Insulin receptor substrates mediate distinct biological responses to insulin-like growth factor receptor activation in breast cancer cells

Abstract

Activation of the type I insulin-like growth factor receptor (IGF-IR) regulates several aspects of the malignant phenotype, including cancer cell proliferation and metastasis. Phosphorylation of adaptor proteins downstream of IGF-IR may couple IGF action to specific cancer phenotypes. In this study, we sought to determine if insulin receptor substrate-1 and -2 (IRS-1 and -2) mediate distinct biological effects in breast cancer cells. Insulin receptor substrate-1 and IRS-2 were expressed in T47D-YA breast cancer cells, which lack IRS-1 and -2 expression, yet retain functional IGF-IR. In the absence of IRS-1 and -2 expression, IGF-IR activation was unable to stimulate proliferation or motility in T47D-YA cells. Expression of IRS-1 resulted in IGF-I-stimulated proliferation, but did not affect motility. In contrast, expression of IRS-2 enhanced IGF-I-stimulated motility, but did not stimulate proliferation. The alphaIR-3, an inhibitor of the IGF-IR, was unable to affect these IGF-stimulated phenotypes unless IRS-1 or -2 was expressed. Thus, IGF-IR alone is unable to regulate important breast cancer cell phenotypes. In these cells, IRS proteins are required for and mediate distinct aspects of IGF-IR-stimulated behaviour. As multiple agents targeting the IGF-IR are currently in early clinical trials, IRS expression should be considered as a potential biomarker for IGF-IR responsiveness.

Figure 1

T47D-YA cells lacked IGF-I-stimulated proliferation and motility, despite expression of a functional IGF-IR. (A) T47D-WT, T47D-Y, T47D-YA and T47D-YB cells were serum-deprived for 24 h, stimulated with 5 nM IGF-I for 10 min, cell lysates separated by 8% SDS-PAGE and protein levels of the IGF-IR determined by Western blot. (B) T47D cell line variants were serum-deprived for 24 h, stimulated for 10 min with 5 nM IGF-I, and cell lysates collected, immunoprecipitated with IGF-IR antibody, and analysed by immunoblotting with an antibody for phosphorylated tyrosine residues. (C) T47D-Y and T47D-YA cells were serum-starved for 24 h, stimulated with 5 nM IGF-I for 10 min, cell lysates separated by 8% SDS-PAGE and immunoblotted with anti-IRS-2 antibody. MDA-MB-231BO cells were included as a positive control for IRS-2 expression. (D) T47D-WT, T47D-CO, T47D-Y, T47D-YA and T47D-YB cells were stimulated with 5 nM IGF-I for 10 min, cell lysates separated by 8% SDS-PAGE and immunoblotted with anti-IRS-1 antibody. Experiments were performed two to four times and representative data shown. (E) T47D-WT and T47D-YA cells were plated in 24-well plates in growth media, switched to SFM after 24 h, treated with or without 5 nM IGF-I for 6 days, and then cell number estimated using an MTT assay. Error bars represent s.e. of the mean and ^* represents a significant difference (P<0.05) in absorbance in samples treated with IGF-I compared to SFM. P-values: T47D-WT (P=0.0409) Results shown are representative of three independent experiments. (F) T47D-WT and T47D-YA cells were plated on gold particle-coated coverslips. MDA-MB-231BO cells were included as a positive control. The cells were allowed to adhere, treatment of SFM with or without 5 nM IGF-I added, and cells incubated at 37° C for 24 h. Coverslip images were captured using a brightfield microscope with a neutral density filter and the area on the coverslip cleared by cell movement was computed using Simple PCI software. Data are presented as mean area cleared. Error bars represent s.e. of the mean and ^* represents a significant difference (P<0.05) in samples treated with IGF-I compared to SFM. P-value: T47D-WT (P<0.001); MDA-MB-231BO (P=0.0072). Results shown are representative of three independent experiments.

Figure 2

IRS-1 and -2 were stably expressed in T47D-YA cells. (A) T47D-YA cells were stably transfected with a cDNA construct encoding HA-tagged human IRS-1. Insulin receptor substrate-1 expression was determined by immunoblot. (B) T47D-YA cells were stably transfected with a cDNA construct encoding HA-tagged human IRS-2. Insulin receptor substrate-2 expression was determined by immunoblot.

Figure 3

IRS-1, but not IRS-2, expression resulted in IGF-I mediated cell proliferation in T47D-YA cells. (A) T47D-YA/IRS-1 cell clones and T47D-YA/IRS-2 cell clones were plated in 24-well plates in serum-containing media, switched to SFM after 24 h, and treated with or without 5 nM IGF-I. After 6 days, cell number was estimated using an MTT assay. Data are represented as fold increase over each cell line's SFM readings. Error bars represent s.e. of the mean and ^* represents a significant difference (P<0.05) in absorbance in samples treated with IGF-I compared to SFM. P-values: T47D-YA/IRS-1 #5 (P=0.0013); T47D-YA/IRS-1 #8 (P=0.0265); T47D-YA/IRS-1 #10 (P=0.0107); T47D-YA/IRS-1 #20 (P=0.0247); T47D-WT (P=0.0055). Data are representative of three independent experiments. (B) T47D-YA/IRS-1 and T47D-YA/IRS-2 cells were plated in 60 mm dishes in serum-containing media and, after 24 h, switched to SFM for 48 h. Cells were then treated with or without 5 nM IGF-I in SFM for 24 h, stained with propidium iodide, and analysed for phase of the cell cycle by flow cytometry. Error bars represent s.e. of the mean and ^* represents a significant difference (P<0.05) in % of cells in S+G₂M phases of the cell cycle in samples treated with IGF-I compared to SFM. P-values: T47D-WT (P=0.045), T47D-YA/IRS-1 #5 (P=0.020), T47D-YA/IRS-1 #10 (P=0.0317), T47D-YA/IRS-1 #20 (P=0.0483). (C) A bottom agar was prepared and overlaid with 800 μl of a 0.45% top agar mixture containing 10 000 cells per well in the presence or absence of 5 nM IGF-I treatment and plates incubated at 37°C. After 14 days, colonies formed in the soft agar assay were counted using a light microscope with an ocular grid. Only colonies larger than two-thirds of a grid square were counted. Five random fields were counted for each well and the average number of colonies per well calculated. Results are presented as fold increase over each cell line's SFM values. Results are representative of three experiments performed in triplicate for each treatment. Error bars represent s.e. of the mean and ^* represents a significant difference (P<0.05) in samples treated with IGF-I compared to SFM. P-values: T47D-YA/IRS-1 #5 (P=0.0078), T47D-YA/IRS-1 #8 (P=0.0086), T47D-YA/IRS-1 #10 (P=0.0119), T47D-YA/IRS-1 #20 (P=0.0034).

Figure 4

Expression of IRS-2, but not IRS-1, resulted in IGF-I-stimulated motility in T47D-YA cells. T47D-YA/IRS-1 and T47D-YA/IRS-2 cells were plated on gold particle-coated coverslips. The cells were allowed to adhere, treatment of SFM with or without 5 nM IGF-I added, and cells incubated at 37°C for 24 h. Coverslip images were captured using a brightfield microscope with a neutral density filter and the area on the coverslip cleared by cell movement was computed using Simple PCI software. Data are presented as mean area cleared. Error bars represent s.e. of the mean and ^* represents a significant difference (P<0.05) in samples treated with IGF-I compared to SFM. P-values: YA/IRS-2 #6 (P=0.0224); YA/IRS-2 #10 (P=0.0481); Results are representative of three independent experiments.

Figure 5

Downregulation of IRS-1 inhibited response to IGF-I in MCF-7 cells. (A) MCF-7 breast cancer cells were transfected with IRS-1, IRS-2 or nonfunctional (NF) control siRNA constructs. After 24 h, cells were lysed and examined for IRS-1 expression by immunoblotting. Total MAPK was used as a loading control. (B) Effect of IRS-1 downregulation on IGF-I-stimulated growth was examined in 96-well plates. Cells were transfected with a NF, IRS-1 or IRS-2 siRNA for 6 h. Non-transfected (NT) cells were also examined. After incubation in SFM overnight, cells were treated in the absence (open bar) or presence (black bar) of 5 nM IGF-I. Cell number was estimated at this time point (day 0) by MTT assay. After 48 h, cell numbers were again determined and results are presented as percent increase over day 0. These treatments were carried out in triplicate and results were repeated. A representative experiment is shown. Statistically significant differences are shown.

Figure 6

IRS-1 and -2 selectively coupled the IGF-IR to downstream pathways. (A) T47D-YA/IRS-1 cells were stimulated with 5 nM IGF-I for 10 min and cell lysates analysed for activated IRS-1 by immunoprecipitating IRS-1, followed by immunoblotting for phosphorylated tyrosine residues. (B) T47D-YA/IRS-2 cells were stimulated with 5 nM IGF-I for 10 min and cell lysates analysed for activated IRS-2 by immunoprecipitating IRS-2, followed by immunoblotting for phosphorylated tyrosine residues. (C) T47D-YA/IRS-1 cells were examined by immunoblot for MAPK and PI3K activation using phosphorylation specific antibodies following treatment with 5 nM IGF-I for 10 min. (D) T47D-YA/IRS-2 cell clones were examined by immunoblot for MAPK and PI3K activation after treatment with 5 nM IGF-I for 10 min. Experiments were repeated a minimum of three times, and representative blots shown.

Figure 7

IRS-1 and -2 were required for sensitivity to anti-IGF-IR strategies. (A) T47D-YA, T47D-YA/IRS-1, and T47D-YA/IRS-2 cells were plated in triplicate in 24-well tissue culture plates at a density of 10 000 cells per well in growth media. After 24 h, cells were washed twice with 1 × PBS and switched to SFM for 24 h. Cells were then treated with SFM, 3 μg ml⁻¹ αIR-3, 5 nM IGF-I or 3 μg ml⁻¹ αIR-3 plus 5 nM IGF-I for 5 days and cell number estimated using the MTT assay. Data are represented as fold increase over each cell line's SFM readings. Error bars represent s.e. of the mean and ^* represents a significant difference (P<0.05) in absorbance in samples treated with IGF-I compared to SFM. P-values: T47D-YA/IRS-1 #20 (P=0.0141). Data are representative of three independent experiments. (B) T47D-YA, T47D-YA/IRS-1 and T47D-YA/IRS-2 cells were plated on gold particle-coated coverslips. Cells were allowed to adhere and then treated with SFM, 3 μg ml⁻¹ αIR-3, 5 nM IGF-I or 3 μg ml⁻¹ αIR-3 plus 5 nM IGF-I for 24 h. Coverslip images were captured using a brightfield microscope with a neutral density filter and the area on the coverslip cleared by cell movement was computed using Simple PCI software. Data are presented as fold increase of the mean area cleared compared to each cell line's SFM readings. Error bars represent s.e. of the mean and ^* represents a significant difference (P<0.05) in samples treated with IGF-I compared to SFM. P-values: YA/IRS-2 #10 (P=0.0169). Results are representative of three independent experiments.