Autocrine IGF-I/insulin receptor axis compensates for inhibition of AKT in ER-positive breast cancer cells with resistance to estrogen deprivation

Abstract

Introduction: Estrogen receptor α-positive (ER+) breast cancers adapt to hormone deprivation and acquire resistance to antiestrogen therapies. Upon acquisition of hormone independence, ER+ breast cancer cells increase their dependence on the phosphatidylinositol-3 kinase (PI3K)/AKT pathway. We examined the effects of AKT inhibition and its compensatory upregulation of insulin-like growth factor (IGF)-I/InsR signaling in ER+ breast cancer cells with acquired resistance to estrogen deprivation.

Methods: Inhibition of AKT using the catalytic inhibitor AZD5363 was examined in four ER+ breast cancer cell lines resistant to long-term estrogen deprivation (LTED) by western blotting and proliferation assays. Feedback upregulation and activation of receptor tyrosine kinases (RTKs) was examined by western blotting, real-time qPCR, ELISAs, membrane localization of AKT PH-GFP by immunofluorescence and phospho-RTK arrays. For studies in vivo, athymic mice with MCF-7 xenografts were treated with AZD5363 and fulvestrant with either the ATP-competitive IGF-IR/InsR inhibitor AZD9362 or the fibroblast growth factor receptor (FGFR) inhibitor AZD4547.

Results: Treatment with AZD5363 reduced phosphorylation of the AKT/mTOR substrates PRAS40, GSK3α/β and S6K while inducing hyperphosphorylation of AKT at T308 and S473. Inhibition of AKT with AZD5363 suppressed growth of three of four ER+ LTED lines and prevented emergence of hormone-independent MCF-7, ZR75-1 and MDA-361 cells. AZD5363 suppressed growth of MCF-7 xenografts in ovariectomized mice and a patient-derived luminal B xenograft unresponsive to tamoxifen or fulvestrant. Combined treatment with AZD5363 and fulvestrant suppressed MCF-7 xenograft growth better than either drug alone. Inhibition of AKT with AZD5363 resulted in upregulation and activation of RTKs, including IGF-IR and InsR, upregulation of FoxO3a and ERα mRNAs as well as FoxO- and ER-dependent transcription of IGF-I and IGF-II ligands. Inhibition of IGF-IR/InsR or PI3K abrogated AKT PH-GFP membrane localization and T308 P-AKT following treatment with AZD5363. Treatment with IGFBP-3 blocked AZD5363-induced P-IGF-IR/InsR and T308 P-AKT, suggesting that receptor phosphorylation was dependent on increased autocrine ligands. Finally, treatment with the dual IGF-IR/InsR inhibitor AZD9362 enhanced the anti-tumor effect of AZD5363 in MCF-7/LTED cells and MCF-7 xenografts in ovariectomized mice devoid of estrogen supplementation.

Conclusions: These data suggest combinations of AKT and IGF-IR/InsR inhibitors would be an effective treatment strategy against hormone-independent ER+ breast cancer.

Figure 1

Catalytic AKT inhibitor suppresses LTED growth and prevents emergence of hormone-independent breast cancer cells. A) LTED cells were treated with 10% DCC-FBS ± 0 to 15 µM AZD5363 for 24 hours. Protein lysates were analyzed by immunoblot using the indicated antibodies. B) LTED cells were treated with 10% DCC-FBS ± 0.4 or 2 µM AZD5363. Media and drugs were replenished every three days. Cells were counted after five to ten days. Data are presented as percent of control; each bar, mean ± SEM (n = 3; *P <0.0001 versus Con, one-way ANOVA). C) Parental cells in 10% DCC-FBS were treated with ± 2 µM AZD5363 or 1 µM selumetinib. Media and drugs were replenished every three days. When control monolayers reached 60% to 80% confluency (after 15 (MCF-7), 36 (ZR75-1 and MDA-361) or 38 (HCC-1428) days, respectively), cells were fixed and stained with crystal violet. Representative images and quantification of integrated intensity (% control) are shown (*P <0.05 versus control, t-test). ANOVA, analysis of variance; DCC-FBS, dextran/charcoal-treated fetal bovine serum; LTED, long-term estrogen deprivation; SEM, standard error of the mean.

Figure 2

Combined inhibition of AKT and ER suppresses hormone-independent tumor growth. A) LTED cells were treated with 10% DCC-FBS ± 2 µM AZD5363 for 24 hours. RNA was extracted, reverse transcribed to cDNA and analyzed by real-time PCR. Threshold cycle values were normalized for actin. Data are presented as fold versus control; each bar, mean ± SEM (n = 2; *P <0.0001 versus Con, two-way ANOVA). B) LTED cells were treated with 10% DCC-FBS ± 2 µM AZD5363, 1 µM fulvestrant or 1 nM E2. Media and drugs were replenished every three days. Cells were counted after five days. Data are presented as percent of control; each bar, mean ± SEM (n = 3; *P <0.0001 versus Con or E2; #P <0.01 versus AZD or fulvestrant; ^P <0.01 versus E2 + fulvestrant, one-way ANOVA). C) MCF-7 cells were injected s.c. into athymic mice supplemented with 14-day release 17β-estradiol pellets. Mice bearing tumors ≥150 mm³ were randomized to vehicle, AZD5363 (150 mg/kg/day bid), fulvestrant (5 mg/wk), or AZD5363 + fulvestrant for six weeks. Data are presented as mean tumor volume ± SEM (*P <0.01 versus vehicle; ^#P <0.001 versus AZD or fulvestrant, t-test). D) Xenografts from C) were homogenized one hour after the last dose and tumor lysates were analyzed by immunoblot using the indicated antibodies. ANOVA, analysis of variance; DCC-FBS, dextran/charcoal-treated fetal bovine serum; LTED, long-term estrogen deprivation; SEM, standard error of the mean.

Figure 3

Inhibition of AKT results in feedback upregulation of RTKs. A) LTED cells were treated with 10% DCC-FBS ± 2 µM AZD5363 for 24 hours. RNA was extracted, reverse transcribed to cDNA and analyzed by real-time PCR. Threshold cycle values were normalized for actin. Average fold changes over control (Con) were calculated and used to generate a heatmap. B) LTED cells were treated for 24 hours in 10% DCC-FBS ± 15 µM AZD5363. Protein lysates were analyzed by immunoblot. Densitometric analysis was performed and the ratio of InsR:actin is shown below the InsR blots. C) LTED cells in 10% DCC-FBS were treated ± 2 µM AZD5363 for 0 to 24 hours. Cell lysates (0.5 mg) were prepared and analyzed by phospho-RTK arrays. D) Mice bearing MCF-7 xenografts ≥150 mm³ were treated with vehicle or AZD5363 (150 mg/kg bid p.o.) for one or three days. Xenografts were harvested four hours after the last dose; tumor lysates were prepared and analyzed by immunoblot. Densitometric analysis was performed and a graphical representation of the average RTK:actin levels is shown below the blot. E) Xenografts from D) were homogenized and RNA was extracted and analyzed by real-time PCR as in A. Data are presented as fold versus control; each bar, mean ± SEM (n = 3; *P <0.05 versus Con, t-test). DCC-FBS, dextran/charcoal-treated fetal bovine serum; LTED, long-term estrogen deprivation; RTK, receptor tyrosine kinase; SEM, standard error of the mean.

Figure 4

Inhibition of IGF-IR/InsR or PI3K abrogates AZD5363-induced AKT membrane localization and phosphorylation. A) MCF-7/LTED cells were transfected with an AKT PH-GFP plasmid. On day four, cells were treated with 100 ng/ml IGF-I in serum-free medium for 15 minutes, or pre-incubated with 10% DCC-FBS ± 1 µM AEW541 or 1 µM BKM120 for 30 minutes followed by treatment with 2 µM AZD5363 for four hours. Cells were viewed in a LSM 510Meta confocal microscope at 40x magnification. B) LTED lines were treated with 10% DCC-FBS ± 1 µM AEW541 or BKM120 for 1 hour, followed by the addition of 2 µM AZD5363 for 24 hours. Protein lysates were analyzed by immunoblot using the indicated antibodies. DCC-FBS, dextran/charcoal-treated fetal bovine serum; IGF-I, insulin-like growth factor-I; LTED, long-term estrogen deprivation.

Figure 5

AZD5363-induced upregulation of IGF-IR, IGF-I and IGF-II is dependent on ER and FoxO3. A) LTED cells were treated with 10% DCC-FBS ± 2 µM AZD5363 for 24 hours. T47D cells were serum starved for 24 hours and then treated with 10% DCC-FBS ± 2 µM AZD5363 for 24 hours. RNA was extracted, reverse transcribed to cDNA and analyzed by real-time PCR. Threshold cycle values were normalized for actin. qPCR reactions with no input RNA were used as negative controls. Data are presented as fold versus control; each bar, mean ± SEM (n = 4; *P <0.05 versus Con, t-test). B) Mice bearing MCF-7 or ZR75-1 xenografts ≥150 mm³ were treated with vehicle or AZD5363 (150 mg/kg bid) for one day. Xenografts were harvested four hours after the last dose and homogenized. RNA was extracted, reverse transcribed to cDNA and analyzed by real-time PCR. Threshold cycle values were normalized for actin (MCF-7) or 36B4 (ZR75-1). Data are presented as fold versus control; each bar, mean ± SEM (n = 3; *P <0.05 versus Con, t-test). C, D) MCF-7/LTED cells were transfected with siRNA targeting FoxO3, ER, both (siboth) or a non-silencing control (siCon). Two days later the cells were treated with 10% DCC-FBS ± 2 µM AZD5363 for 24 hours. RNA was extracted and analyzed by real-time PCR as in A. Data are presented as fold versus control; each bar, mean ± SEM (n = 3). C) *P <0.01 versus siCon, t-test. D) *P <0.05 versus siCon; #P <0.05 versus AZD, t-test. DCC-FBS, dextran/charcoal-treated fetal bovine serum; ER, estrogen receptor; FoxO, forkhead box class O; LTED, long-term estrogen deprivation.

Figure 6

IGFBP-3 blocks AZD5363-induced phosphorylation of IGF-IR/InsR and AKT. A) MCF-7/LTED cells were treated overnight with serum-free medium ± 1.5 µg/ml IGFBP-3, followed by treatment for 15 minutes ± 100 ng/ml IGF-I or IGF-II. Protein lysates were analyzed by immunoblot using the indicated antibodies. B) MCF-7/LTED cells in 10% DCC-FBS were treated ± 1.5 µg/ml recombinant IGFBP-3 for 1 hour followed by 2 µM AZD5363 for 24 hours. Cell lysates (0.5 mg) were prepared and analyzed by phospho-RTK arrays. C) MCF-7/LTED cells in 10% DCC-FBS were treated ± 1.5 µg/ml IGFBP-3 for 1 hour followed by 2 µM AZD5363 for 24 hours. Protein lysates were prepared and analyzed by immunoblot using the indicated antibodies. DCC-FBS, dextran/charcoal-treated fetal bovine serum; IGF, insulin-like growth factor; LTED, long-term estrogen deprivation.

Figure 7

Inhibition of IGF-IR/InsR enhances the anti-tumor effect of the AKT inhibitor AZD5363. A) MCF-7 cells were transfected with siRNA specific for a non-silencing control (siCon), InsR, IGF-IR or HER3 and re-seeded the next day for assessment of growth in monolayer (A) or immunoblot analyses (B). In A, cells were treated with 10% DCC-FBS ± 2 µM AZD5363 and counted after five days. Data are presented as % of control; mean ± SEM (n = 3; *P <0.0001 versus each Con, #P <0.01 versus siCon + AZD; two-way ANOVA). In B, cells were grown in 10% DCC-FBS and harvested three days later; lysates were analyzed by immunoblot using the indicated antibodies. C) MCF-7/LTED cells in 10% DCC-FBS were treated ± 2 µM AZD5363 or 1 µM AZD9362. Media and drugs were replenished every three days. Cells were counted after five days. Data are presented as % of control; mean ± SEM (n = 2; *P <0.0001 versus Con, #P <0.05 versus 5363 or 9362; one-way ANOVA). D) Mice bearing MCF-7 xenografts ≥150 mm³ were randomized to the indicated treatments. Data are presented as mean tumor volume ± SEM (*P <0.05 versus vehicle; ^#P = 0.0041 versus AZD5363, t-test). ANOVA, analysis of variance; DCC-FBS, dextran/charcoal-treated fetal bovine serum; IGF, insulin-like growth factor; LTED, long-term estrogen deprivation; SEM, standard error of the mean.