Inhibition of insulin receptor isoform-A signalling restores sensitivity to gefitinib in previously de novo resistant colon cancer cells

Abstract

Resistance to antiepidermal growth factor (EGFR) strategies is an emerging clinical problem. Using human colorectal cancer (CRC) cells, we evaluated the involvement of the insulin receptor isoform-A (InsR-A) in de novo resistance to gefitinib, an EGFR tyrosine kinase inhibitor. Challenging the EGFR positive LoVo cells with gefitinib (1 microM) resulted in a small ( approximately 18%) inhibition of cell growth and although a modest reduction in phospho (p)EGFR Tyr845 was seen, pEGFR at residues -Tyr1068 and -Tyr1173 were unchanged. LoVo cells produced unprocessed pro-IGF-1R protein, substantial levels of IGF-II mRNA and mature InsR protein, consisting mainly of the InsR-A isoform. Insulin and IGF-II promoted cell growth and pEGFR Tyr845, Tyr1068 and Tyr1173 activity and conversely, the insulin-like growth factor-1 receptor (IGF-1R)/InsR inhibitor ABDP (1 muM) inhibited growth and reduced pEGFR activity at all three tyrosine residues. pInsR and pAkt levels were increased after gefitinib treatment. Blocking of pInsR with ABDP enabled gefitinib to markedly reduce pEGFR Tyr845, Tyr1068 and Tyr1173. Short-term gefitinib/ABDP dual treatment was more effective than either agent alone and chronic exposure to this combination resulted in total cell loss after 9 weeks, preventing acquisition of resistance to ABDP. LoVo cells with acquired resistance to ABDP were acutely sensitive to gefitinib. We concluded that InsR-A reduces sensitivity to gefitinib in LoVo CRC cells, thus its co-targeting alongside EGFR can improve the anti-tumour effect of gefitinib.

Figure 1

Response to gefitinib and basal EGFR expression and phosphorylation status. (A) LoVo cells were grown in DCCM-1 containing varying serum concentrations (0–10%) in the absence and presence of 1 μM gefitinib. Data are mean values of three independent experiments (every point in each individual experiment also being evaluated in triplicate) and error bars represent 95% CIs. Significant differences were assessed at the P<0.05 level between the growth inhibition value obtained in the presence of 10% serum and gefitinib and the growth inhibitory values seen with each of the other serum concentrations in the presence of gefitinib. (B) Cells were grown in DCCM-1 with 0.5% serum in the absence and presence of 1 μM gefitinib for 7 days. Protein (75 μg) of cell lysate was electrophoresed by SDS–PAGE (7.5%) and immunoblotted for total EGFR, phospho (p)-EGFR Tyr845, pEGFR Tyr1068, pEGFR Tyr1173 and pERK1/2. Densitometric analysis was performed and results were normalised to β-actin levels. The data illustrated are representative of three separate experiments. (C) After growing to 70% confluency and serum starvation for 24 h, cells were challenged with EGF (5 min). Samples were electrophoresed and immunoblotted for total EGFR, pEGFR Tyr1068, pEGFR Tyr1173, total Akt, pAkt, total ERK1/2 and pERK1/2 and analysed as detailed in (B). (D) Cells were harvested for RNA, RT–PCR was performed and the resulting cDNA was amplified using primer sets for TGF-α. Fragments were resolved on agarose gels and densitometric scores were normalised to β-actin. Data represents mean values of three experiments and error bars indicating 95% CIs.

Figure 2

Evaluation of IGF-1R and InsR signalling pathway components. LoVo, A549, DU145 and MCF-7 cells were grown in DCCM-1 with 0.5% serum for 7 days. (A) Protein (75 μg) of cell lysate was electrophoresed by SDS–PAGE (7.5%) and immunoblotted for total IGF-1R and InsR. (B) Cells were harvested for RNA, RT–PCR was performed and the resulting cDNA was amplified using primer sets for IGF-II. Fragments were resolved on agarose gels and densitometric scores were normalised to β-actin. Data represents mean values of three experiments and error bars indicating 95% CIs. ^*Shows significant differences assessed between LoVo cells and each of the other cell lines at the P<0.05 level. (C) Amplified InsR-A and InsR-B cDNA was resolved by PAGE (15%) and normalised to β-actin. (D) Following PCR amplification for InsR-A and InsR-B, the PCR products were digested with Ban I restriction enzyme, before resolution by PAGE. Data illustrates nondigested (ND) and digested (D) cDNA fragments in LoVo and DU145 samples. (E) Cells were grown in DCCM-1 with 0.5% serum in the absence and presence of 1 μM gefitinib for 7 days. Protein (75 μg) of cell lysate was electrophoresed by SDS–PAGE (7.5%) and immunoblotted for total and phosphorylated InsR and Akt.

Figure 3

Growth responses of LoVo cells to mitogenic growth factors and the InsR/IGF-1R inhibitor ABDP. (A) LoVo cells were challenged with EGF, IGF-II (both at 10 ng ml⁻¹) and insulin (10 μg ml⁻¹) for 7 days in DCCM-1 serum-free medium. Data are mean values of three independent experiments (every point in each individual experiment also being evaluated in triplicate) and error bars represent 95% CIs. ^*Significant differences were assessed at the P<0.05 level. (B) LoVo cells were grown in the absence and presence of varying concentrations of ABDP (0–5 μM) in DCCM-1 with 0.5% serum for 7 days. Data are mean values of three independent experiments (every point in each individual experiment also being evaluated in triplicate) and error bars represent 95% CIs. ^*Significant differences were assessed at the P<0.05 level.

Figure 4

Modulation of pEGFR activity by insulin and IGF-II and the InsR/IGF-1R inhibitor ABDP. (A) LoVo cells were grown in routine culture medium until 70% confluent and after 24 h in serum-free DCCM-1 were challenged with IGF-II (10 ng ml⁻¹) and insulin (10 μg ml⁻¹) for 5 min. Cell lysates (75 μg) were electrophoresed by SDS–PAGE (7.5%) and immunoblotted for total InsR, pInsR Tyr1146, total EGFR, pEGFR Tyr845, pEGFR Tyr1068, pEGFR Tyr1173. Densitometric analysis was performed and results were normalised to β-actin levels. The data illustrated are representative of three separate experiments. (B) LoVo cells were grown in the absence and presence of 1 μM ABDP in DCCM-1 with 0.5% serum for 7 days and electrophoresed, immunoblotted for pEGFR Tyr845, pEGFR Tyr1068, pEGFR Tyr1173 and analysed as described in (A). (C) LoVo cells were grown in routine culture medium until 70% confluent and after 24 h in DCCM-1, were challenged with insulin (10 μg ml⁻¹) and EGF (10 ng ml⁻¹) with and without 1 μM ABDP for 5 min. Cells exposed to ABDP were preincubated with this inhibitor for 6 h before electrophoresis, immunoblotting for total InsR, pInsR Tyr1146, pEGFR Tyr845, pEGFR Tyr1068, pEGFR Tyr1173 and analysis as described in (A).

Figure 5

Gefitinib and ABDP in combination are more effective than either agent alone. (A) LoVo cells were cultured in varying concentrations of ABDP (0–1 μM) in the absence (filled bar) or presence (spotted bar) of 1 μM gefitinib in DCCM-1 with 0.5% serum for 7 days. Data are mean values of three independent experiments (every point in each individual experiment also being evaluated in triplicate) and error bars represent 95% CIs. ^*Significant differences were assessed at the P<0.05 level. (B) LoVo cells were grown in DCCM-1 with 0.5% serum containing 1 μM ABDP for 4 days and subsequently challenged with ABDP and gefitinib in combination for 24 h. Cell lysate (75 μg) was electrophoresed by SDS–PAGE (7.5%) and immunoblotted for total EGFR, pEGFR Tyr845, pEGFR Tyr1068, pEGFR Tyr1173, pAkt and pERK1/2. Data was assessed by densitometry and are representative of three separate experiments. (C) LoVo cells were chronically exposed to control media i.e. DCCM-1 with 0.5% serum (-♦-), 1 μM gefitinib (^…□^…), 1 μM ABDP (-▴-) or gefitinib/ABDP dual treatment (-•-) until cell loss occurred or resistance developed. The data demonstrates the total passage number with respect to time (weeks).

Figure 6

LoVo-ABDP-R cells have gained sensitivity to gefitinib. (A) LoVo-ABDP-R cells were challenged with EGF (10 ng ml⁻¹), gefitinib (1 μM) and EGF/gefitinib together for 7 days in DCCM-1 serum-free medium containing 1 μM ABDP. Data are mean values of three independent experiments (every point in each individual experiment also being evaluated in triplicate) and error bars represent 95% CIs. ^*Significant differences were assessed at the P<0.05 level. (B) LoVo cells were grown in DCCM-1 with 0.5% serum for 7 days and LoVo-ABDP-R cells were cultured in the absence and presence of 1 μM gefitinib for 7 days in DCCM-1 supplemented with 0.5% serum and 1 μM ABDP for 7 days. Protein (75 μg) of cell lysate was electrophoresed by SDS–PAGE (7.5%) and immunoblotted for total EGFR, pEGFR Tyr845, pEGFR Tyr1068, pEGFR Tyr1173, total ERK1/2, pERK1/2, total Akt and pAkt. Densitometric analysis was performed and results were normalised to total expression levels. The data illustrated are representative of three separate experiments.