Background
Receptor tyrosine kinases (RTKs) are transmembrane proteins with intracellular kinase domains that undergo phosphorylation in response to ligand binding. This group of proteins has a well established role in breast cancer, and thus many RTKs are currently the focus of directed therapeutics with a significant number of these therapies in clinical trials. Two such proteins with validated roles in breast cancer are ErbB2 (also known as Her2/neu), a member of the epidermal growth factor receptor family, and the type I insulin-like growth factor receptor (IGF-IR). A large amount of evidence implicating both in clinical breast cancer is emerging. In addition, both receptors have been validated as oncogenes through the generation and characterization of transgenic mouse models (reviewed in [
1] and [
2]).
The IGF-IR undergoes autophosphorylation on conserved intracellular tyrosine residues after binding its ligands IGF-I and IGF-II which subsequently triggers signal cascades involved in many processes including proliferation and evasion of apoptosis [
3]. Common activated downstream tyrosine kinase cascades include the phosphatidyl inositol-3 kinase (PI-3K)/Akt and mitogen activated protein kinase (MAPK)/Erk1/2 pathways [
4]. It is also widely accepted that the IGF-IR has the capacity to transform normal cells and its expression is required for transformation by other known oncogenes [
5‐
8]. Studies have shown IGF-IR levels are highly expressed in 39-93% of human breast cancers [
9]and in breast tumors expressing high levels of IGF-IR the receptor is expressed 10-14-fold higher on average compared to normal breast tissue [
10,
11]. In addition, phosphorylation of IGF-IR was reported to be 2-4-fold higher in breast cancer tissue, which translated to a 40-fold increase in active IGF-IR in some cases [
11]. Regulation of IGF-IR expression in breast cancer appears quite complex as it has been associated with both a poor [
12] and favorable prognosis [
13] and may change depending on the stage of the tumor [
14]. In mouse xenografts, syngeneic, and chemically induced models of mammary carcinogenesis, inhibition of the IGF-IR through a variety of means has yielded success in limiting tumor growth and proliferation [
15‐
19]. In addition, the transforming potential of this protein has been confirmed with two transgenic mouse models, one using the native IGF-IR and the other employing a constitutively active form of the receptor [
20,
21]. In light of the mounting evidence suggesting a prominent role of the IGF-IR in breast cancer, numerous targeted therapies are currently in clinical trials [
22,
23].
Unlike the IGF-IR and other RTKs, ErbB2 has no known specific ligand [
24]. However, it preferentially forms heterodimers with other members of the EGF receptor family, EGFR (ErbB1), ErbB3 and ErbB4 [
25]. In addition, ErbB2 heterodimers were determined to enhance the mitogenic signal of ErbB1 and ErbB3 [
26]. Gene amplification with subsequent overexpression of ErbB2 has also been shown to contribute to the formation of activated homodimers [
27]. Downstream signaling cascades triggered by ErbB2 homo/heterodimers are very similar to those activated by IGF-IR. ErbB2 is overexpressed in 25-30% of human breast cancer cases and is correlated with poor prognosis and shorter disease free survival [
28]. It has been established that human and mouse ErbB2 variants are capable of transforming normal murine mammary epithelial cells and NIH/3T3 fibroblasts [
29,
30]. As stated previously, a number of transgenic models have validated this observation; animals overexpressing wild type neu formed tumors with an average latency of 7 months while with an activated form of ErbB2, latency was decreased to 3 months [
2]. Because of the clinical implications of ErbB2 overexpression and its transforming potential, a number of directed therapies have been developed for the treatment of breast cancer which target ErbB2 alone or in combination with other RTKs; one of which, trastuzumab (Herceptin) is currently used to treat Her2
+ breast cancer (reviewed in [
31]).
There is a growing body of evidence suggesting an interaction between the IGF-IR and ErbB2 in clinical breast cancer. Different studies have shown a physical interaction between the two receptors through immunoprecipitation [
32‐
34] and immunofluorescence co-localization staining [
34]. It has also been determined that knocking down expression of IGF-IR in human breast cancer cell lines can attenuate ErbB2 phosphorylation; however, the reciprocal was not observed [
34]. In addition, signaling though the IGF-IR has been shown to mitigate the growth inhibitory effects of trastuzumab on human breast cancer cells overexpressing ErbB2 [
35]; in fact, it is becoming widely accepted that signaling through IGF-IR can contribute to resistance to ErbB2-directed therapies (reviewed in [
36]), and it has been noted that a correlation between IGF-IR expression and trastuzumab resistance exists in clinical breast cancer cases [
37]. The existence of IGF-IR/ErbB2 heterodimers has also been established and this association was shown to contribute to Herceptin resistance in human breast cancer cell lines; this interaction could be disrupted by treating with monoclonal antibody based therapy to IGF-IR [
38]. Based on these findings, a number of groups have investigated different combinations of inhibitors of IGF-IR and ErbB2 signaling. Nordihydroguaiaretic acid (NDGA), a dual inhibitor of IGF-IR and ErbB2, was shown to promote cell death in trastuzumab resistant human breast cancer cell lines [
39]. Also, a synergistic growth inhibitory relationship between trastuzumab and inhibitors of IGF signaling including small molecule inhibitors of IGF-IR, IGFBP-3 and a dominant negative IGF-IR has been established in ErbB2 overexpressing human breast cancer cell lines [
40‐
42]. An IGF-IR inhibitor was even found to enhance ErbB2-mediated apoptosis in a human breast cancer cell line with very little IGF-IR expression; evidence of this inhibitor augmenting the suppression of ErbB2 phosphorylation was also discovered [
32].
To examine the role of the IGF-IR in breast cancer, our lab has previously created a doxycycline-inducible transgenic mouse model (MTB-IGFIR). IGF-IR-induced transgenic animals develop multiple tumors with 100% penetrance and an average latency of approximately 50 d [
21], with metastasis occurring in approximately 40% of mice (unpublished observations). From one such primary tumor, the RM11A cell line was established and was shown to maintain doxycycline-inducible overexpression of the IGF-IR [
43]. Enhanced phosphorylation of downstream signaling molecules Akt and Erk1/2 was observed upon transgene activation in both tumor tissue and RM11A cells [
21,
43].
In vivo, these cells were shown to form tumors upon injection into the mammary gland of syngeneic, wild type, FVB mice. Because of its inducible nature, our model can be used to mimic the effects of IGF-IR-directed therapies through the deactivation of the transgene, and therefore provides a unique opportunity to study the potential function of other known oncogenes during IGF-IR-mediated mammary tumorigenesis. As a number of IGF-IR inhibitory compounds are currently in clinical trials [
23,
44], this model can be used to predict how the inhibition of other proteins/pathways could be utilized to augment IGF-IR-directed therapeutics.
It has been observed that ErbB2 overexpression can alleviate the requirements of IGF and EGF for proliferation in a series of human normal and breast cancer cell lines [
45]. However, most studies have focused on IGF signaling as a mechanism through which clinical cancers gain resistance to ErbB2-directed therapies. It remains unclear whether the reciprocal is also true; signaling through ErbB2 can potentially promote resistance to IGF-IR-targeted treatments. Also, with the prevalence of both oncogenes overexpressed in human breast cancer, it is important to investigate their interaction in mammary tumorigenesis to determine whether it would be useful to combine therapeutics directed at both ErbB2 and IGF-IR. Thus, our objective was to determine if ErbB2 overexpression can augment tumorigenesis and compensate for IGF-IR downregulation in RM11A cells, a model of IGF-IR-mediated mammary tumorigenesis. We accomplished this through overexpression of ErbB2 in the RM11A mammary cell line. We then examined how ErbB2 contributed to cell survival/growth
in vitro and
in vivo, cell signaling, primary tumorigenesis, recurrence in the absence of IGF-IR transgene expression and metastasis. In this study, it was determined that a modest increase in ErbB2 expression could accelerate primary tumor growth by enhancing proliferation immediately after cell colonization of the mammary gland. Overexpression of ErbB2 also impaired regression of tumors in the absence of IGF-IR transgene expression and facilitated metastasis.
Discussion
While the contribution of the IGF-IR to ErbB2 signaling and resistance to ErbB2-directed therapies in breast cancer has been studied in several systems, the reciprocal interaction remains almost completely unknown. To study the potential role of ErbB2 during IGF-IR-mediated mammary tumorigenesis we utilized our model of inducible IGF-IR overexpression. The importance of the IGF-axis in proliferation and transformation of a vast number of cells including human mammary epithelial cells is well documented. Our laboratory has shown that IGF-IR overexpression alone is capable of mediating an extremely rapid transformation of mouse mammary epithelial cells [
21]. Most of these tumors remained dependent on transgene expression as IGF-IR transgene downregulation resulted in tumor regression in a majority of these tumors. A small subset of tumors was capable of resuming growth in the absence of IGF-IR transgene expression [
49]. Similarly, RM11A cells grown in the mammary fat pad of syngeneic wild type mice developed tumors more rapidly when IGF-IR transgene expression was induced and IGF-IR transgene downregulation resulted in the regression of most of the tumors. Unlike the transgenic mammary tumors, most of the RM11A induced tumors eventually resumed growth following IGF-IR transgene downregulation [
43]. Therefore, while IGF-IR-induction remains very important for mammary tumorigenesis in our model, clearly alterations in the expression of other genes/proteins can be acquired during tumorigenesis, which allow tumors to grow independent of the IGF-IR transgene. Given the prominent role of ErbB2 in breast cancer and the previously discussed interactions with the IGF-IR, this oncogene was an attractive candidate.
Selection of stable transfectants yielded RM11A cells with approximately 3-fold higher expression of ErbB2 (RM11A+Dox/ErbB2) than control RM11A cells (RM11A+Dox). Phosphorylated ErbB2 was also elevated approximately 3-fold in the ErbB2 overexpressing cells thus indicating the receptor was active. This overexpression was monitored and consistently maintained throughout the duration of the study (representative western blot is shown in Figure
1). It is well understood that breast cancer tissue and cell lines can contain upwards of 50-100-fold higher levels of ErbB2 protein compared to normal tissue and these levels are typically associated with gene amplification and poor prognosis [
50,
51]. While a modest overexpression was achieved in our cell line, two points should be taken into consideration; first, RM11A cells intrinsically contain high levels of ErbB2. In addition, recent studies have shown a correlation between low-level ErbB2 expression (comparable to normal mammary tissue) and an overall unfavorable disease outcome [
52,
53].
To control for phenotypes caused by integration of the plasmid, we utilized two separate clones for both control and ErbB2 overexpressing stable integrants. Interestingly, the ErbB2 antibody detected two additional bands only in clones with pEN1-ErbB2 integration (and not with the control plasmid). These bands were determined to be approximately 95 and 70 kDa. A 95 kDa truncated N-terminal product of Her2, known as p95Her-2, has been previously described [
54]. This degradation product arises from proteolytic cleavage and subsequent release of the extracellular domain [
55,
56]. While the function of this protein is largely unknown, it has been shown to contribute to herceptin resistance [
57]. In addition, it was observed that higher levels of this product correlated with shorter disease free survival and increased lymph node metastasis [
58‐
60]. Thus, the presence of this ErbB2 degradation product appears to contribute to tumorigenesis. In our cell line, the presence of such a product would indicate that stable integration of this expression vector has yielded a threshold level of ErbB2 expression required for significant production of this cleavage product.
Downstream signaling pathways were studied to determine those potentially augmented by ErbB2 overexpression. The levels of phosphorylated Akt and Erk1/2 were similar in RM11A+Dox cells and RM11A+Dox/ErbB2 cells suggesting that upregulation of ErbB2 was incapable of further activating PI-3K or MAPK pathways. Given the magnitude of IGF-IR overexpression and the fact that both of these pathways are known to be activated by this receptor this observation is not surprising; it is anticipated that the high level of IGF-IR expression has already maximized signaling though the PI-3K and MAPK pathways [
43].
In vivo, it was observed that ErbB2 conferred a more rapid tumor onset and tumor incidence was also elevated as indicated by the number of mammary glands injected that actually developed tumors. To verify that ErbB2 shortened tumor latency, mammary glands were collected 14 d post-injection. Average tumor size at this time point was 4-fold greater in the RM11A+Dox/ErbB2 cells compared to RM11A+Dox cells. We then explored possible mechanisms through which ErbB2 augmented tumor growth. First we looked at survival and proliferation
in vitro. Only a small, insignificant increase in survival was observed in RM11A+Dox/ErbB2 cells were compared to RM11+Dox cells and thus it was concluded that ErbB2 overexpression had a negligible effect on RM11A cell survival
in vitro. Despite a minimal effect
in vitro, overexpression of ErbB2 had a marked effect on tumorigenesis
in vivo. Using Ki67 staining to examine proliferation our data suggested that proliferation is only significantly affected by ErbB2 overexpression shortly after tumor cells colonize the mammary tissue (4 d post injection but not 14 d post injection). The lack of difference in proliferation in established tumors was corroborated by evaluating tumor growth rates using two independent methods. For the first method log(tumor volume) was plotted against time and from the resulting slope of the line tumor doubling time was calculated. The second technique, specific growth rate has been mathematically determined to be an accurate means of quantifying tumor growth rate and is less susceptible to negligible or negative changes in volume from one measurement to the next [
61]. In this experiment, tumor volumes were sometimes observed to decrease before ultimately increasing again and thus this model of growth rate was deemed appropriate. Both methods yielded similar results and showed that ErbB2 overexpression did not enhance tumor growth rate. Therefore, it was concluded that in the presence of IGF-IR overexpression, ErbB2 facilitates proliferation during initial tumor establishment but has no effect on tumor growth/proliferation after this point.
Tumor regression following IGF-IR transgene downregulation was studied to model the effects of ErbB2 overexpression during the use IGF-IR-directed therapeutics. Here it was observed that ErbB2 overexpression impaired tumor regression following IGF-IR downregulation thus suggesting that ErbB2 could potentially facilitate resistance to IGF-IR-directed therapies. These results are of obvious clinical importance as ErbB2 status may become an important predictor of response to IGF-IR-directed therapies. In addition, subsequent mutations enhancing ErbB2 expression may render tumors unresponsive to these therapies. It is becoming clear that IGF-IR can mediate resistance to ErbB2-targeting treatments [
33]; therefore, it stands to reason that the reciprocal interaction may also be important.
Metastasis was also studied for multiple reasons; first, ErbB2 expression is well known to correlate with distant metastasis in human clinical breast cancer [
62]. Also, overexpression of this oncogene has been shown to increase the metastatic potential of human breast cancer cell lines [
63]. Lastly, in transgenic models of ErbB2 overexpression, metastasis to the lung is a common occurrence [
2]. Contrary to our expectations, ErbB2 overexpression did not facilitate an increase in lung metastasis. This could be due to the rapid establishment, onset and subsequent growth rate in primary tumors formed with both IGF-IR and ErbB2 overexpression; for this condition time elapsed between tumor onsets and euthanizing the animals was only 31 days on average and sometimes as short as 15 days, while palpable tumor onset occurred as early as 7 days post injection.
ErbB2 overexpression did however facilitate metastasis following IGF-IR downregulation. It is possible that through the delayed process of partial regression and subsequent resumption of growth, metastasic lesions have time to grow to a size where they are detectable histologically. Furthermore, upregulation of ErbB2 was observed in metastatic primary tumors as well as many metastatic lesions from MTB-IGFIR mice compared to non-metastatic primary tumors. Based on the fact that metastasis is only observed in 40% of all MTB-IGFIR animals, it is apparent that other alterations must occur to confer metastatic competency. Our results suggest that upregulation of ErbB2 is one such mechanism through which tumor cells gain this capacity. These observations suggest that ErbB2 can compensate for the loss of IGF-IR signaling during mammary tumorigenesis and further supports a potential advantage in combining ErbB2 and IGF-IR-directed therapies.
In conclusion, this study describes experiments providing information regarding the interaction between two potent oncogenes in mammary tumorigenesis. It has been previously postulated that targeting multiple signaling pathways such as IGF-IR and ErbB2 may be beneficial to the treatment of breast cancer [
35]. Based on our results, it is suggested that ErbB2 can augment IGF-IR-dependent mammary tumorigenesis by enhancing initial colonization and proliferation as well as following IGF-IR suppression.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
CC participated in design and coordination of the study as well as all of the experiments described in this study and drafting of the manuscript. JP participated in design of the study. RM coordinated the study and contributed to drafting of the manuscript. All authors have read and approved the final manuscript.