FAX: (212) 534-1684.
Rapamycin together with herceptin significantly increased anti-tumor efficacy compared to either alone in ErbB2 over expressing breast cancer cells
Article first published online: 15 FEB 2007
Copyright © 2007 Wiley-Liss, Inc.
International Journal of Cancer
Volume 121, Issue 1, pages 157–164, 1 July 2007
How to Cite
Wang, L.-H., Chan, J. L.-K. and Li, W. (2007), Rapamycin together with herceptin significantly increased anti-tumor efficacy compared to either alone in ErbB2 over expressing breast cancer cells. Int. J. Cancer, 121: 157–164. doi: 10.1002/ijc.22606
- Issue published online: 24 APR 2007
- Article first published online: 15 FEB 2007
- Manuscript Accepted: 22 DEC 2006
- Manuscript Received: 26 OCT 2006
- DOD. Grant Number: DAMD17-02-1-0504/BC11101
- NIH. Grant Number: CA29339
- breast cancer;
- tumor growth
The objective of this study was to assess the anti-tumor efficacy of rapamycin alone or in combination with herceptin in breast cancer. A total of 20 human breast cancer lines were examined for expression of various receptor tyrosine kinases and activation of their down stream signaling molecules, as well as for their invasion and colony forming ability. The ErbB2 and PI3 kinase pathway inhibitors were tested for the inhibition on breast cancer cell growth and tumor development. Seven of the 20 lines displayed an elevated level of ErbB2, others had varying level of EGF, IGF-1 or insulin receptor. Over 30% of the lines also had constitutive activation of Akt and MAP kinase. The lines displayed a wide range of colony forming and invasion ability. The PI3 kinase pathway inhibitors LY294002 and rapamycin inhibited the colony forming ability of all of the lines with the ErbB2 overexpressing lines having a higher sensitivity. A similar trend was observed for inhibition of invasion by LY294002. Rapamycin alone and additively together with herceptin inhibited the breast cancer cell growth especially in ErbB2 overexpressing cells. Rapamycin and herceptin synergistically inhibited tumor growth and endpoint tumor load in a xenograft model using a MCF-7 subline and in a MMTV-ErbB2 transgenic model. Rapamycin and herceptin significantly reduced the level of cyclin D1 and D3 and increased the cleavage of caspase 3 suggesting an increased apoptosis. Our results suggest that rapamycin together with herceptin has an enhanced anti-cancer effect and could be developed as an improved therapeutic regimen for breast cancer. © 2007 Wiley-Liss, Inc.
ErbB2, a member of the epidermal growth factor receptor (EGFR) family, has been found to be overexpressed and/or amplified in about 30% of human breast cancers.1 Amplification of ErbB2 is associated with a high relapse rate and poor clinical prognosis.2, 3, 4 Both in-vitro and in-vivo studies have demonstrated the functional role of ErbB2 in cell transformation and mammary cell carcinogenesis. Overexprssion of ErbB2 led to transformation of NIH3T3 cells and mammary epithelial cells.5, 6, 7, 8 Down modulation of ErbB2 by a monoclonal antibody reverted the transformed phenotype of ErbB2 transformed NIH3T3 cells.9 Mammary tumor virus promoter-based c-ErbB2 transgenic mice developed mammary carcinoma.10, 11, 12 Spontaneous ErbB2 amplification or ectopic overexpression of ErbB2 leads to activation of PI3 kinase signaling, which has been demonstrated to play an important role in ErbB2-mediated cell transformation and mammary malignancy, in part via regulating genes affecting cell cycle, growth and drug sensitivity.13, 14, 15, 16, 17, 18 Activation phosphorylation of Akt, mTor and 4E-Bp1 was shown to correlate with ErbB2 overexpression, tumor progression and poor prognosis.17
Thus, the components of the ErbB2/PI3 kinase pathways represent rational targets for human breast cancer therapy.19 A humanized monoclonal antibody against ErbB2, trastuzumab or herceptin was developed20 and used subsequently for therapy in patients with metastatic breast cancer involving overexpression of ErbB2.21, 22, 23, 24 Clinical trials so far indicate that as a first line single agent, herceptin can achieve an objective response of 15–30% and can significantly improve the response rate of tumor progression interval and median survival when combined with chemotherapy such as paclitaxel.25, 26, 27 However, the overall response rate with herceptin alone or in combination with chemotherapy underscores the need of an improved therapeutic regimen for breast cancer. We have shown previously that ErbB2-overexpressing human mammary carcinoma cells display an increased sensitivity toward inhibitors of the PI3 kinase signaling pathway, including rapamycin, for anchorage independent growth and that non-ErbB2-overexpressing cells could be sensitized for the rapamycin inhibition for growth by ectopic overexpression of ErbB2.28 The inhibition was due to, at least in part, the decrease of cyclin-associated CDK activity.28 Rapamycin forms complexes with immunophilin FK506-binding protein 12 leading to the specific inhibition of the serine/threonine protein kinase mTOR or mammalian target of rapamycin.29 mTOR is a downstream component of the PI3 kinase signaling of which the 2 major phosphorylation targets are 40S ribosomal protein kinase, p70 S6 kinase and a translation repressor 4E-Bp1 (eIF4E-binding protein 1).30, 31, 32 Phosphorylation of p70 S6 kinase and 4E-Bp1 results in their activation and inactivation, respectively.30, 31, 32 Therefore inhibition of the mTOR by rapamycin results in decreased translation of the RNAs involved in cell cycle progression and proliferation leading to growth arrest and apoptosis.33, 34, 35 This is supported by the recent report that rapamycin is able to disrupt the formation of cyclin/cyclin-dependent kinase/P21/proliferating cell nuclear antigen complexes by down regulation of P21, which serves as the anchor of the complexes.36 Rapamycin and its analogs are being developed as therapeutic drugs by themselves or in combination with other chemotherapy drugs for various types of human malignancy including breast cancer.37, 38, 39, 40 A recent study showed that rapamycin synergistically enhanced paclitaxel- and carboplatin-induced cytotoxicity in ErbB2-overexpressing breast cancer cells.41 However, to our knowledge there have not been reports on preclinical or clinical studies on the combination therapy for breast cancer using rapamycin and herceptin. Since PI3 kinase pathway can be activated via a variety of receptor protein tyrosine kinases (RPTKs), inhibition of ErbB2 by herceptin most likely could only achieve a partial inhibition of the pathway. A combination of one of the PI3 kinase pathway inhibitors, namely rapamycin and herceptin is expected to be able to block the PI3 kinase signaling more effectively. Rapamycin has been used for various clinical purposes as described above; thus combination of herceptin and rapamycin is a logical choice for potential combined cancer therapy. Besides, combination of the 2 drugs may allow the use of a lower dosage of herceptin to achieve an effective tumor inhibition, thus reducing the toxicity caused by herceptin.
The purpose of this study is to confirm and extend our previous finding that ErbB2-overexpressing breast cancer cells display an increased requirement of PI3 kinase signaling for anchorage independent growth and to evaluate the anti-cancer effect of rapamycin by itself or in combination with herceptin in both in vitro and in vivo systems. The results confirm our previous observation of hypersensitivity of ErbB2 over-expressing breast cancer cells toward rapamycin for anchorage independent growth. Moreover, rapamycin is able to enhance herceptin-mediated growth inhibition in vitro and tumorigenicity in vivo of ErbB2 overexpressing human breast cancer cells.
Material and methods
Cell culture and reagents
Some of the established normal human breast epithelial and breast cancer lines have been described previously.28 Additional lines used in this study have been obtained from ATCC. Cells were grown in DMEM containing 10% FBS. Some of the lines (T47D, Hs57T, MCF7, BT483, BT549, MDA-MB-330 and MDA-MB-415) were maintained in the same medium supplemented with insulin at 50 ng/ml. An ErbB2 overexpressing, highly tumorigenic subline called MCF7-I4 was derived from the original MCF7 line (to be described elsewhere).
Tissue culture grade LY294002, PD98059 and rapamycin were purchased from LC laboratories. The humanized mouse monoclonal antibody against the human ErbB2, herceptin or tratsuzumab, was provided by Genentech via a material transfer agreement. Rabbit antiserum against the cytoplasmic domains of the insulin and IGF-1 receptors have been prepared and described previously.42 Antibodies against ErbB2 and EGF receptors were purchased from Calbiochem and Santa Cruz Biotechnology. Antibodies against AKT1 and AKT2 were also purchased from Santa Cruz Biotechnology. Antibodies against phospho-AKT, MAP kinase, phospho-MAP kinase, p70 S6 kinase and phosphor-p70 S6 kinase were purchased from Cell Signaling.
Assay of cell growth, migration, invasion and colony formation
For testing the effect of drugs on the growth of cell in monolayer, 1 × 105 cells were seeded per 6 cm dish in duplicates in regular maintenance medium. Twenty-four hours later the cultures were replenished with fresh medium with or without the drug to be tested and the numbers of cells were counted 3 days later. The assay for cell migration has been described previously.43 The assay for cell invasion was similar to that of migration except that the cell culture inserts were coated with matrigel as follows: 0.1 ml of DMEM containing 0.1% BSA and 0.5 μg/ml of growth factor reduced matrigel was added to each insert and allowed to sit at room temperature for 15 min. The inserts were then incubated at 37°C for 0.5–2 hr. Thirty minutes before adding cells to the inserts 400 μl of DMEM containing 0.1% BSA was added to hydrate the matrigel. The remainder of the assay was performed as for the migration assay. Colony assays followed the procedures described previously.44, 45, 46
Preparation of total cell lysates, determination of protein concentration, immunoprecipitation and western analysis followed the previously described procedures.47
For xenograft model, the MCF7 I4 line and female BALB/C athymic mice were used. 2 × 106 cells in 0.1 ml PBS were injected subcutaneously at flank per mouse. One week later, mice were injected at the site of tumor with 0.1 ml PBS or 0.1 ml PBS containing 100 ng/ml of rapamycin, 100 μg/ml herceptin or 100 ng/ml of rapamycin plus 100 μg/ml herceptin. The body weight and tumor size were measured twice a week. At 30 days post tumor cell injection, which was set as the endpoint, the mice were dissected and examined for possible metastasis.
For transgenic mouse model, the FVB/N-TGN (CMMTVneu) mice from Jackson laboratories were used. The transgenic mice harbor the human ErbB2 under the control of the mouse mammary tumor virus LTR promoter.11 Eight-week-old mice were purchased. The drug injection was started when mice reached 4.5 months of age. Four groups of mice were individually injected intraperitoneally twice a week with PBS, 100 ng/ml of rapamycin in 0.1 ml PBS, 100 μg/ml herceptin in 0.1 ml PBS or a combination of rapamycin and herceptin. Sixteen injections were performed. The body weight and appearance of tumors and tumor sizes were monitored twice a week. The endpoint was set when mice were 9.5 months old and most of the control mice developed tumors.
Expression of RPTKs, AKT and MAP kinase signaling molecules in various human breast cancer cell lines
To further investigate the relationship between ErbB2 overexpression in breast cancer cells and their sensitivity toward the inhibitors of the PI3 kinase signaling pathway as we reported previously,28 various breast cancer lines were analyzed for the expression and activation of insulin receptor (IR), IGF-1 receptor (IGF-1R), EGFR and ErbB2 as well as their down stream signaling molecules including AKT, p70 S6 kinase and MAP kinase (Fig. 1a). The results showed that out of 19 lines examined about half of them displayed varying degrees of elevated ErbB2 expression. Several of them, including MDA-MB-468, BT20, MDA-MB-231, showed relatively high EGFR expression. All of them had significant levels of IR and IGF-1R. MCF7, T47D, BT20, MDA-MB-436 had particularly high levels of IGF-1R confirming and extending our previous observations.28, 48 There appeared to have no correlation between the expression levels of Akt, MAP kinase and p70 S6 kinase and their states of activation as measured by the degree of phosphorylation (Fig. 1b). The lines displayed varying degrees of constitutive activation of Akt, MAP kinase and p70 S6 kinase.
Anchorage independent growth, invasion activity and growth inhibition by inhibitors against the PI3 and MAP kinase signaling pathways
The colony forming ability and invasion activity of the various breast cancer cell lines were examined. The results show that they exhibit differential abilities for anchorage-independent growth (Fig. 2a) and invasion activity (Fig. 2b and data not shown) irrespective of their expression levels of ErbB2. Confirming our previous observation, the anchorage-independent growth of ErbB2 overexpressing lines displayed a greater sensitivity than the non-ErbB2 overexpressing lines toward inhibitors of the PI3 kinase including LY294002 and rapamycin (Fig. 3a). All of them were relatively resistant to inhibition by PD98059 (Fig. 3a). The invasion ability of the lines shows different degrees of sensitivity toward LY294003, PD98059 and rapamycin (Fig. 3b). Under the concentrations of PD98059 and LY294003 used, P-MAP kinase and P-Akt was effective blocked, respectively (data not shown). However, there was no direct correlation between the potency of anchorage independent growth or invasion activity with the expression level or activation status of the RPTKs including ErbB2 or their downstream signaling molecules examined.
Inhibition of monolayer and anchorage-independent growth of breast cancer cells by rapamycin and herceptin
The inhibitory effect of rapamycin and herceptin individually or in combination on the growth of ErbB2 overexpressing and non-ErbB2 overexpressing breast cancer cell lines was assessed. A dosage with targeted 50% inhibition by herceptin or rapamycin was used with an intention to detect significant additive effect by combination of the drugs. Under the dosage of herceptin used, different lines displayed distinct sensitivity for the inhibition of growth and colony formation (Figs. 4a and 4b). The monolayer growth of ErbB2 overexpressing lines, BT474, HCC1419 and HCC1954, was more sensitive than the non-ErbB2 overexpressing lines, Hs578T, MCF7 and MDA-MB-231, toward the inhibition of rapamycin and especially the combination of herceptin and rapamycin (Fig. 4a). The non-ErbB2 overexpressing lines were relatively resistant to the herceptin inhibition. In all cases except MCF7, combination of rapamycin and herceptin gave a significantly greater inhibition than that of herceptin alone (Fig. 4a). However, the difference between rapamycin alone and the combination of the drugs was not statistically significant in all but BT474, although the trend of greater inhibition with the combined drugs is consistent for all lines tested. Those lines were further tested for inhibition on anchorage-independent growth. The results showed that the colony forming activity of all 3 ERbB2 overexpressing lines was inhibited by rapamycin or herceptin and the combination of both gave a significantly greater inhibition than that of either alone (Fig. 4b,upper panel). Although the inhibition of HCC1419 in terms of number of colonies was not as significant as that for BT474 or HCC1954, the size distribution of the colonies showed that significant inhibition of colony size was observed with rapamycina and especially with the combination of rapamycin and herceptin (Fig. 4b, lower panel).
The effect of herceptin on ErbB2-mediated signaling was examined. Although no significant inhibition on heregulin-induced tyrosine phosphorylation of ErbB2 and Akt was observed possibly because of the high basal level (data not shown), herceptin significantly inhibited heregulin-induced phosphorylation of ERK in BT474, HCC1419 and HCC1954 cells (Fig. 4c).
Rapamycin and herceptin reduced cyclin D1, cyclin D3 and increased caspase 3 cleavage
To explore possible mechanism for the added growth inhibition of rapamyicn and herceptin, their effect on cell cycle and apoptosis regulatory proteins was analyzed. The result shows that rapamycin significantly inhibited the cyclin D1 and D3 level and increased the caspase 3 cleavage and that herceptin enhanced the effect when added together (Fig. 4d).
Inhibition of rapamycin and herceptin on tumorgenicity
Next, we investigated the effect of rapamycin by itself or in combination with herceptin on tumorigenicity. In a xenograft model MCF7-I4 cells were inoculated into BALB/C athymic mice. Subsequently the mice were treated with control vehicle solution, rapamycin, herceptin or combination of rapamycin and herceptin. The results showed that rapamycin or herceptin alone significantly inhibited the growth of tumors. Combination of the 2 drugs dramatically reduced the growth of tumors (Fig. 5a).
Next the FVB/N-TGN (CMMTVneu) mice harboring the unactivated ErbB2 gene under the control of the MMTV LTR promoter11 were used to further assess the inhibition of tumorigenicity by rapamycin and herceptin. It was reported that 50% of the female transgenic mice developed tumor at 205 days of age.11 In our experiment, the earlier onset of overtly visible tumors was observed in control, rapamycin and herceptin groups at 180 days after birth. Drug treatment was started 140 days after birth. At 248 days after birth 50% of the control group, 40% of the herceptin treated group and 10% of the rapamycin plus herceptin treated group developed overtly visible tumors. The tumor growth was significantly inhibited by herceptin and rapamycin and especially by the combined treatment (Fig. 5b). All mice except one in the combined treatment group were alive at the end point which was set at 280 days after birth. Two-third of the tumor-bearing mice in the control group carried multiple tumors at the endpoint, whereas only one-third and one-fifth of the tumor-bearing mice in the herceptin and rapamycin groups, respectively, carried multiple tumors, none of the tumor-bearing mice in the combined drug treated group carried multiple tumors (Table I). The average tumor weight per mouse of the rapamycin, herceptin and combined drug treatment group was 28%, 56% and 16%, of that of the control group, respectively. (Table I). No detectable adverse toxicity effect on the treated mice was observed for all the treatment groups as monitored by body weight and activity of the mice. The results from the xenograft and transgenic models demonstrate that rapamycin significantly inhibits the tumor growth and that combination of herceptin and rapamycin has significantly enhanced anti-tumorigenicity over that of either treatment alone.
|Treatment||Tumor incidence (%)||Total tumor weight (g)||Average tumor weight (g)||Mice with multiple tumors (%)||Average no. of tumor/mouse|
|Control||58 (7/13)||15.8||1.21||46 (6/13)||1.3|
|Rapamycin||50 (5/10)||3.4||0.34||20 (2/10)||0.8|
|Herceptin||44 (4/9)||6.8||0.76||33 (3/9)||0.9|
|Rapamycin/herceptin||40 (4/10)||1.9||0.19||0 (0/10)||0.4|
The study described above confirms and extends our previous observation that ErbB2 over expressing human breast cancer cells are hypersensitive toward inhibition by inhibitors of the PI3 kinase pathway including rapamycin.28 This is reminiscent of another previous observation of ours that inhibition of MAP kinase selectively inhibits cell proliferation of breast cancer cells with enhanced IGF-1R-mediated MAP kinase activation.48 In this study, we demonstrates the enhanced growth inhibition of ErbB2 over expressing breast cancer cells by combination of rapamycin and herceptin, and for the first time, shows an increased tumor inhibitory effect of rapamycin and herceptin, when treated together, against the mammary malignancy in an ErbB2 transgenic and in a xenograft model. This suggests that combined treatment of rapamycin or its analogs with herceptin maybe developed as an improved therapeutic regimen for breast cancer especially those with ErbB2 over expression.
Our results are in agreement with previous reports that rapamycin or its analog CCI-779 was able to inhibit tumor growth in a multiple myeloma xenograft model49 and in an activated ErbB2-bearing transgenic model.50 The latter study suggested that the anti-tumor activity of rapamycin could be due to decreased ErbB3 expression and inhibition of hypoxia-inducible factor 1-dependent responses to hypoxic stress.50 In a Phase II clinical trial with heavily pre-treated breast cancer patients, treatment with CCI-779 was reported to attain a 9.2% objective response with a tolerable dosage of the drug.51 There have been also a number of reports on anti-tumor effect with combined treatment of rapamycin and other cancer drugs in the in vitro and in vivo studies.52, 53, 54, 55, 56 CCI-779 together with cisplatin was reported to significantly inhibit the tumor growth in a medulloblatoma xenograft model.52 It was shown that rapamycin synergistically interacted with paclitaxel and carboplatin in inducing cytotoxicity of breast cancer cells, including ErbB2- over expressing cells, in vitro and tumor growth in xenograft models.53 Combination of rapamycin with inhibitors of estrogen receptor such as letrozole and 4-OH tamoxifen has also been tried and shown to have an additive effect on growth inhibition and apoptosis in ovarian and breast cancer cells.54, 55 Similarly, combined treatment with rapamycin and cotylenin A, a differentiation inducing agent, significantly inhibited the growth of MCF7 cells in vitro and in xenografts.56 However, to our knowledge, no combined treatment of rapamycin and herceptin in the inhibition of tumor growth in transgenic or xenograft models has been reported so far. Our observation underscores the clinical potential of using this combined therapeutic regimen for the treatment of ErbB2 over expressing breast cancer.
As an inhibitor of global translation, it is not surprising that treatment with rapamycin could lead to growth arrest of cells most likely because of effect on those proteins with a shorter half life and are important for cell cycle progression. However, the mechanism for rapamycin-mediated sensitization of cancer cells for growth inhibition and apoptosis induced by other cancer drugs is not fully understood, and is likely to be variable depending on the types of cells and therapeutic agents used. An effect of rapamycin on the level of factors critical for cell cycle progression and survival has been suggested to underlie the basis for sensitization of breast cancer cells for growth inhibition and apoptosis.57, 58, 59, 60 Rapamycin and its analogs, CCI-779 and RAD001, have been shown to inhibit cyclin D1, D3 and p21 level.57, 58, 59, 60 Our observation of the inhibition of cyclins D1 and D3 by rapamycin and herceptin are consistent with those reports. It was suggested that inhibition of p53-induced p21 expression by rapamycin led to sensitization of A549 non-small cell lung cancer cells for DNA damaging agent-induced apoptosis of the cells.59 A recent study showed that rapamycin-mediated anti-proliferative effect correlated with down regulation of p21 level and blockage of the cyclin D/p21/PCNA complexes formation needed for cell proliferation in breast cancer cells.60 The reduced cyclin D/p21/PCNA complexes formation was also observed in tumors of RAD001 treated transgenic mice bearing MMTV-c-Neu.60 In another study, rapamycin was shown to block m-TOR-mediated inhibition of protein phosphatase 2A, which could activate proapoptotic protein Bad, and thus to sensitize breast cancer cells for fostriecin and cyclosporine-induced apoptosis.61 Our observation of enhanced caspase 3 cleavage by rapamycin and herceptin is in agreement with that report and provide a possible explanation for the observed enhanced antigrowth and anti-tumor efficacy.
Herceptin used in this study was provided by Genentech through a material transfer agreement.
- 46Effect of tyrosine mutations on the kinase activity and transforming potential of an oncogenic human insulin-like growth factor I receptor. J Biol Chem 1996; 5:271: 160–7., , ,
- 48Inhibition of mitogen-activated protein kinase kinase selectively inhibits cell proliferation in human breast cancer cells displaying enhanced insulin-like growth factor I-mediated mitogen-activated protein kinase activation. Cell Growth Differ 2000; 11: 655–64., ,