Combinatorial Treatment with Mtor Inhibitors and Streptozotocin Leads

Published OnlineFirst October 19, 2017; DOI: 10.1158/1535-7163.MCT-17-0325

Small Molecule Therapeutics Molecular Cancer Therapeutics Combinatorial Treatment with mTOR Inhibitors and Streptozotocin Leads to Synergistic In Vitro and In Vivo Antitumor Effects in Insulinoma Cells Julien Bollard1,2,Celine Patte1,2, Patrick Massoma2, Isabelle Goddard2, Nicolas Gadot3, Noura Benslama2,Valerie Hervieu1,2,4,5, Carole Ferraro-Peyret2,4,5, Martine Cordier-Bussat2, Jean-Yves Scoazec6,7, Colette Roche1,2, Thomas Walter1,2,5,8, and Cecile Vercherat1,2

Abstract

Streptozotocin-based chemotherapy is the first-line chemo- mTORC1 and mTORC2). Effects on cell viability and apoptosis therapy recommended for advanced pancreatic neuroendocrine were assessed in insulinoma cell lines INS-1E (rat) and MIN6 tumors (pNETs), whereas targeted therapies, including mTOR (mouse) in vitro and were confirmed in vivo by using a mouse inhibitors, are available in second-line treatment. Unfortunate- model of hepatic tumor dissemination after intrasplenic xeno- ly, objective response rates to both treatments are limited. graft. In vitro, all four combinations display synergistic effects. Because mTOR pathway activation, commonly observed in These combinations lead to heterogeneous mTOR pathway pNETs, has been reported as one of the major mechanisms inhibition, in agreement with their respective target, and accounting for chemoresistance, we investigated the potential increased apoptosis. In vivo, tumor growth in the liver was benefit of mTOR inhibition combined with streptozotocin significantly inhibited by combining streptozotocin with ever- treatment in a subset of pNETs, namely insulinomas. To eval- olimus (P ¼ 0.0014), BKM120 (P ¼ 0.0092), or BEZ235 (P ¼ uate the potential of mTOR inhibition in combination with 0.008) as compared to each agent alone. These results suggest streptozotocin, we selected four different inhibitors acting at that targeting the mTOR pathway in combination with strep- various levels of the pathway (everolimus: inhibition of tozotocin could be of potential benefit for insulinomas and mTORC1; MK-2206: inhibition of AKT; BKM120: inhibition pNET patients and thus support further clinical investigations. of PI3K, mTORC1, and mTORC2; and BEZ235: inhibition of Mol Cancer Ther; 17(1); 60–72. 2017 AACR.

Introduction these rare and heterogeneous tumors display variable behavior in term of evolution and response to treatment. First-line treat- Therapeutic care of advanced pancreatic neuroendocrine ment of advanced pNETs is based on different prognostic factors tumors (pNETs) raises challenging clinical questions. Indeed, and two main strategies are currently described (1): (i) no chemotherapy with a watch-and-wait approach, or somatostatin analogues (SSA), for nonfunctional pNETs with low proliferative 1 Groupe des tumeurs neuroendocrines, Departement de Recherche Translation- tumors and stable disease at initial diagnosis; (ii) or a "top-down" 2 nelle et Innovation, Centre Leon Berard, Lyon, France. INSERM U1052/CNRS strategy with a first-line cytotoxic chemotherapy for more aggres- UMR5286/Universite de Lyon, Lyon1 UMR-S1052, Centre de Recherche en 3 sive pNETs. Streptozotocin (STZ)-based chemotherapy, either Cancerologie, Lyon, France. Plateforme Anatomopathologie-Recherche, fl Departement de Recherche Translationnelle et Innovation, Centre Leon Berard, with doxorubicin or 5- uorouracil (5-FU), remains the standard Lyon, France. 4Service Central d'Anatomie et Cytologie Pathologiques, Hospices first-line chemotherapy showing a 40% response rate (2, 3). The Civils de Lyon, Hopital^ Edouard Herriot, Lyon, France. 5Universite de Lyon, novel targeted therapies sunitinib (4) and everolimus (5) have Universite Claude Bernard Lyon 1, Villeurbanne cedex, France. 6Service de improved the progression-free survival in two phase III clinical pathologie morphologique et moleculaire, Departement de biologie et patho- trials, but the objective response rate did not exceed 10%. New logie medicales; AMMICa, Inserm US23/CNRS UMS3655, Gustave Roussy Cancer therapeutic strategies need thus to be developed in order to Campus, Villejuif, France. 7FacultedeM edecine de Bicetre,^ Universite improve the management of patients with unresectable pNETs. Paris Sud, Universite Paris Saclay, Le Kremlin-Bicetre,^ France. 8Service d'hepatogastroent erologie et d'oncologie digestive, Hospices Civils de Lyon, In this context, combination therapies appear as promising Hopital^ Edouard Herriot, Lyon, France. options. Recent studies have evaluated the potential of such combinatorial treatments in neuroendocrine tumors, including pNETs. Note: Supplementary data for this article are available at Molecular Cancer Therapeutics Online (http://mct.aacrjournals.org/). mTOR inhibitors (everolimus or temsirolimus) were combined with SSAs in RADIANT-2 study (6) or with the anti-VEGF antibody T. Walter and C. Vercherat are co-senior authors of this article. bevacizumab (7). However, to date, limited numbers of phase I/II Corresponding Author: Cecile Vercherat, Centre Leon Berard, Neuroendocrine studies have evaluated the combination of mTOR inhibitor fl Tumor Group, Cheney B 3rd oor, 28 rue Laennec, Lyon 69008, France. Phone: with chemotherapy (8). Nevertheless, given (i) the low number 334-6985-6133; Fax: 0033478782955 E-mail: [email protected] of patients, (ii) the problem of safety when combining two types doi: 10.1158/1535-7163.MCT-17-0325 of treatment with their respective toxicity, and (iii) the growing 2017 American Association for Cancer Research. number of different combinatorial or sequential strategies to be

60 Mol Cancer Ther; 17(1) January 2018

mTOR Inhibition Synergizes with Chemotherapy

evaluated, it is crucial to conduct studies with strong rationale and gift from B. Mark Evers, Galveston, TX, obtained in 2004) were to rely on in vitro and in vivo mechanistic studies, in order to better routinely culture in DMEM Glutamax/F-12K medium (1:1) sup- understand the benefit of combinations, evaluate their tolerance, plemented with 10% FBS and 1% penicillin/streptomycin. Rat and choose the best agents. radio-induced insulinoma, INS-1E cells (ref. 17; kind gift from For future studies, knowledge about mechanisms of resistance C. Wollheim and P. Maechler, Geneva, Switzerland, obtained should be taken in consideration when designing a new thera- in 2007) were routinely cultured in 5 mmol/L glucose RPMI peutic approach. On the one hand, chemotherapy has cytotoxic supplemented with 10% FBS, 10 mmol/L HEPES, 1 mmol/L effects which can be diminished by the activation of specific sodium pyruvate, 50 mmol/L b-mercaptoethanol, and 1% peni- survival and proliferation signaling pathways, leading to resis- cillin/streptomycin. MIN6 cells obtained from RIP-Tag mice tance. On the other hand, targeted therapies can inhibit these (kind gift from S. Dalle, Institut de Genomique Fonctionnelle, abnormally activated pathways; however, they do not necessarily Montpellier, France, obtained in 2009) were routinely cultured have a cytotoxic effect, leading to stabilization but low objective in 5 mmol/L glucose DMEM supplemented with 15% FBS, response. Based on these observations, we hypothesized that 50 mmol/L b-mercaptoethanol, and 1% penicillin/streptomycin. targeted therapies could prevent the activation of survival path- All cell lines were not authenticated and were obtained from ways and ultimately suppress chemoresistance. investigators who generated them. Upon reception, cells were Even if streptozotocin-based chemotherapy is the oldest cyto- amplified and early-passaged stocks were constituted. Cells were toxic chemotherapy used in pNETs, little is known about the passaged for fewer than three months after thawing. factors leading to resistance to streptozotocin in pNETs patients. Cellsweretestedforthepresenceofmycoplasma(MycoAlert O'Toole and colleagues showed that activation of the mTOR Mycoplasma Detection Kit; Lonza) on a regular basis (once a pathway was correlated with a low response to streptozotocin. month for in vitro experiments, and one day before injections in Indeed, in tumor tissues, high expression of AKT was correlated animals). with a lack of response to streptozotocin. On the opposite, conserved expression of PTEN (PI3K antagonist) was associated Reagents with response to streptozotocin (9). These observations are in Streptozotocin was purchased from Enzo Life Sciences. accordance with in vivo studies in which streptozotocin is mainly Everolimus was purchased from Selleck Chemicals. BKM120, used to induce diabetes in animal models. In pancreatic b cells, BEZ235, and MK-2206 were purchased from Active Biochem. specific deletion of Pten gene, as well as expression of constitu- For Western blot analysis, we used primary antibodies raised tively active AKT, protects animals from streptozotocin-induced against p-AKT(T308), p-AKT(S473), AKT, P70S6K, p-P70S6K diabetes and apoptosis (10, 11). It has been widely described that (T389), PRAS40, p-PRAS40(T246), 4EBP1, cleaved caspase-3, mTOR pathway activation has a central role in pNETs. At the p-HistoneH3(S10) (Cell Signaling Technology), p-4EBP1 genomic level, modifications of mTOR pathway coding genes (T45) (Abcam), and tubulin (Sigma). For IHC, we used pri- (PIK3CA, PTEN, TSC2) are involved in approximately 15% of mary antibodies directed to chromograninA (ImmunoStar) pNETs (12); moreover, histochemical analyses show that more and insulin (DAKO) p-P70S6K (T421/T424) (Santa Cruz Bio- than 85% of pNETs display altered expression levels of TSC2, technology Inc.). PTEN or both (13). Based on these data, it appears pertinent to evaluate the In vitro evaluation of the therapeutic potential of drug combination of streptozotocin with mTOR pathway inhibitors combinations in pNETs. The effect of combinatorial treatment was determined using the Given the growing number of mTOR inhibitors, we focused "Bliss independence test" (18). Based on cell viability results for on those that are the most clinically characterized, and that uncombined drugs, an expected response was calculated using the target the pathway at different levels: everolimus (mTORC1), formula: C ¼ (STZR þ IR) (STZR IR). C represents the expected MK-2206 (AKT), BKM120 (PI3K), and BEZ235 (PI3K and dual response, STZR is the difference in cell viability with streptozo- mTORC1 and mTORC2 inhibitor; refs. 14–16). We evaluated tocin treatment compared to untreated, and IR is the difference the effect of their combination with streptozotocin in preclin- in cell viability with mTOR inhibitors treatment compared to ical in vitro and in vivo models. As an in vivo tool, we took untreated. For each combination, the expected response was advantage of a xenograft mouse model developed in the lab compared to experimentally obtained response to calculate the which leads to hepatic dissemination mimicking hepatic metas- Delta Bliss. Sum of all Delta Bliss for a combination of two drugs tases observed in pNETs patients. All four combinations syn- corresponds to the Bliss Sum indicating antagonism (Bliss Sum < ergistically diminished cell viability and increased apoptosis 0), additivism (Bliss Sum ¼ 0), or synergism (Bliss Sum > 0). in vitro. Moreover, all combinations but streptozotocin/MK- 2206, led to significant antitumor responses in xenografted In vivo studies mice. Our data suggest that combining mTOR inhibitors with Four-week-old female athymic nude mice obtained from streptozotocin could be a promising therapeutic option for Envigo (Gannat, France) were housed and bred in the patho- advanced pNETs patients. gen-free animal facility "AniCan." Animals were anesthetized (isoflurane) during all surgical procedures. After surgery, animals Materials and Methods were allowed to recover in a sterile atmosphere and were fed ad libitum with a sterile diet. Cell lines Xenografting procedure was carried out as previously described QGP-1 cells (Japan Health Sciences Foundation, obtained in (19). Briefly, 2.5 million INS-1E cells were injected into the spleen, 2006) were routinely cultured in RPMI Glutamax supplemented from where they disseminated into the liver through the portal with 10% FBS and 1% penicillin/streptomycin. BON cells (kind vein to form intrahepatic tumor nodules. To avoid hypoglycemia

www.aacrjournals.org Mol Cancer Ther; 17(1) January 2018 61

Bollard et al.

due to insulin secretion from INS-1E cells, drinking water was streptozotocin combination with mTOR inhibitors. We evalu- supplemented with 10% sucrose. Seven days after cell injection, ated the mRNA expression of the GLUT2 transporter (SLC2A2 animals were randomized into treatment and control groups. gene), which is essential for streptozotocin incorporation Streptozotocin and everolimus were injected intraperitoneally at within the cell (20, 21). We analyzed the expression of SLC2A2 10 and 1.5 mg/kg/day, respectively. MK-2206, BKM120, and mRNA in the human pNETs cells lines QGP-1 and BON, the BEZ235 were orally administrated at 120, 20, and 45 mg/kg/day, rat cell line INS-1E, and the mouse cell line MIN6. SLC2A2 respectively. Control mice were treated with vehicle only. Animals mRNA was detected in both murine cell lines, whereas its were weighted daily before treatment and glycemia was measured expression was faint in BON cells and absent in QGP-1 cells three times a week. All animals were sacrificed after 3 weeks of (Fig. 1A). Primers specificity was confirmed on a control mRNA treatment. In each animal, liver and pancreas were excised and extracted from human pancreatic neuroendocrine tissue. To prepared for histological analysis. Each group of treatment was correlate SLC2A2 expression and streptozotocin sensitivity, we composed of at least five animals and two independent experi- analyzed the level of apoptosis in these four cell lines by ments were performed. Western blot analysis, because streptozotocin is a known apoptosis inducer (22). Caspase-3 cleavage was observed in INS-1E Statistical analysis cells (3 hours after treatment) and MIN6 cells (6 hours after All data were mean values of at least three individual experi- treatment) but not in QGP-1 and BON cell lines (Fig. 1B). ments and were expressed as mean SEM. Statistical analyses Regarding these results, we decided to select the INS-1E cell were done using GraphPad Prism 6.05 software (test Mann– line for further experiments because its genetic background Whitney, two-tailed; two-way ANOVA, Tukey multiple compar- closely recapitulates what has been observed in patients. isons test). Indeed, we ran a whole genome sequencing (WGS) analysis The rest of materials and methods can be found in Supple- on this cell line and mutations characteristic of pNET (MEN1, mentary Data. DAXX, ATRX, mTOR, PTEN, YY1, TSC2, PI3KCA) were interro- gated. Results showed (i) a deleterious frameshift mutation resulting in a premature stop codon (high pathogenicity) for Results ATRX gene, (ii) an inframe deletion in DAXX (moderate path- Sensitivity to streptozotocin is correlated to SLC2A2 ogenicity), (iii) a missense mutation in DAXX (moderate (GLUT2) expression pathogenicity). No variants for YY1 or low impact variants Given the poor availability of pNETs cell lines and knowing (synonymous variants, variants within introns) were found that pNETs consist in a heterogeneous group of tumor, we for MEN1, mTOR, PTEN, YY1, TSC2,norPI3KCA genes (Sup- wanted to choose the most appropriate cell line to study plementary Table S1). This mutational landscape is in

Figure 1. Correlation between GLUT2 mRNA expression (Slc2A2 gene) and streptozotocin (STZ)-induced apoptosis. A, Slc2A2 gene expression was measured by RT-qPCR. Expression data were normalized to the expression of housekeeping gene (b-Actin). Experiments were done in duplicate and repeated three times. B, Caspase-3 activation following streptozotocin treatment is visualized by Western blot analysis in INS-1E, MIN6, QGP1, and BON cells. a-Tubulin was used as loading control. Blots are representative of at least three independent experiments.

62 Mol Cancer Ther; 17(1) January 2018 Molecular Cancer Therapeutics

mTOR Inhibition Synergizes with Chemotherapy

accordance with what has been published so far for insulinoma effectors AKT, P70S6K, and 4EBP1 at different time points, by and pNETs (12, 23). The MIN6 cell line was used to confirm Western blot analysis. In INS-1E cells, we noticed a quick in vitro results. activation of the mTOR pathway from 10 minutes to 2 hours after incubation with streptozotocin. We detected an increased A rapid and transient activation of mTOR pathway follows phosphorylation of AKT, first on the S473 residue and after streptozotocin treatment in vitro 1 hour on T308 residue (Fig. 2B). This activation of AKT was To investigate whether combining mTOR inhibitors with also correlated with the phosphorylation of p70S6K at T389 streptozotocin could be of any benefit for pNETs treatment, residue mainly after 1 hour, probably through the activation we first analyzed the effect of streptozotocin on the mTOR of mTORC1. However, there was no evidence of activated pathway in the two selected cell lines. After treatment of INS-1E 4EBP1, another target of mTORC1 (Fig. 2B). Results for MIN6 and MIN6 cell lines, the mTOR pathway activation was mon- cells show a similar activation after 2 hour of treatment, itored by evaluating the phosphorylation level of its major although more modest (Fig. 2C).

Figure 2. Effect of streptozotocin (STZ) treatment on the mTOR pathway. A, Inhibitors used in this study target the mTOR pathway at different levels. B and C, mTOR pathway and caspase-3 activation during streptozotocin treatment is visualized by Western-blot. INS-1E cells (B) and MIN6 cells (C) were treated with streptozotocin (2 mmol/L).

www.aacrjournals.org Mol Cancer Ther; 17(1) January 2018 63

Bollard et al.

Figure 3. Beneﬁt of mTOR pathway inhibition during streptozotocin (STZ) treatment in INS-1E cells (A–D) and MIN6 cells (D–H). Cell viability after treatment with drugs combination was assessed with CellTiterGlo assay. Results represent the percentage of growth inhibition compared to untreated cells; they are mean values of at least three independent experiments (ﬁve duplicates/experiments). Cells were treated with increasing doses of streptozotocin combined with increasing doses of mTOR pathway inhibitors: everolimus (A–E), MK-2206 (B–F), BKM120 (C–G), and BEZ235 (D–H). I, Bliss sum scores for combinations in INS-1E and MIN6 cells were calculated from delta bliss values (Supplementary Figs. S2 and S3).

64 Mol Cancer Ther; 17(1) January 2018 Molecular Cancer Therapeutics

mTOR Inhibition Synergizes with Chemotherapy

This activation was only transient and followed by an inhi- correlated with an increased apoptosis, as indicated by the bition of the whole pathway. The mTOR pathway was activated level of cleaved caspase-3. in untreated cells 4 hours after changing the media whereas, in parallel, it was abrogated in streptozotocin-treated cells (inhi- Combinations of mTOR inhibitors and streptozotocin have bition of phosphorylation of AKT, p-P70 and 4EBP1). Inter- synergistic effects in vitro estingly, we also noticed that this inhibition of the mTOR Based on data from the literature showing the involvement pathway after prolonged treatment with streptozotocin was ofthemTORpathwayintheresistancetostreptozotocin

Figure 4. Combinations of streptozotocin (STZ) and mTOR inhibitors affect mTOR pathway activation and increase apoptosis in INS-1E cells. Activation of mTOR pathway was assessed 2 hours (A)and4hours(B) after treatment with streptozotocin (2 mmol/L), everolimus (10 nmol/L), MK-2206 (1 mmol/L), BKM120 (5 mmol/L), and BEZ235 (100 nmol/L) by Western blot analysis. C, Apoptosis activation was evaluated after 4 hours of treatment by analysis of caspase-3 cleavage by Western blot analysis. www.aacrjournals.org Mol Cancer Ther; 17(1) January 2018 65

Bollard et al.

66 Mol Cancer Ther; 17(1) January 2018 Molecular Cancer Therapeutics

mTOR Inhibition Synergizes with Chemotherapy

and on our above data showing the streptozotocin-induced (on both T308 and S473 residues), of p70S6K and of 4E-BP1. activation of the mTOR pathway, we suggest that combining At this time point, combining mTOR inhibitors and strepto- mTOR inhibitors with streptozotocin could potentiate strep- zotocin prevented streptozotocin-induced activation of the tozotocin cytotoxicity ability. To evaluate the effect of such mTOR pathway. combinations, we performed a CellTiter-Glo viability test, After 4 hours of treatment, the mTOR pathway was fully followed by a Bliss independence test (18). Viability tests inhibited (absence of phosphorylation of AKT, PRAS40, showed that streptozotocin decreased the viability of both p70S6K, and 4E-BP1) by the combination of streptozotocin INS-1E and MIN6 cells 24 hours after treatment (Supplemen- and MK-2206, BKM120, or BEZ235. However we still observed tary Fig. S1A). Four different inhibitors were used to target the phosphorylated AKT (S473) and PRAS40 (T246) with ever- mTOR pathway: everolimus (mTORC1 inhibitor), MK-2206 olimus while other inhibitors decreased the phosphorylation (Pan-AKT inhibitor), BKM120 (Pan-PI3K inhibitor), and levels of all components of the mTOR pathway (Fig. 4B for BEZ235 (dual mTORC1, mTORC2, and PI3K inhibitor; INS-1E and Supplementary Fig. S4B for MIN6). Fig. 2A). Treatments with these molecules had different effects on INS-1E and MIN6 cell viability (Supplementary Fig. S1B– Combinations enhance apoptosis in vitro S1E). We proceeded to all the combinations between strepto- Knowing that streptozotocin is an apoptotic inducer, we zotocin and these inhibitors at all doses and measured the analyzed the cleavage of caspase-3. Four hours after treatment, reduction in cell viability of INS-1E and MIN6 cells (Fig. 3A–D everolimus and BEZ235 do not trigger apoptosis. Streptozoto- and Fig. 3E–H, respectively). cin, MK-2206, and BKM120 used as single agents led to a From the viability data, we calculated a delta Bliss score (Sup- cleavage of caspase-3. When combined with streptozotocin, plementary Figs. S2A–S2D and S3A–S3D). By adding all delta MK-2206, BKM120, and BEZ235 increased the cleavage of Bliss values, we obtained the Bliss sum to evaluate whether these caspase-3 (Fig. 4C for INS-1E and Supplementary Fig. S4C for combinations had antagonist (<0), additive (¼0), or synergic MIN6). (>0) effects. For both INS-1E and MIN6 cell lines, all combina- – tions were found to be synergistic (Fig. 3A I). For both cell lines, Effect of combinations on tumor development in a xenograft the strongest synergy was obtained by combining streptozotocin model with BEZ235 (bliss sum 171 for INS-1E and 247 for MIN6) To evaluate the effect of these combinations in vivo, we used a whereas other combinations had different effects on either cell mouse model of intrahepatic tumor dissemination after intras- line. For INS-1E, streptozotocin/everolimus and streptozotocin/ plenic xenograft of INS-1E cells. Mice were treated with strepto- MK-2206 Bliss sums were superior to that of streptozotocin/ zotocin, mTOR inhibitors or a combination. To assess the treat- BKM120. On the opposite, streptozotocin/everolimus synergy ment efficacy, we quantified the intrahepatic tumor surface by was mild for MIN6 cells, in contrast to the greater synergy staining for the endocrine marker chromogranin A (Fig. 5A–E). observed for streptozotocin/MK-2206 and streptozotocin/ Everolimus, MK-2206, BKM120 but not BEZ235 induced a non- BKM120 combinations. significant decrease in tumor surface. These decreases were the result of a reduction in the mean size of tumor nodules for mTOR inhibitors and combinations have various effects on everolimus and MK-2206 (Supplementary Fig. S5) and of their mTOR pathway in vitro number for BKM120 (Supplementary Fig. S6). To better evaluate the effects of these combinations, we When used in combination with streptozotocin, everolimus analyzed the mTOR pathway activation during treatments by (P ¼ 0.0014), BKM120 (P ¼ 0.0092), and BEZ235 (P ¼ 0.008), Western blot analysis. As expected, the degree of mTOR inhi- but not MK-2206 led to a significant decrease of tumor surface bition after 2 hours of treatment was variable according to the compared to control mice (Fig. 5B–E). The effects of combina- inhibitor used, whereas streptozotocin activates mTOR path- tions containing everolimus were also significantly higher than way (Fig. 4A for INS-1E and Supplementary Fig. S4A for MIN6). with everolimus or streptozotocin alone (Fig. 5B). The size of A full inhibition was obtained with BKM120 and BEZ235, both tumor nodules was significantly lower after treatment with alone and in combination with streptozotocin. This was con- streptozotocin/everolimus (P < 0.0001) and streptozotocin/ firmed by a decreased phosphorylation of AKT (T308 and BKM120 (P ¼ 0.0101; Supplementary Fig. S5A and S5C). S473) and its target PRAS40 (T246), but also by a lowered Streptozotocin/BKM120 also decreased the number of tumor phosphorylation of the two mTORC1 effectors: p70S6K (T389) nodules as compared to controls (P ¼ 0.0235; Supplementary and 4E-BP1 (T45). As expected, everolimus (mTORC1 inhib- Fig. S6C). Streptozotocin/everolimus and streptozotocin/ itor) inhibited p70S6K and 4E-BP1, but did not affect AKT and BKM120 decreased the number of tumor nodules as compared PRAS40 phosphorylation. Treatment with everolimus or ever- with streptozotocin used as single-agent (P ¼ 0.0498 and olimus/streptozotocin even activated AKT on T308 and S473 0.0072, respectively; Supplementary Fig. S6A and S6C). Inter- residues. On the opposite, MK-2206, alone or combined with estingly, BEZ235 alone did not exert any antitumor effect, streptozotocin, exerts an inhibition of AKT phosphorylation whereas it led to a significant decrease in both the tumor

Figure 5. Hepatic tumor nodule development and/or growth in xenografted mice is variously affected by treatment with streptozotocin (STZ) alone or combined to mTOR inhibitors compared to controls. While controls received the vehicle (A), treated animals received streptozotocin (10 mg/kg/day) alone (B)or combined to everolimus (1.5 mg/kg/day; B), MK-2206 (120 mg/kg/day; C), BKM120 (20 mg/kg/day; D), and BEZ235 (45 mg/kg/day; E)for3weeksTo visualize tumor nodules, liver sections were stained for Chromogranin A (CgA) by IHC (100). Scale bar: 100 mm. Panels are representative of results obtained in at least three different experiments which included at least three animals for each condition.

www.aacrjournals.org Mol Cancer Ther; 17(1) January 2018 67

Bollard et al.

68 Mol Cancer Ther; 17(1) January 2018 Molecular Cancer Therapeutics

mTOR Inhibition Synergizes with Chemotherapy

surface (P ¼ 0.0492, compared to BEZ235 alone; Fig. 5E) and factors that could serve as potential predictive biomarkers for the number of tumor nodules (P ¼ 0.0253, as compared to response to therapy in pNETs (28, 29), meaning that besides BEZ235 alone; Supplementary Fig. S6D) when combined to searching for new targets and new molecules to improve efficacy streptozotocin. of medical care, we need to understand tumor progression as well as resistance mechanisms. Even if combinatorial strategies seem Glycemia is differently affected by mTOR inhibitors very attractive, strong rationales and preclinical studies are needed The assessment of adverse effects of the different treatments to test relevant combinations. on animals was performed by monitoring body weight. No The mTOR pathway is frequently activated in pNETs (12, 13) difference was detected within the different groups of treatment and it is widely known as a key player in chemoresistance. (Fig. 6A–D). Because INS-1E is an insulin-secreting cell line Indeed, the mTOR pathway promotes cell growth, prolifera- and as different clinical trials reported that mTOR inhibitors tion,andsurvival(30).Furthermore,PTENexpressionandAKT could lead to hyperglycemia (24–27), the effect of streptozo- activity are associated with chemoresistance in various cancers tocin and mTOR inhibitors on insulin secretion by INS-1E (31–34) including gastroenteropancreatic NETs (9). Many AKT- cells was checked. To do so, we measured the glycemia of driven mechanisms of chemoresistance are described, including animals every 2 to 3 days during the experiment and noticed NF-kB activation (31), inhibition of apoptosis through the significant differences. Control animals and everolimus- and MDM2-p53 axis (35), inhibition of caspase-mediated apopto- streptozotocin-treated animals showed a significant decreased sis (36). In this study, we show that streptozotocin treatment glycemia between day 7 and day 21 to 25 post-graft. While at of insulinoma cell lines leads to rapid and transient activation day 25, everolimus alone or combined did not significantly of the mTOR pathway. This paradoxical activation of a survival modify glycemia compared to control animals (Fig. 6E). MK- pathway following cytotoxic treatment has been described 2206/streptozotocin showed significantly higher glycemia than in other cell lines (34, 37, 38). Hence, in some human breast streptozotocin (P ¼ 0.0104) or MK-2206 alone (P ¼ 0.0260) cancer cell lines, doxorubicin leads to a rapid and robust treatedanimals(Fig.6F).BKM120 alone and BKM120/strep- activation of AKT in a PI3K-dependent manner, contributing tozotocin also led to higher glycemia compared with strepto- to cellular resistance to chemotherapy (38). Similar observa- zotocin-treated animals (P ¼ 0.0196 and 0.0021, respectively) tions were made in gastric cancer cells in response to etoposide as well as BKM120/streptozotocin compared with control ani- and doxorubicin. Moreover, pretreatment of cells with PI3K mals (P ¼ 0.0274; Fig. 6G). BEZ235/streptozotocin-treated inhibitors, prior to chemotherapy, enhanced etoposide and animals also presented higher glycemia compared to control doxorubicin apoptotic effects (34). Caporali and colleagues (P ¼ 0.0063) or streptozotocin-treated animals (P ¼ also showed that such an activation of AKT is observed after 0.0075; Fig. 6H). temozolomide treatment in lymphoblastoid TK6 cell line as Moreover, streptozotocin is commonly used as a diabetes well as in the human colorectal HCT116/3-6 cell line through inducer in vivo due to its ability to specifically provoke b-cell an ATR-dependent mechanism (37). These observations sup- death. In this study, streptozotocin was used at very low port the idea of combining mTOR inhibitors with chemother- doses compared to protocols for diabetes induction (10 vs. apy. Because streptozotocin is the first-line treatment for 40–80 mg/kg/day). Nevertheless, we checked the absence of advanced pNETs including insulinoma and given that the deleterious effect on b cells by staining pancreatic sections response to streptozotocin is correlated with mTOR pathway for insulin and by assessing b-cell mass (Supplementary activation, we decided to combine streptozotocin with mTOR Fig. S7). No significant difference was observed in b-cell mass inhibitors. of streptozotocin-treated mice compared to controls. Lately, we observed a growing interest for the development of mTOR inhibitors. Although first-generation inhibitors target mTORC1 complex, the ability of mTORC2 to activate AKT Discussion encouraged the development of compounds targeting both Advanced pNETs represent a therapeutic challenge; indeed, mTORC1 and mTORC2, and more globally of dual-inhibitors even if we can control tumor progression with cytotoxic or also targeting PI3K (39). To understand the specificity of such targeted therapies, objective responses are rare and resistance will inhibitors, we designed a study allowing the comparison of eventually occur. There is thus a constant and unmet need for new first-generation inhibitor, everolimus, currently used for pNETs therapeutic options, especially for patients that failed first- and treatment, to second-generation inhibitors targeting specifically second-line therapies. Up-to-date, only few studies identified AKT (MK-2206), PI3K (Pan-class I PI3K inhibitor: BKM120), or

Figure 6. Combination of streptozotocin (STZ) with mTOR inhibitors has different effects on the hepatic tumor surface of xenografted mice and does not lead to hyperglycemia. Animals were treated with streptozotocin (10 mg/kg/day; 5/7 days), everolimus (1.5 mg/kg/day), MK-2206 (120 mg/kg/day), BKM120 (20 mg/kg/day), and BEZ235 (45 mg/kg/day), alone or in combination, during 3 weeks. A–D, Total tumor surface was measured and compared to total hepatic surface after treatment with A. streptozotocin everolimus (CTL vs. streptozotocin þ everolimus: P ¼ 0.0014; everolimus vs. streptozotocin þ everolimus: P ¼ 0.0269; streptozotocin vs. streptozotocin þ everolimus: P ¼ 0.0088). B, Streptozotocin MK-2206. C, Streptozotocin BKM120 (CTL vs. streptozotocin þ BKM120: P ¼ 0.0092; streptozotocin vs. streptozotocin þ BKM120: P ¼ 0.0292). D, streptozotocin BEZ235 (streptozotocin vs. streptozotocin þ BEZ235: P ¼ 0.008; BEZ235 vs. streptozotocin þ BEZ235: P ¼ 0.0492). E–H, Glycemia was controlled three times per week for the whole treatment duration with E. Streptozotocin everolimus. F, Streptozotocin MK-2206 (streptozotocin vs. streptozotocin þ MK-2206: P ¼ 0.0104; MK-2206 vs. streptozotocin þ MK-2206: P ¼ 0.0260). G, Streptozotocin BKM120 (CTL vs. streptozotocin þ BKM120: P ¼ 0.0274; streptozotocin vs. BKM: P ¼ 0.0196 and streptozotocin vs. streptozotocin þ BKM120: P ¼ 0.0021). H, Streptozotocin BEZ235 (CTL vs. streptozotocin þ BEZ235: P ¼ 0.0063; streptozotocin vs. streptozotocin þ BEZ235: P ¼ 0.0075; n ¼ at least nine for each group of animal; two-way ANOVA).

www.aacrjournals.org Mol Cancer Ther; 17(1) January 2018 69

Bollard et al.

both mTORC1/2 and PI3K (BEZ235; Fig. 2A). Logically, inhi- QGP-1 cells, we noticed that they were not sensitive to strep- bition of mTOR pathway in vitro is variable. As expected, tozotocin combined or not to mTOR inhibitors. Because we everolimus inhibits mTORC1 but not AKT which is activated failed to detect SLC2A2, the glucose transporter responsible for in a robust manner, probably through loss of the negative streptozotocin entrance in the cell, in QGP-1 cells, we hypoth- feedback loop exerted by mTORC1 on AKT (40, 41). However, esized that streptozotocin was not able to penetrate into these although everolimus seems to have mild effects in vitro (mod- cells, leading to intrinsic resistance to streptozotocin-induced erate synergy with streptozotocin, no or mild apoptosis when cell death. Given the promising results of BEZ235 in combi- used as single-agent or combined with streptozotocin), it is very nation with streptozotocin in in vitro and in vivo preclinical efficient in our preclinical xenograft model; probably due to studies, it could be interesting to test it in combination with antitumor direct effects as well as on the microenvironment streptozotocin in phase I/II clinical trials, with reduced doses of and angiogenesis regulation. These results are encouraging; BEZ235 in order to get an acceptable toxicity profile for however, sustained activation of AKT suggests that synergy patients. could be even stronger by suppressing this AKT pro-survival Finally, inhibition of PI3K with BKM120 is very efficient signaling. Indeed, on the one hand, in vitro assays show very when used as single agent or combined with streptozotocin. strong effects of AKT inhibitor MK-2206 (global inhibition of BKM120, a pan-PI3K inhibitor, was shown to display a strong the mTOR pathway and strong synergy with streptozotocin on antitumor effect in numerous models including human glio- cell viability and apoptosis). On the other hand, in vivo results blastoma cell lines (16). Several clinical trials reported its are rather disappointing with a nonsignificant decrease of therapeutic potential mainly when combined to other com- tumor surface when used as single-agent but no additional pounds (26). In our experiments, BKM120 was the most benefit when MK-2206 is combined to streptozotocin. MK- efficient inhibitor and led to the optimal antitumor response. 2206 is currently being evaluated in phase II clinical trials and In conclusion, combination of streptozotocin with mTOR gave disappointing results (only 6 of 33 stable diseases) in a inhibitors is a very attractive option to treat pNETs. However, phase I study in patients with solid tumors. However, among further clinical evaluation is necessary to confirm these results the three pNETs patients included in this study, two of them andtoevaluatetheefficiency and safety of such combinations presented a stabilization of the disease (42). Nevertheless, in pNETs patients. The SEQTOR trial is currently recruiting another study focusing on pNETs reported limited response patients to evaluate the benefit of such combinations. More to MK-2206 alone (43). In preclinical models of ovarian and precisely it aims at determining what sequence (everolimus lung cancers, the combination of the AKT inhibitor with the followed by streptozotocin/50-FU vs. streptozotocin/50-FU fol- EGFR inhibitor erlonitib or the dual EGFR/HER2 inhibitor lowed by everolimus) is the most beneficial. Our results strong- lapatinib showed antitumor efficacy (14). However, our study ly suggest that shutting down the mTOR pathway during suggests that a better in vivo characterization of the molecule streptozotocin treatment leads to more potent antitumor efficacy might be necessary. It could be interesting to further effects, supporting thus the rational for a combinatorial treat- study differences in the whole pathway inhibition to under- ment or a pretreatment with everolimus prior to streptozotocin. stand the inability of MK-2206 to synergize with streptozotocin In this setting potential toxicity needs to be considered when in vivo. Knowing that MK-2206 is a specificinhibitorofAKT combining or sequencing therapies. Everolimus and streptozo- (Fig. 2A), we can hypothesize that inhibition of mTORC1 tocin have different toxicity profiles. Main toxicities of mTOR and/or mTORC2 is necessary to reach synergy with streptozo- inhibitors are skin rash, hyperglycemia, and mucositis (26, 50), tocin in vivo. whereas the latest studies using streptozotocin according to To inhibit the whole mTOR pathway, we also evaluated the current guidelines, showed that renal toxicity of STZ is much effect of BEZ235. In vitro inhibition of the pathway seems very lower than initially reported (3). Preclinical studies may help to efficient and global. However, when used as single-agent in vivo, determine the best agents to be combined and to optimize BEZ235 does not exert any antitumor effect. This puzzling doses of each drug in combination. Finally, optimal trial result is in accordance with two clinical trials including pNETs designs will be needed to determine the best therapeutic patients. The first trial was conducted with patients presenting protocol i.e., combination or sequence. resistance to everolimus and the second with patient na€ve for mTOR inhibition. Unfortunately, both trials were prematurely Disclosure of Potential Conflicts of Interest stopped due to unmet statistical endpoint (44). Our data are in No potential conflicts of interest were disclosed. accordance with these results as BEZ235 monotherapy does not show any antitumor effect in vivo.However,weobservedthat the combination of BEZ235 with streptozotocin leads to a Authors' Contributions significant antitumor effect in this xenograft model. The efficacy Conception and design: J.-Y. Scoazec, C. Roche, T. Walter, C. Vercherat of BEZ235 was demonstrated both in vitro and in vivo in Development of methodology: T. Walter, C. Vercherat Acquisition of data (provided animals, acquired and managed patients, hepatocellular carcinoma (45) and in the human NET cell lines provided facilities, etc.): J. Bollard, C. Patte, P. Massoma, I. Goddard, BON and QGP-1 (46). Both cell lines were not used in our N. Gadot, C. Roche, T. Walter, C. Vercherat study as BON cell line is very heterogeneous and subclonal Analysis and interpretation of data (e.g., statistical analysis, biostatistics, modifications could lead to the loss of neuroendocrine tumors computational analysis): J. Bollard, P. Massoma, T. Walter, C. Vercherat characteristics (47–49). Therefore, the use of this cell line as a Writing, review, and/or revision of the manuscript: J. Bollard, P. Massoma, model of pNETs is controversial; moreover, we observed that C. Ferraro-Peyret, M. Cordier-Bussat, J.-Y. Scoazec, C. Roche, T. Walter, C. Vercherat Administrative, technical, or material support (i.e., reporting or organizing BON cells lead to Mixed Adenoneuroendocrine Carcinoma data, constructing databases): C. Patte, P. Massoma, I. Goddard, N. Benslama, (MANEC) tumors rather than pure pNETs in our model of V. Hervieu, J.-Y. Scoazec, C. Vercherat xenograft (unpublished personal observations). Regarding Study supervision: J.-Y. Scoazec, T. Walter, C. Vercherat

70 Mol Cancer Ther; 17(1) January 2018 Molecular Cancer Therapeutics

mTOR Inhibition Synergizes with Chemotherapy

Acknowledgments The costs of publication of this article were defrayed in part by the payment advertisement The authors thank M. Colombe and F. Bourguillaut for technical assis- of page charges. This article must therefore be hereby marked tance; A. Chappot de la Chanonie, E. Servoz, and V. Martin (Laboratoire des in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Modeles Tumoraux) for support with in vivo experiments; B. Mark Evers for BON cell line; C. Wollheim and P. Maechler for INS-1E cell line; and S. Dalle Received April 13, 2017; revised September 29, 2017; accepted October 12, for MIN6 cell line. 2017; published OnlineFirst October 19, 2017.

References 1. Pavel M, O'Toole D, Costa F, Capdevila J, Gross D, Kianmanesh R, et al. 17. Merglen A, Theander S, Rubi B, Chaffard G, Wollheim CB, Maechler P. ENETS consensus guidelines update for the management of distant Glucose sensitivity and metabolism-secretion coupling studied during metastatic disease of intestinal, pancreatic, bronchial neuroendocrine two-year continuous culture in INS-1E insulinoma cells. Endocrinology neoplasms (NEN) and NEN of unknown primary site. Neuroendocri- 2004;145:667–78. nology 2016;103:172–85. 18. Wallin JJ, Guan J, Prior WW, Lee LB, Berry L, Belmont LD, et al. GDC-0941, 2. Fjallskog ML, Janson ET, Falkmer UG, Vatn MH, Oberg KE, Eriksson BK. a novel class I selective PI3K inhibitor, enhances the efficacy of docetaxel in Treatment with combined streptozotocin and liposomal doxorubicin human breast cancer models by increasing cell death in vitro and in vivo. in metastatic endocrine pancreatic tumors. Neuroendocrinology 2008; Clin Cancer Res 2012;18:3901–11. 88:53–8. 19. Pourreyron C, Poncet G, Roche C, Gouysse G, Nejjari M, Walter T, et al. The 3. Dilz LM, Denecke T, Steffen IG, Prasad V, von Weikersthal LF, Pape UF, et al. role of angiogenesis in endocrine liver metastases: an experimental study. Streptozocin/5-fluorouracil chemotherapy is associated with durable J Surg Res 2008;144:64–73. response in patients with advanced pancreatic neuroendocrine tumours. 20. Schnedl WJ, Ferber S, Johnson JH, Newgard CB. STZ transport and Eur J Cancer 2015;51:1253–62. cytotoxicity. Specific enhancement in GLUT2-expressing cells. Diabetes 4. Raymond E, Dahan L, Raoul JL, Bang YJ, Borbath I, Lombard-Bohas C, et al. 1994;43:1326–33. Sunitinib malate for the treatment of pancreatic neuroendocrine tumors. 21. Thulesen J, Orskov C, Holst JJ, Poulsen SS. Short-term insulin treatment N Engl J Med 2011;364:501–13. prevents the diabetogenic action of streptozotocin in rats. Endocrinology 5. Yao JC, Shah MH, Ito T, Bohas CL, Wolin EM, Van Cutsem E, et al. 1997;138:62–8. Everolimus for advanced pancreatic neuroendocrine tumors. N Engl J Med 22. Saini KS, Thompson C, Winterford CM, Walker NI, Cameron DP. Strepto- 2011;364:514–23. zotocin at low doses induces apoptosis and at high doses causes necrosis in 6. Pavel ME, Hainsworth JD, Baudin E, Peeters M, Horsch D, Winkler RE, et al. a murine pancreatic beta cell line, INS-1. Biochem Mol Biol Int 1996; Everolimus plus octreotide long-acting repeatable for the treatment of 39:1229–36. advanced neuroendocrine tumours associated with carcinoid syndrome 23. Scarpa A, Chang DK, Nones K, Corbo V, Patch AM, Bailey P, et al. Whole- (RADIANT-2): a randomised, placebo-controlled, phase 3 study. Lancet genome landscape of pancreatic neuroendocrine tumours. Nature 2017; 2011;378:2005–12. 543:65–71. 7. Hobday TJ, Qin R, Reidy-Lagunes D, Moore MJ, Strosberg J, Kaubisch A, 24. Yao JC, Fazio N, Singh S, Buzzoni R, Carnaghi C, Wolin E, et al. Everolimus et al. Multicenter phase II trial of temsirolimus and bevacizumab in for the treatment of advanced, non-functional neuroendocrine tumours of pancreatic neuroendocrine tumors. J Clin Oncol 2015;33:1551–6. the lung or gastrointestinal tract (RADIANT-4): a randomised, placebo- 8. Chan JA, Blaszkowsky L, Stuart K, Zhu AX, Allen J, Wadlow R, et al. A controlled, phase 3 study. Lancet 2016;387:968–77. prospective, phase 1/2 study of everolimus and temozolomide in 25. Ramanathan RK, McDonough SL, Kennecke HF, Iqbal S, Baranda JC, patients with advanced pancreatic neuroendocrine tumor. Cancer 2013; Seery TE, et al. Phase 2 study of MK-2206, an allosteric inhibitor of 119:3212–8. AKT, as second-line therapy for advanced gastric and gastroesophageal 9. O'Toole D, Couvelard A, Rebours V, Zappa M, Hentic O, Hammel P, et al. junction cancer: a SWOG cooperative group trial (S1005). Cancer Molecular markers associated with response to chemotherapy in gastro- 2015;121:2193–7. entero-pancreatic neuroendocrine tumors. Endocr Relat Cancer 2010;17: 26. Bendell JC, Rodon J, Burris HA, de Jonge M, Verweij J, Birle D, et al. 847–56. Phase I, dose-escalation study of BKM120, an oral pan-Class I PI3K 10. Stiles BL, Kuralwalla-Martinez C, Guo W, Gregorian C, Wang Y, Tian J, inhibitor, in patients with advanced solid tumors. J Clin Oncol 2012; et al. Selective deletion of Pten in pancreatic beta cells leads to increased 30:282–90. islet mass and resistance to STZ-induceddiabetes.MolCellBiol2006; 27. Bendell JC, Kurkjian C, Infante JR, Bauer TM, Burris HA III, Greco FA, et al. A 26:2772–81. phase 1 study of the sachet formulation of the oral dual PI3K/mTOR 11. Tuttle RL, Gill NS, Pugh W, Lee JP, Koeberlein B, Furth EE, et al. Regulation inhibitor BEZ235 given twice daily (BID) in patients with advanced solid of pancreatic beta-cell growth and survival by the serine/threonine protein tumors. Invest New Drugs 2015;33:463–71. kinase Akt1/PKBalpha. Nat Med 2001;7:1133–7. 28. Kulke MH, Hornick JL, Frauenhoffer C, Hooshmand S, Ryan DP, Enzinger 12. Jiao Y, Shi C, Edil BH, de Wilde RF, Klimstra DS, Maitra A, et al. DAXX/ PC, et al. O6-methylguanine DNA methyltransferase deficiency and ATRX, MEN1, and mTOR pathway genes are frequently altered in pancre- response to temozolomide-based therapy in patients with neuroendocrine atic neuroendocrine tumors. Science 2011;331:1199–203. tumors. Clin Cancer Res 2009;15:338–45. 13. Missiaglia E, Dalai I, Barbi S, Beghelli S, Falconi M, della Peruta M, et al. 29. Walter T, van Brakel B, Vercherat C, Hervieu V, Forestier J, Chayvialle JA, Pancreatic endocrine tumors: expression profiling evidences a role for et al. O6-Methylguanine-DNA methyltransferase status in neuroendocrine AKT-mTOR pathway. J Clin Oncol 2010;28:245–55. tumours: Prognostic relevance and association with response to alkylating 14. Hirai H, Sootome H, Nakatsuru Y, Miyama K, Taguchi S, Tsujioka K, et al. agents. Br J Cancer 2015;112:523–31. MK-2206, an allosteric Akt inhibitor, enhances antitumor efficacy by 30. Xu K, Liu P, Wei W. mTOR signaling in tumorigenesis. Biochim Biophys standard chemotherapeutic agents or molecular targeted drugs in vitro Acta 2014;1846:638–54. and in vivo. Mol Cancer Ther 2010;9:1956–67. 31. Caporali S, Levati L, Graziani G, Muzi A, Atzori MG, Bonmassar E, et al. NF- 15. Maira SM, Stauffer F, Brueggen J, Furet P, Schnell C, Fritsch C, et al. kappaB is activated in response to temozolomide in an AKT-dependent Identification and characterization of NVP-BEZ235, a new orally available manner and confers protection against the growth suppressive effect of dual phosphatidylinositol 3-kinase/mammalian target of rapamycin the drug. J Transl Med 2012;10:252. inhibitor with potent in vivo antitumor activity. Mol Cancer Ther 2008; 32. Liu LZ, Zhou XD, Qian G, Shi X, Fang J, Jiang BH. AKT1 amplification 7:1851–63. regulates cisplatin resistance in human lung cancer cells through the 16. Koul D, Fu J, Shen R, LaFortune TA, Wang S, Tiao N, et al. Antitumor activity mammalian target of rapamycin/p70S6K1 pathway. Cancer Res 2007;67: of NVP-BKM120—a selective pan class I PI3 kinase inhibitor showed 6325–32. differential forms of cell death based on p53 status of glioma cells. Clin 33. Wang X, Ling MT, Guan XY, Tsao SW, Cheung HW, Lee DT, et al. Iden- Cancer Res 2012;18:184–95. tification of a novel function of TWIST, a bHLH protein, in the

www.aacrjournals.org Mol Cancer Ther; 17(1) January 2018 71

Bollard et al.

development of acquired taxol resistance in human cancer cells. Oncogene 42. Yap TA, Yan L, Patnaik A, Fearen I, Olmos D, Papadopoulos K, et al. First-in- 2004;23:474–82. man clinical trial of the oral pan-AKT inhibitor MK-2206 in patients with 34. Yu HG, Ai YW, Yu LL, Zhou XD, Liu J, Li JH, et al. Phosphoinositide 3- advanced solid tumors. J Clin Oncol 2011;29:4688–95. kinase/Akt pathway plays an important role in chemoresistance of 43. Dimou A, Syrigos KN, Saif MW. Update on novel therapies for pancreatic gastric cancer cells against etoposide and doxorubicin induced cell death. neuroendocrine tumors: 2013. JOP 2013;14:377–80. Int J Cancer 2008;122:433–43. 44. Fazio N. Neuroendocrine tumors resistant to mammalian target of 35. Abraham AG, O'Neill E. PI3K/Akt-mediated regulation of p53 in cancer. rapamycin inhibitors: a difﬁcult conversion from biology to the clinic. Biochem Soc Trans 2014;42:798–803. World J Clin Oncol 2015;6:194–7. 36. Yang X, Fraser M, Abedini MR, Bai T, Tsang BK. Regulation of apoptosis- 45. Kirstein MM, Boukouris AE, Pothiraju D, Buitrago-Molina LE, Marhenke S, inducing factor-mediated, cisplatin-induced apoptosis by Akt. Br J Cancer Schutt J, et al. Activity of the mTOR inhibitor RAD001, the dual mTOR 2008;98:803–8. and PI3-kinase inhibitor BEZ235 and the PI3-kinase inhibitor BKM120 37. Caporali S, Levati L, Starace G, Ragone G, Bonmassar E, Alvino E, et al. in hepatocellular carcinoma. Liver Int 2013;33:780–93. AKT is activated in an ataxia-telangiectasia and Rad3-related-dependent 46. Valentino JD, Li J, Zaytseva YY, Mustain WC, Elliott VA, Kim JT, et al. manner in response to temozolomide and confers protection against Cotargeting the PI3K and RAS pathways for the treatment of neuroendo- drug-induced cell growth inhibition. Mol Pharmacol 2008;74:173–83. crine tumors. Clin Cancer Res 2014;20:1212–22. 38. Li X, Lu Y, Liang K, Liu B, Fan Z. Differential responses to doxorubicin- 47. Boora GK, Kanwar R, Kulkarni AA, Pleticha J, Ames M, Schroth G, et al. induced phosphorylation and activation of Akt in human breast cancer Exome-level comparison of primary well-differentiated neuroendocrine cells. Breast Cancer Res 2005;7:R589–97. tumors and their cell lines. Cancer Genet 2015;208:374–81. 39. Dienstmann R, Rodon J, Serra V, Tabernero J. Picking the point of 48. Vandamme T, Peeters M, Dogan F, Pauwels P, Van Assche E, Beyens M, et al. inhibition: A comparative review of PI3K/AKT/mTOR pathway inhibitors. Whole-exome characterization of pancreatic neuroendocrine tumor cell Mol Cancer Ther 2014;13:1021–31. lines BON-1 and QGP-1. J Mol Endocrinol 2015;54:137–47. 40. Sun SY, Rosenberg LM, Wang X, Zhou Z, Yue P, Fu H, et al. Activation of Akt 49. Vandamme T, Beyens M, Peeters M, Van Camp G, de Beeck KO. Next and eIF4E survival pathways by rapamycin-mediated mammalian target of generation exome sequencing of pancreatic neuroendocrine tumor cell rapamycin inhibition. Cancer Res 2005;65:7052–8. lines BON-1 and QGP-1 reveals different lineages. Cancer Genet 2015; 41. Wan X, Harkavy B, Shen N, Grohar P, Helman LJ. Rapamycin induces 208:523. feedback activation of Akt signaling through an IGF-1R-dependent mech- 50. Soefje SA, Karnad A, Brenner AJ. Common toxicities of mammalian target anism. Oncogene 2007;26:1932–40. of rapamycin inhibitors. Target Oncol 2011;6:125–9.

72 Mol Cancer Ther; 17(1) January 2018 Molecular Cancer Therapeutics

Combinatorial Treatment with mTOR Inhibitors and Streptozotocin Leads to Synergistic In Vitro and In Vivo Antitumor Effects in Insulinoma Cells

Julien Bollard, Céline Patte, Patrick Massoma, et al.

Mol Cancer Ther 2018;17:60-72. Published OnlineFirst October 19, 2017.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-17-0325

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2017/10/19/1535-7163.MCT-17-0325.DC1

Cited articles This article cites 50 articles, 20 of which you can access for free at: http://mct.aacrjournals.org/content/17/1/60.full#ref-list-1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/17/1/60. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.