erbB-2 Overexpression in Human Mammary Epithelial Cells Confers

7 downloads 0 Views 611KB Size Report
HME cells, the spontaneously immortalized MCF-10A cell line and the HPV-16-immortalized H16N2 cell line were infected with the bicistronic retroviral vector ...
0013-7227/99/$03.00/0 Endocrinology Copyright © 1999 by The Endocrine Society

Vol. 140, No. 8 Printed in U.S.A.

erbB-2 Overexpression in Human Mammary Epithelial Cells Confers Growth Factor Independence* KATHLEEN M. WOODS IGNATOSKI, ALLISON J. LAPOINTE, ERIC H. RADANY, AND STEPHEN P. ETHIER Department of Radiation Oncology, Division of Radiation and Cancer Biology, University of Michigan Medical School, University of Michigan Comprehensive Cancer Center, Ann Arbor, Michigan 48109-0948 ABSTRACT Previously, we demonstrated that human breast cancer cells with progressively elevated levels of constitutively tyrosine phosphorylated erbB-2 are independent of growth factors required by normal human mammary epithelial (HME) cells for proliferation in serumfree medium. To determine whether erbB-2 overexpression alone is sufficient to confer the growth factor-independence phenotype in HME cells, the spontaneously immortalized MCF-10A cell line and the HPV-16-immortalized H16N2 cell line were infected with the bicistronic retroviral vector pTPerbB-2 and tested for their ability to grow in the absence of specific factors. Selection of infected cells in G418-containing medium resulted in moderate levels of erbB-2 overexpression in approximately 40% of cells. The subpopulation of erbB-2 overexpressing cells could be selected for by culturing the cells in medium devoid of insulin. When MCF-10A or H16N2 cells were infected with pTPerbB-2 and directly selected in growth factor-deficient medium over long periods of time, populations of both cell lines emerged that expressed levels of erbB-2 protein equivalent to levels

T

HE erbB-2 GENE is among the most well-characterized human breast cancer oncogenes. Since the discovery of its amplification in human breast cancer (1), numerous papers have documented erbB-2 as a prognostic factor in nodepositive and node-negative disease (2–11). Additionally, much has been learned about erbB-2 as a signal transduction molecule, including its interactions with other members of the erbB family of growth factor receptors (12–14), its activation by the family of growth factors known as heregulins (HRGs) (15–24), and its ability to act as a breast cancer oncogene in transgenic mice (25, 26). This body of work makes the erbB-2 oncogene among the most well-characterized of human cancer genes and indicates clearly its causal role in the pathogenesis of breast and other human neoplasms. Work in our laboratory over the past several years has focused on understanding the physiological consequences of erbB-2 overexpression in human breast epithelial cells. In particular, we have attempted to understand which altered growth phenotypes of human breast cancer cells can be specifically mediated by erbB-2 when it is overexpressed and activated. To approach this question, we first studied the Received October 1, 1998. Address all correspondence and requests for reprints to: Stephen P. Ethier, Ph.D., 1500 East Medical Center Drive, Ann Arbor, Michigan 48109-0948. E-mail: [email protected]. * This work was supported by: 1) NIH Grant CA-70354; and 2) Cancer Biology Training Grant T32-CAO-9676 (to K.M.W.I.).

expressed by breast cancer cells with an erbB-2 gene amplification. Furthermore, overexpressed p185erbB-2 was constitutively tyrosine phosphorylated in these cells. The levels of tyrosine phosphorylated p185erbB-2 differed in the two recipient lines, with H16N2-erbB-2 cells having higher levels of activated receptor than MCF-10AerbB-2 cells. Furthermore, only the H16N2-erbB-2 cells were independent of both insulin and epidermal growth factor for growth in serum-free medium. Overexpression of erbB-2 also resulted in progressively increasing levels of tyrosine-phorphorylated erbB-3, without any significant changes in p180erbB-3 levels. These studies demonstrate a direct relationship between the level of expression and activation of p185erbB-2 and the requirements of HME cells for insulin-like and epidermal growth factor-like growth factors. The results also suggest that genetic alterations present in breast cancer cells, or mediated by HPV-16-induced alterations in pRb and p53, can influence the expression level and activation status of erbB-2 as well as erbB-3 and, in turn, their degree of growth factor independence. (Endocrinology 140: 3615–3622, 1999)

biological activity of the HRGs, which are potent activators of erbB-2. It is now known that HRG does not bind directly to erbB-2 but, rather, activates erbB-2 when present as a heterodimer with erbB-3 and, to a lesser extent, with epidermal growth factor receptor (EGFR) and erbB-4 (19, 20, 24, 27, 28). We demonstrated that HRG has the unique property of functioning as both an insulin-like growth factor (IGF)-like and EGF-like mitogen in mammary epithelial cells (21). Thus, whereas normal human mammary epithelial (HME) cells have a strict dependence on the synergistic interaction between EGF and IGF-I for growth under serum-free conditions (29), HRG (particularly HRG-b) can substitute for either growth factor to stimulate rapid growth of HME cells. This property is not shared by any other members of the EGF or IGF family of growth factors. It seems that activation of erbB-2/erbB-3 heterodimers is responsible for the dual specific nature of the HRG response. Docking molecules that bind activated erbB-2 stimulate the mitogen-activated protein (MAP) kinase pathway, thereby sending an EGFR-like signal (23); whereas, tyrosine phosphorylated erbB-3 is a potent activator of the phosphatidyl inositol (PI) 39-kinase pathway (30 –32), which is activated by the IGF-I receptor under normal conditions (33–38). Based on these observations, we hypothesized that human breast cancer cells with amplified and overexpressed erbB-2 would become independent of IGF-I and EGF for growth under serum-free conditions. Experiments using the 21T se-

3615

3616

erbB-2 AFFECTS GROWTH FACTOR INDEPENDENCE

ries of human breast cancer cell lines were consistent with this hypothesis. 21MT-2 cells express intermediate erbB-2 levels and are IGF-I independent for growth, whereas 21MT1 cells, which express very high levels of activated erbB-2, are independent of both IGF-I and EGF for continuous growth in serum-free medium (39). The growth factor independence expressed by the 21T human breast cancer cells was consistent with predictions made from the biological activity of erbB-2/3 after HRG stimulation. However, these breast cancer cells contain many genetic alterations that could cooperate with, or act independently of, erbB-2 to mediate these altered growth phenotypes. Therefore, the purpose of the present studies was to determine whether transduction of human erbB-2 into two normal HME cell lines, which were immortalized by different mechanisms, results in growth factor-independent proliferation when the gene is expressed at levels comparable with breast cancer cells with an erbB-2 gene amplification. The data indicate that transduction of erbB-2, followed by long-term selection in growth factor-deficient medium, results in the emergence of growth factor-independent cells that dramatically overexpress p185erbB-2. However, the degree of constitutive activation of p185erbB-2 and the extent of growth factor independence in the two cell lines differed. The H16N2 cell line, which was immortalized using the HPV-16 genome, expressed higher levels of tyrosine phosphorylated p185erbB-2 than MCF-10A cells, which were spontaneously immortalized in culture. In addition, H16N2-erbB-2 cells were independent of both insulin and EGF, whereas MCF10A-erbB-2 cells maintained their EGF-dependency for growth. Thus, overexpression and constitutive activation of erbB-2 can, by itself, induce multiple growth factor-independent proliferation of normal HME cells. However, cellular factors, possibly involving HPV-16-mediated disruption of the Rb and p53 pathways, influence the levels of constitutive receptor activation and, in turn, the degree of growth factor independence of the cells.

Endo • 1999 Vol 140 • No 8

X-100, 0.5% Na1deoxycholate, and 0.1% SDS). Protein concentrations were equalized using the Løwry method (Bradford, 1976 no. 878). Laemmeli sample buffer (Sambrook, 1989 no. 42) was added, and the samples were boiled. Equal amounts of protein were separated in 7.5% SDS-PAGE. Proteins were blotted to polyvinylidene diflouride (PVDF) membrane and probed with either a-erbB-2 (clone 9.3), a-Ptyr4G10 (catalog no. 05–321, Upstate Biotechnology, Inc., Lake Placid, NY), or a-erbB-3 (catalog no. PC27, Oncogene Research Products, Cambridge, MA) antibodies.

Flow cytometry analysis Infected cells were washed three times in PBS, and 1 mg/ml antierbB-2 antibody (clone 254) was incubated with the cells for 1 h at room temperature. Cells were then incubated with secondary antibody for 1 h at room temperature, then with fluorescein avidin DCS in 0.1 m NaHCO3 buffered saline (pH 8.2) for 30 min at room temperature. Cells were removed from the tissue culture dish with 10 mm EDTA and fixed with 95% ethanol. Cells that were sterily sorted were not fixed.

Immunocytochemistry Rinsed cells were fixed in 100% methanol at 220 C for 20 min. Cells were incubated with 1 mg/ml erbB-2 antibody (clone 254) for 30 min at room temperature, then with biotintylated secondary antibody for 30 min at room temperature. Cell-surface erbB-2 was visualized using the Vectastain ABC kit (catalog no. PK-4000, Vector Laboratories, Inc., Burlingame, CA).

Southern blot Ten micrograms of genomic DNA was digested with KpnI. The restriction fragments were separated on an agarose gel, transferred by standard methods to nylon membrane, probed with radioactive erbB-2 complementary DNA (cDNA), quantitated using a Molecular Dynamics Storm phosphorimager (Molecular Dynamics, Inc., Sunnyvale, CA), and visualized by autoradiography.

Northern blot Ten micrograms of total cellular RNA was separated on a formaldehyde-containing agarose gel and transferred to nylon membrane by standard methods. Blots were probed with radioactive erbB-2 cDNA and visualized by autoradiography.

Immunoprecipitation Materials and Methods Cell culture The medium for MCF-10A and H16N2 cells was Ham’s F12 supplemented with 0.1% BSA, 0.5 mg/ml fungizone, 5 mg/ml gentamycin, 5 mm ethanolamine, 10 mm HEPES, 5 mg/ml transferrin, 10 mm T3, 50 mm selenium, 5 mg/ml insulin, 1 mg/ml hydrocortisone, and 10 ng/ml EGF. MCF-10A cells were grown on collagen-coated tissue culture plates.

erbB-2 cloning and virus preparation The full-length human erbB-2 gene was subcloned in the XhoI site in the pTP2000 retroviral vector. Five micrograms of DNA was transfected in ØNX-A cells. Forty-eight hours later, conditioned medium was collected, cell debris was pelleted, the supernatant was passed through a 0.45-mm syringe filter, and the virus was frozen at 280 C. Virus was thawed and allowed to infect MCF-10A and H16N2 cells in the presence of polybrene for 24 h.

Membrane preparations and protein blots Cells were dounced in a buffer consisting of 20 mm HEPES (pH 7.4), 5 mm Na3VO4, 1 mm phenylmethylsulfonylflouride, and 10 mm pyrophosphate. Membrane fractions were retrieved by centrifuging at 100,000 3 g for 35 min at 4 C. Membrane fractions were resuspended in lysis buffer (10 mm sodium phosphate (pH 7.5), 100 mm NaCl, 1% Triton

Cells were lysed in a buffer containing 20 mm TriszHCl (pH 8.0), 137 mm NaCl, 1% NP-40, 10% glycerol, 1 mm Na3VO4, 1 mm phenylmethylsulfonylflouride, 1% aprotinin, and 20 mg/ml leupeptin. Protein concentrations were equalized using the Løwry method (Bradford, 1976 no. 878). Laemmeli sample buffer (Sambrook, 1989 no. 42) was added, and the samples were boiled. Equal amounts of protein were separated in 7.5% SDS-PAGE. Proteins were blotted to PVDF membrane and probed with a-Ptyr4G10 (catalog no. 05–321, Upstate Biotechnology, Inc.).

Results Bicistronic retroviral vectors for erbB-2 transduction in HME cells

To test the hypothesis that erbB-2 overexpression in HME cells results in growth factor-independent proliferation, a bicistronic retroviral expression vector was used to transduce erbB-2 into HME cells. This vector yields high-efficiency gene transfer, allows for selection in G418-containing medium, and efficiently coexpresses erbB-2 after selection with G418. We found, in preliminary studies, that the use of plasmid vectors or conventional retroviral vectors resulted in efficient selection of G418-resistant colonies, but that transduced mammary epithelial cells rarely coexpressed the trans-

erbB-2 AFFECTS GROWTH FACTOR INDEPENDENCE

gene. Thus, to overcome this problem, we subcloned fulllength erbB-2 into the bicistronic retroviral expression vector (pTP2000), which was recently developed in the Radany laboratory and has been recently described (40). In this construct, the transgene is upstream of neoR, and both genes are driven from the cytomegalovirus promoter, resulting in efficient expression of erbB-2. Figure 1A shows the general features of the bicistronic retroviral vector used in these studies.

FIG. 1. Expression of erbB-2 in infected MCF-10A cells. A, The fulllength human erbB-2 cDNA was subcloned into the XhoI site in pTP2000; note that the erbB-2 gene is expressed bicistronically with the antibiotic resistance gene via the BiP IRES sequence. B, One hundred micrograms of membrane protein, from erbB-2-infected or vector only-infected (PTP) MCF-10A cells selected in G418-containing medium, was separated on 7.5% SDS-PAGE, blotted to PVDF, and probed with a-erbB-2 antibody. C, Vector-only (PTP) or pTPerbB-2infected MCF-10A cells that were selected in G418-containing or in insulin-free medium containing G418 (bottom right panel) were subjected to immunocytochemistry for erbB-2. The cells in the bottom left panel were selected for two passages after infection, and the cells in the top right panel were selected for 15 passages after infection in G418-containing medium. The cells in the bottom right panel were selected in G418-containing medium for eight passages and then for seven passages in insulin-free medium containing G418. D, One hundred micrograms of membrane preparation from vector-only-infected and erbB-2-expressing cells selected in insulin-free medium containing G418, and varying amounts of membrane protein from 21MT1 and SUM-190PT, were separated on 7.5% SDS-PAGE, blotted to PVDF, and probed with an erbB-2 antibody.

3617

Expression of erbB-2 in retrovirally infected MCF-10A cells

To begin to examine the relationship between erbB-2 expression and growth factor independence in HME cells, MCF-10A HME cells were infected with pTPerbB-2 and selected in G418-containing medium. As reported by us previously, MCF-10A cells have a strict requirement for exogenous EGF and IGF-I for growth in serum-free medium (21). When MCF-10A cells were infected with pTPerbB-2 or pTPneo, and selected directly in G418-containing medium, the cells infected with the erbB-2-containing vector overexpressed p185erbB-2, relative to control cells, as determined by Western blot (Fig. 1B). Examination of the cells by flow cytometry (not shown) and immunocytochemistry demonstrated that, despite the use of the bicistronic vector, only about half of the MCF-10A cells stained positively for erbB-2 (Fig. 1C, left panels). Next, experiments were performed to determine whether the moderate level of erbB-2 overexpression observed in infected MCF-10A cells resulted in cell proliferation in the absence of individual growth factors. MCF-10A cells that had been selected in G418-containing medium with a full complement of growth factors were switched to insulin-free medium and grown to confluence. Although these cells grew slowly, erbB-2-transduced MCF-10A cells did grow continuously and were subculturable. Reanalysis of erbB-2 expression, after selection in insulin-free medium, demonstrated that 100% of the cells stained positively for erbB-2 (Fig. 1C, right panels). This selection for the erbB-2-expressing subpopulation resulted in higher apparent levels of erbB-2, as detected in Western blots, as compared with cells selected with G418 only (Fig. 1D). These results are consistent with the hypothesis that erbB-2 overexpression directly mediates IGFindependent proliferation of HME cells. However, MCF-10A cells expressing these moderate levels of p185erbB-2 were unable to grow in EGF-free medium. To compare the erbB-2 expression levels of the MCF-10A cells described above, with breast cancer cells having an erbB-2 gene amplification, Western blot analyses were carried out using varying amounts of membrane protein obtained from two breast cancer cell lines with an erbB-2 amplification. The 21MT-1 cell line was developed by Band and co-workers (41), and the SUM-190 cell line was recently developed in our laboratory. The data in Fig. 1D show that, despite the erbB-2 overexpression exhibited by the transduced MCF-10A cells (relative to parental cells), the overall levels of erbB-2 protein in these cells was dramatically lower than that observed in the two breast cancer cell lines with an erbB-2 gene amplification. Expression of erbB-2 in retrovirally infected H16N2 cells

Similar experiments were then performed using a different immortalized HME cell line. H16N2 cells were derived by HPV-16 infection of normal HME cell cultures (41). These immortalized cells express luminal cytokeratins and are EGF-dependent for growth in serum-free medium but are only marginally dependent on exogenous insulin. The relative IGF-independence of these cells is likely the result of their immortalization with HPV-16. The data in Fig. 2 show that, as in MCF-10A cells, infection of H16N2 cells with

3618

erbB-2 AFFECTS GROWTH FACTOR INDEPENDENCE

Endo • 1999 Vol 140 • No 8

pTPerbB-2 resulted in moderate overexpression of p185erbB-2 relative to controls. In an attempt to increase the number of cells infected with the viral vector, H16N2 cells were infected three times, over a 12-h period, with pTPerbB-2 followed by selection in G418-containing medium. These triply infected cells expressed marginally higher levels of p185erbB-2 than cells infected singly with the vector. As indicated above, H16N2 cells are relatively IGF-independent for growth in serum-free medium. Thus, growth in insulin-free medium could not be used to directly select p185erbB-2 overexpressing cells. Therefore, experiments were performed to determine whether flow sorting could be used to isolate a subpopulation of cells expressing high levels of p185erbB-2. Accordingly, H16N2 cells that had been infected with pTPerbB-2 and selected for growth in G418-containing medium were harvested and incubated with an erbB-2 monoclonal antibody. The cells were then analyzed by flow cytometry, and the cells expressing the highest levels of erbB-2 were collected and replated. Figure 2B shows the erbB-2 expression profile of the transduced H16N2 cells and indicates the fractions that were sorted. Figure 2C shows that expression of p185erbB-2 in the flow-sorted cells was higher than in the unsorted population of triply infected cells. However, these cells still expressed substantially lower levels of p185erbB-2 than breast cancer cells with an erbB-2 gene amplification. And, like MCF-10A cells described above, these H16N2 cells were also unable to grow in EGF-free medium. These experiments demonstrate that infection of two different HME cell lines with the pTPerbB-2 expression vector, followed by short-term selection in G418-containing medium, results in cells that overexpress p185erbB-2 (relative to the parental cells) but express lower levels than breast cancer cells with an erbB-2 amplification. Furthermore, these transduced HME cells can grow in the absence of insulin but not in the absence of EGF. Generation of HME cells with high-level erbB-2 overexpression

FIG. 2. erbB-2 expression in singly or multiply infected H16N2 cells. A, One hundred micrograms of membrane protein from vector-only infected and erbB-2-overexpressing cells, selected in G418-containing medium, was separated on 7.5% SDS-PAGE, blotted to PVDF, and probed with an erbB-2 antibody. 13 and 33 refer to the number of times the cells were infected with pTPerbB-2. B, Infected H16N2 cells, selected in G418, were sorted by flow cytometry for overexpression of cell surface erbB-2. The arrow indicates the threshold of the cells that were plated. C, One hundred micrograms of membrane protein from SUM-190PT cells, flow-sorted erbB-2-overexpressing H16N2 cells were separated on 7.5% SDS-PAGE, blotted to PVDF, and probed with an erbB-2 antibody.

Experiments were performed to determine whether direct, long-term selection of pTPerbB-2-infected cells for growth factor independence would yield cell populations expressing high levels of p185erbB-2. MCF-10A or H16N2 cells were infected with pTPerbB-2 or pTPneo and switched, 24 h later, to G418-containing medium without insulin or without insulin and EGF. Switching to growth factor-deficient medium, in this way, dramatically reduced the proliferative capacity of both cell lines, regardless of which vector was used to infect the cells. However, cells infected with pTPerbB-2 (but not control virus) and cultured continuously in growth factor deficient-medium slowly grew to confluence and could be subcultured. For both lines, cells had to be maintained for several weeks under these culture conditions before they became confluent. However, once confluent, cells in subsequent passages grew increasingly rapidly until achieving a growth rate comparable with control cells in the presence of growth factors. Thus, long-term selection in growth factordeficient medium did not result in the prompt emergence of growth factor-independent clones but, rather, resulted in the slow adaptation of the entire population of G418-resistant

erbB-2 AFFECTS GROWTH FACTOR INDEPENDENCE

cells to these conditions. Western blot analysis of erbB-2transduced growth factor-independent cells, compared with breast cancer cells with an erbB-2 gene amplification, is shown in Fig. 3A. These data demonstrate that growth factorindependent MCF-10erbB-2 and H16N2erbB-2 cells that emerged after long-term selection in growth factor-deficient medium expressed p185erbB-2 levels that were equivalent to that expressed by breast cancer cells with an erbB-2 gene amplification. Phosphotyrosine Western blot analysis of membrane protein obtained from these cells indicated that growth factor-independent cells expressed high levels of constitutively tyrosine-phosphorylated protein in the 185kDa range (Fig. 3B). Immunoprecipitation, using erbB-2 specific antibodies, followed by phosphotyrosine Western blotting, demonstrated that p185erbB-2 was indeed highly tyrosine phosphorylated in these cells (Fig. 3C). In addition, a tyrosine-phosphorylated p170 was coimmunoprecipitated with erbB-2, which was shown to be the EGF receptor by probing the blots with an EGF receptor-specific antibody (not shown). These data further support the hypothesis of a direct relationship between the level of erbB-2 expression and growth factor-independent proliferation of HME cells. Interestingly, whereas the H16N2erbB-2 cells could be routinely grown in insulin and EGF-free medium, MCF-

3619

10AerbB-2 cells, despite their relatively high levels of p185erbB-2, still required exogenous EGF for growth in culture. The data in Fig. 3, B and C, show that the levels of tyrosine phosphorylated p185erbB-2 in the MCF-10AerbB-2 cells were lower than in either the H16N2erbB-2 cells or the SUM-190 cells. This difference could account for the continued EGF-dependence of the MCF-10AerbB-2 cells. The high level expression and constitutive activation of p185erbB-2 found in H16N2-erbB-2 cells, relative to MCF-10A-erbB-2 cells, may reflect inherent differences between the two cell lines or may be the result of alterations induced by the HPV-16 genome that was used to immortalize the H16N2 cells. Further work will be required to explore this possibility. Constitutive activation of erbB-3 in erbB-2-overexpressing HME cells

Because the growth factor requirements of the HME cell lines were altered by progressive overexpression of erbB-2, experiments to examine the erbB-3 expression and activation status of these cells were performed. Phosphotyrosine Western analysis of erbB-3 immunoprecipitates, derived from the different erbB-2-transduced cell lines and controls, demonstrated that p180erbB-3 became activated to increasingly high levels in the growth factor-independent cells. Indeed, the H16N2erbB-2 cells that were both insulin- and EGF-independent, expressed the highest levels of tyrosine-phosphorylated erbB-3 (Fig. 4A). Under the conditions used for erbB-3 immunoprecipitation, erbB-2 was not coimmunoprecipitated (not shown). Probing of Western blots of erbB-3 immunoprecipitates with erbB-3 antibodies demonstrated that overall levels of p180erbB-3 were not changed in the erbB-2overexpressing cells (Fig. 4B). Thus, progressive overexpression of erbB-2 in HME cells results in the constitutive activation of erbB-3 without changes in the level of p180erbB-3. Furthermore, the level of activation of both erbB-2 and erbB-3 correlates with progressive growth factor independence. Relationship between insert number and erbB-2 expression levels in erbB-2-infected HME cells

FIG. 3. Amount and tyrosine phosphorylation of erbB2 protein in cells selected for extended periods of time in growth factor-depleted media. One hundred micrograms of membrane protein from erbB-2overexpressing MCF-10A or H16N2 cells, selected for at least 15 passages in growth factor-depleted media and from SUM-190PT cells, was separated on 7.5% SDS-PAGE, blotted to PVDF, and probed with a-erbB-2 antibody (A) or a-Ptyr4G10 (catalog no. 05–321, Upstate Biotechnology, Inc.) (B). erbB-2 was immunoprecipitated from 1 mg protein obtained from the various cell lines and separated on 7.5% SDS-PAGE, blotted to PVDF, and probed with a-Ptyr antibody (C).

The above results demonstrate that long-term selection of erbB-2-transduced HME cells in growth factor-deficient medium results in levels of p185erbB-2 expression equivalent to that of breast cancer cells with an erbB-2 gene amplification. To determine whether this long-term selection resulted in increased copy number of erbB-2 inserts, or whether the increase in expression occurred at the transcriptional level, Southern and Northern blot analyses were carried out using cells expressing different levels of p185erbB-2. The results shown in Fig. 5 demonstrate that long-term selection of cells, either singly- or triply-infected with pTPerbB-2, did not result in increases in erbB-2 insert number relative to cells that were transduced and directly flow sorted. By contrast, erbB-2 messenger RNA levels were dramatically increased in the multiple growth factor-independent cells. These results indicate that long-term maintenance of erbB2-transduced cells in growth factor-deprived conditions results in progressively higher levels of erbB-2 expression, at both the transcriptional and translational levels, without

3620

erbB-2 AFFECTS GROWTH FACTOR INDEPENDENCE

Endo • 1999 Vol 140 • No 8

FIG. 4. Activation of erbB-3 in erbB-2overexpressing cells. A, erbB-3 was immunoprecipitated from 1 mg protein from SUM-149PT cells, SUM-190PT cells, control cells (MCF-10Aand vectoronly-infected cells), and erbB-2-overexpressing MCF-10A and H16N2 cells and separated on 7.5% SDS-PAGE, blotted to PVDF, and probed with a-Ptyr antibody; B, One hundred micrograms of membrane protein from SUM-149PT cells, SUM-190PT cells, control cells (MCF-10A and vector-onlyinfected cells), and erbB-2-overexpressing MCF-10A and H16N2 cells was separated on 7.5% SDS-PAGE, blotted to PVDF, and probed with a-erbB-3 antibody.

changes in the number of erbB-2 inserts in the genome. The results of the Southern and Northern blot experiments are consistent with the original observation that growth factorindependent cells emerged slowly as a population rather than as a prompt clonal selection of preexist cells with higher copy numbers of the erbB-2 transgene. The mechanism by which prolonged culture, under growth factor-deficient conditions, results in the progressive increase in steady-state levels of erbB-2 messenger RNA remains to be determined. Discussion

For many years, our laboratory has been interested in the mechanisms by which breast cancer cells acquire independence of growth factors that are absolutely required by normal HME cells for growth under serum-free conditions. We initially studied these phenomena in rat mammary carcinoma cells. We found that the in vivo malignant potential of carcinogen-induced rat mammary carcinoma cells was directly related to their ability to grow in the absence of IGFs and, to a lesser extent, EGF (42, 43). Further study of these growth factor-independent cells demonstrated that IGF-independent growth was associated with the presence of a constitutively tyrosine phosphorylated p185 (40). Subsequent work demonstrated this protein to be erbB-2 and also demonstrated that rat mammary carcinoma cells synthesize and secrete NDF, the rat homologue of HRG (44). The finding that both rat and human HRG can act as potent IGF-like mitogens for mammary epithelial cells further established the link between the constitutive activation of erbB-2 and independence of IGF-like growth factors (21). The situation in human breast cancer cells is analogous to what was seen in rat mammary carcinoma cells, but the mechanism of erbB-2 activation seems different, at least in the setting of gene amplification. Human breast cancer cells with an erbB-2 gene amplification dramatically overexpress the protein, relative to normal HME cells (1, 8, 45). Furthermore, erbB-2 protein is constitutively tyrosine phosphorylated in the absence of any demonstrable autocrine growth factor activity (39). Thus, in this case, constitutive activation of p185erbB-2 seems to be the direct result of the overexpression of the protein, as has been suggested by others (46). In addition, breast cancer cells were found to be independent of

FIG. 5. Transcription of erbB-2 in infected H16N2 cells. A, Southern blot for erbB-2 of genomic DNA from erbB-2-overexpressing H16N2 cells. Genomic DNA from erbB-2-overexpressing H16N2 cells and flow sorted erbB-2-expressing H16N2 cells were compared by Southern blot with control cell DNA (PTP) for the amount of integrated erbB-2. B, Northern blot for erbB-2 message in erbB-2-overexpressing H16N2 cells. Total cellular RNA from erbB-2-overexpressing H16N2 cells was compared, in a Northern blot, with control cell (PTP) RNA for the amount of erbB-2 message. Northern blots were also probed with GAPDH to control for loading.

erbB-2 AFFECTS GROWTH FACTOR INDEPENDENCE

IGF-I for growth when erbB-2 was amplified and overexpressed to moderate levels and both IGF-I- and EGF-independent when erbB-2 was overexpressed to very high levels. In the present studies, we found that transduction of fulllength erbB-2 into immortalized HME cells results in erbB-2 overexpression. However, without long-term selection in the absence of growth factors, the pTPerbB-2 vector induced only moderate increases in p185erbB-2 expression, relative to parental control cells and these levels were not sufficient for rapid growth in the absence of insulin or EGF. When maintained in insulin-free, or insulin and EGF-free medium, MCF10A cells and H16N2 cells slowly expanded and, after being subcultured, grew increasingly well in the absence of growth factors. The growth factor-independent cells, so derived, expressed very high levels of erbB-2 protein, which were equivalent to that expressed by breast cancer cells with an erbB-2 gene amplification. The high-level erbB-2 expression was not the result of an increase in insert copy number but, rather, occurred at the transcriptional level. Furthermore, erbB-2 and erbB-3 expressed by these growth factor-independent cells were constitutively tyrosine phosphorylated, just as we had observed previously with breast cancer cells. Despite the high-level erbB-2 expression obtained in these HME cells, only the H16N2 cells became independent of both insulin and EGF. MCF-10AerbB-2 cells, although insulinindependent for growth, still require exogenous EGF. The levels of p185erbB-2 expressed by the MCF-10AerbB-2 cells are at least as high as the levels found in the SUM-190 breast cancer cells, which are independent of both insulin and EGF. This indicates that overexpression of erbB-2 is sufficient to mediate growth of HME cells in the absence of IGFs but not sufficient to make cells EGF independent. The observation that both 21MT-1 cells and SUM-190PT cells are independent of both insulin and EGF suggests that erbB-2 overexpression activates signaling pathways that act in concert with other alterations present in breast cancer cells, to yield EGF-independence. That H16N2erbB-2 cells are also EGF-independent suggests that specific changes mediated by the HPV-16 E5, E6, or E7 proteins cooperate with erbB-2 overexpression to complete the transformation to EGF-independence. Studies are currently underway to test this hypothesis. The ability of HRG to act as an IGF-like mitogen, and for constitutive erbB-2 activation to result in IGF-I-independent growth of mammary epithelial cells, is a consistent and intriguing observation in our laboratory. The potent activation of the PI 39-kinase pathway by erbB-3 and (to a lesser extent) erbB-2 may explain this important aspect of erbB-2/erbB-3 signal transduction (31, 32). The cytoplasmic domain of erbB-3 differs significantly from that of the other members of the erbB family, by having six YXXM motifs, which serve as docking sites for the p85 regulatory subunit of PI 39-kinase. Therefore, erbB-3 is a potent activator of this enzyme. We have demonstrated high-level PI 39-kinase activity in growth factor-independent, erbB-2-overexpressing human breast cancer cells and in HME cells stimulated with HRG (30). This is important because, in normal HME cells, which require both IGF-I and EGF for growth, activation of the PI 39-kinase pathway is driven predominantly by IGF-I signaling (34, 47, 48). Activation of the IGF-I receptor results in tyrosine phosphorylation of IRS-I, which contains 20 YXXM sites for PI

3621

39-kinase activation. Site-specific mutagenesis studies have demonstrated that the mitogenic activity of the IGF-I receptor is dependent on the integrity of the binding sites for p85 (33). Thus, the synergistic interaction between IGF-I and EGF, seen for most epithelial cell types, is the result of complementary signaling pathways activated by the two mitogens; the ras/raf/MAP kinase pathway by EGF, and the PI 39-kinase pathway by IGF-I. The observation that progressive overexpression and activation of erbB-2 in our cells yields a progressive increase in the levels of erbB-3 tyrosine phosphorylation without changes in erbB-3 protein levels is consistent with the hypothesis that erbB-2 and erbB-3 interact to drive altered phenotypes of breast cancer cells. The ability of activated erbB-2 to stimulate both PI 39kinase and the MAP kinase pathways (23) has profound implications for its behavior as an oncoprotein. Several groups have demonstrated the importance of the PI 39-kinase pathway as a mediator of cell survival and apoptosis resistance (37, 38, 49 –53). Thus, many cell types can be induced to undergo programmed cell death by blocking signals either from the IGF-I receptor, from PI 39-kinase itself, or from the AKT kinase that is activated by PI 39-kinase. In addition, the finding that the proapoptotic protein BAD can be phosphorylated by AKT, effectively preventing its interaction with BCL-2, provides mechanistic support for the hypothesis that PI 39-kinase signaling is important for cell survival (54, 55). Taken together, these findings suggest that erbB-2 is a potent oncogene, because it not only has the ability to drive cell proliferation but also has the ability to signal in ways that prevent programmed cell death. The results obtained from our work support this model and demonstrate directly that overexpression of erbB-2 results in the constitutive activation of both p185erbB-2 and erbB-3, which in turn, results in independence of IGF-I and, in some cases, EGF. Acknowledgment We thank Amy Pace for help in making the figures.

References 1. Slamon DJ, Clark GM, Wong SG, Levin WJ, Ullrich A, McGuire WL 1987 Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 235:177–182 2. Berger MS, Locher GW, Saurer S, Gullick WJ, Waterfield MD, Groner B, Hynes NE 1988 Correlation of c-erbB-2 gene amplification and protein expression in human breast carcinoma with nodal status and nuclear grading. Cancer Res 48:1238 –1243 3. Guerin M, Gabillot M, Mathieu MC, Travagli JP, Spielmann NA, Riou G 1989 Structure and expression of c-erbB-2 and EGF receptor genes inflammatory and non-inflammatory breast cancer: prognostic significance. Int J Cancer 43:201–208 4. Kraus MH, Popescu NC, Amsbaugh SC, King CR 1987 Overexpression of the EGF receptor-related proto-oncogene erbB-2 in human mammary tumor cell lines by different molecular mechanisms. EMBO J 6:605– 610 5. Lacroix H, Iglehart JD, Skinner MA, Kraus MH 1989 Overexpression of erbB-2 or EGF receptor proteins present in early stage mammary carcinoma is detected simultaneously in matched primary tumors and regional metastases. Oncogene 4:145–151 6. Ro J, El-Naggar A, Ro JY, Blick M, Frye D, Fraschini G, Fritsche H, Hortobagyi G 1989 c-erbB-2 Amplification in node-negative human breast cancer. Cancer Res 49:6941– 6944 7. van de Vijver M, van de Bersselaar R, Devilee P, Cornelisse C, Peterse J, Nusse R 1987 Amplification of the neu (c-erbB-2) oncogene in human mammary tumors is relatively frequent and is often accompanied by amplification of the linked c-erbA oncogene. Mol Cell Biol 7:2019 –2023 8. Wright C, Angus B, Nicholson S, Sainsbury JRC, Cairns J, Gullick WJ, Kelly P, Harris AL, Horne CHW 1989 Expression of c-erbB-2 oncoprotein: a prognostic indicator in human breast cancer. Cancer Res 49:2087–2090

3622

erbB-2 AFFECTS GROWTH FACTOR INDEPENDENCE

9. Zeillinger R, Kury F, Czerwenka K, Kubista E, Sliutz G, Knogler W, Huber J, Zielinski C, Reiner G, Jakesz R, Staffen A, Reiner A, Wrba F, Spona J 1989 HER-2 amplification, steroid receptors and epidermal growth factor receptor in primary breast cancer. Oncogene 4:109 –114 10. Fontaine J, Tesseraux M, Klein V, Bastert G, Blin N 1988 Gene amplification and expression of the neu (c-erbB-2) sequence in human mammary carcinoma. Oncology 45:360 –363 11. Borg A, Tandon AK, Sigurdsson H, Clark GM, Ferno M, Fuqua SAW, Killander D, McGuire WL 1990 HER-2/neu amplification predicts poor survival in node-positive breast cancer. Cancer Res 50:4332– 4337 12. Alimandi M, Romano A, Curia MC, Muraro R, Fedi P, Aaronson SA, Difiore PP, Kraus MH 1995 Cooperative signaling of erbB3 and erbB2 in neoplastic transformation and human mammary carcinomas. Oncogene 10:1813–1821 13. Qian X, Dougall WC, Gei Z, Greene MI 1995 Intermolecular association and trans-phosphorylation of different neu-kinase forms permit SH2-dependent signaling and oncogenic transformation. Oncogene 10:211–219 14. Murali R, Brennan PJ, Kieberemmons T, Greene MI 1996 Structural analysis of p185(c-neu) and epidermal growth factor receptor tyrosine kinases: oligomerization of kinase domains. Proc Natl Acad Sci USA 93:6252– 6257 15. Holmes WE, Sliwkowski MX, Akita RW, Henzel WJJL, Park JW, Yansura D, Abadi N, Raab H, Lewis GD, Shepard M, Kuang W-J, Wood WI, Goeddel DV, Vandlen RL 1992 Identification of heregulin, a specific activator of p185erb2. Science 256:1205–1210 16. Peles E, Benlevy R, Tzahar E, Liu NL, Wen DZ, Yarden Y 1993 Cell-type specific interaction of neu differentiation factor (NDF heregulin) with neu/ HER-2 suggests complex ligand receptor relationships. EMBO J 12:961–971 17. Plowman GD, Green JM, Culouscou JM, Carlton GW, Rothwell VM, Buckley S 1993 Heregulin induces tyrosine phosphorylation of HER4/P180(erbB4). Nature 366:473– 475 18. Carraway KL, Sliwkowski MX, Akita R, Platko JV, Guy PM, Nuijens A, Diamonti AJ, Vandlen RL, Cantley LC, Cerione RA 1994 The erbB3 gene product is a receptor for heregulin. J Biol Chem 269:14303–14306 19. Sliwkowski MX, Schaefer G, Akita RW, Lofgren JA, Fitzpatrick VD, Nuijens A, Fendly BM, Cerione RA, Vandlen RL, Carraway KL 1994 Coexpression of erbB2 and erbB3 proteins reconstitutes a high affinity receptor for heregulin. J Biol Chem 269:14661–14665 20. Carraway KL, Cantley LC 1994 A neu acquaintance for ErbB3 and ErbB4: a role for receptor heterodimerization in growth signaling. Cell 78:5– 8 21. Ram TG, Kokeny KE, Dilts CA, Ethier SP 1995 Mitogenic activity of neu differentiation factor/heregulin mimics that of epidermal growth factor and insulin-like growth factor-I in human mammary epithelial cells. J Cell Physiol 163:589 –596 22. Kita YA, Barff J, Luo Y, Wen DZ, Brankow D, Hu S, Liu NL, Prigent SA, Gullick WJ, Nicolson M 1994 NDF/heregulin stimulates the phosphorylation of Her3/erbB3. FEBS Lett 349:139 –143 23. Marte BM, Grausporta D, Jeschke M, Fabbro D, Hynes NE, Taverna D 1995 NDF/heregulin activates MAP kinase and p70/p85 S6 kinase during proliferation or differentiation of mammary epithelial cells. Oncogene 10:167–175 24. Wallasch C, Weiss FU, Niederfellner G, Jallal B, Issing W, Ullrich A 1995 Heregulin-dependent regulation of HER2/neu oncogenic signaling by heterodimerization with HER3. EMBO J 14:4267– 4275 25. Guy CT, Webster MA, Schaller M, Parsons TJ, Cardiff RD, Muller WJ 1992 Expression of the neu protooncogene in the mammary epithelium of transgenic mice induces metastatic disease. Proc Natl Acad Sci USA 89:10578 –10582 26. Muller WJ, Sinn E, Pattengale PK, Wallace R, Leder P 1988 Single-step induction of mammary adenocarcinoma in transgenic mice bearing the activated c-neu oncogene. Cell 54:105–115 27. Karunagaran D, Tzahar E, Beerli RR, Chen XM, Grausporta D, Ratzkin BJ, Seger R, Hynes NE, Yarden Y 1996 ErbB-2 is a common auxiliary subunit of NDF and EGF receptors: implications for breast cancer. EMBO J 15:254 –264 28. Pinkaskramarski R, Soussan L, Waterman H, Levkowitz G, Alroy I, Klapper L, Lavi S, Seger R, Ratzkin BJ, Sela M, Yarden Y 1996 Diversification of Neu differentiation factor and epidermal growth factor signaling by combinatorial receptor interactions. EMBO J 15:2452–2467 29. Ethier SP, Summerfelt RM, Cundiff KC, Asch BB 1990 The influence of growth factors on the proliferative potential of normal and primary breast cancer-derived human breast epithelial cells. Breast Cancer Res Treat 17:221–230 30. Ram TG, Ethier SP 1996 Phosphatidylinositol 3-kinase recruitment by p185(erbB-2) and erbB-3 is potently induced by neu differentiation factor/ heregulin during mitogenesis and is constitutively elevated in growth factorindependent breast carcinoma cells with c-erbB-2 gene amplification. Cell Growth Differ 7:551–561 31. Fedi P, Pierce JH, Difiore PP, Kraus MH 1994 Efficient coupling with phosphatidylinositol 3-kinase, but not phospholipase C gamma or GTPase-activating protein, distinguishes erbB-3 signaling from that of other erbB EGFr family members. Mol Cell Biol 14:492–500

Endo • 1999 Vol 140 • No 8

32. Soltoff SP, Carraway KL, Prigent SA, Gullick WG, Cantley LC 1994 ErbB3 is involved in activation of phosphatidylinositol 3-kinase by epidermal growth factor. Mol Cell Biol 14:3550 –3558 33. Myers MG, Zhang YT, Aldaz GAI, Grammer T, Glasheen EM, Yenush L, Wang LM., Sun XJ, Blenis J, Pierce JH, White MF 1996 YMXM motifs and signaling by an insulin receptor substrate 1 molecule without tyrosine phosphorylation sites. Mol Cell Biol 16:4147– 4155 34. Lam K, Carpenter CL, Ruderman NB, Friel JC, Kelly KL 1994 The phosphatidylinositol 3-kinase serine kinase phosphorylates IRS-1—stimulation by insulin and inhibition by Wortmannin. J Biol Chem 269:20648 –20652 35. Cheatham B, Vlahos CJ, Cheatham L, Wang L, Blenis J, Kahn CR 1994 Phosphatidylinositol 3-kinase activation is required for insulin stimulation of pp70 S6 kinase, DNA synthesis, and glucose transporter translocation. Mol Cell Biol 14:4902– 4911 36. Roche S, Koegl M, Courtneidge SA 1994 The phosphatidylinositol 3-kinase alpha is required for DNA synthesis induced by some, but not all, growth factors. Proc Natl Acad Sci USA 91:9185–9189 37. Alessi DR, Andjelkovic M, Caudwell B, Cron P, Morrice N, Cohen P, Hemmings BA 1996 Mechanism of activation of protein kinase B by insulin and IGF-1. EMBO J 15:6541– 6551 38. Kulik G, Klippel A, Weber MJ 1997 Antiapoptotic signalling by the insulinlike growth factor I receptor, phosphatidylinositol 3-kinase, and Akt. Mol Cell Biol 17:1595–1606 39. Ram TG, Dilts CA, Dziubinski ML, Pierce LJ, Ethier SP 1996 Insulin-like growth factor and epidermal growth factor independence in human mammary carcinoma cells with c-erbB-2 gene amplification and progressively elevated levels of tyrosine phosphorylated erbB-2. Mol Carcinog 15:227–238 40. Harris KF, Christensen JB, Radany EH, Imperiale MJ 1998 Novel mechanisms of E2F induction by BK virus large T antigen: requirement of both the pRb binding and J domains. Mol Cell Biol 18:1746 –1756 41. Band V, Zajchowski D, Swisshelm D, Trask D, Kulesa V, Cohen C, Connolly J, Sager R 1990 Tumor progression in four mammary epithelial cell lines derived from the same patient. Cancer Res 50:7351–7357 42. Ethier SP, Cundiff KC 1987 Importance of extended growth potential and growth factor independence on in vivo neoplastic potential of primary rat mammary carcinoma cells. Cancer Res 47:5316 –5322 43. Ethier SP, Moorthy R 1991 Multiple growth factor independence in rat mammary carcinoma cells. Breast Cancer Res Treat 18:73– 81 44. Ethier SP, Langton BC, Dilts CA 1996 Growth factor-independent proliferation of rat mammary carcinoma cells by autocrine secretion of neu-differentiation factor/heregulin and transforming growth factor-alpha. Mol Carcinog 15:134 –143 45. Hollywood DP, Hurst HC 1993 A novel transcription factor, OB2–1, is required for overexpression of the proto-oncogene c-erbB-2 in mammary tumour lines. EMBO J 12:2369 –2375 46. Weiner DB, Kokai Y, Wada T, Cohen JA, Williams WV, Greene MI 1989 Linkage of tyrosine kinase activity with transforming ability of the p185neu oncoprotein. Oncogene 4:1175–1183 47. Sun XJ, Rothenberg P, Kahn CR, Backer JM, Araki E, Wilden PA, Cahill DA, Goldstein BJ, White MF 1991 Structure of the insulin receptor substrate IRS-1 defines a unique signal transduction protein. Nature 352:73–77 48. Herbst JJ, Andrews G, Contillo L, Lamphere L, Gardner J, Lienhard GE, Gibbs EM 1994 Potent activation of phosphatidylinositol 39-kinase by simple phosphotyrosine peptides derived from insulin receptor substrate 1 containing two YMXM motifs for binding SH2 domains. Biochemistry 33:9376 –9381 49. King WG, Mattaliano MD, Chan TO, Tsichlis PN, Brugge JS 1997 Phosphatidylinositol 3-kinase is required for integrin-stimulated AKT and raf-1/mitogen-activated protein kinase pathway activation. Mol Cell Biol 17:4406 – 4418 50. Cheng JQ, Altomare DA, Klein MA, Lee WC, Kruh GD, Lissy NA, Testa JR 1997 Transforming activity and mitosis-related expression of the AKT2 oncogene: evidence suggesting a link between cell cycle regulation and oncogenesis. Oncogene 14:2793–2801 51. Datta K, Bellacosa A, Chan TO, Tsichlis PN 1996 Akt is a direct target of the phosphatidylinositol 3-kinase. J Biol Chem 271:30835–30839 52. Franke TF, Kaplan DR, Cantley LC, Toker A 1997 Direct regulation of the Akt proto-oncogene product by phosphatidylinositol-3,4-bisphosphate. Science 275:665– 668 53. Burgering BMT, Coffer PJ 1995 Protein kinase B (c-Akt) in phosphatidylinositol-3-OH kinase signal transduction. Nature 376:599 – 602 54. Datta SR, Dudek H, Tao X, Masters S, Fu H, Gotoh Y, Greenberg ME 1997 Akt phosphorylation of BAD couples survival signals to the cell-intrinsic death machinery. Cell 91:231–241 55. del Peso L, Gonzalez-Garcia M, Page C, Herrera R, Nunez G 1997 Interleukin3-induced phosphorylation of BAD through the protein kinase Akt. Science 278:687– 689