;BRARY “3 higan Sum ; nivcrs‘ity 5 ~91)“, ABSTRACT PURIFICATION OF A STEROID BINDING PROTEIN IN MAMMARY GLAND BY Donald H. Steyert Adrenal glucocorticoids are important in several tissues for a number of functions. Among those functions is their apparent essentiality in the lactation process at both the general metabolic and mammary tissue levels. At the tissue level, binding of a hormone is generally thought to indicate a biological function for that hormone in that tissue. This study was to determine whether such binding occurred in mammary tissue and if it did to par— tially purify and study the binding material. Binding from 2.5 X lO—BM corticosterone was ob— served in centrifugally produced nuclear, mitochondrial and microsomal fractions of mammary tissue at 41.1%. 27.4%” and 31.1%.of the total particulate label in those three reSpective fractions. The remaining soluble, or Donald H. Steyert cytOplasmic, fraction also had considerable binding ac- tivity, revealed during Sephadex filtration. That soluble binding fraction was then partially purified by ammonium sulfate precipitation, calcium phos- phate gel and DEAE-cellulose adsorption and Sephadex fil- tration. The binding protein was initially precipitated from solution by 60-80% saturated ammonium sulfate and that fraction, stable to frozen storage, became the focus of further purification and binding studies. Calcium phosphate gel or Sephadex G-100 filtration provided little or no further purification. However, DEAE-cellulose eluted with phOSphate buffers constituting ionic and pH gradients did provide purity beyond that in the ammonium sulfate fraction. Absolute degree of purifi- cation could not be determined due to apparent high levels of low-affinity, nonspecific binding in the starting material. Although DEAE-cellulose fractionation provided the highest purity of binding protein there were still several bands observable by polyacrylamide gel electro- phoresis. Competition and binding studies are facilitated by separation of bound and unbound hormone. Adsorption 2 Donald H. Steyert of unbound steroid to Florisil, charcoal and dextran- and polyvinylpyrrolidone-coated charcoals on Millipore filters proved to be unsatisfactory. Dialysis of samples against charcoal in a buffer suSpension proved unsatisfactory due to dissociation of bound steroid during dialysis. Equil- ibrium dialysis does not separate unbound from bound hor— mone but does produce data for equilibrium conditions and as such was used for both competition and binding constant studies. Competition for corticosterone binding sites was observed for both hydrocortisone and progesterone but progesterone competition appeared less complete. Neither cholesterol nor l7B-estradiol competed with corticosterone. Two sets of binding constant data were obtained. One set was produced with ammonium sulfate fractions from two mammary tissues dialyzed 12 hours against four corti- costerone concentrations. The second set of data was from the same fractions of two other mammary tissues, one of them duplicated, but they were dialyzed 24 hours against 12 corticosterone concentrations. The first data gave high affinity binding constants between 0.96 and 4 X 10-9M from Scatchard and double reciprocal plots. The second 3 Donald H. Steyert data gave, by linear regression calculations on a Scatch— ard plot, a high affinity binding constant of 3.2 X lo-llM. The discrepancy between the two may result from closer Spacing of the low concentration points or from the in- creased dialysis time. Relationships between blood corticosteroid-binding globulin and mammary glucocorticoid—binding protein may exist. Furthermore, relationships between those two and levels of progesterone and glucocorticoid may relate to lactogenesis. During lactogenesis. near parturition, blood levels of progesterone decrease while free gluco- corticoid levels increase. Since the mammary binding protein strongly binds both of these steroids the decreas- ing progesterone concentration may permit increased gluco- corticoid binding which may thus involve this mammary protein in the initiation of lactation. V ITA Donald Harrison Steyert was born December 8, 1940, in Brooklyn, New York. His parents, Harrison and Sybil, and younger sister, Jo-Ann, lived fourteen years in Massa- pequa Park, New York, before moving to Orwell, Vermont. After attending two high schools in New York and two in Vermont he was graduated from Hudson Falls Central High School, New York, in 1958. He attended the University of Vermont from 1958 to 1962 and received a BS degree in Animal Science. He received the MS degree from Michigan State University, Department of Dairy, under the supervi— sion of Dr. R. S. Emery in 1965. The title of his thesis was "Glucose Assimilation by Rumen Microbes and Effect of Penicillin." He was then employed by the Michigan State University Department of Veterinary Surgery and Medicine for work on a canine leukemia project for one year. He will now receive a Ph.D. degree from the Dairy Science Department, Michigan State University, in 1971. Since January of 1971 he has been an assistant professor in the Dairy Science Department at the University of Illinois on temporary appointment for one year. He is a member of the .American Dairy Science Association and Alpha Zeta. PURIFICATION OF A STEROID BINDING PROTEIN IN MAMMARY GLAND BY . i'r/ 'U‘ Donald H? Steyert ' A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Dairy 1971 PI 0 u :o o u !-I at ACKNOWLEDGEMENTS I would like to express my gratitude to Dr. Roy S. Emery, my academic advisor, for his assistance andperse— verence during the term of my graduate studies. I also appreciate the efforts of Dr. Harold Hafs and Dr. Steven .Aust to bring this study to completion. I am indebted to the Dairy Science Department for use of their facilities. I must also thank the United States' taxpayers and Dr. C. A. Lassiter for provision of some financial support in the form of a research assis- tantship. To my parents, without whose financial assistance and continuous encouragement this degree would not have been attained, I express my deepest gratitude. Similar gratitude is expressed to those people whose friendship and encouragement enabled me to continue this study at times when its completion seemed impossible. ii a. I. i 1' o 3‘ '6 TABLE OF CONTENTS Page LIST OF TABLES. . . . . . . . . . . . . . . . . . . vi LIST OF FIGURES . . . . . . . . . . . . . . . . . . ix LIST OF APPENDIX TABLES . . . . . . . . . . . . . . Xi INTRODUCTION. . . . . . . . . . . . . . . . . . . . 1 REVIEW OF LITERATURE. . . . . . . . . . . . . . . . 2 Adrenal Glands and Their Secretions . . . . . . 2 In Vivo Effects of Glucocorticoids. . . . . . . 4 Liver . . . . . . . . . . . . . . . . . . . 4 Mammary Gland . . . . . . . . . . . . . . . 5 In Vitro Effects of Glucocorticoids . . . . . . 11 Mammary Gland . . . . . . . . . . . . . . . 11 Other Tissues . . . . . . . . . . . . . . . 20 Histones. . . . . . . . . . . . . . . . . . 22 Binding Studies . . . . . . . . . . . . . . . . 25 Transcortin, CBG. . . . . . . . . . . . . . 25 Estrogen in Target Tissues. . . . . . . . . 29 Glucocorticoids . . . . . . . . . . . . . . 30 iii TABLE OF CONTENTS (Cont.) General Techniques. . . . . . . . . . . . Time and Temperature Effects. . MATERIAL AND METHODS. . . . . . . . . . . . . . . Tissues . . . . . . . . . . . . . . . . . . . Homogenization. . . . . . . . . . . . . . . . Tissue Fractions. . . . . . . . . . . . . . . Steroid Binding . . . . . . . . . . . . . . . Other . . . . . . . . . . . . . . . . . . . . RESULTS AND DISCUSSION. . . . . . . . . . . . . . Measurement of Binding. . . . . . . . . . . . Dilution Methods. . . . . . . . . . . . . Adsorption Methods. . . . . . . . . . . . Equilibrium Dialysis. . . . . . . . . . . Isolation and Purification of Steroid-binding Protein 0 O O O O O O O O O O O O O O I I 0 Stability . . . . . . . . . . . . . . . . Extraction from Particulate Fractions . . Distribution of Binding Among Cell Fractions 0 O O O O O O O O O O O O O 0 Purification of Solubilized Binding Pro- tein. O O O O I I O I O C O O O O O O 0 iv Page 34 38 41 41 42 42 47 50 53 53 53 58 71 75 75 78 78 84 TABLE or CONTENTS (Cont.) page Ammonium sulfate. . . . . . . . . . . . 84 Calcium phosphate gel . . . . . . . . . 92 Sephadex. . . . . . . . . . . . . . . . 98 DEAE-cellulose. . . . . . . . . . . . . 99 Immunological comparisons . . . . . . . 102 Binding Characteristics . . . . . . . . . . . . 105 Competition Studies . . . . . . . . . . . . 105 Binding Constants Determinations. . . . . . 119 GENERAL DISCUSSION. . . . . . . . . . . . . . . . . 150 LIST OF REFERENCES. . . . . . . . . . . . . . . . . 163 APPENDIX. . . . . . . . . . . . . . . . . . . . . . 175 Table LIST OF TABLES Bound hydrocortisone and corticosterone in the peripheral circulation. . . . . . . . . Unbound serum corticosteroid levels (pg/100 m1) during pregnancy and lactation. Mean subcellular distribution of the corticosterone-H3 fraction in rat brain, thymus, heart and liver 30 minutes after an intravenous injection of 20 microcuries of corticosterone-H3. . . . . . . . . . . . Percentage of total recovered label found in each of four washes of 27,000 x g precip— itates from homogenate incubated with four levels of corticosterone. . . . . . . . . Amounts of glucocorticoid adsorbed by Flor- isil and comparison with Trapp and West data. . . . . . . . . . . . . . . . . . . Adsorption of corticosterone passed through 0.1 g activated charcoal on a 8 u Millipore filter. . . . . . . . . . . . . . . . . . Loss of corticosterone and protein by passage through three coated-charcoal preparations on a 8 u Millipore filter . . . . . . . . . Removal of corticosterone by dialysis against charcoal suspension . . . . . . . . . . . . Effect of mercaptoethanol on binding of corticosterone-H3 to particulate subcel- lular fractions . . . . . . . . . . . . . . vi Page 26 28 31 57 60 63 64 7O 76 LIST OF TABLES (Cont.) Table 10. 11. 12. 13. 14. 15. 16. 17. Distribution of corticosterone—H3 among three subcellular, centrifugally produced particulate fractions . . . . . . . . . . . Recovery of protein and radioactivity from increasing concentrations of calcium phos- phate gel added to the dialyzed NaOH ex- tract of a 30,000 x g particulate fraction previously incubated with corticosterone- H . . . . . . . . . . . . . . . . . . . . . Recovery of radioactivity per mg protein from increasing concentration of calcium phos— phate gel added to labeled 30,000 x g supernatant incubated with corticosterone- H . . . . . . . . . . . . . . . . . . . . Incorporation of corticosterone-H3 into par- ticulate fractions and "competition" by progesterone, hydrocortisone and unlabeled corticosterone. . . . . . . . . . . . . . . Loss of corticosterone from TMK and 750 x g supernatant dialyzed against TMK and char— coal. . . . . . . . . . . . . . . . . . . . Loss of corticosterone-H3, hydrocortisone, progesterone and estradiol from a 750 x g supernatant dialyzed against charcoal suSpended in TMK. . . . . . . . . . . . . . Concentrations of various hormones added to compete with 8.3 X lO’llM corticosterone- H3 in equilibrium dialysis of 60-80% sat- urated ammonium sulfate fraction of 30,000 x g supernatant. . . . . . . . . . . Corticosterone concentrations and reSpective bound to unbound ratios after equilibrium dialysis of two 60-80% saturated ammonium sulfate fractions from different bovine tissues . . . . . . . . . . . . . . . . . . vii Page 80 94 96 107 112 115 118 127 v" s . \ I 1 I l LIST OF TABLES (Cont.) Table Page 18. Twelve corticosterone concentrations used in equilibrium dialysis of 60-80% saturated ammonium sulfate fractions from two bovine mammary tissues . . . . . . . . . . . . . . 142 19. Binding constants, k, and their standard deviations, Sk' calculated for the five segments of the curve shown in Figure 12. . 147 20. High affinity binding constants and their standard deviations determined by linear regression calculations on data represented in Figure 12. . . . . . . . . . . . . . . . 148 viii Figure 10. LIST OF FIGURES . 3 . Loss of corticosterone-H from solutions dialyzed against charcoal in suspension. . Rates of corticosterone uptake by three cell fractions during equilibrium dialysis . . . . . . . . . . . . . . Sephadex G-100 fractionation of 70-80 ammonium sulfate fraction. . . . . . . . . DEAE-cellulose fractionation of 60—80 ammonium sulfate fraction. . . . . . . . . Corticosterone bound in two particulate cell fractions incubated with four levels of corticosterone. . . . . . . . . . . . . Competition for corticosterone-binding sites by cholesterol, estradiol, progesterone and hydrocortisone . . . . . . . . . . . . Scatchard plot of Table 14 data. . . . . . Double reciprocal plot of Tissue 1 60-80 ammonium sulfate fraction corticosterone— binding after equilibrium dialysis . . . . Scatchard plot of data represented in Figure 8 . . . . . . . . . . . . . . . . Double reciprocal plot of Tissue 2 60-80 ammonium sulfate fraction corticosterone- binding after equilibrium dialysis . . . . ix Page 67 73 100 103 109 120 123 129 133 137 LIST OF FIGURES (Cont.) Figure Page 11. Scatchard plot of data represented in Figure 10. . . . . . . . . . . . . . . . . 139 12. Scatchard plot of combined 60-80 ammonium sulfate fraction, equilibrium dialysis corticosterone-binding data from two tissues with duplication of one. . . . . . 144 LIST OF APPENDIX TABLES Table Page 1. Chemical formulation for disc electro— phoresis. . . . . . . . . . . . . . . . . . 175 2. Composition of scintillation fluid. . . . . . 176 xi INTRODUCT ION Adrenal glucocorticoid hormones have great influ- ence upon the body. The research conducted and reported herein was designed to add one more piece of information concerning the relationship of glucocorticoids to one specific tissue, the mammary gland. The importance of glucocorticoids in mammary de— velopment and in the initiation and maintenance of lacta— tion is concluded from a review of the pertinent litera- ture. Binding of a hormone in a tissue is an indication for its biological role in that tissue. Binding of estrogenic hormones and glucocorticoids has been shown in a number of tissues. This study was performed to show the existence of macromolecular binding of glucocorticoid hormone in mammary tissue. ‘a-: “.3; -.=. V. ‘i' ' . z. 5a.: he .‘v n \. N . L: I s,. \‘ 3'. REVIEW OF L ITERATURE Adrenal Glands and Their Secretions The adrenal glands, which lie near the kidneys, are each composed of two distinct portions; each secreting a different type of hormone. The inner portion, or ad- renal medulla, secretes epinephrine and norepinephrine *while the other portion, the adrenal cortex, secretes steroid hormones called corticosteroids. The adrenal cortex also secretes small quantities of androgenic and estrogenic hormones (1). Corticosteroids are further classified as mineralocorticoids and glucocorticoids. The mineralocorticoids are primarily responsible for maintaining electrolyte blance while the glucocorti- coids, of which hydrocortisone (cortisol, or l7a—hydro— xycorticosterone) and corticosterone (4—pregnene—118, 21—diol-3,20-dione) are the major molecules, are important in carbohydrate, protein, and lipid metabolism. The adrenal cortex synthesizes the glucocorticoids from acetyl CoA or cholesterol formed elsewhere in the 2 . IQ." I'I‘ n u l I It‘! to I .!. «\ It'll | .‘m.. 'hdo. "'n~ '0. .. um A. t,n: '5’. VI. .I . . a» n. "‘ s..E '8‘ ' . ‘s: body. Secretion of these hormones is under the control of adrenocorticotrOpin (ACTH) secreted from the anterior pituitary gland (1). Ninety-five percent of the glucocorticoid activity in the human results from hydrocortisone secretion while a small amount comes from corticosterone (2) and a minute amount from cortisone (l). Corticosterone is the major circulating glucocorticoid in the rat (3) and mouse (4,5); however, the bovine has both corticosterone and hydro- cortisone but blood concentration of both is only one- .tenth that of the rat (6,7,8). The best known affect of glucocorticoids on me- tabolism is their stimulation of gluconeogenesis in the liver. By increasing gluconeogenesis and possibly by decreasing cellular glucose utilization, glucocorticoids increase the blood level of glucose. The gluconeogenic effect of glucocorticoids increases the catabolism of cellular protein and decreases the rate of protein syn- thesis in all tissues except the liver where both liver protein content and output of plasma proteins are in- creased (1). Glucocorticoids also promote lipid mobilization. They increase both the rate of incorporation into adipose Q.‘ noN. . n. 'c you, -—- c.‘ . og~~ I, ‘ .~.u ‘ k.‘ . \‘1‘1 "I: fig, t“' “cl \‘3 tissue and the rate of removal from adipose tissue by accelerating many of the different processes involved in lipid metabolism (1). In Vivo Effects of Glucocorticoids Liver The tissue showing the most widespread response to glucocorticoids is the liver which would be expected considering the importance of both glucocorticoids and liver in gluconeogenesis and lipid and carbohydrate me- tabolism. The liver usually responds to glucocorticoids or to increased dietary protein by increasing its enzyme activity. Induction of numerous liver enzymes has been studied by administration of both exogenous glucocorti- coid and increased dietary protein. Szepesi and Freedland (9) cite numerous examples of rat liver enzymes affected by both glucocorticoid treatment and increased dietary protein. Among them are tryptophan pyrrolase, tyrosine- a-ketoglutarate transaminase, serine dehydrase, glutamic- pyruvic transaminase, glutamic-oxalacetic transaminase, 0 n .,. 011 0 0 (II n, ‘ . . "-5 Ii'. . I.~ .- 9» . i".. E-: < glucose-6-phosphatase and certain enzymes of the urea cycle, amino acid catabolism and gluconeogenesis. Hor- monal effects on enzyme turnover rates were also studied. The rates of synthesis of two of the enzymes studied, glucose-6-phosphatase and glutamic-pyruvic transaminase, were increased by hydrocortisone without changes in their rates of degradation. The rate of synthesis of another e1"lzyme, serine dehydrase, was increased by both hydro- cortisone and increased dietary protein while only hydro- cortisone increased its half-life (9). Wm: Gland The develOpment of the mammary gland is a step- wise process starting with duct development followed by formation of the secretory tissue and finally secretory aetivity. The hormonal requirements for duct development were studied by Lyons e_t__a_l. (10) using hypophysectomized- cDvéariectomizedwadrenalectomized rats. They found that $81:rogen plus growth hormone gave some duct growth; how- ea"Gar, a glucocorticoid administered along with the estro- g§h and growth hormone yielded full duct growth. or n. l rev. u 1‘ a. I... "-- baf- 9!“ n—‘v. . Oo- h u.» I ‘9. "wt . h.‘ « 5.. I“: m,“ a .\ ‘S . v in (I) a. My The secretory tissue of the mammary gland, con- sisting primarily of lobules of alveoli, develops after duct develOpment. Some lobule-alveolar development was produced in rats by estrogen plus progesterone plus growth hormone; however, the addition of a glucocorticoid to these hormones resulted in full development and milk Secretion (10). Lobule-alveolar develOpment histologically sim— ilar to midpregnancy development in the intact mouse has been produced in hypOphysectomized-ovariectomized- acarenalectomized virgin mice by in vivo hormone treatment. Growth hormone plus estradiol and progesterone produced the best develOpment. Lactogenesis was induced in the 1T1ice by a fifteen—day treatment with estradiol, proges- 1:erone and growth hormone plus five daily injections of hl’drocortisone acetate and growth hormone. The histolog- iQal appearance of the glands in these rats was very sim- ilar to that of the normal lactating gland (ll) . Similar re sults have also been observed in the goat (12) indicat- ing that growth hormone and prolactin act in conjunction with ovarian and adrenal cortical hormones to produce :Eull alveolar development. "‘P .. ”.1. ... . 1g 0 ., ‘0' In' y. l 'o-.§ . no...- . ‘ . ‘56:. - "va'. ‘In 5... .I" Irv: 2+- ,_. “0.. 0"‘: Hormonal initiation of lactation has also been studied. ACTH or adrenal corticoids will initiate lacta- tion in all species in which they have been tried provid- ing well deve10ped mammary glands are present in the animal prior to treatment. ACTH injections initiated lactation in pseudOpregnant rabbits (13) and corticoid treatment initiated lactation in pregnant rats (14,15) and rabbits (15) and pseudOpregnant rabbits (16). Predef (9-f1uoroprednisilone), a synthetic adrenocortical ster— oid, was used by Tucker and Meites (17) to initiate lac- tation in early, mid-, or late pregnancy heifers. They suggested that "adrenal cortical steroids are limiting fa Ctors for initiation of lactation in pregnant heifers, as in laboratory animals." There is some evidence that 91a sma concentration of free adrenal glucocorticoids it‘Qreases at the time of parturition (7) - Much evidence exists suggesting glucocorticoids are also very important in the maintenance of lactation. The ACTH content of the anterior pituitary was found to a‘eQ:r:ease 68% between the twentieth and thirty-sixth day Q . . . . f lactation in rats (18) . This period is also assoc1ated W ‘ ' , 1th greatly decreased milk production. I” '1 o I-‘Vh n .§< q. I IthI 1,. u..:= 3“; 'b.‘ 1 II ."QI 3‘... ‘IP . § . 3 I. ‘ Q U . a l “ . . . ‘I ~ ' ‘I ‘0 . . .I I In vivo glucocorticoid treatments have been used to study sustentation of lactation. Lactation in adrenal— ectomized rats could not be supported by mineralocorti— coids alone but added glucocorticoids acted synergistic- ally to maintain lactation (19,20) . Cortisone acetate Siignificantly increased lactation during the first 18 days postpartum (21) and hydrocortisone acetate prevented the drastic reduction in litter weight gain normally Observed between 24 and 32 days of lactation in rats (22) . HYdrocortisone acetate was injected with and without ell‘tcbgenous prolactin in order to study effects on mammary cell RNA and DNA. Injections from day 16 to 32, without ED-t‘calactin, maintained mammary cell numbers, measured as 9land DNA content: however, metabolic activity per cell (IRJNA/DNA) decreased (22). Injections of both hydrocorti— Es(Dine acetate and prolactin had the inverse effect. These ht>erones reduced the DNA content but increased the RNA/DNA 1:6i12io. Thatcher (23) also showed that Predef injections Estliirted on day 24 maintained litter weight gains, DNA lea‘U’els and gave a high RNA/DNA ratio to day 32. It would lbfii possible that the decreased lactational performance EQUnd during the latter portions of corticoid treatments Q0mild be due to nutrients becoming limiting and increased caloric intake has improved lactation performance during Predef treatment (24) . ACTH, contrary to its effects in rats, decreased lactational performance in cows (8,25) . Injection of ACTH cxaused a sharp drop in milk yield from which maximum re- cxovery took at least three days (8). There was also a ITipid, marked increase in plasma l7—hydroxy corticosteroid llavels which fell sharply in a few hours and returned to nCDrmal within 24 to 48 hours. Activities of a number of enzymes have been studied with respect to the effect of onset of lactation upon tllmem. Activities of 6-phosphogluconate dehydrogenase, Ipliesphoglucomutase and phosphofructokinase in the rabbit e13~1 increase between one day prepartum and one day post- E36irtum and these increases follow the DNA increase seen c)\’er the same period (26). Acetyl CoA carboxylase follows tllle same pattern but citrate cleavage enzyme continues to irlcrease in activity to at least 10 days of lactation, tille time when DNA levels decrease. Hartmann (26) reported t:})e observed changes in the rabbit were consistent with tzlie pattern seen in the rat, mouse, and guinea pig; how- ea‘Ier, the changes were not as marked as those seen in the 1:‘Eit (27) and guinea pig (28). The marked increases in 10 the activity of enzymes seen with the onset of lactation in those two species was not observed in the cow where the enzymatic potential for milk synthesis was present at least 14 days prepartum (28) . A number of mammary enzymes are affected by in 3319 treatments of prolactin and hydrocortisone. These two hormones in combination significantly increased glucose-6-phosphate dehydrogenase and fatty acid syn— thetase in hypOphysectomized-rat mammary glands (29,30, 31). Levels of 6-phosphogluconate dehydrogenase and phosphoglucomutase were also increased by combined pro- lactin and hydrocortisone, while malate dehydrogenase arind fructose 1,6-diphosphate aldolase were increased by either prolactin or hydrocortisone (29,30) . Citrate Qleavage enzyme and UDPG-perphosphorylase were both snignificantly increased by hydrocortisone and further increased by the addition of prolactin (31) . With the exception of glucose-G—phosphate dehy— drogenase, most of the enzymes mentioned above increased a 8 DNA content increased thereby reflecting gland growth a~11d cell proliferation (29) . Glucose-6-phosphate dehy- dIlz'ogenase activity per mg DNA was increased by hydrocor- tisone plus prolactin but not by either one alone. The 11 activity of another enzyme, aspartate amino transferase, increased slightly per mg DNA by either hydrocortisone or prolactin (29) . In Vitro Effects of Glucocorticoids Ma_nuuary Gland Numerous studies relating lactogenesis to specific hormones have been performed in vitro where experiments are often more convenient than in vivo and the tissue is removed from influences other than those being studied. Much of the in vitro work has used mouse mammary cultures to study histological changes following various hormone treatments. Glands removed from 14 day-pregnant mice.show a-dvanced lobule-alveolar development, but culturing this tissue in hormone-free medium resulted in extensive de- gli‘adation of the parenchymal cells. Neither hydrocorti- 8<>ne, growth hormone, nor prolactin alone maintained this tissue; however, a combination of hydrocortisone and pro- lfictin was effective in maintaining viability and the l"I'Lstological integrity of secretory tissue (32) . 12 Insulin introduced into hormone—free culture medium improved the in vitro survival of 10-12 and 14-18 day—pregnant mouse mammary explants but did not maintain them as well as in combination with hydrocortisone (33, 134). Hydrocortisone alone would not maintain the normal lristological appearance of these explants and some his— tIDlogical differences were noted between tissue in ilusulin-hydrocortisone vs the prolactin—hydrocortisone nRadium (33) used previously (32). The hydrocortisone requirement for maintenance of the mammary explant was also found to increase with the degree of differentiation ( 34). Hormonal treatments have also been used to induce rnEiImarysecretion. The minimum hormone combination re- c«II-'lired for initiation of secretion from early-prelactating ( 1.0-12 day pregnant) mouse mammary was either prolactin ()1? growth hormone plus insulin and hydrocortisone (34). {Irfle insulin-hydrocortisone-prolactin treatment produced differentiated alveolar cells which were not merely mod— 1ified duct cells but arose only as the result of stem cell ciIiLvision in the presence of the three hormones (35). Hydrocortisone, corticosterone, and aldosterone wGare all effective for induction and maintenance of normal 13 secretory histological appearance in mouse mammary ex- plants (36) . Using aldosterone as the model adrenal corticoid histological appearance was maintained with a minimum of l and 5 ug per ml of corticoid and insulin, respectively (37) . The hormones at those levels also allowed prolactin and growth hormone to initiate secre— tion (37) . Steroid structural requirements for mammary gland differentiation have been determined. For intermediate acztivity at 3 x 10.7 M concentration the 4-pregnene steroid nucleus must have 1) a 20-keto group, and 2) an 11 B—ol or 21-01. The requirement for the 3-keto group was not determined. Hydrocortisone, corticosterone, and aldosterone all meet the requirements for intermediate "‘ammary gland differentiation. For increased activity at 3 X 10-7M the 4-pregnene nucleus requires the 20-keto with either a l7a—hydroxyl or 18-aldehyde and either an 11 B-ol or ll-one (38) . Hydrocortisone and aldosterone Deoxycorticosterone, 5M both meet those latter requirements. IE‘DImd inactive at levels of 3 X 10-6M and 1.4 X 10- ( 36), meets the requirements for intermediate differen- tiation (38) . Perhaps the relationships among the groups IPITesent on the nucleus is more important than was realized 14 in determining the requirements. Deoxycorticosterone has the 21-hydroxyl group required. for intermediate activity and the ll-one required for a high degree of differentia- ti on. However, the presence of the ll—one without the accompanying l7a-hydroxyl or 18-aldehyde may be enough to make it inactive. Corticosterone lacks the requirements for high activity although some in vitro work done with high concentrations of corticosterone showed it equivalent to hydrocortisone (36) . There is some evidence that much h:i—gher levels of corticosterone than hydrocortisone were re<1uired in vivo for histological response (see 36) . Histological develOpment of the mammary gland has been correlated with increased nucleolar size, increased RNA synthesis and initiation of casein synthesis (39) . A number of studies related hormonal effects to both his- tological appearance and the onset of casein synthesis. Virgin—mouse and pregnant-mouse mammary glands were treated with insulin, hydrocortisone, and prolactin t0 study incorporation of Pi32 into casein-like phospho- pt‘ Qteins. Virgin mouse mammary tissue showed little in— c(Disporation relative to pregnant mouse mammary (40). Pt’Oliferation of the duct system and more extensive alveolar development are probably required for 15 susceptibility to the triple hormone treatment (40). The pregnant mouse mammary was develOped enough for the three hormones to synergistically elicit "casein" synthesis (40, 41 ) but none of the hormones alone or in pairs elicited that response (40). The histological development of the tissue was closely related to the elicitation of "casein" and both had the same triple hormone requirement for max— imum response (35,40,42). Those phosphoproteins showing a stimulated synthesis were identical to the phosphoprotein Present in mouse milk (43) . The rate of total casein synthesis was augmented t-11:r‘ee to sixfold by the triple hormone treatment (43) . I:lisulin alone was not able to maintain either the charac— t-eristic pattern among the major casein components or the to‘!::.al capacity to make casein; however, the addition of hYdrocortisone to the insulin treatment occasionally main- ta ined the pattern but did not sustain total casein syn- t1'lesis. Insulin plus prolactin gave some small stimula- tion of total casein synthesis but could not maintain the j‘tlitial pattern. The triple hormone treatment produced a casein pattern like that found in the milk from 10-day LaQtating mouse tissue (43) . PhosphOproteins other than l6 casein were'not stimulated by insulin—hydrocortisone— prolactin treatment therefore the effect of the hormones must be very specific and not merely be to increase total protein synthesis or total cell numbers (39) . Further work elucidated the sequence in which the hormones exerted their effects. Insulin is primarily re sponsible for the DNA synthesis observed during cell Proliferation in vitro (35,42) and the cytoplasmic non- mi 1k protein synthesis believed to reflect that prolifer- at; ion (44) . In the absence of exogenous insulin immature ( 3—week old) mouse mammary, in contrast to adult tissue, will undergo DNA synthesis and mitosis (45) . Such pro— L i feration, however, did not lead to functionally differ- e1'11: iated cells (45) . Insulin and hydrocortisone both had to be present during proliferation in order to elicit Ll“creased casein synthesis (42,45). The casein synthesis was induced in the absence of lobule—alveolar development: ho‘Mever, the normal alveolar development usually associ- aIled with initiation of casein synthesis could be induced in the immature mouse mammary gland after priming the atifLmal with estradiol, progesterone and growth hormone. Then alveolar structures could be developed in vitro by LI“Sulin, hydrocortisone and prolactin (45) . 17 The hormone-dependant casein synthesis was di- xr<3<2tly prOportional to the rate of DNA synthesis, a re- ifJLeection of proliferation occurring during that period ( £355). The DNA synthesis resulting from insulin treatment, 511113 subsequent events in the cell cycle, were necessary 15(3): hormone dependent mammary differentiation (35,42). Early reports indicated that cells which prolif— eezraated under the influence of insulin did not make casein unless hydrocortisone and prolactin were present during the proliferative period (42), but later work showed only I‘13.1?’<3rocortisone, and not prolactin, to be necessary during F>J=<31iferation (46). In the more recent work tissue which l"mad proliferated in insulin or insulin-prolactin medium 3l'lcawed no increase in casein synthesis upon post— L:>~'l=‘11rs the rate of synthesis decreased almost 27% and total Km levels also fell. RNA synthesis in the presence of L nsulin gave higher initial incorporation which was main- ta ined during the 12—hour incubation (47,48,49) and tissue RNA was nearly unaltered (47) . However the insulin effect (lid not last for 48 hours of culture (48) . Addition of o Q:I:'ticosterone and prolactin to the insulin-culture-medium pI‘Qduced increased adenine incorporation into RNA starting it 6 hours and finally doubled the initial value with a Le sser increase observed in insulin-hydrocortisone medium ( <1 8). Prolactin has a post-mitotic action during casein induotion as evidenced by the inability of colchicine to block its effect (46,50) . ‘Prolactin post-mitotically lr‘duced casein synthesis in mammary cells differentiated 19 by hydrocortisone. Lactalbumin and B—lactoglobulin syn- thesis was also augmented (50) and may reflect a further role for prolactin during initiation of lactation because the ratio of a-lactalbumin and B-lactoglobulin has been shown to increase as the lactational state approached a. lthough their combined amount increased proportionately t— o casein's increase (44) . Most reports concerning hor— ItIOr'ie requirements for induction of casein synthesis show Prolactin to be necessary but there is some evidence that growth hormone can substitute for prolactin (34) . Closely related to prolactin's effect on induction 0 5 milk protein synthesis is its effect on RNA synthesis ( 41) . Prolactin-stimulated RNA synthesis was required to convert hydrocortisone-differentiated cells into secretory cells. The prolactin was also able to induce RNA syn- thesis in cells which had not been exposed to hydrocor- t' i sone but they were unable to make secretory proteins ‘ $1). The best defined roles of insulin, hydrocortisone and prolactin are those just discussed but one further e Effect of hydrocortisone should be noted. In rat liver, either corticosterone or estradiol caused in vitro forma- tlon of rough endoplasmic reticulum from smooth 'I .V w '51 .- 20 endOplasmic reticulum and polysomes (52) . A similar study 0 f the subcellular appearance of mouse mammary alveolar epithelial cells, also showed hydrocortisone able to in- duce formation of rough endOplasmic reticulum which was necessary for subsequent synthesis of secretory protein C 5 3,54). In the mammary cells the hydrocortisone was used 1. r1 conjunction with insulin, but insulin alone elicited L ittle response (53) . The formation of the rough endo- Plasmic reticulum provided an indication that hydrocor- :- isone may have a post-mitotic effect because the cells cQI1taining rough membranes were formed whether hydrocor— : isone was added with insulin prior to daughter cell E(Damnation, or was added with insulin after proliferation C 54). Ner Tissues Glucocorticoid-induced synthesis of specific pro- t‘a ins in liver hepatoma has also been studied. As men— t ioned previously, hydrocortisone induces numerous enzymes in the liver, two of which are tyrosine-a-ketoglutarate t‘t‘ansaminase and tryptOphan pyrrolase. Tyrosine transa- mlnase was studied in a tissue culture cell line 'u “r '. 21 e stablished from rats with ascites hepatomas (55) . The hepatoma cells were found to be rapidly and substantially induced to produce increased tyrosine transaminase in response to a number of glucocorticoid hormones. The increased transaminase levels were due to increased syn- thesis and not decreased degradation (56) and there were 9 lucocorticoid receptors detected having properties con- 8 iStent with the increased transaminase synthesis (57) . The induction of tyrosine transaminase was found to be due to steroid antagonism of a labile post- transcriptional repressor of tyrosine transaminase (58, 59 ) . Transcription of the tyrosine transaminase gene was repressed by a steroid—insensitive factor during the 3 2 mitosis and early Gl phases of the cell cycle. In late G1 phase both mRNA for tyrosine transaminase and the steroid sensitive repressor appear. During this J‘éster Gl phase the presence of steroid allows translation Q :E the tyrosine transaminase mRNA. Transport of this RNA to the site of translation may be enhanced or its degradation may be decreased by the steroid antagonism Of the repressor (59) . The mention of transport to the site of action should bring to mind the evidence, men— tioned in a previous section, concerning the ability of . ,...4 .d5\ . nan! dad! a O :" r-.‘ .1 pr: m.‘ :5. s U 5‘ . {v “E (I) 22 glucocorticoids to mediate the formation of the rough endOplasmic reticulum required for secretory protein synthesis. The mRNA which directs the protein synthesis must be transported from the nucleus to the endoplasmic reticulum and glucocorticoid enhancement of the transport could promote formation of rough endoplasmic reticulum. One other tissue in which glucocorticoid induc- t ion of an enzyme has been Shown is human skin fibroblasts where prednisolone induced alkaline phosphatase (60). H Me As early as 1943 there were prOposals (61) that 1'11stones (low molecular weight basic nuclear proteins) a Qt as gene regulators or modifiers. Busch (62) notes that although the basis for such a concept is weaker now there are a number of authors who have reported evidence sIzl‘Ebporting histone suppression of nuclear function and Qhe of the most convincing works showed the suppression of RNA synthesis in the nucleus was not merely due to hi stone inhibition of nuclear ATP synthesis (63) . Recent studies of histone synthesis in mammary g:Lvand have provided some indications of possible I V In D ;n . UV ‘0- .1.... nusbb b. “ rm. ‘ I - ~q.‘ o ‘F-H‘] O-‘D‘4 - o u: f Dub 5 .“DA. n-n on...- a“. l" f 23 relationships between histones and initiation of lacta- ticui. Using the mammary culture system of Stockdale and Topper (42), Hohmann and Cole (64) compared histone synthesis in mammary glands cultured in insulin and insulin-hydrocortisone—prolactin. As proliferation of the glands occurred, total synthesis of lysine—rich his- tones paralleled DNA synthesis and chromatographic frac— tions of the lysine—rich histones showed definite effects due to hormone treatment. Synthesis of one (Fraction 1) of the five major fractions was reduced by exposure to the triple hormone treatment for 16-24 hours while expo- sure for 24—32 hours produced a much larger reduction of Fraction 1 but increased synthesis of Fractions 4 and 5. The authors concluded that such dramatic effects preceding the maximum induction of casein synthesis were related to the eventual synthesis of casein. Histone synthesis in the mammary gland has also been studied in vivo (65). With the onset of lactation the lysine-rich and arginine—rich histones were synthe- sized at 2—3 times higher rates than the slightly lysine— rich fraction. The specific activities of all five lysine-rich fractions increased with the onset of lacta- tion but Fractions 1 and 3 increased the least. After 24 Six.mo« Amanovmm.m AvHV©H.H ANHVNO.H cowumuoma mama Amuavoo.m Assas.o Amoco.o acaumuomH mamas Imm-smvam.o Amvaa.o Aomvo~.o squamous mums Aemvue.m ANHVNH.o aaaevma.o uamamuuauaz ¢H.H mm.o mo.o camufl> ucmcmwumcoz Aeronauuoooumxmv Aocoumumooauuoov choumumooaunouv panama mmooz umm coaumuomH no mocmcmmum mo mmmum taonmumooeuuoo 6cm mmeommm .Ame .ns coaumuoma com mocmcmmum mcfluso AHEOOH\maV mao>ma caoumumoowuuoo Eamon GCDOQGDII.N mummy 29 Estrogen in Target Tissues Much of the work concerning steroid binding in target tissues has involved the binding of estrogen. Jensen and Jacobsen (73), studying estrogen binding, have been credited (74) with obtaining the best early evidence that target tissues contain molecules which specifically interact with hormones and subsequently modify biological function. They injected estradiol into immature rats and showed the hormone to be specifically concentrated in the uterus and vagina. A number of other tissues in various species will also bind estrogen. Jensen et a1. (75) cite reports of high estrogen affinity in rat, mouse, sheep, and goat uterus and vagina; human uterus and mammary tumors and rat anterior pituitary and mammary tumor. Rat liver and mammary gland also show such affinity (76). Numerous studies have localized the binding and considered the binding kinetics (76,77). Shyamala and Gorski (74) showed that the binding of 17 B-estradiol in the rat uterus was first associated with a cytOplasmic 98 protein, of at least 100,000 MW (78). The estrogen then moves into the nucleus where it is bound to an acidic (79) SS protein which can be extracted from the chromatin. 30 As the estrogen moved into the nucleus the 98 protein disappeared or lost its ability to bind estradiol. The authors suggested that the estrogen changed the confor— mation of the 9S cytoplasmic receptor protein enabling it to subsequently move into the nucleus (74). Glucocorticoids Investigations of glucocorticoid binding have been performed on a number of tissues, both in vivo and in vitro. In Vivo binding is usually studied after corticosteroid injection of rats. Injected corticosterone is taken up by a number of rat tissues (80). The quan- tities of corticosterone bound in various subcellular fractions of several of those rat tissues are shown in Table 3. Liver was the only tissue to concentrate gluco- corticoid above blood levels. Subcellular distribution of bound glucocorticoid was studied in liver from gluco— corticoid injected normal and adrenalectomized rats. The mitochondrial fraction bound more hydrocortisone or corticosterone than the nuclear (80,81), microsomal or supernatant (87) fractions and in vitro much more corti— costerone than hydrocortisone was bound (82). Binding 31 .mumcomOEOS one as Umcflmucoo >ue>auom Hmuou mo unmoumm mm Ummmwumxw mue>fluo<«« .mumu mo quESZ¥ o.mv h.mn m.©n 0.00 ucmpmcnwmsm ¢.va H.© o.¢ ©.¢ mmEomouUHZ m.o H.m m.o v.0 MHMUCOLUOUAE m.mm N.mH m.ma «*¢.va Hmaosz v v s rm um>flq pummm msfihne cemum acauuauu demo mounds .mmlwcoumumoofiuuoo mo mweusoouofifi om mo cofluommce msocm>muuCH cm umumm monogas om H0>HH cam sumo: .msahnu .Cflmun umn CH coauomum mmlmcoumumoofluuoo one mo coflusnanumao HMHDHHmuan cmeII.m mqmde 32 in the microsomes and mitochrondia has also been shown by others (83,84). In order to determine whether the mitochondrial binding material was an integral part of the structure, rat liver mitochondria were sonically disrupted and cen- trifuged 105,000 x g. Forty-nine percent of the total mitochondrial protein was sedimented and that 49% of the protein contained 95.5 and 91.8 percent, respectively, of hydrocortisone and corticosterone bound to the mito- chondria. In those mitochondria more corticosterone than hydrocortisone was bound per mg protein (81). The binding in the 100,000 x g supernatant frac— tion from rat liver labeled in vivo appeared in the Sephadex G-100 and G-50 exclusion peaks (83,85,86,87,88); however, exclusion peak material prepared from rat liver would not bind hydrocortisone in vitro (86). The gluco- corticoid bound to liver in vivo has been suggested to be a modified hydrocortisone (86) and two cytoplasmic proteins able to bind hydrocortisone metabolites have been isolated (89). The number of Sephadex—separable corticoid—binding macromolecules has been studied in pigs and rats. Pigs injected with hydrocortisone-H3 showed two labeled 33 cytOplasmic macromolecular peaks eluted from G-100 in liver and spleen but only one such peak in thymus (90). Rats showed the two peaks in liver and one in thymus (91). In the pigs, as in the rats, only the liver con- centrated hormone above blood levels (90). Glucocorticoid binding has also been studied in thymus cells. Thymus cells labeled with various gluco- corticoids in vitro yielded a rapidly labeled physiolog- ically saturable binding fraction and many of the steroids competed for binding in proportion to their glucocorticoid activity (92). At least fifty percent of the glucocor- ticoid bound in those cells labeled in vitro was in the nucleus (93,94) and was extractable with 0.6 M KCl, pH 8.0 (95). The steroid bound in rat thymus cells, in contrast to liver, was hydrocortisone and not some modified hydro- cortisone molecule (92). Some tissues which did not bind any significant amounts of glucocorticoid were rat skeletal muscle (82), heart and brain (91) and pig heart muscle (90). Most of the tissue binding measurements have neglected mammary gland. However, in View of recent work concerning the initiation of casein synthesis in mammary explants there has been renewed interest in mammary 34 glucocorticoid binding. In one limited study, the amount of corticosterone bound increased in both the nuclear and extranuclear fraction with the onset of lactation in rat mammary tissue; however, very little increase was shown between pregnancy and lactation in bovine tissue (96). Mammary glucocorticoid binding has also been studied in bovine mammary cell cultures. Mammary cells 12M labeled hydrocortisone showed cultured with 2—16 X 10- 34% of the incorporated label in the nuclear fraction while the 15,000 x g and 105,000 x g precipitates con— tained 2 and 1% of the label, respectively (97,98). The association constant for hydrocortisone was reported to 8M (98). Hydrocortisone uptake and be about 10"9 - 10' binding was interfered with by progesterone, but not diethylstilbestrol or 17 B-estradiol (97,98). General Techniques ,1 The researchers cited have used many ways to mea- sure binding of steroid to particulate and macromolecular cell fractions. In vivo incorporation was measured with animals injected with radioactive steroid. Those animals were killed after a specified time and their tissues 35 removed, rinsed, and fractionated, usually by homogeniza- tion and differential centrifugation. The radioactivity incorporated into the various fractions was then measured. Binding in the 100,000 x g supernatant, or cytoplasmic fraction, was generally measured by molecular seive chromatography. Binding measurements in cells grown in culture were usually performed in similar manner after incubation of the cells with labeled steroid. Incubation with steroid is frequently done after the tissues and cells are disrupted and fractionated. A homogenate can be incubated with radioactive steroid and the labeled particulate fractions then centrifugally separated or the particulate fractions can be separated and then individually incubated with steroid and measured for binding capacity. In vitro work is frequently concerned with the physiochemical aspects of steroid binding. Effects of pH, temperature, and steroid concentration are often of concern. The most frequently used method of measuring binding under various conditions is equilibrium dialysis. Ultrafiltration can also be used because both equilibrium dialysis and ultrafiltration allow measurement to be made at equilibrium steroid concentrations. 36 Other methods of measuring binding after incuba- tion with steroid are subsequent passage through a molec- ular seive column or precipitation with ammonium sulfate. These two methods are frequently used to separate bound from unbound steroid and hence allow measurement of bound steroid. Both methods give more equivocal results than equilibrium dialysis or ultrafiltration. The results are more equivocal because elution through a molecular seive exposes the steroid-binding molecule to a zero steroid concentration which encourages dissociation of the steroid and ammonium sulfate precipitation aggregates and dena- tures protein which could affect binding capacity and tenacity even though the denaturation may be reversible. Kinetic studies of binding dynamics relate degree of binding to amount of steroid present. Data collected in such studies can be plotted in a manner similar to enzyme data. Lineweaver-Burk-type graphical plots (see 99) of the reciprocal of bound steroid vs the reciprocal of steroid concentration will yield binding constants and number of binding sites, respectively equivalent to KM and Vmax in enzyme studies. The y intercept is the reciprocal of n, the number of binding sites, the x intercept is the negative reciprocal of k, the binding 37 constant, and the slope is -1/kn. Scatchard, in justi— fying a "better" method to plot binding data states that the double reciprocal plot has the disadvantage of con— cealing deviations from ideal laws and tempting straight lines where there should be curvature (100). Scatchard's method is the preferred method for plotting binding data. His plot of bound: free vs bound ligand gives a straight line if k is constant. The y (boundzfree) intercept is —kn and the x intercept is n; therefore, the slope equals -k. Often there are two or more binding sites, each with different affinities for steroid, within a mixture or molecule. To remove the influence of one binding site on another further treatment of the Scatchard plot can be performed (71). The portion of the curve due to low affinity binding is geometrically subtracted from the high affinity binding curve supposedly leaving a pure high affinity curve. However, as the high affinity portion of the binding curve merges with the low affinity portion, subtraction of points in the area of the merging curves will cause the derived line to bend back to the origin. Such geometrical subtraction has recently been used in a study of corticosterone binding in rat liver (84) but 38 most data in the literature is usually derived solely from a Scatchard plot. Time and Temperature Effects Most of the studies dealing with the time course of glucocorticoid uptake have used rat liver and thymus. In rat liver slices, uptake of hydrocortisone followed Michaelis—Menten kinetics with a Vmax' at 4°C, of 9 X 10‘.11 moles/min/g wet tissue and a Km of 2.7 X 10—7M (101). Diffusion was thought to control the rate of uptake in those rat liver slices since equilibrium was 9M and 10-6M corticost- reached in 6 hours with both 10— erone at 0°C (84). The time required to achieve equilibrium has been related to the steroid concentration by incubation of hepatoma cells with two steroid concentrations. The hepatoma cells incubated with 10-6M glucocorticoid showed maximum labeling of binding sites in both nuclei and supernatant in less than 5 min while at a lower concen- tration, 5 X 10—9M steroid, the supernatant binding sites required 5 min and the nuclei 30 min to become saturated (57). 39 The time course of hydrocortisone transfer from cytoplasm to isolated nuclei has been studied in rat liver. Cytosol was isolated and labeled with radioactive hydrocortisone. Rat liver nuclei were incubated with the labeled cytoplasmic binder and within 10 min were maxi— mally labeled. The nuclei incubated with the labeled cytosol fraction bound 1.2 times more label than nuclei incubated with either free hydrocortisone or hydrocorti- sone plus albumin (88). The effect of temperature on binding kinetics has also been examined. At 37°C rat liver cytosol within 6 min bound 16% of available hydrocortisone but after 10 min there was a rapid decrease in amount of steroid bound. At 17°C and 4°C more of the available steroid, 21 and 26%, respectively, was bound but the incorporation was much slower; 95 min were required to reach maximum binding at 17°C and 200 min at 4°C (88). Dissociation has also been examined in thymus cells diluted 50-fold after incubation with 6 X 10-9M hydrocortisone. Those cells showed a rapid, marked dis- sociation of hydrocortisone. The cells lost two-thirds of the bound steroid in about 20 min after dilution at 3°C. After the rapid loss of steroid, the rate of 4O dissociation decreased but rapid dissociation occurred again after warming the suspension from 3°C to 37°C (94). The dissociative prOperties of hydrocortisone in the thymus cell have been used to distinguish high and low affinity binding in those cells. From the high affinity sites hydrocortisone dissociated slowly, as a first order process, with a half time of dissociation of less than 3 min at 37°C while the low affinity sites had a disso- ciation half time of less than 15 sec (92,94). MATERIAL AND METHODS Tissues Most of the work done in this study was performed on lactating bovine mammary tissue; however, some experi- ments were performed with lactating rat mammary tissue. All of the bovine tissue was obtained from an abattoir but was subjected to different treatment depending on whether it was to be stored frozen or fractionated before storage. Bovine tissue was always removed from the udder within 20 min of death. Rat tissue was removed immedi- ately following decapitation. If the tissue was to be frozen whole, it was rinsed in 0.15 M.KC1, placed in a polyethylene bag and kept on ice until it was frozen at -55°C. In one case, mentioned in the Results and Dis— cussion section, the bovine tissue was rinsed with KCl containing penicillin and oxytocin which helped remove residual milk from the tissue. Tissue to be fractionated before storage was removed and washed with an ice-cold buffered salt solution (TMK) containing tris (2—amino—2- (hydroxymethyl)-l,3—propanediol) (0.01 M, pH 7.2), 41 42 MgCl (0.0015 M) and KCl (0.01 M), to remove as much milk 2 as possible. Homogenization Tissues were minced, diluted with three or four volumes of cold TMK and homogenized either with a glass tube-teflon pestle homogenizer driven by an electric drill, or with a Willems Polytron Model PT10 (Brinkmann Instru- ments Inc., Westbury, New York) run at medium speed until no pieces of tissue were visible. The homogenate was filtered through four thicknesses of cheesecloth to remove connective tissue if the 750 x g precipitate was to be used for experimental purposes. If the 750 x g precipi— tate was to be discarded, the homogenate was not cheese- cloth filtered but the connective tissue was sedimented and discarded with the 750 x 9 fraction. Tissue Fractions The initial fractions studied were obtained by differential centrifugation. The low-speed fractions were obtained at either 600 or 750 x g. The 27,000 or 30,000 x 9 fractions were the precipitates resulting from 27,000 43 or 30,000 x g centrifugation of the 600 or 750 x g super- natant. Thirty-thousand x g precipitates referred to in this dissertation were produced by 30,000 x g centrifuga— tions of whole homogenate. The centrifugations at 30,000 x g or less were performed in a Sorvall RC2B centrifuge (Ivan Sorvall, Inc., Norwalk, Conn.). The 100,000 x g precipitate of the 27,000 or 30,000 x g supernatant, termed the 100,000 x g fraction, was produced in a Spinco Model L centrifuge (Beckman, Palo Alto, California) by a 90 min centrifugation in a number 50 rotor. Any of the supernatants used for experimental purposes are referred to in the Results and Discussion section as supernatants of the given centrifugation in order to avoid confusion with the particulate fraction produced by that centrifugation. Supernatants of the 750 and 30,000 x g centrifu- gations were usually passed through a glass wool plug in a glass funnel in order to remove as much lipid as pos- sible before the next centrifugation or any other frac— tionation. Particulate fractions produced by the cen- trifugations were washed by addition of cold TMK, dis— ruption of the pellet and recentrifugation. 44 The above pellet fractions were further fraction— ated by extraction with 0.1 N NaOH. The final standard procedure for NaOH extraction was the addition of 0.5 ml of 0.1 N NaOH per 10 ml of original 750 x g supernatant. The pellet was disrupted, extracted for 30 min and then recentrifuged after dilution with an equal volume of cold distilled water and used fresh or stored frozen at ~10°C. The above supernatants were further fractionated by ammonium sulfate precipitation. The ammonium sulfate precipitations were made on a volume to volume basis using cold saturated ammonium sulfate adjusted to approx- imately pH 7.2 [measured with Accutint (Anachemia Chem— icals, Ltd., Champlain, New York) or pHydrion (Micro Essential Laboratory, Brooklyn, New York) pH paper] with ammonium hydroxide. The precipitates were allowed to form for 20 min while stirring in the cold. The suspen- sion was then centrifuged 5 min at 30,000 x g and the pellet produced was redissolved in a volume of 0.1 N NaOH corresponding to 0.01 the volume of the starting super— natant (750 x g or 30,000 x g supernatant of NaOH extract). An equal volume of TMK was then added and the solution dialyzed against TMK to remove residual ammonium sulfate. The ammonium sulfate fractions are named by the 45 percentages of saturated ammonium sulfate which produced them. They are the precipitates, or resuspended precipi— tates, obtained by making a solution to the given percen- tage with saturated ammonium sulfate. Unless specified otherwise, all the ammonium sulfate fractions were pro- duced from the 30,000 x g supernatant of the mammary tissue homogenates. Molecular seive chromatography [Sephadex G-100, 40—120 mesh (Pharmacia Fine Chemicals, Piscataway, New Jersey)] was used to separate bound and unbound steroid and for further fractionation of ammonium sulfate frac- tions. Glass columns used varied in size from modified Pasteur pipettes to a Pharmacia 1.5 x 30 cm column. The columns were eluted with distilled water at room tempera- ture and timed fractions were collected on a fraction collector. Calcium phosphate gel fractionations were performed with 94—96%.moisture gels prepared by published procedures (102). The gel was stored and used in the cold. Frac- tionations by calcium phosphate were performed in two different ways. One way was addition of a given quantity of gel to the material being fractionated. After incuba- tion, the gel was centrifugally removed from solution and 46 a larger quantity of fresh gel added to the solution. After each addition of gel the mixture was stirred fre- quently for 20 min in a 15 ml conical glass centrifuge tube. This procedure was repeated with increasing quan— tities of gel after which the gels were each extracted with a K2HP04 solution of given molarity and pH. Calcium phosphate gel fractionation of the am— monium sulfate fractions was performed differently. Gel was added at l g (wet weight) gel per 1-3 ml dialyzed ammonium sulfate fraction and incubated 20 min in the cold. It was then sedimented at 2,000 RPM in an Inter- national Model V centrifuge (International Equipment Co., Boston, Mass.) and washed 1-3 times with 1 ml 1 mM K HPO4, 2 pH 7.4, per g gel added. After washing, the gel was ex- tracted with similar volumes of K2HP04 of higher concen- tration and pH as given with the results. Cellex—D (Bio—Rad Laboratories, Richmond, Cali- fornia) was used for DEAE—cellulose fractionations. It was equilibrated with cold distilled water and packed in 0.8 X 30 cm glass columns to about 10 ml volume. Columns were prepared and eluted at 4°C. Samples to be fraction— ated were placed on the column and the column then rinsed with 1 mM K2HP04, pH 7.4. The fractions were eluted from 47 the column by increasing concentration and pH K2HP04 as given in the Results and Discussion section. Steroid Binding Steroid binding experiments to measure distribu— tion of binding among cellular fractions were usually performed with the tissue homogenate diluted 1 g to 4 m1 of TMK, except for two experiments when the homOgenate was diluted 1 g to 9 m1 of TMK. All incubations were performed in either plastic or Corex (Corning Glass Works, Corning, New York) centrifuge tubes and labeled steroids were dissolved in ethanol and added to the mixture with a 10 or 50 ul syringe. Ethanol concentrations in the incu— bation mixtures were normally less than 1% with a maximum of 2%.and agitation during addition of the steroid in ethanol ensured complete mixing. When the incubation mixtures were used to measure incorporation of steroid into particulate fractions the incubations were of 15-60 min duration at 0°C, 4°C or room temperature. Radioactively labeled steroid hormones (New Eng— land Nuclear, Boston, Mass.) used in the various incuba- tions included corticosterone-1,2-H3 (30 Ci/mM), 48 corticosterone—4-C14 (55 mCi/mM), hydrocortisone-4—C (51.8 mCi/mM), progesterone-4-Cl4 (52.8 mCi/mM) and 17 B—estradiol-C14 (47.6 mCi/mM). 14 After incubation with any of the labeled steroids the incubation mixture, or portions of it, were subjected to washing with buffer or Sephadex, charcoal or dialysis treatments. These treatments were attempts to remove unbound steroid from the incubation mixture. The buffer washes and Sephadex treatments are discussed in the Re- sults and Discussion section. One of the other methods, charcoal treatment, was performed by layering either Darco (Atlas Chemical Industries, Wilmington, Delaware) or Norit 1 (Sigma Chemical Co., St. Louis, Missouri) char- coals on 25 mm diameter filters (Millipore Filter Corp., Watertown, Mass. Type SC 84). The charcoal was suspended by rapid stirring and an aliquot of the suspension deli- vered onto the filter. The filters were held in a Milli— pore filter apparatus on a vacuum flask. Incubation mix- ture could then be drawn through the charcoal layer by vacuum and delivered into a small test tube inside the vacuum flask. The charcoal used on the filters was in some cases coated to reduce adsorption of protein from the incubation mixture. Coatings used were dextran 49 (60,000-100,000 MW) and polyvinylpyrrolidone (PVP). Amounts of these two materials used for coating the char— coal are given with the results and discussion of the experiments in which they were used. Charcoal was also combined with dialysis to sep- arate unbound from bound steroid. The incubation mixture (5 to 8 ml) was dialyzed against 300—600 ml TMK containing 1—2 9 activated charcoal kept in suspension by magnetic stirring. Either l or 2.5 cm flat width Visking tubing was used for the dialyses which were done at 4°C. Equilibrium dialysis was also used for binding studies. The first equilibrium dialysis experiments were performed with 1 cm Visking dialysis sacs suspended in 500 or 600 ml beakers containing TMK and the steroids being tested. Later equilibrium dialysis experiments, used to determine binding competition and constants, used 1 cm sacs containing 2 ml ammonium sulfate fraction. Those sacs were suspended in TMK to a total volume of. 100 ml in a 110 m1 test tube. Three to four tubes, each with a different steroid concentration, were used to study degree of binding with varied steroid concentration. Attainment of equilibrium in these eXperiments was moni- tored by suspending sacs containing only 2 ml of TMK in 50 the low and high steroid concentration tubes. All of these equilibrium dialysis experiments proceeded about 12 hours with constant magnetic stirring. The final series of binding constant determina— tions also used equilibrium dialysis and the 60-80% sat— urated ammonium sulfate fraction, however, twelve, rather than three or four, steroid concentrations were used and the dialysis time was 24 hours rather than 12. Other Measurements of radioactivity were performed by the liquid scintillation method using a Nuclear-Chicago Model 720 liquid scintillation spectrometer (Nuclear- Chicago, Des Plaines, Illinois). The liquid scintillation fluid contained dioxane, xylene, ethanol, naphthalene, 2,5-diphenyloxazole and 1.4—bis—(2—(4,methyl—S-phenoxazoly1)) -benzene (Appendix 1). Vials used to contain the scintil- 1ation mixture were either polyethylene or glass and 10 m1 of the scintillation fluid was added to one half ml of aqueous sample to be counted. ElectrOphoresis was performed on 7% polyacrylamide gels made according to a Canalco formulation (Appendix 2). 51 The stacking gel was pH 8.9 and the running gel pH 9.5. The gels were electrOphoresed in a Buchler apparatus (Buchler Instruments, Inc., Fort Lee, New Jersey). Samples placed on the gels had sucrose added to increase their density and were delivered to the top of the gel with a microsyringe (Hamilton Co., Whittier, California). Five to 80 ul samples were applied to the gels. In order to visibly observe the electrophoresis in the gels bromphenol blue solution (1 ml of 0.005% solution) was added to the electrophoresis buffer as a tracking dye. The electrophoresis started at 1.25 milliamps (mA) per tube for 20-30 min until the sample entered the running gel and then the amperage was increased to 2.5 mA per tube. The electric current was stOpped when the tracking dye neared the end of the gel after which the gels were removed from the tubes and stained with Buffalo (amido) black in 7.5%‘acetic acid. After 8-12 hours in the stain they were destained by reverse electrOphoresis in 7.5% acetic acid. After destaining the distances traveled by the bands of protein were measured relative to the dis— tance traveled by the tracking dye. The relative dis— tance moved was called the R (R ) value. dye d 52 The one immunodiffusion study was done using Ouchterlony agar diffusion plates (103) made with 0.85% agar in 0.85%isaline buffered with 0.005 M phosphate, pH 7.4. Antibodies used on the plates were produced by two rabbits injected subcutaneously twice, 11 days apart, with 50-80%.saturated ammonium sulfate fraction and again 31 days later with 60-80%.saturated ammonium sulfate fraction mixed 1:1 with Freund's adjuvant. The rabbits were bled from the ear two months after the first injec- tion; 13 days after the third injection. Serum was pre- pared from the blood and the serum was stored cold with 0.0001%.merthiolate added. Determination of protein concentration were per- formed by 260/280 absorbance ratios and the method of Lowry et al. (104). The absorbancies for the 260/280 ratios were read in a Gilford spectrophotometer (Gilford Instrument Laboratories, Inc., Oberlin, Ohio) and the protein concentrations determined from those ratios by reference to a nomograph prepared by the California Corporation for Biochemical Research. RESULTS AND DISCUSSION Measurement of Binding The quantity of steroid which will bind to some material is usually determined by binding a radioactively labeled steroid to that material and subsequently measur- ing the amount of label bound. The quantity of steroid bound can be found by difference if bound plus unbound label are measured and if unbound, or free, label can be determined in some manner. Such is the case in equili- brium dialysis where bound plus free label is measured inside a dialysis sac and free label is measured in the medium surrounding sac. More frequently, however, bound label is determined by procedures which separate bound from free steroid. Dilution Methods One method for separation of bound and free ster- oid is extensive washing which will dilute the free ster- oid away from the bound steroid. This method is 53 54 particularly suited for use on particulate fractions. Another method often used is molecular seive chromatog— raphy, or gel filtration. Generally, Sephadex G—100 or G-SO, or their equivalent are used to allow exclusion of the macromolecular binding protein and retarded passage of the free steroid. The method is well suited for use with soluble binding proteins. However, both extensive washing and gel filtration have a distinct disadvantage to their use for measuring bound steroid. In both cases the wash or elution media have no steroid in them, i.e. steroid concentration is zero and dissociation occurs. One of the earliest experiments in this study indicated dissociation during pellet washing and gel filtration. Although the experiment was not designed to demonstrate dissociation during washing and gel filtra- tion, calculations of the amount of label added to the incubation mixture versus that recovered as bound and free steroid did indicate the possibility that such dis- sociation was occurring. Those calculations showed that the increased unbound steroid measured after washing and gel filtration would probably not be attributable to contaminating supernatant from the original incubation mixture. 55 The experiment from which the data was calculated was performed by adding corticosterone-H3 (2.2 x 106 dpm) to 2 ml of a resuspended 27,000 x g precipitate from bo— vine mammary gland. After that incubation with labeled steroid, the suspension was recentrifuged. The pellet was carefully separated from the supernatant and suspended in 1 ml of phosphate buffer containing 0.05 N NaOH. An aliquot of that suspension was passed through Sephadex G-100 and the Sephadex fractions counted. Counting effi— ciency, normally about 10%, had to be assumed so in order to obtain a minimum contamination value, a maximum count- ing efficiency (20%) was assumed. Using that 20% counting efficiency there were 1.6 x 105 dpm from unbound corti— costerone after passage through G—lOO. For that many dpm to be unbound, 7.5%, or 0.15 ml, of the original 2 ml of incubation mixture would have to have been contaminating the pellet after the post-incubation centrifugation. That much contamination could not have been present and the data were considered to represent dissociation occurring during resuspension and elution through Sephadex. Dissociation occurring during gel filtration has been observed in binding studies reported in the litera- ture. Fiala and Litwack (86) considered the extent of 56 dissociation which will occur on Bio—Gel P-100, which is similar to Sephadex G-100. After passage through P-100, 48% of the total hydrocortisone added to serum in vitro remained bound compared to 80% during ultrafiltration, in which the free steroid concentration remains unchanged. Similarly, ultrafiltration showed that particle-free supernatant from livers of hydrocortisone-injected rats bound 50%.of the steroid but gel filtration showed only 13% bound (86). Dissociation might also occur during washes of particulate fractions. This was indicated in an experi- ment using homogenates incubated with four levels of corticosterone. Four successive washes of the 750 plus 27,000 x 9 fractions obtained from those homogenates gave the data shown in Table 4. The data, expressed as percent of total recovered radioactivity, show that most of the label which appeared in Wash 1 was probably contaminating supernatant. In Wash 2 some of the label could also be contaminating supernatant, but the label in Washes 3 and 4 probably represent dissociation of bound label from the pellet being washed. Dilution should reduce the label in Wash 4 drastically but although the label in Wash 4 fell below that of washes 2 and 3, it does not reach the low 57 H0.0 VV.0 00.0 H0.0 00.0 H0.H 0N.H 05.0 00.0 00.0 00.0 HB.0 00.0 00.0 0h.v 0v.¢ 00.v N0.0 00.H 00.0 £mm3 2 0 OH x coHumuucmocoo Ufloumum HmfiuecH .mcoumumooauuoo mo mam>0a “Dom £ua3 woumnooca mumcmmoeon Eoum moumuameomum 0 x 000.50 00 mmnnm3 noon mo Sumo cw ocsom Hwnma omnm>oomu Hmuou mo mmmucmoummlu.¢ 04008 58 level which would be expected if there was no dissociation occurring. Adsorption Methods Another method for separation of bound and free steroid is the use of some adsorbant with a strong affin- ity for steroid. The adsorbant is usually added to a mixture of bound and free steroid, the free steroid ad- sorbed and then the label not adsorbed, representing bound steroid, is measured. Two adsorbents commonly used in steroid—binding studies determinations are Florisil, an activated magnesium silicate, and charcoal. Both mate- rials have been coated with dextran (105,106,107), to reduce protein adsorption. A number of determinations during this project showed the adsorption methods to be of little value to this study. Florisil was first tested with a NaOH extract of the 27,000 x 9 fraction. The NaOH extract was incubated with corticosterone-H3 and then passed through G—100. The radioactivity recovered from the G-100 showed a total corticosterone concentration of about 8.3 x 10.13 moles in 0.2 ml with 37%.of that bound to a macromolecule. To 59 0.2 m1 of the same labeled NaOH extract 24 mg of Florisil was added and the solution passed through G-100. The amount of steroid bound was unaffected while the free steroid was reduced to only 71% of its pretreatment level. A 29% reduction in free steroid did not justify the rou— tine use of Florisil to remove free corticosterone-H3 from solution. Next, the removal of steroid from buffer was studied to eliminate any effect of the NaOH (high pH) used in the first experiment. Regular and heat-activated Florisil were added to TMK containing corticosterone—H3. The Florisil was mixed with the solution and allowed to sit for one hour at room temperature. Aliquots of that solution were counted and the data obtained is shown in Table 5. The data of Trapp and West (106), included in the table for comparative purposes, was obtained by mixing hydrocortisone in a phosphate buffer with dextran-coated Florisil and shaking one hour at 4°C. Under these condi- tions, 80 and 90% of the steroid in a given volume was adsorbed by 40 and 80 mg Florisil, respectively (106). In contrast, the plain Florisil used in this study, which presumably should bind more steroid than dextran-coated Florisil, bound only 23 and 62% of the available steroid 6O .musuxee cofluommu mo mEDHo>in coma oooa m.o om smumoo reams cam memuev coma oooa m.o ca :amuuxmc mucuauuououcsm 00.0 s.m o.a oa swumum mm.m s.~ m.H oom 00.0 5.0 0.H 0.0a umHS0mm mconmumooeuuoo mm OE E 05 I H mauoa x0 I: osuos xv AH V A v x o omnuomom coHumuucwocoo >uflucmoo m we Houmum saoumum «assao> ceoumum c. . Hamauon .mumo AO0Hv ummz 0cm momma Qua3 consummeoo 0cm Hawauon >9 Umnuomwm UAOUfiUHOUOUSHm mo mucsoemll.0 mamas 61 on 15.6 and 300 mg, respectively. The dextran—Florisil, used by Trapp and West (106), although it was removing such a high prOportion of steroid, was leaving a higher concentration, 400 X 10_10 M, of unbound steroid than that used in this study as the initial concentration. There- fore, Florisil may not be able to adsorb much steroid from the solutions of very low steroid concentration used in this study. Because of that lack of high affinity binding in Florisil higher proportions of unbound steroid are left in solution and the method becomes unsuitable for use in this study of high affinity binding from dilute steroid solutions. The efficiency of adsorption methods and the desire to utilize them for faster analyses led to exper- iments with other adsorbents. In one experiment, 30 mg granular activated charcoal added to 1 m1 TMK containing 1.66 X 10-10 M corticosterone removed 97% of the free steroid within five minutes. In another experiment dextran-coated (0.0125 g activated charcoal, 0.0015 g dextran) charcoal also showed rapid removal of steroid from solution. The coated charcoal was centrifuged out of suSpension but usually contaminated the end of the 62 pipette used to withdraw a sample. This contamination led to highly erratic data. A relatively large amount of steroid in the 50- 80%.saturated ammonium sulfate fraction went through the charcoal filter on first passage and should represent the large amount of steroid—binding protein present in that fraction. That observation stimulated further interest in the highly (over 50%) saturated ammonium sulfate fraction and further study of that fraction will be re- ported in other sections. Coating of charcoal is a means of minimizing protein adsorption to the charcoal. Three different coated charcoal preparations were tested in this study (Table 7). One ml of each coated—charcoal suspension was placed on a Millipore filter. Test solutions were the NaOH extracts of 750 x g and 30,000 x g fractions and resuSpended 0-50 and 50-80% saturated ammonium sul— fact fractions (from 30,000 x g supernatant). These were incubated with corticosterone and passed through the filters. In Table 7 both protein and label recovery after passage through dextran-coated charcoal are ex- pressed as percent of the original amount. 63 s.n0 v.0 0 ommmmmm 0cm 0.H0 s.m 0 m0mmmmm uma coeuomum mummasm EDHCOEEM $00I00 v.50 s.m 0 coeuomum mummasm EsacoEEm.$0010 ucmumcummsm 0 x 000.00 m.mm a.~ 0 uumuuxm momz m x ooo.om H.00 0.0 0 $0Mmmmm 0:0 «.00 0.0 0 00mmmmm umH uomuuxm momz 0 x 00h s.mm H.H m.a s29 ca amm ma mom 0.00 0.0 0 $29 AHMHUHQH M0 “New A: OHIOH v3 Adv: om>oawn coaumnucmocoo mEs~o> Hmfluumo owoumum oHOHmum .Hmuaflm muomfiaaaz a 0 m co Hmooumno owum>fluom 0 H.0 smsounu common wcoumumoofiuuoo mo coeumuomomll.0 mamas 64 .kuHHm co HE H .uaueau maze as ea .aauuxmu m mo.o .Haouumau H pauoz m mauumuaase m>auusuamzeaaua .HmUHHM co HE H “M29 HE om .m>m «H .Hmooumco H uHHoz 0 o>Hm««« co HE H «M29 HE 0H .cmnuxmo 0 H0.0 .Hmooumco H00 comma 0 msucmulmoucete .HmuHHM co cmnwmmH Hmoonmco cmcousu oommmm coHuumum room no HE 039: r o.ee m.ma s.mH Hs\ms mmm.a aauuoua b.0H 0.0 v.H 20HI0H x N vHouwum cOHuomnm muMMHSm EchoEEm 000:00 -1- m.m m.NH Ha\ms mmm.~ tampons 0.0 0 0 20HI0H x m oHoumum GOHuomnm quMHom EchQEEm 000:0 ucmumchmmsm 0 x 000~00 0.0H H.H~ 0.0 HE\0E 0vH.H cHououm 5.0 o m.H Seance x m saoumum uomuuxm momz m x 000.00 0.0 ¢.Nm 0.0 HE\0E 000. chuoum «.0 0.0 0.0 20H|0H x N oHoumum uomuuxo momz m x oms IIIIIIIIIIIIIIIIIIII HmsHOHHo no a Ilullllnlrllnulnnlllt wumuuHHM muMMUHHm muMHUHHm GOHumuucmocoo .«coauomum «tatH UHHOZIcmuuxmo «toonmoucmnuxwo HmHuHcH . «*«H uHHOZIm>m .HmuHHM whomHHHHz : 0 o co mGOHumnmmmHm HMOOHMSOIomumoo mwhnu SODOHSU 00Mmmmm an cwwuoum 0cm oconmumoowuuoo mo mmoqll.h mqmda 65 Some steroid did get through the charcoal and probably represents bound corticoid but a high percentage of protein was lost from every test solution. The very low levels of steroid recovered in the filtrates are best explained, as in the prior experiment, by loss due to protein adsorption to the charcoal. Less protein and more steroid was adsorbed by the PVP—coated Norit than the dextran—coated Norit in all cases but all three of the coated charcoals adsorbed too much protein to be of value in routinely separating free from bound steroid. A relatively large amount of label from the 50-80%.saturated ammonium sulfate fraction passed through the Norit 1-dextran. This was encouraging but the protein adsorbed to the charcoal may have contained more steroid-binding protein than was passed through. Charcoal could not be directly used to separate free and bound steroid; however, charcoal's high affinity for steroid combined with dialysis could offer a method for good separation. The problem with using dialysis for the separation of bound vs unbound steroid is that the time required is so long that dissociation occurs and a true measurement of bound steroid is not obtained. The charcoal should adsorb any steroid dialyzing out of the 66 sac and would not be in contact with the protein and the dialyzing medium would remain essentially zero steroid concentration, hence, speeding dialysis. The ability of charcoal to aid in dialysis was assessed by measuring the removal of corticosterone from a TMK solution. Another control could have been a sac with corticosterone-TMK dialyzed against TMK without charcoal. However, this would only show an approach to equilibrium and would not be very meaningful since the objective was to remove all unbound steroid from the sac. A 2.5 cm flat width Visking sac containing corticosterone-H3 and TMK was dialyzed against TMK and suspended charcoal. The percentages of the original steroid remaining inside the dialysis sac at various sampling times are shown in Figure 1. Included in Figure 1, along with the corticosterone-TMK data, is a 750 x g supernatant fraction dialyzed under similar con— ditions. Figure 1 shows the fast removal followed by a steady though slow removal of steroid after 8 hours dialysis. The slow phase was noticeable in both the buffer and 750 x g supernatant sacs. There is an obvious break in the curve at 8-9 hours and the steroid remaining 67 Fig. 1.--Loss of corticosterone-H3 from solutions dialyzed against charcoal in suspension. Ten ml TMK or 750 x g supernatant containing 8.3 X lO'lOM corticosterone-H3 dialyzed against 300 m1 TMK containing 2 g charcoal in suSpension. o——o TMK r——0 750 x g supernatant 68 L 80 0 r0 mconmmeOHuuoo r 0 O 4. 2 HmHUHCH 00 X 24 16 Hours dialyzed Figure l 69 should represent bound steroid. Eight to twelve hours of dialysis was therefore used for binding data obtained by dialysis against charcoal suspension. Two NaOH ex— tracts and two ammonium sulfate fractions were incubated with 2.1 x 10-10M corticosterone and then dialyzed against TMK containing activated charcoal. Samples were taken at 0, 0.5, 1.0, 7.0, and 9.0 hours and the results are shown in Table 8. After 9 hours dialysis the steroid in the fractions was assumed to be predominantly in the bound form. The 9 hour dialyzed fractions were then passed through charcoal layered on a filter to further study re— moval of bound steroid from solution. Both techniques suggested binding in the 50-80% saturated ammonium sulfate fraction. Removal of label from the dialysis sac containing the 50—80% saturated ammonium sulfate fraction was slower than in any of the other fractions. When the material dialyzed 9 hours was passed through either 0.02 g Darco S51 charcoal and 0.02 g dextran or 0.02 g Norit l charcoal and 0.02 g dextran on a Millipore filter, all the label remaining in the dial- yzed solution was removed except in the 50—80% saturated ammonium sulfate fraction. In that fraction 1.4 and 1.6% of the prefiltration label passed through the Norit- and I011 in! ”inxiflwdu \flnu onanvhoumauvwv q. «dzHfivnv HNPV H €~>flu2~flfi~§ll q 3 u.~.-~ (.3 7O 0.HH 0.0H H.0n 0.00 00H soHuumuw muMMHsm EchoEEm $00100 0.¢ 0.0H 0.0m ~.n0 00H coHuomum muwMHsm EDHCOEEmAXOmuo ucmumcnmmsn 0 x 000.00 0.0 H.0H m.s0 v.am oos uomuuxm moms m x ooo.om a.e a.oa H.0e 0.0m 00H uomuuxm momz m x oms IIIIIIIIIIIIIIIIII HmcflmHuo «0 x.:::::::::::::::::: 0.0 0.5 0.H 0.0 0 vmuhHMHU mason COHuomum .conchmsu Hmooumso umchmm uHmmHmHo >3 mcoumumooHuuoo mo Hm>oEomll.0 mamas 71 Darco-coated filters, respectively. If there was any binding in the other three fractions the protein was all adsorbed by the coated charcoal but the 50-80% saturated ammonium sulfate fraction either had enough binding pro- tein so that some got through or its binding protein was less well adsorbed. Equilibrium Dialysis One of the most sure methods for obtaining valid measurements of extent of binding is equilibrium dialysis, although the method is somewhat time-consuming. The bind- ing material to be tested is placed in a dialysis sac in a solution containing steroid. When equilibrium becomes established between free steroid outside the sac and free steroid inside the sac, the steroid concentration outside can be subtracted from the total concentration inside to determine the concentration of bound steroid. Dissocia- tion need not be considered and the extent of binding at a given steroid concentration is known. To test equilibrium dialysis 5 ml samples were 11 dialyzed against 500 ml TMK containing 1.66 X 10- M corticosterone. Aliquots were withdrawn between one and 72 twenty hours, counted, and the data plotted in Figure 2. The data show apparent sampling error as evidenced by the variation in the buffer curve. The 0-50% saturated ammonium sulfate fraction curve showed little binding but the 30,000 x g NaOH extract showed more extensive binding with approximately the same shape as the 0-50%»saturated ammonium sulfate curve. Since the curves showed only slight increases between 8.5 and 20 hours, 12 hours was chosen as the standard time for cessation of dialysis. The 50-80%,ammonium sulfate fraction curve showed a relatively large increase between 8.5 and 20 hours. The increased time required for saturation of that frac— tion could be a function of protein concentration in that a high concentration of binding protein in the sac would take longer to saturate, due to dialysis being rate— limiting, than a lower concentration. More dilute solu- tions should allow for equilibration within 12 hours. 73 Fig. 2.-—Rates of corticosterone uptake by three cell fractions during equilibrium dialysis. Six m1 of each sample with all samples dialyzed together against 500 ml TMK containing 1.66 X 10’11M corticosterone-H3. One half ml samples withdrawn after 1.0, 1.5, 3.5, 8.5, and 20 hours dialysis. r——o TMK buffer ———— 0-50%.saturated ammonium sulfate fraction o——o 50-80%,saturated ammonium sulfate fraction A——A 30,000 x g NaOH extract 240 cpm/ml 80 74 8 12 Hours dialyzed Figure 2 16 20 75 Isolation and Purification of Steroid-Binding Proteins Stability The influence of a reducing agent, mercaptoethanol, on stability of binding was determined in one experiment. Homogenates with and without 5 x 10-6 M mercaptoethanol were incubated at room temperature for 60 min in the 0 . . M corticosterone—H . Two fractions, presence of 8 x 10- the 750 x g and 30,000 x g particulate fractions, were assayed for incorporated label. Duplicate control and treatment determinations gave the data shown in Table 9 as percent of total recovered label found in each frac- I: ion. The mercaptoethanol-treated material'had 18 and 2 4% more label than the control for the 750 x g and However , when the 3 0.000 x g fractions, respectively. da ta was adjusted for total protein content of the homo— qenate from which the fractions came, the treated material 81'1<>wed 2% fewer cpm per mg protein per cpm recovered. It was concluded that mercaptoethanol did not appreciably affect binding in either the 750 x g or 30,000 x g frac- t e 2‘ Qns and the binding material was apparently stable w- k "e hout an added reducing agent. 76 .HQOH 00Hm>oomu Hmuou mo ucwouom mm 00mmmumxw 0CHUCHQ mumHooHuumms OH X mfv OH X mi? fiOHOerUTH 2&0 a: an tampons uswzdo cm.m mm.e m x 000.00 oo.m .os.a m x ems Hocmnumoummonmz .. . Houucoo coHuumum .nsoHuomum HmHoHHmonm mumHsoHuqu ou mmlmconmumooHuuoo mo 0CH0CHQ co HocmSumoummoumE mo uommmmul.0 mamma 77 To determine stability of the binding material during freezing and thawing, a 2 ml aliquot of resus- pended 30,000 x g fraction from lactating rat mammary M corticosterone-H After was incubated with 1.7 X 10- recentrifugation and extraction of the 30,000 x g sedi- .ment with 1 ml 0.05 N NaOH the extract was frozen for riine days. Some of the same extract was frozen and thawed twice and both were passed through Sephadex G—100. CDlairty-seven percent of the label in the NaOH extract vvaas bound, i.e., appeared in the G—100 exclusion peak, kaeafore freezing the extract. Frozen storage for 9 days ]_eeft 33.4% of the steroid bound and the twice frozen and tilaawed material showed 34% bound, hence, freezing had IL.j.tt1e effect on stability of binding in the NaOH ex- tract. Stability to freezing and thawing was also c1"l-ecked in the ammonium sulfate fractions prepared from t]b1€3 30,000 x g supernatant. The ability of those frac- tzi-<:Mns to bind corticosterone, determined by equilibrium djiuaa_1ysis. was not diminished by months of frozen storage 0:7. :repeated freezing and thawing. Apparently the bind- i C‘SEEF material was very stable during storage at -10°C. 78 Extraction from Particulate Fractions Release of the binding material from particulate fractions was attempted by NaOH extraction. A 27,000 x g fraction representing 2.5 g bovine mammary tissue was "extracted" first with 2 ml 0.1 M phosphate buffer (pH 7.4), then with 2 m1 of 1:1 phosphate buffer: 0.1 N IQaOH. One half ml of each extract was incubated with $1.7 X 10.ll moles of corticosterone-H3 and then passed t:hrough G-100. The macromolecular and free steroid frac- t:ions from the G-100 were counted to determine the per— czentage of label in the macromolecular peak, i.e., bound. (Dnly 1.7% of the total recovered label was bound in the iEDhosphate buffer "extract” while 7.7% of the label in the NaOH extract was bound. The four and one half times more binding seen in the NaOH extract was assumed to represent ITlllch better extraction of binding material by NaOH than 1;>Iil7.4 phosphate buffer. EigHQLL§tribution of Binding ‘-EEE;9ng Cell Fractions Corticosterone binding was studied in a number C) , _ :IEE‘ cellular fractions. Those studied were the 750 x g 79 pellet containing mostly nuclei, the subsequent 30,000 x g pellet containing mitochondria and microsomes, the subse- quent 100,000 x g pellet, containing microsomes, and the 100,000 x g supernatant. In many cases the 100,000 x g centrifugation was neglected and the 30,000 x g super— natant used instead of the 100,000 x g supernatant. The distribution of binding activity in partic— iilate fractions was studied in a homogenate prepared from frozen lactating—bovine mammary gland. The homogenate vvas prepared in 0.1 M Tris (pH 7.5) buffer and incubated vvnith approximately 2.5 X 10_8 M corticosterone. After :iqncubation, the 600 x g, 30,000 x g, and 100,000 x g particulate fractions were centrifugally produced, washed <:>lace, and measured for radioactivity. The data are shown :i.ra Table 10 and are expressed as both the percent of total Particulate radioactivity found in each particulate frac- t: chan and percent of total recovered activity (includes L C30,000 x g supernatant activity) in each particulate firm:‘Eaction. Some of the label in the particulate fractions c2‘:>’1.11d be contamination because of the single pellet wash; 11% race, not too much emphasis should be put on the data. Emery (96) also studied distribution of corti- C<::: aEsterone binding in particulate fractions. He expressed 80 vo.o om.H om.a H.Hm x ooo.ooa ma.o mm.a wh.a v.nm x ooo.om mH.o mo.m v>.o H.Hv x com .m.m H M .m.m H Wm Aucmumcummom moUSHUCHV mcofluomum mumHooHuumm cofluomum Umum>oomu 3an :33 mo s. CH omu0>oomu dogma Hmuou m0 X .mcofiuomum onwasofluumm couscoum adamm Ismauucoo .ansHHmoQSm woman mcoEm mmlmcoumumooflunoo mo cowuzflwhwmenI.QN QNQN% 81 his binding as percent of added activity and found a range of 3.7-6.0% bound in the nuclear fraction (1,000 x g pellet) from rat mammary homogenate. The 30,000 x g pellet from that homogenate contained 0.9 to 16.6% of the added steroid. The 30,000 x g precipitate from bovine :mammary tissue bound 6.9 to 7.1% of added steroid. Emery IJsed lO-6M corticosterone vs 2.5 X 10— M used in this eexperiment and the 40—fold higher concentration he used nnight be reSponsible for much of the difference between the two sets of data. Further study of particulate steroid-binding was Cicone under different conditions. A homogenate was pre— jEDEared in TMK from oxytocin-penicillin-treated frozen J_Eactating-bovine mammary tissue. It showed much more k:rjmnding activity than the previous homogenates prepared i—Ii 0.1M Tris buffer. At 8.3 X 10-9M corticosterone con— centration the 750 x g, 27,000 x g and 100,000 x g frac- tli-CDns bound 9.3, 3.9, and 0.3% of the recovered label, re Spectively. The percents of the total steroid bound 1‘71 the total particulate fractions were 69.2, 28.7, and 2" :1~ for each of the three fractions. Those values are qL“;jL—te different from the 41.1, 27.4, and 31.1% found Pt? , . ‘EE=”Viously. The apparent differences are the treatment 82 of the tissue before freezing, the molarity and the pH of the buffer, better pellet rinsing and the use of less than one half the previously used corticosterone concen- tration. An experiment was performed to compare total particulate steroid binding at four different corticos- ‘terone concentrations. Aliquots of homogenate were in— czubated at 0.83, 1.66, 3.32, and 4.98 X 10-9M corticos- tzerone, centrifuged 27,000)ellets, washes, and supernatants counted. The washed $275000 x g precipitate from the 0.83 X lO-gM corticos- ‘t:erone incubation contained 10% of the total recovered Zlflabel. The 10% represented the most efficient (highest percent of recovered label) binding among the four steroid <::<3ncentrations and indicated that higher prOportions of Steroid are bound at lower concentrations. None of the foregoing experiments considered k:’lending activity which might be present in the 100,000 X g 3 L:l‘pernatant. This soluble, or cytOplasmic, binding was Silfiltbwn by passage of labeled 30,000 x g supernatant through G-._ . 43100. From the label in the G-lOO macromolecular peak t hQ amount of binding due to the 100,000 x g particulate f3:=> Qfiaction must be subtracted in order to determine the 83 soluble binding. Some experiments which could be used to calculate the percentage of particulate binding in the 30,000 x g supernatant had been performed. One experi- ment, using a homogenate incubated with 2.5 X lO-BM corticosterone had shown only 1.5 to 1.7% of the label in the 30,000 x g supernatant to be due to the 100,000 x g pellet. In another experiment the 100,000 x 9 fraction from homogenate incubated at 8.3 X lO-9M corticosterone contained only 0.3%.of the 30,000 x g supernatant label. That represents one-fifth the binding activity of the first experiment at one—fifth the steroid concentration used in the first experiment. This observation was ap- plied to the measurement of the binding protein in a 30,000 x g supernatant of a homogenate which had been 1OM corticosterone and passed incubated with 1.8 X 10- through G-lOO. The macromolecular peak from the G—lOO had 7.9% of the label recovered from the column. The work with the 100,000 x g pellet had shown that only a small prOportion, less than 1.7% of that 7.9% from the G-100 would be due to 100,000 x g sedimentable material. Therefore the 7.9% of the 30,000 x g supernatant label found in the G-lOO exclusion peak apparently represented 84 considerable binding in the "cytoplasmic" (100,000 x g supernatant) portion of the mammary cell. Ehirification of Solubilized Binding Protein Studies which are designed to determine the phys- ical characteristics of a binding protein require that the protein be isolated in pure form. The purpose of tzliis study was to show the existence of a glucocorticoid binding protein in mammary gland and to study some of its binding kinetics. For such purposes a totally pur- i fied preparation is not required and the purification aNttempted and accomplished during this study was incom- Plete. The successful fractionation, purification, steps treeloorted here could serve as the initial procedure for a total purification. Fractionation was performed on sol— 1'1.<':>.‘.I.e cytoplasmic proteins or NaOH-solubilized particulate I:‘:":'::‘\u3 wavy Em a .mmlmcoumumoofluuoo SDNB chMQSUCH >Hm50w>mum cowuomnm mumHsowuumm m x ooo.om m mo uomuuxo momz pwuhamflp map on cocoa Hem wumnmmosm EsHono mo mSOHumuucmocoo mcflmmouocw Eoum muw>fluomoflpmn cam camuoum mo >um>ooom|n.aa mamfia 95 The 30,000 x g supernatant was dialyzed against 2mM KZHPO4 (pH 7.6) containing charcoal. After dialysis 0.36, 0.68, 0.96, 1.22, and 1.45% calcium phOSphate was successively added to 7.1 ml of the dialyzed supernatant. Each frac- tion was then extracted twice with 1 ml 0.5MLK2HPO4 to elute adsorbed protein. Those protein concentrations were determined by 260/280 adsorbance and used to calculate the Specific activities shown in Table 12. As in the calcium phOSphate fractionation of the NaOH extracts, nonbinding protein was removed by the first additions of gel while the gels in concentrations over about 1% ad- sorbed steroid-binding protein. The alternative approach to calcium phOSphate fractionation was used to fractionate an ammonium sulfate fraction. A relatively large quantity of gel was added to the ammonium sulfate fraction and then the protein was selectively eluted from the gel. For that procedure a dialyzed 70-80% saturated ammonium sulfate fraction from a 30,000 x g supernatant was used. One ml of that frac- tion contained about 1.2 mg protein. To 1 ml was added 0.037 g (dry weight) calcium phosphate gel in 1 ml. After incubation the gel was rinsed three times with 2 ml of 1 mM KZHPO4 (pH 7.4) and successively extracted three 96 ®.Nom m.hhm mv.a ©.th m.h¢v NN.H o.NOH ¢.mHN mm.o w.mm m.om m©.o ¢.ON 0.Hm mm.o IIIIIIIIIIIII camuoum m£\an IIIIIIIIIIIIII AmESHo>\u3 mnpv m uomuuxm H uumuuxm 3m x .mmumcoumumoofluuoo SDHB pmumnsoca ucmumcuwmsm m x 000.0m ooamflmH ou poppm Hom mumnmmonm ESHono mo coflumuucwocoo mafimmmHUCH Eoum Gaououm m5 mom hufl>fluomowpmn mo >um>oommII.NH mamas 97 times each with 2 ml of 50 mM KZHPO4 (pH 7.4) and 2 ml of l M KZHPO4. Portions of each extract were subjected to equilibrium dialysis at 1.66 X lo-llM corticosterone-H3 and aliquots counted. Using the 260/280 absorbance pro- tein determinations on the extracts before equilibrium dialysis, the cpm/mg protein for the material not absorb— ing to the gel and the 50 mM and l M extracts were 130, 183, and 110, respectively. Apparently there was little further fractionation of the 70-80% saturated ammonium sulfate fraction by the calcium phOSphate gel. Disc gel electrOphoresis was used to monitor the degree of protein separation by calcium phOSphate gel fractionation. A dialyzed 60-80%.saturated ammonium sulfate fraction was fractionated on calcium phOSphate in the manner previously described for the 70-80% frac- tion. Equilibrium dialysis of the gel supernatant and extracts showed 15, 229, and 35 cpm bound per ml. With that distribution, if the binding protein separates from other proteins into a single band upon electrophoresis and equal volumes are electrophoresed, the 50 mM KZHPO4 extract (229 cpm/ml) should Show a band 15 times as dense as an equivalent band in the supernatant (15 cpm/m1) and 6 times as dense as the equivalent band from the l M 98 extract (35 cpm/m1). Only one band corresponded to that type of binding activity distribution. The material not adsorbing to the phosphate gel showed 11 bands and the 50 mM and l M KZHPO4 extracts of the gel Showed 12 bands. The bands were then compared to find the one corresponding to the concentration of binding activity placed on the electrOphoreSis gels. The 60-80%.saturated ammonium sulfate fraction, put on the gel in small quantity be- cause of its high protein concentration, had a very faint band at Rdye (Rd) 0.63. The material not absorbing to the gel had a band of Rd 0.57, the 50 mM extract a heavier band at Rd 0.60 and the l M extract had a light band of Rd 0.59. The 60—80% fraction also had a faint band at Rd 0.54 but the calcium phOSphate fractions had no hands closer to the 0.57-0.60 bands than Rd 0.47 and 0.66. The band seen between Rd 0.57 and 0.60 was therefore probably the same protein in each of the three fractions and the only one correSponding in density with the distribution of the bound steroid. Sephadex Sephadex, previously used to remove free from bound steroid, was also tested as a fractionation method 99 after ammonium sulfate, by passing 1.0 ml of the 70-80% saturated ammonium sulfate fraction through a 1.5 X 25 cm Sephadex G-100 column eluted with water. Timed fractions were collected and six protein peaks of three fractions each were obtained. Each protein fraction was subjected to equilibrium dialysis against TMK with 1.66 X lO-llM corticosterone-H3. After equilibrium dialysis Specific activities, along with the 280 nm absorbance profile of the Sephadex fractions, are shown in Figure 3. The binding protein apparently eluted from the column with the largest protein fractions. The highest Specific activity among the fractions never reached the 424 cpm/mg protein found in the original 70-80% saturated ammonium sulfate fraction and although there appeared to be some separation of proteins the loss of Specific activity negated any gain from their separation on Sephadex. DEAE-cellulose DEAE-cellulose ion exchange chromatography was also used to fractionate the 60-80% saturated ammonium sulfate fraction. One ml of the 60-80% fraction was put 100 Fig. 3.--Sephadex G-100 fractionation of 70-80 ammonium sulfate fraction. Absorbance (280 nm) profile of 1.0 ml 70-80% saturated ammonium sulfate fraction passed through a 1.5 X 25 cm G-lOO column equilibrated and eluted with water. Six combined fractions then equilibrium dialyzed against 1.66 X 10-11M corticosterone- H3 in TMK and cpm bound per mg protein calcu- lated after dialysis. 101 m musmwm mcofluomnm o m w m ON ow om om cwmun um mfi\Emo OOH AS: omN. mocmfluomnm 102 on a 10 ml DEAE-cellulose column equilibrated as described in the Material and Methods section. The fractions were then eluted with a gradient from 1 mM KZHPO4 (pH 7.4) to l M KZHPO4. Timed fractions were collected and combined to form the six fractions shown in Figure 4. These frac- tions were subjected to equilibrium dialysis against TMK with 1.66 X lo-llM corticosterone-H3. After equilibrium dialysis Specific activities were calculated using Lowry protein determinations for protein concentrations. The specific activities and 280 nm absorbances are shown in Figure 4. Separation of the binding protein was better on the DEAF-cellulose than on the Sephadex since there was less overlap between fractions and much higher Specific activity was found by DEAE separation. The highest Spe- cific activity found after DEAE separation represented 2.3 times that of the 60-80%.saturated ammonium sulfate fraction from which it came. Immunological comparisons Glucocorticoid—binding material had been found in both NaOH extracts of particulate fractions and in the 103 Fig. 4.--DEAE-cellulose fractionation of 60-80 ammonium sulfate fraction. Absorbance (280 nm) profile 1.0 ml 60-80%lsaturated ammonium sulfate frac- tion eluted from 10 ml DEAE- cellulose by a gradient of 1 mM pH 7. 4 KZHPO4 to l M K2 HPO Six combined fractions then equilibrium 2diag- yzed against 1. 66 X 10' 11M corticosterone-H3 in TMK and cpm bound per mg protein calculated after dialysis. 104~ v musmfim mSONUUmum o m -v m N - H » ow H o om N.o ONH m.o 00H «.0 OON m o camuoum mE\EQo AS: omN. mononuomna m.o O¢N 105 ammonium sulfate fractions from 30,000 x g supernatant. The binding materials could conceivably be the same pro- tein due to adsorption of soluble, cytOplasmic, protein to the particulate fractions. The possibility of such cross contamination was examined by immunological tech— niques. Rabbits were immunized against the 60—80%.sat— urated ammonium sulfate fraction and their serum later reacted against both the 60-80% fraction and the 30,000 x g NaOH extract. Though the precipitin bands on the Ouchterlony plates were broad and smeared, much of the reaction area was similar and some identity between precipitin bands could be assumed. At least one precip- itin band present in the 60—80 fraction was absent in the NaOH extract, otherwise the reactions were indis- tinguishable. The similar reactions indicate the possi- bility that "particulate" binding was due to adsorption of soluble binding protein to the particulate fractions. Binding Characteristics Competition Studies Binding of a steroid hormone to a protein is .relevant only if some degree of Specificity for that 106 hormone can be demonstrated. Competition experiments are therefore very important to studies of a particular bind— ing protein. Such experiments were performed in a number of ways on the glucocorticoid binding protein under study. The competition for corticosterone binding sites by various steroids was very difficult to study during the initial portions of this study. In the first attempts, homogenates were incubated with 1.7 X lO-BM corticosterone- H3 and very high (more than lOOO-fold higher) concentra- tions of progesterone, hydrocortisone, or corticosterone and the three particulate fractions produced from those homogenates were assayed for incorporated label. The results are shown in Table 13 as percentage of total re- covered label found in each of the particulate fractions. Neither progesterone nor hydrocortisone would compete with corticosterone, however, unlabeled corticosterone also did not dilute the amount of label bound. The re- sults led to the concept of "infinite," or unsaturable, steroid binding. Another experiment was performed with 2 m1 of 1:4 diluted homogenate incubated with a constant amount, 4.9 X lO-gM, of corticosterone—H3 and concentrations of . l . . - corticosterone-C 4 increa51ng from zero to 9.9 X 10 7M. 107 .Hmnma ownw>oowu Hmuou wo Ram .mmImcoumumooHuuoo mSHm wcouwumoofluuoo UmHQOHCS I 0+0 .mcomfluuoooucmn I Om .wcoumummmoum I m .mmecoumumOUHUHOU I Us m.H e.H e.H m.H m.H m.a m.H m x ooo.ooa e.H «.4 m.a H.H 0.0 a.H o.H m x ooo.om o.~ ~.N m.H e.~ s.H H.m 4*H.m m x 006 0+0 um O on U m *0 m N H cofiuomum ucmsflummxm .mcoumumoowuuoo Umawnmacs pom wcomfluuoooupms .mcoumummmoum >3 :coflufiummeoo: cam mcofluomuw mumHsofluumm oucfl mmImcoumumooHuHoo mo coflumuomnoocHII.mH mamda 108 Double isotOpe counting allowed calculation of the total corticosterone bound. The uu moles corticosterone re- covered in the washed 750 x g and 27,000 x g fractions are shown in Figure 5. The total up moles steroid re— covered are indicative of the increasing steroid added and the steadily increasing uptake of label with increas- ing initial steroid concentration is obvious. The assumption was then made that there were at least two types of binding sites with different affin— ities for corticosterone. Therefore, the only way to Show good competition between labeled and unlabeled corticosterone would be to work in a narrow, low-steroid concentration range where the high affinity, "finite," "saturable" binding sites give a curvalinear uptake of steroid rather than at higher steroid concentrations where the apparently unsaturable, "infinite” binding is linear and tends to mask competition. Before competition could be shown between corticosterone-H3 and other steroids it was necessary to demonstrate competition between corticosterone-H3 and its nonlabeled analog. Two experiments using low steroid concentrations and dialysis against charcoal were done to show competi- O I I 3 tion between corticosterone and corticosterone-H . In 109 Fig. 5.—-Corticosterone bound in two particulate cell fractions incubated with four levels of corticosterone. Two-tenths gram homogenized tissue in 2 ml TMK incubated with corticos- terone (—H3 and —C14) 60 min at room temper— ature. Particulate fractions washed three times each and corticosterone extracted with ethanol for counting. r——v 750 x 9 fraction o——o 27,000 x g fraction 110 Acofluumum m x ooo.s~c mason mcoumumoowuuou mmHoE 1: 0H vN Nm m musmflm Umuo>oomu mcouopmoofluuoo moaoe :1 Hmuoa coma OOOH oom 00¢ NH 0H Acofluomum m x omhv canon mcoumumoofluuoo moaoe 1: 111 the first experiment three 2.5 cm dialysis sacs, each containing 5 ml of a 1:4 dilution of a 750 x g superna- tant or TMK, were suspended in 300 ml TMK containing 1 to 2 g activated charcoal in suSpenSion. Conditions for the second experiment were the same except that each sac contained 10 ml of diluted 750 x g supernatant. Each sac, within the same experiment, contained equal concen- trations of corticosterone—H3. One sac contained only TMK while two sacs, one of which had unlabeled corticos- terone added, contained the diluted 750 x g supernatant. The levels of labeled and unlabeled corticosterone added initially and steroid remaining in the sac after 8 hours dialysis are shown in Table 14. During the experiment samples had been removed from the sacs during the first hour of dialysis in order to check the rate of dialysis. Those samples showed that equal percentages of the initial radioactive label were being dialyzed out of each sac although the "high steroid" sac had much more steroid per unit of label. The sac with the highest steroid concentration lost 5 X 10-11 moleS corticosterone per hour averaged over 8 hours. Because equal percentages of label were dialyzed out of the sacs the percentage of label left in the sac without 112 00.0 mv.o o ASOHIOH xv pflouwum canon n: 0 em.NN m¢.o mv.o AZOHIOA xv Ufloumum canons: HQ m m.s m.oH ~.m ooa x as o\us m vv.Nm Hm.o mv.o AEOHIOH xv coflumuucmocoo pflouwum a: 0 see m.m m.m Axoanoa xv coflumuucmocoo pfloumum u: o N ucmEHHmme oa.hN mm.H o AZOHIOH xv pflonmum 0:509 u: m 00.00 vN.N vN.N AEOHIOH xv Uflouwum pcsoncs HS 0 s.m m.oa 0.0 ooa x up o\ur m 0n.>0 nm.m «N.N ASOHIOH Xv coflumuucmocoo pflouwum us 0 mom mm mm ASOHIOH x0 coflumuucmocoo oaoumum up 0 H ucoefluoaxm N ucmumcnmmsm H ucmumcummsm M29 .Hmooumno 0cm xza umcflmmm Umnhamflp ucmumcummsm m x 005 0cm M29 Eoum wcoumumoofluuoo mo mmoqII.vH mamma 113 750 x g supernatant after 8 hours dialysis was indicative of the unbound steroid in the other two sacs. Therefore an amount of steroid proportional to that in the TMK (no 750 x g supernatant) sac could be subtracted from the residual steroid in the other two sacs in order to cal- culate bound and unbound steroid in those sacs. If the steroid bound increased prOportionate to steroid concen- tration all four sacs containing binding protein would retain the same proportion of initial steroid; i.e., all four 8 hr/D hr ratios would be identical. The fact that the two sacs with the highest steroid concentrations both retained a lower percentage of initial steroid showed that binding sites were being saturated (corticosterone molecules were competing with each other for binding sites). Increasing the steroid concentration should, but did not, give steadily decreasing retention of steroid prOportional to initial steroid concentration. The pro— portions of label retained at increasing steroid levels were 10.9, 10.8, 7.3, and 9.7%. There is no apparent reason for such a discrepancy. The percentage of initial corticosteron-H3 re- maining after 8 hours dialysis clearly showed competition 114 between corticosterone and corticosterone-H3. However, if available binding sites had been saturated with steroid at the low concentrations there Should have been 27~ and 53-fold decreases in label retained in EXperiments 1 and 2, reSpectively. The differences were not nearly that large indicating either failure to saturate one type of binding site or saturation of one type of site and partial saturation of other sites having lower affinity for corticosterone. Successful demonstration of competition, though not of the magnitude expected, by dialysis against char- coal led to a dual label competition experiment with hydrocortisone, progesterone and estradiol. Those hor- mones, each labeled with C14, were added to supernatant of a 750 x g centrifugation. Corticosterone—H3 was then added and the mixtures dialyzed against TMK and charcoal. After dialysis aliquots were counted by double label techniques and total steroid concentration calculated (Table 15). If corticosterone competed equally with any of the other steroids it should have displaced one half of that steroid from its binding sites at equal steroid concentrations and Cl4 and H3 should be dialyzed out at 115 .Hoflpmuumm n m .mconmummmoum u m . mImcoumumooHuHoo u U .mCOmHuHoooupms H Um* m 0.H 0.0 m.o m.m 0.0 m.O HMHUHCfl MO R Nvo.o mm mH0.0 mm 0H0.0 m UHOHmum SOHIOH N mflmwamflfl ML ON HvN.N mow mmm.N mmm mm0.N omha GHOHmum ZOHIOH X mHmNHmHU HQ 0 U m U m U SUE m. UMW N Umm H 00m .MEB ca pmpcodmsm HMOUMMSU umcflmmm pmumamflp ucmumcuomsm m x 005 m Eoum Hoawmuumm 0cm maonmummmoum .mcomflunoooupwn .mmIocoumumoofluuou mo mmoqII.mH mqmfifi 116 approximately equal rates. If there was no competition, but each steroid was binding to a separate site within the 750 x g supernatant, more H3 and Cl4 should have been retained than in the sacs where there was competition. Partial competition should have been intermediate. Be- cause of the unequal concentrations of hydrocortisone, progesterone, and estradiol used in this experiment only the effect of these steroids on corticosterone retention was valid for comparative purposes. Three degrees of competition were observed with hydrocortisone showing the best competition with corti- costerone. Therefore hydrocortisone was apparently com— peting for corticosterone binding sites. Almost as much corticosterone was dialyzed out of the sac containing progesterone as from the sac with hydrocortisone. There— fore progesterone must have been competing although not to the same extent as hydrocortisone. Corticosterone retention was best in the presence of estradiol and al- though competition between the two steroids cannot be ruled out in this experiment, there apparently was much less competition from estradiol. 117 Observations of the Cl4 steroids bound Showed more estradiol bound than any of the others. The least binding was with hydrocortisone although it was present in the highest concentration. The difference could have been due to tenacity of binding or number of binding Sites available to those steroids in the 750 x g supernatant. The quantitative extent of competition for corti- costerone binding sites was determined with the 60-80% saturated ammonium sulfate fraction of the 30,000 x g supernatant using equilibrium dialysis. The steroids used in the competition studies were cholesterol, es- tradiol, progesterone, hydrocortisone, and corticosterone. They were tested for their ability to compete with corticosterone—H3 for binding sites. The equilibrium dialysis was performed with sacs containing 2 ml of diluted (1:9 with TMK) 60-80% saturated ammonium sulfate fraction from fresh lactating bovine mammary. The sacs were placed in test tubes containing buffer and corticosterone (total volume, 100 ml). Corticosterone-H3 was added to 8.3 X 10-11M . The quan- tities of cholesterol, estradiol, progesterone, hydro— cortisone and corticosterone added are shown in Table 16. After equilibrium dialysis aliquots of inside and outside 118 mmva O00mm O0Hmm o¢o¢v 00mmN m hmmv oamm o0m0 ovmn 0mmN m ave mmNN omam 050m mmN N m o O o O H IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII z HHIoH xuuuuIIIIIIIIIIIIIIIIIIIIIIIIII- mcoumumoofluuoo wcomfluuoooupwm mconmummmonm Hoflvmuumm HonmummHO£U # maze .ucmumcuomsm m x ooo.om mo cofluomnm mummasm EsflsoEEm pmumMSDMm.$omIo0 mo mammamflp Esflunflaflsvo ca mmImcoumumoowuuoo Eda OH x m.m nufl3 mummEoo ou owoom mocoEuon msoflum> mo mcoflumuuchSOOII.0H mqmfie 119 solutions were counted and the bound to unbound ratios calculated. A graph of the bound to unbound corticos- terone ratios vs total steroid concentration is shown for all the steroids in Figure 6. Corticosterone and hydro— cortisone both competed with corticosterone-H3 for binding sites. Progesterone competed to a lesser extent while cholesterol and estradiol either did not compete or competed only at very high concentrations. The competi- tion by progesterone was not unexpected since Tucker et a1. (97, 98) had shown progesterone competing with hydrocor- tisone in their cell culture binding studies. BindianConstants The nature of corticosterone binding at various concentrations is elucidated somewhat by calculation of binding constants from data relating the prOportion of bound to unbound steroid at each of several steroid con- centrations. The bound to unbound ratio is frequently plotted against the concentration of bound steroid. That relationship forms the Scratchard plot described in the Review of Literature section. 120 Fig. 6.--Competition for corticosterone binding sites by cholesterol, estradiol, progesterone and hydrocortisone. Two ml diluted (1:9 TMK) 60-80%.saturated ammonium sulfate fractions dialyzed in 100 ml total volume containing 8.3 X 10'11M corticosterone-H and different levels of unlabeled steroid. corticosterone cholesterol estradiol hydrocortisone progesterone 121 --_-. “a--------- 4O tration 30 n 20 M total steroid conce x 10”8 Figure 6 122 The first data which could be used for a Scatchard plot, although it was not designed for such use, was the data shown in Table 14. That data, although there were only two points in each eXperiment, was plotted by Scatchard's method and binding constants were determined as shown in Figure 7. The lines can be used only to indi— cate that one set of points represented higher or lower affinity binding than the other set. The Slopes of those lines are the binding constants. They indicate that the data in Experiment 1 was collected in the concentration range which involves low affinity binding rather than high affinity binding. The binding constant in Experiment 2 was more than 10 times lower and therefore indicated higher affinity binding. High affinity binding usually indicates specificity and as such it should therefore be easier to show competition between labeled and unlabeled corticosterone in Experiment 2. That was the case as the discussion of the Table 14 data indicated. The two experiments Shown in Table 14 were per- formed in similar manner and tempt one to draw a "curve" using all four points on the Figure 7 Scatchard plot. The dashed line in Figure 7 shows such a curve. The last point, the high steroid concentration point, should not 123 Fig. 7.-—Scatchard plot of Table 14 data. Two charcoal- dialysis experiments with the binding constant from each (k1 and k2) and the highest affinity binding constant (k1) from the combined data. 124 h musmflm 0COHOUMOUHUHOU UGDOQ S. OH VA can mm «m om 0H NH AZBIOA X 0.N u Hxv H ucofiquwmxm z I A muoa x m m 0 0 N0. 00. 2 OH x v. m on u #0 N ucmEflummxm . .00. Axoauoa x m 00 maoumumoofluuou ocsoncp\nssom 125 have been higher if the points had been obtained in the same experiment, therefore, the line from Experiment 1 should probably be displaced slightly downward. The initial lepe, between the 8.3 and 33 X lO-loM corti- costerone concentration points, estimates the high affinity binding constant if all the points had been from one experiment. Over that narrow range of initial corticosterone concentration the binding constant was 4.4 x 10'9M. A factor which must be considered in discussing the data in Figure 7 is that as dialysis proceeded the steroid concentration decreased inside the sac and dis- sociation was probably slowly occurring. With slow dis- sociation the binding observed was probably more dependent on initial steroid concentrations than the 8 hour concen- trations and the data was therefore calculated as such. The data in Figure 7 used the 8 hour steroid concentra- tions to determine bound steroid while the initial steroid was used as "unbound" to calculate bound to unbound ratios. In order to relate the bound to unbound ratios to initial steroid concentration that initial concentration is shown at each point on the figure. 126 The uncertainties of unbound steroid concentration in the charcoal dialysis method led to use of equilibrium dialysis to find binding constants under equilibrium con- ditions. The charcoal-dialysis binding constant calcula— tions (data in Table 14) were made using data from 750 x g supernatants: however, binding constants determined by equilibrium dialysis were for the 60-80% saturated ammon- ium sulfate fraction of the 30,000 x g supernatant. The equilibrium dialysis experiments were performed as de- scribed in the Material and Methods section. Two experi- ments were performed initially, each with tissue from a different cow. The steroid concentrations used, the concentration of bound and free steroid after dialysis, and bound to unbound ratios are shown in Table 17. When comparing the data in Table 17 with that on Figure 7 the most obvious difference between the charcoal— dialysis experiment shown in Figure 7 and the equilibrium dialysis experiments was the difference in the bound to unbound ratios. The most plausible explanation is that during charcoal-dialysis dissociation was occurring and any dissociation would lower the bound to unbound ratio when initial steroid concentration was used as the unbound steroid concentration. Therefore the binding constants 127 NHH. mmNm¢ ems. smmv Ame. ave Hem. see New. mam has. m.m s00. m.m mmm. s.a z Hauoa x z sauce x Dum acoumumoofluuoo Dum mcoumumoofluuoo N mfiwmflnh H GDWMflH. .mosmmflu mcfi>on ucmummmap Eouw mcoHuumum mummasm EchoEEm pmumusumm X00I00 030 no mammamflp ESHHQHstvw nmumm moflumu pcsoncs on canon m>Huowmmmu 0cm mSONumuucmocoo odoumumooHuHOUII.nH mamas 128 calculated from the charcoal-dialysis data were probably for steroid concentration intermediate between the initial and the 8 hour steroid concentrations. Also obvious in Table 17 is that the second equi- librium dialysis experiment had higher bound to unbound ratios than for equivalent steroid concentrations in Experiment 1. The higher ratios in Experiment 2 may rep— resent differences in protein concentration since the experiments represented two different 60-80% saturated L” ammonium sulfate fractions prepared from the same weight of different bovine mammary tissues. The ammonium sulfate fraction used in Experiment 2 may have had more total protein, hence, more binding protein, or may have had a higher pr0portion of binding protein. Since stage of lactation of the cows at slaughter was unknown any change in binding protein quantity with stage of lactation, if it does change, could explain the different ratios, as could the milk, moisture, or connective tissue content of the tissue. The data from the two experiments was plotted by both the Lineweaver-Burk, or double reciprocal, and Scatchard methods. The double reciprocal plot from Experiment 1 is shown in Figure 8. The x intercept equals 129 Fig. 8.--Double reciprocal plot of Tissue l 60~80 ammonium sulfate fraction corticosterone- binding after equilibrium dialysis. Two ml diluted (1:9 TMK) 60-80%.Saturated ammonium sulfate fraction per dialysis sac, two sacs per 100 ml total volume containing 8.3 X lO'llM corticosterone-H3 and different levels of un- labeled corticosterone. l/bound corticosterone (x 1010 L/mole) 130 200‘ 160' 120. 80- -l/k = -25 X 107M"1 k = 4 x 10'9M 4o. / A ‘- l A '25 400 800 1200 1/corticosterone concentration (X 107 L/mole) Figure 8 131 the negative reciprocal of the binding constant while the y intercept is the reciprocal of the number of binding sites at infinite (l/b=0) steroid concentration. The problems inherent in the double reciprocal plot are clear in the figure. The high steroid concen- trations gave points clustered around the origin and they therefore contributed less to the derived line than did the low concentration points with their wider Spread. However, the high affinity binding that is seen at very low steroid concentrations is usually of the most interest and those are the points which are sufficiently separated to allow accuracy in drawing a line through them. In Figure 8 those low concentration points definitely did not create a line which encompassed the points derived from lower affinity binding (high steroid concentration). Over a narrow range of steroid concentration the double reciprocal plot would be valid but over a wide range some of the points have to be ignored. The line in Figure 8 was drawn from only the two low-steroid points. The other two points, clustered at the origin, would make calcula- tion of k, the binding constant, for low affinity binding almost impossible using the scale Shown. The line shown in Figure 8 yielded a k of 4 X 10-9M and since the line 132 emphasized high affinity binding, the 4 X lO-gM is a high affinity binding constant. The Scatchard plot of the same data, shown in Figure 9, resolves high and low affinity binding into a more vivid graphical presentation. The low affinity por- tion of the curve, that part of the curve with the least slope, can be geometrically subtracted from the high af— finity portion in order to better determine the high af— finity binding constant (71). In Figure 9 the data ob- tained at 4337 x 10'11 11 M (shown in parenthesis in Figure 9) and 43298 X 10_ M (shown as line 2 in Figure 9) were both considered to represent low affinity binding and each was separately subtracted from the high affinity portion of the curve. Depending on which value was subtracted there was a slight variation in the binding constants calculated. Subtraction of the binding found at 43298 X lo-llM steroid (Shown in Figure 7 only as Line 2) from the others yielded a new curve rather than the straight line which the cal- culations were seeking to produce. By arbitrarily break— ing the derived curve into two straight lines (Lines 3 and 4, Figure 9) binding constants of 2.9 and 5.9 X 10-9M were obtained. The 5.9 X 10—9M binding constant was the result of increased low affinity binding because the 133 Fig. 9.--Scatchard plot of data represented in Figure 8. L1 — L5 = Lines 1-5. Corticos- terone concentrations at each point shown in parentheses. Bound/Unbound corticosterone 134 (8.3 x lO-llM) 5 = 2.1 x 10—9M _ k4 = 2.9 x 10'9M ; k3 = 5.9 x 10-9M (441 x io'llM) 200 400 600 800 X lo-llM corticosterone bound Figure 9 135 portion of the curve between 441 and 4337 X lO-llM steroid was flattening and therefore had a strong influence on calculation of the binding constant. However, when the binding ratios obtained at 4337 X 10_11M corticosterone were subtracted from the curve they had less influence on the derived curve and the Slope of that derived line more closely followed the slope of the initial portion of the original curve. That second derived line gave a bind- ing constant of 2.1 X 10-9M. As noted in the paragraph above the line between 4337 and 43298 X lO-llM steroid concentration was not shown in Figure 7. The line between those two concentra- tions became almost horizontal indicating very low binding affinity beyond 4337 X 10—11M corticosterone concentra- tion and conversely, the low steroid concentration data showed progressively smaller binding constants as the steroid concentration decreased. More points must be measured in that area before a really accurate high af- finity binding constant can be reliably determined and lower steroid concentrations than those used may yield an even lower binding constant for corticosterone. In an attempt to better define the high affinity binding constant the second experiment was conducted with 136 a much narrower range of corticosterone concentration (Table 17). The double reciprocal plot of the data ob- tained over that narrow range is shown in Figure 10. If in Figure 10, as in Figure 8, emphasis was placed only on the high affinity binding the line would have yielded a k equal to l X 10-9M, however, Figure 10 shows the line drawn between the high and low steroid concentration p01nts. Inclusion of the high steroid concentration points was possible because of the narrow range of steroid concentration. Inclusion of the high-steroid, lower affinity points gave a binding constant of 2 X 10—9M. The same data plotted by Scatchard's method is shown in Figure 11. This plot was treated in a couple of ways. Each two successive points were first connected by a straight line, thereby giving three SlOpeS. Those slopes 10M, 3.5 x 10-9M, and 1.3 x lO-BM. gave k's of 9.6 X 10— The steepest lepe, the highest affinity portion of the data curve, gave the 0.96 X 10-9M which was very close to the l X 10-9M determined from the double reciprocal plot when the high affinity binding was emphasized. The least lepe may well represent the initial influence of a lower affinity binding which would be seen at high steroid concentrations, while the intermediate lepe might 137 Fig. 10.--Double reciprocal plot of Tissue 2 60-80 ammonium sulfate fraction corticosterone binding after equilibrium dialysis. Two ml diluted (1:9 TMK) 60-80% saturated ammonium sulfate fraction per dialysis sac, one sac per 100 m1 total volume con— taining 1.66 or 8.3 X 10'11M corticosterone- H and two levels of unlabeled corticosterone. l/bound corticosterone (X 109L/mole) 138 80 60 40 -0.05 X 1010 2 x 1079M r $3 II II 20 M- l -0.05 2 4 6 1/corticosterone concentration (x ioloL/moie) Figure 10 139 Fig. ll.--Scatchard plot of data represented in Figure 10. 140 .B.anlfiw|ai III!» | o .. HH whomflm canon mcoumumOUfluuoo Eda 0H X omm oem oom owe owa om es s oH x m.m u mx 2 ca x 0.m u as ml OH! 0 SmnoH x m.a u mx 0 is ca x sees . Ha- ASHHIOH x mmmv a z . A Hausa x m we is IoH x 5.40 dd euozeqsoorqxoo punoqun/punog 141 represent either a transition between high and low af- finity binding or a family of steroid binding proteins or binding sites having progressively lower affinity for corticosterone. Different affinity binding sites were proposed by Tucker et a1. (98) for mammary cells. They had more points from which to derive lines and concluded there were two sites of relatively high affinity for hydrocortisone in mammary cells in culture. Their highest affinity site had a k of 2 X 10—9M.which is close to the intermediate binding constant in this experiment. The presence of more than two binding constants due to either other proteins or changes in binding affinity with pro- gressive increase in steroid bound cannot be ruled out. The second treatment of the Figure 11 Scatchard plot involved drawing a curve through the first three points and subtracting the low affinity slope. This does little for the data as shown by the circled points in the figure since a new curve was formed rather than a straight line. The binding constants derived from the two sets of data ranged between 0.96 X 10_9M and 5.9 X 10-9M. The double reciprocal and Scatchard plots both gave similar binding constants with the high affinity portions of the 142 curves ranging between 0.96 X 10-9M and 2.1 X lO-gM with the Scatchard plots and l and 4 X 10-9M with the double reciprocal plots. As an attempt to further define the high affinity binding constant, or constants, an experiment was performed using twelve corticosterone concentrations and two differ- ent bovine mammary tissues with one of them duplicated. The twelve corticosterone-H3 concentrations are Shown in Table 18. After equilibrium dialysis at these concentra- tions aliquots were counted for tritium and from that data bound to unbound ratios and concentration of bound corti— costerone were calculated and subjected to statistical analysis. TABLE 18.—-Twelve corticosterone concentrations used in equilibrium dialysis of 60-80%.saturated ammon- ium sulfate fractions from two bovine mammary tissues. ' Tube X lO-llM 1.66 3.32 4.98 9.96 20 53 138.9 226.4 312 10 442.9 11 2174 12 4339 KOQQO‘UIthl-J 143 For statistical purposes each dialysis sac repre- sented one value for bound to unbound ratio and one value for concentration of steroid bound. That made Six points available for each axis of the Scatchard plot for each steroid concentration used. An analysis of variance was performed on each group of six points to determine whether variation between eXperiments was significantly larger than variation within an experiment. A significant dif- ference between eXperiments was seen in only one case where the second lowest steroid concentration in one experiment had points which were obviously different from any of the others. These data were discarded before further calculations were made. After that deletion there were no significant differences between experiments so the points at each concentration were averaged before plotting by Scatchard's method (Figure 12). Figure 12 shows sample standard deviations for both bound to unbound ratio and concentration of steroid bound for only the first five concentration points since after the fifth concentration point the curve was almost horizontal. As can be seen in that figure the standard deviations were rather large. The large variation was traceable to flocculation of some material in the labeled 144 Fig. 12.--Scatchard plot of combined 60-80 ammonium sulfate fraction, equilibrium dialysis corticosterone—binding data from two tissues with duplication of one. Two ml diluted (1:9 TMK) 60-80% saturated ammonium sulfate fraction per dialysis sac, two sacs per 100 ml total volume containing 1.66 X lO‘llM corticosterone—H3 and dif- ferent levels of unlabeled corticosterone. Dialyzed 24 hours at 4°C. Means and stand— ard deviations for bound/unbound and con- centration bound are Shown. 145 wcouwumoofluuoo UCSOQ:D\U:Som 4O 3O 20 10 X lO-llM corticosterone bound Figure 12 146 corticosterone used in these experiments. The floccula— tion produced error in sampling for counting but although the standard deviations were relatively large they were such that they had little effect upon the Scatchard plot lepe. That is, in the high affinity portion of the Scatchard curve most of the variation was in the bound to unbound term and as such there was little effect on the SlOpe. Variation in the low affinity portion of the curve was mostly in the amount bound which produced little effect on the lepe representative of that low affinity binding. The binding constants and their standard devia— tions, calculated from the data represented in Figure 12, are Shown in Table 19. The SlOpe of each segment of the Scatchard curve was then tested (109) to determine whether it was significantly different from the slope of its suc- ceeding segment. From left to right in Figure 12 the first lepe was significantly different from the second, primarily because the means of the duplicates were almost equal, and the second from the third. The SlOpe of the third segment was significantly different from that of the fourth segment (P=0.10) but thereafter none of the 147 TABLE l9.--Binding constants, k, and their standard devia- tions, S , calculated for the five segments of the curve shown in Figure 12. Sk 2.332 x 10'11 i 0.256 x lo’llm 1.516 x 10'11 0.032 x 10’11M 35.238 x 10’11 12.189 x 10'11M 220.504 x 10'11 183.676 x 10'11M 401.140 x 10'11M 402.746 x 10-11 SlOpes was significantly different from the lepe of its adjacent segment. Because of the large standard deviations involved the SlOpe of the first segment of the curve was not used alone to determine the highest affinity binding constant. AS an alternative the data for the first three concentra- tion points of each eXperiment were used to determine a binding constant by calculation of a linear regression equation for each eXperiment. By doing that the differ- ences between experiments could again be tested for sta- tistical significance. The binding constants, which are the reciprocals of the regression coefficients, and their .“ m“ in“.n 148 standard deviations are shown in Table 20. When the slopes from which the binding constants were derived were tested for Significant differences none of the lepeS was significantly different from any of the others. The data could therefore be combined to give an average binding constant of 3.2 X lo-llM. TABLE 20.--High affinity binding constants and their standard deviations determined by linear re- gression calculations on data represented in Figure 12; Tissue k i 8k 7 ___________ x 10-11M__-__-_1-__ l 2.64 1.34 2 3.57 1.35 3 3 40 1.79 The data obtained in these last three experiments did not resolve the question of whether there are several high affinity binding Sites of differing affinities. However, there was a striking aSpect to the data in that the binding constants calculated were much lower than previously observed. They were of the order of lo-llM 149 rather than the 10—9M determined previously. One possible explanation is the greater number of low-steroid concen— tration points used in the latter experiments. These apparently enabled better definition of the high affinity binding constant. There is, however, no reason to believe that lower binding constants could not be determined with even lower steroid concentrations. However, any attempt to use lower steroid concentration would soon be limited by the specific activity of the tritiated corticosterone available. The 30 Ci/mM corticosterone used in these eXperiments could not be used very much below the 1.66 X 10-11M concentration used as the lowest steroid concentration. It therefore became limiting in the determination of the lowest possible binding constant. GENERAL D ISCUSS ION A number of tissues have been shown to bind glucocorticoids, including mammary cells in culture, and glucocorticoids are also known to be essential to the process of lactation. Since binding of a hormone in a cell implies a biological function of that hormone in that cell, and since glucocorticoids are so important to lactation, there is a strong possibility that mammary tissue contains a material able to Specifically bind adrenal glucocorticoids. This study has shown the pres- ence of such binding protein or proteins in mammary tissue. Mammary glucocorticoid binding was observed in centrifugally Separated nuclear. mitochondrial and micro- somal particulate fractions although the real extent of binding in each of the particulate fractions could not be accurately determined in these experiments. The de- sired accuracy was not achieved because the methodology used to study binding in the particulate fractions allowed 150 151 dissociation and some cross-contamination among fractions. Therefore the data concerning distribution among fractions Should be considered only approximate. Two conclusions can be drawn from the work on particulate distribution of binding. One is that such binding did exist and the second is that the binding material was extractable by high pH. The observation that high pH is required to extract the binding protein may indicate that the protein is an integral part of the par- ticulate structure and extractable only by rigorous treatment such as sodium hydroxide. The particulate bind- ing observed in these studies may be Similar to rat liver mitochondrial glucocorticoid binding reported by DeVenuto and Muldoon (81). When their mitochondrial preparations were sonicated and centrifuged 105,000 x g over 91% of the binding material sedimented thereby indicating that such binding protein was an integral part of the mitochon- drial structure. Observations on the stability of the binding pro- tein to freezing were made on two of the fractions, the sodium hydroxide extract and the 60—80%.Saturated ammonium sulfate fraction of the 30,000 x g supernatant. Neither lost their ability to bind steroids. The ammonium sulfate -..u.—... .“v‘u_.—‘; A-.. 152 fractions were used for competition and binding constant studies and when equilibrium dialyzed at equal steroid concentrations (low concentration point in Figure 5) the bound to unbound ratios were all in reasonable agreement thereby indicating the freezing and thawing had little effect on the high affinity binding activity in the 60—80% saturated ammonium sulfate fraction. I Several fractionation methods were attempted dur- ing this study and the most important fractionation step was ammonium sulfate precipitation. It was used both before and after incubation with tritiated steroid but although ammonium sulfate precipitation of steroid-labeled protein has been used (108) to study degree of binding activity, supposedly with little loss of label, this study has shown that very low concentrations of glucocorticoid will precipitate with denatured protein during ammonium sulfate treatment. Therefore, ammonium sulfate fraction- ation should be used before incubation with steroid if accurate measure of binding is to be made. Used in such a manner, it was a very successful method for partial purification of the cytOplasmic high affinity binding protein which became the fraction of primary interest in this study. The soluble high affinity steroid binding 153 activity was concentrated in the 70-80% ammonium sulfate fraction but was also high in the 60—70%»fraction. Therefore, the 60—80%.fraction was used for competition and binding kinetcis studies. After ammonium sulfate precipitation, further fractionation was attempted. Calcium phosphate gel and Sephadex G-100 filtration added little to the purifica- I tion of the 60-80% saturated ammonium sulfate fraction. DEAF-cellulose ion exchange chromatography did provide 5- good separation of a number of protein fractions, one of which had much higher specific binding activity than the others. Even finer separation might have been obtained on the DEAE had a more shallow gradient been used during elution. Studies of binding kinetics and competition among steroids were studied in the 60-80%.saturated ammonium sulfate fraction of the 30,000 x g supernatant. Previous studies, in the particulate fractions, had indicated steroid binding to be linear with steroid concentration. The reason for that linear uptake was probably due to the presence of two or more types of binding sites, each with different affinities for corticosterone. That situation was clarified when the binding data for the 30,000 x g 154 supernatant was plotted by Scatchard's method. The Scatchard plot showed low affinity binding sites which appeared to be unsaturable and have linear steroid uptake. Such sites were apparently reSponsible for the “infinite" binding observed in the particulate fractions (Figure 5 and Table 13). The existence of such a phenomenon has I‘ been observed in other tissues. Gardner and Tomkins (87), r during isolation of a corticosteroid—binding macromolecule from hepatoma cells, observed an almost linear increase in “" binding until added steroid concentration exceeded lO-SM. NonSpecific, low affinity, binding which increased lin- early with increasing steroid concentration was also seen in hepatoma cells (57) with similar observations having been made with aldosterone in rat kidney (110) and in thymus cells using a number of glucocorticoids and other steroids (92). The "infinite" binding observed in this study was probably of the same low affinity, nonSpecific nature as seen in those other tissues. This was further confirmed when Tucker et al. (98) showed low affinity binding in their mammary cell cultures. The competition experiments revealed that hydro- cortisone was bound equally as well as corticosterone although some rat liver work (81,82) showed corticosterone 155 preferentially bound compared to hydrocortisone. One explanation could be tissue differences but species dif- ference is also a feasible explanation since corticos- terone is the glucocorticoid found in the rat while both corticosterone and hydrocortisone are present in the cow. If tissue difference is unimportant, the rat corticosteroid-binding protein must be different from that of the bovine and purification of rat mammary bind— ing fractions and testing them for corticosterone- hydrocortisone competition should answer whether tissue or Species differences predominate. One other factor should be considered when pon- dering rat vs bovine steroid binding. The liver is the main site of metabolic degradation of glucocorticoids and as such it may have different receptors, for different purposes, than the mammary gland. The rat liver receptors may be designed solely for the degradation of corticos- terone. The rat would have no need to degrade hydro- cortisone and it may therefore not have receptors for hydrocortisone. Estrogen uptake by the mammary gland has been shown previously (111) and estradiol-C14 was bound in the 750 x g supernatant in this study but competition 156 experiments on the 60-80% saturated ammonium sulfate fractions showed estradiol noncompetitive for corticos- terone binding sites (Figure 6). These eXperiments do not indicate whether or not estradiol is binding to some portion of that 60—80% saturated ammonium sulfate frac- tion but only that it doesn't compete with corticosterone. _.1...-4z.-.I--I~s- -m..-v . . Progesterone competed with corticosterone for binding sites and the competition was extensive though not complete. Progesterone will bind to hydrocortisone receptors in hepatoma cells (57,87) and bovine mammary cells in culture (97,98) but the physiological signifi- cance is unknown. A physiological relationship is sug- gested by the fact that progesterone rises to a high level during pregnancy but markedly decreases after paturition while at parturition free glucocorticoid levels increase. Perhaps during pregnancy the high progesterone levels promote extensive binding of progesterone to the glucocorticoid receptors and progesterone on the receptors might make the binding protein unable to eXpress any lac- tational influences it is assumed to have. At parturition the combined decrease of progesterone and increase in glucocorticoid would promote glucocorticoid binding. The 157 bound glucocorticoid would then enable the receptor pro- tein to exert its lactogenic effect. The binding constants for the soluble cell frac- tion (using the 60-80 ammonium sulfate fraction) under equilibrium conditions were first determined to be 9 9 0.96 x 10’ to 5.9 x 10’ M for the high affinity portion of the binding curve while the low affinity portion of .3 .I. A. .»'a-u‘.L‘ the Scatchard plots was almost horizontal and a low affinity binding constant was impossible to calculate. The transition area, between the high and low affinity portions of the curve, was difficult to deal with. It could represent either a family of binding proteins or binding sites, or merely transition between high and low affinity binding. Tucker et a1. concluded there were at least two sites with relatively high affinity binding in their mammary cells in culture (98). With more than one site available any binding and competition data col— lected at steroid concentrations in the transition area gives misleading information on the real nature of the binding. The k = 5.9 X 10-9M given above probably repre- sented influence of low affinity binding by considering too much of the low affinity portion in the transition portion of the Scatchard plot. 158 Estimates of the binding constants from the charcoal dialysis experiment Shown in Figure '7 gave k's -9 -8 -9 . of 4.4 X 10 M to 2.5 X 10 M. The 4.4 X 10 M value is within the range found by equilibrium dialysis. Those binding constants can be compared to corticosterone binding globulin (CBG) which has 8 9M at 4°C (6). Those k = 3 x 10" M at 37°C and 2 x 10' first binding constants for the mammary binding protein were obtained at 4°C and are very close to the CBG value. A lower k value in the mammary gland would encourage transfer of glucocorticoid from the blood CBG to mammary gland and the CBG could be considered a vehicle for tranSport of glucocorticoid to the mammary gland: however, the similar binding affinities would not indicate such a role for CBG. With similar affinity for the same steroid they would be in competition for that steroid therefore the decrease in CBG levels observed soon after parturi— tion may be the means by which glucocorticoid is made available to the mammary gland for lactogenesis. Similar binding constants could indicate another, less important, role for CBG in relation to the mammary binding. Blood glucocorticoid levels can be increased by stress and doubled during normal diurnal cycles. Much 159 of those sharp increases in corticoid is probably absorbed by CBG because with similar binding constants distribution of the increase between CBG and the mammary binding pro— tein would be proportional to the quantity of each. Dis- tribution between the two would save the mammary receptors from really large variations in glucocorticoid concentra— tion and if the glucocorticoid receptors are important in enzyme induction such a mechanism protects the cell from rapid or cyclic changes in enzyme levels when such changes would not necessarily be beneficial to the cell or tissue. Such relationships between CBG and the mammary binding protein are very feasible with a k of approxi- mately 10-9M. However, the last data collected indicated a much lower value for k. A lower k value neceSSitateS revision of the hypotheses prOposed to relate CBG and the binding protein. It was mentioned that a k value lower for mammary binding protein than CBG would encourage transfer of corticosteroid from CBG to the mammary binding protein. Because of the approximately lOO-fold higher affinity the mammary protein would be less sensitive to normal physiological changes in blood glucocorticoid con- centration and expression of its physiological effect may 160 therefore be more dependent on glucocorticoid—progesterone ratio than absolute concentration. One important factor should be considered when discussing the lower k value. It should be remembered that the lower binding constant was obtained with very low glucocorticoid concentrations and dialysis of longer duration. If the glucocorticoid exerts any sort of pro- 2 tective effect upon the binding protein those two factors, low steroid concentration and longer time could act in conjunction to produce an artificially low k value. If the lower k is correct, however, the cytOplasmic binding protein has a very high affinity for glucocorti- coid and one can Speculate Whetherjit‘plays a passive or and active role in the initiation of lactation. In a passive role it might act merely as an "intracellular CBG" whose function would be tranSport of glucocorticoid through the cytOplasm to the nucleus similar to the cyto- plasmic estrogen-binder observed in uterine tissue. A nuclear role for glucocorticoid seems to be implied by the requirement for glucocorticoid during cell division preceding casein synthesis. Such a nuclear role implies existence of a nuclear receptor for the glucocorticoid and if the nuclear receptor is not merely the cytOplasmic 161 binding protein in a new role it probably has a binding constant even lower than 3 X 10_11M in order to facilitate transfer of glucocorticoid from cytoplasmic binder to nuclear receptor. Any material with a k even lower than 3 X 10_11M would be very difficult to study accurately with presently available isotOpeS. The cytOplasmic binding protein could also have a direct extranuclear role in lactation. The observation that certain glucocorticoids promote rough endOplasmic reticulum formation makes such Speculation feasible. The binding protein could conceivably play some role in the interaction of polysomes and smooth endoplasmic re- ticulum. Further Speculation envisions the binding pro- tein as some vital enzyme which is inactive when associ— ated with progesterone but active when bound to glucocor— ticoid. Or perhaps the glucocorticoid promotes the asso- ciation between two proteins required for enzyme activity. An example would be the association of UDP—galactosyl transferase and d-lactalbumin to create lactose synthetase. Lactose synthetase activity appears to be sensitive to progesterone-glucocorticoid levels during pregnancy and parturition and although any glucocorticoid effect on those enzyme levels could be nuclear that enzyme would be .-_-. _.1_.__._._¢—4—_-.-.—_‘_ .. . I ; 162 the prime candidate for further study of the role of the cytoplasmic glucocorticoid-progesterone binding protein. Relationships existing among CBG, glucocorticoid, progesterone, mammary gland binding protein and lactation have been partially clarified by this study. The exis- tence of a binding protein(s) in particulate and cyto- 1.__-*.__.-___1____—‘ _. . plasmic fractions has been shown in mammary gland. It will bind corticosterone, hydrocortisone, and progesterone and has a binding constant best estimated as 3.2 X lO-IIM 1, for corticosterone. LIST OF REFERENCES -1- _“.ri_--w. “11 LIST OF REFERENCES Guyton, A. C. Textbook of Medical Physiology. W. B. Saunders Co., Philadelphia, 1961. Peterson, R. E. The identification of corticosterone in human plasma and its assay by isotOpe dilution. J. Biol. Chem. 225:25, 1957. Bush, I. E. Species differences in adrenocortical secretion. J. Endocrin., 9:95, 1953. Wilson, H., J. J. Borris, and R. C. Bahn. Steroids in the blood and urine of female mice bearing an ACTH— producing pituitary tumor. Endocrinology 62:135, 1958. Halberg, F., R. E. Peterson, and R. H. Silber. Phase relations of 24—hour periodicities in blood corticos- terone, mitoses in cortical adrenal parenchyma, and total body activity. Endocrinology 64:222, 1959. Seal, U. S., and R. P. Doe, ip Pincus, G., T. Nakao, and J. F. Tait (eds.), Steroid Dynamics, Academic Press, New York, 1966, p. 63. Gala, R. R. and U. Westphal. Corticosteroid-binding activity in serum of mouse, rabbit and guinea pig during pregnancy and lactation: possible involvement in the initiation of lactation. Acta Endocr. 55:47, 1967. Brush, M. G. The effect of ACTH injections on plasma corticosteroid levels and milk yield in the cow. J. Endocrin. 21:155, 1960. 163 10. 11. 12. 13. 14. 15. 16. 17. 18. 164 Szepesi, B., and R. A. Freedland. A possible method for estimating hormone effects on enzyme synthesis. Arch. Biochem. BiOphyS. 133:60, 1969. Lyons, W. R., C. H. Li, and R. E. Johnson. The hor— monal control of mammary growth. Recent Prog. Horm. Res. 14:219, 1958. Nandi, S. Role of somatotrOpin in mammogenesis and lactogenesis in C3H/HeCRGL mice. Science 128:772, 1958. Cowie, A. T., J. S. Tindal, and A. Yokoyama. The induction of mammary growth in the hypOphysectomized goat. J. Endocrin. 34:185. 1966. Chadwick, A. and S. J. Folley. Lactogenesis in pseudo- pregnant rabbits treated with ACTH. J. Endocrin. 24:XI, 1962. Talwalker, P. K., C. S. Nicoll, and J. Meites. Induc- tion of mammary secretion in pregnant rats and rabbits by hydrocortisone acetate. Endocrinology 69:802, 1961. Meites, J., T. F. Hopkins, and P. K. Talwalker. In— duction of lactation in pregnant rabbits with prolac- tin, cortisol acetate or both. Endocrinology 73:261. 1963. Freison, H. G. Lactation induced by human placental lactogen and cortisone acetate in rabbits. Endocrinology 79:212, 1966. Tucker, H. A. and J. Meites. Induction of lactation in pregnant heifers with 9-fluor0prednisolone acetate. J. Dairy Sci. 48:403, 1965. Thatcher, W. W. and H. A. Tucker. Intensive nursing and lactational performance during extended lactation. Proc. Soc. EXptl. Biol. Med. 128:46, 1968. 19. 20. 21. 22. 23. 24. 25. 26. 27. 165 Anderson, R. R. and C. W. Turner. Effect of adrenal- ectomy and corticoid replacement on lactational per- formance in rats. Proc. Soc. Exptl. Biol. Med. 110:349, 1962. Anderson, R. R. and C. W. Turner. Aldosterone and corticosterone in maintenance of lactation in adren- alectomized rats. Proc. Soc. EXptl. Biol. Med. 112:997, 1963. Johnson, R. M. and J. Meites. Effects of cortisone acetate on milk production and mammary involution in parturient rats. Endocrinology 63:290. 1958. Thatcher, W. W., and H. A. Tucker. Lactational per- formances of rats injected with oxytocin, cortisol- 21-acetate, prolactin and growth hormone during pro- longed lactation. Endocrinology 86:237, 1970. Thatcher, W. W. The maintenance of milk synthesis during extended lactation, Ph.D. Thesis, Michigan State University, 1968. Emery, R. S., J. D. Benson, and H. A. Tucker. Dietary and hormonal effects on extended lactation and lipid metabolism. J. Dairy Sci. 53:519 (abs.), 1970. Campbell, I. L., A. W. F. Davey, F. H. McDowell, G. F. Wilson, and R. E. Munford. The effect of adrenocor- ticotrOphic hormone on the yield, composition, and butterfat properties of cows' milk. J. Dairy Res. 31:71, 1964. Hartmann, P. E., ip_Reynolds, M. and S. J. Folley (eds.). Lactogenesis, University of Pensylvania Press, Philadelphia, 1969, p. 97. Baldwin, R. L., and L. P. Milligan. Enzymatic changes associated with the initiation and maintenance of lactation in the rat. J. Biol. Chem. 241:2058, 1966. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 166 Baldwin, R. L. Enzymatic activities in mammary glands of several species. J. Dairy Sci. 49:1533, 1966. Baldwin, R. L., and R. J. Martin. Protein and nucleic acid synthesis in rat mammary glands during early lactation. Endocrinology 82:1209, 1968. Baldwin, R. L., and R. J. Martin. Effects of hypo- physectomy and several hormone replacement therapies upon patterns of nucleic acid and protein synthesis and enzyme levels in lactating rat mammary glands. J. Dairy Sci. 51:748, 1968. Baldwin, R. L., in Reynolds, M.. and s. J. Folley (eds.). Lactogenesis, University of Pennsylvania Press, Philadelphia, 1969, p. 85. Elias, J. J. Cultivation of adult mouse mammary gland in hormone—enriched synthetic medium. Science 126:842, 1957. Elias, J. J. Effect of insulin and cortisol on organ cultures of adult mouse mammary gland. Proc. Soc. Exptl. Biol. Med. 101:500. 1959. Rivera, E. M. and H. A. Bern. Influence of insulin on maintenance and secretory stimulation of mouse mammary tissues by hormones in organ-culture. Endocrinology 69:340. 1961. Turkington. R. W., and Y. J. TOpper. Androgen inhibi- tion of mammary gland differentiation in vitro. Endocrinology 80:329. 1967. Rivera, E. M. Interchangeability of adrenocortical hormones in initiating mammary secretion in vitro. Proc. Soc. EXptl. Biol. Med. 116:568. 1964. Rivera, E. M. Differential responsiveness to hormones of C3H and A mouse mammary tissues in organ culture. Endocrinology 74:853. 1964. 38. 39. 40. 41. 42. 43. 44. 45. 46. 167 Turkington, R. W., W. G. Juergens, and Y. J. TOpper. Steroid structural requirements for mammary gland differentiation in vitro. Endocrinology 80:1139, 1967. Stockdale, F. E., W. G. Juergens, and Y. J. TOpper. A histological and biochemical study of hormone- dependent differentiation of mammary gland tissue in vitro. DevelOp. Biol. 13:266. 1966. Juergens, W. G.. F. E. Stockdale, Y. J. TOpper, and J. J. Elias. Hormone-dependent differentiation of mammary gland in vitro. Proc. Natl. Acad. Sci. U.S. 54:629. 1965. Turkington, R. W. Induction of milk protein synthesis by placental lactogen and prolactin in vitro. Endocrinology 82:575. 1968. Stockdale, F. E. and Y. J. TOpper. The role of DNA synthesis and mitosis in hormone-dependent differen- tiation. Proc. Natl. Acad. Sci. U.S. 56:1283. 1966. Turkington. R. W., W. G. Juergens, and Y. J. Topper. Hormone-dependent synthesis of casein in vitro. Biochim. BiOphyS. Acta 111:573, 1965. Lockwood, D. H., R. W. Turkington, and Y. J. TOpper. Hormone-dependent development of milk protein syn- thesis in mammary gland in vitro. Biochim. BiOphys. Acta 130:493. 1966. Voytovich, A. E. and Y. J. TOpper. Hormone-dependent differentiation of immature mouse mammary gland lg vitro. Science 158:1326, 1967. Lockwood. D. H., F. E. Stockdale, and Y. J. Topper. Hormone-dependent differentiation of mammary gland: sequence of action of hormones in relation to cell cycle. Science 156:945, 1967. 47. 48. 49. 50. 51. 52. 53. 54. 55. 168 Mayne, R., and J. M. Barry. Insulin maintains the level of RNA in mammary tissue cultures. Biochim. BiOphys. Acta 138:195, 1967. Mayne, R., I. A. Forsyth, and J. M. Barry. Stimula- tion by hormones of RNA and protein formation in organ cultures of the mammary glands of pregnant mice. J. Endocrin. 41:247. 1968. Turkington, R. W., and O. T. Ward. Hormonal stimula- tion of RNA polymerase in mammary gland in vitro. Biochim. Biophys. Acta 174:291, 1969. Turkington, R. W., D. H. Lockwood, and Y. J. TOpper. The induction of milk protein synthesis in post— mitotic mammary cells exposed to prolactin. Biochim. BiOphys. Acta 148:475. 1967. Green, M. R., and Y. J. TOpper. Some‘effects of pro- lactin, insulin and hydrocortisone on RNA synthesis by mouse mammary gland, in vitro. Biochim. Biophys. Acta 204:441, 1970. James, D. W., B. R. Rabin, and D. J. Williams. Role of steroid hormones in the interaction of polysomes with endoplasmic reticulum. Nature 224:371. 1969. Mills, E. S. and Y. J. Topper. Mammary alveolar epithelial cells: effect of hydrocortisone on ultra- structure. Science 165:1127, 1969. Mills. E. s., and Y. J. Topper. Some ultrastructural effects of insulin, hydrocortisone, and prolactin on mammary gland explants. J. Cell Biol. 44:310. 1970. Thompson, E. B., G. M. Tomkins and J. R. Curran. Induction of tyrosine o—ketoglutarate transaminase by steroid hormones in a newly established tissue culture cell line. Proc. Natl. Acad. Sci. U.S. 56:296. 1966. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 169 Granner, D. K., S. Hayashi, E. B. Thompson. and G. M. Tomkins. Stimulation of tyrosine amino transferase synthesis by dexamethasone phOSphate. J. Mol. Biol. 35:291. 1968. Baxter. J. D., and G. M. Tomkins. The relationship between glucocorticoid binding and tyrosine amino transferase induction in hepatoma tissue culture cells. Proc. Natl. Acad. Sci. U.S. 65:709. 1970. i Auricchio. F., D. Martin, Jr., and G. Tomkins. Con— l trol of degradation and synthesis of induced tyrosine y aminotransferase studied in hepatoma cells in culture. 5 Nature 224:806, 1969. Martin, D. W., Jr., G. M. Tomkins. and M. A. Bresler. '7" Control of Specific gene expression examined in syn- chronized mammalian cells. Proc. Natl. Acad. Sci. U.S. 63:842. 1969. Waters, M. D. and G. K. Summer. Prednisolone induc- tion of alkaline phOSphatase in human skin fibro- blasts. Biochim. Biophys. Acta 177:650, 1969. Stedman, E. and E. Stedman. Probable function of histone as a regulator of mitOsis. Nature 152:556. 1943. Busch, H. Histones and other nuclear proteins. Academic Press, New York, 1965. Allfrey, V. G.. V. C. Littau, and A. E. Mirsky. On the role of histones in regulating ribonucleic acid synthesis in the cell nucleus. Proc. Natl. Acad. Sci. U.S. 49:414. 1963. Hohmann, P., and R. D. Cole. Hormonal effects on amino—acid incorporation into lysine—rich histones. Nature 223:1064, 1969. Stellwagen, R. H., and R. D. Cole. Histone biosyn- thesis in the mammary gland during develOpment and lactation. J. Biol. Chem. 244:4878, 1969. 66. 67. 68. 69. 70. 71. 72. 73. 74. 170 Sluyser, M. Binding of hydrocortisone to rat liver histones. J. Mol. Biol. 19:591. 1966. Sluyser, M. Interaction of steroid hormones with histones in vitro. Biochim. Biophys. Acta 182:235, 1969. Sluyser, M. Binding of testosterone and hydrocorti- sone to rat liver histones. J. Mol. Biol. 22:411. 1966. Daughaday, W. H. Binding of corticosteroids by plasma proteins. III. The binding of corticosteroid and related hormones by human plasma and plasma protein fractions as measured by equilibrium dialysis. ; J. Clin. Invest. 37:511. 1958. 5 o‘ll“”_—a-_Jnn ”v. ._ -1 I Slaunwhite, W. R., and A. A. Sandberg. Transcortin: A corticosteroid-binding protein of plasma. J. Clin. Invest. 38:384. 1959. Sandberg. A. A.. H. Rosenthal. S. L. Schneider, and W. R. Slaunwhite, Jr., 13 Pincus, G.. T. Nakao, and J. F. Tait (eds.), Steroid Dynamics, Academic Press. New York, 1966, p. l. Gala, R. R., and U. Westphal. Corticosteroid-binding globulin in the rat: possible role in the initiation of lactation. Endocrinology 76:1079. 1965. Jensen, E. V., and H. I. Jacobson. Basic guides to the mechanism of estrogen action. Recent Progr. Hormone Res. 18:387. 1962. Shyamala, G. and J. Gorski. Estrogen receptors in the rat uterus: studies on the interaction of cytosol and nuclear binding sites. J. Biol. Chem. 244:1097, 1969. 75. 76. 77. 78. 79. 80. 81. 82. 171 Jensen. E. V., H. I. Jacobson, J. W. Flesher. N. N. Saha, G. N. Gupta, S. Smith, V. Colucci, D. Shipla- coff, H. G. Neumann, E. R. DeSombre, and P. W. Jung- blut, in Pincus. G.. T. Nakao, and J. F. Tait (eds.). Steroid Dynamics. Academic Press, New York, 1966, p. 133. Puca, G. A.. and F. Bresciani. Interactions of 6.7-3H-17-8 estradiol with mammary gland and other organs of the C3H mouse in vivo. Endocrinology 85:1. 1969. Puca, G. A.. and F. Bresciani. Association constant and specificity of oestradiol-receptor interaction. Nature 223:745. 1969. Talwar, G. P., M. L. SOpori, D. K. Biswas, and S. J. Segal. Nature and characteristics of the binding of oestradiol-l78 to a uterine macromolecular fraction. Biochem. J. 107:765. 1968. King, R. J. B., J. Gordon. and A. W. Steggles. The prOperties of a nuclear acidic protein fraction that binds (6.7-3H)estradiol-l78. Biochem. J. 114:649. 1969. Bottoms. G.. and D. D. Goetsch. Subcellular distri- bution of the (3H)corticosterone fraction in brain, thymus, heart and liver of the rat. Proc. Soc. Exptl. Biol. Med. 124:662, 1967. DeVenuto, F., and T. Muldoon. Interactions between corticosteroids and fractions of mitochondria and nuclei from normal rat liver cells. EXptl. Cell Res. 50:338. 1968. DeVenuto, F., P. C. Kelleher, and U. Westphal. Inter- actions between corticosteroids and fractions of rat liver and muscle cells as determined by "equilibrium fractionation" and equilibrium dialysis. Biochim. BiOphyS. Acta 63:434. 1962. 83. 84. 85. 86. 87. 88. 89. 90. 91. 172 Morris, D. J., and F. W. Barnes. Jr. On the intra- cellular distribution of (4—14C)cortisol in rat liver. Biochim. BiOphys. Acta 136:67. 1967. Snart, R. S., N. N. Sanyal. and M. K. Agarwal. Bind- ing of corticosterone in rat liver. J. Endocrin. 47:149. 1970. Litwack. G.. E. S. Fiala. and R. J. Filosa. Intra- cellular binding of (14C)hydrocortisone prior to en- zymatic induction. Biochim. BiOphys. Acta 111:569. 1965. Fiala, E. S., and G. Litwack. Binding of a (14C) hydrocortisone metabolite in liver supernatant. Biochim. BiOphyS. Acta 124:260. 1966. Gardner, R. S., and G. M. Tomkins. Steroid hormone binding to a macromolecule from hepatoma tissue cul- ture cells. J. Biol. Chem. 244:4761, 1969. Beato, M.. W. Brandle, D. Biesewig, and C. E. Sekeris. On the mechanism of hormone action XVI- Transfer of (1.2-3H2)cortisol from the cytoplasm to the nucleus of rat-liver cells. Biochim. Biophys. Acta 208:125. 1970. Morey, K. S., and G. Litwack. Isolation and proper- ties of cortisol metabolite binding proteins of rat cytosol. Biochemistry 8:4813. 1969. Bottoms. G. D., R. D. Stith, and R. 0. Burger. Dis- tribution of radioactivity in different tissues of pigs after 3H-hydrocortisone injection. Proc. Soc. Exptl. Biol. Med. 132:1133. 1969. Bottoms, G. D. Intracellular binding of 3H- corticosterone in different tissues of the rat. Fed. Proc. 28:702. 1969. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 173 Munck. A.. and T. Brinck-Johnsen. Specific and non- specific physicochemical interactions of glucocorti- coids and related steroids with rat thymus cells in vitro. J. Biol. Chem. 243:5556. 1968. Wira, C.. and A. Munck. In vitro binding of 3H- cortisol to rat thymus cell nuclei. Fed. Proc. 28:702. 1969. Wira, C.. and A. Munck. Specific glucocorticoid re- ceptors in thymus cells: localization in the nucleus and extraction of the cortisol-receptor complex. J. Biol. Chem. 245:3436, 1970. Wira, C.. and A. Munck. Extraction and preliminary identification of physiological glucocorticoid re- ceptors in rat thymus cell nuclei. Fed. Proc. 29:832. 1970. Emery, R. S., ip Reynolds, M. and S. J. Folley (eds.) Lactogenesis, University of Pennsylvania Press. Phil- adelphia, 1969. p. 105. Tucker, H. A.. B. L. Larson. and J. Gorski. Cortisol binding in cultured bovine mammary cells. Fed. Proc. 29:707. 1970. Tucker. H. A.. B. L. Larson. and J. Gorski. Cortisol binding in cultured bovine mammary cells. Endocrinology 89:152. 1971. Mahler, H. R., and E. H. Cordes. Biological Chemistry. Harper and Row, New York, 1966. Scatchard, G. The attractions of proteins for small molecules and ions. Ann. N.Y. Acad. Sci. 51:660. 1949. Bellamy. D., J. G. Phillips, I. C. Jones. and R. A. Leonard. The uptake of cortisol by rat tissues. Biochem. J. 85:537. 1962. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 174 Colowick, s. 9.. _i_n Colowick. s. P., and .N. 0. Kaplan (eds.). Methods in enzymology, Vol. 1. Academic Press, New York. 1955. p. 98. Stollar, D., and L. Levine. ig_Colowick, S. P., and N. 0. Kaplan (eds.). Methods in enzymologY. Vol. 6. Academic Press, New York, 1963, p. 848. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. Protein measurement with the folin phenol reagent. J. Biol. Chem. 193:265. 1951. Nugent, C- A.. and D. M. Mayes. Plasma corticoster— oids determined by use of CBG and dextran-coated charcoal. J. Clin. Endocrinol. 26:1116. 1966. Trapp. G. A.. and C. D. West. Determination of corticosteroid-binding proteins by an adsorption method. J. Lab. Clin. Med. 73:861. 1969. Milgrom, E. and E. Baulieu. A method for studying binding proteins, based upon differential dissocia- tion of small ligand. Biochim. BiOphys. Acta 194:602, 1969. Herman, T. S., G. M. Fimognari, and I. S. Edelman. Studies on renal aldosterone—binding proteins. J. Biol. Chem. 243:3849. 1968. Steel. R. G. D., and J. H. Torrie. Principles and procedures of statistics. McGraw-Hill Book Co., New York. 1960, p. 173. Fanestil. D. D., and I. S. Edelman. Characteristics of the renal nuclear receptors for aldosterone. Proc. Natl. Acad. Sci. 56:872. 1966. Sander, S., and A. Attramadal. An autoradiographic study of oestradiol incorporation in the breast tissue of female rats. Acta Endocrin. 58:235. 1968. APPENDIX iv... -1- 1 «1.1. ‘“ -(rx 1». ‘o-na ‘ APPENDIX TABLE 1.--Chemica1 Formulation for Disc Electro- phoresis (Canalco Formulation)* Stock Solutions Component Quantity Component Quantity A) l N HCl 48 ml E) Riboflavin 4 mg Tris** 36.3 g H20 100 m1 Temed*** 0.23 ml H20 to 100 ml F) Sucrose 40 9 H20 to 100 ml B) 1 N HCl 48 ml Tris 5.98 g G) Ammonium Temed 0.46 ml persulfate 0.14 9 H20 to 100 ml H20 100 m1 C) Acrylamide 28 g H) 10X buffer Bis**** 0.735 g Tris 3.0 9 H20 to 100 m1 Glycine 14.4 9 H20 to 1000 ml D) Acrylamide 10 g Bis 2.5 9 H20 to 100 m1 Working Solutions Separating Gel 1 part A 2 parts C 1 part H20 (mix 1:1 with G to polymerize) Stacking Gel 1 part B 2 parts D 1 part E 4 parts F (expose to fluorescent light to polymerize) *Standard Gel (7%)--stacks at pH 8.9, runs at pH 9.5. **Tris = 2-amino-2-hydroxymethyl-l,3-pr0panediol. ***Temed = N,N,N',N'-tetramethylethylenediamine. ****Bis = N,N'-methylenebisacrylamide. 176 APPENDIX TABLE 2.--Composition of scintillation fluid. V‘ Component Quantity Dioxane 770 ml Xylene 770 ml Absolute ethanol 460 ml Naphthalene 160 g PPO* 10 g POPOP** 0.1 g *2,5-diphenyloxazole. **l.4-bis-[2-(4-methyl-5-phenyloxazolyl)]—benzene.