Aberrant activation of the PI3K/mTOR pathway is a common feature of many cancers and an attractive target for therapy, but resistance inevitably evolves as is the case for any cancer cell–targeted therapy. In animal tumor models, chronic inhibition of PI3K/mTOR initially inhibits tumor growth, but over time, tumor cells escape inhibition. In this study, we identified a context-dependent mechanism of escape whereby tumor cells upregulated the proto-oncogene transcriptional regulators c-MYC and YAP1. This mechanism was dependent on both constitutive ERK activity as well as inhibition of the stress kinase p38. Inhibition of p38 relieved proliferation arrest and allowed upregulation of MYC and YAP through stabilization of CREB. These data provide new insights into cellular signaling mechanisms that influence resistance to PI3K/mTOR inhibitors. Furthermore, they suggest that therapies that inactivate YAP or MYC or augment p38 activity could enhance the efficacy of PI3K/mTOR inhibitors. Cancer Res; 76(24); 7168–80. ©2016 AACR.

The PI3K/mTOR pathway is a major growth signaling pathway that regulates a wide variety of cellular processes, such as protein, nucleotide, and lipid synthesis (1). Mutations in several oncogenes or loss of tumor suppressors lead to deregulated PI3K/mTOR signaling in tumors (for reviews, see refs. 2, 3), making it an attractive target for therapy. Although single-agent inhibition of PI3K/mTOR typically leads to cytostasis, it has not led to durable response rates, rarely inducing tumor cell death or shrinkage of solid tumors (reviewed in ref. 4). Under cytostasis, tumors can relapse due to emergence of resistant cells that escape proliferative suppression through genetic or epigenetic mechanisms.

Upregulation of receptor tyrosine kinases (RTK; refs. 5–10) is a known resistance mechanism that allows cancer cells to survive PI3K/mTOR inhibition. Upregulation of RAS/MEK/ERK pathway can also promote resistance to PI3K/mTOR inhibitors (6, 11, 12). Other reported mechanisms include modulation of RSK3/4 (13), c-MYC (14, 15), β-catenin (16), eIF4E (15, 17), or deregulation of 4EBP proteins (18–20). Although these alterations promote survival, the mechanism by which tumor cells overcome proliferation arrest induced by PI3K/mTOR inhibitors is not well understood. Here, we identify upregulation of c-MYC and Yes-associated protein (YAP1) as a resistance mechanism that allows proliferation under chronic PI3K/mTOR inhibition. MYC and YAP upregulation is context dependent, occurring in an activated RAS/ERK background, and requires suppression of the stress-activated p38 MAPK (MAPK14).

Antibodies and reagents

Rapamycin (gift from Dr. Blenis, Weill Cornell Medicine, New York, NY), BEZ235 (Axon-Medchem), and Torin1 (Tocris Bioscience) were dissolved in DMSO and used at the indicated concentrations. BEZ235 was used at 0.5 μmol/L except where noted differently. SB203580/SB202190 (Cell Signaling Technology) were used at 5 μmol/L concentration, and cells were preincubated with the inhibitor for 24 hours. UO126 (Sigma) was used at 10 μmol/L. MK2206, BKM120, and GDC0941 (Selleckchem) were all used at 250 nmol/L. All antibodies are described in the Supplementary Materials.

Plasmids/shRNAs

shRNAs were obtained from RNAi Consortium (clone IDs from Open BioSystems). pLKO MYC shB5 (TRCN0000039639), MYC shB6 (TRCN0000039640) and shB8 (TRCN0000039642), YAP1 shF5 (TRCN0000107265) and shF8 (TRCN0000107268), KRAS shA8 (TRCN0000033261), shA9 (TRCN0000033262), and pLKO scrambled were used in shRNA experiments. Plasmids for overexpression of MYC and YAP1 were pBABE YAP1 (Addgene#15682), pWZL MYC (Addgene#10674), and CREB1 WT in pQCXIB. pcDNA-p38CA (p38/Mapk14) was a kind gift from Dr. Aguirre-Ghiso (Icahn School of Medicine at Mount Sinai, New York, NY).

Reverse phase protein array analysis

Reverse phase protein array (RPPA) sample preparation, normalization of data points, and analysis were performed as described previously (21). Cell lysates were arrayed on slides, incubated with antibodies, and quantified. Differentially expressed proteins and phospho-proteins were identified using t test or ANOVA and P < 0.05 threshold. Heatmaps were generated using Cluster3.0 and Java TreeView1.1.1. Proteins were ordered by the rank sum of the normalized values.

Cell culture

MCAS, OVCAR5, OvCa432, MDA-MB-468, MCF7, and TOV21G lines were from Dr. Slamon (UCLA, Los Angeles, CA) between 2008 and 2012 and validated by short tandem repeat (STR) profiling. HeLa, HCT116, and cancer lines in Fig. 6 were from Dr. Sabatini (Whitehead/MIT, Boston, MA) obtained in 2013, and 19 of the 27 lines were validated by STR at the time. HeLa, A431, DLD1, DU-145, PC3, BT549, SK-MEL-28, SW620, and S462 have not been profiled. Jeko is of unknown lineage. Three-dimensional (3D) experiments were performed as described before (8). Detailed description is in the Supplementary Methods. Lentiviral vectors encoding shRNAs in pLKO plasmids were generated in 293T cells and retroviruses were generated in 293GPGs according to standard protocol (22).

Xenograft experiments

HCT116 (5 × 105) or OVCAR5 cells (1 × 106) in 1:1 mix of PBS:Matrigel were injected subcutaneously into two flanks of approximately 24 g 10- to 12-week-old female NOD.CB17-Prkdcscid/J mice (Jackson Laboratories). Once tumors became palpable (∼250 mm3), 12 days (HCT116) or 28 days (OVCAR5), mice were randomized to groups of five for each treatment group (20 animals in total). Five animals per group were calculated to give sufficient statistical power for the purpose of this experiment. Drug was administered daily intraperitoneally. GNE493 (Genentech; 10 mg/kg) was dissolved in 0.5% methylcellulose/0.2% Tween-80. Tumors were harvested on 11- to 13-day posttreatment. All mouse studies were conducted through Institutional Animal Care and Use Committee-approved animal protocols (#04004) in accordance with the Harvard Medical School institutional guidelines.

Immunofluorescence and microscopy

3D spheroids were fixed, stained, and imaged as previously described (23). Paraffin-embedded tumor sections were unmasked by pH6 citrate-buffer and probed overnight with primary antibodies. Secondary antibodies were probed with Alexa-488 and -568 (Invitrogen). Cells were imaged with confocal microscopy. For details, see Supplementary Methods.

Western blot analysis

Cells were harvested for Western blot analysis in RIPA buffer supplemented with protease and phosphatase inhibitors and MG132 (Sigma). Lysates were boiled in 1× sample buffer for 5 minutes, resolved by 4%–20% SDS-PAGE gradient gels, transferred to polyvinylidene difluoride membranes (Millipore), blocked with 5% BSA-TTBS, and probed by primary antibodies overnight. Membranes were probed with secondary antibodies linked to horseradish peroxidase.

We previously showed using 3D spheroid cultures that treatment of matrix-adherent cancer cells with PI3K/mTOR inhibitors results in the inhibition of cell proliferation, but rarely in cell death (8). To model progression under conditions of chronic PI3K/mTOR inhibition in 3D, we cultured MCAS tumor cells under chronic exposure to the dual PI3K/mTOR inhibitor, BEZ235. Cells were cultured in reconstituted basement membrane proteins (3D), during which time the drugs and media were replenished every four days. Because of the sequestration of BEZ235 in 3D cultures, we used BEZ235 at 0.5 to 1 μmol/L concentration to fully inhibit the pathway (Supplementary Fig. S1A). MCAS cells initially displayed cytostasis in the presence of BEZ235. However, after 1 year of chronic exposure, proliferative outgrowths emerged (Fig. 1A, bottom), whereas control cells cultured in 3D for the same amount of time in the absence of drug remained sensitive to BEZ235 (Fig. 1A, top). After proliferation became apparent, the resistant cells (MCAS-R) were passaged by trypsinization every 2 weeks and maintained constitutively as 3D cultures in the presence of BEZ235, and the sensitive cells were maintained in DMSO (MCAS-S).

Figure 1.

MCAS cancer cells develop resistance to chronic PI3K/mTOR inhibition. A, Top, representative images of sensitive (MCAS-S) spheroids incubated with BEZ235 for 2 days. Bottom, representative images of resistant (MCAS-R) spheroids incubated with 1 μmol/L BEZ235 for 1 year. Outgrowths of proliferating cells were detected in multiple structures (arrows). Scale bar, 200 μm. B, Left, Ki67 (red) and laminin-5 (green) staining of MCAS-S and MCAS-R cells in BEZ235. Right, cleaved-caspase-3 (red) and E-cadherin (green) staining of resistant and sensitive cells. Scale bar, 25 μm. C and D, Cells were isolated from spheroid cultures, and similar numbers were replated in 3D. The MCAS-S and MCAS-R cells were incubated with DMSO (C), BEZ235 (0.5 μmol/L), Torin 1 (0.5 μmol/L), or BEZ235 with rapamycin (20 nmol/L), and proliferation was measured by cell counting (D). Data are shown as total cell number and presented as mean SEM ± from representative experiment of five. Statistical analysis, one-way ANOVA (P < 0.001) and post hoc Tukey HSD test. Pairwise P values are shown (***, P < 0.005; **, P < 0.01; *, P < 0.05). E, Western blot analysis of mTOR downstream effectors from MCAS-R and MCAS-S cells in 3D treated with DMSO, BEZ235, Torin, or BEZ + rapamaycin for 48 hours (MCAS-S) or 6 days (MCAS-R).

Figure 1.

MCAS cancer cells develop resistance to chronic PI3K/mTOR inhibition. A, Top, representative images of sensitive (MCAS-S) spheroids incubated with BEZ235 for 2 days. Bottom, representative images of resistant (MCAS-R) spheroids incubated with 1 μmol/L BEZ235 for 1 year. Outgrowths of proliferating cells were detected in multiple structures (arrows). Scale bar, 200 μm. B, Left, Ki67 (red) and laminin-5 (green) staining of MCAS-S and MCAS-R cells in BEZ235. Right, cleaved-caspase-3 (red) and E-cadherin (green) staining of resistant and sensitive cells. Scale bar, 25 μm. C and D, Cells were isolated from spheroid cultures, and similar numbers were replated in 3D. The MCAS-S and MCAS-R cells were incubated with DMSO (C), BEZ235 (0.5 μmol/L), Torin 1 (0.5 μmol/L), or BEZ235 with rapamycin (20 nmol/L), and proliferation was measured by cell counting (D). Data are shown as total cell number and presented as mean SEM ± from representative experiment of five. Statistical analysis, one-way ANOVA (P < 0.001) and post hoc Tukey HSD test. Pairwise P values are shown (***, P < 0.005; **, P < 0.01; *, P < 0.05). E, Western blot analysis of mTOR downstream effectors from MCAS-R and MCAS-S cells in 3D treated with DMSO, BEZ235, Torin, or BEZ + rapamaycin for 48 hours (MCAS-S) or 6 days (MCAS-R).

Close modal

To verify that the MCAS-R outgrowths were because of proliferation, we stained for the proliferation marker Ki67 in 3D (Fig. 1B), and analyzed cell numbers in 3D at indicated timepoints (Fig. 1C and D). Although the baseline proliferation rate of vehicle-treated MCAS-R cells was lower than that of MCAS-S cells (Fig. 1C), the MCAS-R cells displayed significant proliferative advantage in the presence of BEZ235 (Fig. 1D, Tukey HSD, P < 0.01). This effect was not limited to BEZ235, as MCAS-R cells also proliferated in the presence of a structurally distinct mTOR inhibitor, Torin1 (Fig. 1D, Tukey, P < 0.05), and with combination of BEZ235 and rapamycin (Fig. 1D, Tukey, P < 0.01). To eliminate altered drug metabolism or efflux as potential mechanisms of resistance, we evaluated whether BEZ235, Torin, and rapamycin still inhibit mTOR target molecules (p-S6, p-AKT, and p-4EBP1). Indeed, all inhibitors suppressed phosphorylation of these proteins (Fig. 1E). However, MCAS-R cells consistently displayed slightly higher levels of translation regulator 4EBP and phospho-4EBP1 under BEZ235-treated conditions. This is not surprising given previous reports demonstrating that minimal recovery of cap-dependent translation is needed for mTOR inhibitor resistance (15, 18–20).

The simultaneous emergence of MCAS-R cell outgrowths in several spheroids suggests that this is not a rare genetic event, but a time-dependent adaptation to chronic PI3K/mTOR inhibition. Consistent with this hypothesis, full exome-sequencing (>830× coverage/cell line) failed to identify nonsynonymous or splicing variants that are discordant between MCAS-R and MCAS-S cells (NCBI Sequence-Read-Archive Accession #SRP057514; ref. 24). Although epigenetic or non-exon alterations that might contribute to resistance cannot be ruled out, this data, together with the evidence that MCAS-R cells revert to drug-sensitive state after 16 weeks of drug-free culture (Supplementary Fig. S1B), led us to investigate alternative mechanisms of resistance. To examine a broad array of signaling pathways that could mediate resistance, we performed RPPAs on lysates from MCAS-S and MCAS-R cells treated in 3D cultures with DMSO or BEZ235. We then evaluated which proteins were differentially altered under any condition (Fig. 2A). Signals from several proteins or phospho-proteins differed significantly in BEZ235-treated MCAS-R and MCAS-S cells; however, many of these also differed under basal DMSO-treated conditions (Fig. 2A and Supplementary Fig. S1C). Fifty-eight proteins were significantly differentially altered in expression or phosphorylation by BEZ235 treatment in one of the two lines (Fig. 2B, P < 0.05 BEZ235-treated MCAS-R vs. BEZ235-treated MCAS-S cells). Two proteins previously reported to promote resistance to PI3K/mTOR inhibitors, MYC and eIF4E, were significantly upregulated by BEZ235 treatment in MCAS-R cells relative to MCAS-S cells (Fig. 2A and B), but were not significantly changed at baseline. Similarly, several other proteins, including EGFR, paxillin, NF2, Cyclin E, caveolin1, and YAP were also differentially upregulated by BEZ235 in MCAS-R cells with minimal differences at baseline (Fig. 2B). The RPPA results for several upregulated proteins were validated by Western blot analysis (Fig. 2C and Supplementary Fig. S1D).

Figure 2.

MCAS-R cells exhibit increased MYC and YAP expression. A, Heatmap from RPPA analysis showing proteins/phospho-proteins differentially expressed in MCAS-S and MCAS-R lines following DMSO and BEZ235 treatment (ANOVA, P < 0.05). Data are median centered (red: greater than the median, green: less). For A and B, the most discussed proteins in the manuscript are labeled in red font and marked with an asterisk. B, RPPA analysis showing proteins differentially altered by BEZ235 in MCAS-S and MCAS-R cells in 3D. Heatmap visualizes relative differences in signal induced by BEZ235 in MCAS-R cells compared with BEZ235-treated MCAS-S cells. Proteins with significant (t test, P < 0.05) changes are shown. Values from BEZ235-resistant line are normalized to the average of BEZ235-sensitive line. C, Western blot validation of the 3D protein arrays from A and B. D, Parental MCAS cells in 2D were transduced with pBABE-YAP and/or pWZL-MYC or plasmid controls, incubated with BEZ235, and counted. Fold change in cell number was calculated by normalizing the number of cells counted on day 10 in BEZ235 compared with the number of cells on day 10 in their own DMSO controls. Bottom panel shows validation of MYC/YAP expression. Statistical analysis: one-way ANOVA (P = 6.1839e−14) and Tukey HSD test, which showed significant differences between all pairs. **, P < 0.01. E, YAP was knocked down with shRNAs (F5 and F8) in control or MYC-overexpressing MCAS cells and cell numbers were analyzed as in D. The proliferation graphs are a combined experiment of two, with triplicates in each. All data shown as mean ± SEM. Statistical analysis: one-way ANOVA (P = 1.8652e−14) with Tukey HSD test. **, P < 0.01.

Figure 2.

MCAS-R cells exhibit increased MYC and YAP expression. A, Heatmap from RPPA analysis showing proteins/phospho-proteins differentially expressed in MCAS-S and MCAS-R lines following DMSO and BEZ235 treatment (ANOVA, P < 0.05). Data are median centered (red: greater than the median, green: less). For A and B, the most discussed proteins in the manuscript are labeled in red font and marked with an asterisk. B, RPPA analysis showing proteins differentially altered by BEZ235 in MCAS-S and MCAS-R cells in 3D. Heatmap visualizes relative differences in signal induced by BEZ235 in MCAS-R cells compared with BEZ235-treated MCAS-S cells. Proteins with significant (t test, P < 0.05) changes are shown. Values from BEZ235-resistant line are normalized to the average of BEZ235-sensitive line. C, Western blot validation of the 3D protein arrays from A and B. D, Parental MCAS cells in 2D were transduced with pBABE-YAP and/or pWZL-MYC or plasmid controls, incubated with BEZ235, and counted. Fold change in cell number was calculated by normalizing the number of cells counted on day 10 in BEZ235 compared with the number of cells on day 10 in their own DMSO controls. Bottom panel shows validation of MYC/YAP expression. Statistical analysis: one-way ANOVA (P = 6.1839e−14) and Tukey HSD test, which showed significant differences between all pairs. **, P < 0.01. E, YAP was knocked down with shRNAs (F5 and F8) in control or MYC-overexpressing MCAS cells and cell numbers were analyzed as in D. The proliferation graphs are a combined experiment of two, with triplicates in each. All data shown as mean ± SEM. Statistical analysis: one-way ANOVA (P = 1.8652e−14) with Tukey HSD test. **, P < 0.01.

Close modal

Because MYC and YAP, the Hippo pathway effector (for reviews, refs. 25, 26), are well-recognized regulators of cell proliferation, we focused on these two proteins and examined whether their modulation would be sufficient to impact the sensitivity to BEZ235. We first evaluated whether other structurally distinct PI3K/mTOR (GNE493) or mTOR (Torin1) inhibitors induce similar changes as observed with BEZ235 in 3D and observed upregulation in both YAP and MYC (Supplementary Fig. S1E). We next addressed whether overexpression of YAP and/or MYC in the parental MCAS cells would be sufficient to overcome proliferative suppression induced by BEZ235 under standard monolayer culture conditions (2D). Although overexpression of either YAP or MYC alone promoted growth in the presence of BEZ235 (Fig. 2D and Supplementary Fig. S2A, P < 0.01), concurrent overexpression of MYC and YAP significantly enhanced proliferation of MCAS cells compared with overexpression of either one alone (Fig. 2D, P < 0.01).

We also examined whether the ability of MCAS-R cells to evade proliferative suppression induced by BEZ235 was dependent on YAP or MYC. Although shRNA knockdown of YAP or MYC resulted in significant proliferation inhibition of BEZ235-treated MCAS-R cells in 3D (Supplementary Fig. S2B and S2C), knockdown of either one of these genes also impaired proliferation in the DMSO control cells, indicating that MYC and YAP are required for proliferation under 3D culture conditions. Thus, we were unable to assess the effects of YAP or MYC knockdown on BEZ235 resistance in MCAS-R cells in 3D. Interestingly, knockdown of YAP did not significantly impair proliferation of DMSO-treated parental MCAS cells grown as 2D cultures (Supplementary Fig. S2D). Because MYC has previously been reported to promote resistance to PI3K/mTOR inhibitors (14, 15), we examined the requirement for YAP in MYC overexpression–induced BEZ235 resistance in 2D. shRNA knockdown of endogenous YAP significantly sensitized MYC-overexpressing parental MCAS cells to BEZ235-induced proliferative suppression (Fig. 2E and Supplementary Fig. S2E, P < 0.01), indicating that the ability of MYC to drive proliferation under conditions of PI3K/mTOR inhibition is supported by endogenous YAP.

Endogenous YAP levels were unaffected by MYC expression, and vice versa (Supplementary Fig. S2F), indicating that codependence of these proteins is not due to direct regulation of one protein by the other. Therefore, we focused on identifying other potential upstream mechanisms by which these proteins could be coordinately upregulated during the establishment of resistance. The p38/MAPK, which is known to regulate stress responses when activated by phosphorylation (reviewed in ref. 27), was also identified by RPPA analysis as differentially regulated by BEZ235 in MCAS-R and MCAS-S cells (Fig. 2A and B). Given the known role of p38 in regulating response to cellular stress and cell-cycle arrest (28), and its role in regulating YAP levels (29, 30), we examined whether it could contribute to enhanced proliferation of MCAS-R cells in 3D. Of note, although BEZ235 treatment induced activation of p38 (phosphorylation of T180/T182) in both MCAS-R and MCAS-S cells in 3D (Figs. 2A and B and 3A), MCAS-R cells displayed significantly less induction of phospho-p38 than MCAS-S cells (Figs. 2A and B and 3A), raising the possibility that reduced activation of p38 could contribute to the increased proliferative capacity of MCAS-R cells in the presence of BEZ235.

Figure 3.

p38 and CREB regulate MYC and YAP. A, Phospho-p38 (T180/Y182) and p38 were probed by Western blot analysis from MCAS-R and MCAS-S cells in DMSO or BEZ235 in 3D for 6 days. B, Phase contrast images of parental MCAS cells cultured in 2D for 14 days in the presence of BEZ235 (BEZ) or preincubated with 5 μmol/L p38 inhibitor (SB203580) and then coincubated with BEZ235. C, Parental MCAS cells in 2D were treated with BEZ235 alone or BEZ235 + SB203580 for 14 days and cell numbers were counted on day 14; graph represents total cell numbers. D, Parental MCAS cells were incubated in 2D with DMSO, BEZ235, or combined BEZ235 and SB203580 for 14 days, and lysates were probed for CREB, MYC, YAP, p-p38, and p-MAPKAPK by Western blot analysis. E, MCAS-S and MCAS-R cells were cultured in 3D with BEZ235 and DMSO as in A, lysed and probed for CREB. F, Parental MCAS cells expressing wild-type CREB or control were cultured in BEZ235 in 2D. Cells were imaged at day 12 and counted at indicated time points. Graph shows fold change in cell number in BEZ235 on day 8 compared with its own vector control in DMSO on day 8. G, Western blot analysis of cells from F. Lysates were probed for CREB, MYC, and YAP. H, Parental MCAS cells overexpressing WT-CREB and YAP shRNAs were cultured in BEZ235 for the indicated time and fold change in cell number over time in BEZ235 normalized over cell number on day 0 (top) or determined and normalized as in F (bottom). I, Comparative analysis of YAP protein levels (RPPA score) in CREB amplified versus nonamplified TCGA ovarian cancer dataset. J, Breast, endometrioid, colon, and kidney cancer TCGA studies all show significant positive correlations between CREB and YAP mRNA expression in human tumors. The proliferation graphs are a representative experiment of three, and triplicates were analyzed. All data are shown as mean ± SEM. Statistical analysis, Student t test. *, P < 0.05; **, P < 0.01; ***, P < 0.005. Scale bars, 100 μm.

Figure 3.

p38 and CREB regulate MYC and YAP. A, Phospho-p38 (T180/Y182) and p38 were probed by Western blot analysis from MCAS-R and MCAS-S cells in DMSO or BEZ235 in 3D for 6 days. B, Phase contrast images of parental MCAS cells cultured in 2D for 14 days in the presence of BEZ235 (BEZ) or preincubated with 5 μmol/L p38 inhibitor (SB203580) and then coincubated with BEZ235. C, Parental MCAS cells in 2D were treated with BEZ235 alone or BEZ235 + SB203580 for 14 days and cell numbers were counted on day 14; graph represents total cell numbers. D, Parental MCAS cells were incubated in 2D with DMSO, BEZ235, or combined BEZ235 and SB203580 for 14 days, and lysates were probed for CREB, MYC, YAP, p-p38, and p-MAPKAPK by Western blot analysis. E, MCAS-S and MCAS-R cells were cultured in 3D with BEZ235 and DMSO as in A, lysed and probed for CREB. F, Parental MCAS cells expressing wild-type CREB or control were cultured in BEZ235 in 2D. Cells were imaged at day 12 and counted at indicated time points. Graph shows fold change in cell number in BEZ235 on day 8 compared with its own vector control in DMSO on day 8. G, Western blot analysis of cells from F. Lysates were probed for CREB, MYC, and YAP. H, Parental MCAS cells overexpressing WT-CREB and YAP shRNAs were cultured in BEZ235 for the indicated time and fold change in cell number over time in BEZ235 normalized over cell number on day 0 (top) or determined and normalized as in F (bottom). I, Comparative analysis of YAP protein levels (RPPA score) in CREB amplified versus nonamplified TCGA ovarian cancer dataset. J, Breast, endometrioid, colon, and kidney cancer TCGA studies all show significant positive correlations between CREB and YAP mRNA expression in human tumors. The proliferation graphs are a representative experiment of three, and triplicates were analyzed. All data are shown as mean ± SEM. Statistical analysis, Student t test. *, P < 0.05; **, P < 0.01; ***, P < 0.005. Scale bars, 100 μm.

Close modal

To examine the contribution of p38 to BEZ235 sensitivity, we inhibited p38 with SB203580 in parental MCAS cells treated with BEZ235 in 2D. This treatment significantly enhanced proliferation (P < 0.05) of BEZ235-treated cells, but had no effect on proliferation in the absence of BEZ235 (Fig. 3B and C and Supplementary Fig. S3A). After 14-day treatment, few cells remained in the BEZ235-treated samples, whereas small colonies of proliferating cells were apparent in samples treated with combined BEZ235 and p38 inhibitor (Fig. 3B). Combination treatment induced upregulation of YAP and MYC in parental MCAS cells (Fig. 3D and Supplementary Fig. S3B), as did BEZ235 treatment in combination with a structurally distinct p38 inhibitor (SB202190 and Supplementary Fig. S3C), indicating that in the presence of BEZ235, inhibition of p38 is sufficient to induce upregulation of YAP and MYC and result in proliferation.

Given that PI3K inhibition induces DNA damage (31, 32), a known regulator of p38 (27), we examined potential markers of DNA damage in the RPPA. We found that expression of XRCC1, Mre11, and X53BP1 was induced in the BEZ235-treated MCAS-S cells, whereas these proteins were all downregulated in MCAS-R cells (Supplementary Fig. S3D). We next assessed the extent of DNA damage by phospho-γH2AX immunostaining or immunofluorescence staining. BEZ235-treated MCAS-S cells displayed significantly more phospho-γHA2X–positive foci compared with MCAS-R cells (Supplementary Fig. S3E), raising the possibility that DNA damage induced by PI3K/mTOR inhibition drives activation of p38 in MCAS-S cells, and that, through an unknown mechanism, this stress is resolved in MCAS-R cells.

To determine which node of the PI3K/mTOR pathway regulates p38 activation, we treated parental MCAS cells in 2D with Torin1 (mTORi), BEZ235 (PI3Ki/mTORi), MK2206 (AKTi), GDC0941 (PI3Ki), or BKM120 (PI3Ki). Activation of p38 was not detected after short-term treatment (5 minutes), indicating that a long-term inhibition is required for p38 activation. After 72-hour incubation, mTOR inhibition (BEZ235 or Torin1) induced robust phospho-p38, whereas AKT and PI3K inhibition induced it to a lesser extent (Supplementary Fig. S3F). A detailed time-course demonstrated that phospho-p38 was detectable after 3-hour incubation and reached a peak after 72 hours (Supplementary Fig. S3F). These data suggest that p38 activation in response to mTOR inhibition is not mediated by an acute kinase/phosphatase cascade, but rather a mechanism that involves prolonged mTOR inhibition.

Interestingly, inhibition of p38 stabilizes cAMP response element-binding protein (CREB), which has been shown to positively regulate YAP levels (29, 30). To address whether CREB could contribute to YAP upregulation in MCAS-R cells in 3D, we analyzed total CREB levels in the MCAS lines after BEZ235 treatment. We detected a marked increase in CREB protein levels in BEZ235-treated MCAS-R cells, whereas only a slight increase in MCAS-S cells was observed (Fig. 3E). Prolonged BEZ235 treatment (14 days) of the parental MCAS cell line in 2D led to a marked increase in phospho-p38 and a decrease in CREB levels, and conversely p38 inhibition restored CREB levels in the parental BEZ235–treated cells (Fig. 3D and Supplementary Fig. S3C). These results are consistent with a model in which p38 activity, which is induced by BEZ235, suppresses CREB protein expression in the sensitive/parental cells.

To assess whether an increase in CREB protein levels is sufficient to promote resistance to mTOR inhibitors, we overexpressed wild-type CREB in parental MCAS cells and assessed proliferation in 2D. Overexpression of CREB significantly (P < 0.005) increased the proliferation of BEZ235-treated parental MCAS cells and induced upregulation of YAP and MYC but only under BEZ235 treatment (Fig. 3F and G and Supplementary Fig. S3G). These results suggest that reduced p38 activation contributes both to the increase in proliferation and the upregulation of MYC, YAP, and CREB levels observed in BEZ235-treated MCAS-R cells.

To evaluate the requirement for YAP downstream of CREB, we knocked down YAP in CREB-overexpressing parental MCAS cells. YAP depletion prevented proliferation in the presence of BEZ235, suggesting that YAP is required for CREB-induced proliferation under conditions of PI3K/mTOR inhibition (Fig. 3H). Furthermore, a comparative analysis of CREB-amplified versus nonamplified high-grade serous ovarian tumors revealed that YAP protein expression is significantly increased in tumors harboring CREB amplification (P = 0.031; Fig. 3I). We also noted a significant positive correlation between CREB and YAP mRNA expression in colon, kidney, breast, and liver tumors, suggesting that these proteins could be functionally coupled in patient tumors (Fig. 3J).

To assess whether YAP and MYC are required for resistance to PI3K/mTOR inhibitors in other cell lines, we analyzed a wide panel of tumor cell lines and identified three resistant lines that display continuous proliferation in the presence of BEZ235: the colon cancer cell line HCT116 and the ovarian cancer lines OVCAR5 and OvCa432 (Fig. 4A and Supplementary Fig. S4A and S4B). Notably, upregulation of YAP after 6-day BEZ235 treatment was specifically observed in all resistant cell lines, and upregulation of MYC in HCT116 and OVCAR5 cells, but not in the sensitive lines (HeLa, MCF7, and TOV21G; Fig. 4B). Although CREB upregulation was not detected following BEZ235 treatment, high basal CREB levels were specifically observed in the resistant cell lines (Fig. 4C). Furthermore, shRNA knockdown of either MYC or YAP significantly retarded proliferation in BEZ235-treated HCT116 cells (P < 0.001), the most resistant cell line we identified (Fig. 4D and Supplementary Fig. S4C). We also found that transient overexpression of constitutively active p38α significantly reduced the viability of HCT116 cells 2 days after BEZ235 treatment (P < 0.005; Fig. 4E), suggesting that activation of p38 also sensitizes intrinsically resistant cells to PI3K/mTOR inhibition.

Figure 4.

Several intrinsically resistant tumor cell lines upregulate MYC and YAP in BEZ235. A, Cell proliferation was measured for six cancer cell lines in the presence and absence of BEZ235 and is represented as fold change in cell number in BEZ235 on day 8 compared with the respective DMSO control on day 8. B, Western blot analysis of MYC and YAP in cell lines indicated in A. C, Cells were treated with BEZ235 for 8 days and CREB levels were analyzed by Western blot analysis. D, MYC or YAP was knocked down by shRNAs (MYCsh, B6/B8; YAPsh, F5/F8) in HCT116 cells and change in cell number is shown as in A, but each cell type is normalized to its own vector DMSO control. E, HCT116 cells were transfected with pcDNA3-GFP or constitutively active p38 (pcDNA3-p38CA), treated with BEZ235 for 48 hours, and cell proliferation was compared with proliferation in DMSO controls. All proliferation experiments were conducted with biological triplicates and repeated twice. All data are shown as mean ± SEM. Statistical analysis, Student t test. ***, P < 0.005; **, P < 0.01.

Figure 4.

Several intrinsically resistant tumor cell lines upregulate MYC and YAP in BEZ235. A, Cell proliferation was measured for six cancer cell lines in the presence and absence of BEZ235 and is represented as fold change in cell number in BEZ235 on day 8 compared with the respective DMSO control on day 8. B, Western blot analysis of MYC and YAP in cell lines indicated in A. C, Cells were treated with BEZ235 for 8 days and CREB levels were analyzed by Western blot analysis. D, MYC or YAP was knocked down by shRNAs (MYCsh, B6/B8; YAPsh, F5/F8) in HCT116 cells and change in cell number is shown as in A, but each cell type is normalized to its own vector DMSO control. E, HCT116 cells were transfected with pcDNA3-GFP or constitutively active p38 (pcDNA3-p38CA), treated with BEZ235 for 48 hours, and cell proliferation was compared with proliferation in DMSO controls. All proliferation experiments were conducted with biological triplicates and repeated twice. All data are shown as mean ± SEM. Statistical analysis, Student t test. ***, P < 0.005; **, P < 0.01.

Close modal

To assess whether PI3K/mTOR inhibition induces YAP and MYC upregulation in vivo, NOD-SCID mice were subcutaneously injected with resistant HCT116 and OVCAR5 cells and, once tumors were formed, were treated with vehicle or PI3K/mTOR inhibitor GNE493 (10 mg/kg; ref. 33). GNE493 treatment significantly increased intensity of both YAP (Fig. 5A and C) and MYC staining (Fig. 5B and D) in the Ki67-positive areas of HCT116 and OVCAR5 tumors (Fig. 5E), indicating that both YAP and MYC are upregulated in proliferating tumor cells in vivo under PI3K/mTOR inhibition.

Figure 5.

YAP and MYC are upregulated in mTOR inhibitor–treated xenograft tumors. HCT116 (A and B) and OVCAR5 cells (C and D) were injected subcutaneously into five NOD-SCID mice and following tumor formation, mice were treated every 24 hours with PI3K/mTOR inhibitor GNE493 (10 mg/kg) or vehicle for 8 to 14 days. Tumors were harvested and stained for Ki67 (red) and MYC (B and D, green) or YAP (A and C, green). E, Quantification of fluorescence intensity values of MYC and YAP staining in vehicle- and GNE493-treated tumors. Data are shown as mean ± SEM. Statistical analysis, Student t test. *, P < 0.05; **, P < 0.01; ***, P < 0.005. Scale bar, 50 μm.

Figure 5.

YAP and MYC are upregulated in mTOR inhibitor–treated xenograft tumors. HCT116 (A and B) and OVCAR5 cells (C and D) were injected subcutaneously into five NOD-SCID mice and following tumor formation, mice were treated every 24 hours with PI3K/mTOR inhibitor GNE493 (10 mg/kg) or vehicle for 8 to 14 days. Tumors were harvested and stained for Ki67 (red) and MYC (B and D, green) or YAP (A and C, green). E, Quantification of fluorescence intensity values of MYC and YAP staining in vehicle- and GNE493-treated tumors. Data are shown as mean ± SEM. Statistical analysis, Student t test. *, P < 0.05; **, P < 0.01; ***, P < 0.005. Scale bar, 50 μm.

Close modal

To determine whether resistance to mTOR inhibitors is correlated with mutational status, we performed a proliferation analysis on a panel of cancer cell lines in the presence of mTOR inhibition (Fig. 6A). The proliferation index and resistance to mTOR inhibitors correlated with the mutational status of the KRAS/ERK pathway (Fig. 6B). Of the resistant cell lines we previously identified, MCAS, HCT116, OVCAR5, and OVCA432 harbor activating mutations in KRAS. Consistent with these results, knockdown of KRAS sensitized HCT116 cells to PI3K/mTOR inhibition (Fig. 6C). Moreover, combined treatment of HCT116 cells with BEZ235 and 10 μmol/L MEK inhibitor (UO126) prevented upregulation of MYC and YAP, suggesting that RAS/MEK/ERK pathway is required for MYC and YAP upregulation (Fig. 6D).

Figure 6.

Activated KRAS pathway promotes resistance. A, Cell lines were cultured in 2D with mTOR inhibitor Torin 1 for 6 days and relative growth in Torin 1 compared with DMSO is shown. Cell lines with BRAF, RAS, or NF1 mutation are indicated. *, cell line with activated RAS with no known mutation (56). B, Relative cell number in Torin 1 compared with DMSO in RAS-activated cells versus non-RAS–activated cells. C, KRAS was knocked down in HCT116 cells and data are shown as fold change in cell number in BEZ235 on day 6 compared with each vectors DMSO control on day 6. Bottom, validation of RAS knockdown. KRAS shRNA A7 resulted in cell death in DMSO and could not be used. Proliferation experiment was done twice with triplicates. D, HCT116 cells were cultured in 2D for 6 days in the presence of DMSO, BEZ235, or BEZ235 and 10 μmol/L UO126 and probed for MYC and YAP. E, HCT116 in 2D and MCAS-R and -S cells in 3D were cultured for 48 hours with DMSO or BEZ235 and probed for p-ERK. F, MCAS-R and HCT116 cells were grown with DMSO, 0.5 μmol/L BEZ235, or BEZ235 and UO126 (10 μmol/L). Lysates were collected after 48 hours and probed for CREB and actin. G and H, Parental MCAS cells (G) and HCT116 cells (H) were cultured in 2D with indicated inhibitors and counted on day 0 and day 5 (HCT116) or day 7 (MCAS). Fold change in cell number was calculated by comparing the cell number at the end of the experiment to that on day 0. Experiments were repeated twice with triplicates. Cells were lysed on day 2 (MCAS) and day 4 (HCT116) and probed for MYC, YAP, and actin. All data are shown as mean ± SEM. Statistical analysis, Student t test. *, P < 0.05, **, P < 0.01, ***, P < 0.005.

Figure 6.

Activated KRAS pathway promotes resistance. A, Cell lines were cultured in 2D with mTOR inhibitor Torin 1 for 6 days and relative growth in Torin 1 compared with DMSO is shown. Cell lines with BRAF, RAS, or NF1 mutation are indicated. *, cell line with activated RAS with no known mutation (56). B, Relative cell number in Torin 1 compared with DMSO in RAS-activated cells versus non-RAS–activated cells. C, KRAS was knocked down in HCT116 cells and data are shown as fold change in cell number in BEZ235 on day 6 compared with each vectors DMSO control on day 6. Bottom, validation of RAS knockdown. KRAS shRNA A7 resulted in cell death in DMSO and could not be used. Proliferation experiment was done twice with triplicates. D, HCT116 cells were cultured in 2D for 6 days in the presence of DMSO, BEZ235, or BEZ235 and 10 μmol/L UO126 and probed for MYC and YAP. E, HCT116 in 2D and MCAS-R and -S cells in 3D were cultured for 48 hours with DMSO or BEZ235 and probed for p-ERK. F, MCAS-R and HCT116 cells were grown with DMSO, 0.5 μmol/L BEZ235, or BEZ235 and UO126 (10 μmol/L). Lysates were collected after 48 hours and probed for CREB and actin. G and H, Parental MCAS cells (G) and HCT116 cells (H) were cultured in 2D with indicated inhibitors and counted on day 0 and day 5 (HCT116) or day 7 (MCAS). Fold change in cell number was calculated by comparing the cell number at the end of the experiment to that on day 0. Experiments were repeated twice with triplicates. Cells were lysed on day 2 (MCAS) and day 4 (HCT116) and probed for MYC, YAP, and actin. All data are shown as mean ± SEM. Statistical analysis, Student t test. *, P < 0.05, **, P < 0.01, ***, P < 0.005.

Close modal

A role for ERK in mediating resistance to PI3K inhibitors has been reported (6, 11, 12). Given our data showing a correlation between PI3K/mTOR inhibitor resistance and KRAS mutational status, and that MYC and YAP upregulation was blocked by inhibition of the MEK/ERK pathway, we probed for ERK activation in HCT116, MCAS-S, and MCAS-R cells. BEZ235 treatment induced significant increase in phospho-ERK levels, including MCAS-S cells (Fig. 6E), suggesting that ERK activation alone is not sufficient to induce MYC and YAP upregulation. Because ERK can regulate CREB (34), we examined whether ERK inhibition positively or negatively affects CREB expression in resistant HCT116 and MCAS-R cells. Combined treatment with BEZ235 and UO126 inhibited CREB expression (Fig. 6F), suggesting that ERK activation is required for CREB expression in the resistant cells, and that p38 and ERK have opposing roles in the regulation of CREB protein levels.

Given the opposing effects of p38 and ERK on BEZ235 responsiveness, we next assessed the degree to which these pathways contribute to the acquired and intrinsic resistance. To do this, we evaluated the effects of p38 inhibition in combination with BEZ235 and UO126 on survival and MYC and YAP upregulation in HCT116 and parental MCAS cells. Unlike in the MCAS cells, the addition of p38 inhibitor with BEZ235 did not result in any additional upregulation of MYC and YAP in HCT116 cells. However, the addition of p38 inhibitor to BEZ235 and UO126 treatment resulted in significantly improved cell survival and upregulation of MYC and YAP in both cell lines (Fig. 6G and H). These data imply that, even in a KRAS-mutant background, the efficacy of combined inhibition of the PI3K/mTOR and MEK/ERK pathways can be enhanced by the activation of p38, and conversely, that suppression of p38 can promote resistance through upregulation of MYC and YAP. Interestingly, combined treatment with p38 inhibitor and UO126 also resulted in improved cell survival (Fig. 6G and H and Supplementary Fig. S4D), suggesting that activation of p38 could also be an effective strategy to improve responsiveness to MEK/ERK inhibitors in KRAS-mutant tumors.

Taken together, these data support a model (Fig. 7) in which ERK and p38 play opposing roles in regulating the transcription factor CREB in tumor cells carrying mutations in the RAS/ERK pathway, with ERK promoting CREB protein expression and p38 destabilizing it. Elevated CREB expression positively regulates YAP and MYC, allowing tumor cells to escape from proliferative suppression imposed by PI3K/mTOR inhibitors.

Figure 7.

Model of signaling events leading to MYC and YAP upregulation. A, In KRAS WT cells, inhibition of PI3K/mTOR leads to upregulation of p38 through activated DNA damage pathway (and inhibition of CREB and MYC) and upregulation of receptor tyrosine kinases, with only a minor increase in p-ERK. B, In KRAS-mutant background, increased MEK/ERK signaling leads to slightly higher MYC levels, but high p-p38 still opposes MYC, CREB, and YAP. C, In KRAS-mutant cells that have downregulated p38 activity by relieving DNA damage response through an unknown mechanism, suppression of MYC and CREB is relieved and YAP is upregulated.

Figure 7.

Model of signaling events leading to MYC and YAP upregulation. A, In KRAS WT cells, inhibition of PI3K/mTOR leads to upregulation of p38 through activated DNA damage pathway (and inhibition of CREB and MYC) and upregulation of receptor tyrosine kinases, with only a minor increase in p-ERK. B, In KRAS-mutant background, increased MEK/ERK signaling leads to slightly higher MYC levels, but high p-p38 still opposes MYC, CREB, and YAP. C, In KRAS-mutant cells that have downregulated p38 activity by relieving DNA damage response through an unknown mechanism, suppression of MYC and CREB is relieved and YAP is upregulated.

Close modal

Our data provide evidence for a previously unrecognized mode of drug resistance to PI3K/mTOR inhibitors, which is observed in both acquired and intrinsically resistant cell lines, that involves rewiring several parallel signaling pathways that coordinately promote induction of YAP and MYC, both of which are required for proliferation under conditions of PI3K/mTOR inhibition (Fig. 7). Furthermore, our data support a model (Fig. 7) in which both the ERK and p38/MAPK pathways play critical, dichotomous roles in regulating resistance.

In cells that are sensitive to PI3K/mTOR inhibitors, drug treatment induces cytostasis (Fig. 7A). This is likely mediated in part through p38 activation, which regulates cell-cycle arrest at G1–S and G2–M checkpoints (28, 35). Our data indicate that suppression of p38 activity (Fig. 7C) allows resistant cells to escape proliferation inhibition induced by PI3K/mTOR inhibitors. This conclusion is supported by several findings: (i) p38 activation positively correlated with proliferation suppression by PI3K/mTOR inhibitors; (ii) p38 activity was significantly lower in resistant cells compared with sensitive cells; and (iii) p38 kinase inhibitors promoted proliferation, and YAP and MYC expression in the presence of BEZ235. Furthermore, p38 has been shown to regulate MYC expression at the posttranscriptional level (36), potentially explaining why MYC levels increased after combined BEZ235 and p38 inhibitor treatment. These data suggest that reduction in p38 activation in tumor cells could both release negative controls on proliferation as well as induce positive regulators of proliferation.

Our findings indicate that the mechanism of YAP upregulation upon p38 inhibition involves the transcription factor CREB. A recent report described an auto-regulatory feedback loop between p38, CREB, and YAP whereby p38 inhibition leads to CREB stabilization (37). It was also shown that CREB binds to the YAP promoter and enhances its transcription (29, 30). Indeed, we observed that overexpression of CREB increased YAP levels, and inhibition of p38 led to increased CREB and YAP expression, suggesting a functional link between p38, CREB, and YAP. Furthermore, TCGA database analysis showed positive correlation between CREB and YAP expression in several tumor datasets, suggesting that these proteins could also be functionally coupled in patient tumors.

p38 is activated in response to cellular stress, such as DNA damage (reviewed in ref. 27), and also in response to inhibition of oncogenic kinases, such as EGFR, Bcr-Abl, and Src (38). Although our data do not delineate how p38 activity is suppressed in the resistant cells, our results and that of others suggest that acute PI3K/mTOR inhibition leads to DNA damage and activation of DNA repair pathways (31, 32). However, in the resistant cells, p38 activity was significantly suppressed, as were markers of DNA repair, suggesting that these cells have resolved the DNA damage stress leading to p38 activation.

Our study also implicates the RAS/ERK pathway in promoting drug resistance and upregulation of YAP and MYC by three findings: (i) the strong correlation between mutations in KRAS/BRAF and resistance to PI3K/mTOR inhibition, (ii) the suppression of YAP, MYC, and CREB expression in resistant cells treated with a MEK inhibitor; and (iii) the loss of resistance induced by downregulation of mutant KRAS in HCT116 cells. These results suggest that the activity of MEK/ERK pathway positively influences resistance, and that mutations in the RAS pathway provide a context in which a MYC–YAP resistance mechanism can be operative. Furthermore, ERK pathway has been shown to stabilize MYC protein (39, 40), and the RAS pathway has been implicated in YAP protein stabilization (41). However, YAP has been shown to promote resistance to RAF/MEK–targeted cancer therapies in RAS-mutant tumors (42, 43) and to rescue cancer cells after KRAS loss (44, 45), suggesting that YAP can be regulated at several levels to promote resistance to multiple targeted therapies.

Interestingly, both p38 and ERK pathway activity, as well as reduced PI3K/AKT pathway activity, have been linked to tumor cell dormancy (46–49). Results from the Aguirre-Ghiso group and others have shown that dormancy of tumor cells is dependent on the ratio of ERK and p38 activity, where a higher ERK/p38 ratio leads to cell-cycle re-entry, whereas a higher p38/ERK ratio leads to tumor dormancy (50–53). Using a tumor model driven by HER2, Chodosh and colleagues showed that although withdrawal of this oncogene results in tumor regression, mice ultimately develop recurrent tumors that have become HER2 independent (54), suggesting the existence of residual dormant tumor cells. Although Notch signaling was implicated in maintenance of dormancy, its inhibition did not inhibit tumor regrowth (55), suggesting that another pathway regulates the growth of recurrent tumors.

Chronic PI3K/mTOR inhibition might induce a state resembling dormancy, perhaps through activation of p38 and cell-cycle arrest (Fig. 7A). However, in the context of a mutated RAS/ERK pathway, and another unidentified factor that results in downregulation of phospho-p38, the balance between p38 and ERK is shifted, and ERK becomes more dominant under PI3K/mTOR inhibition. The reduction in p38 leads to an increase in CREB, YAP, and MYC levels, allowing a low level of proliferation resembling “awakening” from dormancy (Fig. 7B and C). We speculate that although the acquired resistance is mainly driven by suppressed p38 signaling, the intrinsic resistance would be more dependent on active ERK signaling. However, p38 inhibition in combination with PI3K/mTOR and MEK inhibitors resulted in improved fitness in both models, suggesting that activation of p38 would also benefit KRAS-mutant tumors, a possibility worth exploring because of its translational implications.

In conclusion, these data suggest that in the absence of PI3K/mTOR signaling, MYC and YAP can orchestrate the required steps for cellular growth and proliferation normally driven by the PI3K/mTOR pathways. We also demonstrate that activation of the RAS/ERK pathway alone is not sufficient to allow proliferation in the presence of chronic PI3K/mTOR inhibition, but also requires suppression of the stress-activated p38/MAPK. This implies that tumors treated with PI3K/mTOR inhibitors enter a state similar to tumor dormancy, and can evade therapies aimed at targeting proliferating cells.

G.B. Mills reports receiving a commercial research grant from AstraZeneca, Critical Outcome Technologies, Komen Research Foundation, Nanostring, Breast Cancer Research Foundation, and Karus. G.B. Mills has received speakers bureau honoraria from Symphogen, Nuevolution, AstraZeneca, ISIS Pharmaceuticals, Lilly, and Novartis, has ownership interest (including patents) in Catena Pharmaceuticals, PTV Ventures, Spindletop Ventures, and Myriad Genetics, and is a consultant/advisory board member for Adventist Health, AstraZeneca, Provista Diagnostics, Signalchem Lifesciences, Symphogen, Blend, Catena Pharmaceuticals, Critical Outcome Technologies, HanAl Bio Korea, ImmunoMET, Millenium Pharmaceuticals, Nuevolution, and Precision Medicine. No potential conflicts of interest were disclosed by the other authors.

Conception and design: T. Muranen, D.M. Sabatini, G.B. Mills, J.S. Brugge

Development of methodology: T. Muranen, L.L. Gallegos, D.M. Sabatini

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): T. Muranen, J.L. Coloff, C.C. Thoreen, S.A. Kang, D.M. Sabatini, G.B. Mills

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): T. Muranen, L.M. Selfors, S.A. Kang, D.M. Sabatini, G.B. Mills, J.S. Brugge

Writing, review, and/or revision of the manuscript: T. Muranen, L.L. Gallegos, C.C. Thoreen, D.M. Sabatini, G.B. Mills, J.S. Brugge

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): J. Hwang, D.M. Sabatini

Study supervision: T. Muranen, J.S. Brugge

We thank Dennis Slamon for providing cell lines, Jennifer Waters and Nikon Imaging Center at Harvard Medical School for assistance with microscopy, Deepak Sampath (Genentech) for providing GNE493, Julio Aguirre-Ghiso for providing p38α construct, Angie Martinez-Gakidis for critical reading of the manuscript, Valerie Pireaux for technical help, Shomit Sengupta and members of the Brugge laboratory for helpful discussions.

This work was supported by grant NIC-5K99CA180221 and Laura Ziskin Memorial Award (Entertainment Industry Foundation to T. Muranen). J.S. Brugge received a gift from the Dr. Miriam and Sheldon G. Adelson Medical Research Foundation. D.M. Sabatini was supported by NIH grants R01 CA103866 and AI47389. G.B. Mills has been supported by grants 0099031 U54CA112970, BCRF 01-06-00332, Komen Foundation (KG08169404 and SAC110052), MDACC CCSG grant P30 CA016672, and a gift from the Dr. Miriam and Sheldon G. Adelson Medical Research Foundation.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Dibble
CC
,
Manning
BD
. 
Signal integration by mTORC1 coordinates nutrient input with biosynthetic output
.
Nat Cell Biol
2013
;
15
:
555
64
.
2.
Engelman
JA
. 
Targeting PI3K signalling in cancer: opportunities, challenges and limitations
.
Nat Rev Cancer
2009
;
9
:
550
62
.
3.
Samuels
Y
,
Velculescu
VE
. 
Oncogenic mutations of PIK3CA in human cancers
.
Cell Cycle
2004
;
3
:
1221
4
.
4.
Rodon
J
,
Dienstmann
R
,
Serra
V
,
Tabernero
J
. 
Development of PI3K inhibitors: lessons learned from early clinical trials
.
Nat Rev Clin Oncol
2013
;
10
:
143
53
.
5.
Chandarlapaty
S
,
Sawai
A
,
Scaltriti
M
,
Rodrik-Outmezguine
V
,
Grbovic-Huezo
O
,
Serra
V
, et al
AKT inhibition relieves feedback suppression of receptor tyrosine kinase expression and activity
.
Cancer Cell
2011
;
19
:
58
71
.
6.
Serra
V
,
Scaltriti
M
,
Prudkin
L
,
Eichhorn
PJA
,
Ibrahim
YH
,
Chandarlapaty
S
, et al
PI3K inhibition results in enhanced HER signaling and acquired ERK dependency in HER2-overexpressing breast cancer
.
Oncogene
2011
;
30
:
2547
57
7.
Garrett
JT
,
Olivares
MG
,
Rinehart
C
,
Granja-Ingram
ND
,
Sánchez
V
,
Chakrabarty
A
, et al
Transcriptional and posttranslational up-regulation of HER3 (ErbB3) compensates for inhibition of the HER2 tyrosine kinase
.
Proc Natl Acad Sci U S A
2011
;
108
:
5021
6
.
8.
Muranen
T
,
Selfors
LM
,
Worster
DT
,
Iwanicki
MP
,
Song
L
,
Morales
FC
, et al
Inhibition of PI3K/mTOR leads to adaptive resistance in matrix-attached cancer cells
.
Cancer Cell
2012
;
21
:
227
39
.
9.
Sergina
NV
,
Rausch
M
,
Wang
D
,
Blair
J
,
Hann
B
,
Shokat
KM
, et al
Escape from HER-family tyrosine kinase inhibitor therapy by the kinase-inactive HER3
.
Nature
2007
;
445
:
437
41
.
10.
O'Reilly
KE
,
Rojo
F
,
She
Q-B
,
Solit
D
,
Mills
GB
,
Smith
D
, et al
mTOR inhibition induces upstream receptor tyrosine kinase signaling and activates Akt
.
Cancer Res
2006
;
66
:
1500
8
.
11.
Carracedo
A
,
Ma
L
,
Teruya-Feldstein
J
,
Rojo
F
,
Salmena
L
,
Alimonti
A
, et al
Inhibition of mTORC1 leads to MAPK pathway activation through a PI3K-dependent feedback loop in human cancer
.
J Clin Invest
2008
;
118
:
3065
74
.
12.
Engelman
JA
,
Chen
L
,
Tan
X
,
Crosby
K
,
Guimaraes
AR
,
Upadhyay
R
, et al
Effective use of PI3K and MEK inhibitors to treat mutant Kras G12D and PIK3CA H1047R murine lung cancers
.
Nat Med
2008
;
14
:
1351
6
.
13.
Serra
V
,
Eichhorn
PJ
,
Garcia-Garcia
C
,
Ibrahim
YH
,
Prudkin
L
,
Sanchez
G
, et al
RSK3/4 mediate resistance to PI3K pathway inhibitors in breast cancer
.
J Clin Invest
2013
;
123
:
2551
63
.
14.
Liu
P
,
Cheng
H
,
Santiago
S
,
Raeder
M
,
Zhang
F
,
Isabella
A
, et al
Oncogenic PIK3CA-driven mammary tumors frequently recur via PI3K pathway-dependent and PI3K pathway-independent mechanisms
.
Nat Med
2011
;
17
:
1116
20
.
15.
Ilic
N
,
Utermark
T
,
Widlund
HR
,
Roberts
TM
. 
PI3K-targeted therapy can be evaded by gene amplification along the MYC-eukaryotic translation initiation factor 4E (eIF4E) axis
.
Proc Natl Acad Sci USA
2011
;
108
:
E699
708
16.
Tenbaum
SP
,
Ordóñez-Morán
P
,
Puig
I
,
Chicote
I
,
Arqués
O
,
Landolfi
S
, et al
β-Catenin confers resistance to PI3K and AKT inhibitors and subverts FOXO3a to promote metastasis in colon cancer
.
Nat Med
2012
;
18
:
892
901
.
17.
Cope
CL
,
Gilley
R
,
Balmanno
K
,
Sale
MJ
,
Howarth
KD
,
Hampson
M
, et al
Adaptation to mTOR kinase inhibitors by amplification of eIF4E to maintain cap-dependent translation
.
J Cell Sci
2014
;
127
:
788
800
.
18.
Alain
T
,
Morita
M
,
Fonseca
BD
,
Yanagiya
A
,
Siddiqui
N
,
Bhat
M
, et al
eIF4E/4E-BP ratio predicts the efficacy of mTOR targeted therapies
.
Cancer Res
2012
;
72
:
6468
76
.
19.
Ducker
GS
,
Atreya
CE
,
Simko
JP
,
Hom
YK
,
Matli
MR
,
Benes
CH
, et al
Incomplete inhibition of phosphorylation of 4E-BP1 as a mechanism of primary resistance to ATP-competitive mTOR inhibitors
.
Oncogene
2014
;
33
:
1590
600
.
20.
Dowling
RJO
,
Topisirovic
I
,
Alain
T
,
Bidinosti
M
,
Fonseca
BD
,
Petroulakis
E
, et al
mTORC1-mediated cell proliferation, but not cell growth, controlled by the 4E-BPs
.
Science
2010
;
328
:
1172
6
.
21.
Hennessy
BT
,
Lu
Y
,
Gonzalez-Angulo
AM
,
Carey
MS
,
Myhre
S
,
Ju
Z
, et al
A technical assessment of the utility of reverse phase protein arrays for the study of the functional proteome in non-microdissected human breast cancers
.
Clin Proteomics
2010
;
6
:
129
51
.
22.
Ory
DS
,
Neugeboren
BA
,
Mulligan
RC
. 
A stable human-derived packaging cell line for production of high titer retrovirus/vesicular stomatitis virus G pseudotypes
.
Proc Natl Acad Sci U S A
1996
;
93
:
11400
6
.
23.
Debnath
J
,
Muthuswamy
SK
,
Brugge
JS
. 
Morphogenesis and oncogenesis of MCF-10A mammary epithelial acini grown in three-dimensional basement membrane cultures
.
Methods
2003
;
30
:
256
68
.
24.
Carson
AR
,
Smith
EN
,
Matsui
H
,
Braekkan
SK
,
Jepsen
K
,
Hansen
JB
, et al
Effective filtering strategies to improve data quality from population-based whole exome sequencing studies
.
BMC Bioinform
2014
;
15
:
125
.
25.
Tumaneng
K
,
Russell Ryan
C
,
Guan
K-L
. 
Organ size control by hippo and TOR pathways
.
Curr Biol
2012
;
22
:
R368
79
.
26.
Yu
F-X
,
Guan
K-L
. 
The Hippo pathway: regulators and regulations
.
Genes Dev
2013
;
27
:
355
71
.
27.
Coulthard
LR
,
White
DE
,
Jones
DL
,
McDermott
MF
,
Burchill
SA
. 
p38(MAPK): stress responses from molecular mechanisms to therapeutics
.
Trends Mol Med
2009
;
15
:
369
79
.
28.
Reinhardt
HC
,
Hasskamp
P
,
Schmedding
I
,
Morandell
S
,
van Vugt
MA
,
Wang
X
, et al
DNA damage activates a spatially distinct late cytoplasmic cell-cycle checkpoint network controlled by MK2-mediated RNA stabilization
.
Mol Cell
2010
;
40
:
34
49
.
29.
Wang
J
,
Ma
L
,
Weng
W
,
Qiao
Y
,
Zhang
Y
,
He
J
, et al
Mutual interaction between YAP and CREB promotes tumorigenesis in liver cancer
.
Hepatology
2013
;
58
:
1011
20
.
30.
Zhang
T
,
Zhang
J
,
You
X
,
Liu
Q
,
Du
Y
,
Gao
Y
, et al
Hepatitis B virus X protein modulates oncogene Yes-associated protein by CREB to promote growth of hepatoma cells
.
Hepatology
2012
;
56
:
2051
9
.
31.
Juvekar
A
,
Burga
LN
,
Hu
H
,
Lunsford
EP
,
Ibrahim
YH
,
Balmana
J
, et al
Combining a PI3K inhibitor with a PARP inhibitor provides an effective therapy for BRCA1-related breast cancer
.
Cancer Discov
2012
;
2
:
1048
63
.
32.
Ibrahim
YH
,
Garcia-Garcia
C
,
Serra
V
,
He
L
,
Torres-Lockhart
K
,
Prat
A
, et al
PI3K inhibition impairs BRCA1/2 expression and sensitizes BRCA-proficient triple-negative breast cancer to PARP inhibition
.
Cancer Discov
2012
;
2
:
1036
47
.
33.
Sutherlin
DP
,
Sampath
D
,
Berry
M
,
Castanedo
G
,
Chang
Z
,
Chuckowree
I
, et al
Discovery of (thienopyrimidin-2-yl)aminopyrimidines as potent, selective, and orally available pan-PI3-kinase and dual pan-PI3-kinase/mTOR inhibitors for the treatment of cancer
.
J Med Chem
2010
;
53
:
1086
97
.
34.
Deak
M
,
Clifton
AD
,
Lucocq
LM
,
Alessi
DR
. 
Mitogen- and stress-activated protein kinase-1 (MSK1) is directly activated by MAPK and SAPK2/p38, and may mediate activation of CREB
.
EMBO J
1998
;
17
:
4426
41
.
35.
Manke
IA
,
Nguyen
A
,
Lim
D
,
Stewart
MQ
,
Elia
AE
,
Yaffe
MB
. 
MAPKAP kinase-2 is a cell cycle checkpoint kinase that regulates the G2/M transition and S phase progression in response to UV irradiation
.
Mol Cell
2005
;
17
:
37
48
.
36.
Cannell
IG
,
Kong
YW
,
Johnston
SJ
,
Chen
ML
,
Collins
HM
,
Dobbyn
HC
, et al
p38 MAPK/MK2-mediated induction of miR-34c following DNA damage prevents Myc-dependent DNA replication
.
Proc Natl Acad Sci U S A
2010
;
107
:
5375
80
.
37.
Wang
J
,
Park
J-S
,
Wei
Y
,
Rajurkar
M
,
Cotton Jennifer
L
,
Fan
Q
, et al
TRIB2 acts downstream of Wnt/TCF in liver cancer cells to regulate YAP and C/EBP&alpha; Function
.
Mol Cell
2013
;
51
:
211
25
38.
Sharma
SV
,
Gajowniczek
P
,
Way
IP
,
Lee
DY
,
Jiang
J
,
Yuza
Y
, et al
A common signaling cascade may underlie “addiction” to the Src, BCR-ABL, and EGF receptor oncogenes
.
Cancer Cell
2006
;
10
:
425
35
.
39.
Sears
R
,
Leone
G
,
DeGregori
J
,
Nevins
JR
. 
Ras enhances Myc protein stability
.
Mol Cell
1999
;
3
:
169
79
.
40.
Sears
R
,
Nuckolls
F
,
Haura
E
,
Taya
Y
,
Tamai
K
,
Nevins
JR
. 
Multiple Ras-dependent phosphorylation pathways regulate Myc protein stability
.
Genes Dev
2000
;
14
:
2501
14
.
41.
Hong
X
,
Nguyen
HT
,
Chen
Q
,
Zhang
R
,
Hagman
Z
,
Voorhoeve
PM
, et al
Opposing activities of the Ras and Hippo pathways converge on regulation of YAP protein turnover
.
EMBO J
2014
;
33
:
2447
57
.
42.
Lin
L
,
Sabnis
AJ
,
Chan
E
,
Olivas
V
,
Cade
L
,
Pazarentzos
E
, et al
The Hippo effector YAP promotes resistance to RAF- and MEK-targeted cancer therapies
.
Nat Genet
2015
;
47
:
250
6
.
43.
Kim
MH
,
Kim
J
,
Hong
H
,
Lee
SH
,
Lee
JK
,
Jung
E
, et al
Actin remodeling confers BRAF inhibitor resistance to melanoma cells through YAP/TAZ activation
.
EMBO J
2016
;
35
:
462
78
.
44.
Shao
DD
,
Xue
W
,
Krall
EB
,
Bhutkar
A
,
Piccioni
F
,
Wang
X
, et al
KRAS and YAP1 converge to regulate EMT and tumor survival
.
Cell
2014
;
158
:
171
84
.
45.
Kapoor
A
,
Yao
W
,
Ying
H
,
Hua
S
,
Liewen
A
,
Wang
Q
, et al
Yap1 activation enables bypass of oncogenic Kras addiction in pancreatic cancer
.
Cell
2014
;
158
:
185
97
.
46.
Jo
H
,
Jia
Y
,
Subramanian
KK
,
Hattori
H
,
Luo
HR
. 
Cancer cell-derived clusterin modulates the phosphatidylinositol 3′-kinase-Akt pathway through attenuation of insulin-like growth factor 1 during serum deprivation
.
Mol Cell Biol
2008
;
28
:
4285
99
.
47.
Humtsoe
JO
,
Kramer
RH
. 
Differential epidermal growth factor receptor signaling regulates anchorage-independent growth by modulation of the PI3K/AKT pathway
.
Oncogene
2010
;
29
:
1214
26
.
48.
Balz
LM
,
Bartkowiak
K
,
Andreas
A
,
Pantel
K
,
Niggemann
B
,
Zanker
KS
, et al
The interplay of HER2/HER3/PI3K and EGFR/HER2/PLC-gamma1 signalling in breast cancer cell migration and dissemination
.
J Pathol
2012
;
227
:
234
44
.
49.
Schewe
DM
,
Aguirre-Ghiso
JA
. 
ATF6alpha-Rheb-mTOR signaling promotes survival of dormant tumor cells invivo
.
Proc Natl Acad Sci U S A
2008
;
105
:
10519
24
.
50.
Diehl
NL
,
Enslen
H
,
Fortner
KA
,
Merritt
C
,
Stetson
N
,
Charland
C
, et al
Activation of the p38 mitogen-activated protein kinase pathway arrests cell cycle progression and differentiation of immature thymocytes invivo
.
J Exp Med
2000
;
191
:
321
34
.
51.
Aguirre-Ghiso
JA
,
Estrada
Y
,
Liu
D
,
Ossowski
L
. 
ERK(MAPK) activity as a determinant of tumor growth and dormancy; regulation by p38(SAPK)
.
Cancer Res
2003
;
63
:
1684
95
.
52.
Aguirre-Ghiso
JA
,
Ossowski
L
,
Rosenbaum
SK
. 
Green fluorescent protein tagging of extracellular signal-regulated kinase and p38 pathways reveals novel dynamics of pathway activation during primary and metastatic growth
.
Cancer Res
2004
;
64
:
7336
45
.
53.
Bragado
P
,
Estrada
Y
,
Parikh
F
,
Krause
S
,
Capobianco
C
,
Farina
HG
, et al
TGF-beta2 dictates disseminated tumour cell fate in target organs through TGF-beta-RIII and p38alpha/beta signalling
.
Nat Cell Biol
2013
;
15
:
1351
61
.
54.
Moody
SE
,
Sarkisian
CJ
,
Hahn
KT
,
Gunther
EJ
,
Pickup
S
,
Dugan
KD
, et al
Conditional activation of Neu in the mammary epithelium of transgenic mice results in reversible pulmonary metastasis
.
Cancer Cell
2002
;
2
:
451
61
.
55.
Abravanel
DL
,
Belka
GK
,
Pan
TC
,
Pant
DK
,
Collins
MA
,
Sterner
CJ
, et al
Notch promotes recurrence of dormant tumor cells following HER2/neu-targeted therapy
.
J Clin Invest
2015
;
125
:
2484
96
.
56.
Eckert
LB
,
Repasky
GA
,
Ulku
AS
,
McFall
A
,
Zhou
H
,
Sartor
CI
, et al
Involvement of Ras activation in human breast cancer cell signaling, invasion, and anoikis
.
Cancer Res
2004
;
64
:
4585
92
.