUM OEAE CELLULOSE
B
k~,~PM
2000
E
;i ~= 15oo
8
)°
4 I
6
.OIM
O
I0 12 I
14 .05M
I
18 2 0 I
4 2B 30 32 FRACTION NUMBER .IM M PO 4
I
5 4 3 6 38 4 0 4 2 4 4 4 6 4B ,3M
I
2.0M
50 I
FIG. 4. DEAE cellulose elution profile of [3H] cortisol b o u n d to 700 × g s u p e r n a t a n t of m a m m a r y tissue slices from lactating cows. B. DEAE cellulose elution profile of [3 H] cortisol b o u n d to bovine sera. Counts per min (CPM) profiles in Fig. 4A and 4B each represent the average o f three cows whereas optical density (OD) in b o t h figures is representative o f one e x p e r i m e n t each. Counting efficiency was 50%. Journal of Dairy Science Vol. 59, No. 7
1252
GOREWIT AND TUCKER
associated with serum protein(s) which migrated 5 to 6 cm from the origin (Fig. 3B). N,N-diethylaminoethyl Cellulose (DEAE) Chromatography
Electrochemical properties of glucocorticoid-binding proteins of mammary tissue and blood sera were determined by DEAE chromatography. The [3Hlcortisol bound to mammary protein(s) in the presence of .01 M potassium phosphate buffer was eluted subsequently from 700 × g supernatant fluids with .3 M potassium phosphate (Fig. 4A). On the other hand, [3 H] cortisol bound to bovine serum protein(s) under similar conditions eluted with .05 and .1 M potassium phosphate (Fig. 4B). Binding of Tritiated Cortisol, Dexamethasone, and Progesterone to Bovine Serum
In experiments designed to determine the relative magnitude and order of binding of tritiated steroids to bovine sera at 37 C, sera bound 3.23 x 10-13 /lmol of [3 H] cortisol, 2.19 × 10-13 /.tmol of [3 H] progesterone, and .0599 x 10-13 /amol of [3H] dexamethasone per mg of protein.
DISCUSSION
Results suggested that the glucocorticoidbinding component of mammary tissue was not a contaminant from blood. For example, gel filtration chromatography and sucrose gradient analysis of mammary supernatant fractions showed that the protein(s) binding cortisol was a macromolecule (~--11.0 S) with a molecular weight of approximately 2.5 × l 0 s to 3 x 106 (Fig. 1 and 2) whereas the major protein which bound cortisol in serum was a macromolecule (~5.0 S) with an approximate molecular weight of 6 × 10 a to 8 × 10 a (Fig. 1 and 2). These characteristics of the major protein binding component for glucocorticoids in blood sera of cattle were similar to those for CBG in blood of other species (10). Radioactive cortisol which associated with the 1 × 106 molecular weight component of sera (Fig. 1) may represent [3H] cortisol binding to immune globulins or other "high" molecular weight serum proteins. The Sephadex G-200 elution patterns of the mammary component binding glucocorticoid Journal of Dairy Science Vol. 59, No. 7
were similar to those described for "specific" cortisol binding protein(s) in mammary tissue by Gorewit and Tucker (4). Moreover, the chromatographic properties of bovine serum components which bound glucocorticoids were similar to those reported for blood of humans by Seal and Doe (10). Disc gel electrophoresis and DEAE cellulose chromatography confirmed that glucocorticoid binding protein(s) of bovine mammary tissue had different electrochemical characteristics than bovine serum binding components. Similar electrophoretic patterns for glucocorticoid receptor proteins in rat brain tissue and rat blood sera have been reported by McEwen (6). Experiments in our laboratory (4) also suggested that the specific corticoid binding component of mammary tissue from lactating cows was not a result of tissue contamination, with CBG of blood. For example, tritiated cortisol or tritiated dexamethasone binding to mammary tissue could not be inhibited competitively with progesterone at 37 C (4). In contrast, CBG has a higher affinity for progesterone at 37 C than for corticoids (9, 12). Moreover, dexamethasone is a moderate to poor competitive inhibitor of cortisol binding to CBG (1, 5, 8), but dexamethasone proved to be a strong competitive inhibitor of cortisol binding in the bovine mammary slice system (4). In the present study, blood sera from lactating cows bound more cortisol and progesterone at 37 C as compared with dexamethasone, confirming further that the glucocorticoid receptor in mammary tissue was not CBG. The biochemical properties of the mammary glucocorticoid-binding protein(s) (i.e., their behavior upon gel filtration and electrophoresis) suggested that it was a large molecule. Several workers have shown similar results for glucocorticoid binding proteins in other tissues (13). Most of these macromolecutar binding proteins were composed of subunits. It is reasonable to speculate that the mammary glucocorticoid binding protein described in this paper is composed also of subunits. This molecular structure may explain the binding protein's chromatographic and electrophoretic properties. REFERENCES
1 Baxter, J. E., and G. M. Tomkins. 1971. Glucocorticoid hormone receptors. Page 331 in Advances in
GLUCOCORTICOID BINDING PROTEINS the biosciences. Vol. 7. G. Raspe, ed. Pergamon Press, Vieweg. 2 Gardner, D. G., and J. L. Witliff. 1973. Characterization of a distinct glucocorticoid-binding protein in the lactating mammary gland of the rat. Bioclaim. Biophys. Acta 320:617. 3 Gardner, D. G., and J. L. Witliff. 1973. Demonstration of a glucoeorticoid hormone-receptor complex in the cytoplasm of a hormone responsive tumour. Br. J. Cancer 27:441. 4 Gorewit, R. C., and H. A. Tucker. 1976. Corticoid binding in mammary tissue slices from lactating cows. J. Dairy Sci. 59:232. 5 Kolanowski, J., and M. A. Pizarro. 1969. Critical evaluation of competitive protein-binding radioassay for cortisol. Ann. Endocrinol (Paris) 30:177. 6 McEwen, B. S., C. Magnus, and G. Wallach. 1972. Soluble corticosterone binding macromolecule from rat brain. Endocrinology 90:217. 7 Munck, A., and C. Wira. 1971. Glucocorticoid receptors in rat thymus cells. Page 301 in Advances in the biosciences. Vol. 7. G. Raspe, ed. Pergamon Press, Vieweg. 8 Peets, E. A., M. Staub, and S. Symchowicz. 1969. Plasma binding of betamethasone-3 H, dexamethasone-3 H, and cortiso1-14 C--A comparative study. Biochem. Pharmacol. 18:1655.
12 53
9 Sandberg, A. A., H. Rosenthal, and S. L. Schneider. 1966. Protein steroid interactions and their role in the transport and metabolism of steroids. Page 1 in Steroid dynamics. G. Pincus, J. F. Tait and T. Nakao, ed. Academic Press, New York. 10 Seal, U. S., and R. P. Doe. 1966. Corticosteroid binding globulin: Biochemistry, physiology and phylogeny. Page 63 in Steroid dynamics. G. PincuS, J. F. Tait, and T. Nakao, ed. Academic Press, New York. 11 Shyamala, G. 1973. Specific cytoplasmic glueocorticoid hormone receptors in lactating mammary glands. Biochemistry 12: 3085. 12 Shyamala, G. 1974. Glucocortieoid receptors in mouse mammary tumours. Specific binding of glucocortieoids in the cytoplasm. J. Biol. Chem. 249:2160. 13 Thomas, P. J. 1973. Steroid hormones and their receptors. J. Endocrinol. 57: 333. 14 Tucker, H. A., B. L. Larson, and J. Gorski. 1971. Cortisol binding in cultured bovine mammary cells. Endocrinology 89:152. 15 Turnell, R. W., P. C. Beers, andJ. L. Wifliff. 1974. Glucocorticoid-binding macromoleeules in the lactating mammary gland of the vole. Endocrinology 95:1770.
Journal of Dairy Science Vol. 59, No. 7
INTRODUCTION
ABSTRACT
Binding of glucocorticoids has been described for thymus (7), liver (1), brain (6), and blood (10). In addition, investigators partially have characterized glucocorticoid binding to bovine (4, 14), mouse (11), vole (15), and rat (2, 3) mammary tissue and rat (3) and mouse (12) mammary tumors. Results of these studies suggested that tissue binders of glucocorticoids were proteins since they were destroyed by proteolytic enzymes, mercurials, and heating at 100 C for 10 rain. Furthermore, sedimentation coefficients of these binding proteins ranged from 4 to 8 S on sucrose gradients (13). The molecular weight of tissue binding proteins ranged from 50,000 to 250,000 depending upon experimental conditions. The following study was undertaken to determine if glucocortieoid receptor protein(s) in mammary tissue from lactating cattle (4) was unique and not a result of contamination from corticosteroid binding globulin (CBG) of blood origin (10).
Supernatant fractions (700 x g) isolated from homogenates of mammary tissue slices from three lactating Holstein cows (slices previously incubated with [hydrogen-3lcortisol for 1 h at 37 C) were electrophoresed on 7% polyacrylamide gels. The majority of radioactivity was in a protein(s) 2.5 to 3 cm from the origin whereas bovine serum incubated with [hydrogen-3]cortisol showed the majority of radioactivity in a protein 5 to 6 cm from the origin. N,N-diethylaminoethyl cellulose chromatography of 700 x g supernatant fluids from three lactating cows revealed that [hydrogen-3]cortisol was associated with a protein component that eluted with .3 M potassium phosphate, pH 8.0. In contrast, [hydrogen-3] cortisol bound to bovine sera eluted as two protein components with .05 and .1 M potassium phosphate, pH 8.0. Approximate molecular weights determined from gel filtration (four lactating cows) and sucrose density studies (two lactating cows) were estimated to be 2.5 x l 0 s to 3 × 106 for the 700 x g mammary receptor of cortisol and 6 x 104 to 8 × 104 for the primary protein which binds cortisol in serum. We conclude that lactating bovine mammary tissue contains a protein(s) capable of binding tritiated cortisol which is unique from the corticosteroid binding proteins of blood.
MATERIALS AND METHODS Routine Procedures
Received January 15, 1976. Published with the approval of the Director of the Michigan Agricultural Experiment Station as Journal Article No. 7542. This research was supported in part by USPHS Grant HD-05750. 2Department of Animal Science, Cornell University, Ithaca, NY 14853.
Tissue samples (approximately 50 g each) were removed from eight random sites in the mammary glands of 12 lactating Holstein cows within 15 to 30 rain after slaughter. Samples were transferred immediately to ice-cold .01 M Tris-EDTA buffer [.01 M Trizma base and .01 M ethylenediamine tetra-acetic acid (Sigma Chemical) adjusted to pH 7.4 with 6 N HC1]. Within 1 to 2 h, each of the eight tissue samples was cut further into approximately .2 to .4 mm 3 slices. Single slices from each of the original eight tissue samples were placed in tubes for use in individual experiments as described below. One cow was used on a given day. Whole blood from the same 12 cows as above was collected at slaughter, and sera were
1247
1248
GOREWIT AND TUCKER
harvested for later use. Unless otherwise specified, mammary tissue and blood sera were incubated for 1 h at 37 C in 3 ml of .01 M Tris-EDTA buffer (pH 7.4) containing 1.3 × 10-7 M [ 1 , 2 ) H ] cortisol (44 Ci/mmol, New England Nuclear). After incubation, mammary tissue slices were washed five times with 5 ml of Tris-EDTA buffer at 4 C and homogenized in 2 ml of Tris-EDTA buffer as previously described (4). Homogenates were centrifuged at 700 × g for 10 min at 4 C. Radioactivity was quantified in a NuclearChicago Mark I liquid scintillation spectrometer. Protein concentrations were monitored as a function of optical density at 280 nm. Gel Filtration
Approximate molecular weights of glucocorticoid binding components within 700 × g supernatant fractions of mammary tissue slices and blood sera were compared by gel filtration. Eight .2 to .4 mm 3 mammary slices from each of four lactating cows were incubated with [3 HI cortisol, washed, homogenized, and centrifuged as described in Routine Procedures. The lipid layer of the 700 × g supernatant fluid was aspirated and discarded. A .6-ml sample of lipid-free supernatant fluid was applied to a 1.5 × 25 cm column of Sephadex G-200 and eluted at 37 C with Tris-EDTA buffer at a flow rate of .25 ml/min with a hydrostatic pressure head of 6 to 10 cm. Three milliliter fractions were collected; protein concentrations and radioactivity were determined in each fraction. The elution volume for the protein component associated with [3 H] cortisol in 700 × g supernatant fluids was recorded. Three milliliters of sera were incubated with [3 H] cortisol under the same experimental conditions as for mammary tissue slices. After incubation, a .6-ml sample of serum was eluted on Sephadex G-200 as described above. Elution volumes of protein fractions of 700 × g mammary supernatants and bovine sera which bound [3 H] cortisol were compared with a series of standard protein solutions which had been chromatographed in the same manner. The standard protein solutions were Dextran 2000 (2 x 106 molecular weight); 2.0% ovalbumin (4.3 × 104 molecular weight); 2.0%/3-1actoglobulin (1.8 × 104 molecular weight); and 2.0% ribonuclease A (1.5 × 104 molecular Journal of Dairy Science Vol. 59, No. 7
weight). The 700 × g mammary supernatant and serum binding proteins were fitted on a standard curve based upon their respective elution volumes. Sucrose Density Gradient Analysis
Eight representative mammary tissue slices from each of two lactating cows were homogenized and centrifuged at 700 x g as described in Routine Procedures. Lipid material was aspirated and discarded, and the remaining supernatant fractions were centrifuged at 45,000 × g for 10 rain to remove mitochondria. Supernatant fluids then were centrifuged at 100,000 × g for 2 h to isolate cytosol fractions. We chose 2 h centrifugations at 100,000 x g to assure that supernatant fluids were particle-free. Three milliliters of the 100,000 x g cytosol fractions were incubated with 1.3 x 10 -7 M (final concentration) [3 H] cortisol for 1 h at 37 C. Five milliliters bovine sera from each of three lactating cows were centrifuged at 100,000 x g for 2 h. After centrifugation, 3 ml of sera supernatant were .incubated with [3HI cortisol as described for mammary cytosol. Mammary cytosol and sera (100,000 x g supernatant fluids) were adjusted to contain 25 mg protein/ml. Samples of .3 ml of 100,000 x g mammary cytosol or bovine sera, previously incubated with [3 H] cortisol, were layered on 4.5 ml gradients of 10 to 30% sucrose in Tris-EDTA buffer and were centrifuged at 4 C for 18 h at 100,000 × g in a Beckman Model L-2 ultracentrifuge. Bottoms of tubes were pierced after centrifugation, 32 fractions of .1-ml each were collected directly into scintillation vials, and radioactivity were determined. Approximate sedimentation coefficients were determined by centrifuging standards (Sigma Chemical) of yeast alcohol dehydrogenase (8.0 S), liver alcohol dehydrogenase (5.0 S), and catalase (11.0 S) under the same conditions as the 100,000 × g mammary cytosols and sera. Migration of standards was estimated from optical density determinations at 280 nm. Polyacrylamide Disc-Gel Electrophoresis
Eight mammary slices from each of three lactating cows were incubated with [3H]cor-
GLUCOCORTICOID
tisol, washed, homogenized, and centrifuged as described in Routine Procedures. Samples (50 /~1) of lipid-free 700 x g supernatants were electrophoresed on 7% polyacrylamide gels containing 10% glycerol (running gel pH 9.3; electrode buffer pH 8.6). Electrophoresis was at 2 mA/disc at 4 C for 4 or 5 h. Gels were stained 10 min for protein with 1.0% amido Schwartz and destained with several washes of 7.0% glacial acetic acid for 24 h. Gels were scanned at 460 nm for protein in a Gilford Spectrophotometer with linear transporter. Gels were frozen on dry ice, cut into .5 cm slices, dissolved in 1 ml of 30% H202 for 12 h at 56 C, and counted for [3 H] cortisol. Whole bovine sera from each of three lactating cows were incubated with [3H] cortisol, diluted 1:5 with .85% NaC1, and electrophoresed under the same experimental conditions as mammary 700 × g supernatant fluids.
1249
BINDING PROTEINS
Binding of Tritated Cortisol, Dexamethasone, and Progesterone to Bovine Sera
To determine relative magnitudes and orders of binding of tritiated steroids to bovine sera at 37 C, the following was done. Three tubes containing 3 ml each of bovine sera from each of three lactating cows were incubated with 10/Jl of either 2.7 × 10 -9 M [3H] cortisol (44 Ci/m mol), [3H] dexamethasone (28 Ci/m mol), or [3 H] progesterone (48 Ci/m mol) for 1 h at 37 C. After incubation, .6 ml of serum containing [3H] steroid were added to .4 ml of 40% sucrose, and this mixture was applied to Sephadex G-200 columns and eluted at 37 C with Tris-EDTA buffer. Three-milliliter fractions were collected, counted for radioactivity, and assayed for protein at 280 nm. Data were expressed as /~mole steroid bound per. mg protein. R ESU LTS Gel Filtration
N,N-diethylaminoethyl Cellulose (DEAE) Chromatography
Approximately 1 g of mammary tissue slices from each of three lactating cows was incubated for 1 h at 37 C in 2.0 ml of .01 M potassium phosphate buffer, pH 8.0, containing 1.3 × 10- 7 M [ 3 H] cortisol. After incubation, tissues were washed five times with 5 ml .01 M potassium phosphate buffer at 4 C, homogenized in 2 ml of this buffer, and centrifuged at 700 × g for 30 min. The 700 × g lipid-free supernatant fractions were eluted (desalted) with .01 M potassium phosphate buffer through a Sephadex G-25 column. Three milliliters of the eluate containing protein-bound [3HI cortisol were applied to DEAE cellulose columns (Whatman, DE-52, Whatman Co.) which had been pre-equilibrated with .01 M potassium phosphate buffer pH 8.0. Mammary cytoplasmic proteins were eluted by a stepwise .01 M to 2.0 M potassium phosphate gradient, pH 8.0. Ten milliliters of bovine sera from each of three cows were dialyzed against 100 ml of .01 M potassium phosphate buffer for 18 h then centrifuged at 2000 × g for 10 min. The supernatant fluid (15 ml in swelled dialysis tubing) were incubated with [3 H] cortisol; and 3 ml of this solution were applied to DEAE cellulose columns and eluted, as described above for mammary cytoplasmic proteins (16).
Approximate molecular weights of the major [3 H] cortisol binding proteins in 700 x g mammary supernatants and bovine sera averaged (+ SE) 3 + .86 × 106 and 6 + .94 × 10 a, respectively (Fig. 1). There was a second protein component in sera which bound [3 H] cortisol and eluted with the void volume. This component had a molecular weight of approximately 1 + .53 × 106. Radioactive cortisol associated with this component was
~ 700xg e • F ~',,,~Mammary Cortisol Binding Protein mw~3xlO L Dextron 2 0 0 0 ~ S e c o n d o r y Serum Cortisol Rmding Protein 6.0[- ~ row- I xlO Q 70
"~
Primary
J
Ovalbumm " " ~ - lact*globulin
; 5.
'%lO....r
Rib
....
....
....
¢o ....
es ....
....
;0 ....
z'"
Volumeof Effluent (rot) FIG. t. Gel filtration (Sephadex G-200) elution
volumes of 700 X g mammary proteins binding cortisol, serum proteins binding cortisol, and various standard proteins of known molecular weight. J o u r n a l o f D a i r y S c i e n c e Vol. 59, N o . 7
1250
GOREWIT AND TUCKER
'°~ogl: ,'~o° ~
,ooo[
/ ,r / /l
/
I 100~
jj
x x x'lr
-
.c
6
...=.oJl,
~. ,Ao(~os)\Z,~ ,-,r'k i .oo[ / cArA..s~.,o= ¢ // /] II
,ooL
1",, "~ \
'°°V/ i .... 2
4
6
Bottom
8
,/"
x
I
.%.:' .t ....
I
I
"t
Polyacrylamide Disc-Gel Electrophoresis
Electrophoretic properties of glucocorticoid-binding proteins of mammary tissue and blood sera were compared by polyacrylamide disc-gel electrophoresis. The majority of radioactivity in the 700 × g mammary supernatants after electrophoresis was associated with protein(s) which migrated 2.5 to 3 cm from the gel origin (Fig. 3A). In contrast [aH] cortisol was
7Q
A
I0 12 14 16 18 20 22 24 26 2 8 ~ 3 2 FRACTION NUMBER
Top
FIG. 2. Sucrose density gradient patterns of 100,000 × g mammary proteins binding cortisol, serum proteins binding cortisol, and various standard proteins of known molecular weights. MCBC = [aHlcortisol bound to mammary cortisol binding component. SBC = [3H]cordsol bound to serum binding component. Milligrams of catalase, yeast alcohol dehydrogenase (YAD), and liver alcohol dehydrogenase (LAD) standard proteins in each fraction were plotted for comparison. The MCBC and SBC curves were the average of two and three cows, respectively. Catalase, YAD, and LAD curves were the average of two experiments. Counting efficiency was 50%.
700 X G MAMMARY SUPERNATANT
Z 25
5£
cPM ~
0
4( ~0
0
ZO
.?5
"', / , :
I0
\
,I
', I0
210
]iO
,oo ---,
410 5~0 60 Get LenQlh (cm)
B
]I 70
,,,,., 80
l-l'"
3 O0
I
9o
only 10% of that associated with the 6 × 104 molecular weight serum binding component.
8o
i
BOVINESERUM I 2 25
Sucrose Density Gradient Analysis
To determine more precisely molecular weights of glucocorticoid binding proteins, sucrose density gradient analysis of 100,000 × g supernatant fluids from mammary tissue and proteins of sera was used. The mammary cytosol binding component (MCBC) which bound [aH] cortisol sedimented in the 11.0 S region of sucrose gradients (Fig. 2). This suggested that the molecular weight of the mammary receptor for glucocorticoids in 100,000 x g supernatant fluids was 250,000 or greater since it sedimented in the region of catalase which has a molecular weight of 250,000. In contrast, the serum binding component (SBC) which bound [3HI cortisol sedimented in the 5.0 S region of the sucrose gradient. This suggested that the molecular weight of the SBC was approximately 80,000 since it sedimented in the region of liver alcohol dehydrogenase (LAD) which has a molecular weight of 83,000. Journal of Dairy Science Vol. 59, No. 7
70
CPM
,0
;".,
I
30
'~%
20
! I
~
8
* I
/
I
/
",-/
~C
I '
/!
0 D ot 4 6 0 nM I I0
i ~'0
i
~10 , ; 0
I ~i0 ' 610 '%%70--~r'~~17
IO
Gel Length (cm) FIG. 3. A. Polyacrylamide disc-gel electrophoretic profile of [3H] cortisol bound to protein in 700 X g supernatant fractions of mammary tissue slices from lactating cows. B. Polyactylamide disc-gel electrophoretic profile of [ 3 l-.l cortisol bound to proteins of bovine sera. Counts per min (CPM) profiles in Fig. 3A and 3B each represent the average o f three cows, whereas optical density (OD) in both figures is representative of one experiment each. Counting efficiency was 50%.
GLUCOCORTICOID BINDING PROTEINS
12 51
.30 A
700 x g MAMMARY SUPERNATANT OEAE CELLULOSE
400
300
c~
.ISO
200 .075
I00 3
5 7
9 II 13 IS 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 FRACTION NUMBER
,01M
I
.05 M
|
'.1M
I
.3 M
I
20 M
I
M PO 4
8000 7000 6000 5000 4000 3000
.30 ~
~
BOVINE SERUM OEAE CELLULOSE
B
k~,~PM
2000
E
;i ~= 15oo
8
)°
4 I
6
.OIM
O
I0 12 I
14 .05M
I
18 2 0 I
4 2B 30 32 FRACTION NUMBER .IM M PO 4
I
5 4 3 6 38 4 0 4 2 4 4 4 6 4B ,3M
I
2.0M
50 I
FIG. 4. DEAE cellulose elution profile of [3H] cortisol b o u n d to 700 × g s u p e r n a t a n t of m a m m a r y tissue slices from lactating cows. B. DEAE cellulose elution profile of [3 H] cortisol b o u n d to bovine sera. Counts per min (CPM) profiles in Fig. 4A and 4B each represent the average o f three cows whereas optical density (OD) in b o t h figures is representative o f one e x p e r i m e n t each. Counting efficiency was 50%. Journal of Dairy Science Vol. 59, No. 7
1252
GOREWIT AND TUCKER
associated with serum protein(s) which migrated 5 to 6 cm from the origin (Fig. 3B). N,N-diethylaminoethyl Cellulose (DEAE) Chromatography
Electrochemical properties of glucocorticoid-binding proteins of mammary tissue and blood sera were determined by DEAE chromatography. The [3Hlcortisol bound to mammary protein(s) in the presence of .01 M potassium phosphate buffer was eluted subsequently from 700 × g supernatant fluids with .3 M potassium phosphate (Fig. 4A). On the other hand, [3 H] cortisol bound to bovine serum protein(s) under similar conditions eluted with .05 and .1 M potassium phosphate (Fig. 4B). Binding of Tritiated Cortisol, Dexamethasone, and Progesterone to Bovine Serum
In experiments designed to determine the relative magnitude and order of binding of tritiated steroids to bovine sera at 37 C, sera bound 3.23 x 10-13 /lmol of [3 H] cortisol, 2.19 × 10-13 /.tmol of [3 H] progesterone, and .0599 x 10-13 /amol of [3H] dexamethasone per mg of protein.
DISCUSSION
Results suggested that the glucocorticoidbinding component of mammary tissue was not a contaminant from blood. For example, gel filtration chromatography and sucrose gradient analysis of mammary supernatant fractions showed that the protein(s) binding cortisol was a macromolecule (~--11.0 S) with a molecular weight of approximately 2.5 × l 0 s to 3 x 106 (Fig. 1 and 2) whereas the major protein which bound cortisol in serum was a macromolecule (~5.0 S) with an approximate molecular weight of 6 × 10 a to 8 × 10 a (Fig. 1 and 2). These characteristics of the major protein binding component for glucocorticoids in blood sera of cattle were similar to those for CBG in blood of other species (10). Radioactive cortisol which associated with the 1 × 106 molecular weight component of sera (Fig. 1) may represent [3H] cortisol binding to immune globulins or other "high" molecular weight serum proteins. The Sephadex G-200 elution patterns of the mammary component binding glucocorticoid Journal of Dairy Science Vol. 59, No. 7
were similar to those described for "specific" cortisol binding protein(s) in mammary tissue by Gorewit and Tucker (4). Moreover, the chromatographic properties of bovine serum components which bound glucocorticoids were similar to those reported for blood of humans by Seal and Doe (10). Disc gel electrophoresis and DEAE cellulose chromatography confirmed that glucocorticoid binding protein(s) of bovine mammary tissue had different electrochemical characteristics than bovine serum binding components. Similar electrophoretic patterns for glucocorticoid receptor proteins in rat brain tissue and rat blood sera have been reported by McEwen (6). Experiments in our laboratory (4) also suggested that the specific corticoid binding component of mammary tissue from lactating cows was not a result of tissue contamination, with CBG of blood. For example, tritiated cortisol or tritiated dexamethasone binding to mammary tissue could not be inhibited competitively with progesterone at 37 C (4). In contrast, CBG has a higher affinity for progesterone at 37 C than for corticoids (9, 12). Moreover, dexamethasone is a moderate to poor competitive inhibitor of cortisol binding to CBG (1, 5, 8), but dexamethasone proved to be a strong competitive inhibitor of cortisol binding in the bovine mammary slice system (4). In the present study, blood sera from lactating cows bound more cortisol and progesterone at 37 C as compared with dexamethasone, confirming further that the glucocorticoid receptor in mammary tissue was not CBG. The biochemical properties of the mammary glucocorticoid-binding protein(s) (i.e., their behavior upon gel filtration and electrophoresis) suggested that it was a large molecule. Several workers have shown similar results for glucocorticoid binding proteins in other tissues (13). Most of these macromolecutar binding proteins were composed of subunits. It is reasonable to speculate that the mammary glucocorticoid binding protein described in this paper is composed also of subunits. This molecular structure may explain the binding protein's chromatographic and electrophoretic properties. REFERENCES
1 Baxter, J. E., and G. M. Tomkins. 1971. Glucocorticoid hormone receptors. Page 331 in Advances in
GLUCOCORTICOID BINDING PROTEINS the biosciences. Vol. 7. G. Raspe, ed. Pergamon Press, Vieweg. 2 Gardner, D. G., and J. L. Witliff. 1973. Characterization of a distinct glucocorticoid-binding protein in the lactating mammary gland of the rat. Bioclaim. Biophys. Acta 320:617. 3 Gardner, D. G., and J. L. Witliff. 1973. Demonstration of a glucoeorticoid hormone-receptor complex in the cytoplasm of a hormone responsive tumour. Br. J. Cancer 27:441. 4 Gorewit, R. C., and H. A. Tucker. 1976. Corticoid binding in mammary tissue slices from lactating cows. J. Dairy Sci. 59:232. 5 Kolanowski, J., and M. A. Pizarro. 1969. Critical evaluation of competitive protein-binding radioassay for cortisol. Ann. Endocrinol (Paris) 30:177. 6 McEwen, B. S., C. Magnus, and G. Wallach. 1972. Soluble corticosterone binding macromolecule from rat brain. Endocrinology 90:217. 7 Munck, A., and C. Wira. 1971. Glucocorticoid receptors in rat thymus cells. Page 301 in Advances in the biosciences. Vol. 7. G. Raspe, ed. Pergamon Press, Vieweg. 8 Peets, E. A., M. Staub, and S. Symchowicz. 1969. Plasma binding of betamethasone-3 H, dexamethasone-3 H, and cortiso1-14 C--A comparative study. Biochem. Pharmacol. 18:1655.
12 53
9 Sandberg, A. A., H. Rosenthal, and S. L. Schneider. 1966. Protein steroid interactions and their role in the transport and metabolism of steroids. Page 1 in Steroid dynamics. G. Pincus, J. F. Tait and T. Nakao, ed. Academic Press, New York. 10 Seal, U. S., and R. P. Doe. 1966. Corticosteroid binding globulin: Biochemistry, physiology and phylogeny. Page 63 in Steroid dynamics. G. PincuS, J. F. Tait, and T. Nakao, ed. Academic Press, New York. 11 Shyamala, G. 1973. Specific cytoplasmic glueocorticoid hormone receptors in lactating mammary glands. Biochemistry 12: 3085. 12 Shyamala, G. 1974. Glucocortieoid receptors in mouse mammary tumours. Specific binding of glucocortieoids in the cytoplasm. J. Biol. Chem. 249:2160. 13 Thomas, P. J. 1973. Steroid hormones and their receptors. J. Endocrinol. 57: 333. 14 Tucker, H. A., B. L. Larson, and J. Gorski. 1971. Cortisol binding in cultured bovine mammary cells. Endocrinology 89:152. 15 Turnell, R. W., P. C. Beers, andJ. L. Wifliff. 1974. Glucocorticoid-binding macromoleeules in the lactating mammary gland of the vole. Endocrinology 95:1770.
Journal of Dairy Science Vol. 59, No. 7