JOURNAL OF CELLULAR PHYSIOLOGY 152:553-557 (1992)
ECF Receptor Regulation in Normal Mouse Mammary Gland SANDRA Z. HASLAM,* LAURA J. COUNTERMAN, AND KATHERINE A. N U M M Y
Physiology Departmcnt, Michigan State University, East Lansing, Michigan 48823- 1 10 I Estrogen (El, progesterone (P), and epidermal growth factor (EGF) are known to regulate growth and development of the normal mammary gland, and it is possible that ECF may interact with E and/or P. Estrogen (ER), progesterone (PR), and ECF receptors (EGF-R) have been detected in both mammary epithelial and stroma1 cells, and the relative roles of the various cells types in horrnone-dependent growth regulation are not known. The present studies were undertaken to determine if E and/or P influence EGF action hy exerting a regulatory effect on EGF-R levels and which cell types are affected. The comparative effects of ovariectomy and hormone treatments on EGF-R levels were examined in immature, pubertal 5-week-old and sexually mature 1O-week-old female mice. EGF-R were characterized as a single class of high affinity sites and EGF-R concentration was 2-fold higher in glands of 5-week-old mice. Ovariectomy had no significant effect an EGF-R concentration in either age group, and treatment with E and/or P had no effect on EGF-R levels in either epithelial or stromal cells in 5-week-old mice. In contrast, E+P treatment caused a 2-fold increase in receptor Concentration in 10-week-old mice in the mammary epithelium. Thus it appears that the developmental state of the gland may determine the nature and extent of the interaction of of ECF, E, and P. 0 1992 Wiley Ll55, Inc.
The ovarian hormones estrogen (El and progesterone
(P) are known to be important for the in vivo growth and development of the normal mammary gland. Recently, growth factors such as epidermal growth factor (EGF) have also been implicated in normal mammary gland growth and development (Coleman et al., 1987; Scheffield and Welsch, 1987; Vonderhaar, 1987). EGF has been shown to stimulate proliferation of normal and neoplastic mammary tissues in vitro (Yang e t al., 1980; Taketani and Oka, 1983) and in vivo, and there is some evidence that EGF may act synergistically with E and/or P (Scheffield and Welsch, 1987; Vonderhaar, 1987). In this context, it is possible that one of the ways in which ovarian hormones can affect cell proliferation is by regulating EGF receptor (EGF-R) Concentration. The normal mouse mammary gland possesses membrane receptors for EGF and the concentration of receptors varies a t different developmental states. Receptor levels, present in both epithelial and stromal cells, are high in the immature pubertal mammary gland and then decrease with age. Receptor levels increase again with the onset of pregnancy, reaching a peak level a t 10 days followed by a rapid decline to very low levels during lactation (Edery et al., 1985). The relative distribution of EGF-R between epithelial and stromal cell compartments in the mature mammary gland is not known. Not much is known about how EGF-R concentration is regulated in the normal mammary gland in vivo. However, it has been reported that in another ovarian hormone target tissue, immature rat uterus, estrogen stimulates an increase in EGF-R concentration (Mukku and Stancel, 19851, whereas in certain human mam0 1992 WILEY-LISS. INC.
mary carcinoma cell lines, progesterone has been shown to regulate EGF-R concentration (Ewing et al., 1989). The present studies were undertaken to investigate the specific roles of E and/or P in regulating EGF-R in the epithelial and stromal cells of the normal mammary gland a s a function of developmental state. The mouse mammary gland undergoes a developmental process that results in the sequential acquisition of responsiveness to the proliferative effects of E and/or P (Haslam, 1989). Responsiveness to E is first detectable a t 3 weeks of age. whereas responsiveness to P is acquired much later, a t 7 weeks. Recent studies in our laboratory have demonstrated that mammary stroma plays a n important role in the in vivo acquisition of hormonal responsiveness to P (Haslam and Counterman, 1991). To elucidate further the mechanisms by which ovarian hormones influence mammary gland growth at different developmental states, the potential roles of E and/or P in regulating EGF receptor concentration were analyzed in epithelial and stromal cells in mammary glands of immature pubertal vs. sexually mature mice. MATERIALS AND METHODS Chemicals Mouse '"'1-EGF (100 pCi/pg) was purchased from Amersham Corp. (Arlington Heights, 11). Unlabeled
Received March 2, 1992; accepted April 24, 1992. *To whom reprint requests/correspondence should be addressed.
554
HASLAM ET AL
EGF and all other hormones were purchased from Sigma (St. Louis, MO).
Animals Balbic mice from our own colony were kept intact or bilaterally ovariectomized under nembutal anesthesia a t 5 or 10 weeks of age and 1 week prior to receiving a single sc injection of E (1 pg), or P (1mg) or E + P (1 pg + 1mg) in a solution of 0.85% NaCl-1% ethanol. At indicated times after treatment, mammary glands were removed, crude microsomal membranes were prepared, and high affinity EGF receptor binding was determined by '"I-EGF ligand bindin assay method or froZen tissues were analyzed by gI-EGF autoradiography. For certain experiments in which mammary stroma was assayed separately from intact gland, the stromal component was either surgically separated from the intact gland or epithelium devoid stroma was obtained by the cleared fat pad technique of De Ome et al. (1959).
Statistical analysis
*
All data are expressed as the mean SEM and were analyzed for significance using Student's t-test or analysis of variance where applicable.
RESULTS Effect of E and/or P on the concentration of lZ5I-EGFbinding sites Scatchard analysis of lZ51-EGFequilibrium binding to membrane preparations of mammary tissues from ovary intact 5- and 10-week-old mice revealed a single class of high affinity receptors with Kd values of 1.2 2 0.08 and 1.6 & 0.10 x 10 M and receptor concentrations of 15.9 +- 0.8 (n = 5) and 7.5 ? 0.4 (n = 7) fmoleimg DNA, respectively. The concentration of '251-EGF binding sites in mammary glands of 5-weekold mice was 2-fold greater (p = 0.02) than that of 10week-old mice. Ovariectomy had no significant effect on the concentration of EGF receptors in either 5- or 10-week-old mice. Time-course analysis of specific "'I-EGF binding Binding assays after treatment of 5-week-old mice with E, P, or E + P Membrane protein (50-400 pg) was incubated with showed that hormone treatment had no significant ef1-25 nM lZ5I-EGFfor 16-18 h r at 23°C in the presence fect on EGF receptor concentration (Fig. 1A). In 10or absence of 1 pg unlabeled EGF (Vonderhaar, 1986). week-old mice, although treatment with either E or P Specific binding was calculated by subtracting nonspe- alone had no effect, E + P treatment resulted in a signifcific binding obtained in the absence of unlabeled EGF; icant 2-fold increase in EGF receptor concentration a s nonspecific binding was less than 5% of total added compared with ovariectomized control mice (Fig. 1B). radioactivity. As previously reported, under these con- The change in amount of 1251-EGFbinding was due to ditions maximal binding was observed a t 6 h r and was a n actual increase in receptor concentration since the constant between 6 and 24 hr. The specificity of '"Iamount of "'I-EGF binding was not significantly alEGF binding was established by the lack of binding tered even after in vitro desaturation of the membranes competition by mammogenic hormones and other of endogenous growth factors with 3 M MgCl, treatgrowth factors (insulin, prolactin, and fibroblast ment (Table 1). growth factor). Specific binding was linear with increasing amounts of membrane protein in the range of Cellular distribution of E G F receptors 5 0 4 0 0 pg (data not shown). For some studies, desatuSince the mammary gland is composed of epithelial ration of the receptor to remove endogenous EGF was and stromal cell components, and receptors for estrogen performed with 3 M MgC1, as previously described (Ed- and progestins are present in both epithelial and stroery et al., 1985). Affinity constants and the number of ma1 cells (Haslam and Shyamala, 1981), it was of interbinding sites were analyzed according to the method of est to investigate the potential effects of E and/or P on Scatchard (1948) and Scatchard plots were interpolated the cellular distribution of EGF-R in mammary tissues using Biostatistics 3PC Program (A2 Devices, Alameda a s a function of developmental state. EGF-R have been CA). DNA was quantified a s previously described (Ce- previously identified by histoautoradiographic analyriotti, 1952) using calf thymus DNA as the standard. sis of '"I-EGF binding to be present in both epithelial and stromal cells of immature 5-week-old mice (Colel2'1-EGF histoautoradiography man et al., 1988). Since stromal tissue predominates at Mammary tissue embedded in OCT was snap-frozen this stage of development, it was possible that a n inin isopentane chilled with liquid N,. Sections (6 pm) crease in epithelial EGF-R in response t o hormone were mounted on slides coated with tissue adhesive treatment could have been obscured by a lack of in(Abbott Labs) and stored dry a t -20°C a s previously crease in the stromal cell EGF-R concentration. To addescribed (Coleman et al., 1988).Sections were washed dress this question, mammary tissue from inguinal in Media 199 (Gibco) with 1% BSA for 20 min a t room mammary glands was divided into two parts, one contemperature (RT). Tissues were then incubated with taining both epithelial and stromal cells and the other Media 199 + 1%BSA containing 1nM l2'1-EGF alone containing only stromal cells. That portion of the inor with 1 pgiml unlabeled EGF for 4 h r a t RT. Sections guinal gland between the nipple and lymph node conwere washed in cold PBS twice for 5 min, then fixed in tained both epithelial and stromal cells, whereas that PBS/10% formalin for 10 min, ice-cold absolute metha- portion of the gland distal from the lymph node connol for 3 min, ice-cold acetone for 1min, rinsed twice in tained only stromal cells. Measurement of lZ5I-EGF PBS, and air-dried overnight. Slides were processed for binding showed that EGF-R concentration present in autoradiography as previously described (Haslam, stromal cells alone was equivalent t o EGF-R concentra1988) and developed after 2,4, or 6 weeks exposure and tion present in the epithelium + stroma (Table 2). An analysis of the effects of E and P on cellular distribution counterstained with 1%methyl green.
555
MAMMARY EGF RECEPTOR REGULATION
20T
TABLE 3. Effect of hormoncs on cellular distribution of EGF-R in
A
10-week-old mice Specific ""I-EGF binding (fm'me DNA) Hormone treatment Intact control OVX conLrol E-treated E + P treated
Epithelium
+ Stroma
7.2 * 0 3' 6.3 ? 0 2 6.6 ? 0 3 15.4? 0.4*
Stroma
* *
7.3 0.2 6.9 i 0.4 7.6 0.4 7.9 t 0.3
'M ? SEM Each value represents the mean of 2 experiments. In each experiment pooled tissue from 3 mice was assayed. ';P= 0.02 E i P group significantly greater than control or E-treated groups.
6
12
15
24
T i m e A f t e r T r e c h e r i : (i-rs) Fig. 1. Effect of E and P treatment on EGF-R concentration. Ovariectomized mice, 5 (A) or 10 (B) weeks old, were treated with salinc (01, E (1 Kgi (*i,P (1 mg) (W or E + P (1 pg + 1 mgi ( A )and assayed a t specified times after treatment for specific '""I-EGF binding a s described in Materials and Methods. Each point represents t h e M ? SEM of 2 4 experiments. * P = 0.02 that E + P treated group had higher EGF-R levels t h a n saline, E, or P-treated groups.
TABLE 1. Concentration of total and free '"I-EGF binding sites in hormone-treated 10-week-old mice Hormone treatment
OVX control E + P-treated
Free receptors Total receptors' ifm/mg protein) 4.67 ? 0.05' 7.90 Z 0.06
4.09 ? 0.04 8.25 2 0.09
'Total receptor concentration was determined using a MgC12 desaturation method as described in Materials and Methods. 'M I SEM Each value represents the mean of 2 experiments In each experiment pooled tissue from 3 mice was assaycd.
TABLE 2. Effect of hormones on rellular distribution of EGF-R in 5-week-old mice Specific lZ5I-EGFbinding (fmime DNA1 Hormone treatment Intact control OVX control E-treated E + P treated
Epithelium + Stroma
Stroma
15.9 ? 0.8l 12.1 ? 2.0 15.3 ? 0.9
13.5 i; 1.8
13.4 & 0.5
11.1 Z 2.3 14.3 Z 0.8 12.5 ? 0.9
'M 2 SEM. Each value represents the mean of 3 experiments. In each experiment pooled tissue from 3 mice was assa.yed.
of EGF-R showed that hormone treatment had no effect on EGF-R concentration in either epithelial or stromal cells (Table 2). A similar analysis was carried out in 10-week-old mice. Due to the uniform distribution of epithelium throughout the gland at this age, it is not possible to isolate pure stromal tissue as described above for 5-week-old mice. Thus in order to obtain only stromal tissue, the epithelial rudiment was removed from one
inguinal gland a t 3 weeks of age, leaving only the stroma1 fat pad (DeOme et al., 1959); the contralateral inguinal gland was left intact, i.e., containing both epithelium and stroma. At 10 weeks of age, the epithelium-free stroma and intact glands were assayed separately for ' 251-EGFbinding. EGF-R concentrations in mammary stroma and in the epithelium+stroma were similar. In contrast to the results obtained in 5-week-old mice, a 2-fold increase in EGF-R was observed after E + P treatment in the intact gland that contained both epithelium and stroma (Table 3). These results indicate that EGF-R were increased by hormone treatment in the epithelial compartment only. Histoautoradiographic analysis of lZ5I-EGFbinding in 5- and 10-week-old mammary glands (Fig. 2) confirmed the presence of EGF-R in both epithelial and stromal cells a t both developmental states.
DISCUSSION The role of E and/or P in the regulation of EGF-R concentration in the mammary glands of either pubertal immature or sexually mature mice was investigated. Using a n l2'1-EGF ligand binding assay, we found a single class of low capacity, high affinity binding sites highly specific for EGF a t both stages of mammary gland development. These findings are in good agreement with the report of Vonderhaar (1985). However, Edery et al. (1985) have reported the presence of two classes of EGF binding sites in mouse mammary gland with Kd values of lo-'' M and lo-' M, respectively. The presence of one vs. two classes of binding sites in a variety of tissues has been reported to vary considerably and the significance of these observations is presently not known. However, it is generally believed that high affinity binding sites mediate the biological effects of EGF (Defeze et al., 1989). Our finding that EGF-R concentration in the immature mammary gland is at least 2-fold higher than t h a t of the sexually mature, 10-week-old gland is in agreement with the observations of Edery et al. (1985). Our present studies on hormonal regulation of EGF-R concentration have shown that in the immature mammary gland, EGF-R levels appear to be independent of regulation by E and/or P. By contrast, in the mature 10-week-old mammary gland, administration of exogenous E+P results in a significant 2-fold increase in EGF binding sites. Scatchard analysis of the binding data indicates that the increase in EGF-R levels in E + P treated mammary glands is due to a n actual increase in the number of binding sites rather than to a change in binding affinity.
556
HASLAM ET AL.
Fig. 2. Histoautoradiography of 'T-EGF binding in the 5- and 10week-old mammary gland. Specific '"I-EGF binding was localked in both epithelial and stromal cells (arrows) of 5 (A,B)- and 10 (C,D)week-old mammary glands. Binding carried out in the presence of excess unlabeled EGF (B,D) abolished cellular labeling. No reduction
in labeling was observed in the presence of excess unrelated growth factors (fibroblast growth factor) (not shown). Exposure was for 2 and 4 weeks for 5- and 10-week-old mammary glands, respectively. Magnification X200.
It has been previously reported that EGF-R levels increase during pregnancy (Edery et al., 1985)at a time when circulating levels of E and P are known to be elevated. Interestingly, the existing level of EGF-R in the mature gland was not decreased by ovariectomy, indicating that constitutive levels of EGF-R are increased only when E + P is added. Thus our observations provide a plausible explanation for the basis for elevated EGF-R levels during pregnancy, a time when
circulating levels of estrogen and progesterone are significantly elevated. Progesterone has also been reported to increase EGF-R levels in human mammary carcinoma cell lines that are estrogen and progesterone receptor positive (Ewing et al., 1989).Thus it is possible that progesterone-dependent increase of EGF-R may be a mammary-specific regulatory mechanism. Recently, we have reported that P has differing effects on mammary proliferation in immature 5-week-
MAMMARY EGF RECEPTOR REGULATION
old mice a s compared with its effects in sexually mature 10-week-old mice. No stimulatory effect on cell proliferation was observed in 5-week-old mice, whereas P in combination with E provides the major mitogenic stimulus in 10-week-old mice (Haslam, 1989). An analysis of in vivo effects of EGF in 5-week-old ovariectomized mice has shown that EGF promotes normal ductal morphogenesis by stimulating proliferation in the end buds (Coleman et al., 1988). This proliferative effect of EGF occurs in the absence of E or P. Thus our observations that EGF-R levels are independent of ovarian hormone regulation in immature mouse mammary gland agree well with the reported in vivo response to EGF. At variance with these results is a report that after 3 days of exposure to slow release pellets containing E+P, ovary intact immature mice exhibited increased levels of EGF-R (Vonderhaar, 1985). This suggests that the developmental process that leads to the condition of E + P regulation of EGF-R in the mature mammary gland has a n underlying hormonal component. In this regard, we have recently found that P responsiveness similar to that observed in mature mice can be precociously induced in immature mice by repeated injections of E (unpublished observations). However, the specific mechanisms controlling EGF-R expression in the immature gland is not well understood; one possibility is that increased levels of EGF itself cause a down regulation of EGF-R (Coleman and Daniel, 1990). The in vivo response of the mature mammary gland to EGF is not known; however, a role for salivary gland EGF has been implicated in promoting E +P-induced lobuloalveolar development similar to that which occurs during pregnancy (Scheffield and Welsch, 1987). Further potential evidence for a biological role of EGF comes from the findings that EGF appears to be produced in vivo in mammary epithelial cells (Brown et al., 1989). Our results on the cellular distribution of EGF-R demonstrate that the receptors are present in epithelial and stromal cells in both pubertal and mature mammary glands. These results agree with previous localization of EGF-R in stromal and epithelial cells of 5-week-old mammary glands (Coleman et al., 1988). We have extended these observations to the mature mammary gland and, furthermore, find t h a t upon treatment with E + P , EGF-R are preferentially increased in the epithelial compartment of the gland. In this context it is noteworthy that in the mature mammary gland estrogen-inducible progesterone receptors are also localized preferentially in the epithelial cells. Since E alone was without effect on EGF-R concentration, it is quite likely that EGF-R regulation is mediated through progesterone action in the epithelium. It should be noted that in the present study the effects of E and/or P on stromal EGF-R concentration were measured in epithelium-devoid stromal fat pads. Thus we cannot rule out the possibility that the proximity of mammary epithelium is needed for the induction of receptors in the stroma. In summary, we report that EGF-R concentration appears to be developmentally regulated in the mouse mammary gland. In the immature mammary gland, epithelial and stromal cell EGF-R are regulated independent of ovarian hormones. By contrast, in the ma-
557
ture mammary gland, a significant increase in epithelial but not stromal EGF-R is induced by E + P most likely via estrogen-inducible progesterone receptors. These results suggest the certain proliferative effects of E and/or P in vivo may be mediated indirectly via growth factors. However, the role of stromal EGF-R in the growth and differentiation of the mammary gland remains to be determined. The proliferative effects of EGF in epithelial and stromal cells of the mature mammary gland are currently being investigated.
ACKNOWLEDGMENTS This work was supported by NIH Grant CA 40401.
LITERATURE CITED Brown, C.F., Teng, C.T., Pentecost, B.T., and DiAugustene, R.P. (1989)Epidermal growth factor precursor in mouse lactating mammary gland alveolar cells. Mol. Endocrinol., 3,1077-1083. Ceriotti, G.A. (1952) Microchemical determination of DNA. J. Biol. Chem., 198t297-315. Coleman, S.: and Daniel, C.W. (1990) Inhibition of mouse mammary ductal morphogenesis and down regulation of the EGF receptor by epidermal growth factor. Dev. Biol., 137t425433. Coleman, S., Silberstein, G.B., and Daniel, C.W. (1988) Ductal morphogenesis in the mouse mammary gland: Evidence supporting a role for epidermal growth factor. Dev. Biol., 127t304-315. Defeze, L.H.K., Boonstra, J., Musenheldic, J., Kruijer, W., Tertoolen, L.G.J., Tilly, B.C., Hunter, T., van Bergen en Henegorewen, P.M.P.. Moolenaar, W.H., and Laat, S.W. (1989) Signal transduction by EGF occurs through the subclass of high affinity receptors. J. Cell Biol., 109,2495-2507. DeOme, K.B., Foulken, L.J.,Bern, H.A., and Blair, P.B. (1959)Development of mammary tumors from HAN transplanted into glandfree mammary fat pads of C3H mice. Cancer Res., 19t515624. Edery, M., Pang, K., Larson, L., Colosi, T., and Nandi, S. (1985) Epidermal growth factor receptor levels in mouse mammary glands in various physiological states. Endocrinol., 11 7t405-411. Ewing, T.M., Murphy, L.J., Ng, M-L., Pang, G.Y.N., Lee, C.S.L., Atts, C.K.W., Sutherland, R.L. (1989) Regulation of epidermal growth factor receptor by progestins and glucocorticoids in human breast cancer cell lines. Int. J . Cancer, 44:744-752. Haslam, S.Z. (1988)Progesterone effects of DNA synthesis in normal mouse mammary glands. Endocrinol., 122:464470. Haslam, S.Z. (1989) The ontogeny of mouse mammary gland responsiveness to ovarian steroid hormones. Endocrinol., 125t2766-2772. Haslam, S.Z., and Counterman, L.J. (1991) Mammary stroma modulates hormonal responsiveness of mammary epitheliuni in Vivo in the mouse. Endocrinol., 129,2017-2023. Haslam, S.Z., and Shyamala, G., (1981)Relative distribution of estrogen receptors and progesterone receptors among epithelial, adipose and connective tissues of normal mammary gland. Endocrinol., I O8:82&830. Mukku, V.R., and Stancel, G.M. (1985) Regulation of epidermal growt,h factor receptor by estrogen. J. Biol. Chem., 260:9820-9824. Scatchard, G. (1949)The attraction of proteins for small molecules and ions. Ann. NY Acad. Sci., 51;600472. Scheffield, L.G., and Welsch, C.W. 11987)Influence of submandibular salivary glands on hormone responsiveness of mouse mammary glands. Proc. SOC.Exptl. Med., 186;36%377. Taketani, Y., and Oka, T. (1983)Epidermal growth factor stimulates cell proliferation and inhihits functional differentiation of mouse mammary epithelial cells in culture. Endocrinol., 113371-877. Vonderhaar, B.K. (1985)Hormones and growth factors in mammary gland development. In: Control o f Cell Growth and Proliferation. C.M. Veneziale, ed. Van Nostrand-Reinhold, New York, pp. 1 1 3 3 . Vonderhaar, B.K. 11987) Local effects of EGF, a-TGF and EGF-like growth factors on lobuloalveolar development of the mouse mammary gland in vivo. J. Cell Physiol., 132581-584. Vonderhaar, B.K., Tang, E., Lyster, R.L., and Nascimento, M.C.S. (1986)Thyroid hormone regulation of EGF receptor levels in mouse mammary glands. Endocrinol., 119r580-585. Yang, J., Guzman, R., Richards, J.,Imagawa, W., McCormick, K., and Nandi, S., (1980) Growth factor- and cyclic nucleotide-induced proliferation of normal and malignant mammary epithelial cells in primary culture. Endocrinol., 107t35-40.
ECF Receptor Regulation in Normal Mouse Mammary Gland SANDRA Z. HASLAM,* LAURA J. COUNTERMAN, AND KATHERINE A. N U M M Y
Physiology Departmcnt, Michigan State University, East Lansing, Michigan 48823- 1 10 I Estrogen (El, progesterone (P), and epidermal growth factor (EGF) are known to regulate growth and development of the normal mammary gland, and it is possible that ECF may interact with E and/or P. Estrogen (ER), progesterone (PR), and ECF receptors (EGF-R) have been detected in both mammary epithelial and stroma1 cells, and the relative roles of the various cells types in horrnone-dependent growth regulation are not known. The present studies were undertaken to determine if E and/or P influence EGF action hy exerting a regulatory effect on EGF-R levels and which cell types are affected. The comparative effects of ovariectomy and hormone treatments on EGF-R levels were examined in immature, pubertal 5-week-old and sexually mature 1O-week-old female mice. EGF-R were characterized as a single class of high affinity sites and EGF-R concentration was 2-fold higher in glands of 5-week-old mice. Ovariectomy had no significant effect an EGF-R concentration in either age group, and treatment with E and/or P had no effect on EGF-R levels in either epithelial or stromal cells in 5-week-old mice. In contrast, E+P treatment caused a 2-fold increase in receptor Concentration in 10-week-old mice in the mammary epithelium. Thus it appears that the developmental state of the gland may determine the nature and extent of the interaction of of ECF, E, and P. 0 1992 Wiley Ll55, Inc.
The ovarian hormones estrogen (El and progesterone
(P) are known to be important for the in vivo growth and development of the normal mammary gland. Recently, growth factors such as epidermal growth factor (EGF) have also been implicated in normal mammary gland growth and development (Coleman et al., 1987; Scheffield and Welsch, 1987; Vonderhaar, 1987). EGF has been shown to stimulate proliferation of normal and neoplastic mammary tissues in vitro (Yang e t al., 1980; Taketani and Oka, 1983) and in vivo, and there is some evidence that EGF may act synergistically with E and/or P (Scheffield and Welsch, 1987; Vonderhaar, 1987). In this context, it is possible that one of the ways in which ovarian hormones can affect cell proliferation is by regulating EGF receptor (EGF-R) Concentration. The normal mouse mammary gland possesses membrane receptors for EGF and the concentration of receptors varies a t different developmental states. Receptor levels, present in both epithelial and stromal cells, are high in the immature pubertal mammary gland and then decrease with age. Receptor levels increase again with the onset of pregnancy, reaching a peak level a t 10 days followed by a rapid decline to very low levels during lactation (Edery et al., 1985). The relative distribution of EGF-R between epithelial and stromal cell compartments in the mature mammary gland is not known. Not much is known about how EGF-R concentration is regulated in the normal mammary gland in vivo. However, it has been reported that in another ovarian hormone target tissue, immature rat uterus, estrogen stimulates an increase in EGF-R concentration (Mukku and Stancel, 19851, whereas in certain human mam0 1992 WILEY-LISS. INC.
mary carcinoma cell lines, progesterone has been shown to regulate EGF-R concentration (Ewing et al., 1989). The present studies were undertaken to investigate the specific roles of E and/or P in regulating EGF-R in the epithelial and stromal cells of the normal mammary gland a s a function of developmental state. The mouse mammary gland undergoes a developmental process that results in the sequential acquisition of responsiveness to the proliferative effects of E and/or P (Haslam, 1989). Responsiveness to E is first detectable a t 3 weeks of age. whereas responsiveness to P is acquired much later, a t 7 weeks. Recent studies in our laboratory have demonstrated that mammary stroma plays a n important role in the in vivo acquisition of hormonal responsiveness to P (Haslam and Counterman, 1991). To elucidate further the mechanisms by which ovarian hormones influence mammary gland growth at different developmental states, the potential roles of E and/or P in regulating EGF receptor concentration were analyzed in epithelial and stromal cells in mammary glands of immature pubertal vs. sexually mature mice. MATERIALS AND METHODS Chemicals Mouse '"'1-EGF (100 pCi/pg) was purchased from Amersham Corp. (Arlington Heights, 11). Unlabeled
Received March 2, 1992; accepted April 24, 1992. *To whom reprint requests/correspondence should be addressed.
554
HASLAM ET AL
EGF and all other hormones were purchased from Sigma (St. Louis, MO).
Animals Balbic mice from our own colony were kept intact or bilaterally ovariectomized under nembutal anesthesia a t 5 or 10 weeks of age and 1 week prior to receiving a single sc injection of E (1 pg), or P (1mg) or E + P (1 pg + 1mg) in a solution of 0.85% NaCl-1% ethanol. At indicated times after treatment, mammary glands were removed, crude microsomal membranes were prepared, and high affinity EGF receptor binding was determined by '"I-EGF ligand bindin assay method or froZen tissues were analyzed by gI-EGF autoradiography. For certain experiments in which mammary stroma was assayed separately from intact gland, the stromal component was either surgically separated from the intact gland or epithelium devoid stroma was obtained by the cleared fat pad technique of De Ome et al. (1959).
Statistical analysis
*
All data are expressed as the mean SEM and were analyzed for significance using Student's t-test or analysis of variance where applicable.
RESULTS Effect of E and/or P on the concentration of lZ5I-EGFbinding sites Scatchard analysis of lZ51-EGFequilibrium binding to membrane preparations of mammary tissues from ovary intact 5- and 10-week-old mice revealed a single class of high affinity receptors with Kd values of 1.2 2 0.08 and 1.6 & 0.10 x 10 M and receptor concentrations of 15.9 +- 0.8 (n = 5) and 7.5 ? 0.4 (n = 7) fmoleimg DNA, respectively. The concentration of '251-EGF binding sites in mammary glands of 5-weekold mice was 2-fold greater (p = 0.02) than that of 10week-old mice. Ovariectomy had no significant effect on the concentration of EGF receptors in either 5- or 10-week-old mice. Time-course analysis of specific "'I-EGF binding Binding assays after treatment of 5-week-old mice with E, P, or E + P Membrane protein (50-400 pg) was incubated with showed that hormone treatment had no significant ef1-25 nM lZ5I-EGFfor 16-18 h r at 23°C in the presence fect on EGF receptor concentration (Fig. 1A). In 10or absence of 1 pg unlabeled EGF (Vonderhaar, 1986). week-old mice, although treatment with either E or P Specific binding was calculated by subtracting nonspe- alone had no effect, E + P treatment resulted in a signifcific binding obtained in the absence of unlabeled EGF; icant 2-fold increase in EGF receptor concentration a s nonspecific binding was less than 5% of total added compared with ovariectomized control mice (Fig. 1B). radioactivity. As previously reported, under these con- The change in amount of 1251-EGFbinding was due to ditions maximal binding was observed a t 6 h r and was a n actual increase in receptor concentration since the constant between 6 and 24 hr. The specificity of '"Iamount of "'I-EGF binding was not significantly alEGF binding was established by the lack of binding tered even after in vitro desaturation of the membranes competition by mammogenic hormones and other of endogenous growth factors with 3 M MgCl, treatgrowth factors (insulin, prolactin, and fibroblast ment (Table 1). growth factor). Specific binding was linear with increasing amounts of membrane protein in the range of Cellular distribution of E G F receptors 5 0 4 0 0 pg (data not shown). For some studies, desatuSince the mammary gland is composed of epithelial ration of the receptor to remove endogenous EGF was and stromal cell components, and receptors for estrogen performed with 3 M MgC1, as previously described (Ed- and progestins are present in both epithelial and stroery et al., 1985). Affinity constants and the number of ma1 cells (Haslam and Shyamala, 1981), it was of interbinding sites were analyzed according to the method of est to investigate the potential effects of E and/or P on Scatchard (1948) and Scatchard plots were interpolated the cellular distribution of EGF-R in mammary tissues using Biostatistics 3PC Program (A2 Devices, Alameda a s a function of developmental state. EGF-R have been CA). DNA was quantified a s previously described (Ce- previously identified by histoautoradiographic analyriotti, 1952) using calf thymus DNA as the standard. sis of '"I-EGF binding to be present in both epithelial and stromal cells of immature 5-week-old mice (Colel2'1-EGF histoautoradiography man et al., 1988). Since stromal tissue predominates at Mammary tissue embedded in OCT was snap-frozen this stage of development, it was possible that a n inin isopentane chilled with liquid N,. Sections (6 pm) crease in epithelial EGF-R in response t o hormone were mounted on slides coated with tissue adhesive treatment could have been obscured by a lack of in(Abbott Labs) and stored dry a t -20°C a s previously crease in the stromal cell EGF-R concentration. To addescribed (Coleman et al., 1988).Sections were washed dress this question, mammary tissue from inguinal in Media 199 (Gibco) with 1% BSA for 20 min a t room mammary glands was divided into two parts, one contemperature (RT). Tissues were then incubated with taining both epithelial and stromal cells and the other Media 199 + 1%BSA containing 1nM l2'1-EGF alone containing only stromal cells. That portion of the inor with 1 pgiml unlabeled EGF for 4 h r a t RT. Sections guinal gland between the nipple and lymph node conwere washed in cold PBS twice for 5 min, then fixed in tained both epithelial and stromal cells, whereas that PBS/10% formalin for 10 min, ice-cold absolute metha- portion of the gland distal from the lymph node connol for 3 min, ice-cold acetone for 1min, rinsed twice in tained only stromal cells. Measurement of lZ5I-EGF PBS, and air-dried overnight. Slides were processed for binding showed that EGF-R concentration present in autoradiography as previously described (Haslam, stromal cells alone was equivalent t o EGF-R concentra1988) and developed after 2,4, or 6 weeks exposure and tion present in the epithelium + stroma (Table 2). An analysis of the effects of E and P on cellular distribution counterstained with 1%methyl green.
555
MAMMARY EGF RECEPTOR REGULATION
20T
TABLE 3. Effect of hormoncs on cellular distribution of EGF-R in
A
10-week-old mice Specific ""I-EGF binding (fm'me DNA) Hormone treatment Intact control OVX conLrol E-treated E + P treated
Epithelium
+ Stroma
7.2 * 0 3' 6.3 ? 0 2 6.6 ? 0 3 15.4? 0.4*
Stroma
* *
7.3 0.2 6.9 i 0.4 7.6 0.4 7.9 t 0.3
'M ? SEM Each value represents the mean of 2 experiments. In each experiment pooled tissue from 3 mice was assayed. ';P= 0.02 E i P group significantly greater than control or E-treated groups.
6
12
15
24
T i m e A f t e r T r e c h e r i : (i-rs) Fig. 1. Effect of E and P treatment on EGF-R concentration. Ovariectomized mice, 5 (A) or 10 (B) weeks old, were treated with salinc (01, E (1 Kgi (*i,P (1 mg) (W or E + P (1 pg + 1 mgi ( A )and assayed a t specified times after treatment for specific '""I-EGF binding a s described in Materials and Methods. Each point represents t h e M ? SEM of 2 4 experiments. * P = 0.02 that E + P treated group had higher EGF-R levels t h a n saline, E, or P-treated groups.
TABLE 1. Concentration of total and free '"I-EGF binding sites in hormone-treated 10-week-old mice Hormone treatment
OVX control E + P-treated
Free receptors Total receptors' ifm/mg protein) 4.67 ? 0.05' 7.90 Z 0.06
4.09 ? 0.04 8.25 2 0.09
'Total receptor concentration was determined using a MgC12 desaturation method as described in Materials and Methods. 'M I SEM Each value represents the mean of 2 experiments In each experiment pooled tissue from 3 mice was assaycd.
TABLE 2. Effect of hormones on rellular distribution of EGF-R in 5-week-old mice Specific lZ5I-EGFbinding (fmime DNA1 Hormone treatment Intact control OVX control E-treated E + P treated
Epithelium + Stroma
Stroma
15.9 ? 0.8l 12.1 ? 2.0 15.3 ? 0.9
13.5 i; 1.8
13.4 & 0.5
11.1 Z 2.3 14.3 Z 0.8 12.5 ? 0.9
'M 2 SEM. Each value represents the mean of 3 experiments. In each experiment pooled tissue from 3 mice was assa.yed.
of EGF-R showed that hormone treatment had no effect on EGF-R concentration in either epithelial or stromal cells (Table 2). A similar analysis was carried out in 10-week-old mice. Due to the uniform distribution of epithelium throughout the gland at this age, it is not possible to isolate pure stromal tissue as described above for 5-week-old mice. Thus in order to obtain only stromal tissue, the epithelial rudiment was removed from one
inguinal gland a t 3 weeks of age, leaving only the stroma1 fat pad (DeOme et al., 1959); the contralateral inguinal gland was left intact, i.e., containing both epithelium and stroma. At 10 weeks of age, the epithelium-free stroma and intact glands were assayed separately for ' 251-EGFbinding. EGF-R concentrations in mammary stroma and in the epithelium+stroma were similar. In contrast to the results obtained in 5-week-old mice, a 2-fold increase in EGF-R was observed after E + P treatment in the intact gland that contained both epithelium and stroma (Table 3). These results indicate that EGF-R were increased by hormone treatment in the epithelial compartment only. Histoautoradiographic analysis of lZ5I-EGFbinding in 5- and 10-week-old mammary glands (Fig. 2) confirmed the presence of EGF-R in both epithelial and stromal cells a t both developmental states.
DISCUSSION The role of E and/or P in the regulation of EGF-R concentration in the mammary glands of either pubertal immature or sexually mature mice was investigated. Using a n l2'1-EGF ligand binding assay, we found a single class of low capacity, high affinity binding sites highly specific for EGF a t both stages of mammary gland development. These findings are in good agreement with the report of Vonderhaar (1985). However, Edery et al. (1985) have reported the presence of two classes of EGF binding sites in mouse mammary gland with Kd values of lo-'' M and lo-' M, respectively. The presence of one vs. two classes of binding sites in a variety of tissues has been reported to vary considerably and the significance of these observations is presently not known. However, it is generally believed that high affinity binding sites mediate the biological effects of EGF (Defeze et al., 1989). Our finding that EGF-R concentration in the immature mammary gland is at least 2-fold higher than t h a t of the sexually mature, 10-week-old gland is in agreement with the observations of Edery et al. (1985). Our present studies on hormonal regulation of EGF-R concentration have shown that in the immature mammary gland, EGF-R levels appear to be independent of regulation by E and/or P. By contrast, in the mature 10-week-old mammary gland, administration of exogenous E+P results in a significant 2-fold increase in EGF binding sites. Scatchard analysis of the binding data indicates that the increase in EGF-R levels in E + P treated mammary glands is due to a n actual increase in the number of binding sites rather than to a change in binding affinity.
556
HASLAM ET AL.
Fig. 2. Histoautoradiography of 'T-EGF binding in the 5- and 10week-old mammary gland. Specific '"I-EGF binding was localked in both epithelial and stromal cells (arrows) of 5 (A,B)- and 10 (C,D)week-old mammary glands. Binding carried out in the presence of excess unlabeled EGF (B,D) abolished cellular labeling. No reduction
in labeling was observed in the presence of excess unrelated growth factors (fibroblast growth factor) (not shown). Exposure was for 2 and 4 weeks for 5- and 10-week-old mammary glands, respectively. Magnification X200.
It has been previously reported that EGF-R levels increase during pregnancy (Edery et al., 1985)at a time when circulating levels of E and P are known to be elevated. Interestingly, the existing level of EGF-R in the mature gland was not decreased by ovariectomy, indicating that constitutive levels of EGF-R are increased only when E + P is added. Thus our observations provide a plausible explanation for the basis for elevated EGF-R levels during pregnancy, a time when
circulating levels of estrogen and progesterone are significantly elevated. Progesterone has also been reported to increase EGF-R levels in human mammary carcinoma cell lines that are estrogen and progesterone receptor positive (Ewing et al., 1989).Thus it is possible that progesterone-dependent increase of EGF-R may be a mammary-specific regulatory mechanism. Recently, we have reported that P has differing effects on mammary proliferation in immature 5-week-
MAMMARY EGF RECEPTOR REGULATION
old mice a s compared with its effects in sexually mature 10-week-old mice. No stimulatory effect on cell proliferation was observed in 5-week-old mice, whereas P in combination with E provides the major mitogenic stimulus in 10-week-old mice (Haslam, 1989). An analysis of in vivo effects of EGF in 5-week-old ovariectomized mice has shown that EGF promotes normal ductal morphogenesis by stimulating proliferation in the end buds (Coleman et al., 1988). This proliferative effect of EGF occurs in the absence of E or P. Thus our observations that EGF-R levels are independent of ovarian hormone regulation in immature mouse mammary gland agree well with the reported in vivo response to EGF. At variance with these results is a report that after 3 days of exposure to slow release pellets containing E+P, ovary intact immature mice exhibited increased levels of EGF-R (Vonderhaar, 1985). This suggests that the developmental process that leads to the condition of E + P regulation of EGF-R in the mature mammary gland has a n underlying hormonal component. In this regard, we have recently found that P responsiveness similar to that observed in mature mice can be precociously induced in immature mice by repeated injections of E (unpublished observations). However, the specific mechanisms controlling EGF-R expression in the immature gland is not well understood; one possibility is that increased levels of EGF itself cause a down regulation of EGF-R (Coleman and Daniel, 1990). The in vivo response of the mature mammary gland to EGF is not known; however, a role for salivary gland EGF has been implicated in promoting E +P-induced lobuloalveolar development similar to that which occurs during pregnancy (Scheffield and Welsch, 1987). Further potential evidence for a biological role of EGF comes from the findings that EGF appears to be produced in vivo in mammary epithelial cells (Brown et al., 1989). Our results on the cellular distribution of EGF-R demonstrate that the receptors are present in epithelial and stromal cells in both pubertal and mature mammary glands. These results agree with previous localization of EGF-R in stromal and epithelial cells of 5-week-old mammary glands (Coleman et al., 1988). We have extended these observations to the mature mammary gland and, furthermore, find t h a t upon treatment with E + P , EGF-R are preferentially increased in the epithelial compartment of the gland. In this context it is noteworthy that in the mature mammary gland estrogen-inducible progesterone receptors are also localized preferentially in the epithelial cells. Since E alone was without effect on EGF-R concentration, it is quite likely that EGF-R regulation is mediated through progesterone action in the epithelium. It should be noted that in the present study the effects of E and/or P on stromal EGF-R concentration were measured in epithelium-devoid stromal fat pads. Thus we cannot rule out the possibility that the proximity of mammary epithelium is needed for the induction of receptors in the stroma. In summary, we report that EGF-R concentration appears to be developmentally regulated in the mouse mammary gland. In the immature mammary gland, epithelial and stromal cell EGF-R are regulated independent of ovarian hormones. By contrast, in the ma-
557
ture mammary gland, a significant increase in epithelial but not stromal EGF-R is induced by E + P most likely via estrogen-inducible progesterone receptors. These results suggest the certain proliferative effects of E and/or P in vivo may be mediated indirectly via growth factors. However, the role of stromal EGF-R in the growth and differentiation of the mammary gland remains to be determined. The proliferative effects of EGF in epithelial and stromal cells of the mature mammary gland are currently being investigated.
ACKNOWLEDGMENTS This work was supported by NIH Grant CA 40401.
LITERATURE CITED Brown, C.F., Teng, C.T., Pentecost, B.T., and DiAugustene, R.P. (1989)Epidermal growth factor precursor in mouse lactating mammary gland alveolar cells. Mol. Endocrinol., 3,1077-1083. Ceriotti, G.A. (1952) Microchemical determination of DNA. J. Biol. Chem., 198t297-315. Coleman, S.: and Daniel, C.W. (1990) Inhibition of mouse mammary ductal morphogenesis and down regulation of the EGF receptor by epidermal growth factor. Dev. Biol., 137t425433. Coleman, S., Silberstein, G.B., and Daniel, C.W. (1988) Ductal morphogenesis in the mouse mammary gland: Evidence supporting a role for epidermal growth factor. Dev. Biol., 127t304-315. Defeze, L.H.K., Boonstra, J., Musenheldic, J., Kruijer, W., Tertoolen, L.G.J., Tilly, B.C., Hunter, T., van Bergen en Henegorewen, P.M.P.. Moolenaar, W.H., and Laat, S.W. (1989) Signal transduction by EGF occurs through the subclass of high affinity receptors. J. Cell Biol., 109,2495-2507. DeOme, K.B., Foulken, L.J.,Bern, H.A., and Blair, P.B. (1959)Development of mammary tumors from HAN transplanted into glandfree mammary fat pads of C3H mice. Cancer Res., 19t515624. Edery, M., Pang, K., Larson, L., Colosi, T., and Nandi, S. (1985) Epidermal growth factor receptor levels in mouse mammary glands in various physiological states. Endocrinol., 11 7t405-411. Ewing, T.M., Murphy, L.J., Ng, M-L., Pang, G.Y.N., Lee, C.S.L., Atts, C.K.W., Sutherland, R.L. (1989) Regulation of epidermal growth factor receptor by progestins and glucocorticoids in human breast cancer cell lines. Int. J . Cancer, 44:744-752. Haslam, S.Z. (1988)Progesterone effects of DNA synthesis in normal mouse mammary glands. Endocrinol., 122:464470. Haslam, S.Z. (1989) The ontogeny of mouse mammary gland responsiveness to ovarian steroid hormones. Endocrinol., 125t2766-2772. Haslam, S.Z., and Counterman, L.J. (1991) Mammary stroma modulates hormonal responsiveness of mammary epitheliuni in Vivo in the mouse. Endocrinol., 129,2017-2023. Haslam, S.Z., and Shyamala, G., (1981)Relative distribution of estrogen receptors and progesterone receptors among epithelial, adipose and connective tissues of normal mammary gland. Endocrinol., I O8:82&830. Mukku, V.R., and Stancel, G.M. (1985) Regulation of epidermal growt,h factor receptor by estrogen. J. Biol. Chem., 260:9820-9824. Scatchard, G. (1949)The attraction of proteins for small molecules and ions. Ann. NY Acad. Sci., 51;600472. Scheffield, L.G., and Welsch, C.W. 11987)Influence of submandibular salivary glands on hormone responsiveness of mouse mammary glands. Proc. SOC.Exptl. Med., 186;36%377. Taketani, Y., and Oka, T. (1983)Epidermal growth factor stimulates cell proliferation and inhihits functional differentiation of mouse mammary epithelial cells in culture. Endocrinol., 113371-877. Vonderhaar, B.K. (1985)Hormones and growth factors in mammary gland development. In: Control o f Cell Growth and Proliferation. C.M. Veneziale, ed. Van Nostrand-Reinhold, New York, pp. 1 1 3 3 . Vonderhaar, B.K. 11987) Local effects of EGF, a-TGF and EGF-like growth factors on lobuloalveolar development of the mouse mammary gland in vivo. J. Cell Physiol., 132581-584. Vonderhaar, B.K., Tang, E., Lyster, R.L., and Nascimento, M.C.S. (1986)Thyroid hormone regulation of EGF receptor levels in mouse mammary glands. Endocrinol., 119r580-585. Yang, J., Guzman, R., Richards, J.,Imagawa, W., McCormick, K., and Nandi, S., (1980) Growth factor- and cyclic nucleotide-induced proliferation of normal and malignant mammary epithelial cells in primary culture. Endocrinol., 107t35-40.
