Original Article ERBB3, IGF1R, and TGFBR2 Expression Correlate With

Am J Cancer Res 2018;8(5):792-809 www.ajcr.us /ISSN:2156-6976/ajcr0077452

Original Article ERBB3, IGF1R, and TGFBR2 expression correlate with PDGFR expression in glioblastoma and participate in PDGFR inhibitor resistance of glioblastoma cells

Kang Song1,2*, Ye Yuan1,2*, Yong Lin1,2, Yan-Xia Wang1,2, Jie Zhou1,2, Qu-Jing Gai1,2, Lin Zhang1,2, Min Mao1,2, Xiao-Xue Yao1,2, Yan Qin1,2, Hui-Min Lu1,2, Xiang Zhang1,2, You-Hong Cui1,2, Xiu-Wu Bian1,2, Xia Zhang1,2, Yan Wang1,2

1Department of Pathology, Institute of Pathology and Southwest Cancer Center, Southwest Hospital, Third Military Medical University, Chongqing 400038, China; 2Key Laboratory of Tumor Immunology and Pathology of Ministry of Education, Chongqing 400038, China. *Equal contributors. Received April 6, 2018; Accepted April 9, 2018; Epub May 1, 2018; Published May 15, 2018

Abstract: Glioma, the most prevalent malignancy in brain, is classified into four grades (I, II, III, and IV), and grade IV glioma is also known as glioblastoma multiforme (GBM). Aberrant activation of receptor tyrosine kinases (RTKs), including platelet-derived growth factor receptor (PDGFR), are frequently observed in glioma. Accumulating evi- dence suggests that PDGFR plays critical roles during glioma development and progression and is a promising drug target for GBM therapy. However, PDGFR inhibitor (PDGFRi) has failed in clinical trials, at least partially, due to the activation of other RTKs, which compensates for PDGFR inhibition and renders tumor cells resistance to PDGFRi. Therefore, identifying the RTKs responsible for PDGFRi resistance might provide new therapeutic targets to synergetically enhance the efficacy of PDGFRi. In this study, we analyzed the TCGA glioma database and found that the mRNA expressions of three RTKs, i.e. ERBB3, IGF1R, and TGFBR2, were positively correlated with that of PDGFR. Co-immunoprecipitation assay indicated novel interactions between the three RTKs and PDGFR in GBM cells. Moreover, concurrent expression of PDGFR with ERBB3, IGF1R, or TGFBR2 in GBM cells attenuated the toxicity of PDGFRi and maintained the activation of PDGFR downstream targets under the existence of PDGFRi. Thus, ERBB3, IGF1R, and TGFBR2 might participate in PDGFRi resistance of GBM cells. Consistent with this notion, combination of PDGFRi with inhibitor targeting either ERBB3 or IGF1R more potently suppressed the growth of GBM cells than each inhibitor alone. The positive correlations of PDGFR with ERBB3, IGF1R, and TGFBR2 were further confirmed in 66 GBM patient samples. Intriguingly, survival analysis showed that ERBB3 predicted poor prognosis in GBM patients with high PDGFRA expression. Altogether, our work herein suggested that ERBB3, IGF1R, and TGFBR2 were responsible for PDGFRi resistance and revealed that ERBB3 acted as potential prognostic marker and therapeutic target for GBM with high PDGFRA expression.

Keywords: Glioblastoma, resistance, PDGFR, ERBB3, IGF1R, TGFBR2

Introduction 3]. The Cancer Genome Atlas (TCGA) project has unveiled critical genetic alterations in glio- Glioma is a prevalent malignancy in brain and ma [4] and provides important rationales for pathologically classified into four grades, i.e. I, target therapies. In most of GBM cases, genetic II, III, and IV, according to the 2016 World profiling reveals aberrant activation of signaling Health Organization (WHO) classification of pathways mediated by receptor tyrosine kinas- central nervous system tumors [1]. Grade IV es (RTKs) [4-6], including the platelet-derived glioma is the most malignant form of glioma growth factor receptor (PDGFR) [5, 6]. PDGFR is and also known as glioblastoma multiforme a transmembrane receptor with 5 immunoglob- (GBM) [1]. Despite the progression of surgical ulin-like repeats in the extracellular domain and and pharmacological therapies, GBM is still an a tyrosine kinase domain in the intracellular intractable disease and the average survival domain. Two PDGFR members have been iden- time of GBM patients is only about one year [2, tified: PDGF receptor α (PDGFRA) and PDGF RTKs involved in resistance to PDGFR inhibitor

Table 1. Clinicopathological information of tion of AXL (AXL Receptor Tyrosine Kinase) patients causes resistance to EGFR-targeted therapy in Clinical Feature Sample Amount lung cancer [23]. Therefore, the activation of alternative RTKs might be also responsible for Grade IV 66 the failure of PDGFRi and identifying these Gender Male 41 alternatively RTKs could provide new therapeu- Female 25 tic targets to synergetically enhance the inhibi- Age ≤ 50 31 tory effect of PDGFRi. > 50 35 Total 66 So far, few RTKs have been reported to participate in the development of PDGFRi resistance in GBM cells, but TCGA database provides us a receptor β (PDGFRB) [7]. Upon the binding of powerful tool to systematically evaluate the PDGF ligand, PDGFRA and PDGFRB form homo- relationships of PDGFR with other RTKs. In this or hetero-dimer and undergo autophosphoryla- study, we hypothesized that RTKs concurrently tion to activate downstream targets, inclu- expressed with PDGFR might contribute to ding PI3K (phosphatidylinositol-4,5-bisphos- PDGFRi resistance with high probability. The phate 3-kinase)/AKT (protein kinase B) and analyses on TCGA glioma database and a MAPK (mitogen-activated protein kinases)/ cohort containing 66 GBM patients revealed ERK1/2 (extracellular signal-regulated kinases tight relationships of PDGFR with three RTKs, i.e. ERBB3, IGF1R, and TGFBR2. Further cellu- 1/2), which results in cell proliferation, survival, lar experiments indicated that the three RTKs migration, and oncogenesis [8]. Remarkably, were indeed involved in PDGFRi resistance and PDGFRA overexpression has been widely de- the combination of PDGFRi with inhibitor tar- tected in all stages of glioma, and the activa- geting either ERBB3 or IGF1R might represent tion of PDGF/PDGFR signaling pathway is pivot- a novel therapeutic strategy to treat GBM ally involved in the initiation and progression of patients. glioma [9, 10]. Materials and methods The critical involvement of PDGFR in glioma makes PDGFR inhibitor (PDGFRi) promising Patient samples drug to treat glioma, especially PDGFR-dependent GBM. So far, several anti-tumor agents 66 GBM samples were obtained from patients targeting PDGFR have been developed, such as diagnosed and treated in Southwest Hospital Imatinib (Gleevec®), Sorafenib (Nexavar®), (Chongqing, China) (Table 1). All patients under- Nilotinib (Tasigna®), and Sunitinib (Sutent®). went surgical resection from January 2014 Although the data from in vitro and animal through December 2016. Specimens were experiments support the potent inhibitory fixed in 4% buffered formaldehyde solution effects of PDGFRi on GBM cells [11, 12], clini- after surgical removal, and then paraffin- cal trials of single PDGFRi have failed to show embedded. Pathohistological diagnoses were encouraging anti-tumor effects [13-16], which independently made by two neuropathologi- might result from the rapid emergence of resis- sts according to “The 2016 World Health tance to PDGFRi [17-19]. Multiple mechanisms Organization Classification of Tumors of the on resistance to RTK-targeted therapy have Central Nervous System” [1]. This study was been identified, including mutation of the active carried out according to the principles of the site, amplification of the targeted RTK, and acti- Helsinki Declaration and all protocols have been approved by the ethics committee of vation of alternative RTKs [18-20]. Indeed, co- Southwest Hospital, Third Military Medical activation of alternative RTKs renders tumor University (TMMU). cells resistance to inhibitor targeting original RTK [21]. Additionally, activation of c-MET (MET TCGA glioma database Proto-Oncogene, Receptor Tyrosine Kinase) and ERBB3 (Erb-B2 Receptor Tyrosine Kinase A group of 669 patient specimens from TCGA_ 3) leads to tolerance of lung cancer cells to GBMLGG database and 538 patient specimens Gefitinib, an inhibitor targeting EGFR (Epidermal from TCGA_GBM database (http://gliovis.bioin- Growth Factor Receptor) [22]. Similarly, activa- fo.cnio.es/) were utilized to evaluate the expres-

793 Am J Cancer Res 2018;8(5):792-809 RTKs involved in resistance to PDGFR inhibitor sion correlation, the clinical relevance, and the and recorded by fluoroanalyzer with OD 450 nm Gene Set Enrichment assay of RTKs. at day 3.

Cell culture Cell migration and invasion

LN229 and 293FT cell lines were purchased The upper chamber (Millicell 8.0 μm PET) was from ATCC (VA, USA), and primary GBM cells coated with 15 µl of Matrigel (CORNING, (GBM-1) were established from the tumor spec- 354234, Bedford, MA 01730, USA) for invasion imen of GBM patient treated in Southwest assay, and not coated for migration assay. Hospital (TMMU, Chongqing, China) [24]. All 3×104 cells in 200 µl of serum-free medium cells were grown in Dulbecco’s modified Eagle’s were seeded into the upper chamber and the medium (Gibco, NY14072, USA) supplemented lower chamber was added with 500 µl growth with 10% (v/v) fetal bovine serum (Gibco, medium supplemented with 10% FBS. After

10270-106, EU-Approved) and 1% (v/v) incubation at 37°C with 5% CO2 for 24 hr for Penicillin-Streptomycin (HyClone, SV30010, invasion assay and 12 hr for migration assay, Austria) at 37°C in a humidified incubator with the cells were fixed with 4% paraformaldehyde

5% CO2. (PFA) followed by crystal violet staining. Non- invading cells were removed with a cotton Plasmid constructs, lentivirus production, and swab, and the stained cells were counted under transient transfection a light microscope (Olympus Corporation, Tokyo, Japan). Full length cDNAs of PDGFRA and PDGFRB were cloned in pCDH-CMV-MCS-EF1-Puro and Co-immunoprecipitation pCDH-EF1-MCS-IRES-Neo lentivirus vectors, respectively (System Biosciences, CA, USA) GBM cells overexpressing PDGFRA-Flag wi- using NheI and NotI sites (New ENGLAND th IGF1R-HA, PDGFRA-Flag with ERBB3-HA, BIOLABS, 6387681, US). The sequences of PDGFRB-Myc with TGFBR2-HA, or PDGFRB-Myc Flag/Myc were also constructed into lentivirus with FGFR1-Flag were collected in cold PBS vectors as empty vector control. The constructs using cell scraper (Corning, NY, USA) and lysed were verified by sequencing (Invitrogen, Shang- with NP40 lysis buffer (Beyotime, P0013F, hai, China). For lentivirus production, 293FT China) with PMSF (Beyotime, ST505, China), cells were co-transfected with the targeted Protease/Phosphatase Inhibitor Cocktail (CST, plasmids and package plasmids (Addgene, MA, 5872S, USA), in ice for 60 min. Then, the lysates USA) with Lipofectamine-2000 (Thermo Fisher were centrifuged at 25000×g for 15 minutes at Scientific, IL, US). Virus-containing superna- 4°C to remove cell debris. Quantitative lysates tants were collected after 48-72 hr and glioma were immediately pre-cleared with Protein A/G cells were infected with the purified superna- magnetic beads (GE Healthcare, Beijing, China) tants and polybrene 8 µg/ml (Sigma, MA, USA). for 1 hr. For affinity enrichment, pre-cleared The stable cells were selected in 4 µg/ml puro- lysates were incubated with beads and anti- mycin (Invivogen, CA, USA) or 130 µg/ml G418 bodies at 4°C overnight. The treated beads (Sangon Biotech, A100859, China) and con- were washed three times with 500 µl of cold firmed by western blot assay. lysis buffer, and then heated in SDS-PAGE sample buffer for 5 min at 95°C. Cell proliferation and IC50 measurement Western blotting Cells were counted and seeded into 96-wells plates at a density of 2000 cells/well with four Cells were harvested by centrifuged at 1000×g replicate wells per group. After 6-8 hours, cells for 5 min at 4°C, then resuspended by RIPA were changed by new medium with the gradient buffer with PMSF, Protease/Phosphatase In- RTK inhibitors as needed. Cell proliferation was hibitor Cocktail for 30 min in cold ice, centri- measured by assay of Cell Counting Kit-8 fuged at 25000×g for 15 minutes at 4°C. The reagent (CCK-8) (DOJINDO, Japan) and record- supernatant was measured by BCA (Bio-Rad, ed by fluoroanalyzer (Floskan Ascent) with OD CA, USA) and 50 μg of total protein was ana-

450 nm for continuous 5 or 6 days. IC50 values lyzed by western blot assay as previously of the inhibitors were measured by CCK-8 assay described [25]. The antibodies used in this

794 Am J Cancer Res 2018;8(5):792-809 RTKs involved in resistance to PDGFR inhibitor

Table 2. Correlation of indicated RTKs in 78419, MA, USA). The secondary antibodies TCGA_GBMLGG database were Dako REALTM EnVision HRP RABBIT/ PDGFRA PDGFRB MOUSE (DAKO, 20039641, Denmark). All slides RTK Name R P R P were counterstained with hematoxylin. The staining signal was visualized using Dako PDGFRA -0.127 0.001 REAL™ EnVision™ Detection System. To score PDGFRB -0.127 0.001 the IHC staining data, five images of each sam- EGFR -0.144 0.000 0.129 0.001 ple were taken and the average integrated opti- ERBB2 -0.359 0.000 0.217 0.000 cal density (IOD) was measured by image pro ERBB3 0.368 0.000 -0.293 0.000 plus 6.0 software. ERBB4 0.072 0.063 -0.342 0.000 FGFR1 -0.074 0.056 0.310 0.000 Statistical analysis FGFR2 -0.024 0.542 -0.167 0.000 All experiments were performed at least three FGFR3 -0.245 0.000 0.036 0.350 times with triplicate samples. Data were pre- FGFR4 0.077 0.045 -0.019 0.624 sented as the mean ± SD. The correlation anal- IGF1R 0.452 0.000 -0.191 0.000 ysis between the gene expression of RTKs was IGF2R 0.297 0.000 0.222 0.000 determined by means of SPSS (V16.0) with TGFBR1 -0.009 0.815 0.194 0.000 Pearson’s test. Statistical significance was set TGFBR2 -0.204 0.000 0.448 0.000 at **P < 0.01 and *P < 0.05. Student’s t-test TGFBR3 -0.021 0.588 -0.123 0.001 was used to compare gene expressions in dif- Note: Genes with bold font are RTKs significantly and ferent grades of glioma. The survival curves positively correlated with PDGFRA or PDGFRB. Genes were described through Kaplan-Meier analysis, with bold font and covered by grey color are top two correlated RTKs with PDGFRA or PDGFRB. and P values were calculated by Log-rank test. All graphs in the study were conducted by Graphpad Prism 6.0. study were as follows: anti-Flag (Sigma, MA, USA), anti-Myc (CST, 2276S, USA), anti-HA (CST, Results 3724S, USA), anti-β-Actin (Beyotime, AF0003, Analysis on TCGA glioma database identified China), anti-IGF1R (CST, 14534S, USA), anti- RTKs positively correlated with PDGFR ERBB3 (CST, 12708S, USA), anti-TGFRβ2 (Bioss, bs-0117R, China), anti-FGFR1 (CST, It has been reported that the activation of alter- 9740S, USA), anti-PDGFRα (CST, 3174S, USA), native RTKs contributes to the resistance of anti-phospho-PDGFRα (Tyr754) (CST, 2292S, RTK inhibitors, including PDGFRi [17], and we USA), anti-PDGFRβ (CST, 3169S, USA), anti- hypothesized that RTKs concurrently expressed phospho-PDGFRβ (Tyr751) (CST, 4549S, USA), with PDGFR might contribute to PDGFRi resis- anti-AKT1 (CST, 2938S, USA), anti-phospho- tance with high probability. To identify the RTKs AKT1 (Ser473) (CST, 9018S, USA), anti-S6 highly correlated with PDGFR in glioma, we ana- Ribosomal Protein (CST, 2217S, USA), anti- lyzed the TCGA glioma database and evaluat- phospho-S6 Ribosomal Protein (Ser235/236) ed the expression correlations of PDGFRA or (CST, 4858S, USA), anti-mouse IgG HRP-linked PDGFRB with several RTK family members criti- Antibody (CST, 7076S, USA), and anti-rabbit IgG cally involved in glioma, including EGFR, ERBB2, HRP-linked Antibody (CST, 7074S, USA). ERBB3, ERBB4, FGFR1, FGFR2, FGFR3, FGFR4, Immunohistochemistry (IHC) IGF1R, IGF2R, TGFBR1, TGFBR2, and TGFBR3 [26-32]. Eight RTK genes were found to be sig- All formalin-fixed tissues were embedded in nificantly concurrently expressed with PDGFRA paraffin and sectioned at 4 μm for IHC. The or PDGFRB (Table 2). Among of them, ERBB3 whole process of IHC was performed as previ- and IGF1R were top two PDGFRA-correlated ously described [24]. Primary antibodies includ- RTK genes (Figure 1A and 1B), and TGFRB2 ed: anti-PDGFRα (CST, 3174S, MA, USA), anti- and FGFR1 were top two PDGFRB-correlated PDGFRβ (CST, 3169S, MA, USA), anti-IGF1R RTK genes (Figure 1C and 1D). Thus, we (CST, 14534S, MA, USA), anti-ERBB3 (CST, focused on the four RTKs for the following 12708S, MA, USA), anti-TGFBR2 (Abcam, experiments.

795 Am J Cancer Res 2018;8(5):792-809 RTKs involved in resistance to PDGFR inhibitor

Figure 1. Correlation analysis of indicated genes using TCGA_GBMLGG database. (A-D) Pearson correlation analysis of PDGFRA and IGF1R expressions (A), PDGFRA and ERBB3 expressions (B), PDGFRB and TGFBR2 expressions (C), or PDGFRB and FGFR1 expressions (D).

ERBB3, IGF1R, and TGFBR2 interacted with action between PDGFRB and FGFR1 (Figure PDGFR in GBM cells 2F). Therefore, the data revealed novel interactions between PDGFR and concurrently- Generally, interactions between PDGFR and expressed RTKs, i.e. ERBB3, IGF1R, and alternative RTKs are required for the alterna- TGFBR2. tive RTKs-mediated resistance to PDGFRi treatment [29, 33, 34], which promoted us to exam- ERBB3, IGF1R, and TGFBR2 contributed to the ine whether there were interactions between resistance of GBM cells to PDGFRi PDGFR with ERBB3, IGF1R, TGFBR2, and FGFR1. For this purpose, we constructed plas- Next, we explored whether ERBB3, IGF1R, mids expressing PDGFRA-Flag, PDGFRB-Myc, TGFBR2, and FGFR1 affected the sensitivity ERBB3-HA, IGF1R-HA, TGFBR2-HA, or FGFR1- of GBM cells to Imatinib, a small-molecule

Flag, and transfected them into a GBM cell line PDGFRi. At first, we calculated IC50 of Imatinib (LN229) and a primary human GBM cell line on growth of LN229 cells (16.94 µM) and (GBM-1) [24] to ensure the successful expres- GBM-1 cells (12.3 µM) through CCK-8 assay sion of the plasmids (Figure 2A and 2B). Then, [35] (Figure 3A and 3B). Then, we treated the we performed co-immunoprecipitation (co-IP) GBM cells with Imatinib at concentrations assay and clearly detected the interactions around the IC50 (Figure 3C and 3D) and identi- between PDGFRA and IGF1R (Figure 2C), fied 15 µM as appropriate concentration to PDGFRA and ERBB3 (Figure 2D), and PDGFRB treat GBM cells in the following assays. To eas- and TGFBR2 (Figure 2E), but no obvious inter- ily assess the inhibitory effect of Imatinib on

796 Am J Cancer Res 2018;8(5):792-809 RTKs involved in resistance to PDGFR inhibitor

ment (Figure 3E). Interesting- ly, although Imatinib treatment potently inhibited the growth of GBM-1 cells expressing PDGFRA (RGR = 0.45), IGF1R (RGR = 0.46), or ERBB3 (RGR = 0.4), it just mildly reduced the growth of GBM-1 cells transfected with PDGFRA plus IGF1R (RGR = 0.73) or PDGFRA plus ERBB3 (RGR = 0.79) (Figure 3E). In addition, we observed similar growth curves for GBM-1 cells transfected with PDGFRB, TGFBR2, FGFR1, PDGFRB plus TGFBR2, or PDGFRB plus FGFR1 under vehicle treatment (Figure 3F). With Ima- tinib treatment, GBM-1 cells with PDGFRB plus TGFBR2 seemed insensitive to Imati- nib (RGR = 0.8) compared to GBM-1 cells with other trans- fections (RGR = 0.41~0.64) (Figure 3F). We performed the same set of experiments in LN229 cells, and consistently observed Imatinib tolerance in LN229 cells transfected with PDGFRA plus IGF1R (RGR = 0.88), PDGFRA plus ERBB3 (RGR = 0.82) (Figure 3G), or PDGFRB plus TGFBR2 (RGR = 0.78) (Figure 3H) compared to PDGFRA (RGR = 0.6), IGF1R (RGR = 0.54), ERBB3 (RGR = 0.51), PDGFRB (RGR = 0.33), TGFBR2 (RGR = 0.34), FGFR1 (RGR = 0.47), or PDGFRB plus Figure 2. Co-IP experiment of PDGFR with PDGFR-correlated RTKs. (A and FGFR1 (RGR = 0.53) (Figure B) Western blotting evaluation on transfection efficiency of indicated genes 3G and 3H). Together, these in GBM-1 (A) and LN229 cells (B). Actin (β-actin) is used as loading con- data suggested that ERBB3, trol. (C-F) Co-IP assay between Flag-tagged PDGFRA and HA-tagged IGF1R IGF1R, and TGFBR2, but not (C), Flag-tagged PDGFRA and HA-tagged ERBB3 (D), Myc-tagged PDGFRB and HA-tagged TGFBR2 (E), as well as Myc-tagged PDGFRB and Flag-tagged FGFR1, were potentially res- FGFR1 (F). ponsible for Imatinib resistance in GBM cells.

GBM cells, we used CCK-8 assay to calculate ERBB3, IGF1R, or TGFBR2 maintained ac- the Relative Growth Rate (RGR) of Imatinib ver- tivation of PDGFR downstream targets with sus vehicle (RGR = OD value at day-5 under Imatinib treatment Imatinib treatment/OD value at day-5 under vehicle treatment). In GBM-1 cells, forced To further evaluate the effects of IGF1R, expression of PDGFRA, IGF1R, ERBB3, PDGFRA ERBB3, TGFBR2, and FGFR1 on PDGFR down- with IGF1R, or PDGFRA with ERBB3 produced stream targets in GBM cells, we examined the comparable growth curves under vehicle treat- activation of AKT, as well as S6, a typical AKT

797 Am J Cancer Res 2018;8(5):792-809 RTKs involved in resistance to PDGFR inhibitor

Figure 3. Relative growth rate of GBM cells with or without Imatinib (IMA) treatment. A. Growth rate of GBM cells under treatment of IMA with different concentrations for 96 hr. B. IC50 calculation of IMA in GBM cells. C and D. Growth curves of GBM cells under treatment of IMA with different concentrations from 0 to 4 days. E and F. Growth curves and relative growth rate of GBM-1 cells with exogenous expression of indicated genes under treatment of Vehicle or IMA. G and H. Growth curves and relative growth rate of LN229 cells with exogenous expression of indicated genes under treatment of Vehicle or IMA. RA: PDGFRA; I1R: IGF1R; EB3: ERBB3; RB: PDGFRB; TR2: TGFBR2; FR1: FGFR1. The data are shown as mean ± SD, n = 3 per group. substrate. In GBM-1 cells, transfection of AKTS473 (p-AKT) (Figure 4A, lane 1-lane 3). As PDGFRA, PDGFRA plus IGF1R, or PDGFRA plus expected, Imatinib treatment obviously de- ERBB3 resulted in comparable activation of pressed AKT activation in GBM-1 cells with AKT marked by the level of phosphorylated PDGFRA (Figure 4A, lane 4), but failed to do so

798 Am J Cancer Res 2018;8(5):792-809 RTKs involved in resistance to PDGFR inhibitor

Figure 4. Western blot assay of GBM cells with or without Imatinib (IMA) treatment. A and B. GBM-1 cells with exogenous expression of indicated genes under treatment of Vehicle (-) or IMA (+). C and D. LN229 cells with exogenous expression of indicated genes under treatment of Vehicle (-) or IMA (+). RA: PDGFRA; I1R: IGF1R; EB3: ERBB3; RB: PDGFRB; TR2: TGFBR2; FR1: FGFR1. Actin (β-actin) is used as loading control. in GBM-1 cells with PDGFRA plus IGF1R or maintained the activation of PDGFR down- PDGFRA plus ERBB3 (Figure 4A, lane 5 and stream targets under the existence of PDGFRi. lane 6). Consistently, the phosphorylation of S6 (p-S6) was resistant to Imatinib treatment in Co-expression of PDGFR with ERBB3, IGF1R, GBM-1 cells with PDGFRA plus IGF1R or or TGFBR2 promoted migration and invasion PDGFRA plus ERBB3 (Figure 4A, lane 5 and of GBM cell in vitro lane 6), but reduced by Imatinib in GBM-1 cells with PDGFRA (Figure 4A, lane 4). Unexpectedly, One of the typical characteristics of GBM cells Imatinib treatment seemed ineffective for is highly invasive growth, which is derived from GBM-1 cells with exogenous expression of the increased migration and invasion ability PDGFRB (Figure 4B, lane 4), but co-transfec- [36-38]. Moreover, the development of resis- tion of PDGFRB and TGFBR2 further increased tance is usually accompanied with the in- p-AKT and p-S6, especially under existence of creased mobility and invasibility of GBM cells Imatinib (Figure 4B, lane 5), which was not obvi- [39]. Thereby, we asked whether the PDGFR- ous for the co-transfection of PDGFRB and correlated RTKs could regulated the migration FGFR1 (Figure 4B, lane 6). Supporting the data and invasion of GBM cells. In GBM-1 cells, we in GBM-1 cells, we acquired similar results found that co-expression of PDGFRA with using LN229 cells (Figure 4C and 4D). ERBB3 or IGF1R effectively enhanced the Therefore, the western data indicated that migration and invasion as indicated by cham- ERBB3, IGF1R, and TGFBR2, but not FGFR1, ber assay without Matrigel and with Matrigel,

799 Am J Cancer Res 2018;8(5):792-809 RTKs involved in resistance to PDGFR inhibitor

800 Am J Cancer Res 2018;8(5):792-809 RTKs involved in resistance to PDGFR inhibitor

Figure 5. Migration and invasion assay of GBM-1 cells with exogenous expression of indicated genes. A and B. Migration statistics and representative pictures for GBM-1 cells transfected with indicated genes through Matrigel- free chamber experiment. C and D. Invasion statistics and representative pictures for GBM-1 cells transfected with indicated genes through Matrigel-based chamber experiment. The data are shown as mean ± SD, *P < 0.05 vs. EV, **P < 0.01 vs. EV, n = 3 per group. E. Geneset enrichment assay of PDGFRBhigh/TGFBR2high versus PDGFRBhigh/TGF- BR2low in GO_CELLULAR_EXTRAVASATION geneset. F. Geneset enrichment assay of PDGFRBhigh/TGFBR2high versus PDGFRBhigh/TGFBR2low in ANASTASSIOU_MULTICANCER_INVASIVENESS_SIGNATURE geneset. EV: empty vector; RA: PDGFRA; I1R: IGF1R; EB3: ERBB3; RB: PDGFRB; TR2: TGFBR2; FR1: FGFR1. respectively (Figure 5A and 5B). Additionally, (Figure 6C and 6D) and found that 2 µM co-expression of PDGFRB with TGFBR2 also AZD3463 or 20 µM Sapitinib did not show high significantly promoted the migration and inva- cytotoxicity on cell growth (Figure 6A-D) and sion of GBM-1 cells (Figure 5C and 5D). were appropriate to test combination with Interestingly, we observed that among the Imatinib. Western data showed that Imatinib three combinations, PDGFRB and TGFBR2 (10 µM) treatment eliminated the activation of exhibited the strongest promoting effect on AKT in PDGFRA overexpression cells (Figure migration and invasion. In accordance with this 6E), while overexpression of PDGFRA with observation, GSEA [40, 41] on TCGA_GBM IGF1R or ERBB3 maintained the activation of database further confirmed that concurrent AKT under the existence of Imatinib (Figure upregulation of both PDGFRB and TGFBR2 dra- 6E). As expected, combination of Imatinib (10 matically enriched the genes responsible for µM) with AZD3463 (2 µM) or Sapitinib (20 µM) cell migration and invasion in the context of effectively blocked the activation of AKT (Figure GO_CELLULAR_EXTRAVASATION geneset (Ac- 6E). Growth assay from two GBM cell lines fur- cession GO:0045123) (Figure 5E) and ANAS- ther showed that application of Imatinib, TASSIOU_MULTICANCER_INVASIVENESS_ AZD3463, or Sapitinib alone partially reduced SIGNATURE geneset [42] (Figure 5F). Thus, co- the cell growth, but combination of Imatinib expression of PDGFR with ERBB3, IGF1R, or with AZD3463 or Sapitinib almost complete- TGFBR2 played positive roles for the migration ly inhibited the cell growth (Figure 7A-D). and invasion of GBM cells. Interestingly, analysis on TCGA database suggested that both PDGFRA and ERBB3 were Simultaneous inhibition of PDGFR and IGF1R highly expressed in proneural-subtype GBM or ERBB3 potently dampened the growth of compared with classical- and mesenchymal- GBM cells subtype GBM, but the expression of IGF1R was similar in all three subtypes of GBM (Figure 7E). It has been known that proneural subtype of Through GSEA on TCGA_GBM database and GBM is featured with PDGFRA overexpression using geneset of MAHADEVAN_IMATINIB_ and insensitive to radio-chemotherapy in com- RESISTANCE [44] as background, we found parison with other subtypes [5, 6]. Although that the genes upregulated in Imatinib-resis- PDGFRA is considered as a therapeutic target tant cells were significantly enriched in GBM for proneural GBM, PDGFRAi did not produce samples with PDGFRAhigh/ERBB3high versus benefit for the patients in clinical trials, at least PDGFRAhigh/ERBB3low (Figure 7F), which strong- partially, due to the resistance to PDGFRAi [17- ly supported the participation of ERBB3 in 19, 43]. Our data revealed that the ERBB3 and Imatinib resistance. Therefore, the combina- IGF1R were interacted with PDGFRA and con- tion of inhibitors targeting PDGFR and related- tributed to PDGFRi resistance, and we specu- RTKs was a novel strategy to suppress high- lated that combination of Imatinib with ERBB3 PDGFR GBM cells, and in particular, PDGFRA inhibitor (Sapitinib) or IGF1R inhibitor (AZD- and ERBB3 were hopeful therapeutic targets 3463) would have more potent inhibitory for proneural GBM. effects on the growth of GBM cells than application of each inhibitor alone. To test this The correlations of PDGFR with ERBB3, IGF1R, hypothesis, we examined whether the combina- and TGFBR2 were confirmed in human GBM tion of Imatinib with Sapitinib or AZD3463 tissues could effectively reverse the Imatinib resistance. We first performed growth assays using The cellular analysis showed expression and AZD3463 (Figure 6A and 6B) or Sapitinib function correlations of PDGFRA with ERBB3,

801 Am J Cancer Res 2018;8(5):792-809 RTKs involved in resistance to PDGFR inhibitor

Figure 6. Treatments using AZD- 3463 or Sapitinib in GBM cells.

A. IC50 calculation of AZD3463 in GBM cells transfected with RA + I1R. B. Growth curves of GBM cells transfected with RA + I1R under treatment of AZD3463 with different concentrations from 0 to

4 days. C. IC50 calculation of Sa- pitinib in GBM cells transfected with RA + EB3. D. Growth curves of GBM cells transfected with RA + EB3 under treatment of Sapitinib with different concentrations from 0 to 4 days. E. Western blot assay of GBM-1 cells with exogenous expression of indicated genes under treatment of vehicle (-), single inhibitor (+) or combination of inhibitors (+). Single inhibitor: A3463 2 µM, SAP 20 µM, IMA 15 µM; Combination of inhibitors: A3463 2 µM, SAP 20 µM, IMA 10 µM. RA: PDGFRA; I1R: IGF1R; EB3: ERBB3. Actin (β-actin) is used as loading control.

PDGFRA with IGF1R, and PDGFRB with TGFBR2 in GBM cells, which promoted us to evaluate their expressions in a cohort containing 66 human GBM tissues (Cohort-66) (Tab- le 1) through IHC. For scoring the IHC staining, five images of each sample were taken and the average integrated optical density (IOD) was measured by image pro plus 6.0 software. Immunostaining on successive slides (Figure 8A) clearly showed positive correlations of PDGFRA with ERBB3, PDGFRA with IGF1R, and PDGFRB with TGFBR2 (Figure 8B). To assess the prognostic significance of IGF1R or ERBB3 in the context of high PDGFRA expression and TGFBR2 in the context of high PDGFRB expression, we used quartile of all IOD values as cutoff to define high and low expressions of interested proteins (Figure 9A). Kaplan-Meier survival analysis showed that in the context of high PDGFRA expression (PDGFRAhigh), high ERBB3 expression (ERBB3high) predicted poor survival in com-

802 Am J Cancer Res 2018;8(5):792-809 RTKs involved in resistance to PDGFR inhibitor

Figure 7. Cell growth assay of GBM cells under treatment of inhibitors. A. Cell growth assay of GBM-1 cells with exogenous expression of RA plus I1R under indicated treatment. B. Cell growth assay of GBM-1 cells with exogenous expression of RA plus EB3 under indicated treatment. C. Cell growth assay of LN229 cells with exogenous expression of RA plus I1R under indicated treatment. D. Cell growth assay of LN229 cells with exogenous expression of RA plus EB3 under indicated treatment. Single inhibitor: A3463 2 µM, SAP 20 µM, IMA 15 µM; Combination of inhibitors: A3463 2 µM, SAP 20 µM, IMA 10 µM. The data are shown as mean ± SD, *P < 0.01 vs. control, **P < 0.001 vs. control, n = 3 per group. E. Box-plots showing the level of PDGFRA, ERBB3, and IGF1R in patients with different molecular classifications using TCGA_GBM database. F. Geneset enrichment assay of PDGFRAhigh/ERBB3high versus PDGFRAhigh/ERBB3low in MAHADEVAN_IMATINIB_RESISTANCE geneset using samples in TCGA_GBM database. parison with low ERBB3 expression (ERBB3low) of PDGFR with ERBB3, IGF1R, TGFBR2, and (i.e. PDGFRAhigh/ERBB3high versus PDGFRAhigh/ moreover, emphasized ERBB3 as an indicator ERBB3low, P = 0.015) (Figure 9C). However, of poor prognosis for GBM patients with high no survival correlations were observed for PDGFRA expression. PDGFRAhigh/IGF1Rhigh versus PDGFRAhigh/IGF- 1Rlow (P = 0.8212) (Figure 9B) or PDGFRBhigh/ Discussion TGFBR2high versus PDGFRBhigh/TGFBR2low (P = 0.8477) (Figure 9D). Hence, the analysis on In this work, we reported ERBB3, IGF1R, and human GBM tissues confirmed the correlations TGFBR2 concurrently expressed with PDGFR in

803 Am J Cancer Res 2018;8(5):792-809 RTKs involved in resistance to PDGFR inhibitor

804 Am J Cancer Res 2018;8(5):792-809 RTKs involved in resistance to PDGFR inhibitor

Figure 8. Correlation analysis between the expression of PDGFR with IGF1R, ERBB3, or TGFBR2 in human GBM tissues. A. Representative immunohistochemical images for PDGFRA/IGF1R, PDGFRA/ERBB3, and PDGFRB/TGFBR2 in successive slides of human GBM tissues, respectively. Scale bar = 100 µm (original image) and 50 µm (inset). B. Pearson correlation analysis of indicated proteins in the cohort containing 66 human GBM patients.

Figure 9. Survival analysis of IGF1R, ERBB3, and TGFBR2 in the context of high PDGFR expression in human GBM tissues. (A) Protein expression intensity above and below the quartile of all samples were categorized as high and low, respectively. (B-D) Kaplan-Meier curves of overall survival rate for GBM patients with PDGFRAhigh/IGF1Rhigh versus PDGFRAhigh/IGF1Rlow (B), PDGFRAhigh/ERBB3high versus PDGFRAhigh/ERBB3low (C), and PDGFRBhigh/TGFBR2high versus PDGFRBhigh/TGFBR2low (D). glioma, including GBM, through analysis on tested in gliomas [4-6]. These inhibitors, pre- TCGA database and a cohort of GBM patients. dominantly targeting PDGFR and EGFR, showed Further experiments revealed novel interac- preclinical benefit, but finally failed with insuffi- tions between PDGFRA with ERBB3 or IGF1R, cient therapeutic efficacy. For example, Imatinib and between PDGFRB and TGFBR2. Moreover, showed promising anti-tumor activities in pre- co-expression of PDGFRA with ERBB3, PDGFRA clinical studies but could not lead to survival with IGF1R, or PDGFRB with TGFBR2 not only improvement in patients with recurrent GBM, enhanced the Imatinib tolerance but also pro- partially due to the resistance to the inhibitors moted the migration and invasion in GBM cells. [15, 16]. One of the mechanisms underlying the resistance is the activation of alternative RTKs, It has been well-known that RTKs pivotally mod- and thereby, simultaneous inhibition of the ulate important pathological properties of originally targeted RTK and the alternative tumor cells, such as proliferation, resistance to RTKs could attenuate, even eliminate, the apoptosis, and cell motility. Aberrant activation resistance and suppress the tumor growth. Our of RTKs has been linked to the initiation, main- work herein revealed ERBB3, IGF1R, and tenance, and progression of many different TGFBR2 as alternative RTKs related with tumor types including GBM [45-47]. A wide PDGFR. Moreover, we treated the GBM cells range of inhibitors targeting RTKs, also known using the combination of Imatinib with ERBB3 as tyrosine kinase inhibitors (TKI), has been inhibitor or IGF1R inhibitor, and indeed

805 Am J Cancer Res 2018;8(5):792-809 RTKs involved in resistance to PDGFR inhibitor observed that the combination almost com- ing pathway, but also implied novel combina- pletely repressed the growth of GBM cells, tion targets for PDGFRi-based therapy in GBM, which confirmed that the alternative RTKs were especially in proneural subtype. functionally involved in the PDGFRi resistance. Acknowledgements Actually, ERBB3, IGF1R, and TGFBR2 have also been reported to function in GBM. For example, This work was supported by: The National Key ERBB3 is coordinated with other ERBB mem- Research and Development Program of China bers to promote growth of GBM cells [27], and (2017YFC1309004), and The Basic and App- also plays a role in the resistance to EGFR- lied Fund of First Affiliated Hospital of Third targeted therapy for GBM [48]. IGF1R plays Military Medical University (SWH2016JCZD-08) important roles in the development and pro- to Yan Wang; The National Key Research and gression of GBM [28, 49], and has been found Development Program of China (2016YFC- to participate in the resistance to EGFR and 1201801), and The National Natural Science PDGFR-targeted treatment [29, 30]. In addi- Foundation of China (81372273) to Xia Zhang; tion, TGFBR2 is a mediator for TGF-β-induced The National Key Research and Development signaling pathway in GBM and functions as Program of China (2016YFA0101203) to Xiu- oncogene in glioma cells as well as glioma stem Wu Bian; The National Natural Science Founda- cells [31, 32, 50]. Accordingly, our data linked tion of China (81402080) to Yan-Xia Wang. these RTKs with PDGFR and expanded the net- Disclosure of conflict of interest work of RTKs in GBM. None. Transcriptomic analysis classified GBM into four clinically relevant subtypes: proneural, Address correspondence to: Drs. Yan Wang and Xia neural, classical, and mesenchymal, and each Zhang, Department of Pathology, Institute of Pa- of these subtypes is defined by a specific thology and Southwest Cancer Center, Southwest molecular signature [5]. For example, proneu- Hospital, Third Military Medical University, Chong- ral GBM is featured with overexpression of qing 400038, China. Tel: +86-23-68754431; Fax: PDGFRA and enhanced activation of PDGF/ +86-23-65397004; E-mail: wang_yan1977@hot- PDGFR signaling pathway [5, 6]. Interestingly, mail.com (YW); [email protected] (XZ) proneural GBM is insensitive to radio- and che- mo-therapy in comparison with other subtypes References [5, 6], which implies that PDGFRA is a hopeful drug target in proneural GBM. Our work indicat- [1] Louis DN, Perry A, Reifenberger G, von ed that ERBB3 was tightly correlated and inter- Deimling A, Figarella-Branger D, Cavenee WK, acted with PDGFRA, and moreover, ERBB3 was Ohgaki H, Wiestler OD, Kleihues P and Ellison identified as an indicator for poor survival in DW. The 2016 world health organization clas- GBM patients with high PDGFRA expression. sification of tumors of the central nervous sys- Intriguingly, simultaneous inhibition of PDGFRA tem: a summary. Acta Neuropathol 2016; 131: and ERBB3 showed potent inhibitory effects on 803-820. cell growth, which emphasized that inhibitors [2] Ostrom QT, Gittleman H, de Blank PM, Finlay JL, Gurney JG, McKean-Cowdin R, Stearns DS, targeting PDGFRA and ERBB3 might act as a Wolff JE, Liu M, Wolinsky Y, Kruchko C and novel independent regimen or adjuvant agents Barnholtz-Sloan JS. American brain tumor as- for standard treatment in proneural GBM. sociation adolescent and young adult primary brain and central nervous system tumors diag- Altogether, this study identified candidate RTKs nosed in the United States in 2008-2012. responsible for the development of PDGFRi Neuro Oncol 2016; 18 Suppl 1: i1-i50. resistance via a TCGA database-based analy- [3] Tanaka S, Louis DN, Curry WT, Batchelor TT sis, which we thought might be applicable to and Dietrich J. Diagnostic and therapeutic av- other RTK inhibitors. Clinical data further enues for glioblastoma: no longer a dead end? high high emphasized that PDGFRA /ERBB3 predict- Nat Rev Clin Oncol 2013; 10: 14-26. ed poor prognosis for GBM patients. Moreover, [4] Cancer Genome Atlas Research Network. the identification of PDGFRi-related alternative Comprehensive genomic characterization de- RTKs, i.e. ERBB3, IGF1R, and TGFBR2, not only fines human glioblastoma genes and core provided new insights on PDGF/PDGFR signal- pathways. Nature 2008; 455: 1061-1068.

806 Am J Cancer Res 2018;8(5):792-809 RTKs involved in resistance to PDGFR inhibitor

[5] Verhaak RG, Hoadley KA, Purdom E, Wang V, Qi proves tumor growth delay and survival in irra- Y, Wilkerson MD, Miller CR, Ding L, Golub T, diated mouse models of glioblastoma. Int J Mesirov JP, Alexe G, Lawrence M, O’Kelly M, Radiat Oncol Biol Phys 2006; 64: 263-271. Tamayo P, Weir BA, Gabriel S, Winckler W, [16] Wen PY, Yung WK, Lamborn KR, Dahia PL, Gupta S, Jakkula L, Feiler HS, Hodgson JG, Wang Y, Peng B, Abrey LE, Raizer J, Cloughesy James CD, Sarkaria JN, Brennan C, Kahn A, TF, Fink K, Gilbert M, Chang S, Junck L, Schiff Spellman PT, Wilson RK, Speed TP, Gray JW, D, Lieberman F, Fine HA, Mehta M, Robins HI, Meyerson M, Getz G, Perou CM, Hayes DN; DeAngelis LM, Groves MD, Puduvalli VK, Levin Cancer Genome Atlas Research Network. V, Conrad C, Maher EA, Aldape K, Hayes M, Integrated genomic analysis identifies clinical- Letvak L, Egorin MJ, Capdeville R, Kaplan R, ly relevant subtypes of glioblastoma character- Murgo AJ, Stiles C and Prados MD. Phase I/II ized by abnormalities in PDGFRA, IDH1, EGFR, study of imatinib mesylate for recurrent malig- and NF1. Cancer Cell 2010; 17: 98-110. nant gliomas: North American Brain Tumor [6] Ozawa T, Riester M, Cheng YK, Huse JT, Consortium Study 99-08. Clin Cancer Res Squatrito M, Helmy K, Charles N, Michor F and 2006; 12: 4899-4907. Holland EC. Most human non-GCIMP glioblas- [17] Wilson TR, Fridlyand J, Yan Y, Penuel E, Burton toma subtypes evolve from a common proneu- L, Chan E, Peng J, Lin E, Wang Y, Sosman J, ral-like precursor glioma. Cancer Cell 2014; Ribas A, Li J, Moffat J, Sutherlin DP, Koeppen 26: 288-300. H, Merchant M, Neve R and Settleman J. [7] Fredriksson L, Li H and Eriksson U. The PDGF Widespread potential for growth-factor-driven family: four gene products form five dimeric resistance to anticancer kinase inhibitors. isoforms. Cytokine Growth Factor Rev 2004; Nature 2012; 487: 505-509. 15: 197-204. [18] Mrugala MM, Kesari S, Ramakrishna N and [8] Blume-Jensen P and Hunter T. Oncogenic ki- Wen PY. Therapy for recurrent malignant glio- nase signalling. Nature 2001; 411: 355-365. ma in adults. Expert Rev Anticancer Ther [9] Maher EA, Furnari FB, Bachoo RM, Rowitch 2004; 4: 759-782. DH, Louis DN, Cavenee WK and DePinho RA. [19] Drappatz J, Norden AD and Wen PY. Thera- Malignant glioma: genetics and biology of a peutic strategies for inhibiting invasion in grave matter. Genes Dev 2001; 15: 1311- glioblastoma. Expert Rev Neurother 2009; 9: 1333. 519-534. [10] Hermanson M, Funa K, Hartman M, Claesson- [20] Logue JS and Morrison DK. Complexity in the Welsh L, Heldin CH, Westermark B and Nister signaling network: insights from the use of tar- M. Platelet-derived growth factor and its recep- geted inhibitors in cancer therapy. Genes Dev tors in human glioma tissue: expression of 2012; 26: 641-650. messenger RNA and protein suggests the pres- [21] Stommel JM, Kimmelman AC, Ying H, ence of autocrine and paracrine loops. Cancer Nabioullin R, Ponugoti AH, Wiedemeyer R, Res 1992; 52: 3213-3219. Stegh AH, Bradner JE, Ligon KL, Brennan C, [11] Ranza E, Mazzini G, Facoetti A and Nano R. In- vitro effects of the tyrosine kinase inhibitor Chin L and DePinho RA. Coactivation of recep- imatinib on glioblastoma cell proliferation. J tor tyrosine kinases affects the response of Neurooncol 2010; 96: 349-357. tumor cells to targeted therapies. Science [12] Ren H, Tan X, Dong Y, Giese A, Chou TC, Rainov 2007; 318: 287-290. N and Yang B. Differential effect of imatinib [22] Engelman JA, Zejnullahu K, Mitsudomi T, Song and synergism of combination treatment with Y, Hyland C, Park JO, Lindeman N, Gale CM, chemotherapeutic agents in malignant glioma Zhao X, Christensen J, Kosaka T, Holmes AJ, cells. Basic Clin Pharmacol Toxicol 2009; 104: Rogers AM, Cappuzzo F, Mok T, Lee C, Johnson 241-252. BE, Cantley LC and Janne PA. MET amplifica- [13] De Witt Hamer PC. Small molecule kinase in- tion leads to gefitinib resistance in lung cancer hibitors in glioblastoma: a systematic review of by activating ERBB3 signaling. Science 2007; clinical studies. Neuro Oncol 2010; 12: 304- 316: 1039-1043. 316. [23] Zhang Z, Lee JC, Lin L, Olivas V, Au V, [14] Paulsson J, Lindh MB, Jarvius M, Puputti M, LaFramboise T, Abdel-Rahman M, Wang X, Nister M, Nupponen NN, Paulus W, Soderberg Levine AD, Rho JK, Choi YJ, Choi CM, Kim SW, O, Dresemann G, von Deimling A, Joensuu H, Jang SJ, Park YS, Kim WS, Lee DH, Lee JS, Ostman A and Hasselblatt M. Prognostic but Miller VA, Arcila M, Ladanyi M, Moonsamy P, not predictive role of platelet-derived growth Sawyers C, Boggon TJ, Ma PC, Costa C, Taron factor receptors in patients with recurrent glio- M, Rosell R, Halmos B and Bivona TG. blastoma. Int J Cancer 2011; 128: 1981-1988. Activation of the AXL kinase causes resistance [15] Geng L, Shinohara ET, Kim D, Tan J, Osusky K, to EGFR-targeted therapy in lung cancer. Nat Shyr Y and Hallahan DE. STI571 (Gleevec) im- Genet 2012; 44: 852-860.

807 Am J Cancer Res 2018;8(5):792-809 RTKs involved in resistance to PDGFR inhibitor

[24] Shi Y, Chen C, Zhang X, Liu Q, Xu JL, Zhang HR, [34] Abouantoun TJ and MacDonald TJ. Imatinib Yao XH, Jiang T, He ZC, Ren Y, Cui W, Xu C, Liu blocks migration and invasion of medulloblas- L, Cui YH, Yu SZ, Ping YF and Bian XW. Primate- toma cells by concurrently inhibiting activation specific miR-663 functions as a tumor sup- of platelet-derived growth factor receptor and pressor by targeting PIK3CD and predicts the transactivation of epidermal growth factor re- prognosis of human glioblastoma. Clin Cancer ceptor. Mol Cancer Ther 2009; 8: 1137-1147. Res 2014; 20: 1803-1813. [35] Guo H, Hu G, Yang Q, Zhang P, Kuang W, Zhu X [25] Liu M, Lin Y, Zhang XC, Tan YH, Yao YL, Tan J, and Wu L. Knockdown of long non-coding RNA Zhang X, Cui YH, Liu X, Wang Y and Bian XW. CCAT2 suppressed proliferation and migration Phosphorylated mTOR and YAP serve as prog- of glioma cells. Oncotarget 2016; 7: 81806- nostic markers and therapeutic targets in glio- 81814. mas. Lab Invest 2017; 97: 1354-1363. [36] Furnari FB, Fenton T, Bachoo RM, Mukasa A, [26] Kinsella P, Howley R, Doolan P, Clarke C, Stommel JM, Stegh A, Hahn WC, Ligon KL, Madden SF, Clynes M, Farrell M and Amberger- Louis DN, Brennan C, Chin L, DePinho RA and Murphy V. Characterization and response of Cavenee WK. Malignant astrocytic glioma: ge- newly developed high-grade glioma cultures to netics, biology, and paths to treatment. Genes the tyrosine kinase inhibitors, erlotinib, gefi- Dev 2007; 21: 2683-2710. tinib and imatinib. Exp Cell Res 2012; 318: [37] Xie Q, Mittal S and Berens ME. Targeting adap- 641-652. tive glioblastoma: an overview of proliferation [27] Berezowska S and Schlegel J. Targeting ErbB and invasion. Neuro Oncol 2014; 16: 1575- receptors in high-grade glioma. Curr Pharm 1584. Des 2011; 17: 2468-2487. [38] Lamouille S, Xu J and Derynck R. Molecular [28] Maris C, D’Haene N, Trepant AL, Le Mercier M, mechanisms of epithelial-mesenchymal transi- Sauvage S, Allard J, Rorive S, Demetter P, tion. Nat Rev Mol Cell Biol 2014; 15: 178-196. Decaestecker C and Salmon I. IGF-IR: a new [39] Cloughesy TF, Cavenee WK and Mischel PS. prognostic biomarker for human glioblastoma. Glioblastoma: from molecular pathology to tar- Br J Cancer 2015; 113: 729-737. geted treatment. Annu Rev Pathol 2014; 9: [29] Almiron Bonnin DA, Ran C, Havrda MC, Liu H, 1-25. Hitoshi Y, Zhang Z, Cheng C, Ung M and Israel [40] Mootha VK, Lindgren CM, Eriksson KF, Su- MA. Insulin-mediated signaling facilitates re- bramanian A, Sihag S, Lehar J, Puigserver P, sistance to PDGFR inhibition in proneural hP- Carlsson E, Ridderstrale M, Laurila E, Houstis DGFB-driven gliomas. Mol Cancer Ther 2017; N, Daly MJ, Patterson N, Mesirov JP, Golub 16: 705-716. TR, Tamayo P, Spiegelman B, Lander ES, [30] Ma Y, Tang N, Thompson RC, Mobley BC, Clark Hirschhorn JN, Altshuler D and Groop LC. PGC- SW, Sarkaria JN and Wang J. InsR/IGF1R path- 1alpha-responsive genes involved in oxidative way mediates resistance to EGFR Inhibitors in phosphorylation are coordinately downregu- glioblastoma. Clin Cancer Res 2016; 22: 1767- lated in human diabetes. Nat Genet 2003; 34: 1776. 267-273. [31] Narushima Y, Kozuka-Hata H, Koyama-Nasu R, [41] Subramanian A, Tamayo P, Mootha VK, Tsumoto K, Inoue J, Akiyama T and Oyama M. Mukherjee S, Ebert BL, Gillette MA, Paulovich Integrative network analysis combined with A, Pomeroy SL, Golub TR, Lander ES and quantitative phosphoproteomics reveals trans- Mesirov JP. Gene set enrichment analysis: a forming growth factor-beta receptor type-2 knowledge-based approach for interpreting (TGFBR2) as a novel regulator of glioblastoma genome-wide expression profiles. Proc Natl stem cell properties. Mol Cell Proteomics Acad Sci U S A 2005; 102: 15545-15550. 2016; 15: 1017-1031. [42] Anastassiou D, Rumjantseva V, Cheng W, [32] Joseph JV, Balasubramaniyan V, Walenkamp A Huang J, Canoll PD, Yamashiro DJ and Kandel and Kruyt FA. TGF-beta as a therapeutic target JJ. Human cancer cells express Slug-based in high grade gliomas-promises and challeng- epithelial-mesenchymal transition gene ex- es. Biochem Pharmacol 2013; 85: 478-485. pression signature obtained in vivo. BMC [33] Akhavan D, Pourzia AL, Nourian AA, Williams Cancer 2011; 11: 529. KJ, Nathanson D, Babic I, Villa GR, Tanaka K, [43] Lu F, Chen Y, Zhao C, Wang H, He D, Xu L, Wang Nael A, Yang H, Dang J, Vinters HV, Yong WH, J, He X, Deng Y, Lu EE, Liu X, Verma R, Bu H, Flagg M, Tamanoi F, Sasayama T, James CD, Drissi R, Fouladi M, Stemmer-Rachamimov AO, Kornblum HI, Cloughesy TF, Cavenee WK, Burns D, Xin M, Rubin JB, Bahassi EM, Canoll Bensinger SJ and Mischel PS. De-repression of P, Holland EC and Lu QR. Olig2-dependent re- PDGFRbeta transcription promotes acquired ciprocal shift in PDGF and EGF receptor signal- resistance to EGFR tyrosine kinase inhibitors ing regulates tumor phenotype and mitotic in glioblastoma patients. Cancer Discov 2013; growth in malignant glioma. Cancer Cell 2016; 3: 534-547. 29: 669-683.

808 Am J Cancer Res 2018;8(5):792-809 RTKs involved in resistance to PDGFR inhibitor

[44] Mahadevan D, Cooke L, Riley C, Swart R, [48] Clark PA, Iida M, Treisman DM, Kalluri H, Simons B, Della Croce K, Wisner L, Iorio M, Ezhilan S, Zorniak M, Wheeler DL and Kuo Shakalya K, Garewal H, Nagle R and Bearss D. JS. Activation of multiple ERBB family recep- A novel tyrosine kinase switch is a mechanism tors mediates glioblastoma cancer stem-like of imatinib resistance in gastrointestinal stro- cell resistance to EGFR-targeted inhibition. mal tumors. Oncogene 2007; 26: 3909-3919. Neoplasia 2012; 14: 420-428. [45] Lemmon MA and Schlessinger J. Cell signaling [49] Osuka S, Sampetrean O, Shimizu T, Saga I, by receptor tyrosine kinases. Cell 2010; 141: Onishi N, Sugihara E, Okubo J, Fujita S, Takano 1117-1134. S, Matsumura A and Saya H. IGF1 receptor sig- [46] Casaletto JB and McClatchey AI. Spatial regu- naling regulates adaptive radioprotection in lation of receptor tyrosine kinases in develop- glioma stem cells. Stem Cells 2013; 31: 627- ment and cancer. Nat Rev Cancer 2012; 12: 640. 387-400. [50] Han J, Alvarez-Breckenridge CA, Wang QE and [47] Carrasco-Garcia E, Saceda M and Martinez- Yu J. TGF-beta signaling and its targeting for Lacaci I. Role of receptor tyrosine kinases and glioma treatment. Am J Cancer Res 2015; 5: their ligands in glioblastoma. Cells 2014; 3: 945-955. 199-235.

809 Am J Cancer Res 2018;8(5):792-809