Cotargeting the PI3K and RAS Pathways for the Treatment ofNeuroendocrine Tumors
Joseph D. Valentino1,4, Jing Li1,4, Yekaterina Y. Zaytseva1,4, W. Conan Mustain1,4, Victoria A. Elliott1,Ji Tae Kim1,4, Jennifer W. Harris1,4, Katherine Campbell1,4, Heidi Weiss3,4, Chi Wang3,4, Jun Song1,4,Lowell Anthony2,4, Courtney M. Townsend Jr5, and B. Mark Evers1,4
AbstractBackground: The precise involvement of the PI3K/mTOR andRAS/MEKpathways in carcinoid tumors is
not well defined. Therefore, the purpose of our study was to evaluate the role these pathways play in
carcinoid cell proliferation, apoptosis, and secretion and to determine the effects of combined treatment on
carcinoid tumor inhibition.
Methods: The human neuroendocrine cell lines BON (pancreatic carcinoid), NCI-H727 (lung carci-
noid), andQGP-1 (somatostatinoma)were treatedwith either the pan-PI3K inhibitor, BKM120, or the dual
PI3K–mTOR inhibitor, BEZ235, alone or in combination with the MEK inhibitor, PD0325901; prolifer-
ation, apoptosis, and protein expression were assessed. Peptide secretion was evaluated in BON andQGP-1
cells. The antiproliferative effect of BEZ235, alone or combined with PD0325901, was then tested in vivo.
Results: Both BKM120 and BEZ235 decreased proliferation and increased apoptosis; combination with
PD0325901 significantly enhanced the antineoplastic effects of either treatment alone. In contrast,
neurotensin peptide secretion was markedly stimulated with BKM120 treatment, but not BEZ235. The
combination of BEZ235 þ PD0325901 significantly inhibited the growth of BON xenografts without
systemic toxicity.
Conclusions: Both BKM120 and BEZ235 effectively inhibited neuroendocrine tumor (NET) cell
proliferation and stimulated apoptosis. However, inhibition of the PI3K pathway alone with BKM120
significantly stimulated neurotensin peptide secretion; this did not occur with the dual inhibition of both
PI3K and mTOR using BEZ235 suggesting that this would be a more effective treatment regimen for NETs.
Moreover, the combination of BEZ235 and the MEK inhibitor PD0325901 was a safe and more effective
therapy in vivo compared with single agents alone. Clin Cancer Res; 1–11. �2014 AACR.
IntroductionCarcinoid tumors are well-differentiated neuroendocrine
tumors (NET) most commonly occurring in the gastroin-testinal tract (1). Although rare, accounting for only 0.49%of all malignancies, their incidence has been increasing (1).Effective treatment of carcinoid tumors remains difficult, asthey are often resistant to traditional, cytotoxic chemother-apeutic agents (2). As a result, surgery remains themainstayof treatment for carcinoid tumors (3). Unfortunately, theindolent nature of these tumors often results inpresentation
after the development of metastases (3, 4). This usuallyprecludes curative resection and is associated withdecreased5-year survival (1). The poor response to availablesystemic treatment modalities underscores the need formore effective and targeted therapies in the treatment ofcarcinoid disease.
The PI3K complex is a ubiquitous lipid kinase composedof an 85-kDa regulatory subunit and a 110 kDa catalyticsubunit (5). It exerts its influence through its downstreameffectors Akt and mTOR (6). Abnormal signaling in thispathway frequently occurs through constitutive activationof the PI3K subunits or downregulation or mutation ofPTEN (7). Upregulation of the PI3K pathway has beenidentified as a critical component in the growth and pro-gression of numerous cancer types (8, 9) and has beenimplicated as a key contributor to the development ofmetastatic disease (6, 10, 11). Recent evidence suggests thatabnormal signaling through these pathways is a factor inNETs as well (8, 9). In recent clinical trials, the mTORinhibitor everolimus was used to treat pancreatic NETs andresulted in significantly prolonged progression-free survival(12).
Authors' Affiliations: Departments of 1Surgery, 2Internal Medicine, and3Biostatistics; 4Markey Cancer Center, University of Kentucky, Lexington,Kentucky; and 5Department of Surgery, University of Texas MedicalBranch, Galveston, Texas
Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).
Corresponding Author: B. Mark Evers, Markey Cancer Center, Universityof Kentucky, 800 Rose Street, CC140, Lexington, KY 40536-0093. Phone:859-323-6556; Fax: 859-323-2074; E-mail: [email protected]
The RAS/MEK pathway also has a well-defined role incancer development and progression (13) and is oftenclosely integrated with the PI3K pathway (14). Further-more, cotargeting of these pathways can enhance the ther-apeutic response (15–17). Although the precise pathwaymutations in carcinoid tumors are less clear, the activationof RAS/MEKpathway components, such as the downstreameffector extracellular signal–regulated kinase (ERK), hasbeen identified. For example, in a study evaluating KRASand BRAFmutations in NETs, only one BRAFmutation andno KRAS mutations were identified; however, the majorityof these tumors showed constitutive activation of ERK (18).Despite the recognition of these pathways as potentialcontributors to carcinoid disease, their specific role incarcinoid tumors is not well defined.
Management of carcinoid tumors is further complicat-ed by their ability to secrete bioactive amines and pep-tides. These secretory products can cause carcinoid syn-drome, which is characterized by severe flushing, diar-rhea, bronchospasm, and cardiac fibrosis, and can resultin death (3). The PI3K pathway has been linked to thecontrol of microtubule dynamics, indicating its potentialinvolvement in vesicle trafficking and the secretion ofthese bioactive products (19). Furthermore, PI3K hasbeen identified as a potential negative regulator of pep-tide secretion. In a recent study, inhibition using thep110a-specific inhibitor, PIK-75, resulted in upregulationof neurotensin peptide secretion from the human carci-noid cell line, BON (20). These findings imply that theuse of PI3K inhibitors for the treatment of carcinoid
disease may increase secretion, which could result inadverse sequelae. The purpose of our study was to definethe roles the PI3K and RAS/MEK pathways play in carci-noid cell proliferation and secretion as well as to deter-mine the effects of combined inhibition on carcinoidtumor growth, apoptosis, and peptide secretion.
Materials and MethodsMaterials
Dulbecco’s Modified Eagle Medium (DMEM)/F12Kwas obtained from Invitrogen and RPMI medium fromAmerican Type Culture Collection (ATCC). FBS was pur-chased from Atlanta Biologicals. The mTOR inhibitorrapamycin was obtained from Calbiochem. The CellDeath Detection ELISAplus Kit was obtained from Roche.Annexin V Alexa Fluor 488 was purchased from Invitro-gen. The pBabe-puro p110a plasmids were obtained fromAddgene. Lipofectamine 2000 was acquired from Invitro-gen. NuPAGE 4% to 12% Bis–Tris gels were also pur-chased from Invitrogen and Sequi-blot polyvinylidenedifluoride (PVDF) membranes from Bio-Rad. The follow-ing antibodies were obtained from Cell Signaling Tech-nology: pAkt, Akt, pERK, ERK, p-p70S6K, p70S6K, pS6,S6, p4E-BP1, 4E-BP1, PARP, and cleaved caspase-3. Anenhanced chemiluminescence (ECL) detection systemwas purchased from GE healthcare. The neurotensinenzyme immunoassay (EIA) Kit was acquired from Phoe-nix Pharmaceuticals. The serotonin ELISA Kit wasobtained from Immuno-Biological Laboratories Inc. Ofnote, 1-methyl-2-pyrrolidinone (NMP), polyethylene gly-col 300, hydroxypropyl methycellulose, and 0.2% Tween80 were purchased from Sigma-Aldrich.
Small-molecule inhibitorsPIK-75, a p110a-specific inhibitor of the PI3K complex;
rapamycin, an mTOR inhibitor; and PD0325901 andPD98059, MEK inhibitors, were purchased from CaymanChemical. BKM120, a pan-PI3K inhibitor, was purchasedfrom Active Biochem, and BEZ235, a dual PI3K/mTORinhibitor, was purchased from LC Laboratories.
Cell lines and cultureThe human carcinoid cell lines BON (pancreatic car-
cinoid; ref. 21) and NCI-H727 (lung carcinoid; refs. 22,23) were used in this study. The NET cell line, QGP-1(human somatostatinoma; ref. 24) was also used because,similar to BON cells, it produces and secretes the neuro-tensin peptide (24). The cell lines (BON, QGP-1, andNCI-H727) were authenticated in May 2012 at GeneticaDNA Laboratories, Inc. and 17 autosomal short tandemrepeat loci and the sex identity locus for the cells wereprofiled using the PowerPlex 18D PCR Amplification Kit(Promega Corporation). BON cells were maintained in a1:1 mixture of DMEM and nutrient mixture F12K,supplemented with 5% FBS. NCI-H727 and QGP-1 cellswere maintained in ATCC-formulated RPMI-1640 medi-um with 10% FBS. Cells were incubated in 5% CO2 at37�C.
Translational RelevanceThe increasing incidence of neuroendocrine tumors
(NET), the frequently advanced stage at the time ofpresentation, and the poor response to traditional che-motherapeutic agents indicate an important need formore effective systemic therapies. Molecular targetedtherapies, such as mTOR inhibitors, have shown prom-ise as a therapeutic alternative. In this study, we evalu-ated the efficacy of a pan-PI3K inhibitor (BKM120) and adual PI3K/mTOR inhibitor (BEZ235), either alone or incombination with an MEK inhibitor (PD0325901) onproliferation and peptide secretion in human NET cells.Our findings indicate that the dual inhibition of PI3KandmTOR, using BEZ235, effectively increased NET cellapoptosis and, in contrast with PI3K inhibition alonewith BKM120, did not stimulate NET peptide secretion.Furthermore, the combination of BEZ235 with the MEKinhibitor represented a safe andmore effective treatmentregimen in vivo compared with single agents alone.Therefore, our findings suggest that the dual inhibitionof PI3K and mTOR pathways may represent a noveltreatment strategy for advanced NETs and that combi-nation with an MEK inhibitor may further enhance theantiproliferative effect.
Generation of stable cell linesTo produce retrovirus, 293FT packaging cells cultured in
60-mmdishes were cotransfected with pBabe-puro plasmid(1 mg) and Ampho packaging plasmid by lipofectamine2000 and incubated in growth medium overnight. Subse-quently, the cells were cultured in complete medium[growth medium plus 1 mmol/L minimum essential medi-um (MEM) sodium pyruvate] for 24 hours; the supernatantcontaining the retrovirus was collected, filtered through a0.45-mm surfactant-free cellulose acetate (SFCA) sterilesyringe filter, and used to infect target cells. BON cells in6-well plates (5 � 105 cells/well) were incubated with theviral supernatant for 24 hours; cells were then incubatedwith growth medium for an additional 24 hours. Theinfected cells were subcultured in 100-mm dishes in freshmedium containing puromycin (2.5 mg/mL). Puromycin-resistant cell pools were collected and the overexpressionlevels were monitored by Western blot analysis.
Cell proliferationCells were plated in 24-well plates at a density of 3� 104
cells/cm2. Drug treatments were initiated the followingmorning after allowing the cells to adhere. Media wasexchanged with fresh drug media at 36 hours. Cell prolif-eration was assessed at 72 hours directly by cell countingusing a Beckman Coulter Cell Viability Analyzer.
DNA fragmentation ELISACells were plated in 24-well plates at a density of 5� 104
cells/cm2. Drug treatments were initiated the followingmorning and continued for 24 hours. Apoptosis was mea-sured by DNA fragmentation using the Cell Death Detec-tion ELISAplus as previously described (25).
Annexin VBON cells were plated in 6-well plates at a density of 9 �
105 cells per well. Drug treatments were initiated the nextday and continued for 24 hours. Cells were collected,washed, stained with Annexin V Alexa Fluor 488 accordingto the manufacturer’s protocol (A13201; Molecular Probes;Invitrogen) and analyzed by flow cytometry (UK FlowCytometry Core Facility).
Western blot analysisCells were plated in 6-well plates at a density of 5 � 104
cells/cm2. Whole-cell lysates were collected following 24hours of drug treatment. Total protein was resolved onNuPAGE 4% to 12% Bis–Tris gels and transferred toSequi-blot PVDF membranes. Membranes were incubatedwith specific primary antibodies and subsequently horse-radish peroxidase–conjugated secondary antibody. Pro-teins were visualized using an ECL detection system.For the in vivo study, three tumors per treatment group
were selected at random for analysis. Briefly, pieces of frozentumors were placed in tubes with ceramic beads. Ice-coldlysis buffer with protease inhibitors was added to the tubes,and tissues were homogenized in a tissue homogenizer(Bullet Blender Advance) at 4�C. Homogenized samples
were centrifuged and supernatants collected. Protein con-centration of the samples was detected by Bradford assay.Western blots were performed as described above.
Neurotensin EIACells were plated in 24-well plates at a density of 1� 105
cells/cm2 and grown for 48 hours. Cells were treated withinhibitors in growth medium for 1 hour. Media werecollected and stored at �80�C. Secreted neurotensin pep-tide was measured by neurotensin EIA as previouslydescribed (26, 27).
Serotonin ELISACells were plated in 24-well plates at a density of 1� 105
cells/cm2. After 24 hours, cells werewashed andmaintainedwith serum-free medium. The next day, cells were treatedwith inhibitors in serum-free medium for 1 hour. Mediawere collected and stored at�80�C. Secreted serotonin wasmeasured by a serotonin ELISA according to the manufac-turer’s instructions. The data for secreted serotonin werenormalized by protein concentration from parallel celllysates.
In vivo studiesTwo-month-old athymic nude male mice weighing
approximately 25 g were used for this study. Mice wereacclimated for 1 week at which time xenografts were estab-lished by injection of 1� 107 BON cells (in 100 mL of sterilePBS) subcutaneously in the flank. The mice were thenrandomized into four groups (n¼ 15 per group): (i) vehiclecontrol, (ii) BEZ235 (45mg/kg), (iii) PD0325901 (5mg/kg),or (iv) BEZ235 (45 mg/kg) and PD0325901 (5 mg/kg).BEZ235 was dissolved in one part NMP to nine parts poly-ethylene glycol 300. PD0325901 was formulated in 0.5%hydroxypropyl methycellulose plus 0.2% Tween 80. Vehiclecontrol consisted of a combination of NMP, polyethyleneglycol 300, and 0.5% hydroxypropyl methycellulose plus0.2% Tween 80 prepared in the same ratios as used for thedrugpreparations.Allmicewere treatedbydailyoral gavage5days perweek for 6weeks.Micewereweighed three times perweek during the experiment to monitor for toxicity. Tumorsize was measured using vernier calipers and volume calcu-
lated using the equation: ðlength�width2Þ=2. At the com-
pletion of the experiment, mice were sacrificed and tumorsexcised and weighed. All animal procedures were performedaccording to protocols approved by the Institutional AnimalCare and Use Committee.
Statistical analysisStudy endpoints from the in vitro experiments, including
cell proliferation, apoptosis, and neurotensin secretion, aresummarized using bar graphs withmeans� SEM. Compar-isons across treatment groups used the ANOVAmodel withcontrasts generated from the model to perform specificcomparisons, including linear trend for increasing doselevels, pairwise comparisons of treatment versus controland combination versus monotherapy. The model alsoincluded experiment as an additional factor to account for
PI3K and RAS Inhibitor Treatments for Carcinoid Tumors
repeat experiments in these studies. Tumor growth curvesand body weight over time were plotted. Two-sample t testswere used to compare tumor weight, fold change in tumorvolume, and fold change in bodyweight between treatmentgroups.
ResultsPI3K inhibition decreases carcinoid cell proliferationbut increases signaling through the RAS/MEK pathway
We first determined whether PI3K inhibitors alone wereeffective in the treatment of NETs. BON carcinoid cells wereplated in equal numbers and treated with the pan-PI3Kinhibitor BKM120 at various doses (1.0, 2.5, and 5.0mmol/L) or the dual PI3K/mTOR inhibitor BEZ235 at 10,100, or 1,000 nmol/L. Cells were counted at 72 hours toevaluate responsiveness to treatment (Fig. 1A and B). Cellnumber was significantly decreased at all doses for bothBKM120 andBEZ235. Furthermore, a dose-dependent trend
wasobservedacross the rangeofdoses tested. TwoadditionalNET cell lines, NCI-H727 and QGP-1, were tested as wellwith a similar decrease in proliferation identified (Supple-mentary Fig. S2). The role ofmTORwas also evaluated usingrapamycin treatments (Supplementary Fig. S1). Significantinhibition was identified; however, the effects were lesspronounced than with PI3K inhibition. To further confirmthe benefit of targeted PI3K inhibition, effects on prolifer-ation were evaluated using short hairpin RNA directedagainst the alpha subunit of the PI3K complex. Analysis wasperformed directly by cell counting at 24 hours intervals upto 96 hours (Supplementary Fig. S2C). Significant reductionin proliferation was identified at all time points.
Western blot analysis using whole BON cell lysates wasperformed to confirm that these effects were the result oftargeted inhibition of the PI3K pathway (Fig. 1C and D).pAkt levels were progressively decreased with increasingdoses of BKM120. Conversely, treatment with BEZ235 at
Figure 1. PI3K inhibition decreasesBONcell proliferationbut increasessignaling through the RAS/MEKpathway. Experiments wereperformed in triplicate andrepeated three times. BON cellswere plated in 24-well plates at adensity of 3 � 104 cells per well.Treatments were initiated thefollowing morning with increasingdoses of either (A) BKM120 (0–5mmol/L) or (B) BEZ235 (0–1,000nmol/L). Media was exchanged forfresh drugmedia again at 36 hours.Proliferation was assessed directlyby cell counting at 72 hours(�,P<0.05 vs. control). ForWesternblot analysis, cells were plated in 6-well plates at a density of 5 � 104
cells/cm2. Whole-cell lysates werecollected following 24 hours ofdrug treatment with either (C)BKM120 (0–5 mmol/L) or (D)BEZ235 (0–1,000 nmol/L). b-Actinwas used as a loading control.
10 nmol/L resulted in a slight induction of pAkt, possibly asa result of mTOR inhibition with associated loss of negativePI3K inhibition through the S6K feedback loop; this mech-anism is supported by the finding of increased pAkt fol-lowing treatment with mTOR inhibition alone (Supple-mentary Fig. S1). However, decreased pAkt was observedat the 100 and 1,000 nmol/L BEZ235 doses. The down-stream targets p70S6K, S6, and 4E-BP1 were also analyzed.Phosphorylation of all of these targets was decreased withincreasing doses of BKM120. For BEZ235, phosphorylationof these targets was minimally affected at 10 nmol/L;however, phosphorylation of all three targets was decreasedat 100 and 1,000 nmol/L doses.InhibitionofmTORC1has beendemonstrated to activate
RAS/MEK signaling through a S6K–PI3K–RAS feedbackloop in several cancer types (28). Given such interactionsbetween thePI3K andRAS/MEKpathways,we also analyzedthe levels of pERK to evaluate cross-talk between these twopathways in NETs. Interestingly, levels of pERK wereincreased at all doses for both the BKM120 and BEZ235treatment groups indicating compensatory signalingthrough the RAS/MEK pathway following treatment witha PI3K inhibitor alone.
PI3K inhibition results in an increase in apoptosisWe next determined whether the response to treatment
was the result of a cytostatic effect or, conversely, whether itrepresented an increase in apoptosis. BON cellswere treatedwith BKM120 at doses of 1.0, 2.5, and 5.0 mmol/L or withBEZ235 at 10, 100, and 1,000 nmol/L for 24 hours andapoptosis measured by DNA fragmentation (Fig. 2). DNAfragmentation was significantly increased at themiddle andhighest doses tested for both treatment groups. mTORinhibition alone following treatment with rapaymycin wasalso assessed (Supplementary Fig. S1). No increase in apo-ptosis was identified.
To corroborate this apoptotic effect, Western blot analysisof PARP cleavage was also performed. Increased PARP cleav-age corresponded with increases in DNA fragmentation.NCI-H727andQGP-1 cellswere tested to confirm the resultsin other NET cells. NCI-H727 cells showed a dose-depen-dent trend similar to that identified in BON cells with DNAfragmentation significantly increased at both the middleand highest doses tested; increased DNA fragmentation wasalso identified in QGP-1 cells (Supplementary Fig. S3).
Combination of PI3K þ MEK inhibitor treatmentenhances inhibition of carcinoid cell proliferation
After identifying the upregulation of pERK noted withPI3K treatment alone, we next hypothesized that the addi-tion of anMEK inhibitor (PD0325901) would augment theantiproliferative response. To test this hypothesis, BONcellswere treated with BKM120 (5 mmol/L), PD0325901 (100nmol/L), or the combination of BKM120þ PD0325901. Inanother experiment, BON cells were treated with BEZ235(1,000 nmol/L), PD0325901 (100 nmol/L), or the combi-nation of BEZ235 þ PD0325901. Cell numbers wereassessed using the Coulter counter at 72 hours (Fig. 3). Allindividual and combination treatments significantlydecreased cell number relative to control (vehicle treat-ment). Furthermore, the combination significantly reducedBON cell numbers relative to either of the individual treat-ments. Similar to the BON cell line, all individual andcombination treatments significantly decreased NCI-H727 and QGP-1 cell numbers relative to control (Supple-mentary Fig. S4). mTOR inhibition alone with rapamycinshowed an enhanced effect in BON and NCI-H727 cells;however, the response was not as pronounced as with PI3KþMEK inhibition and was absent for QGP-1 cells (Supple-mentary Fig. S1). Together, these data support enhancedinhibition of cell proliferation with combination of PI3KþMEK treatments.
Figure 2. PI3K inhibition increasesapoptosis in BON cells.Experiments were performed intriplicate and repeated three times.Representative graph is shown.Cells were plated in 24-well platesat a density of 5 � 104 cells/cm2.Treatments were initiated thefollowing morning with increasingdoses of either (A) BKM120 (0–5mmol/L) or (B) BEZ235 (0–1,000nmol/L) and continued for 24hours. ApoptosiswasmeasuredbyDNA fragmentation using the CellDeath Detection ELISAplus
(�, P < 0.05 vs. control). Whole-celllysates were also collected after 24hours of treatment and Westernblot analysis was performed toanalyze for PARPcleavage.b-Actinwas used as a loading control.
PI3K and RAS Inhibitor Treatments for Carcinoid Tumors
Western blot analysis of BON cell lysate was also per-formed to evaluate the effects on PI3K and RAS/MEKpathway components. Treatment with a PI3K inhibitorindividually or in combination resulted in decreased pAktand reduced phosphorylation of the downstream effectorsp70S6K, S6, and 4E-BP1. Treatment with PD0325901 (100nmol/L), both alone and in combination, decreased pERKexpression. Importantly, the upregulation of pERKobserved following treatment with BKM120 or BEZ235alone was abrogated when combined with the MEKinhibitor.
Combination of PI3K þ MEK inhibition increasesapoptosis relative to individual agents
After confirming a decrease in cell number with combi-nation treatment, we next determined whether theenhanced response associated with the MEK inhibitor wasdue to increased apoptosis. BON cells were treated witheither BKM120 (5 mmol/L) or BEZ235 (1,000 nmol/L) assingle agents or in combination with PD0325901 (100
nmol/L) as described above, and apoptosis measured byDNA fragmentation (Fig. 4). Both the PI3K and MEKinhibitors significantly increased apoptosis when adminis-tered as individual treatments. Furthermore, combinationof PI3KþMEK treatment resulted in a significant increase inapoptosis compared with either individual agents. Similarto the BON cell line, both individual and combinationtreatments significantly increased DNA fragmentation inNCI-H727 andQGP-1 cells relative to control (Supplemen-tary Fig. S5A and S5B). This enhanced effect was not presentfollowing treatment with mTOR þ MEK inhibitors (Sup-plementary Fig. S1).
To corroborate these findings, BON cells were againtreated with single and combination treatments asdescribed. Cells were stained using Annexin V Alex Fluor488and sortedbyflowcytometry (Supplementary Fig. S5C).Consistent with findings from theDNA fragmentation stud-ies, an increase in cell deathwas identifiedwith all treatmentgroups. Increased cell death was also significantly increasedfor BEZ235 þ PD0325901 compared with BEZ235 or
Figure 3. Combination treatmentwith PI3K and MEK inhibitorsresults in more pronouncedinhibition of BON cell proliferation.Experiments were performed intriplicate and repeated three times.BON cells were plated in 24-wellplates at a density of 3 � 104 cellsper well. Treatments were initiatedthe following morning with either(A) control, BKM120 (5 mmol/L),PD0325901 (100 nmol/L), orcombination of BKM120 þPD0325901 (5 mmol/L/100 nmol/L),or (B) control, BEZ235 (1,000nmol/L), PD0325901 (100 nmol/L),or combination of BEZ235 þPD0325901 (1,000/100 nmol/L).Media was exchanged for freshdrug media at 36 hours.Proliferation was assessed directlyby cell counting at 72 hours(�, P < 0.05 vs. control; †, P < 0.05vs.monotherapy). ForWestern blotanalysis, cells were plated in 6-wellplates at a density of 5 � 104
cells/cm2. Whole-cell lysates werecollected following 24 hours ofdrug treatment with either (A)control, BKM120 (5 mmol/L),PD0325901 (100 nmol/L), orcombination of BKM120 þPD0325901 (5 mmol/L/100 nmol/L),or (B) control, BEZ235 (1,000nmol/L), PD0325901 (100 nmol/L),or combination of BEZ235 þPD0325901 (1,000/100 nmol/L).b-Actin was used as a loadingcontrol.
PD0325901 alone and for BKM120 þ PD0325901 com-pared with PD0325901 alone (Supplementary Fig. S5C).Subsequent Western blot analysis also showed increasedPARP cleavage with individual PI3K or MEK inhibitor treat-ment (Fig. 4). The increase was even more prominent withcombination of PI3K þ MEK inhibition. In addition,increased cleavage of caspase-3 was identified in the com-bination treatment groups.
Treatment with BKM120, but not BEZ235 orPD0325901, increases peptide secretion fromNET cellsAny treatment regimen considered for NETs must
decrease cell proliferation and, importantly, cannotincrease peptide secretion, which can lead to untowardtreatment sequelae. Li and colleagues (20), in our labora-tory has shown that treatment with the p110a PI3K inhib-itor PIK75 increases neurotensin secretion, possibly as aresult of a loss of inhibitory feedback from the PI3K path-way. We next determined whether the broader spectrumPI3K inhibitors, BKM120 and BEZ235, or the MEK inhib-itor, PD0325901, are associated with increased peptidesecretion. BON andQGP-1 cells were treated with BKM120(5 mmol/L) or BEZ235 (1,000 nmol/L) as single agents or incombination with PD0325901 (100 nmol/L); media wascollected and analyzed for neurotensin secretion (Fig. 5).Individual treatment with BKM120 significantly stimulatedneurotensin release in both BON and QGP-1 cells. Con-versely, BEZ235 did not increase neurotensin secretion ineither cell line, whereas PD0325901 significantly decreasedsecretion.When tested in combination, PD0325901didnotcompletely abrogate the increased secretion associated withBKM120; however, the decreased secretion identified with
PD0325901 persisted with combination of BEZ235 þPD0325901 treatments.
The ability of carcinoid tumors to secrete serotonin is alsowell described and has been associated with carcinoidsyndrome. Therefore, BON and QGP-1 cells were againtreated with single and combination agents as describedabove and media collected for analysis of serotonin secre-tion (Supplementary Fig. S6A and 6B). Similar to analysis ofneurotensin secretion, serotonin secretion was decreasedfollowing treatment with PD0325901. This decreasewas maintained following combination of BEZ235 þPD0325901 and BKM120 þ PD0325901 treatments. Ex-pression of CgA, another secretory product of carcinoidtumors, was also analyzed byWestern blot analysis in BONcells (Supplementary Fig. S6C). Decreased expression ofCgAwasnoted in all treatment groups; however, the greatestdecrease was identified following treatment with BEZ235and BEZ235 þ PD0325901.
Combination of BEZ235 þ PD0325901 treatment mosteffectively suppresses in vivo subcutaneous BONxenograft growth
Both BEZ235 and BKM120 significantly increased NETcell death in vitro; however, BKM120 resulted in increasedNET peptide secretion. Therefore, we next focused ourstudies on BEZ235 and the efficacy and safety of BEZ235treatment alone or combined with PD0325901 in vivo.Subcutaneous BON xenografts were established in athymicnude mice and allowed to grow over a 14-day period.Groups of 15 mice were then randomized to treatmentwith vehicle (control), BEZ235 (45 mg/kg), PD0325901(5 mg/kg), or the combination of BEZ235 (45 mg/kg) þ
Figure 4. Combination of PI3K þ MEK inhibition enhances apoptosis in BON cells compared with either agent individually. Experiments were performed intriplicate and repeated three times. Representative graph is shown. Cells were plated in 24-well plates at a density of 5 � 104 cells/cm2. Treatments wereinitiated the following morning with either (A) control, BKM120 (5 mmol/L), PD0325901 (100 nmol/L), or combination of BKM120þ PD0325901 (5 mmol/L/100nmol/L), or (B) control, BEZ235 (1,000 nmol/L), PD0325901 (100 nmol/L), or combination of BEZ235 þ PD0325901 (1,000/100 nmol/L) and continuedfor 24 hours. Apoptosis was measured by DNA fragmentation using the Cell Death Detection ELISAplus (�, P < 0.05 vs. control; †, P < 0.05 vs. monotherapy).Whole-cell lysates were also collected after 24 hours of treatment and Western blot analysis performed for PARP and caspase-3 cleavage. b-Actinwas used as a loading control.
PI3K and RAS Inhibitor Treatments for Carcinoid Tumors
PD0325901 (5 mg/kg). Tumor volumes at the initiation oftherapy were comparable between all treatment groups(average volume52.3–54.2mm3).Micewere sacrificed after6 weeks of treatment when control tumors had reached themaximal size limitations (average volume ¼ 653 mm3) asindicated by our institutional guidelines.
Measurement of tumor volumes showed significant inhi-bition of tumor growth for individual BEZ235 andPD0325901 treatments compared with control (Fig. 6A).In addition, combination treatment with BEZ235 þPD0325901 significantly inhibited tumor growth com-paredwith both control and individual treatments. At studycompletion, tumor volumes in the combination treatmentgroup increased by only 2.4-fold compared with the 12.5-fold increase in the control group. Comparisons of tumorweight were not significant for BEZ235 relative to control(P ¼ 0.133) or for combination treatment compared withPD0325901 alone (P ¼ 0.051), although they did follow asimilar pattern as tumor volume (Fig. 6B).
To confirm on-target effects following treatments in vivo,the BON xenografts were harvested and analyzed by West-ern blot analysis (Fig. 6C). Phosphorylation of Akt, p70S6K,S6, 4E-BP1, and ERK were evaluated. Although pAkt hadvariable expression, decreased phosphorylation of thedownstream targets p70S6K, S6, and 4E-BP1were identifiedfollowing treatment with BEZ235 and BEZ235 þPD0325901. Decreased phosphorylation of ERK was alsonoted for groups treated with PD0325901 and BEZ235 þ
PD0325901. These findings support the targeted effects ofthese inhibitors in vivo.
Four mice died early in the treatment period, 2 in theBEZ235 treatment group and 2 in the combination treat-ment group. Mortality occurred shortly after oral gavage inthese mice, and necropsy showed mottling of the lungs,which is consistent with administration of the treatmentinto the trachea. As a more accurate assessment of toxicity,there was no significant reduction in weight between themice in the various treatment groups compared with con-trol nor were other signs of acute or delayed toxicity present(Fig. 6D). Collectively, these data indicate that the dose andtreatment schedule tested in this experiment are well tol-erated by the mice and demonstrate effective BON tumorinhibition.
DiscussionIn this study, we evaluated the role of the PI3K and RAS/
MEK pathways in NET proliferation, apoptosis, and secre-tion. First, we showed that PI3K pathway inhibition resultsin both decreased proliferation and increased apoptosis inmultiple NET cell lines, whereas concomitantly increasingsignaling through the RAS/MEK pathway. Second, we dem-onstrated that using the MEK inhibitor, PD0325901, toblock this compensatory signaling enhanced the antineo-plastic effects over either agent alone. Interestingly, wefound that, similar to the p110a-specific PI3K inhibitorPIK75, the pan-PI3K inhibitor BKM120, stimulated peptide
Figure 5. BKM120 increasesneurotensin (NT) secretion,whereas BEZ235 does not causean increase in NT secretion, andPD0325901 is associated with adecrease in NT secretion. A, BONand (B) QGP-1 cells were plated in24-well plates at a density of1� 105 cells/cm2 and grown for 48hours. Cells were treated witheither (top) control, BKM120 (5mmol/L), PD0325901 (100 nmol/L),or combination of BKM120 þPD0325901 (5 mmol/L/100 nmol/L),or (bottom) control, BEZ235 (1,000nmol/L), PD0325901 (100 nmol/L),or combination of BEZ235 þPD0325901 (1,000100 nmol/L) ingrowth medium for 1 hours. Mediawere collected and secreted NTpeptide measured by NT EIA(�, P < 0.05 vs. control).
release from the functioning BON carcinoid cell line; how-ever, the dual PI3K/mTOR inhibitor, BEZ235, did notincrease secretion. Treatment with the MEK inhibitor,PD0325901, decreased neurotensin secretion; this decreasewas sustained when used in combination with BEZ235.Finally, we demonstrated that combination treatment withBEZ235 and PD0325901 is both safe and effective for the invivo treatment of mice bearing BON xenografts.Mutations resulting in aberrant activation of signaling
pathways provide a selective advantage for cancer cellscompared with their normal counterparts. Upregulation ofthe PI3K pathway, in particular, has been demonstrated as akey contributor to tumor progression in a number of cancertypes (9). In carcinoid tumors, compromised regulation ofthe PI3K complex has been attributed to activation of Akt orloss of PTEN expression (29–31). In addition, global inhi-bition of PI3K with LY294002 or selective siRNA knock-down of Akt1 has previously been shown to inhibit carci-noid cell proliferation (32, 33). In this study, we extendthese findings by confirming that the pan-PI3K inhibitor,BKM120, and the dual PI3K/mTOR inhibitor, BEZ235,decrease proliferation inmultiple NET cell lines. We furthershow that, at least in vitro, the mechanism may include
increased cell death rather than strictly a cytostatic effect;however, these cytotoxic effects are less evident in vivo.
The abnormal activation of the RAS/MEK pathway alsooccurs in many cancers, including carcinoid tumors(18, 34). This was demonstrated in a study of 40 primarygastroenteropancreatic NETs, in which activated ERK wasidentified in all specimens (18). In addition to the individ-ual contribution of theRAS/MEKpathway, significant cross-talk with the PI3K pathway can also occur. For example,treatmentwith themTORC inhibitor, RAD001, can increaseRAS/MEK pathway signaling through a S6K–PI3K–RASfeedback loop (28). Zitzmann and colleauges (15) alsoshowed that combination treatment with an mTOR inhib-itor and a high-dose RAF inhibitor is capable of enhancingantitumor effects in neuroendocrine cells. In our study, wedemonstrate that, in addition to mTOR, inhibition of thePI3K complex alone also results in compensatory signalingthrough the RAS/MEK pathway, as indicated by an increasein pERK expression following treatment with both BKM120and BEZ235 at all doses tested. We further show thatcombining PI3K andMEK inhibition significantly increasesDNA fragmentation and is associated with increasedPARP and caspase-3 cleavage. Furthermore, combination
Figure 6. Combination treatment with BEZ235 and PD0325901 most effectively suppresses in vivo growth of subcutaneous carcinoid xenografts.Subcutaneous injections of 1� 106 BON cells in 100 mL of sterile PBSwere performed in athymic nudemice. After allowing time for xenograft establishment,mice were randomized into either vehicle control, BEZ235 (45 mg/kg), PD0325901 (5 mg/kg), or combination of BEZ235 þ PD0325901 (45 and 5 mg/kg)treatment groups (n ¼ 15/group). Treatments were administered by oral gavage once daily, 5 days per week, for 6 weeks. A, tumor size was
measured with Vernier calipers and the equation ðlength�width2Þ=2 was used to calculate volume (�, P < 0.05 vs. control; †, P < 0.05 vs. monotherapy).B, tumors were excised at the end of the treatment period andweighed (�,P < 0.05 vs. control; †,P < 0.05 vs. BEZ235monotherapy). C, protein was collectedfrom harvested xenografts and expression analyzed by Western blot analysis. b-Actin was used as a loading control. D, to monitor for toxicity,mice were weighed weekly for 6 weeks.
PI3K and RAS Inhibitor Treatments for Carcinoid Tumors
treatment using BEZ235 and PD0325901 in vivo is both asafe and effective treatment and offers a greater therapeuticbenefit than either agent alone. Together these data providea strong argument for the use of dual PI3K and MEKinhibition as treatment for NETs.
A unique feature of carcinoid tumors is their ability tosecrete bioactive substances. When the tumor is localized tothe bowel, these substances are metabolized by the liver,thereby preventing sequelae; however, when liver metasta-ses are present, these substances are released systemicallyand can result in carcinoid syndrome with potentially life-threatening effects (35). As shownbyLi and colleagues (20),the loss of negative regulation from the PI3K pathwayresults in an increase in NET peptide secretion, a processthat seems to be dependent upon the p110a subunit. In ourstudy, we show that while the pan-PI3K inhibitor, BKM120,significantly increases carcinoid cell apoptosis, peptidesecretion is also stimulated. Furthermore, the decreasedsecretion observed with PD0325901 treatment only par-tially counteracts this increase when used in combinationwith BKM120. This augmentation of carcinoid cell secretionhas the potential to produce adverse effects in the clinicalsetting. In contrast, we demonstrate that the dual PI3K/mTOR inhibitor, BEZ235, significantly inhibits NET growthbut, in contrast with BKM120, does not stimulate peptiderelease. These results argue that dual inhibition withBEZ235 and PD0325901 is unlikely to be associated withadverse effects related to enhanced secretion and mayactually provide a therapeutic benefit in those patients withcarcinoid syndrome.
In summary, we show that PI3K inhibition effectivelydecreases carcinoid cell growth and induces apoptosis;however, compensatory signaling occurs through theRAS/MEK pathway. Dual inhibition of the PI3K and RAS/
MEKpathways blocks this feedback activation and results inan enhanced therapeutic benefit compared with eitheragent alone. Furthermore, we show that the increasedsecretion that can occur following treatment with PI3Kcomplex inhibitors does not occur with the dual PI3K/mTOR inhibitor BEZ235 and, in addition, that MEK inhi-bition can actually decrease secretion. Finally, we confirmthe safety and efficacy of dual PI3K andRAS/MEK inhibitionin vivo. Collectively, our data strongly support a combina-tion strategy of dual PI3K/mTOR and MEK inhibition as aneffective treatment for NETs.
Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.
Authors' ContributionsConception and design: J.D. Valentino, B. Mark EversDevelopment of methodology: J.D. Valentino, J. Li, Y.Y. Zaytseva, J. Song,B. Mark EversAcquisitionofdata (provided animals, acquired andmanagedpatients,provided facilities, etc.): J.D. Valentino, Y.Y. Zaytseva, W.C. Mustain, V.A.Elliott, J.T. Kim, C.M. Townsend, B. Mark EversAnalysis and interpretation of data (e.g., statistical analysis, biosta-tistics, computational analysis): J.D. Valentino, V.A. Elliott, H. Weiss, C.Wang, L. Anthony, B. Mark EversWriting, review, and/or revision of the manuscript: J.D. Valentino, Y.Y.Zaytseva, W.C. Mustain, J.W. Harris, H. Weiss, L. Anthony, C.M. Townsend,B. Mark EversStudy supervision: J.D. Valentino
Grant SupportThis work is supported by the P20CA153043 (GI SPORE),
R37AG010885-21, and R01 DK048498-18 grants from the NIH.The costs of publication of this article were defrayed in part by the
payment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.
Received July 10, 2013; revisedNovember 27, 2013; acceptedDecember 2,2013; published OnlineFirst January 17, 2014.
References1. Modlin IM, Lye KD, Kidd M. A 5-decade analysis of 13,715 carcinoid
tumors. Cancer 2003;97:934–59.2. KulkeMH. Clinical presentation andmanagement of carcinoid tumors.
Hematol Oncol Clin North Am 2007;21:433–55.3. Valentino J, Evers BM. Recent advances in the diagnosis and treat-
ment of gastrointestinal carcinoids. Adv Surg 2011;285–300.4. Rorstad O. Prognostic indicators for carcinoid neuroendocrine tumors
of the gastrointestinal tract. J Surg Oncol 2005;89:151–60.5. Cantley LC. The phosphoinositide 3-kinase pathway. Science 2002;
296:1655–7.6. Gulhati P, Cai Q, Li J, Liu J, Rychahou PG, Qiu S, et al. Targeted
inhibition of mammalian target of rapamycin signaling inhibitstumorigenesis of colorectal cancer. Clin Cancer Res 2009;15:7207–16.
7. Silva SR, Zaytseva YY, Jackson LN, Lee EY, Weiss HL, Bowen KA,et al. The effect of PTEN on serotonin synthesis and secretionfrom the carcinoid cell line BON. Anticancer Res 2011;31:1153–60.
8. Luo J, Manning BD, Cantley LC. Targeting the PI3K–Akt pathway inhuman cancer: rationale and promise. Cancer Cell 2003;4:257–62.
9. Vivanco I, Sawyers CL. The phosphatidylinositol 3-Kinase AKT path-way in human cancer. Nat Rev Cancer 2002;2:489–501.
10. Rychahou PG, Jackson LN, Silva SR, Rajaraman S, Evers BM. Tar-getedmolecular therapy of the PI3K pathway: therapeutic significanceof PI3K subunit targeting in colorectal carcinoma. Ann Surg 2006;243:833–42.
11. Rychahou PG, Kang J, Gulhati P, Doan HQ, Chen LA, Xiao SY, et al.Akt2 overexpression plays a critical role in the establishment ofcolorectal cancer metastasis. Proc Natl Acad Sci U S A 2008;105:20315–20.
12. Yao JC, Shah MH, Ito T, Bohas CL, Wolin EM, Van Cutsem E, et al.Everolimus for advanced pancreatic neuroendocrine tumors. N Engl JMed 2011;364:514–23.
13. McCubrey JA, Steelman LS, Chappell WH, Abrams SL, Wong EW,Chang F, et al. Roles of the Raf/MEK/ERK pathway in cell growth,malignant transformation and drug resistance. Biochim Biophys Acta2007;1773:1263–84.
15. Zitzmann K, Ruden J, Brand S, Goke B, Lichtl J, Spottl G, et al.Compensatory activation of Akt in response to mTOR and Raf inhibi-tors - a rationale for dual-targeted therapy approaches in neuroendo-crine tumor disease. Cancer Lett 2010;295:100–9.
16. She QB, Halilovic E, Ye Q, ZhenW, Shirasawa S, Sasazuki T, et al. 4E-BP1 is a key effector of the oncogenic activation of the AKT and ERKsignalingpathways that integrates their function in tumors. Cancer Cell2010;18:39–51.
17. Valentino JD, Elliott VA, Zaytseva YY, Rychahou PG, MustainWC, Wang C, et al. Novel small interfering RNA cotargetingstrategy as treatment for colorectal cancer. Surgery 2012;152:277–85.
Valentino et al.
Clin Cancer Res; 2014 Clinical Cancer ResearchOF10
18. Tannapfel A, Vomschloss S, Karhoff D, Markwarth A, Hengge UR,Wittekind C, et al. BRAF gene mutations are rare events in gastro-enteropancreatic neuroendocrine tumors. Am J Clin Pathol 2005;123:256–60.
19. Akiyama H, Kamiguchi H. Phosphatidylinositol 3-kinase facilitatesmicrotubule-dependent membrane transport for neuronal growthcone guidance. J Biol Chem 2010;285:41740–8.
20. Li J, Song J, Cassidy MG, Rychahou P, Starr ME, Liu J, et al. PI3Kp110alpha/Akt signaling negatively regulates secretion of the intestinalpeptide neurotensin through interference of granule transport. MolEndocrinol 2012;26:1380–93.
21. EversBM, TownsendCMJr,Upp JR,AllenE,Hurlbut SC,KimSW, et al.Establishment and characterization of a human carcinoid in nudemiceand effect of various agents on tumor growth. Gastroenterology 1991;101:303–11.
22. Takahashi T, Nau MM, Chiba I, Birrer MJ, Rosenberg RK, Vinocour M,et al. p53: a frequent target for genetic abnormalities in lung cancer.Science 1989;246:491–4.
23. Brandt DW, Pandol SJ, Deftos LJ. Calcium-stimulated parathyroidhormone-like protein secretion: potentiation through a protein kinase-C pathway. Endocrinology 1991;128:2999–3004.
24. Doihara H, Nozawa K, Kojima R, Kawabata-Shoda E, Yokoyama T, ItoH. QGP-1 cells release 5-HT via TRPA1 activation: a model of humanenterochromaffin cells. Mol Cell Biochem 2009;331:239–45.
25. Rychahou PG, Murillo CA, Evers BM. Targeted RNA interference ofPI3K pathway components sensitizes colon cancer cells to TNF-related apoptosis-inducing ligand (TRAIL). Surgery 2005;138:391–7.
26. Li J, O'Connor KL, Cheng X, Mei FC, Uchida T, Townsend CM Jr,et al. Cyclic adenosine 50-monophosphate-stimulated neurotensinsecretion is mediated through Rap1 downstream of both Epac andprotein kinase A signaling pathways. Mol Endocrinol 2007;21:159–71.
27. Li J, Chen LA, Townsend CM Jr, Evers BM. PKD1, PKD2, and theirsubstrate Kidins220 regulate neurotensin secretion in the BON humanendocrine cell line. J Biol Chem 2008;283:2614–21.
28. Carracedo A, Ma L, Teruya-Feldstein J, Rojo F, Salmena L, Alimonti A,et al. Inhibition of mTORC1 leads toMAPK pathway activation througha PI3K-dependent feedback loop in human cancer. J Clin Invest2008;118:3065–74.
29. Pitt SC, Davis R, Kunnimalaiyaan M, Chen H. AKT and PTEN expres-sion in human gastrointestinal carcinoid tumors. Am J Transl Res2009;1:291–9.
30. Voortman J, Lee JH, Killian JK, Suuriniemi M, Wang Y, Lucchi M, et al.Array comparative genomic hybridization-based characterization ofgenetic alterations in pulmonary neuroendocrine tumors. Proc NatlAcad Sci U S A 2010;107:13040–5.
31. Wang L, Ignat A, Axiotis CA. Differential expression of the PTEN tumorsuppressor protein in fetal and adult neuroendocrine tissues andtumors: progressive loss of PTEN expression in poorly differentiatedneuroendocrine neoplasms. Appl Immunohistochem Mol Morphol2002;10:139–46.
32. Pitt SC, Chen H, Kunnimalaiyaan M. Inhibition of phosphatidylinositol3-kinase/Akt signaling suppresses tumor cell proliferation and neuro-endocrine marker expression in GI carcinoid tumors. Ann Surg Oncol2009;16:2936–42.
33. Pitt SC, Chen H, Kunnimalaiyaan M. Phosphatidylinositol 3-kinase–Akt signaling in pulmonary carcinoid cells. J Am Coll Surg 2009;209:82–8.
34. Shah T, Hochhauser D, Frow R, Quaglia A, Dhillon AP, Caplin ME.Epidermal growth factor receptor expression and activation in neuro-endocrine tumours. J Neuroendocrinol 2006;18:355–60.
35. Modlin IM, Shapiro MD, Kidd M. Carcinoid tumors and fibrosis:an association with no explanation. Am J Gastroenterol 2004;99:2466–78.
PI3K and RAS Inhibitor Treatments for Carcinoid Tumors
Published OnlineFirst January 17, 2014.Clin Cancer Res Joseph D. Valentino, Jing Li, Yekaterina Y. Zaytseva, et al. Neuroendocrine TumorsCotargeting the PI3K and RAS Pathways for the Treatment of
Updated version
10.1158/1078-0432.CCR-13-1897doi:
Access the most recent version of this article at:
Material
Supplementary
http://clincancerres.aacrjournals.org/content/suppl/2014/01/20/1078-0432.CCR-13-1897.DC1Access the most recent supplemental material at:
E-mail alerts related to this article or journal.Sign up to receive free email-alerts
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Permissions
Rightslink site. (CCC)Click on "Request Permissions" which will take you to the Copyright Clearance Center's
.http://clincancerres.aacrjournals.org/content/early/2014/02/11/1078-0432.CCR-13-1897To request permission to re-use all or part of this article, use this link
