United States Patent (19) 11 Patent Number: 4859,585 Sonnenschein et al. 45 Date of Patent: Aug. 22, 1989

54 IN-VITRO METHODS FOR IDENTIFYING Schatz et al., Endocrinology, 115: 501-506 (1984). COMPOSITIONS WHICH ARE AGONSTS Paz et al., J. Cell Biol, 99: 339a, No. 1256 (1984). AND ANTAGONSTS OF ESTROGENS Sonnenschein & Soto, Proc. AACR, 26: 198, No. 783 (1985). (75 Inventors: Carlos Sonnenschein; Ana M. Soto, Soto et al., Proc. AACR, 26: 198, No. 784 (1985). both of Boston, Mass. Soto & Sonnenschein, Endo. Soc., 67th Annual Mtg., 73 Assignee: Trustees of Tufts College, Medford, No. 1152 (1985). Mass. Sonnenschein & Soto, Endo. Soc., 67th Annual Mtg., No. 663 (1985). 21 Appl. No.: 853,240 Sonnenschein et al., Life Sciences, 37:387-394 (1985). 22 Filed: Apr. 17, 1986 Soto & Sonnenschein, J. Steroid. Biochem., 23: 87-94 51) Int. Cl* ...... C12O 1/02 (1985). 52 U.S. Cl...... 435/29 Schatz et al., J. Cell. Physiol., 124: 386-390 (1985). 58) Field of Search ...... 435/29 Papendorp et al., J. Cell. Physiol, 125: 591-595 (1985). 56 References Cited Primary Examiner-Esther M. Kepplinger Assistant Examiner-David A. Saunders PUBLICATIONS Attorney, Agent, or Firm-David Prashken Coezy et al, Cancer Research, 42, 317-323, 1982. 57 ABSTRACT Reddeletal, Cancer Research, 43, 4618-4624, 1983. Soto et al, Biochem. Biophys. Res. Communic, 122, The present invention provides two general protocols 1097-1103, 1984. by which a substance may be identified and character Biotechnology, Fisher Scientific, 1983, p. 97. ized as an estrogen agonist and/or an estrogen antago Sonneschein et al., J. Cell Biol, 99, 3392, 1984. nist. The protocols are in-vitro methods which utilize Sonneschein et al, Biol. Abstr., 80, Abstr. No. 69567, estrogen dependent cells in culture and a medium com 1985. prising an inhibitor endogenous to the sera of adult Stack & Gorski, Endocrinology, 115: 1141-1150 (1984). humans which is able to prevent the proliferation of Sonnenschein & Pierce, Cancer Research, 41: 4742-4743 cultured cells in-vitro. The methodology is rapid, repro (1981). ducible, accurate and provides the major advantage of Laugier et al., Proc. Natl. Acad. Sci., USA, 80: 1621-1625 being able to screen large numbers of substances for (1983). their primary or secondary side effects in a large variety Soto & Sonneschein, J. Cell Biol, 97: 393a (1983). of medical and environmental applications. Soto & Sonnenschein, Proc. Am. Assoc. Can. Res., 25: 213 (1984). 13 Claims, 6 Drawing Sheets

---O-O-O---a : A to a - O Iol2O-ll IO-IOIO-9 O-8M CONCENTRATION PER WELL O is ESTRADIOL 8 a NACTIVE A a FULL ANTAGONST U.S. Patent Aug. 22, 1989 Sheet of 2 4859,585

A-A-A- IO S 9 8 7 6 a 5 4 3. O------p 2 ---O-O-O----a a - - - - f -- - - O IO-2 IO-ll IO-IO IO-9 O-8M O O-12 O-O-IOIO-9 O-8M CONCENTRATION PER WELL CONCENTRATION PER WELL O POSITIVE CONTROL (CELSESTRADIOL-CDHS) O a ESTRADIOL Or NEGATIVE CONTROL (CELLS+CDHSWITHOUTESTRADIOL) O = |NACTIVE &-PARTIAL ANTAGONIST (CELLS+ESTRADIOL+TEST SUBSTANCE) A - FULL ANTAGONIST Cy-FULLANTAGONIST (CELLS+ESTRADIOL+TEST SUBSTANCE) F G FG.2

- EMPERICAL DATA 5 ------EXTRAPOLATED DATA

5 O 5 20 25 3O f M CONCENTRATION PERCENTAGE (PROCEDUREL) 3 6 9 O5 O 5 CALENDARDAYS MLLIGRAMS OF PROTEIN (PROCEDUREI) FG.3 F. G. 4 U.S. Patent Aug. 22, 1989 Sheet 2 of 2 4859,585

-O-O-O-O-O-O-O-O."

O O.2 O4 O6 O.8 O .2 INHIBITOR CONCENTRATION (MILLIGRAMS PROTEIN) () = CELS-- INHIBITOR PREPARATION O = CELLS-- INHIBITOR PREPARATION+ESTRADIOL F.G. 5

10% CHARCOAL - 10% CHARCOAL - DEXTRAN STRIPPE) DEXTRAN STRIPPED MALE HUMAN FEMALE HUMAN SERA SERA

NO ESTRADOL

3 x 10" ESTRADIOL 4,859,585 2 Macmillan Publishing Company, 1985, pages N-VITRO METHODS FOR IDENTIFYING 1421-1423. Accordingly, there has been increasing COMPOSITIONS WHICH AREAGONISTS AND interest in finding newly synthesized compositions and ANTAGONSTS OF ESTROGENS new uses for previously known compounds which are demonstrably estrogenic, that is, able to induce estrus. FIELD OF THE INVENTION Probably the best known examples are: ethinyl estradiol The present invention is concerned generally with (Estinyl); 3-methyl-ethinyl estradiol (Mestranol); and methods for identifying chemical compositions whose diethylstilbestrol (DES); methallenestril (Vallestril); pharmacological activities mimic or contradict the ac and doisynoestrol (Fenocylin). tion of naturally occuring estrogens and is particularly 10 Despite the apparent wide interest in compositions directed to in-vitro protocols which will classify phar said to be estrogenic, surprisingly few assay methods macologically active compositions as being either full are available to detect and identify those compositions or partial agonists and/or full or partial antagonists of which have the pharmacological ability to initiate es estrogens. trus. To date, only three are recognized: the cornifica 15 tion of vaginal cells in the spayed rodent; the change in BACKGROUND OF THE INVENTION weight of excised wet uterine tissue; and the ability of Naturally occurring or endogeneous estrogens con the composition to bind competitively with radiola stitute one class of steroid sex hormones which are beled estradiol in cytosol preparations, and homoge produced in the ovaries and other tissues in the body mates of immature uterine cells. Each of these will be and which stimulate the growth and development of the 20 briefly summarized. secondary sex characteristics in female animals. The The cornification of vaginal cells is the classical naturally occurring estrogens are estrone (also known methodology by which the naturally occurring endoge as E), estradiol-17B (also known as E2), and estriol nous estrogens were discovered. The assay methodol (also known as E3) The secretion of such estrogens ogy has remained substantially unaltered since its first controls in major part the normal sexual cycle in hu 25 use Allen and Doisy, J. Am. Med. Assoc. 81:819 (1923). mans and animals which appears characteristically as The assay is based upon the ability of a substance to the changes of estrus, a period of mating activity produce the typical estrus response in-vivo when in marked by intense sexual urge (sexual heat). Historically, it has been demonstrated that removal of jected into castrated rats or mice. A positive reaction is both ovaries from an adult normal female mammal abol 30 characterized by the distinct change in the structure of ishes estrus, that is the period of intense sexual urge and the cells lining the vagina in which the cells acquire a mating activity; equally important, it was found that unique, cornified appearance easily distinguishable from entering into the state of estrus is accompanied and cells in the resting state. Microscopic examination of characterized by visually identifiable and distinct vaginal smears provides empirical evaluation of the changes in the cell structure of the cells lining the va 35 estrus condition of the living animal. Traditionally, the gina. At the height of estrus, the cell lining acquires a activity of a substance under test is expressed in terms of unique, cornified character easily distinguishable from mouse or rat units. A mouse unit is the quantity of en those cells present when the animal is resting, in a non dogenous estrogen that just suffices to produce estrus in estrus condition, in an immature animal which does not a castrated animal. Typically, the mouse unit varies display full sexual activity of an adult, or has been cas 40 from about 0.04-0.1 microgram (hereinafter "ug”) of trated. On this historical and technical basis, the natu pure estrone; the international unit of activity has been rally occurring estrogens were first isolated and puri set as 0.1 ug of estrone. fied. Concomittantly, the term "estrogen' is defined The uterine wet weight assay identifies substances only in operational form as-any substance able to in which induce estrus by measuring the increase in the duce estrus in a living animal. In consequence, all other 45 mass weight of uterine tissues, the increase in weight pharmacologically active compositions, steroidal and being directly attributable to the interaction between nonsteroidal which were subsequently synthesized and those substances and the secondary sex organs. The test which were demonstrably able to induce estrus or to utilizes immature and/or adult ovariectomized mice and mimic the induction of estrus have been conventionally rats which are administered a test substance subcutane classified as "estrogenic' regardless of their chemical 50 ously using an inert fluid such as corn oil. Groups of six structure or their mechanism of action. to ten animals are typically injected daily for three days The value of naturally occuring estrogens and syn with the putative estrogenic substance and sacrificed on thetic compositions demonstrating "estrogenic' activ the fourth day. The uteri are removed, dissected free of ity has been in their medical and therapeutic uses, many extraneous tissue and the fluid contents expelled. Each of which continue to be the subject of considerable 55 excised uterine tissue is then blotted and the wet weight controversy. A traditional listing of the therapeutic determined individually. The increased wet weight of applications for estrogens alone or in combination with the uterine tissue in comparison to those from control other active agents includes: oral contraception; relief animals receiving only the inert fluid identifies a potent for the symptoms of menopause; prevention of threat estrogenic composition Astwood, Anat. Rec. 70:5 ened or habitual abortion; relief of dysmenorrhea; relief 60 (1938); Kay et al., Biochem. Biophys. Acta. 261:475 of dysfunctional uterine bleeding; an aid in ovarian (1972); Black et al., Life Sciences 26:1453-1458 (1980). development; treatment of acne; diminution of exces The competitive binding assay is based on the exis sive growth of body hair in women (hirsutism); the tence of a specific estrogen receptor (hereinafter "ER”) prevention of heart attacks; treatment of osteoporosis; on the cell or tissue extract to which all putative estro an alternative to surgery for metastatic breast cancer; 65 genic compositions will specifically bind in some mea treatment of prostatic carcinoma; and suppression of surable degree. The binding of the substance to the ER post-partum lactation Goodman and Gilman, The Phar site is believed by many investigators to be responsible macological Basis Of Therapeutics (Seventh Edition), for initiating both protein synthesis and cell prolifera 4,859,585 3 4. tion in those cells and tissues having demonstratable ER and Thompson, Proc. Soc. Exp. Biol. Med. 91:623-625 sites. Although initially described as an in-vivo tech (1956)). Subsequently, the related non-estrogenic com nique Jensen and Jacobson, Rec. Prog. Horm. Res. pound, ethanoxytriphenol, was found to be strikingly 18:387-414 (1962)) the preferred technique is an in-vitro antiestrogenic. It inhibited the activity of endogeneous methodology in which binding of the estrogenic sub- 5 estrogen as well as of synthetic estrogens such as DES stance occurs in a homogenate of rat uterine tissue and chlorotrianisene. Subsequently, a large number of which has been prepared as a cytosol, the supernatant other antiestrogens have been identified and classified fraction containing the soluble proteins after sedimenta according to their chemical structure into two different tion at 105,000Xgravity for one hour Noteboom and classes: steroidal antiestrogens such as RU16117 which Gorski, Arch. Biochem. Biophys, 111:559-568 (1965); 10 dissociates rapidly from the ER site and RU39411 Talwar et al., Proc. Natl. Acad. Sci. U.S.A. 52:1059-1066 which forms a more stable complex with the ER pro (1964)). This supernatant fraction, the cytosol, is typi tein; and non-steroidal antiestrogens including the tri cally combined with the test substance for 30 minutes at phenylethylenes such as tamoxifen and its various deriv 4. C. in a test tube. Controls are prepared using inert atives, enclomiphene, clomiphene, nafoxidine, fluid carriers and non-specific binding is determined in 15 LY117018 and Keoxifene Jordan, V., Pharmacological parallel preferrably using a synthetic estrogen such as Reviews 36:245-276 (1984); Jordan et al., Cancer Treat. DES. Subsequently, radiolabeled Hestradiol is added Rep. 64:745-759 (1980); Heel et al., Drugs 16:1 (1978); to all samples at a predetermined concentration. The Legha et al., Ann. Intern. Med. 88:69-77 (1978)). incubation is continued after which a suspension of An "antiestrogen' is usually defined as a compound dextran-coated charcoal in a suitable buffer is added to 20 that will inhibit the vaginal cornification produced by all the samples and allowed to react for approximately estradiol in ovarectomized rats and/or will inhibit the 20 minutes' duration. Each control and test sample is increase in uterine weight produced by estradiol in then centrifuged to yield a discardable precipitant and a immature rats. The existence of a specific antiestrogenic supernatant whose radionuclide content is measured by binding site in addition to the traditionally accepted ER liquid scintillation counting. Jordan et al., J. Endocr. 25 binding site on specific cells and tissues has been an area 75:305-316 (1977)). of intense controversy and study. Solely to test this It is noteworthy that all the presently accepted assay theory, two in-vitro test systems have been developed: techniques, whether in-vivo or in-vitro, are based upon the growth of MCF-breast cancer cells - to specifically a single common mechanism of action for all estrogens determine the effects of tamoxifen and its metabolites on as a class (including endogeneous estrogens, syntheti- 30 cell proliferation Coezy et al., Cancer Res. 42:317-323 cally prepared estrogens and other substances demon (1982); Reddel et al., Cancer Res. 43:4618-4624 (1983) strating estrus inducing activity). The theory accepted and estrogen-stimulated prolactin synthesis by primary almost exclusively is the existence of putative intracellu cell cultures of immature rat pituitary glands-for the lar receptor proteins for the estrogenic substance in the ability of antiestrogens to inhibit prolactin synthesis cells of estrogen responsive tissues such as the vaginal 35 Lieberman et al., J. Biol. Chem. 258:4734-4740 (1983); lining, uterine tissue, the female breast, the pituitary, Lieberman et al., Proc. Natl. Acad. Sci. U.S.A. and the hypothalamus. Estrogens and estrogenic com 75:5946-5949 (1978)). positions bind with high affinity to the intracellular The ability to identify and to distinguish between receptor protein, termed the estrogen receptor or “ER” estrogenic compounds and antiestrogens is recognised site. After binding the formed estrogen-protein complex 40 as critical by physicians and pharmacologists. Naturally is said to be converted into a species that is physically occurring endogenous estrogens are able to induce a translocated to the nucleus of the cell where further variety of multiple changes in their target organs: in binding of the estrogen to the genetic material occurs. creased cellular proliferation or hyperplasis; induction Some recent publications have confirmed the data re of specific protein synthesis; and an increased cellular ported previously Mester et al., Exp. Cell Res. 45 mass or hypertrophy. From the viewpoint of clinicians 81:447-452 (1973) suggesting that estrogens bind di and human disease, the hyperplastic effects are the most rectly to unoccupied nuclear ER sites avoiding the important because it is central to the problem of hor translocation step King and Greene, Nature mone-sensitive cancers and their control. Estrogen an 307:745-747 (1984); Welshons et al. Nature 307:747-749 tagonists thus are of primary interest. From the view (1984). Subsequent to these events, a general increase 50 point of pharmacologists interested in developing new cell proliferation is observed. Gorski and Gannon, drugs useful for the treatment of human diseases and Annu. Rey, Physiol. 38:425-450 (1976); Gorski et al., specific pathological conditions, it is most important to Recent Prog. Horm. Res. 24:45-72 (1968); Jordan, Phar procure compounds with some demonstrable estrogen macological Reviews 36:245-276 (1984). like function but which are devoid of proliferative side It is ironic that the effort to procure new estrogenic 55 effects. Exemplifying this latter view, osteoporosis, a compositions and to obtain evidence which reinforces disease in which bone becomes increasingly more frag the hypothesis that a single ER site for estrogen is pres ile, is greatly ameliorated by the use of fully active ent in estrogen sensitive cells and tissues has led directly estrogens; however, due to the recognized increased to the recognition and isolation of several different risk of cancer in patients chronically treated with active kinds of compounds that in fact inhibit or neutralize the 60 estrogens, it is not clinically advisable to treat osteopor action of endogenous estrogens. These have been osis with fully active estrogens. Accordingly estrogen termed "antiestrogens'. Historically, the weakly estro agonists are the primary interest and focus. genic compound chlorotrianisene, unlike most estro The difficulties of identifying and evaluating estrogen gens, was observed not to cause enlargement of the agonistic compositions and estrogen antagonistic com pituitary when given to rats in large doses. Estrogens 65 pounds are recognized. In-vivo methods to detect and normally cause pronounced enlargement of the pitui identify such compositions have been difficult to per tary, but when chlorotrianisene was given concurrently form in a consistent manner and are open to indepen with estradiol, the effect was greatly reduced Segal dent biological variations which cannot be entirely 4,859,585 5 6 controlled or eliminated. Equally recognized as a con ology is rapid, technically easy to perform, and pro tinuing problem are the mechanisms of action for estro vides meaningful data in a precisely controlled in-vitro genic and antiestrogen compositions-the evidence for assay environment. which is often conflicting, extremely complex, and fre quently inconsistent. Each class of composition has DETAILED DESCRIPTION OF THE FIGURES been shown to have different properties in estrogen The present invention may be more easily and com target tissues and to induce radically different pharma pletely understood when taken in conjunction with the cological reactions among alternate species of test ani accompanying drawing, in which: mals. The best example of the latter is the number of FIG. 1 is a graph illustrating the interpretative basis estrogenic and antiestrogenic compositions which have 10 for assessing empirical data to determine whether a been demonstrated to be active in rats and mice but substance is an estrogen agonist; which have subsequently been found to be inactive or FIG. 2 is a graph illustrating the interpretative data poorly effective in clinical trials using human subjects. by which a substance may be identified as an estrogen Underlying and inherent in each of these problems and antagonist; and difficulties, is the seemingly widely accepted require 15 FIG. 3 is a graph showing the exponential growth ment to conform to the generally accepted theory of rate of cells in culture; intracellular ER sites on the target cells and tissues FIG. 4 is a graph illustrating the correlation between before any in-vivo or in-vitro assay methodology may number of cell doublings as a function of inhibitor activ be deemed effective. As can be readily appreciated, ity; there is a continuing need for novel, reliable, accurate, 20 FIG. 5 is a graph illustrating the ability of estradiol to and sensitive assay protocols which will identify those neutralize the activity of the inhibitor; and pharmacological substances which are demonstrably FIG. 6 illustrates the actual visual display resulting active as estrogens from those which are inhibitors or from the culture of MCF7 cells for 10 days in a medium neutralizers of estrogens and from those substances containing charcoal-dextran stripped human male and which are completely quiescent with respect to estro 25 female sera at 10% concentration. gens. SUMMARY OF THE INVENTION DETAILED DESCRIPTION THE PREFERRED EMBODEMENTS The present invention comprises in-vitro methods for identifying and characterizing a test substance as an 30 The present invention is a general, in-vitro test meth estrogen agonist and as an estrogen antagonist. The odology which is used in alternative modes and pro method for identifying an estrogen agonist comprises vides distinct protocols for the identification and char the steps of obtaining a plurality of cells cultured in acterization of unknown substances for their capacity to vitro, these cells being estrogen dependent for prolifera serve as agonists and/or antagonists of endogenous tion in-vivo; maintaining a known quantity of these 35 estrogens. cultured cells in a medium comprising a fluid and an There are multiple advantages provided by the meth inhibitor endogenous to the sera of adult humans, this odologies comprising the present invention. These in endogenous inhibitor being present in the medium at a clude: concentration effective to prevent proliferation of the 1. The ability to identify available or newly synthe cells in-vitro; adding the test substance to the cultured 40 sized compounds as estrogen agonists and/or estrogen cells and the maintaining medium to form a reaction antagonists in a verifiable and reproducible manner mixture; incubating the reaction mixture for a prese within a controled in-vitro assay environment. All of lected period of time; and determining the number of the experimental parameters and factors are predeter cultured cells in the incubated reaction mixture, a mea mined and carefully prepared to lie within selected surable increase in the number of cultured cells serving 45 limits-thereby eliminating the effects of individual to identify the test substance as an estrogen agonist. biological variation common to all in-vivo studies. The general method for identifyinq an estrogen an 2. The capability to test statistically significant num tagonist comprises the steps of obtaining a plurality of bers of different substances over a relatively short per cells cultured in-vitro, these cells being estrogen depen iod of time. Although the optimal time for each test is dent for proliferation in-vivo; maintaining a known 50 ten days in duration, each day's testing is performed in quantity of these cultured cells in a medium comprising staggered series; accordingly, several hundred individ a fluid and an inhibitor endogenous to the sera of adult ual test samples can be prepared within one average humans, the endogenous inhibitor being present in the working day and a different compound can be tested medium at a concentration effective to prevent prolifer each day following in succession. Over a two week time ation of the cells in-vitro; adding the test substance and 55 period, at least five different days' testing can be empiri an active estrogen to the cultured cells and the main cally evaluated. taining medium to form a reaction mixture; incubating 3. The means for reducing the present costs of testing the reaction mixture for a preselected period of time; are provided by each protocol of the present invention. and determining the number of cultured cells in the In comparison with present in-vivo animal studies and incubated reaction mixture, the failure of said cultured 60 their concomittant high cost per animal, their limited cells to measurably increase in number serving to iden availability in numbers, and their inconvenience (hous tify the test substance as being an estrogen antagonist. ing, maintenance, and handling), the described in-vitro Although each method comprising the invention may methods are very inexpensive to perform. Furthermore, be performed to advantage individually and indepen in view of the major budget constraints now commonly dently, it is preferred that both general methodologies 65 in effect as a limiting factor in the realization of bringing be performed concurrently or sequentially in order to a new therapeutic composition to market, the present obtain the broader range of information and data about invention offers a cost-effective procedure for evaluat the substance being evaluated. In addition, the method ing the many presently synthesized compounds which 4,859,585 7 8 were considered marginal and thus unlikely of ever compositions examplified by dienestrol, benzestrol, hex being evaluated via in-vivo testing. estrol, methestrol, diethylstilbestrol (DES), quinestrol 4. The means for correlating the pharmacological (Estrovis), chlorotrianisene (Tace), and methallenestril properties of the substance under test directly with its (Vallestril). The pharmacological activities of these individual structural organization and chemical formu compositions often present apparently conflicting prop lation. The protocols of the present invention are com erties. For example, DES is noted for its ability to pletely unlike the known in-vivo assay method and do mimic naturally occuring estrogens and has little dem not rely upon the appearance of gross anatomical onstrated capacity to inhibit or neutralize the pharma changes such as cornification of vaginal cells or the cological action of naturally occuring estrogens; on the increase in weight of excised uterine tissues. Moreover, O other hand, chlorotrianisene only weakly mimics the unlike the competitive protein binding assays, the de activity of naturally occuring estrogens but is demon scribed protocols are not based upon the pre-existence strated to be a potent inhibitor or neutralizing agent of identifiable estrogen receptors proteins and do not when combined with naturally occuring estrogens. Ac require that ER-induced new protein synthesis be di cordingly, following the preferred terminology and rectly linked with cell proliferation. The present inven 15 usage employed herein, DES is properly identified as a tion, instead, permits the user to take into account the full or complete agonist with no antagonistic properties; mechanism of action of the substance under test as it while chlorotrianisene is properly characterized as a demonstrates its individual pharmacological activity; partial agonist and a complete antagonist of naturally and it allows a more detailed evaluation of the struc occuring estrogens. ture-function relationship of the test substance. 20 Full or complete agonist: A compound that either As a result of the historical development in this art, a produces or mimics the effect of naturally occuring number of terms have come into common usage which estrogens. are non-informative at best and misrepresentative and Partial agonist: A compound that either produces or distortive in the worst instances. These terms include mimics the effects of naturally occuring estrogens in "estrogenic", "estrogenicity', and "antiestrogen'. 25 some degree but which is markedly inferior in activity Equally difficult and distortive is the common practice and effect to the results obtained with a full agonist even of directly linking the ability of potent estrogens to at the highest attainable dosage. initiate new protein synthesis in a target cell or tissue Full or complete antagonist: A compound which with the distinctly different ability of the active estro completely inhibits or neutralizes the effect of a natu gen to induce cell proliferation. Contrary to this com 30 rally occuring estrogen when administered simulta mon habit, the present invention relies and depends neously with such an estrogen. upon the ability to separate each of these characteristic Partial antagonist: A compound which only partially abilities and to evaluate them independently without neutralizes or inhibits the effect of a naturally occuring direct interdependence between them. estrogen when administered simultaneously with that To avoid confusion and miscommunication, and for 35 estrogen, even at the highest attainable dosage. greater ease and clarity of understanding of both the For purposes of the present invention, it is only the basis of the present invention and the major advance in ability of naturally occurring estrogens to induce the this art which it represents, a set of strictly-adhered-to proliferation of their target cells and tissues through an definitions and terms will be employed herein. It will be indirect mechanism of action which is of interest. For expressly noted, however, that the terms "estrogenic' this reason, all substances under test are evaluated for and "antiestrogen' have no well-defined meaning and proliferative activity only with respect to naturally will not be used at any time in the description which occurring estrogens. The present invention is based follows. Instead, the following terms and definitions are upon and utilizes the empirical finding that human employed exclusively. serum and plasma contain a demonstrable endogenous Estrogens: A class of compounds including naturally 45 inhibitor which prevents the proliferation of estrogen occuring and synthetically made compositions which dependent cells and tissues in-vivo and in-vitro Son have a demonstrated ability to induce cell proliferation nenschein and Pierce, Cancer Research 41:4742-4743 in-vivo and in-vitro. Cells and tissues said to be estrogen (1981); Soto and Sonnenschein, J. Cell Biol. 97:393a dependent are those cells and tissues which require one (1983); Paz et al., J. Cell Biol. 99:339a (1984); Schatz et or more of the naturally occurring estrogens to be pres 50 al., Endocrinology 115:501-506 (1984); Soto and Sonnen ent for the cells and tissues to proliferate in-vivo. Syn schein, Biochem. Biophys. Res. Comm. 122:1097-1103 thetic estrogens mimic the characteristics of naturally (1984); Soto and Sonnenschein, Proc. Am. Assoc. Cancer occuring estrogens to induce proliferation of cells, tis Res. 25:213 (1984)). Each of the in-vitro methods for sues, and organs in-vivo and in-vitro. As used herein, identification of estrogen agonists and estrogen antago the ability of naturally occuring estrogens and/or syn 55 nists makes use of the indirect-negative mechanism thetic estrogens to induce or to fail to initiate new pro present in-vivo, in which estrogens demonstrably can tein synthesis in specific targeted cells and tissues is of cel the effect of plasma-borne specific inhibitors of es no consequence. trogen-sensitive cell proliferation. The use of this spe Naturally occuring estrogens: The three most com cific mechanism of action, within an in-vitro methodol mon, naturally occuring estrogens are: estrone (E), 60 ogy for the characterization of unknown test substances estradiol-17B (E2), and estriol (E3) Of these, estradiol is as agonists and/or antagonists of naturally occuring the most active pharmacologically and is the estrogen estrogens, was previously unknown and unappreciated of choice in the test methodologies comprising the pres in this art. ent invention. The present invention is preferably utilized as a com Synthetic estrogens: A class of compounds not occur 65 bination of two different in-vitro methodologies per ring in nature which duplicate or mimic the activity of formed concurrently or in sequence. One test methodol endogenous estrogens in some degree. These com ogy identifies a test substance as an estrogen agonist pounds include a variety of steroidal and non-steroidal that is, a test substance which duplicates or mimics the 4,859,585 10 effects of naturally occuring estrogens completely or mately four hours duration. The sera are then clarified partially. The second methodology evaluates the test by centrifugation (3,000 rpm for ten minutes) and subse substance as an estrogen antagonist-that is, the ability quently “heat inactivated’ (30 minutes at 56' C.), fil to inhibit or neutralize a naturally occuring estrogen tered through a 0.45 micron pore filter (Millipore Cor when both are administered simultaneously. The indi 5 poration, Bedford, Massachusetts,) aliquoted into said vidual methods to identify an agonist or antagonist of volumes, and stored at -20 C. for future use. Subse estrogens employ common cells, reagents, and follow quently, the sera are stripped of endogenous estrogens similar protocols. For this reason, the constituents com by extraction with 0.5% charcoal in the following man monly shared and used in both methods, albeit for dif ferent purposes and goals, will be described. O A 0.5% charcoal (Norit A, Acid washed, Sigma Cor Cells poration)-0.05% dextran T70 (Pharmacia Corp., Pis cataway, N.J.) suspension is prepared. The charcoal A wide variety of different cells in culture may be dextran aliquots in volumes similar to the volume of the employed in the identification methods described herein serum aliquots to be processes are centrifuged at 3000 Such cell lines are maintained by tissue culture methods 15 and medium commonly available and used in the art, rpm for 10 minutes to pack the charcoal into a pellet. many of which are commercially prepared and sold. The supernatants are aspirated to yield the formed char The sole essential requirement to be demonstrated by coal pellet. Each serum aliquot is mixed with the char the culture cell line is the demonstrated need for a natu coal pellet and maintained in suspension by means of a rally occuring estrogen to be present before the cells 20 roller apparatus set at 3 cycle/minute at 37.5 C. for 60 proliferate in-vivo-that is, in their naturally occuring minutes duration. To monitor the extraction efficiency, state within the living animal. Mammalian cell lines are comparable volumes of each sera are equilibrated for 16 preferred, with human cell lines being more preferred hours at room temperature with radiolabeled estradiol than cell lines from mice, rats, and the like. In addition, at 10-9M and 10-8M concentrations respectively, prior it is preferable in many instances that the cultured cell 25 to charcoal extraction. It is found that 99% of the radi line be an abnormal or tumor cell line derived from olabeled estradiol is removed by this treatment. cancers or tumors in the living animal. Desirable cell Each of the prepared sera are combined with the lines include: MCF7 cells, passage number 173 (Michi charcoal pellets and are consequently stripped of en gan Cancer Foundation, Detroit, Mich.) and the dogenous estrogens via this extraction method. Subse CMCF7-173 clone which are derived from human 30 quent testing demonstrates that the endogenous inhibi breast tumor cells (Tufts University, School of Medi tor remains stable in the treated serum even after 5 cine, Boston, MA); T47D-All cells, a clonal population successive charcoal-dextran extractions. from the uncloned T47D cells originally described by Within the preferred protocols to identify agonists Keydar et al. Eur. J. Cancer 15:659-670 (1979); In and antagonists of estrogens, the CDHS is added to the stituto de Biologiay Medicina Experimental, Buenos 35 DME in quantities sufficient to form a final concentra Aires, Argentina; Rockville, MD.); and clones of the tion ranging from 40-10% by volume. This CDHS and H301 cell line of the Syrian hamster kidney adenocarci DME in combination comprises the preferred maintain noma originally described by Sirbasku and Kirkland ing medium for use in the protocols. The actual concen Endocrinology 98:1260 (1976); Tufts University, trations of endogenous inhibitor present in the CDHS School of Medicine, Boston, MA). Each of these cul thus are also diluted in concentration to a working tures cell lines are demonstrably estrogen dependent for range of between 40-10% of its original concentration proliferation in-vivo. Prior to use in the protocols of the in adult human serum. present invention, each of the cell lines are routinely grown in 5% (v/v) fetal bovine serum (hereinafter Estrogens "FBS') commercially obtained (Sterile Systems Inc., 45 The estrogens preferred for use with the present in Logan, Utah) in Dulbecco's modification of Eagle's vention comprised the natural endogenous estrogens Medium (hereinafter “DME') in an atmosphere of 5% CO2/95% air under saturating humidity at 37° C. At the (estrone, estradiol, and estriol) and their various chemi time of test, each cell line is preferably introduced at the cal forms such as alpha-E2, beta-E2, and the like. If stated aliquot concentrations into the wells of multiwell 50 desired, it is possible to utilize any of the wide variety of plates (Costar 3512 plates, Cambridge, MA). The cell synthetically made estrogens as a comparative basis. aliquot in each well is allowed to attach to the surface of However, this latter approach is not preferred because the well for a period of 6-24 hours after which the of the wide range of steroidal and non-steroidal compo seeding medium (5% FBS in DME) is removed and sitions which, although able to mimic the effect of en replaced by specific quantities of the maintaining me 55 dogenous estrogens, also demonstrate a variety of an dium used during the testing phases. Preferably, the tagonistic effects. The estrogen of choice for each of the fluid comprising the maintaining medium was DME in-vitro methods to identify agonists and antagonists of supplemented with various concentrations of charcoal estrogens is estradiol (E2). dextran stripped human serum (hereinafter "CDHS'). The Preferred Protocols Alternatively, one may use 2% charcoal-dextran 60 stripped human serum supplemented with known con It is most preferred that the present invention be centrations of an active inhibitor preparation as will be performed in two test phases in order that a single sub described hereinafter. stance under test be evaluated and identified in terms of its agonistic and antagonistic properties with respect to Charcoal-Dextran Stripped Human Serum (CDHS) 65 estradiol. For ease and simplicity, the phase 1 test iden Venous blood is drawn from healthy human adult tifies agonistic characteristics; in turn, the phase 2 test men and women. The blood is allowed to clot on sterile identifies the antagonistic characteristics. Each test 50 milliliter (hereinafter "ml”) plastic tubes for approxi protocol will be described in detail. 4,859,585 11 12 5. To the other prepared (seeded) wells, 100 ul of the Phase 1 Protocol To Identify Estroqen Agonist test substance to be evaluated is added as a 10X stock Properties solution to achieve final concentrations ranging from This protocol measures the potential agonistic prop 10-5M to at least 10-12M. The substance under test is erties of the substance under test. All manipulations are preferably prepared as a stock solution in ethanol, most to be carried out under sterile conditions using dispos preferrably as a 103M solution. All dilutions are made able plastic tissue culture multiwell plates with all ma using DME. The ethanol concentration in the culture nipulations preferrably being performed in a tissue cul medium should be kept below 0.01% by volume. This ture hood. preparative technique avoids the need to filter the vari 1. The cell line of choice is the CMCF7-173 culture, 10 ous volumes of test substance to achieve sterility. This is a cloned estrogen-sensitive breast tumor cell line. The particularly valuable because steroidal compositions are cell line is routinely grown in 5% fetal bovine serum-s- known to become adsorbed to the commercially ob upplemented Dulbecco's modified Eagle minimal essen tained filter membranes used for sterilization. The addi tial medium (hereinafter “5% FBS/DME'), at 37° C. in tion of the test substance in appropriate concentrations a 95% air-5% CO2 atmosphere. 15 to the cultured cells and the maintaining medium forms 2. The cultured cells growing in 5% FBS/DME are a series of individual reaction mixtures, each of which is harvested by treatment with trypsin (1:250)-EDTA varied only by the concentration of the test substance (0.2%). The cells are detached by brief exposure to present in the well. trypsin for five minutes. To avoid excessive proteolysis, 6. After the series of individual reaction mixtures are the reaction is quenched by addition of three volumes of 0 formed, all of the cultured cells under test in each well 5% FBS/DME. The resulting cell suspension is col of the multiwell plate are incubated at 37 C. preferably lected into a sterile centrifuge tube and centrifuged at in an atmosphere of 5% CO2/95% air under saturating 800 x gravity for five minutes' duration. The superna humidity. During this incubation time, it is desired and tant is discarded and the cell pellet resuspended in 5% preferred that the cultured cells employed in the test be 25 maintained in an atmosphere most advantageous for FBS/DME by pipetting gently. The cell suspension survival and growth of that cell line. The incubation thus obtained is diluted into prepared aliquots using 5% period preferably is in the range of from 10-14 days' FBS/DME. Each prepared suspension preferably con duration, although shorter periods of incubation are tains 40,000 cells in a volume of 0.10 ml and it is desir useful under specific circumstances. The 10 day incuba able that the number of cells be counted and verified by 30 tion period is considered optimal in order to allow suffi means of a Coulter Counter (Coulter Electronics, Hia cient proliferation of the cells serving as controls so that leah, Fla.). substantial differences among the positive and negative Subsequently, 1.0 ml aliquots of 5% FBS/DME are controls and the test cell suspensions may exist on a placed in each well of a multiwell (Costar 3512) plate. regular and recurring basis. The number of wells in each plate will correspond to 35 7. After the preferred incubation period of 10-14 days the number of test substances to be evaluated and thus is has elapsed, the cells from each test and control well are merely a question of choice or convenience to the user. individually harvested and counted. Cells are harvested Into each well, 25 microliter (hereinafter "ul') aliquots using alysing solution such as 10% zapoglobin (Coulter of the prepared cell suspension are added using a 1.0 ml Electronics, Hialeah, Fla.) that disrupts the plasma syringe fitted with a 22 gauge needle attached to a re membranes while leaving the cell nuclei intact without petitive Dunn pipetter (Skatron Dispenser). The micro clumping. A preferred lysing solution comprises 10% titer plate is then gently shaken to avoid unequal cell zapoglobin, 0.5% Triton X100, 2 mM mg Cl2, and 15 distribution within each individual well. The seeded mMNaCl in 5 mM phosphate buffer, pH 7.4. The nuclei cells are then allowed to attach to the plastic surface of suspensions are individually counted using an efficient the well proper over a period of 24 hours' duration after 45 cell counter (Coulter Counter Apparatus, model ZF) which the fluid is changed to the experimental maintain and the cell numbers per well (mean-standard devia ing medium. tion) plotted on a logarithmic scale and correlated 3. At the time of test, all the seeded cells in each well against time in days. of the microtiter plate receive 20% heat inactivated 8. To obtain reproducible, accurate, and reliable data charcoal-dextran human stripped serum (CDHS) in 50 consistently using this protocol, a series of quality assur DME at a volume of 0.9 ml per well. Although higher ance measures should be employed. Under the de concentrations of CDHS may be utilized up to and scribed preferred conditions, the cell number per well including 40% concentrations, the 20% concentration obtained from the control containing 3X 10-11M estra is most preferred. diol should preferably be about tenfold higher than the 4. To those wells or plates designated as controls, E2 55 cell numbers obtained in those reaction mixtures which is added in a range of ever decreasing concentration. did not contain any estradiol. These goals are achieved Preferably, a series of controls are prepared containing: using the optimal 10 day incubation period. In addition, no estradiol (the negative control); 3X 10-8 M E2; a separate performance estimate should be obtained by 3X 10-9M E2: 3X10-10M E2: 3X10-11M E2; and measuring the proliferation rate of a separate 10 cell 3X 10-12M E2 (the positive control). Each of these 60 innoculum in a 20% CDHS with and without estradiol respective concentrations of estradiol are achieved by in a 3X 10-11M concentration. From this innoculum, adding a 10X stock solution in a preferred total volume cells are harvested every 48 hours (preferably as tripli of 100 ul of DME. If desired experimental controls cate samples) for the entire 10 day incubation period. A having concentrations of E2 up to 105M may be uti good estimate of the proliferation rate is the mean gen lized to advantage. However, E2 causes maximum cellu 65 eration time, "tD'. "tD' is defined as the time interval in lar proliferation at an optimum concentration of which an exponentially growing culture doubles its cell 3X 10-11M. For this reason, the preferred controls number. td is calculated from the equation a = 1/txln contain the described concentrations of E2. C/Co, where C is the initial cell number, C is the cell 4,859,585 13 14 number at time=t, and a is the instantaneous cell prolif taining media which contains the endogenous inhibitor eration constant (tD is the value in Ct/Cox a would at a concentration effective to prevent proliferation have when C=2C). "tD' is expressed in time units, i.e., in-vitro of the seeded cells in each well. hr. The slopes of the different growth curves are calcu Accordingly, steps 1 through 3 are performed as lated by fitting the experimental data to a straight line stated earlier within the protocol of Phase 1, the full by regression analysis of the pairs in cell numbers/time. description of each individual step not being repro tois defined as the moment at which the seeding medium duced again here. In essence, 10' cultured cells are is replaced by the maintaining medium. In these condi seeded into each well, test and control. The medium is tions, the doubling time (tD) of CMCF7-173 is 36h for changed to 20% CDHS in DME as the maintaining the cultures in 20% CDHS-3X 10-11ME2, and 300 h. 10 medium; and the following combination of estradiol and for the control (estrogenless) cultures. test substance added to the cultured cells to form a 9. The interpretation of the empirically derived data series of reaction mixtures: in which the number of cultured cells present in each of (a) 3x10-11M estradiol; the incubated reaction mixtures is determined, is di (b) 3x10-11M estradiol and 3X 10-6M test sub rectly based on the premise that a measurable increase 15 stance; in the number of cultured cells in those reaction mix (c) 3X 10-11M estradiol and 3X 107M test sub tures containing the test substance identifies the sub stance; stance under test as an estrogen agonist. In other words, (d) 3x10-11M estradiol and 3X 10-8M test sub the substance under test has empirically demonstrated stance; the ability to interact with the endogenous inhibitor 20 (e) 3X 10-11M estradiol and 3X 10-9M test sub present in charcoal-dextran stripped human serum and Stance. has neutralized the effects of the endogenous inhibitor It will be noted that varying concentrations of the test in such degree that the cells proliferate. The ability to substance ranging from approximately 10-6 to 10-9M interact and neutralize the effect of the endogenous are combined with one uniform concentration of estra inhibitor present in the CDHS in a partial or complete 25 diol, 3X 10-11M. This concentration of estradiol has degree also serves to identify the substance under test as been found to be most useful in the majority of test a full agonist or as a partial agonist. The differences are instances; however, the use concentration of estradiol in illustrated by the data represented in FIG. 1. this Phase 2 protocol may be varied to meet the individ As is revealed by the graph of FIG. 1, the basis of ual user's needs or desires. It is required only that one comparison is the demonstrated ability of estradiol to 30 uniform concentration of estradiol (or other naturally induce proliferation of the cell line in-vitro and is empir occuring estrogen) be employed in sufficient but not ically provided by the data in the control wells. A test exceedingly higher concentrations than that required substance tested in the manner similar to that for estra for maximal cell proliferation in combination with vary diol in this protocol, will demonstrate one of three dif ing concentrations of the substance under test. For this ferent growth phenomena: as a full or complete agonist 35 reason, the preferred concentrations for the subtance to which behaves in a manner similar to estradiol although be evaluated are as described; nevertheless, these spe the maximum effect may occur at a use concentration cific test concentrations may be increased or decreased above or below the maximal growth effect induced by in accordance with the individual needs or desires of the estradiol at the 3x 10-11M concentration; as a partial Sc. agonist which shows some ability to induce cell prolif Subsequent to adding prepared concentrations of the eration in comparison to the number of cells maintained test substance and the estradiol to each of the wells in the control wells which do not proliferate at all, but containing aliquots of cells in known quantity to form a this measurable increase never rises to the numerical series of individual reaction mixtures, each of the reac values comparable to those provided by estradiol itself; tion mixtures is incubated in the manner identical to that and as a quiescent composition which provides no sub 45 described in the Phase 1 protocol for the preferred stantial increase in the number of cultured cells over period of between 10 and 14 days. As before, the true those numbers yielded by the control, and accordingly incubation period of time may be varied; nevertheless it is said to have no ability to induce or initiate cell prolif is preferred that the incubation period be not less than eration under the test circumstances. In this manner, by 10 days routinely. comparing the ability of the test substance to initiate or 50 After the preselected incubation period has elapsed, to fail to initiate cell proliferation under the described the number of cells in each of the wells, test and control, test circumstances, such agonistic characteristics as are are determined in the manner identical to that described inherent in the substance under test are demonstrated for Phase 1 testing. The cells are harvested with a lysing and evaluated on a comparative basis with estradiol. solution that disrupts the plasma membranes leaving the 55 cell nuclei intact without clumping. The individual Phase 2 Testing For Identification Of Estrogen nuclei suspensions from each of the wells are then Antagonistic Properties counted separately using a Coulter Counter and the cell This test protocol identifies and measures (on a com number per well (mean-standard deviation) are calcu parative basis) the antagonistic properties of test sub lated accordingly. By definition, an antagonist is a sub stances when simultaneously added with estradiol to a 60 stance which when combined with estradiol results in a prepared cell culture. The protocol is similar to the reduction in the final number of cells in comparison Phase 1 protocol in: the selection of a cell line for cul with the final cell count obtained using estradiol alone. ture in-vitro which is demonstrably estrogen dependent The empirically obtained final cell counts from any test for proliferation in-vivo; the preparation of aliquots of series should be interpreted in accordance with the cells which are subsequently seeded into a plurality of 65 results graphically plotted in FIG. 2. wells in a multiwell plate; the use of 5% FBS in DME As seen in the graph of FIG. 2, the cultured cells as a culture medium until the test phase has begun; and incubated in-vitro in 20% CDHS alone, the negative the use of 20% CDHS in DME as the preferred main controls containing no estradiol or test substance what 4,859,585 15 16 soever, maintain a substantially small number. Accord 2 protocols, a series of illustrative experiments evaluat ingly, the endogenous inhibitor present in the CDHS ing the variety of pharmacologically active substances inhibits the proliferation of the cells in culture thereby which are well known in the literature was undertaken. providing a substantially constant number throughout The substances evaluated as test samples include the the test period. In comparison, the positive controls following: estradiol; 11-B chloromethyl estradiol; Mox comprising 20% CDHS and estradiol without any test estrol; tamoxifen; LY117018; LY1156758; RU3941.1; substance, provides a much higher number of cells and progesterone. Estradiol is Estra-1,3,5(10)-triene throughout the entirety of the test period, a number 3, 17 diol (U.S. Pat. Nos. 2,096,744; 2,225,419; and about one order of magnitude greater than that of the 2,361,847 11B-chloromethyl estradiol is a synthetic negative control in a 10 day assay. In comparison to 10 derivative of endogenous estradiol. Moxestrol is 11B both these positive and negative controls, a test sub methoxy-19-nor-17a-1,3,5(10)-trien-20-yne-3,17 diol stance combined with estradiol may demonstrate a U.S. Pat. No. 3,579,5450. Tamoxifen is 2-[4-(1,2-diphe small decrease in cell number or a substantive decrease nyl-1-butenyl) phenoxyl- N,N-dimethylethanamine in cell number, respectively; a large, substantive de Richardson, Nature 212:733 (1966); British Patent No. crease in cell number identifies a full or complete antag 15 1,064,629; French Addition Patent No. 90,418). onist; such a substance is able to lower the cell growth LY117018 is a triphenylethylene derivative useful as an rate from the maximal to the minimal level comparable antitumor composition Black et al., Life Sciences to the negative controls containing no estradiol whatso 26:1453-1458 (1980). LY1156758 having demonstrated ever. Correspondingly, a partial antagonist will de antitumor activity Bloom et al., Br. Med. J. 2:7-10 crease the cell growth rate to intermediate values, val 20 (1974). RU39411 is said to be an antitumor compound ues lower than the positive control yields containing by its manufacturer Roussel-UCLAF, Paris, France). estradiol alone, yet higher than those of the negative Progesterone is a hormone secreted by the corpus lu controls containing no estradiol whatsoever. It will be teum during the latter half of the menstrual cycle U.S. recognized, by the manipulative steps comprising the Pat. Nos. 2,379,832; 2,232,438; 2,314, 185). Each of these Phase 2 protocol and by the interpretation of empirical 25 compositions were evaluated using the protocols de data, that a test substance able to suppress the ability of scribed earlier for Phase 1 and Phase 2 tests. The results estradiol to interact with and to neutralize the effects of are presented in Table 1 below. TABLE 1. Characteristics Optimal Phase 1 Optimal Phase 2. With Respect To Intended Primary Test Cell Test Cell To Endogenous Use Prior To Compound Concentration Count Concentration Count Estrogens Testing No additions 0.00 20,000 0.00 20,000 None None (control) Estradiol 3 x 10-11 M 200,000 3 x 10-11 M 200,000 Full Agonist Estrogen No Antagonistic Activity 11B-Chloromethyl 3 x 1012M 200,000 3 x 10-12M 200,000 Full Agonist Estrogen Estradio No Antagonistic Activity Moxestrol 3 x 10-12M 200,000 3 x 10-12M 200,000 Full Agonist Estrogen No Antagonistic Activity Tamoxifen 3 x 10-6M 60,000 3 x 10-6M 40,000 Partial Agonist Antitumor Partial Antagonist Agent LY11708 3 x 10" 10 M 40,000 3 x 108 M 20,000 Partial Agonist Antitumor Full Antagonist Agent LY156758 3 x 10-9 M 50,000 3 x 10-8M 20,000 Partial Agonist Antitumor Full Antagonist Agent RU39411 No effect in 20,000 3 x 10 ll M 20,000 No Agonistic Activity Antitumor range from 102 Full Antagonist Agent to 0-6M Progesterone No effect in 20,000 No effect in 200,000 Quiescent: Hormone range from 1012 range from No Agonistic activity to 10-6 M 1012 to 10-6M No Antagonistic Activity the endogenous inhibitor present in the serum, is an 50 estrogen antagonist. The degree with which the sub stance suppresses the effects of estradiol (or any other It will be recognized that the empirical results are endogenous estrogen), will identify and characterize the expressed as the final cell number obtained per well. test substance as being either a complete antagonist or The optimal concentration reported in Table 1 is the merely a partial antagonist. It should be noted that these 55 smallest concentration at which a maximal effect is antagonistic characteristics are separate and distinct detected, be it agonistic or antagonistic. By comparing from any agonistic properties identified previously the empirically obtained results of Table 1 with the using the Phase 1 protocol. The existence of agonistic interpretive graphs of FIGS. 1 and 2 respectively, each and antagonistic properties in a single substance is not of the substances tested is classified as follows: estradiol unusual and is not in any way contradictory or evidence 60 is a full or complete agonist; 11 B-chloromethyl estradiol of artifact in the test methodology. To the contrary, the and Moxestrol both are full agonists; tamoxifen is both coexistence of properties which identify a single sub- a partial agonist and concurrently a partial antagonist; stance as being a partial agonist and simultaneously a LY 1 17018 is a partial agonist and a full antagonist; complete antagonist to estrogens is a phenomenon well LY 156758 is a partial agonist and a full antagonist; described in the art. 65 RU39411 has no agonistic activity but is a full antago To demonstrate the phenomena of coexisting agonis nist; and progesterone is not active either as an agonist tic and antagonistic properties and to illustrate the accu nor an antagonist and for this reason is said to be quies racy and reproducibility of both the Phase 1 and Phase cent. It will be recognized and appreciated that the 4,859,585 17 18 characteristics identified for each of the substances medium used in the test protocols is required to contain under test in Table 1 correspond to the reports in the only two essential components; a fluid carrier, prefera published literature regarding their pharmacological bly DME, containing a requisite amount of nutrients, characteristics with respect to naturally occurring es vitamins, salts, and the like to maintain the cell culture trogens. The ability of the present invention to accu in a healthy, viable state; and a measurable concentra rately identify and distinguish between estrogen ago tion of an inhibitor endogenous to the serum of adult nists and estrogen antagonists in a wide variety of sub male and female humans which is present in the fluid at stances is thus deemed to be unequivocally demon a concentration effect to prevent the proliferation of the strated. cells in-vitro. Both essential requirements are satisfied As a practical matter, it should be recognized that the 10 by the use of CDHS in more than 5% final concentra empirical results presented in Table 1 are based on the tion by volume. Alternatively, the maintaining medium use of 20% CDHS prepared as described earlier herein. may comprise any of the well known natural or syn While it is theoretically possible to use 100% serum thetic preparations now available for the maintenance taken from humans which has been adsorbed with char of cells in culture in combination with a predetermined coal-dextran, this is deemed impractical because such 15 quantity of concentrated, semi-purifed, proteinaceous concentrated serum must be dialysed to equilibrium inhibitor which has been isolated from the plasma of against the fluid comprising the maintaining medium adult humans. (preferably DME) to provide the nutrients and vitamins necessary for cell survival and multiplication during the Protocol For Isolation Of Purified Inhibitor Able To incubation period of the test protocol. The 20% concen 20 Prevent Proliferation Of Estrogen Dependent Cells tration of serum is preferred because this concentration In-Vitro provides consistent results and is much more economi 1. Source: It is preferred that outdated frozen plasma cal than higher concentrations. Other CDHS concen from adult humans be used. The plasma is not to be kept trations up to 40% provide comparable or better results at 4 C. for more than 24 hours from the time it was depending upon the sensitivity of the cells in culture to 25 drawn until the moment it is utilized in the purification the endogenous inhibitor in the serum; accordingly, the protocol. control values for those cells unable to proliferate are This precaution is taken to avoid degradation of the usually lower in number when higher concentrations of inhibitor by the serine proteases present in plasma. CDHS are used, but this is not a significant factor so Serum may be employed in lieu of plasma if desired; long as consistency of CDHS concentration chosen is 30 however, outdated plasma is preferred because of its maintained throughout the test methodology as a lesser cost. whole. 2. Removal of coagulation factors: To the plasma sample maintained at 4 C. estradiol is added at a con Qualitative Assay centration of 100 picograms (1012 grams) per ml of The phase I and II assays may be read by visual com 35 plasma. To this mixture 1.0M sodium citrate solution is parison of the cell density achieved by similar inocula added until a total volume concentration of 15 mM is grown with different test compositions. The assay pro obtained. To this total volume 1.0M benzamidine is tocol follows exactly the descriptions of the quantita added until a final concentration of 5.0 mM is achieved. tive assays already described. The only modification is Subsequently, a 1.0 M BaCl2 solution is added very that by the end of the incubation period, the cells are slowly in sufficient quantity to achieve a final volume not harvested. Instead, the cells are fixed and stained "in concentration of 80 mM in the fluid mixture. With the situ' as described below. addition of BaCl2 solution, a precipitate is formed which 1. The medium in each well is aspirated; each well is is stirred very gently to form a suspension. The reaction carefully rinsed with phosphate-buffered saline solution mixture is then placed in a container having an air-tight (50 mM. Na phosphate buffer, pH 7.0, 150 mM. NaCl or 45 closure and positioned on a roller apparatus (Bellco Co., comparable isotonic solution). Vineland, N.J.), and rotated for approximately 60 min 2. The isotonic solution is removed by aspiration, and utes at 4 C. Subsequently, the reaction mixture is cen 1 ml 10% formalin in phosphate-buffered saline (or trifuged at 3000Xgravity for 10 minutes to form a su comparable isotonic buffer) is placed in each well. This pernatant and a precipitant layer. After separating the solution will fix the cells in a 3 minute incubation per 50 fractions, the precipitant is discarded and the superna iod. After fixation, the formalin solution is removed by tant fraction combined with ammonium sulfate. aspiration. 3. Ammonium sulfate precipitation: Initially, the su 3. 1 ml of 0.01 N sodium hydroxide is placed in each pernatant is brought to a 55% saturation concentration well, and aspirated after 30 seconds-1 minute interval. of ammonium sulfate. This is achieved by adding 0.351 4. 0.5 ml of a 0.5% crystal violet aqueous solution is 55 grams of crystaline ammonium sulfate to each milliliter pipetted into each well, and removed by aspiration after of supernatant fluid. The ammonium sulfate crystals are 1 minute. first prepared as regular, pulverized particles using a 5. The plates are washed with tap water until the mortar and pestle. These small crystals are added background is clear. slowly to the supernatant at 4 C. with gentle stirring of 6. The plates are dried by placing them upside down 60 the mixture. Precipitation is allowed to proceed for 30 on an adsorbent mat. minutes duration while the fluid is held at 4 C. The 7. The results may be recorded by xerox copy as precipitated material is then isolated by centrifugation shown in FIG. 6. at 10,000Xgravity for 15 minutes. The supernatant fluid is retained while the pellet of precipitated material is Alternatives To The Use Of CDHS 65 discarded. Functional equivalents of CDHS in the maintaining Subsequently, the supernatant fluid is brought up to a medium may be employed in either Phase 1 and/or 70% saturation of ammonium sulfate. The supernatant Phase 2 protocols freely as desired. The maintaining fluid is then combined with an additional 0.01 grams of 4,859,585 19 20 ammonium sulfate per milliliter of fluid with gentle bly as duplicate samples) for Procedure II; and 1.0 ml mixing while the crystalline material is being added. per well of different concentrations of the precipitated This results in a second precipitant being formed which inhibitor preparation in 5% CDFBS for Procedure III. is isolated by centrifugation of the mixture at All cell cultures are then harvested and counted on the 10,000x gravity for 15 minutes duration. The isolated 9th day of the test period in accordance with the meth precipitant contains the active inhibitor which is iso ods described for the Phase 1 and 2 protocols. lated by resuspending the precipitant material in 50 um From the empirically obtained cell count data, a de Tris buffer pH 7.0 containing 5.0 mM benzamidine. termination of how many times the cells in each in Alternatively, the precipitated pellet material may be nocula have doubled during the nine calendar days in stored indefinitely at -70' C. in 70% saturated ammo 10 culture is made with respect to: the different concentra nium sulfate solution. tions of serum (Procedure 2); or to the different concen When the ammonium sulfate precipitated inhibitor is trations of precipitated inhibitor preparation (Proce to be utilized, it is necessary to remove the excess salt by dure 3). This determination is made as follows: dialysis against a buffer suitable for use in tissue culture, (i) the number of cells in the innocula at the beginning such as 25 mM Hepes buffer containing 100 mM. NaCl, 15 of the experiment are obtained from Procedure 1 (the pH 7.4. The preferred ratio of inhibitor to dialysate fluid cell number on day 1). should be in the order of 1:50 by volume. In the pre (ii) the number of cells in existence on day 9 of the test ferred method, the dialysis step is performed with two are taken from the series of samples comprising Proce changes of Hepes-NaCl buffer. dures II and III respectively. The maximal number of This method of preparation results in a 3-5 fold puri 20 cells is taken from the count of Procedure I, day 9. fication with a 60-70% yield of inhibitor protein. It (iii) the number of doublings for the cell innocula is should be noted that the inhibitor protein may degrade over time; accordingly, the inhibitor is preferably kept calculated by the term: in a 70% ammonium sulfate saturated solution contain 3.3X (log cell no. at day 9-log cell no. at day 1) ing 5 mM benzamidine until the time of use. 25 As may be seen in FIG. 4, the concentration of inhibi Assay To Determine The Potency Of The tor protein at which the number of cell doublings Concentrated Inhibitor achieved is half maximal represents 1 unit of inhibitory The potency of the inhibitor isolated from human activity. For CDHS, this value is attained at 15-20% serum is defined in activity units. An activity unit is 30 concentration; accordingly there are 5-6 units of inhibi defined as that amount of purified inhibitor protein tor per ml of undiluted serum, or 0.07 units per mg required in 1.0 ml of medium to increase the doubling protein. The preferred concentration for use in the time of a 10 cell innoculum (the time required for a cell Phase 1 and Phase 2 protocols is thus 20% serum-the concentration to double in quantity) by a factor of 2 at equivalent of 1 unit per ml of inhibitor in the maintain the end of a 9 calendar day period. To determine the 35 ing medium. Mathematically this is stated as: potency of any preparation comprising purified inhibi tor, the following measurements should be performed, 3.3 (log cell no. at day 9 - log cell no. at day I) preferably in duplicate using simultaneously seeded unit = 2 wells: Procedure I: Determination of the optimal prolifera (iv) A comparison of the total cell number at each tion rate for a 10 cell innocula in 20% CDHS in combi inhibitor concentration, in the presence (b) series and nation with 3x10-11M E2 (for CMCF7-173 cells the absence (a) series of estradiol may now be made. Such expected optimal rate is 36-37 hours). a comparison is graphically illustrated by FIG. 5. The Procedure II: Determination of the standard dose graphic comparison establishes whether or not the pre response curve to different concentrations of CDHS in 45 cipitated preparation contains an active inhibitor which combination with varying concentrations of estradiol is able to be neutralized by estradiol; and whether or not such as (a) no estradiol and (b) 3X 10-11M estradiol. spurious toxic materials are present in the preparation. Procedure III: Determination of the dose-response In the latter instance, the action of the inhibitor is not curve to different concentrations of (NH4)2SO4 precipi reversed by endogenous estrogens. tated inhibitor preparations which have been added to 50 It is specifically noted that the quantitative difference 5% charcoal-dextran stripped fetal bovine serum, here between the cell count data obtained in the presence inafter "CDFBS' (containing no detectable quantity of and absence of estradiol at set concentrations of inhibi endogenous inhibitor) in combination with varying tor above the equivalent effect of 5% CDFBS should be concentrations of estradiol such as (a) no estradiol and between 5 and 10 fold; i.e., those cells without any (b) 3X 10-11M estradiol. 55 estradiol should achieve no more than one half the To perform Procedure I, 10 cell innocula preferably number of doublings achieved by those innocula receiv are seeded in 5% FBS into each well as described in the ing estradiol. Phase 1 and 2 assay protocols. 24 hours later, the me dium is changed to 1.0 ml per well of 20% CDHS con Final Considerations taining 3X 10.1M estradiol. Triplicate cultures of cells 60 It will be recognized that a major advantange of the are harvested on calendar days 1, 3, 5, 7, and 9 of the described in-vitro methods is the ability to identify a test period. The numerical cell count data are plotted substance as having agonistic or antagonistic properties using a log-linear scale as shown in FIG. 3. From the in comparison to endogenous estrogens. Based on this graphic data plot, the doubling time is extrapolated. ability, the protocols can serve as an initial screening To perform Procedures II and III, innocula compris 65 determination to characterize presently available or ing 10 cells in 5% FBS are seeded into a series of wells. newly synthesized compositions for their primary or 24 hours later, the medium is changed: 1.0 ml per well secondary side effects. For example, the present inven of 5%, 10%, 20%, or 40% CDHS respectively (prefera tion will evaluate new compounds whose intended use 4,859,585 21 22 are as pesticides; as intermediate or final compositions cells to measurably increase in number identifying suspected of having estrogen-like effects due to their the substance of interest as an estrogen antagonist. manufacturing process; as therapeutic drugs which 3. An in-vitro method for identifying a substance of might have undesirable side effects in addition to their interest as having in-vivo estrogen agonistic and estro primary pharmacological effect; and as antitumor (anti 5 gen antagonistic properties, said method comprising the cancer) agents which are suspected of having unwanted steps of: agonistic properties in addition to the desired antagonis obtaining a plurality of cells cultured in-vitro, said tic properties with respect to endogenous estrogens. As cells being estrogen dependent for proliferation an example of the last instance, the use of the drug in-vivo; 10 maintaining a known quantity of said cultured cells in tamoxifen for the clinical treatment of estrogen sensitive a medium without estrogens comprising a fluid and breast cancers in women is dependent upon this drug an inhibitor of cell proliferation endogenous to the behaving as a partial agonist and as a partial antagonist serun of adult humans, said endogenous inhibitor with respect to the estrogens in the body. Ideally, the being present in said maintaining medium at a con most powerful anticancer compositions against estro 15 centration effective to prevent proliferation of said gen-sensitive tumors should demonstrate not only total cells in-vitro in the absence of an estrogen; inactivity as an agonist but also full antagonistic activity adding the substance of interest to a first aliquot of against estrogens. Using the present methodology, a said cultured cells in said maintaining medium to wide variety of compositions could be initially identi form a first reaction mixture; fied and evaluated prior to being utilized in clinical field 20 adding the substance of interest and an estrogen to a trials involving human subjects. The invention is not to second aliquot of said cultured cells in said main be restricted in form nor limited in scope except by the taining medium to form a second reaction mixture; claims appended hereto. incubating said first reaction mixture and said second What we claim is: reaction mixture for a preselected time period; 1. An in-vitro method for identifying a substance of 25 determining the number of cultured cells in said incu interest as an in-vivo estrogen agonist, said method bated first reaction mixture, a measurable increase comprising the steps of: in the number of cultured cells identifying the sub obtaining a plurality of cells cultured in-vitro, said stance of interest as an estrogen agonist; and cells being estrogen dependent for proliferation determining the number of cultured cells in said incu in-vivo; 30 bated second reaction mixture, the failure of said maintaining a known quantity of said cultured cells in cultured cells to measurably increase in number a medium without estrogens comprising a fluid and identifying the substance of interest as an estrogen an inhibitor of cell proliferation endogenous to the antagonist. serum of adult humans, said endogenous inhibitor 4. The in-vitro method as recited in claim 1, 2, or 3 being present in said maintaining medium at a con 35 wherein said cultured cells comprise tumor cells de centration effective to prevent proliferation of said rived from the female human breast. cells in-vitro in the absence of an estrogen; 5. The in-vitro method as recited in claim 4 wherein adding the substance of interest to said cultured cells said cultured cells are C7 MCF7-173 cells, having in said maintaining medium to form a reaction mix ATCC accession number CRL 9836. ture; 6. The in-vitro method as recited in claim 2 or 3 incubating said reaction mixture for a preselected wherein said estrogen is selected from the group con time period; and sisting of estrone, estradiol, and estriol. determining the number of cultured cells in said incu 7. The in-vitro method as recited in claim 1, 2, or 3 bated reaction mixture, a measurable increase in the wherein said maintaining medium comprises fetal bo number of cultured cells identifying the substance 45 vine serum. of interest as an estrogen agonist. 8. The in-vitro method as recited in claim 1, 2, or 3 2. An in-vitro method for identifying a substance of wherein said maintaining medium comprises Dulbec interest as an in-vivo estrogen antagonist, said method co's modification of Eagle's Minimal Essential Medium. comprising the steps of: 9. The in-vitro method as recited in claim 1, 2, or 3 50 wherein said maintaining medium comprises charcoal obtaining a plurality of cells cultured in-vitro, said dextran stripped human serum. cells being estrogen dependent for proliferation 10. The in-vitro method as recited in claim 9 wherein in-vivo; said serum is used in a concentration ranging from maintaining a known quantity of said cultured cells in 40-10% concentration by volume. a medium without estrogens comprising a fluid and 55 11. The in-vitro method as recited in claim 1, 2, or 3 an inhibitor of cell proliferation endogenous to the wherein said wherein said inhibitor is added to said serum of adult humans, said endogenous inhibitor maintaining medium as a incompletely proteinaceous being present in said maintaining medium at a con preparation. centration effective to prevent proliferation of said 12. The in-vitro method as recited in claim 1, 2, or 3 cells in-vitro in the absence of an estrogen; 60 wherein said reaction mixtures are incubated for a time adding the substance of interest and an estrogen to period ranging from 9-14 calendar days. said cultured cells in said maintaining medium to 13. The in-vitro method as recited in claim 1, 2, or 3 form a reaction mixture; wherein said fluid of said maintaining medium com incubating said reaction mixture for a preselected prises at least one selected from the group consisting of time period; and 65 buffers, amino acids, sugars, vitamins, and ionizable determining the number of cultured cells in said incu salts. bated reaction mixture, the failure of said cultured s e s s