Monoclonal antibodies to estrophilin: Probes for the study of estrogen ...

Proc. Natl. Acad. Sci. USA

Vol. 77, No. 1, pp. 157-161, January 1980 Biochemistry

Monoclonal antibodies to estrophilin: Probes for the study of estrogen receptors (cell fusion/hybridomas/antiestrophilin antibody/antibody cloning)

GEOFFREY L. GREENEt, FRANK W. FITCHf, AND ELWOOD V. JENSENt tBen May Laboratory for Cancer Research and fCommittee on Immunology and Department of Pathology, The University of Chicago, Chicago, Illinois 60637 Contributed by Elwood V. Jensen, October 1, 1979

ABSTRACT Splenic lymphocytes from a Lewis rat, immunized with purified estradiol-receptor complex of calf uterine nuclei, were fused with cells of three different mouse myeloma lines (P3-X63-Ag8, P3-NSI/1-Ag4-1, and Sp2/0-Ag4) to yield hybridoma cultures, 9% of which produced antibodies to the receptor protein (estrophilin) When cloned by limiting dilution, approximately 70% of the viable cultures secreted antiestrophilin antibody. When expanded in suspension culture, three clones derived from Sp2/0-Agl4 were found to secrete rat IgG (72a class), whereas seven other clones (from all three myeloma lines) secreted IgM. Monoclonal IgG shows comparable affinity for nuclear and extranuclear receptors, whereas IgM reacts preferentially with the nuclear form. Both classes of antibody react with unoccupied as well as with occupied receptor and do not interfere with its ability to bind to estradiol. By growing IgG-secreting clones in the presence of [a5Simethionine, radiolabeled monoclonal antiestrophilin has been prepared. Unlike antiestrophilin antibody previously generated in the rabbit or the goat, which crossreacts with estrogen receptors from every animal species tested, antibodies produced by the Lewis rat and by hybridomas derived from its spleen cells react specifically with estrophilin from calf tissues. These monoclonal antibodies provide reagents for the application of immunochemical techniques to study estrogen receptors in calf target tissues. During the past two years, we have reported (1-3) the successful immunization of rabbits and a goat with highly purified [3H]estradiol-receptor complex (E*R) isolated from calf uterine nuclei. Antiestrophilin antibodies (i-Ig) obtained from the sera of these animals were found to crossreact with both the nuclear and the cytosol forms of estrogen receptors from hormonedependent tissues and tumors of a wide variety of mammalian species, but no reaction was observed with receptors for other classes of steroid hormones. Recently we (4), as well as others (5), have found that rabbit antibodies to calf estrophilin crossreact with estrogen receptor from hen oviduct, establishing an immunochemical similarity among mammalian and nonmammalian estrophilins. The hormone specificity and crossreactivity of rabbit and goat antiestrophilins make them attractive as probes for examining the structure and function of estrogen receptors and as reagents for their radioimmunoassay, but such studies are limited by the heterogeneity of these antibody preparations. To obtain antibodies uncontaminated with other immunoglobulins, we have utilized the techniques of Kohler and Milstein (6), as modified by McKearn et al. (7), to obtain monoclonal hybridoma cell lines secreting specific antibodies to estrophilin. § This paper describes the polyethylene glycol-mediated fusion of splenic lymphocytes from a male Lewis rat, immunized with purified E*R from calf uterine nuclei, with three different mouse myeloma cells lines and the cloning of the resulting hybridoma cells to yield cultures secreting monoclonal anThe publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "ad-

vertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact.

tiestrophilin antibody (either IgG or IgM) that reacts specifically with estrogen receptors from calf tissues. MATERIALS AND METHODS Reagents. [6,7-3H]Estradiol-17f3 (57 Ci/mmol; 1 Ci = 3.7 X 1010 becquerels), [2,4,6,7-3H]estradiol-17,B (108 Ci/mmol, used for the human breast cancer experiments), and [f5S]methionine (600 Ci/mmol) were obtained from New England Nuclear. Phosphate-buffered saline contained 150 mM sodium chloride in 10 mM sodium phosphate, pH 7.9. All buffers contained 0.02% (wt/vol) sodium azide. Polyethylene glycol 1500 was obtained from Fisher Scientific and fetal calf serum was from Microbiological Associates, Walkerville, MD. Methionine-free medium 199 and Dulbecco's modified Eagle's medium (H-21) were obtained from GIBCO, and the latter was supplemented with amino acids (8). Selective growth medium (HAT) contained hypoxanthine, (0.1 mM), amethopterin (0.4 ,M), and thymidine (3 ,uM) in Dulbecco's modified medium with 20% fetal calf serum (9). Inbred Lewis rats were purchased from Microbiological Associates and used for immunizations and as a source of thymocytes, control serum,_ and rat IgG. Antiserum to Lewis rat IgG was obtained by intramuscular immunization of a goat with a highly purified preparation of this immunoglobulin. Mutant myeloma cell lines P3-X63-Ag8 (P3), P3-NSI/1-Ag4-1 (NSI), and Sp2/0-Agl4 (Sp2/0) were kindly provided by C. Milstein, G. Kohler, and R. Kennett. Immunization. The E*R used for immunization was isolated from calf uterine nuclei and purified as described (1, 2). Immunizations were carried out with preparations of nuclear E*R containing 12-20% of the radioactivity expected for a pure steroid-receptor complex consisting of one [3H]estradiol per protein molecule of molecular weight 68,000. The amount of estrophilin administered was estimated on the basis of the radioactivity injected. A 2-month old male Lewis rat was immunized at monthly intervals with a series of three intraperitoneal injections of E*R (36 ,ug each) and finally with an intradermal injection (65 ,tg) at multiple sites on the back. As described (1, 10), Freund's complete adjuvant was used for the primary immunization and incomplete adjuvant was used for booster injections. Three days prior to removal of the spleen for fusion with myeloma cells, the rat was given an intravenous injection of E*R (110,ug) in 1 ml of saline. A partially purified immunoglobulin fraction was prepared from immune serum by two precipitations from ammonium sulfate, 40% of saturaAbbreviations: E*, [3H]estradiol; E*R, [3H]estradiol-receptor complex; i-Ig, immunoglobulin from serum of immunized rat; n-Ig, immunoglobulin from serum of nonimmunized rat; i-IgG and i-IgM, antiestrophilin antibodies from hybridoma culture media; HAT, hypoxanthine/amethopterin/thymidine; P3, P3-X63-Ag8; NSI, P3-NSI/ 1-Ag4-1; Sp2/0, Sp2/0-Agl4. § Greene, G. L., Jensen, E. V. & Fitch, F. W. (1979) Sixty-First Meeting of The Endocrine Society, Anaheim, CA, June 13-15, p. 19

(abstr.).

157

158

Biochemistry: Greene et al.

tion (11). Control immunoglobulin was obtained similarly from nonimmunized rats. Cell Fusion. Spleen cells were taken from the immunized rat 3 days after the intravenous injection of antigen and fused with myeloma cells according to published procedures (7). Spleen cells were suspended in Dulbecco's modified medium by using a loose-fitting glass tissue homogenizer; viable cells were then separated from erythrocytes, dead cells, and debris by Ficoll-Hypaque gradient centrifugation (12). For each fusion, 5 X 107 splenic lymphoid cells and 5 X 106 myeloma cells (P3, NSI, or Sp2/0) were mixed in a 60-mm petri dish, and the dishes were centrifuged (250 X g, 3 min) to produce an adherent monolayer of cells. After aspiration of medium, cell fusion was effected by flooding the dishes with 1 ml of 50% polyethylene glycol 1500 in Dulbecco's modified medium. After 30 sec the cells were washed with two 5-ml portions of the medium; the dishes were then flooded with 5 ml of medium containing 20% fetal calf serum and antibiotics (penicillin plus streptomycin) and incubated overnight at 370C. The cells were removed by gentle scraping with a sterile rubber syringe plunger and pelleted by centrifugation (50 X g, 5 min). After dispersing each pellet in 30 ml of HAT medium containing antibiotics, 100-Ml aliquots of each cell suspension were pipetted into three 96-well Costar plates (no. 3596) and incubated at 370C in a humidified 5% CO2 atmosphere. Wells were fed 6-7 days later with an additional 100 Al of Dulbecco's modified medium containing 20% fetal calf serum. Wells that contained visible cell clusters (10-21 days after fusion) were assayed for antiestrophilin antibody as they became acidic by using the double-antibody precipitation technique described below. Cloning and Expansion of Hybridomas. Monoclonal hybridoma cell lines were obtained by limiting dilution. Antiestrophilin antibody-secreting hybridomas were diluted with Dulbecco's modified medium/20% fetal calf serum, and aliquots containing approximately one cell per 100 ,l of medium were pipetted into 96-well Costar plates previously seeded with 106 irradiated (1200 rad; 1 rad = 0.01 gray) Lewis rat thymocytes per well. Wells containing single clusters of hybridomas (10-21 days after limiting dilution) were assayed for antiestrophilin antibody as described below. Cloned cells were then expanded in suspension culture, ultimately without the aid of thymocytes. Cell lines were stored for future use by freezing in liquid nitrogen at 1-10 X 106 cells per ml in Dulbecco's modified medium/20% fetal calf serum/10% dimethyl sulf oxide. Purification of Monoclonal Ig. Cells were removed from suspension culture by centrifugation for 10 min at 1500 X g. Crude immunoglobulin fractions were prepared from the clarified media by a sequence of two precipitations from ammonium sulfate (40% of saturation). Further purification of Ig was achieved either by chromatography on DEAE-cellulose for IgG (11) or by filtration through Bio-Gel A-1.5 M agarose for IgM. The class and subclass of the monoclonal immunoglobulin produced were determined by Ouchterlony analysis, using antisera specific for ,u chains and for IgG subclasses, kindly provided by H. Bazin (13). Hormone-Receptor Complexes. Cytosol and nuclear E*Rs from calf and rat uterus, as well as cytosol E*R from human breast cancer, were prepared as described (1). Nuclear extract from MCF-7 human breast cancer cells, incubated 1 hr in 10 nM E*, was a gift from Chris Nolan of Abbott Laboratories. Sedimentation Studies. Various E*Rs (0.2-1.0 pmol in 100-200 Al of 10 mM Tris*HCl cytosol or, 10 mM Tris-HCl, pH 7.4/400 mM KCl nuclear extract) were incubated at 40C for 1 hr either with Ig from serutm of nonimmunized rat (n-1g) or with i-Ig from the immunized rat (200 Ag) or from cloned hy-

Proc. Natl. Acad. Sci. USA 77 (1980)

bridomas (10-200 pAg) in a final volume of 220-250 Mtl of 10 mM Tris1HCl (pH 7.4) or 10 mM Tris-HCl, pH 7.4/400 mM KC1. Each mixture was then layered on 3.5 ml of a 10-30% or 1050% sucrose gradient containing either 10 mM Tris-HCl, pH

7.4/10 mM KCI/1.5 mM EDTA (low salt) or 10 mM Tris-HCl, pH 7.4/400 mM KCl/1.5 mM EDTA (high salt) and centrifuged at 0C for 15 hr at 253,000 X g. Successive 100-Mtl fractions were collected from the tube bottom, and radioactivity was measured in Triton X-100 scintillation mixture.

'4C-La-

beled ovalbumin (3.6 S), rabbit IgG (6.6 S), and f-amylase (9.2 S) were used as internal markers. For sedimentation studies involving 35S-labeled monoclonal antibody, the E*R was incubated with radiolabeled i-Ig for 1 hr at 2°C prior to gradient centrifugation. Separate 35S and 3H sedimentation profiles were generated by simultaneous two-channel counting in a Packard Tri-Carb scintillation counter. Immunoprecipitation Assays. Hybridoma-containing culture wells were assayed for antiestrophilin antibody by a double-antibody precipitation technique. Solutions containing crude nuclear E*R of calf uterus (0.1 pmol), normal rat serum (10 Ml), and hybridoma medium (50 ,ul) in a final volume of 1.0 ml of phosphate-buffered saline/10 mM EDTA were incubated for 4-5 hr at 4°C. E*R-Ig complexes were then precipitated by adding sufficient goat antiserum against Lewis rat IgG to precipitate all rat Ig. After the pellets were dissolved in 0.2 M NaOH and neutralized with 1 M HCl, radioactivity was measured in 10 ml of scintillation mixture. Nonspecific binding was determined in tubes containing E*R and n-Ig. The reliability of each screening assay was verified by including tubes containing sequential dilutions of Ig from the serum of the immunized Lewis rat. Preparation of [35S]Ig. [35S]Ig was prepared as described (7). Cloned hybridoma cells in exponential growth phase (1-5 X 106 cells) were pelleted by centrifugation, resuspended in 1-10 ml of medium 199 containing 50-500 MuCi of [35S]methionine, and incubated overnight at 370C. Fetal calf serum was then added (2% final concentration) and the suspension was centrifuged. Free methionine was removed from the supernatant fraction by filtration through a column of Sephadex G-25 equilibrated in 50 mM Tris-HCl, pH 7.4/2% fetal calf serum. Fractions containing incorporated 35S were pooled and further purified by precipitation from ammonium sulfate (40% of saturation) in the presence of rabbit IgG as a carrier. RESULTS

Antibodies to calf uterine estrophilin were obtained by immunizing a male Lewis rat with highly purified nuclear E*R from calf uterus. Significant antibody titer was observed 3 months after the primary injection. As described (1, 2), the interaction of these antibodies with calf estrophilin was detected and characterized both by sucrose density gradient centrifugation and by double-antibody precipitation techniques, using E* as a marker for the receptor complex. The interaction of Lewis rat i-Ig with the E*Rs of calf uterus caused an increase in their rates of sedimentation in salt-containing sucrose gradients from 4 S to 7 S in the case of the cytosol complex (Fig. la) and from 5 S to about 8 S with the nuclear complex (Fig. lb). Similar shifts were observed in the sedimentation of the unoccupied receptors when E* was used to postlabel the gradient fractions. No effect on the sedimentation of E*R was produced by the control immunoglobulin (n-Ig). In contrast to the antiestrophilin antibodies generated in the rabbit and the goat, which show crossreactivity with nuclear and extranuclear receptors of all species tested, the i-Ig from the Lewis rat showed no evidence of any interaction with cytosol or nuclear E*R from rat uterus, MCF-7 cancer cells, or human breast cancer.

Biochemistry:

Greene et al.


6.6i 3.6 I a

159

6.6 3.6

b b

a

E

E

. 2000

0. w

10

20

30 Top 10o 20 30 Top3 Fraction FIG. 1. Sedimentation profiles in sucrose/400 mM KClgradients of E*R from: calf uterine cytosol (10-30%o sucrose) (a) and salt extract of calf uterine nuclei (10-50%o sucrose) (b after incubation with E* in calf uterine cytosol in the presence of Lewis rat n-Ig (O) or i-Ig (-).

The results of the fusion and cloning experiments are given

in Table 1. Over 50% (485/912) of the original microtiter wells contained proliferating hybridomas, and approximately 9% (44/485 wells) of the derived hybrid lines secreted antiestro-

philin antibody as determined by double antibody precipitation with crude nuclear E*R from calf uterus as the labeled antigen. In each case, wells that were scored positive for i-Ig contained enough antibody in 50 ,ul of medium to precipitate at least twice the E*R precipitated in nonspecific control tubes or in negative wells. Each positive well was retested in a subsequent assay to verify the presence of i-Ig. Hybrid lines were successfully derived from all three mouse myeloma cell lines, although the P3 and NSI myelomas produced more viable hybrids than did the

Sp2/0 myeloma. Several of the hybrids, including lines derived from all three myeloma, mutants, were cloned by limiting dilution. About 15% of the total wells (79/510) produced viable clusters of hybridomas. Approximately 71% (56/79 wells) of the proliferating clones observed after 10-21 days of growth secreted antiestrophilin antibody (Table 1). To test the feasibility of growing antiestrophilin antibody-producing hybmMds on a large scale, clones denuved from all three myeloma cell lines were expanded in spinner culture to volumes of 350 ml or greater,

10

20

30 Top Fraction

10

20

30 Top

FIG. 2. Sedimentation profiles in sucrose/400 mM KCl gradients of E*R (0.6 pmol) from: calf uterine cytosol (10-30%o sucrose) (a) and calf uterine nuclear extract (10-50% sucrose) (b) in the presence of rat n-Ig (0), 20 lAg of clonal i-IgG (-), or 200 ,ug of clonal i-IgM (0). To saturate binding to cytosol receptor, the amount of i-IgM is 20-fold greater than that used in the experiment of Fig. 3.

producing milligram amounts of rat immunoglobulin. Viable antiestrophilin antibody-secreting hybridomas have been recovered after storage in liquid nitrogen for at least 6 months. The characteristics of several derived antiestrophilin antibody-secreting clones are summarized in Table 1. Three of these clones (Sp2/0 hybrids) were found to secrete rat IgG (,y2a class) that, like i-Ig, reacts with 4S cytosol E*R (Fig. 2a) and 5S nuclear E*R (Fig. 2b) of calf uterus to give 7S-8S immune complexes in salt-containing sucrose gradients. In low-salt gradients, the sedimentation of the 8S cytosol complex shifted to 10-11 S in the presence of monoclonal i-IgG (data not shown). Seven additional clones, derived from all three myeloma cell lines, were found to secrete rat IgM, which interacted with both cytosol E*R (Figs. 2a and 3a) and nuclear E*R (Figs. 2b and 3b) to form immune complexes sedimenting at 12S-13S in high-salt gradients. In salt-free gradients, 8S cytosol receptor also formed a 12S-13S complex in the presence of i-IgM. Monoclonal i-IgG and i-IgM resemble the crude immunoglobulin of the immunized Lewis rat in being species-specific for calf estrophilin; no interaction was observed with E*R from rat uterus, MCF-7 cancer cells, or human breast cancer. Antibodies isolated from the immune rat serum and from

Table 1. Monoclonal hybridoma lines obtained by fusion of spleen cells of immunized Lewis rat with cells of three different mouse myeloma lines Mouse myeloma used Hybridomas Ig Positive Positive Ig Reactivity Cell line chainst wellst clones§ classl with calf E*R P3 Heavy + 14/187 16/23 1 IgM Nuclear> cytosol light NSI Light 20/188 19/23 2 IgM Nuclear > cytosol Sp2/0 None 10/110 21/33 4 IgM Nuclear > cytosol 3 IgG2a Nuclear = cytosol t Immunoglobulin produced by myeloma cells themselves. I Based on viable cultures. Total wells were 366 for P3 and 288 each for NSI and Sp2/0. § Based on viable clones. Three P3, three NSI, and six Sp2/0 cultures were cloned. Total cloning wells were 144 for P3 and NSI and 288 for Sp2/0. So far, the type of immunoglobulin produced has been characterized for 10 of the 56 hybridoma clones found to make antiestrophilin antibody.

160

Biochemistry: Greene et al.


9.2

9.2

;?^

4000

a

D

4000

IgG 5 S

E

E 2000

C: w

In

a

09

_° X 10

20

l

2000 E

10 20 30 Top

30 Top

10

20

30 Top

Fraction 30 Top Fraction

Fi(.. .3. Comparative reactivity of monoclonal IgM with cytosol and nuclear forms of calf uterine estrophilin. Sedimentation was in 10-50%, sucrose/400 mM KCI gradients of E*R (0.5 pmol) from: calf uterine cytosol (a) and calf uterine nuclear extract (b) in the presence of rat n-Ig (0) or 10 ,ug of clonal i-IgM (0).

IgG-secreting clones appear to have similar avidity for nuclear and cytosol forms of calf uterine E*R, but monoclonal i-IgM reacts preferentially with the nuclear form. Under the same conditions (10,g of i-IgM and 0.5 pmol of E*R) in which i-IgM reacts completely with the nuclear complex (Fig. 3b), only 15-20% of the cytosol complex was bound (Fig. 3a). Only if its concentration was markedly increased (200 ,g) did i-IgM react completely with cytosol E*R (Fig. 2a). Antibody titers for partially purified preparations of monoclonal i-IgG and i-IgM, determined against nuclear E*R from calf uterus, were considerably higher than those in either the rat antiserum or the medium taken from overgrown suspension cultures of cloned hybridomas (Fig. 4). Titration curves for unpurified monoclonal Ig were variable and shifted to the left of the antiserum curve, as shown for an IgM-secreting clone, indicating that antibody concentrations in the hybridoma culture media were lower than in rat antiserum obtained 1 week after splenectomy. The affinities of rat serum antibodies and monoclonal i-IgG and i-IgM for calf uterine nuclear E*R were determined by saturation analysis with the double-antibody precipitation technique as described (2). In each case the Kd value, as determined by Scatchard plot, is approximately 10-10 M. These monoclonal antibodies do not interfere with the binding of E* to the Ig-receptor complex, as determined by

-"

1001

50\

(1/dilution) 1g, g

X

10-

1 20

10

100

1000

2

0.2

0.02

Fi(u. 4. Titration curves for rat antiserum (A), culture medium from an IgM-secreting clone (o), purified monoclonal i-IgG (@), and purified monoclonal i-IgM (0) against crude calf uterine nuclear E*R (0.1 pmol). B, specifically bound E*R; B(, bound E*R at lowest dilution of antibody. For purified i-Ig preparations, 1/dilution = dilution of a solution containing 10 mg of i-Ig per ml, which corresponds to the approximate IgG concentration in rat serum; 200 ,l of i-Ig solution was added to each incubation tube (total volume = 1.0 ml).

Fl(;. 5. Interaction between monoclonal [:35SJIgG and excess calf nuclear E*R. Sedimentation profiles in salt-containing, 10-30% sucrose gradients of nuclear calf uterine E*R (0.8 pmol) (0) and a limiting amount of monoclonal [35S]IgG (0): in separate tubes (a) or after incubation together for 1 hr at 40C (b). The arrows in b indicate the positions of E*R (5 S) and 7S [35SJIgG (IgG), corresponding to the peaks in a.

postlabeling (with E*) fractions from sucrose gradients containing unoccupied receptor and either i-IgG or i-IgM. By inclusion of [a5S]methionine in the culture medium of an IgG-secreting hybridoma clone, we have obtained radiolabeled antiestrophilin antibody, recognized by sedimentation of the isotope at about 7 S (Fig. Sa). When treated with an excess of nuclear E*R, the sedimentation peak of the [35S]IgG was completely shifted to the 8S-9S region, along with that portion of the E*R that reacted (Fig. Sb). DISCUSSION In the application of immunochemical techniques for the study of estrogen receptors (1), the availability of antiestrophilin antibody uncontaminated by other antibodies is advantageous or, for some purposes, obligatory. The experiments described here demonstrate the feasibility of preparing monoclonal lines of hybridoma cells, capable of producing antiestrophilin antibody in substantial amounts, by the fusion of rat splenic lymphocytes with various mouse myeloma lines. Of the mutant mouse myelomas employed, the P3 line synthesizes and secretes a complete myeloma immunoglobulin (IgG), NSI synthesizes only a light chain and therefore does not secrete any Ig (14), and Sp2/0 does not synthesize any immunoglobulin chains (15). Thus, the Sp2/0 myeloma is particularly favorable for the preparation of uncontaminated antibody, because the resulting hybrid cells synthesize and secrete only the immunoglobulin produced by lymphocytes from the immunized rat. However hybridomas from P3 and NSI lines, which may secrete mouse or mixed mouse-rat immunoglobulins, are still practical sources of monospecific antiestrophilin antibody, because immunoglobulin with myeloma components can be removed by immunoadsorption with antibody to myeloma protein. Hybrid cell lines obtained from all three myelomas appear to be stable and can be expanded in suspension cultures. Antiestrophilin antibody-secreting hybridomas can readily be recovered after storage in liquid nitrogen, providing the opportunity for generating additional antibody when needed. As in previous experiments with rabbit and goat antiestrophilin antibodies (1-3), E* can serve as a marker for the receptor, either before or after its association with rat antibody. The interaction of E*R with rat i-Ig or with monoclonal antibodies does not cause the release of significant amounts of hormone (Figs. 1-3), nor does the interaction of unoccupied receptor with monoclonal IgG or IgM prevent or decrease the binding of E* to the receptor in the immune complex. The relative titers of two purified monoclonal Ig preparations (Fig. 4), although much higher than the crude hybridoma

Biochemistry:

Greene et al.

culture media or rat antiserum, were lower than might be expected for pure preparations of antibody in which all molecules are capable of binding receptor. For the i-IgM used in this experiment, this result is not surprising, because it was obtained from a clone of NSI hybrid cells that may also secrete inactive mouse-rat hybrid immunoglobulin molecules. However, the monoclonal IgG antibody was obtained from Sp2/0 hybridomas that do not synthesize myeloma immunoglobulin, so only rat IgG should be secreted. The possibility that more than one clone is present in the IgG-secreting cell line has been minimized by repeated cloning of these cells, as well as by the fact that all the immunoglobulin secreted by cells grown in the presence of [35S]methionine is capable of reacting with estrophilin, or at least with some component of calf uterine nuclei (Fig. 5). Genetic drift is also unlikely because antiestrophilin antibody can be detected in 100% of the viable recloned cells, even when the limiting dilution is done several weeks after the previous cloning. Although these rat antibodies do not discriminate between nuclear and cytosol forms of calf estrophilin in terms of the shift in sedimentation rate, the monoclonal i-IgM shows greatly reduced affinity for cytosol receptor compared to the nuclear form (Fig. 3). The basis of this selectivity is not clear, but it is likely that such differences in immunochemical reactivity can be exploited to gain insight concerning structural or conformational changes that accompany receptor activation, such as dimerization of the native receptor molecule (16, 17) or formation of a complex between receptor and another cytosol component (18). The fact that neither the immunoglobulin obtained from the serum of the immunized Lewis rat nor the monoclonal i-IgG and i-IgM antibodies secreted by the hybridoma cells appears to crossreact with estrophilin from species other than the calf indicates that they bind to antigenic determinant(s) specific for calf estrophilin and not to the common determinant(s) recognized by the crossreacting antibodies generated in the rabbit or the goat (1-3). From the magnitude of the effect on the sedimentation properties of the receptor, it would appear that only one molecule of rat antibody binds to each molecule of either the nuclear or the cytosol form of the receptor. This behavior differs from that of the rabbit antibody, in which case one molecule of immunoglobulin binds to each molecule of cytosol receptor but two or more antibody molecules bind to the nuclear receptor (1). The formation of a 12S-13S immune complex between monoclonal IgM and either cytosol or nuclear E*R (Figs. 2 and 3) is unexpected, because rat IgM itself is reported to sediment at 18 S-19 S (19). However, the i-IgM produced by the hybridoma cells was found to sediment in the 12S-13S region in the same gradient system. Although the IgM molecule should be multivalent, binding several antigen molecules when antibody is present in excess, the gradient data suggest that a 1:1 immune complex is obtained with estrophilin. Because the antibodies to calf estrophilin produced in the Lewis rat do not crossreact with estrogen receptors from other species, the monoclonal antiestrophilin antibody so far obtained can be utilized only for studies of receptors in calf tissues. As specific probes for the receptor protein, independent of its


161

binding to radioactive estrogen, these antibodies are proving useful in the immunocytochemical localization of estrophilin in calf uterine cells and cell dispersions as well as for the purification of calf uterine receptor by immunoadsorption. The monoclonal antibody labeled biosynthetically with [a5S]methionine provides an especially valuable reagent for the study of receptor unoccupied by steroid hormone.

Recently we have found that antiserum from ACI rats immunized with purified calf uterine nuclear E*R does crossreact with estrophilin of human breast cancers and MCF-7 cancer cells. Fusion of splenic lymphocytes from the immunized ACI rat with mouse myeloma cells should provide lines of hybridomas secreting crossreacting monoclonal antiestrophilin antibody, which would be applicable for studying estrogen receptors in target cells of different animal species. Valuable assistance in this investigation was provided by Peter Engler, Nancy Sobel, and Sophia Mirviss. These studies were supported by research grants from the American Cancer Society (BC-86) and Abbott Laboratories, by a research grant (CA-02897) and contract (CB-43969) from the National Cancer Institute, and by the Women's Board of the University of Chicago Cancer Research Foundation and the Helen Adams Selfridge Fund. 1. Greene, G. L., Closs, L. E., Fleming, H., DeSombre, E. R. & Jensen, E. V. (1977) Proc. Nati. Acad. Sci. USA 74, 36813685. 2. Greene, G. L., Closs, L. E., DeSombre, E. R. & Jensen, E. V. (1979) J. Steroid Biochem. 11, 333-341. 3. Jensen, E. V., Greene, G. L., Closs, L. E. & DeSombre, E. R. (1979) in Steroid Hormone Receptor Systems, eds. Leavitt, W. W. & Clark, J. H. (Plenum, New York), pp. 1-16. 4. Greene, G. L., Closs, L. E., DeSombre, E. R. & Jensen, E. V. (1979) J. Steroid Biochem. 11, 1807-1814. 5. Radanyi, C., Redeuilh, G., Eigenmann, E., Lebau, M. C., Massol, N.,.Secco, C., Baulieu, E. E. & Richard-Foy, H. (1979) C. R. Acad. Sci. Ser. D 288, 255-258. 6. Kohler, G. & Milstein, C. (1975) Nature (London) 256, 495497. 7. McKearn, T. J., Fitch, F. W., Smilek, D. E., Sarmiento, M. & Stuart, F. P. (1979) Immunol. Rev. 47,91-115. 8. Cerottini, J. C., Engers, H. D., McDonald, H. R. & Brunner, K. T. (1974) J. Exp. Med. 140, 703-717. 9. Littlefield, J. W. (1964) Science 145, 709. 10. Vaitukaitis, J., Robbins, J. B., Nieschlag, E. & Ross, G. T. (1971) J. Clin. Endocrinol. Metab. 33,988-991. 11. Palacios, R., Palmiter, R. D. & Schimke R. T. (1972) J. Biol. Chem. 247, 2316-2321. 12. Davidson, W. F. & Parish, C. R. (1975) J. Immunol. Methods 7, 291-300. 13. Bazin, H., Beckers, A. & Querinjean, P. (1974) Eur. J. Immunol.

4,44-48. 14. Cowan, N. J., Secher, D. S. & Milstein, C. (1974) J. Mol. Biol. 90, 691-701. 15. Shulman, M., Wilde, C. D. & Kohler, G. (1978) Nature (London)

276,269-270. 16. Notides, A., Hamilton, D. E., & Auer, H. E. (1975) J. Biol. Chem.

250,3945-3950. 17. Little, M., Szendro, P., Teran, C., Hughes, A. & Jungblut, P. W. (1975) J. Steroid Biochem. 6, 493-500. 18. Yamamoto, K. R. (1974) J. Biol. Chem. 249,7068-7075. 19. Bloch, K. J., Morse, H. C. & Austin, F. (1968) J. Immunol. 101, 650-657.