Article pubs.acs.org/molecularpharmaceutics
Nanoparticle-Delivered Antisense MicroRNA-21 Enhances the Effects of Temozolomide on Glioblastoma Cells Jeyarama S. Ananta, Ramasamy Paulmurugan,* and Tarik F. Massoud* Laboratory of Experimental and Molecular Neuroimaging, Molecular Imaging Program at Stanford (MIPS), and Bio-X Program, Stanford University School of Medicine, Stanford, California 94305-5427, United States S Supporting Information *
ABSTRACT: Glioblastoma (GBM) generally exhibits high IC50 values for its standard drug treatment, temozolomide (TMZ). MicroRNA-21 (miR-21) is an oncomiR overexpressed in GBM, thus controlling important aspects of glioma biology. We hypothesized that PLGA nanoparticles carrying antisense miR-21 to glioblastoma cells might beneficially knock down endogenous miR-21 prior to TMZ treatment. PLGA nanoparticles encapsulating antisense miR-21 were effective in intracellular delivery and sustained silencing (p < 0.01) of miR21 function in U87 MG, LN229, and T98G cells. Prior antisense miR-21 delivery significantly reduced the number of viable cells (p < 0.001), and increased (1.6-fold) cell cycle arrest at G2/M phase upon TMZ treatment in U87 MG cells. There was overexpression of the miR-21 target genes PTEN (by 67%) and caspase-3 (by 15%) upon cotreatment. This promising PLGA nanoparticle-based platform for antisense miR-21 delivery to GBM is an effective cotherapeutic strategy in cell culture, warranting the need for further studies prior to future clinical translation. KEYWORDS: GBM, temozolomide, nanoparticle, PLGA, microRNA-21, antisense miR-21
G
egies are emerging as a promising new generation of molecular therapies for cancer, including GBM. MiRNA-21 (miR-21) is one of the first miRNAs to be described as an oncomiR. It is overexpressed in a wide range of cancers,9 and has been identified as an oncogene controlling apoptosis, cell proliferation, and migration.10−13 Recently, using high-throughput miRNA profiling, miR-21 has been shown to be strongly overexpressed in nearly all analyzed human GBM specimens.14 Knockdown of miR-21 also results in significant reduction in cell viability in glioblastoma cell lines.14 The exact mechanisms that control these changes are not yet completely understood. However, considering the general chemoresistance of GBM to its standard treatment using TMZ, and the significant role of miR-21 in cancer cell biology, the question arises as to whether a potential beneficial therapeutic effect may be possible by adding miR-21 knockdown to the therapeutic response of glioblastoma cells to TMZ. Recently, antisense miRNA assisted miR-21 inhibition has been shown to enhance the chemosensitivity of TMZ-resistant glioblastoma cells.15−17 Many of these studies relied on less clinically applicable cationic, polyamino compounds for efficient translocation of antisense miRNAs across the cell membrane. Commercially available transfection agents, such as Oligofectamine, and polycationic dendrimer based delivery vehicles
lioblastoma (GBM) remains an aggressive malignancy with high morbidity and mortality.1 Despite advances in surgical resection, radiotherapy, and chemotherapy, the prognosis for GBM has not significantly improved in recent years.2 The ineffectiveness of current treatments can be attributed in part to tumor heterogeneity, invasion, and chemoresistance. Although temozolomide (TMZ) is currently the standard adjuvant chemotherapy for GBM, glioblastoma cells in general exhibit high IC50 values for TMZ (in the millimolar range). Moreover, recent studies have shown that GBM is resistant to the cytotoxic effects of TMZ.3,4 Hence, there is a need to develop better optimized and more effective chemotherapeutic strategies for GBM that take into account new drugs or drug formulations in combination with its standard therapeutic agent, TMZ. MicroRNAs (miRNA) are small (approximately 18−22 nucleotides) noncoding RNAs that regulate gene expression by binding to partially or fully complementary recognition sequences of target mRNA (mRNA), resulting in mRNA degradation or translational suppression.5 By negatively regulating their target mRNAs, miRNAs can act as either tumor suppressors or oncogenes. Dysregulated miRNA expression is commonly reported in various human cancers including GBM. In addition, genomic regions coding for more than 50% of miRNAs are located near cancer associated genomic break points.6 Altering the expression of miRNAs in general has significant implications on cell viability as well as strategies to overcome cancer cell resistance to chemotherapeutic drugs.7,8 Hence, miRNA-targeted treatment strat© XXXX American Chemical Society
Received: September 9, 2015 Revised: October 28, 2015 Accepted: November 2, 2015
A
DOI: 10.1021/acs.molpharmaceut.5b00694 Mol. Pharmaceutics XXXX, XXX, XXX−XXX
Article
Molecular Pharmaceutics
according to the manufacturer’s protocol. For PLGA assisted transfection of cells, we incubated the GBM cells with PLGA nanoparticles encapsulating 20 pmol of antisense miR-21 (polymer concentration: 20 μg of PLGA) in serum free medium for 4 h at 37 °C and 5% CO2. After 4 h of incubation, we supplemented the medium with 10% FBS and allowed the cells to grow overnight. For subcellular localization of PLGA nanoparticles in U87 MG cells, we transfected the cells with a Cy5 labeled antisense miR-21 (Cy5-UpCpApACAUCAGUCUGAUAApGpCpUpA) using the same procedure as above and imaged using confocal microscopy (see Supporting Information). MicroRNA Extraction and Quantitative Real-Time PCR. We isolated the total RNA (mRNA and miRNA) using mirVana RNA extraction Kit (Life Technologies, CA) using the manufacturer’s protocol. For miRNA quantitation, we reverse transcribed 15 ng of total RNA using RT-primers of miR-21 using TaqMan MicroRNA Reverse Transcription Kit (Life Technologies, Carlsbad, CA, USA). We performed real-time PCR using 5 μL of cDNA (750 pg of RNA equivalent) combined with of TaqMan real time PCR reagents of miR-21. We performed all reactions in a 20 μL volume. PCR parameters consisted of 2 min incubation at 50 °C, followed by activation of the DNA polymerase at 95 °C for 10 min and 60 cycles of 95 °C × 15 s, 60 °C × 60 s in an Eppendorf real-time PCR system. We normalized the miR-21 expression to RNU66. We used a single factor ANOVA method for determining statistical significance, and considered a p value of ≤0.05 as significant. Cell Viability Studies. We used an MTT [3-(4,5dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] assay to evaluate the cell proliferation and cytotoxicity of TMZ on antisense miR-21 transfected and nontransfected GBM (U87 MG, LN229, T98G) cells. We seeded the cells in a 96-well plate at 1 × 104 cells/well density. We exchanged the medium for 100 μL of fresh medium (2% FBS) supplemented with TMZ in concentrations ranging from 100 μM to 500 μM. We then incubated the cells for a further 48 h. At 48 h post TMZ exposure, we aspirated the cell culture medium and incubated the cells with 100 μL of medium (containing 12 mM MTT) for 2 h at 37 °C in 5% CO2. We then carefully removed the medium and dissolved the reduced formazan crystals in 100 μL of DMSO by incubating at 37 °C for 30 min in the dark. Next, we calculated cell viability by measuring the absorbance at 540 nm using a microplate reader (Infinite M1000 Pro, Tecan Group, Switzerland) and comparing with the control cells. The relative cell viability (%) compared to control cells was calculated as follows: cell viability (%) = {[Abs(sample) − Abs(blank)]/[Abs(control) − Abs(blank)]} × 100. We used a single factor ANOVA method for determining statistical significance, and considered a p value of ≤0.05 as significant. Cell Cycle Analysis. For cell cycle distribution studies, we harvested the transfected and control U87 MG cells after the specified experimental conditions, and washed them with PBS. We then fixed the cells in 70% ethanol and then stained them with 0.5 mL of PBS containing 0.5 μg/mL propidium iodide, 10 μg/mL RNase A, and 0.1% Triton X-100. We incubated the cells for 30 min at room temperature in the dark and performed the analysis using a FACS Aria III (BD Biosciences, CA) cell sorter. We analyzed the data using FlowJo FACS analysis software (Tree Star, Ashland, OR). Immunoblot Analysis. For immunoblot analysis, we washed the cells with PBS after the specified experimental conditions to remove any traces of culture medium and debris,
form stable polyamino lipoplexes with sense and antisense miRNAs, to then deliver them across cell membranes. However, such agents are not suitable for in vivo applications owing to degradation by circulating serum nucleases, systemic toxicity, off-target effects, and rapid renal clearance. Despite efforts to improve the plasma stability of antisense miRNAs, therapeutic applications require an effective delivery system to protect the miRNAs from systemic degradation and achieve localized concentration at the target site. Poly(lactic-co-glycolic acid) (PLGA) nanoparticles are attractive platforms for encapsulation of antisense miRNAs and their sustained release at the target site. Here we use a poly[lactic-co-glycolic]acid (PLGA) based nanoparticle vehicle to encapsulate antisense miR-21 and deliver it in a “Trojan horse” manner to GBM cells in culture. We show that PLGA assisted delivery of antisense miR-21 is effective in silencing endogenous miR-21, producing a sustained knockdown effect over time. PLGA nanoparticle assisted miR-21 suppression prior to TMZ treatment significantly reduces the viability of cells. Moreover, silencing of endogenous miR-21 with PLGA nanoparticles significantly increases cell cycle arrest in G2/M phase for TMZ-treated cells, increases phosphatase and tensin homologue (PTEN) gene expression, and increases apoptosis associated caspase-3 expression. Our results demonstrate that PLGA nanoparticles encapsulating antisense miR-21 are very effective in enhancing the antiproliferative and apoptotic response of GBM cells to TMZ.
■
METHODS Preparation of Antisense MiR-21 Encapsulated PLGA Nanoparticles. We prepared PLGA nanoparticles encapsulating antisense miR-21 using a double emulsion solvent evaporation technique as previously reported, with minor modifications.18 The detailed nanoparticle preparation procedure can be found in the Supporting Information. We measured particle size, size distribution, and surface charge (ζ-potential) using the dynamic light scattering (DLS) technique (Zetasizer Nano ZS, Malvern Instruments, U.K.). We analyzed the size, shape, and surface morphology of antisense miR-21 loaded nanoparticles using scanning election microscopy (SEM) (FEI XL30 Sirion, FEI, Hillsboro, OR, USA). We then measured the encapsulation efficiency of antisense miR-21 within PLGA nanoparticles using an organic/aqueous extraction method. We measured the amount of antisense miR21 present in the aqueous fraction using the Quant-iT RNAquantitation kit (Invitrogen, Carlsbad, CA), according to the manufacturer’s guidelines. We calculated the “entrapment efficiency” using the following formula: entrapment efficiency (%) ⎛ amount of antisense miR‐21 quantified in NPs ⎞ ⎟ × 100 =⎜ ⎝ ⎠ total amount of antisense miR‐21 used
Antisense MiR-21 Transfection of Glioblastoma Cells. We used an antisense miR-21 oligonucleotide with the following sequence: UpCpApACAUCAGUCUGAUAApGpCpUpA modified to possess phosphorothioate linkage for our endogenous miR-21 target blocking studies. The oligonucleotide was custom synthesized in the PAN facility at Stanford University with >90% purity. We transfected GBM cells (U87 MG, LN229, T98G [see cell culture details in Supporting Information]) seeded in 6-well plates at 80% confluence using Oligofectamine (Invitrogen, Carlsbad, CA) B
DOI: 10.1021/acs.molpharmaceut.5b00694 Mol. Pharmaceutics XXXX, XXX, XXX−XXX
Article
Molecular Pharmaceutics and lysed the cells in RIPA buffer (Pierce Biotechnology, Rockford, IL). We estimated the protein content of the supernatant solution using a Pierce 660 protein assay kit (Life Technologies, Carlsbad, CA). We then resolved 15 μg of protein in 4−12% gradient SDS/PAGE (Invitrogen) and electroblotted onto a 0.2 μm pore size nitrocellulose membrane (Schleicher & Schuell Biosciences, GmbH). We used SeeBlue (Invitrogen, Carlsbad, CA) protein marker to confirm the molecular mass and complete transfer of protein to the membrane. Next, we incubated the membrane in the blocking solution (5% nonfat dry milk in TBS-T (TBS with 0.1% Tween 20) buffer) with respective antibody (rabbit mAb PTEN, rabbit mAb β-actin, rabbit mAb caspase-3, rabbit Ab bcl-2, rabbit Ab bax; Cell Signaling Technology, Danvers, MA) at the manufacturer suggested dilution, overnight and at 4 °C on a rotating platform. We washed the membrane 3 times with TBST and incubated with HRP-conjugated goat antirabbit antibody for 2 h at room temperature. We washed the membrane a further 3 times with TBS-T buffer before incubation with the chemiluminescent HRP substrate LumiGlo (Cell Signaling), following the manufacturer’s instructions. We stripped the same membrane and reprobed with rabbit anti-human β-actin antibody (Sigma, St. Louis, MO) to control for protein loading. We then detected the luminescence signal using an IVIS optical CCD camera (Caliper, CA) and quantitated the signals in total photon counts by drawing ROIs over the bands.
Figure 1. MiR-21 expression analysis by Taqman-based RT-PCR. (a) Endogenous levels of miR-21 expression in glioblastoma (U87 MG, LN 229, and T98 G) cells, hepatocellular carcinoma (HepG2) cells, and cortical neuronal (HCN 2) cells. Compared to control HCN-2 cells, U87 MG cells overexpressed miR-21 by 10-fold (p < 0.01), and LN 229 cells did so by 3-fold (p < 0.01). There was no statistically significant overexpression of miR-21 in T98G cells (p > 0.05). (Results are presented as mean ± standard error of the mean; ** indicates p ≤ 0.01.)
the particles is to be expected since the hydrodynamic diameter is measured by assuming a hydration or solvation layer surrounding each particle, whereas the SEM images reveal the core size of the nanoparticles. The particles had a negatively charged surface and exhibited a high antisense miR-21 encapsulation efficiency of 76% (Figure 2b). We analyzed the expression of endogenous miR-21 at 24 and 48 h after antisense miR-21 transfection using Taqman-based qRT-PCR (Figure 3a). At 24 h after transfection, the antisense miR-21 delivered by both Oligofectamine (97% suppression) and PLGA (77% suppression) exhibited significant downregulation of endogenous miR-21 (p < 0.01) in U87 MG cells. At 48 h, Oligofectamine transfected cells exhibited higher levels of miR-21 (only 34% suppression compared to nontransfected U87 MG cells), whereas PLGA nanoparticles showed sustained suppression of endogenous miR-21 levels (73% suppression; p < 0.01) (Figure 3a). PLGA nanoparticles encapsulating antisense miR-21 were also very effective in suppressing the endogenous miR-21 levels in LN229 cells (60% suppression at 48 h) and T98 G (>90% suppression at 48 h) (Figure 3b). These results confirm that antisense miR-21 used in our study was effective in specifically blocking endogenous miR-21, and PLGA nanoparticles encapsulating antisense miR-21 were very effective at sustained downregulation of miR-21 compared to commercially available lipid based transfection agents. Antiproliferative and Cytotoxic Effects of Antisense MiR-21 and TMZ Cotreatment in Glioblastoma Cells. After confirming the downregulation of miR-21 in U87 MG, LN 229, and T98G cells, following transfection with antisense miR-21 encapsulated PLGA nanoparticles, we studied the effect of miR-21 downregulation on antiproliferation and cell viability of TMZ treated glioblastoma cells. The cells were treated with 20 pmol of antisense miR-21 loaded PLGA nanoparticles and tested for antiproliferative effects in the presence of varying doses of TMZ (0 to 500 μM). We observed a significant reduction in the number of U87 MG cells when they were transfected with antisense miR-21 before treatment with TMZ (Figure 4a). We also evaluated the proliferation of cells by indirectly measuring mitochondrial function using an MTT assay (Figure 4b). At all concentrations of TMZ, and especially
■
RESULTS MiR-21 Is Overexpressed in GBM Cells. Previous studies have shown that many of the established glioblastoma cell lines (A172, U373, LN229, LN428) exhibit elevated miR-21 levels.14 MiR-21 overexpression has been associated with tumor growth, evasion of apoptosis, tumor invasion, and migration. We thus evaluated miR-21 expression levels in the U87 MG, LN229, and T98G cells used in our studies. We performed Taqman-based quantitative RT-PCR (qRT-PCR) analysis for miR-21 expression in U87 MG, LN 229, and T98 G cells, as well as HepG2 (hepatocellular carcinoma cells) and HCN-2 (brain cortical neuronal cells) for comparative controls. Compared to control HCN-2 cells, U87 MG cells overexpressed miR-21 by 10-fold (p < 0.01), and LN 229 cells did so by 3-fold (p < 0.01). There was no statistically significant overexpression of miR-21 in T98G cells (p > 0.05). Of note, compared to control HCN-2 cells and GBM (U87 MG, LN 229, T98G) cells, liver cancer Hep-G2 cells had significantly lower expression of endogenous miR-21 (p < 0.01) (Figure 1). Therefore, miR-21 is an oncomiR selectively overexpressed in glioblastoma cells (U87 MG, LN 229). These cell lines were deemed suitable to test the PLGA nanoparticle assisted silencing of miR-21 combined with TMZ as a molecular therapeutic strategy for GBM. PLGA Nanoparticle Delivered Antisense MiR-21 Silences Endogenous MiR-21 Function in GBM Cells. We evaluated antisense miR-21-mediated suppression of endogenous miR-21 function in cells by two different delivery techniques. We transfected the miR-21 overexpressing glioblastoma cells with a commercially available transfection agent Oligofectamine and with PLGA nanoparticles encapsulating antisense miR-21 oligonucleotide. Antisense miR-21 loaded PLGA nanoparticles were spherical in shape and exhibited a mean hydrodynamic diameter of 207 nm (Figure 2a). SEM images (Figure 2c) of the nanoparticles showed the nanoparticle diameters to be in the range 150 to 170 nm. This observed difference between two measurements in diameters of C
DOI: 10.1021/acs.molpharmaceut.5b00694 Mol. Pharmaceutics XXXX, XXX, XXX−XXX
Article
Molecular Pharmaceutics
Figure 2. Physicochemical characterization of antisense miR-21 loaded PLGA nanoparticles. (a) Representative DLS result of antisense miR-21 loaded PLGA nanoparticles showing particle size distribution. (b) Size, surface charge, and encapsulation efficiency of antisense miR-21 loading in PLGA nanoparticles. (c) Representative scanning electron microscopic images of antisense miR-21 loaded PLGA nanoparticles. (d) Confocal microscopic images of Cy5-labeled antisense miR-21 loaded nanoparticles internalized in U-87 MG cells showing the nanoparticle distribution in the cytoplasm, likely in endosomes. White arrows indicate the Cy5-antisense miR-21 loaded nanoparticles (red).
Figure 3. (a) Comparison of antisense miR-21 assisted knockdown of endogenous miR-21 in U87 MG cells at 24 and 48 h post transfection. (b) Efficiency of PLGA nanoparticle assisted miR-21 knockdown in glioblastoma cells at 48 h post transfection. (Results are presented as mean ± standard error of the mean; ** represents p ≤ 0.01.)
Antisense MiR-21 Treatment Induces Cell Cycle Arrest at G2/M Phase in U87 MG Cells. To better understand the decrease in cell viability, we performed cell cycle analysis using FACS. We used U87 MG cells, as they exhibited the highest level of endogenous miR-21 expression and also significantly enhanced sensitivity to TMZ upon antisense miR-21 transfection. We did not observe any significant accumulation in G2/M phase when U87 MG cells were transfected with antisense miR-21 in the absence of TMZ treatment. Moreover, TMZ treatment alone induced only a 50% increase in G2/M cell accumulation compared to untreated control cells. However, PLGA nanoparticle-mediated antisense miR-21 transfection prior to TMZ treatment induced a 1.6-fold increase in G2/M phase cell accumulation compared to nontreated cells (Figure 5a,b, Table S1). This nearly 2-fold increase in cell cycle arrest at G2/M phase upon miR-21 downregulation suggested that antisense-miR-21 transfection
at higher concentrations, PLGA nanoparticle-mediated antisense miR-21 transfection prior to TMZ treatment resulted in significant decreases in cell viability compared to TMZ alone treatment (Figure 4b) in U87 MG cells. At 250 μM TMZ, miR21 downregulation induced 15% reduction in cell viability (p < 0.01), and at 500 μM TMZ it increased to 20% (p < 0.01) when the cells were transfected with antisense miR-21. We also observed a linear dose-dependent response to TMZ in transfected cells (R2 = 0.99). Similarly in LN 229 cells, at 500 μM TMZ, miR-21 downregulation induced 16% reduction in cell viability (p < 0.01) compared to nontransfected control cells (10% decrease in cell viability) (Figure 4c). However, antisense miR-21 transfection did not induce any significant decrease in the viability of T98G cells (Figure 4d). The observed lack of effect on cell viability upon antisense miR-21 transfection in T98G cells was likely owing to the lower levels of endogenous miR-21 expression in T98G cells. D
DOI: 10.1021/acs.molpharmaceut.5b00694 Mol. Pharmaceutics XXXX, XXX, XXX−XXX
Article
Molecular Pharmaceutics
Figure 4. Antiproliferative and cell viability studies on antisense miR-21 and TMZ cotreated glioblastoma cells. (a) Phase contrast microscopic images of control and antisense miR-21 transfected cells treated with 500 μM TMZ. In vitro cell viability MTT assays on control and antisense miR21 transfected (b) U87 MG cells, (c) LN 229 cells, and (d) T98G cells treated with different concentrations of TMZ. (Results are presented as mean ± standard error of the mean; * represents p ≤ 0.05 ** represents p ≤ 0.01.)
Figure 5. Cell cycle analysis of antisense miR-21 and TMZ cotreatment on U87 MG cells. (a) Flow cytometry analysis of cells with propidium iodide staining for cell cycle status. (b) Percentage cell distribution at different phases of cell cycle for control cells and antisense miR-21 transfected cells upon treatment with 500 μM TMZ.
before TMZ treatment had a large effect on inhibiting the proliferation of U87 MG cells. MiR-21 Downregulation Prior to TMZ Treatment Increases PTEN and Caspase-3 Expression in U87 MG Cells. To further explore the effects of antisense miR-21 and TMZ cotreatment on different cellular pathways, we performed immunoblot analysis for key targets of miR-21. Figure 6a shows
the immunoblot results for PTEN, Bcl-2, Bax, and caspase-3. TMZ treatment in nontransfected cells induced only an approximate 30% increase in PTEN expression. However, we observed a more than 2-fold increase in the expression of PTEN when U87 MG cells were transfected with antisense miR-21 followed by treatment with TMZ (65%) compared to TMZ treatment alone (Figure 6a). Similarly, nontransfected cells E
DOI: 10.1021/acs.molpharmaceut.5b00694 Mol. Pharmaceutics XXXX, XXX, XXX−XXX
Article
Molecular Pharmaceutics
Figure 6. Cellular pathway analysis of antisense miR-21 and TMZ cotreatment on U87 MG cells. (a) western blot analysis showing PTEN, Bax, Bcl2, caspase-3, and β-actin protein levels after different treatment conditions. (b) Bax/bcl-2 ratio calculated from western blot analysis after different treatment conditions.
upon treatment with 500 μM TMZ exhibited no significant increase (
