Background: Glioma is the most common primary brain tumor in adults with a poor prognosis. As a member of ARF subfamily GTPase, ARL2 plays a key role in regulating the dynamics of microtubules and mitochondrial functions. Recently, ARL2 has been identified as a prognostic and therapeutic target in a variety range of malignant tumors. However, the biological functional role of ARL2 in glioma still remains unknown. The aim of this study was to explore the expression and functional role of ARL2 in glioma. Methods: In this study, we investigated the expression of ARL2 in glioma samples by using RT-PCR, immunohistochemistry and western blot. The correlation between ARL2 expression and the outcomes of glioma patients was evaluated with survival data from TCGA, CGGA and Rembrandt dataset. Lentiviral technique was used for ARL2 overexpression in U87 and U251 cells. CCK8 assay, colony formation assay, wound healing test, transwell invasion assay and in vivo subcutaneous xenograft model were performed to investigated the biological functions of ARL2. Results: ARL2 expression was down-regulated in glioma, and was inversely associated with poor prognosis in glioma patients. Furthermore, exogenous ARL2 overexpression attenuated the growth and colony-formation abilities of glioma cells, as well as their migration and invasive capabilities. Moreover, elevated expression of ARL2 inhibited in vivo tumorigenicity of glioma cells. Mechanistically, ARL2 regulated AXL expression, which was known as an important functional regulator of proliferation and tumorigenicity in glioma cells. Conclusion: Our study suggests that ARL2 inhibits the proliferation, migration and tumorigenicity of glioma cells by regulating the expression of AXL and may conduct as a new prognostic and therapeutic target for glioma. Keywords: ARL2, Glioma, AXL, Tumorgenecity, Brain cancer Background dismal situation, great efforts have been made to find out Glioma is the most common primary brain tumor in adults effective approaches to halt the progression of this aggres- . Although a standard treatment including extensive sur- sive cancer. Besides this, recent studies have showed a gical resection followed by radiation and temozolomide tremendous understanding of the genetic and molecular chemotherapy has been adopted, the outcomes for glioma mechanisms of glioma, leading to a renewed understanding patients are still poor . Median survival of glioblastoma about potential new therapeutic strategies, including onco- multiforme (GBM), the most common and aggressive form genic signal transduction inhibition/targeted therapy, of glioma, is 14–15 months and median progression-free anti-angiogenesis treatment, therapy targeting glioma stem survival (FPS) is approximately 6 months [1, 2]. Due to this cells, and immunotherapy . Small G-proteins also known as the Ras superfamily * Correspondence: email@example.com; firstname.lastname@example.org structurally classified into 5 families: Ras, Rho, Rab, Sar/ Yulin Wang and Peng Cheng contributed equally to this work. Arf, and Ran, which are involved in multiple cell signal- Department of Neurosurgery, The First Hospital of China Medical University, ing pathways and various cellular functions, including 155 Nanjingbei Street, Heping, Shenyang, Liaoning 110001, People’s Republic of China differentiation, proliferation, vesicle transport, nuclear Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Wang et al. BMC Cancer (2018) 18:599 Page 2 of 13 assembly, and regulation of the cytoskeleton [4, 5]. Methods Recent studies have identified Ras mutations in some Patients and samples human carcinomas [6–8]. It has been reported that acti- Twenty-three patient samples were collected at the First vating mutations of KRAS-4B, within the mutated Ras Hospital of China Medical University from February to family, occurs in approximately 21% of all human can- June in 2016, including 20 glioma samples (grade II, 3 cers, and accounts for approximately 90% of pancreatic cases; grade III, 9 cases; grade IV, 8 cases) and 3 cancers, 45% of colon cancers, and 30% of lung cancers, non-tumor brain tissue samples (from partial lobectomy respectively . The mutated forms of KRAS-4B not in patients with epilepsy). Nine glioma tissues (grade only activate their downstream signaling cascades, but II-IV, 3 cases for each grade) and 3 non-tumor brain tis- also interact with each other and subsequently promote sue samples were used for qPCR and western blot. To the proliferation of cancer cells and induce resistance to further confirm the data of qPCR and western blot, IHC standard cancer therapies . staining were performed with these 12 samples and As a member of the ADP-ribosylation factor (ARF) sub- other 11 glioma samples (grade III 6 cases and grade IV family, ADP ribosylation factor-like GTPase 2 (ARL2) is 5 cases). All glioma patients underwent surgical resec- highly conserved and ubiquitously expressed in eukaryotes tion and the histological diagnosis was verified by 2 . Previous studies show that ARL2 regulates neuropathologists according to 2016 World Health microtubule dynamics through the interaction with Organization (WHO) guidelines. All of the samples used tubulin-folding cofactor D (TBC-D), which is required for for this study were primary tumor samples, except 3 re- multiple mitochondrial functions including mitochondrial current samples used for IHC staining. This study was morphology, motility, asymmetric division, and mainten- approved by the Medical Ethics Committee of the First ance of ATP levels [10–12]. Similarly, the trimer Hospital of China Medical University, and written in- consisting of ARL2, tubulin-specific chaperone D and formed consent was obtained from each patient. The beta-tubulin is required for the maintenance of micro- clinical characteristics of 20 glioma patients were tubule network [13–15]. In addition, ARL2 has been listed in Table 1. proved to be a fundamental regulator of farnesylated cargo and mitochondrial fusion [16, 17]. ARL2 is also involved Cell culture in regulating nuclear retention of STAT3 with binder of U87-MG (catalogue number TCHu58) and U251 (cata- ADP-ribosylation factor-like two (BART) [18–20]. Fur- logue number TCHu138) cell lines were obtained from the thermore, ARL2 inhibition induces the apoptosis of neural progenitor cells derived from human embryonic stem cells Table 1 The clinical characteristics of 20 glioma patients . However, the function role of ARL2 in cancer is still Characteristics Number of patients (n = 20) controversial. It has been reported that ARL2 expression Age(years) level modifies cell morphology and influences mitotic and <50 8 (40%) cytokinetic progression in breast cancer . Recent study ≥ 50 12 (60%) demonstrates that ARL2 expression is dramatically ele- vated in hepatocellular carcinoma and might be poten- Gender tially utilizable as a prognostic marker . Similarly, Male 12 (60%) another study reports that ARL2 functions as an oncogene Female 8 (40%) in cervical cancer . Nevertheless, there is a study WHO grade showing that breast tumor cells with increased ARL2 con- II 3 (15%) tent present reduced aggressivity, both in vitro and in vivo III 9 (45%) . Decreased ARL2 expression is associated with the regulation of p53 localization and results in a chemoresis- IV 8 (40%) tant phenotype in breast cancer via a protein phosphatase Tumor size 2A (PP2A) mediated mechanism . Moreover, the path- <4 cm 11 (55%) ophysiologic role of ARL2 in glioma remains unclear. ≥ 4 cm 9 (45%) In this study, we investigated the expression and func- Primary/Recurrent tional role of ARL2 in glioma. We firstly proved that de- Primary 17 (85%) creased ARL2 expression level was clinically correlated to the higher grades and poorer outcomes of glioma patients. Recurrent 3 (15%) Secondly, we found that ARL2 overexpression attenuated IDH1 state the proliferation, clone formation, migration, invasive and IDH1(−) 11 (55%) tumorigenic capabilities of glioma cells by regulating the IDH1(+) 9 (45%) expression of receptor tyrosine kinase AXL. Wang et al. BMC Cancer (2018) 18:599 Page 3 of 13 Chinese Academy of Sciences Cell Bank (Shanghai, China) 95 °C. Endogenous peroxidase activity was blocked with and maintained in high-glucose Dulbecco’s Modified Eagle’s 3% hydrogen peroxide for 10 min and the sections were Medium (DMEM) (Hyclone, SH30022.01) supplemented incubated with normal goat serum to reduce nonspecific with 10% fetal bovine serum (FBS, Hyclone, SV30087), binding for 15 min. Sections were incubated with pri- 100 U/ml of penicillin, and 100 U/ml of streptomycin mary antibody in blocking solution at 4 °C overnight (Hyclone, SV30010) at 37 °C with 5% CO . in a humidified chamber. After washing three times with PBS, sections were incubated with biotinylated RNA isolation and quantitative RT-PCR (qPCR) goat anti-rabbit IgG (SP-9001, ZSGB-BIO) for 15 min Total RNA was isolated from U87, U251 cells and 12 at room temperature. After washing in PBS, 3, 3′-di- clinical samples using TRIzol reagent (Invitrogen), ac- aminobenzidine (DAB) was used for developing. Slides cording to the manufacturer’s protocol. Total RNA was were counterstained with hematoxylin for 3 min. reversely transcribed into cDNA and used for PCR amp- Then the sections were dehydrated and mounted with lification. Real-time PCR were performed in thermal coverslips. German immunohistochemical score (GIS) cycler (Roche LightCycler 480) using TransStart Top was applied to evaluate the expression of ARL2 . Green qPCR SuperMix Assays (Transgen Biotech, Percentage of positive cells was classified as 0 (nega- AQ131). PCR conditions were as follows: 1 cycle of 95 °C tive), 1 (up to 10%), 2 (11–50%), 3 (51–80%), or 4 (> 80% for 30s, followed by 40 cycles of a two-step cycling pro- positive cells), staining intensity was classified as 0 gram (95 °C for 5 s; 60 °C for 30s). The mRNA expression (no staining), 1 (weak), 2 (moderate), or 3 (strong). was normalized to the expression of GAPDH mRNA and The final immunoreactive GIS were defined as the -ΔΔCt calculated by the 2 method. Specific primers for multiplication of both grading results (percentage of ARL2, AXL and GAPDH were: ARL2 forward: GGGA positive cells × staining intensity). The IHC expres- GGACATCGACACCA and reverse: AGGACCGCAGGG sion value of ARL2 and related sample information ACTTCT ; AXL forward: 5-GTTTGGAGCTGTGA were listed in Table 2. TGGA AGGC-3 and reverse: 5-CGCTTCACTCAGGA For immunocytochemistry, 5 × 10 cells were seeded AATCCTCC-3 ; GAPDH forward: GAAGGTGAA into confocal dish per well and incubated at 37 °C with GGTCGGAGTCA and reverse: TTGAGGTCAATGAAG 5% CO for 24 h. Then the cells were fixed with 4% GGGTC , respectively. paraformaldehyde and permeated with 0.3% Triton X-100 for 20 min. After blocking with 5% BSA for 1 h, Protein extraction and western blot analysis primary antibody was added and incubated at 4 °C over- Total proteins from tissue and cells were extracted by night. Following incubation with rhodamine(TRITC)- whole cell lysis buffer (Wanleibio) and quantified using conjugated affinipure goat anti-rabbit IgG (Proteintech, the bicinchoninic acid (BCA) method. 30 μg of protein SA00007–2) and DAPI (BOSTER, AR1176), the samples from each sample was electrophoresed by 12% were detected using fluorescence microscope (OLYM- SDS-PAGE and transferred to PVDF membranes PUS, BX53). (0.45 μm, Millipore). After being blocked with 5% skimmed milk or 5% BSA (used for phosphorylated pro- tein), the PVDF membranes were incubated overnight at ARL2 expression data mining in GEO dataset, TCGA and 4 °C with the primary antibody. Membranes were then Rembrandt dataset washed three times with TBST (5 min each), and incu- ARL2 expression data of TCGA and Rembrandt dataset bated with peroxidase-conjugated affinipure goat and the patients’ survival data of Rembrandt dataset anti-rabbit (1:5000; Proteintech) or anti-mouse (1:10000; were extracted from Project Betastasis (http://betastasis. Proteintech) IgG at 37 °C for 1 h. Protein expression com/). The patients’ survival data of TCGA were down- was visualized with a chemiluminescence ECL kit loaded from GlioVis portal (http://gliovis.bioinfo.cnio.es). (Tanon, 5500). GAPDH served as a loading control, and In addition, GEO datasets (GSE50161, Griesinger dataset; band intensity was quantified using Image J software. GSE4290, Sun dataset) were applied to analyze the expres- sion level of ARL2 in glioma and normal brain [31, 32]. Immunohistochemistry and immunocytochemistry Gene Set Enrichment analysis (GSEA, www.broadinstitu For immunohistochemistry, all samples were fixed in te.org/gsea/index.jsp) was applied to obtain the functional 10% neutral formalin and embedded in paraffin. Sections information on ARL2 as previously described [33, 34]. (4 μm thick) were cut from paraffin blocks and mounted Moreover, the data from Chinese Glioma Genome Atlas on Poly-L-Lysine-coated glass slides. The sections were (CGGA) were used to analyze ARL2 expression and the deparaffinized in xylene and rehydrated in gradient etha- patients’ survival time [35, 36]. The relevant signaling nol. Antigen retrieval was performed in 0.01 mol/L cit- pathways of ARL2 from KEGG and Reactome were ana- rate buffer (pH 6.0) by microwave oven for 15 min at lyzed by pathDIP (http://ophid.utoronto.ca/pathdip/). Wang et al. BMC Cancer (2018) 18:599 Page 4 of 13 Table 2 The ARL2 expression value and information of samples and following 4 h incubation at 37 °C. Then OD values of used for IHC each well were measured by microplate reader (BIO-RAD Sample tissues Sample Age range GIS P value 15033) at the absorbance of 450 nm. Growth curves were (23 cases in total) NO. (years) score (vs. non-tumor) plotted according to the OD value of each well. Non-tumor tissue N1 38–65 12 (3 cases) N2 12 Colony formation assay After transduced with ARL2, AXL overexpression or N3 12 control lentiviruses for 72 h, the cells were collected and Glioma tissues Grade II G0201 26–60 4 0.0074 resuspended as single cells. Cells were seeded into the (3 cases) G0202 9 wells of a six-well plate and incubated at 37 °C with 5% G0203 6 CO . After 2 weeks, the cells were washed with PBS Grade III G0301 32–70 2 0.0001 twice and stained with crystal violet staining solution. (9 cases) G0302 2 The number of colonies (more than 50 cells) was counted under a microscope (Leica, 090–135.001). G0303 1 G0304 3 Wound healing test G0305 8 The cells (5 × 10 per well) were seeded into six-well G0306 2 plates. After 24 h, the cells overspread the bottom and G0307 1 were scratched by a 200 μl pipette tip. PBS was used to G0308 0 wash out cell debris and suspension cells. Fresh serum-free medium was added, and the cells incubated G0309 0 at 37 °C with 5% CO to allow the wound to heal. Pho- Grade IV G0401 37–79 4 0.0001 tographs of the wound were taken at 0 and 24 h at the (8 cases) G0402 2 same position. The percentage of wound closure was G0403 0 measured according to previous reports [37, 38]. In G0404 1 brief, the wound areas were evaluated by image J G0405 2 software and the percentage of wound closure were cal- culated via the formula as follow: (original wound area - G0406 4 actual wound area)/area of the original wound × 100. G0407 4 G0408 1 Transwell invasion assay Transwell chambers with a pore size of 8 μm filter mem- Lentivirus mediated ARL2 and AXL over-expression brane (Corning, 3422) were used to perform invasion Lentiviruses carrying overexpressing ARL2, AXL and assay. 100 μl Matrigel (Corning, 356,234) (diluted with control vectors were purchased from GeneChem serum-free DMEM by 1:8) was plated in transwell cham- (Shanghai, China). The lentivirus transduction was per- ber, and preserved in an incubator. Four hours later, formed according to the protocol provided by the com- 200 μl of serum-free medium containing 10 cells was pany. In brief, after the cells (10 cells/well in 1 ml added into top chambers of transwell inserts, and 750 μl high-glucose DMEM medium supplemented with 10% DMEM containing 10% FBS was added into bottom FBS) were seeded in a 6-well plate for 24 h, 20 μlof chambers. The cells were incubated at 37 °C in 5% CO lentivirus solution (10 IU/mL) were added to each well for 16 h. Matrigel and cells in top chambers were re- and the cells were incubated at 37 °C with 5% CO for moved by cotton swab. After fixation with 4% parafor- 12 h. The medium was replaced with fresh DMEM maldehyde, the cells traversing the membrane were medium containing 10% FBS. After 48 h of transduction, stained with crystal violet staining solution and counted the cells were selected with puromycin (10 μg/mL). under five different high-power microscope fields per Medium was changed every 3 days. Real-time PCR well. The experiment was performed in triplicate. and western blot were performed to assess the trans- fected efficiency. In vivo subcutaneous tumor transplantation All animal procedures were conformed to protocols ap- In vitro cell proliferation assays proved by the Animal Care Committee of China Medical 5×10 cells in 100 μl medium were seeded into 96-well University. For xenograft subcutaneous transplantation, plates per well and incubated at 37 °C with 5% CO for 6-week-old male immune-deficient nude mice (BALB/ 6 days. The cell proliferation was measured at day 0, 2, 4 C-Null) were purchased from Beijing Vital River Labora- and 6 by adding 10 μl CCK8 (DojinDo) into the wells tory Animal Technology Company. Mice were raised in Wang et al. BMC Cancer (2018) 18:599 Page 5 of 13 laminar flow cabinets under specific pathogen free (SPF) Classical) of GBM, compared to non-tumor samples conditions and were fed ad libitum. U87 cells (trans- (Fig. 1f, P < 0.05). Data from Sun (Additional file 2: duced with ARL2 overexpression or control vector) were Figure S2C, P < 0.001) and Griesinger dataset (Additional injected into the back flanks of nude mice at a density of file 2: Figure S2D, P < 0.05) also demonstrated the consist- 10 cells per 0.3 ml as previous described [39, 40]. The ent results. Taken together, these results demonstrated tumor size was measured using a Vernier caliper per that ARL2 expression significantly decreased in glioma. 4 days, and the tumor volume was calculated using the formula: V = (length x width )/2. The mice were ARL2 expression is clinical relevant with the poor sacrificed at day 28 after implantation, and the tumors prognosis of glioma patients were weighed and photographed. Recently, ARL2 has been identified as a potential prog- nosis marker for hepatocellular carcinoma . To Statistical analysis examine whether ARL2 expression is associated with Data are presented as mean ± SD. The number of repli- glioma patient outcomes, we analyzed the data from cates for each experiment is stated in the figure legend. CGGA, Rembrandt database, and TCGA to investigate Statistical differences between and among groups were the clinical relevance of ARL2. The data from CGGA determined by two tailed t-test or one-way analysis of showed decreased ARL2 expression was clinical relevant variance (ANOVA) followed by Dunnett’s post-test, re- to the poor prognosis of glioma patients (Fig. 1g, P = spectively. Statistical analysis was performed by Microsoft 0.0003). Similar results were obtained from Rembrandt, Excel 2013 and Graphpad Prism 6.0, unless mentioned and the patients with a higher ARL2 expression had a fa- otherwise in the figure legend. P < 0.05 was considered as vorable survival (Fig. 1h, P = 0.011). Finally, we verified statistically significant. the result in TCGA. The consistent elevated expression of ARL2 was also associated with prolonged survival Results (Fig. 1i, P = 0.048). ARL2 expression is decreased in glioma The expression of ARL family members in TCGA were ARL2 attenuated the growth and colony-formation abilities studied through GlioVis, and ARL2 was significantly dif- of glioma cells ferentially expressed between glioma and non-tumor To investigate the physiological role of ARL2 in glioma, samples (Additional file 1: Figure S1). To further investi- we first overexpressed ARL2 through transducing lenti- gate ARL2 expression in glioma, we assessed mRNA and viral ARL2 vector in glioma cell lines (U87 and U251), protein levels in a series of clinical glioma specimens and then examined the effects on cell growth. qPCR and and cell lines. Twelve human clinical specimens were Western blotting assay showed that both of ARL2 collected including 9 glioma tissues (grade II-IV, 3 cases mRNA and protein expression level were significantly el- separately) and 3 non-tumor brain tissue samples. Quan- evated at 48 h after transduction (Fig. 2a, P = 0.0099, titative PCR was performed on these specimens. The re- and b, P = 0.0316; Additional file 3: Figure S3A P < sult showed that ARL2 mRNA levels decreased with the 0.0001, and S3B, P = 0.0075). As a result, ARL2 overex- increase in grade of tumor tissues (Fig. 1a, P < 0.01), as pression significantly suppressed the proliferation of well as U87 and U251 glioma cells (Additional file 2: U251 and U87 cells, compared with cells transduced Figure S2A, P < 0.0001). Similarly, ARL2 protein expres- with the control vector (Fig. 2c and d). Moreover, the sion levels were down-regulated in grade IV glioma sam- colony formation assay was performed to examine the ples (Fig. 1b, P < 0.05), as well as U251 and U87 cells foci formation ability of these cells. As expected, foci for- than non-tumor samples (Additional file 2: Figure S2B, mation abilities of U251 or U87 cells infected with lenti- P < 0.001). We then examined ARL2 expression in 20 viral ARL2 overexpression vector were dramatically gliomas tissue samples (grade II, 3 cases; grade III, 9 decreased in comparison with control cells (Fig. 2e and f). cases; grade IV, 8 cases) and 3 non-tumor brain tissue In addition, we measured the cell cycles in U251 cells samples by immunohistochemistry. The data showed transfected with or without ARL2 overexpression vector. that ARL2 protein expression was reduced in high grade The percentage of cells at G0/G1 phase in U251 cells with glioma samples (Fig. 1c, grade III, P < 0.01; grade IV, ARL2 overexpression was increased and the proportion at P < 0.0001). We also collected ARL2 expression data S and G2/M phase was decreased, oppositely (Additional from CGGA, Rembrandt database and TCGA. The re- file 3: Figure S3C, G0/G1 phase P = 0.0014, S phase P = sults confirmed that ARL2 expression level significantly 0.0049, and G2/M phase P = 0.0111). This data indicated decreased in GBM (grade IV) (CGGA, Fig. 1d, P < ARL2 overexpression induced G0/G1 arrest in glioma 0.0001; Rembrandt database, Fig. 1e, P < 0.0001). In cells and inhibited their proliferation. Taken together, TCGA, a decreased ARL2 expression could be observed these results indicated that ARL2 overexpression inhibited in all subtypes (Proneural, Mesenchymal, Neural and the growth and clonogenicity of glioma cells. Wang et al. BMC Cancer (2018) 18:599 Page 6 of 13 Fig. 1 (See legend on next page.) Wang et al. BMC Cancer (2018) 18:599 Page 7 of 13 (See figure on previous page.) Fig. 1 Decreased ARL2 expression is clinically relevant with poor prognosis of glioma patients. a qRT-PCR analyses of ARL2 mRNA in WHO grade II-IV glioma and non-tumor samples (grade II, n = 3; grade III, n = 3; grade IV, n =3, non-tumor n = 3) (non-tumor vs. grade II, P = 0.0489; non-tumor vs. grade III, P = 0.0075; non-tumor vs. grade IV, P = 0.0046; one-way ANOVA). b Western blot analyses of ARL2 protein in WHO grade II-IV glioma and non-tumor samples (grade II, n =3; grade III, n = 3; grade IV, n =3, non-tumor n = 3) (non-tumor vs. grade II, P = 0.0761; non-tumor vs. grade III, P = 0.0512; non-tumor vs. grade IV, P = 0.0033; one-way ANOVA). c Representative immunohistochemistry images and analyses of ARL2 protein in WHO grade II-IV glioma and non-tumor brain samples (grade II, n = 3; grade III, n = 9; grade IV, n = 8; non-tumor n =3). Scale bar, 50 μm. (non-tumor vs. grade II, P = 0.0074; non-tumor vs. grade III, P < 0.0001; non-tumor vs. grade IV, P < 0.0001; one-way ANOVA). d Data from CGGA showed ARL2 mRNA expression decreased in grade IV compared to grade II (grade II, n = 33; grade III, n = 21; grade IV, n = 106) (grade II vs. grade IV, P < 0.0001; grade II vs. grade III, P = 0.8438, one way ANOVA). e, f Data from Rembrandt database (e,non-tumor, n =28; astrocytoma, n = 148; oligodendroglioma, n = 67; GBM, n = 228) (non-tumor vs. Astrocytoma, P = 0.0007; non-tumor vs. oligodendroglioma, P < 0.0001; non-tumor vs. GBM, P < 0.0001; one-way ANOVA) and TCGA (F, normal, n = 11; classical, n = 54; mesenchymal, n = 58; neural, n = 33; proneural, n =57) (normal vs. classical, P < 0.0001; normal vs. mesenchymal, P < 0.0001; normal vs. neural, P = 0.0188; normal vs. proneual,P <0.0001 one-way ANOVA) revealed that ARL2 mRNA expression decreased in glioblastoma, compared with non-tumor brain tissues. g-i Data from CGGA (G, P = 0.0003, low, n = 148; high, n = 147), Rembrandt database (h,low, n = 171; high, n = 158) and TCGA (i,low, n =287; high, n = 238) indicated ARL2 was opposite relevant to the poor prognosis of glioma patients ARL2 inhibited the migration and invasive capabilities of (epidermal growth factor and epidermal growth factor glioma cells. stimulus) (Additional file 4: Figure S4A and S4B). It has Since microtubule network plays a crucial role in the been reported that AXL is closely relevant to EGFR sig- regulation of cell migration and invasion, wounding naling pathway and mediates the resistance to EGFR healing test and Transwell invasion assay were per- inhibition in lung cancer and GBM [42–44]. We then formed to examine whether ARL2 overexpression inhib- explored the relationship among AXL and enriched ited the migration and invasion of glioma cells. As genes in these two datasets separately via STRING shown in Fig. 2g and h, glioma cells with ARL2 overex- (https://string-db.org/). The data also confirmed the pression migrated significantly more slowly than control close relationship between AXL and these enriched cells (Fig. 2g, P < 0.05, and H, P < 0.01). Similar results genes (Additional file 4: Figure S4C and S4D). Moreover, were obtained from Transwell assay, ARL2 overex- pathway analysis through pathDIP (http://ophid.utoron pressed cells exhibited decreased invasive capabilities to.ca/pathdip/) was performed to inquire the relevant (Fig. 2i, P < 0.01, and j, P < 0.01). These data indicated signaling pathways of ARL2 in KEGG and Reactome that ARL2 diminished the migration and invasion abil- (Additional file 4: Figure S4E). It was revealed that ARL2 ities of glioma cells. was relevant to several downstream signaling pathways, including PI3K-Akt, ERK/MAPK and EGFR. These ARL2 suppressed the tumorigenicity of glioma cells in vivo pathways were also downstream targets relevant to AXL To determine whether ARL2 is important to the tumori- [42, 43]. In addition, previous report and our previous genicity of glioma cells in vivo, we injected U87 cells study proved that AXL played a critical role in the func- infected with ARL2 overexpression vector or control tional regulation of glioma cells [29, 45]. Based on these vector into the flank regions of nude mice and measured observations, we further investigated the effect of ARL2 tumor volumes every 4 days. The results demonstrated expression on AXL in glioma cells. Firstly, western blot that the upregulation of ARL2 expression resulted in a was applied to examine AXL expression in glioma cells reduction in subcutaneous growth of U87 glioma cells infected with ARL2 overexpression or control vector. (Fig. 3a). Consistently, after 4 weeks of xenograft trans- The results demonstrated that ARL2 overexpression de- plantation, although ARL2 overexpression didn’t alter creased AXL protein expression (Fig. 4b, P =0.038,and c the tumorigenesis, the mean volume and weight of sub- P = 0.0053). Secondly, immunocytochemistry confirmed cutaneous tumors in ARL2 overexpression group were that ARL2 overexpression attenuated AXL expression in obviously smaller and lighter than the control group U251 cells (Fig. 4d). Thirdly, IHC staining of ARL2 and (Fig. 3b-d). Collectively, these results showed that ARL2 AXL in U87 xenograft were consistent with the results of could suppress glioma tumorigenicity in vivo. western blot and ICC (Fig. 4e). The upregulated ARL2 ex- pression induced a reduced AXL expression. Finally, ARL2 decreased AXL expression in glioma cells phospho-AXL (Tyr702) protein expression in U251 cells Due to the functional role of ARL2 in glioma, we inves- was also inhibited after ARL2 overexpression (Fig. 4f, P = tigated the downstream target of ARL2 via Gene Set 0.0018). We also observed that the expression level of Enrichment Analysis (GSEA). Adhesion dependent cell phospho-ERK decreased in U251 cells with ARL2 overex- spreading signaling pathway were enriched, including 33 pression (Fig. 4g, P < 0.01). In contrast, there was no signifi- genes, such as ILK, ITGA8, and AXL (Fig. 4a). Further- cant change in total ERK, total AKT and phospho-AKT more, another two signaling pathway were enriched expression (Fig. 4f and Additional file 5:FigureS5). Wang et al. BMC Cancer (2018) 18:599 Page 8 of 13 Fig. 2 ARL2 regulated the growth, colony formation, migration and invasion capabilities of glioma cells. a, b The representative western blot images and analyses of ARL2 in U251 (a) and U87 (b) glioma cells transduced with lentiviral ARL2 vector or its control vector (U251, P = 0.0099; U87, P = 0.0316; n =3; t test). c, d In vitro growth assay showed that ARL2 overexpression inhibited the proliferation of U251 (c)and U87(d) (** P < 0.01; ****P < 0.0001; n =5, t test). e, f Colony formation assay revealed that ARL2 overexpression reduces clone formation ability of U251 (e)and U87 (f) cell lines (500 cells/well; U251, P = 0.0004; U87, P = 0.0014; n =3, t test). g, h Woundhealingtest showedthat ARL2 overexpression decreased the migration abilities of U251 (g) and U87 (h) cells (U251, P = 0.0111; U87, P =0.0015; n =3, t test). i, j Transwell assay demonstrated that ARL2 overexpression inhibited the invasion capabilities of U251 (i)and U87 (j) cells. Scale bar, 100 μm. (P < 0.01, n =3, with t test) We further performed qPCR to detect ARL2 and AXL that the expression level of ARL2 was increased signifi- mRNA level in U251 cells transfected with ARL2 overex- cantly after transduction (Additional file 6:FigureS6A, P < pression vector or control vector. The data demonstrated 0.0001). But ARL2 overexpression didn’t lead to significant Wang et al. BMC Cancer (2018) 18:599 Page 9 of 13 Fig. 3 ARL2 inhibits in vivo tumor formation capability of glioma cells. a The size analysis of subcutaneous tumors measured every four days from nude mice transduce with ARL2 overexpression or control vector transduced U87 cells. b, c Images of mice (b) and subcutaneous (c) tumors at 28th day after subcutaneous transplantation of ARL2 overexpression or control vector transduced U87 cells. d The weight of tumor from nude mice at 28th day after transplantation of U87 cells transduced with ARL2 overexpression or control vector decrease in AXL mRNA expression (Additional file 6: Discussion Figure S6B, P = 0.7087). In addition, Ubibrowser (http:// Microtubule network dynamics is crucial to the regula- ubibrowser.ncpsb.org/) were applied to analyze the high tion of physiological processes like cell mitosis and mi- confidence E3 ligases that interacted with AXL. The result gration. As a key regulator of microtubule, ARL2 has showed that STUB1 was one of high confidence E3 ligases been implicated in several malignant tumors, such as that interacted with AXL (Additional file 6: Figure S6C). breast cancer, cervical cancer, and pancreatic cancer Finally, TCGA data were used to investigate whether these [24, 26, 46]. But the pathophysiologic role and ex- genes were coexpressed with ARL2. The result showed pression pattern of ARL2 in cancer is still controver- that STUB1 was positively correlated to ARL2 expression sial [23, 24], and the function of ARL2 in glioma (Additional file 6: Figure S6D). Altogether, these results in- remains unknown. In this study, we identified ARL2 dicated that ARL2 overexpression suppressed the expres- expression pattern and its clinical significance in gli- sion of AXL and the activation of ERK in glioma cells. oma. The downregulation of ARL2 implies the poor prognosis in glioma patients. Furthermore, our results confirmed that ARL2 reduced the growth, clone for- AXL overexpression partially rescued the phenotype mation, migration and invasive abilities of glioma induced by ARL2 overexpression in U251 cells cells, as well as in vivo tumorigenicity. These data in- To explore the physiological role of ARL2-AXL axis in dicate a promising potential role of ARL2 in malig- glioma cells, we evaluated whether AXL overexpression nant glioma treatment. rescue the phenotype induce by ARL2 overexpression in Another novel finding in this study is that AXL ex- glioma cells. Therefore, we transduced ARL2 overex- pression is regulated by ARL2. AXL is a member of the pression U251 cells with lentiviral AXL overexpression TAM (TYRO3, AXL, MER) subfamily of receptor tyro- vector (Fig. 5a and b). As a result, the reduced in vitro sine kinases . Earlier report and our previous works cell growth and clone formation capabilities of U251 have described the function of AXL in regulating cell cells by ARL2 overexpression were partially restored by growth, migration and tumorigenesis of glioma [29, 45]. AXL overexpression (Fig. 5c and d). Consistently, we Our study provides the first evidence for the role of found that their inhibited migration and invasive abilities ARL2 upregulation in modifying AXL expression. To by ARL2 overexpression were also partially rescued by clarify the mechanism that ARL2 reduced the expression AXL overexpression, yet not completely (Fig. 5e and f). of AXL, qPCR was performed to detect ARL2 and AXL Wang et al. BMC Cancer (2018) 18:599 Page 10 of 13 Fig. 4 ARL2 decreases AXL expression in glioma cells. a GSEA analysis showed that the expression of ARL2 is associated with substrate adhesion dependent cell spreading signaling pathway (33 genes were enriched and AXL was involved, http://www.broadinstitute.org/gsea/index.jsp). The normalized enrichment scores (NES) and the p values are shown in the plot. b, c Representative images and analysis of western blot demonstrated that ARL2 overexpression decreased AXL protein expression in U251 (b)and U87 (c) cells. (U251, P = 0.038, n = 3; U87, P = 0.0053; n =4, t test). d Representative images of immunocytochemistry showed that ARL2 overexpression decreased AXL expression. e The IHC staining of U87 xenograft confirmed that AXL expression decreased after ARL2 overexpression. Scale bar = 100 μm. f ARL2 overexpression decreased phospho-AXL protein expression in U251 cells (P =0.0018, n =3, t test). g ARL2 overexpression inhibited phospho-ERK expression (total ERK, P = 0.9975; phospho-ERK, P = 0.0019, n =4, t test) mRNA level in U251 cells after ARL2 overexpression. Rab35, Rac1, Cdc42 and Rnd3 mediate ubiquitin modifi- The result showed that ARL2 overexpression didn’t lead cation [48–50]. Based on these observations, we used to significant decrease in AXL mRNA expression. There- Ubibrowser to analyze the high confidence E3 ligases fore, we concluded that ARL2 might regulate AXL that interacted with AXL. We also applied TCGA data expression through post-transcriptional mechanism. Pre- to explore whether these genes were coexpressed with vious studies showed that Ras family members like ARL2. The result showed that STUB1 was not only one Wang et al. BMC Cancer (2018) 18:599 Page 11 of 13 Fig. 5 AXL overexpression partially restores the phenotype change induced by ARL2 overexpression. a qPCR analyses of AXL in U251 cells transduced with lentiviral ARL2 overexpression vector together with AXL overexpression or control vector. (P < 0.0001, n =3, t test). b Western blot images and analyses of AXL in U251 glioma cells transduced with lentiviral ARL2 overexpression vector together with AXL overexpression or control vector. (P < 0.0001, n =3, t test). c In vitro growth assay showed that the inhibition of U251 proliferation induced by ARL overexpression could be partly rescued by AXL overexpression. (Day 2,P = 0.0055;Day 4,P < 0.0001; Day 6, P < 0.0001; n =6, t test). d Colony formation assay revealed that the inhibition of clone formation capabilities in U251 cells induced by ARL2 overexpression could be partially rescued by AXL overexpression. (1000 cells/well, P < 0.001, n =3, t test). e Wound healing test showed that the inhibition of migration capabilities in U251 cells induced by ARL2 overexpression could be partially rescued by AXL overexpression. (P < 0.01, n =3, t test). f Transwell assay confirmed that AXL overexpression can partially rescued the invasion capability inhibition in U251 cells induced by ARL2 overexpression. Scale bar, 100 μm. (P < 0.01, n =3, t test). g The diagram illustrated that ARL2 up-regulation decreased the capabilities of proliferation, invasion and tumorigenesis via inhibiting the expression of AXL in glioma cells Wang et al. BMC Cancer (2018) 18:599 Page 12 of 13 of high confidence E3 ligases that interacted with AXL, Funding This study was supported by Natural Science Foundation of China (grant no. but also positively correlated to ARL2 expression. Taken 30901781, P. C.) and Liaoning Science and Technology Plan Projects (grant no. together, these results indicate that ubiquitination and 2012225014, P. C., and no. 2012225070, Z. Z. G.). None of the funding bodies degradation may be a possible mechanism how ARL2 had a role in the design of this study and the collection, analysis, and interpretation of data and in the preparation of the manuscript. regulate AXL expression. Further studies are needed to fully elucidate the detail mechanism of ARL2-AXL axis. Availability of data and materials In addition, the restoration of AXL by exogenous ex- The datasets used and/or analyzed during the current study are available pression did not fully rescue the defects in U251 glioma from the corresponding author on reasonable request. cells caused by ARL2 overexpression, which suggested Authors’ contributions that there might be additional molecular downstream YW and PC conceived and designed the study; YW, GG, WC, FS, and YJ targets associated with ARL2 overexpression. performed the experiments and collected data; PC, ZG, and AW were responsible for the analysis and interpretation of data; PC, ZG, and AW drafted the manuscript; ZG and AW supervised the study and revised Conclusion the manuscript. All authors read and approved the final manuscript. In conclusion, this study described the downregulation of ARL2 in clinical glioma samples and its clinical rele- Ethics approval and consent to participate This study was approved by the Medical Ethics Committee of the First Hospital vance to poor prognosis in glioma patients. Secondly, of China Medical University (2017–50-2), and written informed consent this study provided the evidence that elevated ARL2 ex- was obtained from each patient. Animal studies were conducted according to pression in glioma cell lines inhibits the abilities of protocols approved by the Animal Care Committee of China Medical University. proliferation, clone formation, migration and invasion. Competing interests Thirdly, we demonstrated that ARL2 was associated with The authors declare that they have no competing interests. the regulation of tumorigenicity of glioma cells in vivo. Finally, it was proved in this study that ARL2 regulated Publisher’sNote AXL expression and activated phospho-ERK in glioma. Springer Nature remains neutral with regard to jurisdictional claims in Altogether, our data suggest that ARL2 serves as an im- published maps and institutional affiliations. portant suppressor for the proliferation, migration and Author details tumorigenicity of glioma cells by regulating the expres- Department of Neurosurgery, The First Hospital of China Medical University, sion of AXL. Therefore, it may conduct as a new prog- 155 Nanjingbei Street, Heping, Shenyang, Liaoning 110001, People’s Republic nostic and therapeutic target for glioma. Supplementary of China. Department of Immunology, School of Basic Medical Science, China Medical University, Shenyang 110122, Liaoning, China. methods are available in Additional file 7. Received: 2 November 2017 Accepted: 18 May 2018 Additional files References Additional file 1: Figure S1. The mRNA expression level of ARL family 1. Wen PY, Kesari S. Malignant gliomas in adults. N Engl J Med. 2008;359(5): members in GBM and non-tumor samples from TCGA. (TIF 266 kb) 492–507. Additional file 2: Figure S2. ARL2 expression decreased in GBM. 2. Prados MD, Yung WK, Wen PY, Junck L, Cloughesy T, Fink K, Chang S, (TIF 459 kb) Robins HI, Dancey J, Kuhn J. Phase-1 trial of gefitinib and temozolomide in Additional file 3: Figure S3. ARL2 overexpression increased the patients with malignant glioma: a north American brain tumor consortium proportion of cells at G0/G1 phase and decreased the proportion of cells study. Cancer Chemother Pharmacol. 2008;61(6):1059–67. at S and G2/M phase. (TIF 416 kb) 3. Fine HA. New strategies in glioblastoma: exploiting the new biology. Clin Cancer Res. 2015;21(9):1984–8. Additional file 4: Figure S4. The relevant signaling pathway analysis 4. Matozaki T, Nakanishi H, Takai Y. Small G-Protein networks: their crosstalk showed that ARL2 expression was correlated with EGFR and AXL signaling. and signal cascades. Cell Signal. 2000;12(8):515–24. (TIF 1460 kb) 5. Paduch M, Jelen F, Otlewski J. Structure of small G proteins and their Additional file 5: Figure S5. Western blot showed that ARL2 regulators. Acta Biochim Pol. 2001;48(4):829–50. overexpression in U251 cells didn’t affect the expression of total and 6. Zhang F, Cheong JK. The renewed battle against RAS-mutant cancers. phospho-form AKT. (TIF 842 kb) Cellular and molecular life sciences : CMLS. 2016;73(9):1845–58. Additional file 6: Figure S6. ARL2 overexpression didn’t increase the 7. Stephen AG, Esposito D, Bagni RK, McCormick F. Dragging ras back in the expression of AXL mRNA. (TIF 674 kb) ring. Cancer Cell. 2014;25(3):272–81. 8. Cox AD, Fesik SW, Kimmelman AC, Luo J, Der CJ. Drugging the undruggable Additional file 7: Supplementary Methods. (DOCX 17 kb) RAS: mission possible? Nat Rev Drug Discov. 2014;13(11):828–51. 9. Clark J, Moore L, Krasinskas A, Way J, Battey J, Tamkun J, Kahn RA. Selective Abbreviations amplification of additional members of the ADP-ribosylation factor (ARF) ARF: ADP-ribosylation factor; ARL2: ADP ribosylation factor-like GTPase 2; family: cloning of additional human and Drosophila ARF-like genes. Proc BART: Binder of ADP-ribosylation factor-like two; CGGA: Chinese Glioma Natl Acad Sci U S A. 1993;90(19):8952–6. Genome Atlas; FPS: median progression-free survival; GSEA: Gene Set 10. Bhamidipati A, Lewis SA, Cowan NJ. ADP ribosylation factor-like protein 2 Enrichment Analysis; PP2A: Protein phosphatase 2A; STUB1: E3 ubiquitin- (Arl2) regulates the interaction of tubulin-folding cofactor D with native protein ligase CHIP; TCGA: The Cancer Genome Atlas tubulin. J Cell Biol. 2000;149(5):1087–96. 11. Newman LE, Zhou CJ, Mudigonda S, Mattheyses AL, Paradies E, Marobbio Acknowledgements CM, Kahn RA. The ARL2 GTPase is required for mitochondrial morphology, We thank all the members in Dr. Wu AH’s lab for helpful discussion to our study. motility, and maintenance of ATP levels. PLoS One. 2014;9(6):e99270. Wang et al. BMC Cancer (2018) 18:599 Page 13 of 13 12. Chen K, Koe CT, Xing ZB, Tian X, Rossi F, Wang C, Tang Q, Zong W, Hong 33. Cheng W, Zhang C, Ren X, Jiang Y, Han S, Liu Y, Cai J, Li M, Wang K, Liu Y, WJ, Taneja R, et al. Arl2- and Msps-dependent microtubule growth governs et al. Bioinformatic analyses reveal a distinct notch activation induced by asymmetric division. J Cell Biol. 2016;212(6):661–76. STAT3 phosphorylation in the mesenchymal subtype of glioblastoma. 13. Francis JW, Newman LE, Cunningham LA, Kahn RA, Trimer A. Consisting of J Neurosurg. 2017;126(1):249–59. the tubulin-specific chaperone D (TBCD), regulatory GTPase ARL2, and beta- 34. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, tubulin is required for maintaining the microtubule network. J Biol Chem. Paulovich A, Pomeroy SL, Golub TR, Lander ES, et al. Gene set enrichment 2017;292(10):4336–49. analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102(43):15545–50. 14. Nithianantham S, Le S, Seto E, Jia W, Leary J, Corbett KD, Moore JK, Al- 35. Cheng W, Li M, Jiang Y, Zhang C, Cai J, Wang K, Wu A. Association between Bassam J. Tubulin cofactors and Arl2 are cage-like chaperones that regulate small heat shock protein B11 and the prognostic value of MGMT promoter the soluble alphabeta-tubulin pool for microtubule dynamics. eLife. 2015;4: methylation in patients with high-grade glioma. J Neurosurg. 2016;125(1):7–16. e08811. 36. Jiang T, Mao Y, Ma W, Mao Q, You Y, Yang X, Jiang C, Kang C, Li X, Chen L, 15. Zhou C, Cunningham L, Marcus AI, Li Y, Kahn RA. Arl2 and Arl3 regulate et al. CGCG clinical practice guidelines for the management of adult diffuse different microtubule-dependent processes. Mol Biol Cell. 2006;17(5):2476–87. gliomas. Cancer Lett. 2016;375(2):263–73. 16. Newman LE, Schiavon CR, Zhou C, Kahn RA. The abundance of the ARL2 37. Cormier N, Yeo A, Fiorentino E, Paxson J. Optimization of the wound scratch GTPase and its GAP, ELMOD2, at mitochondria are modulated by the assay to detect changes in murine mesenchymal stromal cell migration after fusogenic activity of mitofusins and stressors. PLoS One. 2017;12(4): damage by soluble cigarette smoke extract. J Vis Exp. 2015;106:e53414. e0175164. 38. Chen J, Tang J, Chen W, Gao Y, He Y, Zhang Q, Ran Q, Cao F, Yao S. Effects 17. Ismail SA, Chen YX, Rusinova A, Chandra A, Bierbaum M, Gremer L, Triola G, of syndecan-1 on the expression of syntenin and the migration of U251 Waldmann H, Bastiaens PI, Wittinghofer A. Arl2-GTP and Arl3-GTP regulate a glioma cells. Oncol Lett. 2017;14(6):7217–24. GDI-like transport system for farnesylated cargo. Nat Chem Biol. 2011;7(12): 39. Wu WS, Chien CC, Liu KH, Chen YC, Chiu WT. Evodiamine prevents glioma 942–9. growth, induces glioblastoma cell apoptosis and cell cycle arrest through 18. Muromoto R, Sekine Y, Imoto S, Ikeda O, Okayama T, Sato N, Matsuda T. JNK activation. Am J Chin Med. 2017;45(4):879–99. BART is essential for nuclear retention of STAT3. Int Immunol. 2008;20(3): 40. Xu X, Cai N, Zhi T, Bao Z, Wang D, Liu Y, Jiang K, Fan L, Ji J, Liu N. MicroRNA- 395–403. 1179 inhibits glioblastoma cell proliferation and cell cycle progression via 19. Zhang T, Li S, Zhang Y, Zhong C, Lai Z, Ding J. Crystal structure of the ARL2- directly targeting E2F transcription factor 5. Am J Cancer Res. 2017;7(8):1680–92. GTP-BART complex reveals a novel recognition and binding mode of small 41. Naito S, von Eschenbach AC, Giavazzi R, Fidler IJ. Growth and metastasis of GTPase with effector. Structure. 2009;17(4):602–10. tumor cells isolated from a human renal cell carcinoma implanted into 20. Bailey LK, Campbell LJ, Evetts KA, Littlefield K, Rajendra E, Nietlispach D, Owen D, different organs of nude mice. Cancer Res. 1986;46(8):4109–15. Mott HR. The structure of binder of Arl2 (BART) reveals a novel G protein 42. Guo G, Gong K, Ali S, Ali N, Shallwani S, Hatanpaa KJ, Pan E, Mickey B, Burma S, binding domain: implications for function. J Biol Chem. 2009;284(2):992–9. Wang DH, et al. A TNF-JNK-Axl-ERK signaling axis mediates primary resistance 21. Zhou Y, Jiang H, Gu J, Tang Y, Shen N, Jin Y. MicroRNA-195 targets ADP- to EGFR inhibition in glioblastoma. Nat Neurosci. 2017;20(8):1074–84. ribosylation factor-like protein 2 to induce apoptosis in human embryonic 43. Zhang Z, Lee JC, Lin L, Olivas V, Au V, LaFramboise T, Abdel-Rahman M, stem cell-derived neural progenitor cells. Cell Death Dis. 2013;4:e695. Wang X, Levine AD, Rho JK, et al. Activation of the AXL kinase causes 22. Beghin A, Honore S, Messana C, Matera EL, Aim J, Burlinchon S, Braguer D, resistance to EGFR-targeted therapy in lung cancer. Nat Genet. 2012; Dumontet C. ADP ribosylation factor like 2 (Arl2) protein influences microtubule 44(8):852–60. dynamics in breast cancer cells. Exp Cell Res. 2007;313(3):473–85. 44. Brand TM, Iida M, Corrigan KL, Braverman CM, Coan JP, Flanigan BG, Stein 23. Hass HG, Vogel U, Scheurlen M, Jobst J. Gene-expression analysis identifies AP, Salgia R, Rolff J, Kimple RJ, et al. The receptor tyrosine kinase AXL specific patterns of dysregulated molecular pathways and genetic subgroups mediates nuclear translocation of the epidermal growth factor receptor. Sci of human hepatocellular carcinoma. Anticancer Res. 2016;36(10):5087–95. Signal. 2017;10(460):eaag1064. 24. Peng R, Men J, Ma R, Wang Q, Wang Y, Sun Y, Ren J. miR-214 down-regulates 45. Vajkoczy P, Knyazev P, Kunkel A, Capelle HH, Behrndt S, von Tengg-Kobligk ARL2 and suppresses growth and invasion of cervical cancer cells. Biochem H, Kiessling F, Eichelsbacher U, Essig M, Read TA, et al. Dominant-negative Biophys Res Commun. 2017;484(3):623–30. inhibition of the Axl receptor tyrosine kinase suppresses brain tumor cell 25. Beghin A, Belin S, Hage-Sleiman R, Brunet Manquat S, Goddard S, Tabone E, growth and invasion and prolongs survival. Proc Natl Acad Sci U S A. 2006; Jordheim LP, Treilleux I, Poupon MF, Diaz JJ, et al. ADP ribosylation factor 103(15):5799–804. like 2 (Arl2) regulates breast tumor aggressivity in immunodeficient mice. 46. Taniuchi K, Nishimori I, Hollingsworth MA. Intracellular CD24 inhibits cell PLoS One. 2009;4(10):e7478. invasion by posttranscriptional regulation of BART through interaction with 26. Beghin A, Matera EL, Brunet-Manquat S, Dumontet C. Expression of Arl2 is G3BP. Cancer Res. 2011;71(3):895–905. associated with p53 localization and chemosensitivity in a breast cancer cell 47. O'Bryan JP, Frye RA, Cogswell PC, Neubauer A, Kitch B, Prokop C, Espinosa R line. Cell Cycle. 2008;7(19):3074–82. 3rd, Le Beau MM, Earp HS, Liu ET. Axl, a transforming gene isolated from 27. Wang K, Li P, Dong Y, Cai X, Hou D, Guo J, Yin Y, Zhang Y, Li J, Liang H, primary human myeloid leukemia cells, encodes a novel receptor tyrosine et al. A microarray-based approach identifies ADP ribosylation factor-like kinase. Mol Cell Biol. 1991;11(10):5016–31. protein 2 as a target of microRNA-16. J Biol Chem. 2011;286(11):9468–76. 48. Minowa-Nozawa A, Nozawa T, Okamoto-Furuta K, Kohda H, Nakagawa I. 28. Gioia R, Leroy C, Drullion C, Lagarde V, Etienne G, Dulucq S, Lippert E, Roche S, Rab35 GTPase recruits NPD52 to autophagy targets. EMBO J. 2017;36(18):2790–807. Mahon FX, Pasquet JM. Quantitative phosphoproteomics revealed interplay 49. Ma J, Xue Y, Liu W, Yue C, Bi F, Xu J, Zhang J, Li Y, Zhong C, Chen Y. Role of between Syk and Lyn in the resistance to nilotinib in chronic myeloid activated Rac1/Cdc42 in mediating endothelial cell proliferation and tumor leukemia cells. Blood. 2011;118(8):2211–21. angiogenesis in breast cancer. PLoS One. 2013;8(6):e66275. 29. Cheng P, Phillips E, Kim SH, Taylor D, Hielscher T, Puccio L, Hjelmeland AB, 50. Liu B, Dong H, Lin X, Yang X, Yue X, Yang J, Li Y, Wu L, Zhu X, Zhang S, et al. Lichter P, Nakano I, Goidts V. Kinome-wide shRNA screen identifies the RND3 promotes snail 1 protein degradation and inhibits glioblastoma cell receptor tyrosine kinase AXL as a key regulator for mesenchymal glioblastoma migration and invasion. Oncotarget. 2016;7(50):82411–23. stem-like cells. Stem cell reports. 2015;4(5):899–913. 30. Cheng P, Wang J, Waghmare I, Sartini S, Coviello V, Zhang Z, Kim SH, Mohyeldin A, Pavlyukov MS, Minata M, et al. FOXD1-ALDH1A3 signaling is a determinant for the self-renewal and Tumorigenicity of mesenchymal glioma stem cells. Cancer Res. 2016;76(24):7219–30. 31. Sun L, Hui AM, Su Q, Vortmeyer A, Kotliarov Y, Pastorino S, Passaniti A, Menon J, Walling J, Bailey R, et al. Neuronal and glioma-derived stem cell factor induces angiogenesis within the brain. Cancer Cell. 2006;9(4):287–300. 32. Griesinger AM, Birks DK, Donson AM, Amani V, Hoffman LM, Waziri A, Wang M, Handler MH, Foreman NK. Characterization of distinct immunophenotypes across pediatric brain tumor types. J Immunol. 2013;191(9):4880–8.
BMC Cancer – Springer Journals
Published: May 29, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera