| BACKGROUND:Glioblastoma (GBM) is one of the most lethal diseases in the central nervous system of adults and the median survival time of GBM patients is12months. There are various therapeutic methods for GBM, including surgery, chemotherapy and radiotherapy. However, the median survival time of patients with GBM was only modestly increased to15months. Major limitations of therapies for GBM are tumor recurrence after surgery, tumor infiltration into surrounding tissues, and intrinsic or acquired resistance to chemotherapy and radiotherapy.Although the DNA-methylating agent temozolomide (TMZ) has been developed for treatment of gliomas, several growth factor receptors such as PDGFR and EGFR have been used as therapeutic targets. Treatment with the PDGFR/c-KIT/abl kinase inhibitors dramatically inhibited the viability and anchorage-independent growth of tumor cells. But only10-20%of the patients had a clinical response to these inhibitors, and most of these patients subsequently exhibited rapid tumor progression due to drug resistance. Wilson et al found that inhibition of RTK ligands could reverse both innate and acquired resistance. However, the mechanisms underlying the resistance to RTK inhibitors have not yet been fully elucidated. Imatinib is one of the representative RTK inhibitors. Antagonism of imatinib in glioma models has been demonstrated to successfully inhibit tumor growth both in vitro and in vivo. We constructed an imatinib-resistant GBM cell line U251AR in our previous study and used the two-dimensional difference gel electrophoresis (2D-DIGE) and mass spectrometry (MS)-based proteomic approaches to study the chemoresistance-associated proteins in GBM cells. Proteomics is a powerful and effective tool to evaluate protein profiles.2D-DIGE is a sensitive gel-based method for protein separation and quantification. Proteins are pre-labeled with different fluorescent dyes, mixed, and separated on gels. Proteomics offers the potential ability to find unknown mechanism involved in MDR of cancers and provides new opportunities to find biomarkers and therapeutic targets for tumors.Polymerase I and transcript release factor (PTRF), also known as cavinl, is originally identified as a protein involved in dissociation of transcription complexes in vitro. PTRFs in cell surface are associated with processes of vesicular transport, cholesterol homeostasis, and lipolysis control. PTRF mutations are associated with congenital generalized lipodystrophy in humans. Interactome analyses suggest that PTRF has unknown functions besides the roles described above. Loss of PTRF expression in prostate cancer and lung cancer has been demonstrated to be related with tumor progression. The caveolae structural proteins of PTRF and caveolinl are essential for MDR of breast cancer. PTRF induces formation of abundant caveolae in various cultured cells and in zebrafish embryos. PTRF and caveolinl are closely associated on the plasma membrane. The caveolin proteins have been reported to be located in caveolae and essential for the presence of caveolae. Quann and his colleagues reported that over expression of caveolinl in the GBM cell line U87negatively regulated cell growth and survival pathways. Expression of caveolinl is up-regulated in GBM cell lines and tumors compared to primary human astrocytes and normal brain tissues. The cells resistant to TMZ affect caveolinl expression in vitro and in vivo in human GBM models. However, there is no study on expression of PTRF in GBMs. Thus, in this study, we investigated expression and function of PTRF in GBM cell lines and patients. The role PTRF in chemoresistance of GBM cell lines was also analyzed. Our data indicate that PTRF may be used as valuable targets for developing new therapeutic strategies for GBM patients.OBJECT:METHODS:1. Tissue specimensPatient specimen samples were obtained from Zhujiang and Nanfang Hospital (Southern Medical University, Guangzhou, China). Patients enrolled in this study included8grade I astrocytoma cases,13grade II astrocytoma cases,10grade III astrocytoma cases, and27GBM cases. Among27GBM cases,6GBM cases were relapsed6months after TMZ therapy. All patients gave prior written and informed consent prior to collection of specimens according to institutional guidelines of Southern Medical University. Tissue samples were snap-frozen in the operation room immediately after surgery. Non-tumor tissues were diagnosed by a board-certified neuropathologist. Normal tissues were confirmed to be tissues surrounding tumor and free of cancer cells according to pathologic examination. For each patient, a frozen tumor sample (stored at-80oC) and a paraffin-embedded tissue specimen was available.2. Cell lines and cell cultureHuman GBM cell line U251was obtained as a gift from College of Public Health, Southern Medical University, Guangzhou, China. The MDR cell line U251AR was established and maintained in our laboratory. The cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM/H) containing10%(v/v) fetal bovine serum (FBS), penicillin (200U/mL) and streptomycin (100μg/mL). Cells were cultured at37oC in a humidified incubator with an atmosphere of5%CO2. The U251AR cell line was established by exposing the U251cell line continuously to increasing concentrations of imatinib (STI571) over a period of12months in our lab. To maintain the MDR phenotype of U251AR cells, imatinib was added to the medium at a final concentration of122μg/mL during U251AR cell culture.3. Immunofluorescence A total of3x105cells per chamber were placed into Lab-Tek two-chamber slides and incubated overnight. On the next day, when cells were50-70%confluent, they were washed with PBS twice, fixed in4%paraformaldehyde (Sigma, St. Louis, Missouri, USA) and permeabilized in0.1%Triton X-100(Sigma, St. Louis, Missouri, USA) at4oC for30min. The cells were then washed3times with PBS and incubated with blocking solution (10%horse serum in PBS). After blocking, cells were incubated with primary antibodies against PTRF or caveolinl overnight at4oC. After washing with PBS for three times, cells were incubated with the secondary antibody of goat anti-rabbit-Alexa Fluor488(1:1,000; Molecular Probes, Invitrogen, USA) for1h at room temperature in the dark. Finally, the cells were washed three times with PBS and incubated with0.25mg/ml DAPI (Roche, Mannheim, Germany) for1min at room temperature in the dark. After extensively washing with PBS, samples were imaged on a confocal laser scanning microscope (Olympus Fluoview, Tokyo, Japan) using a60x oil immersion objective, with identical exposure times.4. Protein extractionCell lysates were prepared from U251and U251AR cell lines by mechanical disruption in ice-cold lysis buffer (Tris20mM, pH7.5, CHAPS4%, urea8M (Sigma, St Louis, USA)) and antiproteases cocktail (Complete EDTA-free tablets, Roche Diagnostics, Mannheim, Germany). Samples were sonicated (6cycles of ten seconds with relapse of30seconds in ice-bath) and centrifuged (15000g,30minutes,4oC). Supernatants were ultracentrifuged at108,000g for60minutes at4oC. Protein concentration was determined using the Bradford protein assay and the extracted protein (100μg) was kept at-80℃.5. Protein labeling with cyanin dyesCytosolic extracts were labeled with CyDyes DIGE Fluors (GE Healthcare, Bucks, UK) according to the manufacturer’s recommended protocol. Briefly,50μg of each sample were minimally labeled with400pmol amine-reactive cyanine dyes, Cy3or Cy5, on ice for30minutes, in the dark. U251and U251AR were all labeled with Cy5or Cy3for different gels. An internal pool, labeled with Cy2fluorescent dye, was generated by combining equal amounts of U251and U251AR cells together. The labeling reaction was quenched by incubation with1μL of10mM lysine (Sigma-Aldrich, ST Louis, USA) on ice in the dark for10minutes. Following the labeling reaction, the U251cell extracts and the U251AR cell extracts were combined together with the internal pool, and DestreakTM IEF buffer (GE Healthcare) was added to make the volume up to450μl prior to IEF (isoelectric focalisation) on five24cm gel strips.6. Two-dimensional SDS-PAGEThe isoelectric electrophoresis was carried out using an IPGphorTM system (GE Healthcare). Pre-cast immobilized pH gradient strips (pH3-10NL,24cm) were used for the one-dimensional separation with a total focusing time of60kV-h. After IEF, the IPG strips were incubated two times at ambient temperature for15minutes in an equilibration solution (0.05M Tris-HCl pH8.8,6M Urea,30%glycerol,2%SDS and bromophenol blue) containing65mM DTT and250mM iodoacetamide. Strips were directly applied on top of pre-cast12%SDS-PAGE gels (GE Healthcare) and ran in a vertical Ettan DaltSix system (GE Healthcare) for approximately5hours. Another gel ran in the same way for picking of protein pots. Four gels were processed simultaneously.7. Gel imaging and data analysisAfter SDS-PAGE, cyanine-labeled proteins were directly visualized using a TyphoonTM9400imager (GE Healthcare) in a fluorescence mode. Cy2, Cy3and Cy5images were scanned using488nm,532nm, and633nm laser, respectively. Each gel was scanned at200μm (pixel size) resolution and was processed using the DeCyder software V5.01(GE Healthcare), followed by quantification, gel matching and statistical analyses. To exclude artifacts from gel images and differentially quantify the protein spots in the images, the Differential In-gel Analysis module (DIA) was used for pair-wise comparison of the two samples (U251and U251AR) on each gel. The Biological Variation Analysis module (BVA) was used to match the entire set of protein-spot maps from comparable gels simultaneously. Student’s test (p<0.05) was performed for statistical analyses. Protein spots with at least1.5-fold changes in volume after normalization were defined as differentially regulated. The statistical power of the analysis was calculated similarly to results reported by Engelen K. et al and Karp N. et al. The standard deviation of the log10(standardized abundance) per condition was calculated for each spot that have been matched across the2gels of the analysis. The median of these standard deviations was calculated in each condition to estimate the global variance of the replicates. After2D-DIGE imaging and analysis, another gel was stained with Coomassie-blue. Gels were scanned (Image Scanner TM GE Healthcare) and stored in1%acetic acid at4oC until spot excision. Matching between Coomassie-blue stained gels and fluorescence maps was performed manually and the pick lists were generated using the Image MasterTM2D Elite software (GE Healthcare).8. MALDI-TOF/TOF mass spectrometryCoomassie Blue-stained protein spots were excised from2-D gels and processed using an EttanTM Spot Handling Workstation (GE Healthcare). Gel plugs were washed3times in MilliQ water, followed by a rinse in50%methanol/50mM ammonium bicarbonate and a rinse in75%ACN to ensure complete removal of dye and detergent. After drying, gel pieces were re-hydrated for60minutes in20mM NH4HCO3with16.6μg/ml porcine trypsin (Promega, Charbonnieres-lesbains, France). Extraction was performed in two successive steps by addition of50%ACN and0.1%TFA, respectively. The digestion products were dried out and dissolved in2mg/mL a-cyano-4-hydroxycinnamic acid in70%ACN/0.1%TFA, before spotting onto MALDI targets (600μm384Scout MTP AnchorChipTM; Bruker Daltonics, GmbH, Bremen, Germany). Peptide mass fingerprints were obtained using a MALDI-TOF/TOF mass spectrometer (UltraflexTM; Bruker Daltonics, GmbH) and processed using the FlexAnalysisTM software (version2.2; Bruker Daltonics, GmbH) for generation of peak list and an internal calibration with trypsin auto-digestion peptides. Peak lists were then transferred to ProteinScapeTM software (version1.3; Bruker Daltonics, GmbH) for another automatic calibration based on a calibration list (related to the sample type and treatment) containing autolysis peaks and contaminants (keratins, polymers and background peaks). After re-calibration, an automatic trypsin and contaminants filtering and removal were performed in order to get the m/z ratio and to obtain high identification rates (Score-Booster). Only the monoisotopic masses of tryptic peptides were then used to query NCBInr sequence databases using the Mascot search algorithm (Mascot server version2.1.04; http://www.matrixscience.com). Search conditions were as follows:an initial mass window of70ppm for the internal calibration, only one missed cleavage acceptable, modification of cysteines by iodoacetamide and methionine oxidation as variable modifications. Results were scored using the probability-based Mowse score (the protein score is-10×log (P)). P is the probability that the observed match is a random event. In our experiment, a score greater than90was considered as a significant identification (p<0.05).9. Immunoblot analysesCytosolic protein extracts (10-30μg) were loaded on12%polyacrylamide gels for performing1D-SDS-PAGE. The biotinylated ECL western blotting molecular weight markers (Amersham-GE-Healthcare) were used. The proteins were transferred onto PVDF membranes (Millipore, USA). Equal amount of proteins that were obtained from U251, U251AR and other cells and quantified by Bradford protein assay were loaded on each gel. Non-specific sites were blocked in Tris-buffered saline (TBS) containing5%(w/v) non-fat dry milk and blots were incubated with diluted primary antibodies in0.1%Tween20and1%nonfat dry milk TBS. The primary antibodies included rabbit monoclonal anti-human PTRF (dilution1:1000, Abcam, Cambridge, Massachusetts, USA), rabbit monoclonal anti-human caveolinl (dilution1:1500, Cell Signaling Technology, Danvers, Massachusetts, USA), rabbit monoclonal anti-human VIM (dilution1:1000, Cell Signaling Technology, Danvers, Massachusetts, USA) and goat monoclonal anti-human P-gp (dilution1:200, Santa Cruz Biotechnology, Santa Cruz, California, USA).β-actin was used as an internal control. After washing in TBS, blots were incubated with secondary antibodies of peroxidase-conjugated IgG (dilution1:5000, Santa Cruz, California, USA) and Streptavidin-HRP (for biotinylated markers). The enhanced chemiluminescence system ECL+(GE Healthcare) was used for color development.10. Reverse transcription-quantitative PCR Total RNA was extracted using Trizol reagent (Invitrogen, USA). Total RNA was reversely transcribed using prime Script RT reagent Kit (Takala, Dalian, China). Quantitative RT-PCR was carried out in an MX7500sequence detection system (Stratagene, USA) using SYBR Green according to the manufacturer’s instructions. Primers were listed in Table1. Glyceraldehyde3-phosphate dehydrogenase (GAPDH) was used as an internal control. All samples were normalized to internal controls and fold changes were calculated through relative quantification (2ΔΔCT).11. PTRF knockdownBLOCK-iT Pol II miR RNAi expression vector kit (Invitrogen Co., Carlsbad, California, USA) was used to induce knockdown of PTRF. Briefly, single-stranded miRNAs were annealed to form double-strands, and inserted into the pcDNA6.2-GW/EmGFP-miR vector (Invitrogen, USA). Short hairpin RNA (shRNA) targeting PTRF was named as shPTRF. The empty vector was named as shNC. The nucleotide sequences of the target miRNA and unrelated miRNA were shown in Table2. The plasmids were transfected into U251AR and U251cells by LipofectamineTM2000(Invitrogen, USA) according to the manufacturer’s instructions. After incubation for24h,500ng/mL Blasticidin S HCl (Invitrogen, USA) was added into medium. After transfection, PTRF mRNA level was determined by quantitative RT-PCR. Subsequently, several clones with lower expression levels of PTRF mRNA were analyzed further for their PTRF protein levels by Western blotting. Finally, clones with effective PTRF knockdown were selected for further analyses. Cells transfected with an empty vector were used as a control.12. In vitro drug sensitivity assayCells were placed in96-well plates at a density of2×103per well in a final volume of100μL and transfected with shNC and shPTRF. Cell viability was analyzed after incubation with100μg/mL TMZ for24h,48h,72h,96h and120h. The cell viability assay was performed using a CCK8kit (Dojindo Molecular Technologies, Japan). In drug sensitivity analysis, cells were reseeded in96-well plates24h post-transfection with a density of1.5x104per well and treated with imatinib, VP-16, or TMZ (50to200μg/mL) for48h. 13. Immunohistochemical analysisExpression levels of PTRF and caveolinl in tissues were detected by an ultrasensitive S-P kit (Zhongshan Biotechnology Co. Ltd, Beijing, China) according to the manufacturer’s recommendation. Rabbit monoclonal primary antibodies against human PTRF (dilution,1:100; Abcam, Cambridge, Massachusetts, USA) and caveolinl (dilution,1:150; Cell Signaling Technology, Danvers, Massachusetts, USA) were used. Polyperoxidase rabbit IgG was used as the secondary antibody (Zhongshan Biotechnology Co. Ltd, Beijing, China). Sections were analyzed with bright field microscopy (Olympus BX51, Tokyo, Japan). Negative controls were also detected with the primary antibody. Immunostained sections were examined by light microscopy using x40objective lens and x10eyepieces. Immunostaining intensity (IS) was counted by the Image pro-Plus6.0software.14. Statistical analysisAll experiments were performed in triplicate. The results were given as means±standard deviations (SDs). Statistical analyses were performed using either an analysis of variance (ANOVA) or Student’s t test. The relationship between the PTRF and Caveolin-1mRNA levels in the same GBM specimens were investigated by Pearson correlation. The difference was considered statistically significant when the P value was less than0.05. All statistical analyses were carried out with SPSS13.0software.RESULTS:1. Imatinib-resistant GBM cell line U251AR was established successfullyBy using the parental cell line U251, we previously established the imatinib-resistant GBM cell line U251AR, which had a cross-resistance to VP-16and TMZ. The U251AR was cultured in medium with imatinib (122μg/mL) to maintain the MDR phenotype. The MDR phenotypes of imatinib-resistant cell line U251AR included up-regulation of some cellular genes. In this study, we tested the ATP-dependent drug efflux pump (P-gp) expression by Western blotting and the mRNA levels of P-gp, MRP1and BCRP by quantitative RT-PCR in U251AR in comparison with the parental cell line U251. The P-gp, MRP1and BCRP were significantly increased in drug-resistant cell line U251AR (*, P<0.05). These results suggest that the imatinib-resistant GBM cell line U251AR was established successfully.2. Proteome profiling of U251and U251AR cell linesTo obtain a global protein image of U251and U251AR cells, we performed three2D-DIGE gels to detect differently expressed proteins. For each gel, a merged image was generated from three images of the U251, U251AR, and the internal standard samples. A representative DIGE gel with merging of Cy3and Cy5-labeled images was shown. A total of2516to2735spots were detected in the DIA workspaces using DeCyder software. In the BVA module,41spots were found to be differentially expressed based on the criteria that an average ratio was more than1.5or less than1.5(P value<0.05). Among them,23spots were found to be down-regulated and18spots up-regulated in the chemoresistant U251AR when compared with U251. Some protein spots might be undetectable in gel stained with Coomassie Blue because of their low expression levels. Twenty-one protein spots with high abundance were found with significantly altered expression in both cell lines as indicated by the MALDI-TOF/TOF MS analysis. Among the21differentially expressed proteins,9proteins were up-regulated and12proteins were down-regulated in U251AR cell line, including PTRF and VIM. The3-D view of PTRF and VIM proteins were showed.3. Validation of high-expression protein PTRF and VIM in U251ARTo test our proteomic results, we performed Western blot and quantitative RT-PCR to detect the expression of PTRF and VIM. The Western blot and quantitative RT-PCR results confirmed that PTRF and VIM were both highly expressed in U251AR compared with its parental cell U251(*, p<0.05), which were consistent with our previous results. To gain a comprehensive view of cellular changes induced upon PTRF expression, we used cell immunofluorescence to detect the cellular localization of PTRF and caveolinl in both U251AR and U251cells. PTRF was detected in nucleus and cytoplasm in both cells, with more fluorescence detected in cytoplasm of U251AR than in that of U251. Caveolinl was also detected in cell membrane and cytoplasm in both cell lines with more fluorescence detected in cytoplasm of U251AR, suggesting that U251AR cells may possess more caveolae than U251cells.4. Knockdown of PTRF in GBM cell lines increases chemosensitivityTo further investigate the effect of PTRF on chemoresistance of GBM cells, we knocked down the expression of PTRF using pcDNA6.2-GW/EmGFP-miRNA in both U251and U251AR cell lines. The morphologies of transfected cells were showed in Fig.5A and5B. The interference efficiency was confirmed by Western blotting and quantitative RT-PCR (*, P<0.05). Interestingly, silencing PTRF significantly reduced the mRNA and protein levels of caveolinl and P-gp (**, P<0.05).Both PTRF and caveolinl, the two caveolae structure proteins, have been shown to be relevant to chemoresistance. To test the effect of PTRF on cell viability, we treated both cell lines with or without knockdown of PTRF with TMZ (100μg/mL) for (24h,48h,72h,96h, and120h) and cell viability assay was performed. In this assay, cells with PTRF knockdown showed decreased cell viability when compared with the control cells under the same concentration of TMZ (*, P<0.05,). To test the roles of PTRF in GBM chemical drug sensitivities, the IC50values of U251cells, U251AR cells and the transfected cells after treatment with imatinib, VP-16, and TMZ were determined by CCK8assay kit. The IC50values of shPTRF transfected U251and U251AR cells after treatment with imatinib, VP-16, and TMZ were significantly decreased by2.05-3.92folds (**, P<0.01), indicating that down-regulation of PTRF sensitizes GBM cells to chemotherapeutic drugs.5. PTRF is up-regulated in relapsed GBM specimens and positively correlated with caveolinlIn this study, PTRF expression was further detected by immunohistochemistry in tissues from58cases of patients with astrocytoma and6cases of patients with relapsed GBM. In addition,8cases of non-tumor tissues were used as control samples in the immunohistochemistry analysis. The immunohistochemistry assay of PTRF in astrocytoma and normal brain tissue specimens revealed that PTRF was lowly expressed in normal brain tissue and low-grade astrocytoma (grade I and II), but highly expressed in high-grade astrocytoma (grade III and IV). We also found that the expression level of PTRF in relapsed GBM patients with treatment of TMZ for6months was higher than that in primary GBM patients without treatment of TMZ. Consistent with the expression of PTRF, caveolinl was also highly expressed in the relapsed GBM patients. Furthermore, the mRNA levels of PTRF and caveolinl in relapsed GBM patients were significantly higher than those in patients with primary GBM (*, P<0.01). PTRF and caveolinl are two essential components in the biogenesis and function of caveolae. Then, we did correlation analysis between mRNA level of PTRF and caveolinl in the same GBM specimens. Correlation analysis showed that there was a positive correlation between PTRF mRNA levels and caveolinl mRNA levels (2-tailed Pearson correlation, r=0.766, P<0.01). All these results indicate that the average expression level of PTRF in the same GBM specimens may be correlated with that of caveolin1.CONCLUSION:1. By proteomics, we found21biomarkers of GBM chemoresistance.2. We revealed that PTRF play an important role in GBM chemoresistance.3. The expression of PTRF in relapsed GBM is significantly increased than the primary one.4. PTRF may used for GBM early diagnose and prognosis analysis, furthermore, PTRF may act as newly target of GBM therapy. |
PDF Full Text Request