| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLE |
Department of Physiology, Michigan State University, East Lansing, Michigan 48824
Address all correspondence and requests for reprints to: Sandra Z. Haslam, Ph.D., Department of Physiology, 108 Giltner Hall, Michigan State University, East Lansing, Michigan 48824. E-mail: shaslam{at}msu.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
knockout mice indicate that estrogen-induced proliferation is dependent upon the presence of estrogen receptor in mammary stromal cells, but not in epithelial cells. The purpose of the present study was to identify the underlying mechanism of estrogen-dependent stroma-derived effects on mammary epithelium. We have developed a minimally supplemented serum-free medium, collagen gel primary mammary coculture system to address the issue of stroma-derived, estrogen-dependent effects on epithelial cell proliferation. Conditioned medium from mammary fibroblasts or coculture with mammary fibroblasts caused increased epithelial cell proliferation and produced tubular/ductal morphology. Hepatocyte growth factor (HGF) was identified as the mediator of this effect, as the proliferative activity in fibroblast-conditioned medium was completely abolished by neutralizing antibody to HGF, whereas neutralizing antibodies to either epidermal growth factor or IGF-I had no effect. Treatment of mammary fibroblasts with estrogen increased the production of HGF. From these results we conclude that estrogen may indirectly mediate mammary epithelial cell proliferation via the regulation of HGF in mammary stromal cells and that HGF plays a crucial role in estrogen-induced proliferation in vivo. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(ER) is present in both mammary epithelial and stromal cells and that estrogenic effects on proliferation and progesterone receptor induction in the epithelium are modulated by mammary stroma both in vivo and in vitro (2). Recent studies of ER knockout mice indicate that estrogen-induced proliferation is dependent upon the presence of ER in mammary stromal cells, but not in epithelial cells (3). However, the underlying mechanism of estrogen-dependent stroma-derived effects on mammary epithelium has not been elucidated.
Cell culture models have been useful to elucidate the underlying mechanisms of cell-cell interactions and the cell types involved. In many instances the culture conditions were designed to stimulate long-term, maximal epithelial cell proliferation (4). To accomplish this culture studies were carried out in the presence of serum, high concentrations of growth factors such as epidermal growth factor (EGF), or impure culture components or supplements (serum albumin, collagen, fetuin, Matrigel, pituitary extracts, crude soybean lectin, trypsin inhibitor, -tocopherol succinate, and cholesterol) (4, 5). In addition, high concentrations of insulin (2.5–10 µg/ml) have been widely used (5, 6, 7). It has since been shown that at these doses insulin has a potent proliferative effect that is most likely mediated through the IGF-I receptor, because the insulin effect can be replaced by IGF-I (8). Furthermore, IGF-I can synergize with EGF and other mammogenic hormones in serum-free medium to stimulate proliferation (9). Because numerous growth factors and other unidentified potential growth-promoting or growth-inhibiting components are present under these conditions, our understanding of the specific contribution of stromal cells and the underlying mechanism has been limited. The use of extracellular matrix gel preparations (collagen I and Matrigel) has allowed mammary epithelial cells to be cultured as three-dimensional structures that more closely resemble their architecture in vivo and provide a more physiological context to study growth and morphogenesis (10). However, despite these advances, estrogen-induced proliferation of normal mammary epithelial cells has not been demonstrated in vitro.
In the present studies we used a minimally supplemented, serum-free, collagen gel primary culture system. Under these conditions, untreated control cultures exhibit a low basal proliferative state and an absence of morphological change, similar to those observed in ovariectomized mice (11). This is particularly relevant because a majority of studies performed to define the growth-promoting and morphological effects of hormones and growth factors in vivo have been carried out in ovariectomized mice. Another distinguishing feature of our culture system is the short, 3-d period in which the studies are carried out. This period is similar to that in which the acute proliferative effects of hormones and growth factors have been observed in vivo in ovariectomized mice (12, 13, 14). Using this culture model we have made the novel observation that a soluble factor produced by stromal cells mediates estrogen-induced proliferation in epithelial cells. We report that mammary fibroblasts derived from mature mammary gland produce hepatocyte growth factor (HGF) in vitro and that HGF production is increased by estrogen.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cell culture
Mammary epithelial cells and fibroblasts were isolated from adult virgin (10- to 14-wk-old) BALB/c female mice using enzymatic dissociation methods as previously described (15). Purified fibroblasts were obtained in the supernatant fraction by differential centrifugation at 80 x g for 30 sec. Cell viability was approximately 95% as determined by trypan blue exclusion. Fibroblasts were plated in 5% fetal bovine serum (FBS)-DMEM/F-12 and allowed to attach for 2 h, and any contaminating epithelial cells were removed by gently rinsing with DMEM/F-12. The purity of the fibroblast cultures was greater than 95% (15). Fibroblast contamination of epithelial cultures was less than 5% as determined by immunocytochemical assay with antivimentin antibody (16).
Coculture of mammary epithelial cells and fibroblasts
First passage fibroblasts (0.5 x 105 cells/well, 96-well plates) were plated in 5% FBS-DMEM/F-12; 24 h later serum-containing medium was removed, and two rinses with serum-free DMEM/F-12 were performed. The cells were then overlaid with neutralized collagen I (35 µl/well), which was allowed to gel. Freshly isolated epithelial cells (1 x 105/well) were suspended in neutralized collagen I (2 mg/ml, 75 µl/well), plated on top of the fibroblast layer, and allowed to gel for 30 min at 37 C. Preparation of collagen gel was according to the manufacturer’s instructions. All serum-free cultures were carried out in basal medium (BM): serum- and phenol red-free DMEM/F-12 supplemented with 0.1 mM nonessential amino acids (product 11140-050, Life Technologies, Inc., Grand Island, NY), 2 mM L-glutamine, 0.1 µg/ml insulin, 1 mg/ml fatty acid-free BSA (fraction V), 100 µg/ml penicillin, and 50 µg/ml streptomycin. Cultures were kept in 5% CO2 at 37 C for up to 15 d, with medium changes every 2 d. Treatments with conditioned medium, growth factors, or hormones were added at the time of plating and included 50 ng/ml HGF, 25 ng/ml EGF, 100 ng/ml IGF-I, or 20 nM 17ß-estradiol (E2).
Preparation of conditioned medium
Freshly isolated mammary fibroblasts were plated in 100-mm plates in BM containing 5% FBS as described above and allowed to reach confluence. The cultures were then rinsed twice and cultured in serum-free BM, and conditioned medium was collected 48 h later. Conditioned medium was immediately centrifuged at 900 rpm for 5 min to remove cell debris, filtered through a 0.22-µm pore size filter, and kept at 4 C for up to 1 month. A 4-fold concentration of conditioned medium was obtained using a 3-kDa cut-off membrane (YM3, Millipore Corp., Marleborough, MA). Conditioned medium was diluted 1:1 with BM containing 0.2 µg/ml insulin before assay in epithelial cultures. Medium was changed every other day. Antibody neutralization of conditioned medium was performed at 37 C for 1 h before its addition to epithelial cell cultures. Antibodies were used at the following concentrations: anti-HGF, 1.2–4.8 µg/ml; anti-EGF, 4.75–9.8 µg/ml; and anti-IGF-I, 2.5–20 µg/ml; isotype-specific nonimmune IgG (goat or rabbit) were used as controls in antibody neutralization experiments
[3H]Thymidine incorporation assay and analysis of cell proliferation
After 3 d of treatment, cells were incubated with 0.1 µCi/well [3H]thymidine (SA, 50 Ci/mmol; ICN Pharmaceuticals, Inc., Irvine, CA) at 37 C for 6 h (15). Collagen gels containing epithelial cells were removed, dissolved with 3.5 mM acetic acid, and transferred to GF/C Whatman filters (Clifton, NJ), followed by washing with Hank’s Balanced Salt Solution, ice-cold 10% trichloroacetic acid, and 90% ethanol. Radioactivity was determined by liquid scintillation counting. Organoid sizes under various culture conditions were determined by computer-assisted morphometry. Digitized phase contrast images of organoids were captured using an inverted microscope at a magnification of x40, and the area per organoid was determined using NIH Image software as previously described (13).
Immunoblot analysis of HGF
Recombinant human HGF (Calbiochem; 25 ng) was separated on 8% SDS-PAGE under nonreducing conditions and transferred to a nitrocellulose membrane. The membrane was blocked overnight with 0.3% gelatin in NET buffer (0.5 M NaCl, 0.025 M EDTA (pH 8), and 0.1 M Tris (pH 7.5), and 0.05% Triton X-100) and probed for 1 h with anti-HGF antibody (dilution, 1:1000; Sigma). The membrane was washed three times for 15 min each time with NET buffer, incubated with horseradish peroxidase-labeled secondary antigoat antiserum at a dilution of 1:10,000 for 45 min, and washed three times with NET buffer for 15 min each time. All incubations and washes were performed at room temperature. Immune complexes were detected using a Super Signal Chemiluminescence Kit (Pierce Chemical Co., Rockford, IL) by mixing equal parts of the stable peroxide solution and the luminol/enhancer solution and incubating the blot in the solution for 5 min. The blot was then exposed to film to visualize protein bands.
ER immunohistochemistry
Mammary fibroblasts cultured on coverslips and epithelial cells cultured within collagen gels were fixed in 3.7% buffered formalin. Gels were prepared for paraffin sections, and coverslips were assayed directly after fixation. After antigen retrieval by boiling for 10 min in 10 mM citrate buffer (pH 6.0), samples were incubated with either rabbit antimouse ER antibody (1:800 dilution; Santa Cruz Biotechnology, Inc., Santa Cruz, CA) or goat antimouse ERß antibody (1:150 dilution; Santa Cruz Biotechnology, Inc.) overnight at 4 C, followed by incubation in appropriate rabbit antigoat or goat antirabbit immunoperoxidase-conjugated secondary antibody as previously described (17). Antibody binding was detected by the chromagen diaminobenzidine tetrahydrochloride.
Statistical analysis
Data were expressed as the mean ± SEM, and statistical significance was determined using t test or ANOVA.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). The appearance of the organoids did not change up to 15 d in culture (data not shown). In contrast, when cocultured with mammary fibroblasts, within 3 d the epithelial organoids produced numerous projections, characteristic of a tubular/ductal morphology (Fig. 1B). The tubular nature of the projections was confirmed by histological analysis of the cells within the gels (Fig. 1, E and F).
|
). This was accompanied by a 4.5-fold increase in organoid size [epithelial cells alone, 5099 ± 255 pixels/organoid (n = 50); epithelial cells cocultured with fibroblasts, 22,684 ± 1361 pixels/organoid (n = 55); P < 0.01]. The increase in organoid size was the result of an increase in size of the organoid body as well as the formation of tubules. Comparing organoid sizes in Fig. 1A
with those in Fig. 1, B–D, indicates that [3H]thymidine incorporation reflects an increase in cell number. Time-course studies showed that maximal [3H]thymidine incorporation occurred on d 3 and decreased thereafter through d 6 (data not shown).
|
).
Characterization of fibroblast-derived growth factor
It has been widely reported that cultured mammary fibroblasts synthesize and secrete HGF (7, 18, 19). HGF has also been shown to stimulate mammary epithelial cell proliferation and to induce tubular/ductal morphogenesis in collagen gels (7, 18, 19). Thus, HGF was a likely candidate for the growth factor present in our FCM. To confirm the identity of the growth factor, FCM was pretreated with neutralizing antibody to HGF and assayed for proliferative activity in epithelial cultures. The anti-HGF antibody completely blocked proliferative activity in FCM (Fig. 3A), whereas neither anti-EGF nor anti-IGF-I neutralizing antibodies caused a significant decrease in FCM activity (Fig. 3, B and C). The inhibitory effect of the anti-HGF antibody could be reversed by the addition of excess HGF to the culture medium (Fig. 4). The anti-HGF antibody was specific for HGF and did not decrease the proliferative activity of either EGF or IGF-I (Table 1). Furthermore, similar tubular/ductal morphology was induced in epithelial cells cultured alone, in the presence of HGF or FCM, or in coculture with mammary fibroblasts (Fig. 1).
|
|
|
). Attempts to demonstrate HGF in FCM by immunoblot analysis were not successful. This was most likely due to the fact that the highest concentration of HGF in conditioned medium (concentrated 4-fold = 0.23 ng/µl) was 7-fold lower than the minimum amount of HGF that could be detected by the antibody (1.7 ng/µl) by immunoblot. Attempts at further concentration of the conditioned medium were not successful due to the lability and poor recovery of HGF as measured by the loss of proliferative activity (data not shown).
Mammary fibroblasts mediate E-induced proliferation in epithelial cells
To determine whether mammary fibroblasts cultured in serum-free medium could mediate estrogenic effects in epithelial cells, their ER status was analyzed by immunocytochemistry using methods previously described (20). Cultured fibroblasts exhibited nuclear staining for ER (19.4 ± 0.7% positive cells; n = 511 cells), indicating that the estrogenic effects observed could be due to an ER-mediated mechanism. ER analysis also revealed positive nuclear staining for ERß (21.5 ± 0.05% positive cells; n = 572 cells). Epithelial cells cultured within collagen gels exhibited nuclear staining for ER (50.2 ± 5.3% positive cells; n = 563 cells) and ERß (62.3 ± 8.2% positive cells; n = 595 cells).
To determine whether HGF produced by mammary fibroblasts was under estrogenic regulation, serum-free conditioned medium was obtained from fibroblasts cultured separately in the absence (control FCM) or presence of E2 (E2FCM). These conditioned medium were then tested on epithelial cells cultured alone in collagen I gels. As mammary epithelial cells also contain ER, the antiestrogen ICI was added with E2FCM to epithelial cells. Thus, the effect of E2 observed was an effect of E2 on the fibroblasts. Figure 5 shows a representative experiment in which control FCM increased proliferation of mammary epithelial cells approximately 6-fold; E2FCM produced a significantly greater response and increased proliferation 10.5-fold. The proliferative activities of both FCM and E2FCM were consistently neutralized completely by anti-HGF antibody (Fig. 3A). Antibody titration of HGF activity in FCM and E2FCM against a known amount of HGF (50 ng/ml) showed that conditioned medium obtained from E2-treated fibroblasts (E2FCM) contained 1.7-fold more HGF (113 ng/ml) than conditioned medium obtained from control-treated fibroblasts (FCM; 67 ng/ml; Fig. 3A).
|
and ERß by immunocytochemical analysis (data not shown). Thus, it would be expected that E2 present in E2FCM should not have a direct effect on these cells. Figure 6 shows that proliferation in the presence of three separate preparations of E2FCM was 2-fold greater than control FCM, further confirming that E2 treatment increases HGF in conditioned medium obtained from mammary fibroblasts. It was possible that the Mv1Lu cells were responding to growth factors other than or in addition to HGF produced by mammary fibroblasts such as EGF and/or IGF-I. However, antibody neutralization experiments showed that pretreatment of FCM and E2FCM with antibody to HGF completely abolished the proliferative response of the Mv1Lu cells to FCM and E2FCM (data not shown).
|
shows that antiestrogen reduced the HGF activity in conditioned medium by 58% (ICI vs. FCM) and that subsequent treatment with exogenous E2 produced a 4.5-fold increase in HGF activity over that produced by ICI-treated fibroblasts (ICI
E2 vs. ICIBM).
|
, HGF, EGF, or IGF-I alone each increased epithelial cell proliferation. The morphological responses of the cultures to treatment with HGF vs. EGF were markedly different (Fig. 9). Cultures treated with EGF alone produced only a minimal tubulogenic response, whereas extensive branched tubule formation was observed with HGF. Furthermore, the morphological response to IGF-I was distinct from that obtained with either EGF or HGF (Fig. 9). In contrast to the pronounced tubulogenic/ductal response observed with HGF (Fig. 9) and coculture with fibroblasts (Fig. 1B), only cell spreading was observed in the presence of IGF-I.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and ERß, it was not possible to determine which ER isoform was responsible for this effect. HGF has been shown to be produced by human, rat, and mouse fibroblasts (7, 18, 19). HGF can be produced in high amounts by certain mouse fibroblast cell lines and has been demonstrated to be present in conditioned medium by immunoblot analysis (21). HGF has also been shown previously to be produced by mouse primary mammary fibroblasts (22). Demonstration of mouse mammary fibroblasts HGF by Sasaki et al. (22) was accomplished by SDS-PAGE analysis and silver staining, but required the presence of serum in the conditioned medium, a 20-fold concentration, and partial purification of 200 ml of the concentrated conditioned medium, attesting to the low level and/or lability and low recovery upon concentration of HGF present in the conditioned medium. Thus, it was not surprising that we could not demonstrate HGF by immunoblot analysis in our FCM. Previously, cloning of mouse HGF cDNA from cultured mammary fibroblasts was accomplished by RT-PCR using primers designed from the human and rat HGF cDNA sequences (22). The cloned mouse HGF expression plasmid was transfected to COS-1 cells, and conditioned medium from these cells stimulated proliferation similar to conditioned medium from mammary fibroblasts. The proliferative activity was completely abolished by anti-HGF antiserum.
A stimulatory effect of estrogen on HGF production by mammary fibroblasts cells has not been previously reported (18). The reasons for this are unclear. In our present study a basal level of HGF production by fibroblasts was observed after transfer of fibroblasts from serum containing to serum-free medium. This basal level of HGF activity was reduced 58% by treatment of fibroblasts with antiestrogen. Subsequent treatment with estrogen increased the HGF activity 4.5-fold over the HGF level in ICI-treated cultures. Thus, it is possible that an estrogen effect carried over from serum was responsible for the basal level of HGF, which obscured the actual extent of the stimulatory effect of exogenously added estrogen on fibroblast-derived HGF. The promoter region of the HGF gene has been reported to contain an ER response element, and HGF production in the ovary by estrogen is transcriptionally regulated (23). The mechanism of estrogen regulation of HGF production in mammary fibroblasts is under active investigation and remains to be determined.
Ribonuclease protection analysis of HGF expression in whole mouse mammary gland extracts during development has previously shown that low levels of HGF are present at 2 wk of age and increase 4-fold between 6 and 12 wk of age, coinciding with increased estrogen levels due to the onset of estrous cycles; this is also the period when ductal morphogenesis takes place (24). Based upon the results presented herein, we hypothesize that proliferation and ductal morphogenesis in the normal mammary gland are due to HGF produced in the mammary stroma and that estrogen increases proliferation by increasing HGF production by stromal cells. Thus, it is highly likely that the role of estrogen in ductal morphogenesis is mediated by HGF in vivo.
We have shown that the proliferative activity obtained from FCM and E2FCM was not reduced by neutralizing antibodies to either EGF or IGF-I, two other growth factors produced in the mammary gland and believed to play a role in the regulation of mammary epithelial proliferation. Furthermore, in the present study mammary epithelial cells exhibited proliferative and morphological responses to EGF or IGF-I that were distinct from those obtained with HGF or FCM. In this regard, previous studies of mouse mammary epithelial cell morphology in collagen gels in the presence of serum and HGF, under serum-free conditions with HGF, or in control cultures without HGF but in the presence of other growth factors plus high insulin concentrations and additional supplements have all demonstrated significant branching tubulogenesis (5, 6, 7, 8, 18, 19, 25). In contrast to these findings we observed extensive branching tubulogenesis only in the presence of HGF. This underscores the advantages of using serum-free culture conditions in which control-treated cultures do not exhibit significant proliferation or morphogenesis. To delineate the specific effects of individual growth factors and/or mammogenic hormones, it is necessary to identify and reduce the confounding variables.
In summary, we have demonstrated in vitro, under serum-free conditions, that an indirect effect of estrogen on mammary epithelial cell proliferation is mediated by HGF produced in mammary fibroblasts. Furthermore, HGF production by fibroblasts is increased by estrogen. These studies provide a conceptual basis for understanding the previous observation that ductal proliferation in the mammary gland requires the presence of ER in stromal cells and not in the epithelium (3). These results also demonstrate that other growth factors produced by the mammary gland, such as EGF and IGF-I, are not likely to be merely interchangeable or redundant with regard to proliferation and morphogenesis occurring in vivo. Our results support the concept that the temporal and cell type-specific expression of individual growth factors in coordination with steroid hormone receptor expression and function are required for the developmental stage-specific process of ductal morphogenesis. Advancing our understanding of the complexities of growth regulation by estrogen in relation to epithelial-stromal cell interactions in the normal mammary gland may provide important information about the nature of changes in hormone-regulated growth that occur in mammary cancer. Furthermore, identification of the specific responses of epithelial vs. stromal cells to steroid hormones may lead to new therapeutic strategies in breast cancer treatment.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: BM, Basal medium; E2, 17ß-estradiol; E2FCM, FCM obtained from fibroblasts cultured in the presence of 20 nM estradiol; EGF, epidermal growth factor; ER, estrogen receptor; FBS, fetal bovine serum; FCM, fibroblast-conditioned medium; HGF, hepatocyte growth factor; ICI, ICI 182,780.
Received January 7, 2002.
Accepted for publication May 29, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Z. Haslam, A. Drolet, K. Smith, M. Tan, and M. Aupperlee Progestin-Regulated Luminal Cell and Myoepithelial Cell-Specific Responses in Mammary Organoid Culture Endocrinology, May 1, 2008; 149(5): 2098 - 2107. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. R. Wira, R. M. Rossoll, and R. C. Young Polarized Uterine Epithelial Cells Preferentially Present Antigen at the Basolateral Surface: Role of Stromal Cells in Regulating Class II-Mediated Epithelial Cell Antigen Presentation J. Immunol., August 1, 2005; 175(3): 1795 - 1804. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. F. Koehler, L. A. Helguero, L.-A. Haldosen, M. Warner, and J.-A. Gustafsson Reflections on the Discovery and Significance of Estrogen Receptor {beta} Endocr. Rev., May 1, 2005; 26(3): 465 - 478. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. S. Grant-Tschudy and C. R. Wira Hepatocyte Growth Factor Regulation of Uterine Epithelial Cell Transepithelial Resistance and Tumor Necrosis Factor {alpha} Release in Culture Biol Reprod, April 1, 2005; 72(4): 814 - 821. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Parrinello, J.-P. Coppe, A. Krtolica, and J. Campisi Stromal-epithelial interactions in aging and cancer: senescent fibroblasts alter epithelial cell differentiation J. Cell Sci., February 1, 2005; 118(3): 485 - 496. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Parmar and G. R Cunha Epithelial-stromal interactions in the mouse and human mammary gland in vivo Endocr. Relat. Cancer, September 1, 2004; 11(3): 437 - 458. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. S. Grant and C. R. Wira Effect of Mouse Uterine Stromal Cells on Epithelial Cell Transepithelial Resistance (TER) and TNF{alpha} and TGF{beta} Release in Culture Biol Reprod, September 1, 2003; 69(3): 1091 - 1098. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
) or FCM obtained from fibroblasts treated with 20 nM E2 for 48 h (E2FCM; 
) and pretreated with the indicated concentrations of neutralizing antibody to HGF (A), EGF (B), or IGF-I (C). The data are presented as a percentage of the control [3H]thymidine incorporation; 100% represents [3H]thymidine incorporation in the absence of antibody for each growth factor, E2FCM, or FCM. For each antibody, E2FCM was also treated with the indicated concentrations of isotype-specific nonimmune IgG (
; HGF and IGF-I, goat IgG; EGF, rabbit IgG). The neutralizing ability of each antibody was determined by titration of a known amount of growth factor with its specific antibody (
; HGF, 50 ng/ml; EGF, 50 ng/ml; IGF-I, 100 ng/ml). D, Immunoblot of rhHGF with anti-HGF antibody. Twenty-five nanograms of rhHGF were separated on an 8% SDS-PAGE under nonreducing conditions and transferred to a nitrocellulose membrane. The membrane was probed with anti-HGF antibody (dilution, 1:1000), followed by incubation with horseradish peroxidase-labeled secondary antigoat antibody (dilution, 1:10,000). Immune complexes were detected by chemiluminescence.
