Combinatorial targeting of FGF and ErbB receptors blocks growth and metastatic spread of breast cancer models

Abstract

Introduction: Targeting receptor tyrosine kinases (RTKs) with kinase inhibitors is a clinically validated anti-cancer approach. However, blocking one signaling pathway is often not sufficient to cause tumor regression and the effectiveness of individual inhibitors is often short-lived. As alterations in fibroblast growth factor receptor (FGFR) activity have been implicated in breast cancer, we examined in breast cancer models with autocrine FGFR activity the impact of targeting FGFRs in vivo with a selective kinase inhibitor in combination with an inhibitor of PI3K/mTOR or with a pan-ErbB inhibitor.

Methods: Using 4T1 or 67NR models of basal-like breast cancer, tumor growth was measured in mice treated with an FGFR inhibitor (dovitinib/TKI258), a PI3K/mTOR inhibitor (NVP-BEZ235) or a pan-ErbB inhibitor (AEE788) individually or in combination. To uncover mechanisms underlying inhibitor action, signaling pathway activity was examined in tumor lysates and transcriptome analysis carried out to identify pathways upregulated by FGFR inhibition. Anti-phosphotyrosine receptor antibody arrays (P-Tyr RTK) were also used to screen 4T1 tumors.

Results: The combination of dovitinib + NVP-BEZ235 causes tumor stasis and strong down-regulation of the FRS2/Erk and PI3K/Akt/mTOR signaling pathways. P-Tyr RTK arrays identified high levels of P-EGFR and P-ErbB2 in 4T1 tumors. Testing AEE788 in the tumor models revealed that the combination of dovitinib + AEE788 resulted in blockade of the PI3K/Akt/mTOR pathway, prolonged tumor stasis and in the 4T1 model, a significant decrease in lung metastasis. The results show that in vivo these breast cancer models become dependent upon co-activation of FGFR and ErbB receptors for PI3K pathway activity.

Conclusions: The work presented here shows that in the breast cancer models examined, the combination of dovitinib + NVP-BEZ235 or dovitinib + AEE788 results in strong inhibition of tumor growth and a block in metastatic spread. Only these combinations strongly down-regulate the FGFR/FRS2/Erk and PI3K/Akt/mTOR signaling pathways. The resultant decrease in mitosis and increase in apoptosis was consistently stronger in the dovitinib + AEE788 treatment-group, suggesting that targeting ErbB receptors has broader downstream effects compared to targeting only PI3K/mTOR. Considering that sub-classes of human breast tumors co-express ErbB receptors and FGFRs, these results have implications for targeted therapy.

Figure 1

Effect of fibroblast growth factor receptor (FGFR) and phosphatidyl inositol 3'kinase (PI3K)/mammalian target of rapamycin (mTOR) inhibitors on signaling and tumor outgrowth. T1 (A) or 67NR (B) cell cultures were treated for 1 hr with 0.5 μM NVP-BEZ235 or control dimethyl sulfoxide (DMSO)-containing medium, lysates were prepared and a western analysis for the indicated proteins and phospho-proteins was performed. (C) Starting 7 days after injection of 4T1 cells in mammary fat pads, groups of tumor-bearing mice (n = 6) were treated for 14 days with vehicle (PEG300), dovitinib (TKI, 20 mg/kg), NVP-BEZ235 (10 mg/kg) or a combination of both, and tumor volume was determined; representative of two different experiments. (D) 4T1 tumor-bearing mice were treated with a single dose of the vehicle (PEG300), dovitinib (TKI, 20 mg/kg), NVP-BEZ235 (10 mg/kg) or a combination of both (TKI + BEZ) and sacrificed 2 hrs later. Tumor lysates were prepared from three independent mice per group and a western analysis for the indicated proteins and phospho-proteins (P) was performed. (E) Quantification of metastatic foci number in the lung tissue taken from animals at the end of the experiment in panel C. N = 6, representative of two separate experiments. *P < 0.05 **P < 0.01 (Mann-Whitney U-test).

Figure 2

Analyses of 4T1 tumors and lung metastases in mice treated with dovitinib, NVP-BEZ235 or the combination. (A) 4T1 tumors from mice treated for 14 days with the inhibitors described in Figure 1C were harvested, and frozen sections were stained for proliferation (phospho-histone H3; blue), apoptosis (cleaved caspase 3; red) and endothelial cells (CD31; green). Scale bars, 100 μm. (B) Phosphorylated histone-3 (PH3)-positive cells were manually counted and expressed as the average number of cells per field. Proportion of tumor area expressing cleaved Caspase-3 and CD31 immunoreactivity was quantified using ImageJ. N = 3 to 4 tumors for each group with 5 fields analyzed per tumor (15 to 20 images from each treatment). (C) 2.5 × 10⁵ 4T1 cells were injected into tail veins and indicated treatments were started 7 days later for a total of 11 days. Lungs were harvested and metastases were quantified. N = 5, representative of two separate experiments. (D) After 14 days treatment with the combination of dovitinib (TKI, 20 mg/kg) and NVP-BEZ235 (10 mg/kg) or with vehicle, 4T1-tumor bearing mice were sacrificed, and blood was collected and plated. The graph shows the number and SD of tumor cell colonies growing/mouse/ml of blood; N = 5 to 9 animals per group. *P < 0.05 **P < 0.01 (Mann-Whitney U-test).

Figure 3

4T1 and 67NR tumors retain prolonged sensitivity to treatment with the dovitinib + NVP-BEZ235 combination. (A) Groups of 67NR tumor-bearing mice (n = 5) were treated daily with vehicle (PEG300), or with a combination of dovitinib (TKI, 20 mg/kg) and NVP-BEZ235 (10 mg/kg) for 7 days. Treatment was stopped for 5 days then resumed for 6 days and tumor growth was monitored. (B) Groups of 4T1 tumor-bearing mice (n = 6) were treated daily with vehicle (PEG300) or the combination of dovitinib (TKI, 20 mg/kg) + NVP-BEZ235 (10 mg/kg). After 10 days treatment, residual tumors were enzymatically digested, hematopoietic cell-depleted and 0.5 × 10⁶ were re-injected into naïve mice (scheme in top panel). Starting 7 days after injection, groups of mice (n = 5) were treated daily for 11 days with (PEG300), dovitinib (TKI, 20 mg/kg), NVP-BEZ235 (10 mg/kg) or a combination of both, and tumor growth was monitored (lower portion of B). *P < 0.05 (Mann-Whitney U-test).

Figure 4

Anti-phosphotyrosine receptor antibody (P-Tyr RTK) array analysis of 4T1 cells and tumors. (A-B) P-RTK analysis on lysates of 4T1 cultures either growing in low serum (Vehicle) or stimulated for 15 minutes with 20 ng/ml EGF, 12 nM HRG or 20 ng/ml platelet-derived growth factor (PDGF) (Ligands). Lysates from individual ligand-treated cells were pooled prior to analysis. (B) Quantification of the signal intensity of the indicated p-RTKs is shown in comparison to positive control, set as 1. (C-D) Three control (Vehicle) and 50 mg/kg dovitinib-treated (TKI258) 4T1 tumors were harvested after 3 days treatment, and lysates were pooled and analyzed on P-RTK filters. (D) Quantification of the intensity of signal of indicated p-RTKs is shown, in comparison to positive control set as 1.

Figure 5

Effect of dovitinib, AEE788 or the combination of both on 4T1 and 67NR tumors. (A) Groups of 4T1 tumor-bearing mice (n = 6) were treated for 12 days with PEG300 (Vehicle), AEE788 (50 mg/kg) or a combination of dovitinib (TKI, 40 mg/kg) + AEE788. AEE788 was administered 3×/week; dovitinib daily, and tumor volume was determined; representative of two experiments. (B) Groups of 67NR tumor-bearing mice (n = 5) were treated with PEG300 (Vehicle), AEE788 (50 mg/kg) or a combination of dovitinib (TKI, 40 mg/kg) + AEE788. Treatment was performed through days 7 to 13 and days 18 to 23. Dovitinib and vehicle were dosed daily; AEE788 was administered on days 7, 9, 11 and 13, then on day 18, 20 and 22, and tumor volume was determined. (C) Quantification of the number of metastatic foci covering lungs of mice from the experiment in Panel A. N = 6, representative of two separate experiments. (D) 4T1 tumor-bearing mice were treated with PEG300, a single dose of AEE788 (50 mg/kg) or a combination of dovitinib (TKI, 40 mg/kg) + AEE788, then sacrificed 2 hrs later. Tumor lysates were prepared from three mice per group and a western analysis for the indicated proteins and phospho-proteins (P) was performed. *P < 0.05 (Mann-Whitney U-test).

Figure 6

Analysis of 4T1 tumors from mice treated with AEE788 + dovitinib. (A) Frozen sections from 4T1 tumors harvested after 14 days of treatment as described in Figure 5A were stained for proliferation (phospho-histone H3; blue), apoptosis (cleaved caspase 3; red) and endothelial cells (CD31; green). Scale bars, 100 μm. (B) Phosphorylated histone-3 (PH3)-positive cells were manually counted and expressed as the average number of cells per field. Proportion of tumor area expressing cleaved Caspase-3 and CD31 immunoreactivity were quantified using ImageJ. N = 3 to 4 tumors for each group with five fields analyzed per tumor (15 to 20 images from each treatment). Bars represent mean ± SD in each treatment group. (C) 2.5 × 10⁵ 4T1 cells were injected into tail veins and indicated treatments were started 7 days later for a total of 11 days. Lungs were harvested and metastases were quantified. N = 5, representative of two separate experiments. *P < 0.05; **P < 0.01 (Mann-Whitney U-test).