PPP3CB overexpression mediates EGFR TKI resistance in lung tumors via calcineurin/MEK/ERK signaling

Abstract

Despite initial high response rates to first-line EGFR TKI, all non-small-cell lung cancer (NSCLC) with EGFR-activating mutation will ultimately develop resistance to treatment. Identification of resistance mechanisms is critical to adapt treatment and improve patient outcomes. Here, we show that a PPP3CB transcript that encodes full-length catalytic subunit 2B of calcineurin accumulates in EGFR-mutant NSCLC cells with acquired resistance against different EGFR TKIs and in post-progression biopsies of NSCLC patients treated with EGFR TKIs. Neutralization of PPP3CB by siRNA or inactivation of calcineurin by cyclosporin A induces apoptosis in resistant cells treated with EGFR TKIs. Mechanistically, EGFR TKIs increase the cytosolic level of calcium and trigger activation of a calcineurin/MEK/ERK pathway that prevents apoptosis. Combining EGFR, calcineurin, and MEK inhibitors overcomes resistance to EGFR TKI in both in vitro and in vivo models. Our results identify PPP3CB overexpression as a new mechanism of acquired resistance to EGFR TKIs, and provide a promising therapeutic approach for NSCLC patients that progress under TKI treatment.

Figure S1.. Validation of exon skipping events in PC9/OR clones.

RT–PCR analysis of altered exon skipping events (in gray) common to PC9/GR and PC9/DR clones showing inclusion (ACOT9, ARFGAP2, CD46, ESYT2, GK, SMC5) or exclusion (RAB17) of exons in PC9/OR clones compared with sensitive cells (PC9). GAPDH was used as a control. Percent spliced-in (PSI) is indicated below the PCR blots. n = 2 biological replicates. Source data are available for this figure.

Figure 1.. Cells with acquired resistance to EGFR TKIs accumulate a PPP3CB splice variant isoform that retains exon 16.

(A) RT–PCR showing accumulation of PPP3CB-Ex16 in clones with acquired resistance to EGFR TKI (PC9/GR, PC9/DR, PC9/OR) as compared to parental sensitive cells (PC9). Percent spliced-in (PSI) is indicated below the PCR blot. Data are presented as the mean ± SD. n = 3 biological replicates. The dotted line highlights cropped images from the same gel. (B) Western blot showing accumulation of PPP3CB protein in resistant cells (PC9/GR1, PC9/DR1, PC9/OR3) compared with sensitive cells (PC9). GAPDH was used as a loading control. (C) Resistant cells were transfected with a siRNA targeting exon 16 (siEx16) or the exon 15/17 junction (siEx15/17) of PPP3CB or with a control siRNA (siCtl), and treated with appropriate EGFR TKI (either gefitinib, dacomitinib, or osimertinib, 0.1 μM) for 72 h. Quantification of apoptotic rates was determined by active caspase 3 staining and flow cytometry (left panel). Data are presented as the mean ± SD, n = 3 biological replicates. ****P < 0.0001, ***P < 0.001, **P < 0.01, unpaired t test. Neutralization of PPP3CB isoforms was studied by RT–PCR in PC9/OR (right panel). GAPDH was used as a control. (D) Neutralization of PPP3CB Ex-16 (siEx16) inhibits cell viability (MTS assay) in response to EGFR TKIs (0.1 μM, 72 h). Data are presented as the mean ± SD, n = 3 biological replicates. ****P < 0.0001, ***P < 0.001, **P < 0.01, ns, not significant, unpaired t test. (E) PC9/GFP and PC9/PPP3CB-GFP cells were cultured for 36 h with gefitinib (0.1 μM), dacomitinib (0.05 μM), or osimertinib (0.1 μM). Western blotting was performed to assess PPP3CB-GFP expression (left panel). Tubulin was used as a loading control. (C) Quantification of apoptotic rates was determined as in (C) (right panel). Data are presented as the mean ± SD, n = 4 biological replicates, **P < 0.01, *P < 0.05, unpaired t test. (F) Representative images of a clonogenic assay using PC9-GFP or PC9-PPP3CB-GFP cells cultured with osimertinib (osi, 0.03 μM) for 13 d. Source data are available for this figure.

Figure 2.. Expression of PPP3CB mRNA in human non–small-cell lung cancer biopsies.

Detection of PPP3CB transcript (pink dots) in PC9/OR cells using RNA BaseScope. (A) Minimal 50-bp 1ZZ probe containing complementary sequences to the 30 bp of exon 16 of PPP3CB plus 10 bp in exons 15 and 17 was designed according to the manufacturer’s recommendations (Bio-Techne). (B, C) PC9/OR cells were transfected with control siRNA (siCtl) or with siRNA against exon 16 of PPP3CB for 72 h. (B) Cells were processed for RNA BaseScope with the PPP3CB 1ZZ probe. Arrows illustrate pink dots. Scale bars: 10 μm. (C) Quantification of RNA BaseScope data. Hematoxylin-counterstained tumor cells (blue) with or without pink dots (PPP3CB) were counted in 11 distinct fields within the slide (160 cells at least per condition). Results were expressed as % of cells with at least one dot. Data are presented as the mean ± SD, ****P < 0.0001, unpaired t test. Neutralization of the PPP3CB exon 16 transcript was checked by RT–PCR. GAPDH was used as a loading control. (D) Representative images of the RNA BaseScope assay in matched biopsies showing the increased expression of PPP3CB Ex16 mRNA levels (pink dots, arrows) in post-treatment sample. Scale bars: 10 μm. (E) Accumulation of PPP3CB Ex16 in 11 paired tumor biopsies from the global cohort of 43 patients (upper panel) (7–10 images within each section, mean = 491 total cells per sample). The statistical significance of PPP3CB Ex16 increases in the 11 paired samples (lower panel). PD, progression disease. ***P = 0.001, paired t test (Wilcoxon’s test). (F) Mechanisms of acquired resistance in 25 paired biopsy subcohort of patients treated with first-line EGFR TKIs. Source data are available for this figure.

Figure 3.. CsA sensitizes resistant cells to EGFR TKI treatment.

(A, B, C, D) Cells were cultured with indicated EGFR TKIs in the presence or absence of CsA for 96 h (A), 72 h (B), 2 wk (C), or 10 d (D). Concentrations used for each drug in each assay are indicated. (A) Cell viability was measured by an MTS assay (upper panel). CsA toxicity is shown in lower panels. Data are presented as the mean ± SD, n = 4 biological replicates. ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05, ns, not significant. (B) Apoptosis was quantified by active caspase 3 staining and flow cytometry. Data are presented as the mean ± SD, n = 3 or 4 biological replicates. ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05, ns, not significant, unpaired t test. (C) Representative images of a clonogenic assay. (D) Representative images of PC9/OR spheroids cultured with osimertinib (osi) and/or CsA. Data are presented as the mean ± SD, n = 4 biological replicates; two technical replicates for each biological replicate. **P < 0.01, ns, not significant, unpaired t test. Source data are available for this figure.

Figure 4.. Osimertinib activates a Ca²⁺/calcineurin/MEK/ERK1/2 pathway that prevents apoptosis in response to osimertinib.

(A) Western blot analysis of proteins of the EGFR pathway in cells treated with osimertinib (osi) for 72 h. (B) Representative immunoblots of indicated proteins in PC9/OR cells transfected with siRNA against ERK1/2 (siERK1/2) or with control siRNA (−) and cultured (+) or not (−) with osimertinib (osi) for 72 h. c-Casp3 = cleaved caspase 3. (C, D) PC9/OR cells were cultured with osimertinib (osi) and/or trametinib (trame) for 72 h. (C) Representative immunoblots of indicated proteins. (D) Quantification of apoptotic rates was determined by active caspase 3 staining and flow cytometry. Data are presented as the mean ± SD, n = 3 biological replicates. ***P < 0.001, *P < 0.05, ns, not significant, unpaired t test. (E) Representative images of the colony formation assay in PC9/OR cells cultured with osimertinib (osi) and/or trametinib (trame) for 2 wk. (F) Representative images of PC9/OR spheroids cultured for 10 d with osimertinib (osi) and/or trametinib (trame). Data are presented as the mean ± SD, n = 4 biological replicates; two technical replicates for each biological replicate. **P < 0.01, *P < 0.05, ns, not significant, unpaired t test. (G) Representative immunoblots of pERK/ERK in PC9/OR cells transfected with PPP3CB siEx16 or siEx15/17 or control siRNA (siCtl) and treated with osimertinib (osi) for 72 h. (H) Representative immunoblots of pERK/ERK in PC9/OR cells cultured with osimertinib (osi) and/or CsA for 72 h. (I) Representative immunoblots of pERK/ERK in PC9/GFP and PC9/PPP3CB-GFP cells cultured with osimertinib (osi) for 18 h. (J, K) Measurement of iCa2+ using the fluorescence indicator, cell-permeant Fluo-4/AM. (J) Dose-dependent effect of osimertinib on the iCa²⁺ level in PC9/OR cells treated for 48 h. ns, not significant. Data are presented as the mean ± SD, n = 4 biological replicates; two technical replicates for each biological replicate. ****P < 0.0001, **P < 0.01, unpaired t test. (K) Live-cell imaging of iCa²⁺ using epifluorescence microscopy (green signal) in PC9/OR cells cultured for 48 h with osimertinib (osi). Bar = 10 μm. (L) Representative immunoblots of pERK/ERK in PC9/OR cells cultured with osimertinib (osi) and/or APS-2-79 for 72 h. Data, GAPDH, or tubulin was used as a loading control. The concentration used for each drug in each assay is indicated. (A, B, C, G, H, I, L) Three biological replicates. Source data are available for this figure.

Figure S2.. Effect of EGFR TKI on EGFR signaling pathways in sensitive (PC9) and resistant (PC9/GR and PC9/DR) cells.

(A, B) Representative immunoblots of proteins of the EGFR pathway in cells treated with (A) gefitinib (gef, 0.1 μM) or (B) dacomitinib (daco, 0.05 μM) for 72 h. GAPDH was used as a loading control. Source data are available for this figure.

Figure S3.. Gefitinib activates a Ca²⁺/calcineurin/MEK/ERK pathway in resistant cells.

(A) Representative immunoblots of PC9/GR cells treated with gefitinib (gef) and/or trametinib (trame) for 72 h cCasp3 = cleaved caspase 3. (B) Representative immunoblots of pERK/ERK in PC9/GR cells transfected with PPP3CB siEx16 or siEx15/17 or control siRNA (siCtl) and treated with gefitinib (gef) for 72 h. (C) Representative immunoblots of PC9/GR cells treated with gefitinib (gef) and/or cyclosporin A (CsA) for 72 h. (D) Representative immunoblots of pERK/ERK in PC9/GFP and PC9/PPP3CB-GFP cells cultured with gefitinib (gef) for 18 h. (E, F) Measurement of iCa²⁺ using the fluorescence indicator, cell-permeant Fluo-4/AM. (E) Dose-dependent effect of gefitinib on iCa²⁺ in PC9/GR cells after 48 h of drug treatment. Data are presented as the mean ± SD, n = 4 biological replicates; two technical replicates for each biological replicate. **P < 0.001, unpaired t test, ns, not significant. (F) Live-cell imaging of iCa²⁺ using epifluorescence microscopy (green signal) in PC9/GR cells cultured for 48 h with gefitinib (gef). Bar = 10 μm. (G) Representative immunoblots of pERK/ERK in PC9/GR cells treated with APS-2-79 and/or gefitinib (gef) for 72 h. (H, I) Quantification of apoptotic rates determined by active caspase 3 staining and flow cytometry in PC9/GR (H) and PC9/OR (I) cells transfected with siKSR2 or with a control siRNA (siCtl) and treated with appropriate EGFR TKI (gef or osi) for 72 h. Data are presented as the mean ± SD, n = 3 or 5 biological replicates. ****P < 0.0001, ***P < 0.001, ns, not significant, unpaired t test. Efficient neutralization of KSR2 was checked by quantitative PCR ((H), right panel). (A, B, C, D, G) Three biological replicates. Data, GAPDH, or tubulin was used as a loading control. In all experiments, the concentration used for each drug is indicated. Source data are available for this figure.

Figure 5.. Combined CsA/trametinib treatment restores sensitivity to osimertinib in 2D and 3D cell culture assays.

(A) Representative MTS cell viability assays in PC9/OR cells treated with osimertinib, CsA, and trametinib. Data are presented as the mean ± SD, n = 4 biological replicates; three technical replicates for each biological replicate. IC50 values were calculated using GraphPad Prism. Data are presented as the mean ± SD. (B) Representative images of a clonogenic assay in PC9/OR cells cultured with osimertinib (Osi) and/or CsA and/or trametinib (trame) for 2 wk. (C, D, E) 3D culture of PC9/OR cells. (C) Representative images of PC9/OR spheroids cultured for 10 d with osimertinib (osi), CsA, and/or trametinib (trame). (D) Spheroid area, mean ± SD, **P < 0.01, unpaired t test, and (E) cell viability, mean ± SD, **P < 0.01, unpaired t test, were assessed after 10 d of drug treatment. (D, E) n = 4 biological replicates; two technical replicates for each biological replicate. In all experiments, the concentration used for each drug is indicated. Source data are available for this figure.

Figure 6.. In vivo efficacy of the cyclosporin A/trametinib/osimertinib combination in reversing acquired resistance to osimertinib.

(A) Mean tumor volume (± SD) of the PC9/OR3 cell line model treated with cyclosporin A (25 mg/kg, 4x/w, PO) ± trametinib (0.1 mg/kg, 4x/w, PO) ± osimertinib (5 mg/kg, 5x/w, PO). n = 8–10 animals per group. (B) Individual tumor growth curves. n = 8–10 animals per group. (C) Statistical analysis of the tumor volume performed by an unpaired t test (Mann–Whitney test). n = 8–10 mice per group. (D) Percentage of volume change from initial volume after 29 d of treatment, percentages in gray correspond to a percentage of change lower than 50%, and percentages in black correspond to a percentage of change lower than −50%. n = 8–10 animals per group. Source data are available for this figure.

Figure S4.. Probability of progression after each tested treatment in the PC9/OR3 cell line model.

(A, B) Time to reach relative tumor volume (RTV) = 2 (A) and RTV = 4 (B) for each treated mouse has been calculated. (C) Mean relative weight (± SD) of mice treated with cyclosporin A ± trametinib ± osimertinib. Source data are available for this figure.

Figure S5.. In vivo efficacy of lower doses of the cyclosporin A/trametinib/osimertinib combination in the PC9/OR3 cell line model.

(A) Mean tumor volume (± SD) of PC9/OR3 xenografts treated with cyclosporin A (2.5 mg/kg, 4x/w, PO) ± trametinib (0.1 mg/kg, 4x/w, PO) ± osimertinib (1.25 mg/kg, 5x/w, PO). (B) Percentage of volume change from initial volume after 36 d of treatment, percentages in gray correspond to a percentage of change lower than 50%, and percentages in black correspond to a percentage of change lower than −50%. Source data are available for this figure.

Figure S6.. Viability of cells with acquired resistance to osimertinib.

Representative curves are illustrated. Parental sensitive PC9 cells and resistant PC9/OR3 cells were treated with increasing amounts of osimertinib for 96 h. Survival was studied using an MTS assay. Data are presented as the mean ± SD, n = 4 biological replicates. Source data are available for this figure.