Functional interaction between mouse erbB3 and wild-type rat c-neu in transgenic mouse mammary tumor cells

Human breast cancer cell lines SKBR-3 and BT-474 were obtained from the American Type Culture Collection (Rockville, MD, USA) and maintained in DMEM and Ham's F-12 medium (1:1, v/v) (Invitrogen Corp, Grand Island, NY, USA) supplemented with 10% FBS (Invitrogen Corp). These cell lines were cultured in a 37°C humidified atmosphere containing 95% air and 5% CO₂ and were split twice a week. These human breast cancer cells were used primarily as controls.

Mammary tumors were obtained from the transgenic mice by surgical removal immediately following euthanasia, according to our approved IACUC protocol. The histological pattern and tumor diagnoses were confirmed by microscopic analysis. These methods have been previously described in detail [35], although the specific cell lines described in this work have not been previously published. In brief, solid tumor tissue was transferred into a tissue culture dish containing PBS. After removal of mammary fat and connective tissues, tumors were minced into small pieces and treated with 0.25% trypsin-EDTA (Invitrogen Corp) at 37°C for 30 min. Cells were subsequently centrifuged at 1,200 rpm for 5 min. After discarding supernatant, cells were suspended in DMEM/F12 medium supplemented with 10% FBS and 1% antibiotics and antimycotics (Invitrogen Corp). These mammary tumor cells (~1.0 × 10⁶ cells/plate) were seeded in tissue culture dishes and kept in a 37°C humidified atmosphere containing 95% air and 5% CO₂. The media was changed twice a week to maintain cells in culture. Each line was passaged approximately 20 times before stability was assumed.

Soft agar cloning was performed as described previously [35] with some modification. The bottom agar was prepared with a mixture of 1.6 ml of 1 × DMEM/F12 (complete medium), 3.2 ml of 2 × DMEM/F12 (complete medium), and 3.2 ml of 1.25% Noble agar (Sigma Co, St Louis, MO, USA) and maintained at 42°C. From this, 2 ml was pipetted into each well of six-well cell culture plates and agar was allowed harden in the hood. For each well, top agar was a mixture of 0.2 ml of 1 × DMEM/F12, 0.4 ml of 2 × DMEM/F12, and 0.4 ml of 0.95% Noble agar. Five thousand cells (in 80 μl complete medium) were added into the top agar mixture. After vortexing gently, the cell containing top agar was added in a drop-wise fashion onto the bottom agar containing six-well plates (in triplicate per cell line). After resting for 10 min in the hood, the six-well plates were cultured in a 37°C incubator for 3 weeks. Colony counts were obtained under an inverted microscope, from three wells per cell line counting all colonies >50 μM in diameter.

Measurement of cell growth rate in culture was determined using sulforhodamine B (SRB; Sigma Co) assays as previously described [35]. Two thousand cells were seeded into each well of a six-well plate with complete medium. Cells were fixed with 50% trichloroacetic acid (TCA) at 24 h intervals for 3 days. TCA-fixed cells were then stained with 0.4% SRB for 30 min followed by three washes. Protein-bound dye was dissolved in 10 mM Tris base. Plates were read at 565 nM using a micro-plate reader. Cell-doubling time was calculated based on proliferation curves that resulted from changes in SRB absorbance over time. Data represent the means of at least three independent experiments.

A CellTiter96™ AQ non-radioactive cell proliferation kit (Promega Corp, Madison, WI, USA) was used to determine the responsiveness of cells to various growth factors. Cells were plated onto 96-well plates, 5,000 cells/well for each cell line. Twenty-four hours later, the culture media were replaced by 0.5% FBS in DMEM/F12 fresh medium or the same medium containing 25 ng/ml HRG (R&D Systems, Inc, Minneapolis, MN, USA), 10 ng/ml EGF (Sigma Chemical Co), or 40 ng/ml IGF-1 (R&D Systems, Inc) for another 72 h incubation with 5% CO₂ at 37°C. After reading at 490 nM with the micro-plate reader, the percentages of viable cells were determined by reduction of MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium; inner salt) relative to controls. Data reflect the means of at least three independent experiments.

RT-PCR analyses were performed as previously described [37]. The primers specific for rat neu were synthesized according to the literature [38]. Forward primer AB2913, 5'-CGG AAC CCA CAT CAG GCC-3' and reverse primer AB1310, 5'-TTT CCT GCA GCA GCC TAC GC-3' amplify the region corresponding to nucleotides 1492 to 2117 of rat neu cDNA [38]. The PCR products purified from agarose gel using QIAquick Gel Extraction Kit (Qiagen, Inc, San Valencia, CA, USA) were submitted to the core facility at the Oklahoma Medical Research Foundation for direct sequencing.

Immunohistochemical staining of mammary tumor tissues was performed as previously described [39]. Briefly, after deparaffinization and rehydration, tissue sections were steamed in a 10 mM citrate buffer, pH 6.0, for 30 min. Non-specific reactivity was blocked with 0.3% H₂O₂ in buffer. For erbB3 immunoassays, CAS Block (Zymed Laboratories, Inc, South San Francisco, CA, USA) and 10% normal horse serum (Vector Laboratories, Inc, Burlingame, CA, USA) were used sequentially. For phospho-Akt immunostaining, we used 1% H₂O₂ and 5% normal goat serum (Vector Laboratories, Inc) sequentially. Primary antibodies included an anti-erbB2 (reactive with rat c-neu/erbB2 rabbit polyclonal, dilution 1:1000 (Dako, Carpinteria, CA, USA) for 2 h incubation at room temperature), anti-erbB3 (cross-reacts with mouse and human, mouse monoclonal, dilution 1:50 (NeoMarkers, Inc, Fremont, CA, USA), overnight incubation at 4°C), anti-phospho-Akt (rabbit polyclonal, diluted in 5% normal goat serum 1:12.5 (Cell Signaling Technology, Beverly, MA, USA), overnight at 4°C), or anti-phospho-MAPK (E10 monoclonal antibody, diluted in 5% normal goat serum 1:25 (Cell Signaling Technology), overnight at 4°C). After multiple washes with buffer, tissue sections were sequentially incubated for 30 min at room temperature with diluted biotinylated secondary antibody (1:500; Dako) and VECTASTAIN Elite ABC reagent (Vector Laboratories, Inc) diluted in PBS. After reaction with diaminobenzidine (Dako) and counterstaining with hematoxylin, tumors were individually examined. Each slide was evaluated in its entirety for antigen expression, cell type and histopathological diagnoses.

Immunoprecipitation and Western blot assays were performed as previously described [40]. Briefly, cells were lysed in NP-40 lysis buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.5% NP-40, 50 mM NaF, 1 mM Na₃VO₄, 1 mM phenylmethylsulfonyl fluoride, 25 μg/ml leupeptin, 25 μg/ml aprotinin). The supernatants were cleared by centrifugation. Protein concentrations were measured using the Coomassie plus protein assay reagent (Pierce Chemical Co, Rockford, IL, USA). Total cell lysates containing 200 μg of protein were subjected to immunoprecipitation in the presence of 1 μg anti-erbB2 antibody (mouse monoclonal antibody, Ab-4; Oncogene Science Products, Cambridge, MA, USA) for 2 h at 4°C, followed by incubation with immobilized protein A-agarose (Roche Diagnostics Corp, Indianapolis, IN, USA) at 4°C overnight with rotation. For Western blot analyses, the immunoprecipitates or equal amounts of crude extracts were boiled in Laemmli SDS-sample buffer, resolved by SDS-polyacrylamide gel electrophoresis, transferred to nitrocellulose (Bio-Rad Laboratories, Hercules, CA, USA), and probed with different primary antibodies. After the blots were incubated for another 1 h at room temperature with horseradish peroxidase-labeled secondary antibody (goat anti-rabbit IgG or goat anti-mouse IgG; Perkin Elmer, Boston, MA, USA), the signals were detected using the Enhanced Chemiluminance assay (Amersham Life Science Inc., Arlington Heights, IL, USA) according to the manufacturer's instructions.

Western blot analyses were used to determine erbB2 and erbB3 protein expression in tumor-derived cell lines (and the control SKBR-3 human breast cancer cell line). The majority of tumor-derived cell lines expressed moderate to high levels of both erbB3 and erbB2 (Fig. 1). In general, lines with the highest erbB2 expression showed the highest levels of erbB3 protein. Tyrosine phosphorylation (activation) of these receptors was examined by Western blots using antibodies specific for phophorylated erbB2 (P-erbB2) or phosphorylated erbB3 (P-erbB3). Tumor lines with co-overexpression of both proteins showed higher P-erbB2 and P-erbB3 levels (Fig. 1). The intensity of P-erbB2 and P-erbB3 signals did not necessarily correlate with their corresponding protein levels. The expression of either receptor protein was undetectable in only one of our novel, derived tumor cell lines (78423). AIB-1 (also called SRC3, RAC3, ACTR and p/CIP), a co-activator of estrogen receptor commonly amplified in breast cancer cells [41], was used as a loading control. Expression of AIB-1 further established the origin of these cells as mammary-derived.

To confirm the transformed characteristics of these lines, soft agar cloning assays (which quantitate anchorage-independent cloning capability) were used. All six tumor-derived cell lines formed colonies in soft agar. Colony formation was variable when comparing one cell line with another (range 17–180, Table 1). There was no correlation between the ability of a cell line to form anchorage-independent clones and the expression levels of erbB2 or erbB3.

Immunohistochemical methods were used to visualize RTK expression and downstream signaling (protein activation) by tumors in situ. Tumors showed strong and typically diffuse co-expression of both erbB2 and erbB3. The only exception to this was the mammary tumor 78423 R1, the progenitor of the cell line that did not co-express erbB2 and erbB3 discussed above. We also studied RTK signaling activation in situ, using phosphospecific antibodies. Phosphorylated-Akt (P-Akt) showed cytoplasmic and membranous staining, which was less diffuse than the erbB-2 expression. Phosphorylated-MAPK (P-MAPK) was the most selectively expressed, typically expressed by clustered or isolated tumor cells as shown in Fig. 2 (left panel) with tumor 78617 R3. The majority of tumor cells from 78423 R1 were erbB3 negative, although some cells showed weak erbB2 protein expression. In this later tumor, P-Akt staining was weak with clustered or isolated tumor cells and no staining for P-MAPK was observed (Fig. 2, right panel). The histological, cytological and biological features of these tumors have been reported elsewhere [36, 42]. As a control, we also studied cytokeratin expression and all tumors were positive. This confirmed the epithelial nature of these tumors (data not shown).

As alluded to above, in-frame deletions of 7–12 amino acids have been reported in the extracellular region of the transgene, proximal to the transmembrane domain [38]. To study the mutational status of tumor-derived lines, we performed RT-PCR amplification of exactly the same region followed by direct sequencing analysis. The PCR primers used were specific for rat neu and were designed to amplify the 603 bp extracellular region [38]. Of six tumor-derived cell lines used in this manuscript and therefore studied for mutation, only four showed PCR gene amplification (Fig. 3a). Of these, the strongest PCR signal was seen in 85819 cells. These data are consistent with our Western blot results that showed overexpression of the rat neu/erbB2 in only the four PCR-positive lines (Fig. 1). Direct sequencing of the PCR products revealed no deletion mutations in the amplified product. Sequencing showed three of the four were wt rat neu cDNA sequence. Sequencing data from the 83923 cells indicated a mixture of two kinds of neu cDNA. Using a reverse primer, we verified that both wt and point mutation neu transcripts co-existed in 83923 cells (Fig. 3b). This suggests biclonal populations or a heterozygous mutation. Further studies and sub-cloning are in process.

To study the functionality and interactions of the erbB receptors, 78423 and other three representative mouse mammary tumor-derived lines with the highest expression of wt erbB2 and co-expression of erbB3 were chosen for further study. Baseline proliferation was determined using monolayer culture conditions and the SRB assay (Fig. 4a). Some variability in the basal doubling time was observed between these cell lines. The mouse mammary tumor cell lines 78423, 78617, 85815 and 85819 showed population doubling times of 15.15 ± 1.10, 16.25 ± 1.40, 30.85 ± 2.31 and 20.35 ± 1.89 h, respectively. Using an MTS assay, we then tested the response of these lines to EGF, HRG and insulin-like growth factor (IGF)-1 (Fig. 4b). HRG strongly stimulated the proliferation of three of the four mouse mammary tumor cell lines (78617, 85815, 85819) with overexpression of both erbB2 and erbB3. Proliferation was not induced by EGF or IGF-1, which bind to EGFR and IGF-1 receptor, respectively. HRG also promoted the growth of SKBR-3 and BT-474 human breast cancer cells (controls). These data strongly support a functional interaction between the wt-rat neu/ErbB2 and endogenous mouse erbB3.

It is well documented that the MEK/MAPK and PI-3K/Akt pathways are the two major signal transduction pathways downstream of the erbB receptors [11, 17‐19]. To determine which signaling pathways were activated in the mouse-derived mammary tumor cells exposed to HRG, we performed Western blots to detect P-MAPK or P-Akt. With 2 h of HRG treatment, both P-Akt and P-MAPK increased in the 85815 and 85819 mouse mammary tumor cell lines (Fig. 5a). This study included a series of HRG concentrations, and stimulation was maximal at a concentration of 2.5 ng/ml. Next, we performed a time-course analysis to further verify these results. HRG stimulated both Akt and MAPK in 85815 and 85819 cells, whereas it had no effect on Akt or MAPK activation in the 78423 cells (Fig. 5b). These data were consistent with the results of minimal stimulation by HRG in this cell line (Fig. 4b). In aggregate, these data suggest that HRG induces activation of both MEK/MAPK and PI-3K/Akt signaling transduction pathways in mammary tumor cells with elevated expression levels of both the transgene rat c-neu/ErbB2 and the endogenous mouse ErbB3 gene. This activation was both dose- and time-dependent.

To study cross-species functional interactions between the rat c-neu/ErbB2 transgene and mouse ErbB3, we evaluated tumor and tissue expression in vivo, ligand-associated interactions, and signaling in vitro. Immunohistochemical studies showed cytoplasmic P-Akt and P-MAPK expression in tumor cells with erbB2 and erbB3 co-expression, predominantly a perivascular distribution. In rare tumors without erbB2 and erbB3 expression (e.g. 78423 R1), the perivascular distribution was not identified and only rare cells showed immunoreactivity. This evidence of perivascular pathway activation suggests that ligand-associated signaling via erbB3 may be involved. Ligand-associated signaling probably provides enhanced growth or pro-tumorigenic signaling, in addition to ligand-independent, transgene activation. Our data, and those from others showing frequent erbB3 upregulation in transgenic mice bearing activated neu/ErbB2, suggest that the concomitant upregulation of erbB3 and ligand-associated signaling may be an important additional factor in both wt and activated neu/ErbB2-associated mammary tumor development. To further define the role of HRG (ligand)-associated signaling, we utilized derived cell lines and specific inhibitors in vitro. The PI-3K inhibitor LY294002 was significantly more potent than the MEK inhibitor PD98059 in blocking the stimulatory effects of HRG (Fig. 7). Hence, while the MEK/MAPK and PI-3K/Akt signaling cascades both contribute HRG induced proliferation, the PI-3K/Akt pathway appears to provide the dominant response.

The erbB2/erbB3 complex is believed to be the most biologically active erbB heterodimer [43, 44], with potent activation of the downstream signaling cascade [13, 14]. Since both erbB2 and erbB3 were highly expressed by our mammary tumor cell lines and HRG-promoted tumor cell proliferation, we sought physical evidence that the wt-rat-neu/ErbB2 could form a complex with the endogenous mouse erbB3. Immunoprecipitation of erbB2, followed by Western blot analysis for erbB2 and erbB3 (Fig. 8a) showed a low level of complex formation between these receptors in untreated cell lines. HRG treatment significantly increased the physical interaction between the rat transgene and mouse erbB3 in two out of three cell lines. The antibody we used for immunoprecipitation (c-neu Ab-4) appeared to be wt-rat-neu/ErbB2-specific, because human erbB2 was not immunoprecipitated from SKBR-3 cell lysates (Fig. 8a, upper panel), although it was expressed by SKBR-3 cells (Fig. 8b, Western blot analysis with c-neu Ab-3). HRG treatment (for 2 h) did not increase the total protein levels of erbB2 or erbB3 as compared with untreated cell lines (Fig. 8b).