Pictilisib-Induced Resistance Is Mediated through FOXO1-Dependent Activation of Receptor Tyrosine Kinases in Mucinous Colorectal Adenocarcinoma Cells

Abstract

The phosphatidylinositol (PI3K)/AKT/mTOR axis represents an important therapeutic target to treat human cancers. A well-described downstream target of the PI3K pathway is the forkhead box O (FOXO) transcription factor family. FOXOs have been implicated in many cellular responses, including drug-induced resistance in cancer cells. However, FOXO-dependent acute phase resistance mediated by pictilisib, a potent small molecule PI3K inhibitor (PI3Ki), has not been studied. Here, we report that pictilisib-induced adaptive resistance is regulated by the FOXO-dependent rebound activity of receptor tyrosine kinases (RTKs) in mucinous colorectal adenocarcinoma (MCA) cells. The resistance mediated by PI3K inhibition involves the nuclear localization of FOXO and the altered expression of RTKs, including ErbB2, ErbB3, EphA7, EphA10, IR, and IGF-R1 in MCA cells. Further, in the presence of FOXO siRNA, the pictilisib-induced feedback activation of RTK regulators (pERK and pAKT) was altered in MCA cells. Interestingly, the combinational treatment of pictilisib (Pi3Ki) and FOXO1i (AS1842856) synergistically reduced MCA cell viability and increased apoptosis. These results demonstrate that pictilisib used as a single agent induces acute resistance, partly through FOXO1 inhibition. Therefore, overcoming PI3Ki single-agent adaptive resistance by rational design of FOXO1 and PI3K inhibitor combinations could significantly enhance the therapeutic efficacy of PI3K-targeting drugs in MCA cells.

Keywords: FOXO; mucinous colorectal adenocarcinomas; pictilisib.

Figure 1

Nuclear enrichment of FOXOs in PI3Ki-treated MCA cells. (A,A’,C,C’): DMSO vehicle-treated LS174T and RW7213 cells expressed FOXO1 after 96 h. (B,B’,D,D’): LS174T and RW7213 cells exhibit FOXO1 staining in the nuclear compartment after 96 h of pictilisib treatment. Nuclei were assigned with pseudo-colored blue and FOXO1 with red. Nuclear enrichment of FOXO1 was seen at 96 h in both MCA cell lines (nuclei appear pink due to the red–blue overlap (white arrows). Cytoplasmic FOXO1 peri nuclear red-stained in (A’,C’) (white arrows). Scale bar: (A–D): 25 μm; (A’,B’,D’): 7.5 μm; (C’):10 μm. Statistical significance of nuclear FOXO1 staining intensity was measured in both MCA cell lines using a two-tail t-test (E,F): LS174T (vehicle, n = 230 cells; PI3Ki, n = 249 cells), RW7213 (vehicle = DMSO control, n = 236 cells; PI3Ki, n = 114 cells). p-value significance p < 0.001; error bars: standard error mean values were plotted.

Figure 2

Enrichment of FOXO proteins in nuclear fractions after PI3Ki treatment. FOXO1 (A), FOXO3 (B), and FOXO4 (C) proteins from nuclear and cytoplasmic fractions of the vehicle or PI3Ki-treated (for 72 or 96 h) LS17T cells were resolved, blotted, and probed with respective antibodies: FOXO1, FOXO3, FOXO4, mSIN3A, and β-tubulin. Nuclear FOXOs were normalized to mSIN3A, and cytoplasmic FOXOs were normalized to β-tubulin. Notably, the nuclear fraction of FOXO1 at 72–96 h was elevated in the PI3Ki-treated cells (A), while FOXO4 proteins were restricted only to the nucleus with or without PI3K inhibition (1C). C = cytoplasmic; N = nuclear; vehicle = DMSO control.

Figure 3

FOXO1, FOXO3, and FOXO4 were deleted in MCA cells by siRNA-mediated knockdown. (A–C): LS174T cells were transfected with siRNA smart pools (Dharmacon), targeting FOXO1, FOXO3, FOXO4, and non-silencing RNA as off-target transfection control. Western blot analysis showed effective knockdown of FOXO1, 3, and 4 in LS17T cells. Expression levels of proteins on Western blots were quantified by densitometry, and the percentages of FOXO protein expression relative to untreated (−) controls are shown below the blots. All values were normalized to β-actin as a loading control.

Figure 4

Rebound activation of pERK1/2 and pAKT seen in response to PI3K inhibition and FOXO knockdown. LS174T cells were transfected with FOXO1, FOXO3, or FOXO4 siRNA alone or in combination, grown for 24 h, and treated with PI3Ki for 96 h. (A) Western blots were performed using anti-pERK (pERK1 and ERK2), and (B) anti-pAKT and anti-total-ERK (tERK1 and ERK2), and total-AKT antibodies. The blot signals were quantified using densitometry and were normalized to non-silencing controls. Phospho-ERK1/2 levels were reduced in all PI3Ki-treated cells with FOXO knockdown compared to non-silencing controls (A). A more pronounced reduction in FOXO1 (A) and FOXO4 (A) siRNA-treated cells was seen than in FOXO3 siRNA-treated cells (A). Phospho-ERK1/2 levels were reduced when all three FOXO genes were knocked down, but this reduction was comparable to FOXO1 single knockdown, suggesting that FOXO1 may be the most critical FOXO protein for mediating rebound activation of pERK levels in response to PI3Ki-treatment. Similarly, phospho-AKT levels were reduced in FOXO1 (B), FOXO3 (B), and FOXO1 + FOXO4 siRNA-treated PI3Ki-resistant cells compared to FOXO3, non-silencing siRNA treated cells. Phospho-AKT levels were reduced with FOXO1 gene knockdown (B). Compared to all other FOXOs alone or together, knockdown indicates that the FOXO1 protein could be a critical modulator to influence PI3K inhibitor-treated resistant cell survival through AKT phosphorylation.

Figure 5

FOXO1-mediated differential expression of receptor tyrosine kinases (RTKs) in PI3Ki-induced resistant MCA cells. (A). Quantitative RT-PCR results show a reduction in IGFR1, IR, ErbB2, ErbB3, and EphA10 mRNA relative gene expression levels in LS174T cells transfected with FOXO1 siRNA subjected to PI3Ki treatment. ErbB2 and ErbB3 mRNA relative gene expression was not significantly reduced. Error bars indicate the standard error of the mean of relative gene expression levels (n = 4). p-value significance (* p < 0.01, ** p < 0.001, and *** p < 0.0001). Error bars: standard error mean values were plotted. (B–C’): FOXO transcriptional activity monitored in PI3Ki-treated LS174T cells and DMSO vehicle control cells. (B,B’): FOXO1 transcriptional promoter element contained GFP reporter construct transfected 24 h, upon DMSO vehicle treatment for 96 h. Cytoplasmic GFP expression (white arrows) was seen in LS174T control cells. (C,C’): FOXO1 responsive promoter element contained GFP reporter construct transfected 24 h followed by PI3Ki treatment. After 96 h, nuclear GFP expression (white arrows) was seen in PI3Ki-resistant LS174T cells. Scale bar: (B–C’) 20 μm.

Figure 6

RTK phosphorylation in response to FOXO1 knockdown and prolonged exposure to PI3K inhibitor. (A,B): After 96 h, FOXO1 and non-silencing siRNA followed by PI3Ki treated LS174T cell lysates were incubated with RTK profiler antibody array containing 49 RTKs (R&D Biosystems), followed by the anti-phospho-tyrosine antibody. The RTKs that were elevated and downregulated in the boxes indicate RTK expression. (C): Pixel densities were quantified using ImageJ analysis software 1.44. The numbers on the x-axis correspond to RTKs and fold elevation relative to non-silencing siRNA + PI3Ki controls. The x-axis number 1 to 13 corresponds to EGFR, ErbB2, ErbB3, FGFR3, Insulin R, MSPR, RCR2, Tie 2, Eph A2, Eph A7, ALK, and Eph A10. These RTKs activity was normalized with non-silencing siRNA + PI3Ki.