Mesenchymal stem cells and glioma cells form a structural as well as a functional syncytium in vitro

Share Embed


Descrição do Produto

Experimental Neurology 234 (2012) 208–219

Contents lists available at SciVerse ScienceDirect

Experimental Neurology journal homepage: www.elsevier.com/locate/yexnr

Mesenchymal stem cells and glioma cells form a structural as well as a functional syncytium in vitro Christian Schichor a,⁎, Valerie Albrecht a, Benjamin Korte a, Alexander Buchner b, Rainer Riesenberg b, Josef Mysliwietz c, Igor Paron a, Helena Motaln d, Tamara Lah Turnšek d, Kathrin Jürchott e, Joachim Selbig e, Joerg-Christian Tonn a a

Tumorbiological Laboratory, Neurosurgical Clinic, Ludwig-Maximilians-University Munich, Germany LIFE Center, Department of Urology, University Clinic Grosshadern, Ludwig-Maximilians-University Munich, Germany c Institute for Molecular Immunology, National Research Center for Environment and Health, Munich, Germany d Department of Genetic Toxicology and Cancer Biology, National Institute of Biology, Ljubljana, Slovenia e AG Bioinformatics, Institute for Biochemistry and Biology, University of Potsdam, Germany b

a r t i c l e

i n f o

Article history: Received 20 July 2011 Revised 9 December 2011 Accepted 19 December 2011 Available online 29 December 2011 Keywords: Mesenchymal stem cell Glioma Syncytium Gap junction Fusion

a b s t r a c t The interaction of human mesenchymal stem cells (hMSCs) and tumor cells has been investigated in various contexts. HMSCs are considered as cellular treatment vectors based on their capacity to migrate towards a malignant lesion. However, concerns about unpredictable behavior of transplanted hMSCs are accumulating. In malignant gliomas, the recruitment mechanism is driven by glioma-secreted factors which lead to accumulation of both, tissue specific stem cells as well as bone marrow derived hMSCs within the tumor. The aim of the present work was to study specific cellular interactions between hMSCs and glioma cells in vitro. We show, that glioma cells as well as hMSCs differentially express connexins, and that they interact via gap-junctional coupling. Besides this so-called functional syncytium formation, we also provide evidence of cell fusion events (structural syncytium). These complex cellular interactions led to an enhanced migration and altered proliferation of both, tumor and mesenchymal stem cell types in vitro. The presented work shows that glioma cells display signs of functional as well as structural syncytium formation with hMSCs in vitro. The described cellular phenomena provide new insight into the complexity of interaction patterns between tumor cells and host cells. Based on these findings, further studies are warranted to define the impact of a functional or structural syncytium formation on malignant tumors and cell based therapies in vivo. © 2011 Elsevier Inc. All rights reserved.

Introduction Patients harboring malignant gliomas still have a dismal prognosis despite multimodal treatment strategies. Glioblastoma multiforme is an aggressive and highly vascularized tumor of the brain, which is known to interact extensively with the host environment. Malignant gliomas are supposed to recruit stem cells both from the entire brain itself as well as systemically from the circulation (Kim et al., 2011; Schichor et al., 2006). In human gliomas, bone marrow derived hMSCs have not been detected, yet. This might be due to the fact, that hMSCs cannot be determined easily by immunohistochemistry. They are not defined by a single, specific surface marker, but are characterized by the presence of a typical marker profile including strong expression of e.g. CD105, CD90, CD29, CD73 and others (Keating, 2006). However, the definite role of host stem cells in tumor growth, progression and recurrence is still not completely understood. In ⁎ Corresponding author at: Tumorbiological Laboratory, Neurosurgical Clinic, Marchioninistr. 15, 81377 Munich, Germany. Fax: + 49 3212 1306680. E-mail address: [email protected] (C. Schichor). 0014-4886/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.expneurol.2011.12.033

experimental systems it has been shown that various stem cells including hMSCs might inhibit tumor growth (Pisati et al., 2007; Staflin et al., 2004). In contrast, other studies indicate that they may contribute to tumor formation (Djouad et al., 2003; Liu et al., 2011; Ramasamy et al., 2007; Ricci-Vitiani et al., 2008). Despite these findings, hMSC-based cell-therapies aim at utilizing their affinity for diseased areas as cellular vectors (Choi et al., 2012; Ricci-Vitiani et al., 2008; Sonabend et al., 2008). The putative advantage of this approach would be the administration of a motile multipotent cell type in order to target single, infiltrative glioma cells within functional brain parenchyma. Both, endogenously recruited or exogenously administered hMSCs get into close and intense interaction with malignant glioma cells. The predictability of this interaction remains unclear, when concerns about adverse side-effects are addressed. Several other publications already described the intense intercellular crosstalk and unwanted cell fate in non-malignant brain diseases (Jaderstad et al., 2010; Snyder, 2011). A reasonable approach to identify detailed mechanisms of cross-talk between hMSCs and glioma cells would be a simple co-culture setting. Surprisingly, we found unexpected, complex interaction patterns between glioma cells and hMSCs that

C. Schichor et al. / Experimental Neurology 234 (2012) 208–219

are distinct from paracrine interactions. A special focus of the present study is syncytium formation, which can be seen in two ways: communication between two different cell types via gap junctional coupling (functional syncytium) or fusion of two different cells (structural syncytium). Gap junctions are plasma membrane protein channels connecting cells and allowing communication through electrical coupling and transfer of small molecules (b1 kD, functional syncytium). Structural syncytium formation of hMSCs has gained attention due to their transdifferentiation potential into lineage restricted cells of cardiogenic or neurogenic tissues (Alvarez-Dolado et al., 2003). In fact, there is growing evidence that fusion of hMSCs with host cells serves as an alternative mechanism of transdifferentiation (Kamijo et al., 2006; Wurmser and Gage, 2002). In this study, we describe how adult hMSCs intensively interact with glioma cells via functional as well as structural syncytium formation, altering migratory potential and proliferative behavior. These phenomena give an impression of the complex interaction modes between different cell types and should be taken into account in designing in vitro studies or in interpreting results of in vivo cell therapeutic studies. Materials and methods Cell types Isolation of primary hMSCs from bone marrow of healthy donors HMSCs were generated out of bone marrow aspirates from healthy donors after written informed consent. Isolation was performed as described elsewhere (Otsu et al., 2009). In brief, mononuclear cell fraction was isolated by Ficoll density gradient centrifugation (1.077 g/mL, SIGMA-Aldrich) at 400 g for 25 min. The cells were cultured in DMEM (FG0415, Biochrom) supplemented with 2 mM glutamine, 100 U/ml penicillin, 0.1 mg/ml streptomycin (Gibco) and 20% fetal bovine serum (Biochrom) at 37 °C. Only hMSCs of passages 2–4 were used for experiments. Cells were checked by flow cytometry and resulted positive for CD105, CD90, CD166, CD29 and CD40, and negative for typical hematopoietic and endothelial markers (CD45, CD34). Cell lines The human glioma cells lines U87 and U373 (ATCC) as well as green fluorescent protein (GFP) transfected U373 cells (U373-GFP) were used. Cells were grown in complete growth medium consisting of DMEM, supplemented with 10% FBS, non-essential amino acids, 4 mM L-glutamin, penicillin (100 U/ml) and streptomycin (100 μg/ml) (Gibco). The U373-GFP cells were routinely propagated under the same conditions with addition of 1.2 mg/ml G-418 (SIGMA-Aldrich) to the medium. As control cells, we used an astrocyte cell line SVG p12 (LGC Promochem) cultured in MEM Earle's (FG0325, Biochrom) with above mentioned supplements. Differentially regulated expression of connexins Gene expression data In order to identify differentially regulated expression of connexins, we analyzed microarray data as follows: Gene expression data were generated from the following samples in 3 biological replicates each: hMSC monoculture (MSC), U87 glioma cells (U87), hMSC from indirect co-culture with U87 (hMSC grown on the bottom-plate of the Boyden chamber, MSC-IC) and U87 from indirect co-culture with hMSC (U87 grown above in the insert of the Boyden chamber, U87-IC) on Illumina HumanWG-6 v3.0 expression beadchips. These chips contain probes that detect the expression of 16 out of the 20 connexins annotated with protein-coding transcripts in Ensembl (Table S2). The expression of the remaining four connexins (Cx30.2/Cx31.9, Cx36, Cx40.1, Cx62) cannot be measured by this chip. All protein-coding transcripts of 10 connexins (Cx26, Cx30, Cx30.2, Cx45, Cx47/46.6, Cx50) can be analyzed by these chips while only

209

some of the annotated transcripts can be measured for 6 connexins (Cx25, Cx32, Cx37, Cx40, Cx43, Cx59/Cx58). Analysis of the expression data was performed with R (http://www.r-project.org). Data was quantile normalized. Sequence information for the Illumina probes was received from the corresponding Illumina annotation file (http:// www.illumina.com/support/annotation_files.ilmn) and data concerning connexins was extracted. The sequences were blasted against a database containing all available transcript sequences for connexins based on Ensembl Release 61 (1 February 2011, GRCh37, http:// www.ensembl.org). The results of this analysis are summarized in the Supplementary Material (Table S1). Alternative connexin names were received from the HUGO Gene Nomenclature Committee (HGNC, http://www.genenames.org/) and Genecards (http://www.genecards. org/). In addition, all available Ensembl transcripts for connexin genes were tested for their potential to be recognized by probes on the Illumina HumanWG6 v3.0 array (Table S2). Mean group values for the expression of connexin probes and genes are given in Supplementary Material (Table S3). All microarray data were deposited to the NCBI's Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo/), a public functional genomics data repository (GEO series Accession Number: GSE26283). Co-culture and conditioned medium culture of hMSC and glioma cell lines U87 and hMSC lines were chosen for analysis. The co-culture experiments were set up in duplicates using 6-well transwell inserts (Costar) with 0.4 μm pore size. For indirect co-cultures, hMSCs were plated on the bottom of the 6-well plate at a density of 1 × 10 5 cells/ 2 ml, whereas the U87 cells were plated in the transwell at a density of 8 × 10 4 cells/ml. Cells were left to adhere for 3 h before being put together for 72 h. Conditioned media was harvested after 24 h from subconfluent cells just prior to trypsinization. For conditioned media culture, cells of both types were plated as for indirect co-culture, then conditioned media (diluted 1:1 with fresh medium) was added to the cells. For direct co-cultures both types of cells were mixed at the same ratio (1:1). Cells were left to grow for 72 h before RNA isolation. RNA isolation and quantitative Real-time PCR for Connexin 43 expression Total RNA was isolated from U87 cells and hMSCs either from monoculture, indirect co-culture, conditioned media culture or direct coculture (1:1) using Trizol reagent (Invitrogen) following manufacturers' instructions. cDNA was generated from 1 μg of total RNA using HighCapacity cDNA Reverse Transcription Kit (Applied Biosystems) in a 50 μl final reaction volume, according to the manufacturers' protocol. Gene expression of Gap junction protein alpha 1/Connexin43 (GJA/ Cx43) was quantified using real-time quantitative PCR (ABI 7900 HT Sequence Detection System, Applied Biosystems). Real-time PCR reactions were performed using 1:10 dilution (1 μl/well) of each cDNA, TaqMan Universa PCR Master Mix (Applied Biosystems) and the TaqMan Gene Expression Assay for GJA/Cx43, Hs00748445_s1. Amplification of GAPDH was performed as internal control. The SDS v2.2 software was used to analyze data obtained from TaqMan Gene Expression Assays with the Comparative Ct Method (ΔΔCt algorithm). Independent experiments were performed in duplicates and repeated at least three times. Statistical significance was determined by two tailed student's t-test and p-valueb 0.05 was considered significant. Protein extraction from cultured cells The culture medium was removed and the cell monolayers were rinsed twice with cold PBS and then scrapped in cold lysis buffer (10 mM Tris, 0.5% NP-40, 150 mM NaCl, 10% Glycerol, mM EDTA, 1 mM DTT, 0.5 mM PMSF, pH 8.0) supplemented with protease inhibitor cocktail (P2714 and P8340, SIGMA-Aldrich). Cell lysates were passed through a 26 G needle several times and cellular debris was removed by centrifugation at 15,000 × g for 10 min at 4 °C. Protein

210

C. Schichor et al. / Experimental Neurology 234 (2012) 208–219

concentration was determined using the Bradford protein assay (BioRad) prior to use. Western blot analysis Cell lysates containing 15 μg of total protein were subjected to SDSPAGE with 10–20% polyacrylamide ready gels (161-1160, Bio-Rad) in Tris–glycine buffer (pH 8.3). As positive control, 15 μg rat brain extract (sc-2392, Santa Cruz) was loaded on the gel according to the manufacturers' instructions. Resolved proteins were transferred to Immun-Blot PVDF-Membranes (162-0177, Bio-Rad) in transfer buffer (48 mM Tris, 39 mM Glycine, 0.01% SDS, 20% Methanol, pH8.3). Nonspecific antibody binding was blocked by incubation of the membranes in 5% blottinggrade milk (Carl-Roth) in TTBS buffer (20 mM Tris, 136 mM NaCl, 0.1% Tween20, pH6,8), for 1 h at room temperature. Membranes were incubated overnight at 4 °C with the antibody against Cx43 (sc-9059, Santa Cruz), diluted 1:200 in TTBS with 5% blotting-grade milk. After three washing steps in TTBS, membranes were then probed for 1 h at room temperature with HRP-conjugated secondary antibody (A0545, SIGMA-Aldrich) diluted 1:10.000 in TTBS with 5% blotting-grade milk, and washed again three times. HRP-conjugated antibody against β-Actin (sc-47778, Santa Cruz) was used (1:20,000) as an internal control. Signals were visualized using the Immun-Star HRP Chemiluminescent Kit (170–5041, Bio-Rad) and by light-sensitive imaging film (34089, Pierce). Immunofluorescence To determine the presence of connexin 43 (Cx43), hMSCs and glioma cells (U87) were fixed with buffered formaldehyde (3%), spinned down on a microscope slide, air-dried and subsequently permeabilized by submersion in 0.2% Triton X-100 in PBS for 2 min. Samples were rinsed three times with PBS, blocked with 0.2% bovine serum albumin in PBS for 15 min, and incubated with the respective affinity-purified, primary antibody (anti-Cx43, Transduction Laboratories) for 45 min. After washing, secondary antibody (anti-rabbitA488, Molecular Probes) was added, and samples were incubated for 30 min at room temperature. Cells were observed by fluorescence microscopy. Patterns of cellular interactions (syncytium formation) Fluorescent dye transfer via gap junctions (functional syncytium formation) Functional relevance of gap junction formation was shown by fluorescent dye transfer. Briefly, hMSCs or glioma cells were incubated with 0.1 μM calcein AM (green:C3099 or red:C34851, Molecular Probes) for 15 min. This fluorescent dye ester is taken up by the cells and subsequently chemically modified intracellularly by esterases. Thus, the dye becomes membrane-impermeant and cannot leave intact cells on short-term except via gap junctions. Extracellular calcein was removed by three washing steps. Calcein labeled cells were then added to astrocytes or glioma cells, which have been prestained with the opposite fluorescence (DiI or DiO, respectively). After co-incubation, dye transfer was documented by fluorescencemicroscopy. Nuclear staining was performed using Hoechst 33342 (H3570, Invitrogen). AGA- and Carbenoxolone—mediated inhibition of fluorescent dye transfer 18-alpha-glycyrrhetinic acid (AGA)-treatment of different cell types (glioma cell lines, hMSCs) was performed to prevent the Cx-mediated gap-junctional permeability. In brief, cells were treated with increasing concentrations of AGA (SIGMA-Aldrich). AGA was prepared as 1 mM stocks in dimethylsulfoxide (DMSO) and diluted as needed. The final DMSO concentration was always less than 0.1%. In addition to AGAmediated inhibition, Carbenoxolone (Cbx)-treatment of different celltypes (glioma cell lines and hMSCs) was performed to prevent the Cxmediated gap-junctional permeability (Beraneck et al., 2009). Cells were treated with 200 μM Cbx (SIGMA-Aldrich).

Cell fusion in co-culture systems (structural syncytium formation) For detection of fusion events in co-cultures of different cell types, four different labeling protocols were used, as follows: Vybrant DiI/DiO labeling. Before co-culture, cells were labeled with two different fluorescent dyes—Vybrant DiI and Vybrant DiO (Molecular Probes) according to the manufacturers' instructions. Briefly, cells were incubated with 2.5 μl/ml Vybrant DiO (DiI, respectively) cell solution for 30 min, protected from light at 37 °C in a 5% CO2-humidified atmosphere, followed by PBS washing steps. These fluorochromes are lipophilic carbocyanine dyes binding to intracellular phospholipid bilayer membranes. They are non-translocating, low cytotoxic dyes, which are considered to be non-gap-junction permeable due to their high molecular weight. Vybrant DiI and DiO have different absorption and fluorescence emission maxima separated by about 65 nm enabling two-color labeling (absorption/fluorescence emission maxima; DiO: 484/501 nm and DiI: 549/566 nm). After labeling, cells were cocultured for up to three days and then examined with a Zeiss Axiovert25 inverted fluorescence microscope. Nanocrystal labeling (Qtracker). Qtracker (Quantum Dot Corp.) is a nanocrystal labeling marker, incorporated into vesicles of the cytoplasm. hMSCs were labeled according to the manufacturers' instructions with Qtracker 655, glioma cells with Qtracker 525. Because nanocrystals are large molecules, which can only be introduced into the cells by pretreatment, they are not transferred between cells. Structural syncytia (fused cells) were detected by the presence of both fluorescence signals in the same cell. Dextran tetramethylrhodamine. The loading of dextran tetramethylrhodamine (MW 10 kDa) neutral (Molecular Probes) was performed by scrape-loading of subconfluent adherent cells as described previously (el-Fouly et al., 1987). Cells were labeled with dextran tetramethylrhodamine (MW 10 kDa; red, cytoplasmic localization) before adding them to DiO-labeled (green, incorporated into plasma membrane) cells. High-molecular weight dextran cannot be transferred between cells by connexin mediated gap junctional transfer. Syncytium formation was detected by the presence of both colors (DiO and dextran tetramethylrhodamine) in the same cell. Mitochondrial subcellular compartment staining. U87 glioma cells were transfected with a baculovirus directing the expression of a red fluorescent protein targeted to the mitochondria (Organelle Lights™ Mito-OFP, Invitrogen). Cells were plated at a density of 2 × 10 5 cells/ cm 2 and allowed to grow in complete medium for 24 h. Medium was discarded and cells were transfected with 5.5 ml of transduction solution and incubated at room temperature in the dark for 3 h according to the manufacturers' instructions. After incubation, the transduction solution was discarded and cells were incubated at 37 °C and 5% CO2 in 6 ml cell culture medium completed with a 1:1000 dilution of the enhancer provided with the kit. After 2 h the enhancer solution was aspirated and replaced with 10 ml complete cell culture medium. Cells were then incubated for 16 h prior to trypsinization to allow for expression of the red mitochondrial fluorescent protein. Functional assays Proliferation U373-GFP cells and MSCs, pre-stained with DiI, were cultured in 12-well-plates for up to 4 days in monoculture, indirect co-culture and direct co-culture. Cells were seeded at a density of 21.000 hMSCs and 9000 U373-GFP-cells per well. Cell counts were performed on day 1 and day 4 after seeding using a Neubauer chamber and trypan blue staining to exclude dead cells.

C. Schichor et al. / Experimental Neurology 234 (2012) 208–219

211

Generation of U373-GFP and MSC co-cultured spheroids Generation of spheroids was performed by co-culturing U373 expressing GFP and DiI-labeled hMSC (ratio 1:1) with the hanging drop method (Corcoran and Del Maestro, 2003; Timmins and Nielsen, 2007; Werbowetski et al., 2004). Briefly, 3000 cells were cultured in a hanging drop (30 μl) for 72 h. The evolved spheroids were picked under sterile conditions with a microscope for further studies. Migration of U373 and hMSCs out of spheroids For migration experiments 96-well plates (TPP) were coated with Laminin (2 μg/cm 2, SIGMA-Aldrich) for 2 h at 37 °C and washed twice with PBS. The generated spheroids were placed in the middle of each well in growth media containing either AGA (50 μM), Cbx (200 μM) or no inhibitor. Migration was assayed taking images after 1 h and 46 h, measuring the area covered by migrating cells using WegaImage Viewer Software (M.O.S.S. Computer Grafik Systeme Taufkirchen). Experiments were performed in triplicates and results are shown as change in μm 2 [relative migration]. Statistical significance was determined by two tailed student's t-test; p-value b 0.03 was considered significant. Results Whole-genome gene expression data identify Cx43 as predominantly expressed gap junction protein in hMSC and U87 glioma cells In order to detect differentially regulated connexins in hMSCs and glioma cells, we analyzed whole genome expression data generated on Illumina HumanWG6 v3.0 expression beadchips. We identified Cx43 as predominantly expressed connexin in all groups, even higher in hMSC than in glioma cells. Indirect co-culturing of hMSCs and glioma cells led to an increased expression of Cx43 in U87 glioma cells but not in hMSCs (Fig. 1A). Furthermore, we found evidence for weak expression of Cx47/Cx46.6, especially in hMSC single culture and in hMSC in indirect co-culture with U87. The two Illumina probes measuring Cx45 indicated expression, although at extremely different levels. Expression signals for other connexins like Cx26, Cx30.2/Cx29/ Cx31.3, Cx30.3, Cx31, Cx32 and Cx50 were detectable, although these signals were weak and close to the background. No evidence for expression was found for Cx30, Cx31.1 and Cx46 (all proteincoding transcripts covered by the array) and Cx25, Cx37 and Cx59/ Cx58. Since not all annotated transcripts for these three genes are covered by the Illumina HumanWG6 v3.0 array probes, the expression of uncovered transcript variants of these genes cannot be excluded. The expression values are summarized in the Supplementary Material (Table S3). mRNA expression of Connexin 43 changes significantly in co-cultured glioma and hMSC cells Aiming at the differential regulation of Cx43 expression, glioma cells and hMSC were put in either monoculture, indirect, direct or conditioned media co-culture. The mRNA level of Cx43 was evaluated in the cells with qPCR and is presented in Fig. 1B. The mRNA derived from direct co-culture setting could not be analyzed for each cell type separately due to assumed effects of syncytium formation. However, total Cx43 mRNA expression in direct co-culture increased significantly in relation to hMSC monoculture and decreased significantly in relation to U87 monoculture. Immunofluorescence and Western blotting confirm the expression of Connexin 43 on glioma cells and hMSCs As Cx43 was predicted by microarray gene expression data as predominantly expressed, we focused on this connexin for further

Fig. 1. A Whole genome expression data showed predominant expression of Cx43 in hMSCs and U87 glioma cells in monoculture (MSC, U87) and increased expression in U87 indirect co-culture (MSC_IC, U87_IC) (3 samples per group). Other connexins tested were weakly or not expressed (Table S3). B mRNA expression of Cx43 in hMSC and U87 cells grown alone or in indirect (IC), direct (DC, ratio 1:1) or conditioned media (CM) culture. Cx43 expression significantly decreases in co-cultured MSC cells, whereas it significantly increases in U87-MG co-cultures when compared to monocultures (black bars). The experiments were performed in triplicates, the error bars represent the standard error of the mean, * p-value b 0.05 was considered significant.A Whole genome expression data showed predominant expression of Cx43 in hMSCs and U87 glioma cells in monoculture (MSC, U87) and increased expression in U87 indirect coculture (MSC_IC, U87_IC) (3 samples per group). Other connexins tested were weakly or not expressed (Table S3). B mRNA expression of Cx43 in hMSC and U87 cells grown alone or in indirect (IC), direct (DC, ratio 1:1) or conditioned media (CM) culture. Cx43 expression significantly decreases in co-cultured MSC cells, whereas it significantly increases in U87-MG co-cultures when compared to monocultures (black bars). The experiments were performed in triplicates, the error bars represent the standard error of the mean, * p-value b 0.05 was considered significant.

studies. Immunofluorescence staining showed that glioma cell lines as well as hMSCs express connexin 43 (Fig. 2A). When total protein extracts of hMSCs and of glioma cell lines were analyzed by immunoblotting, Cx43 could be detected in all samples (Fig. 2B).

Dye-transfer experiments reveal gap-junctional coupling between MSCs and glioma cells Connexins are necessary for transfer of different non-membranepermeable molecules, functionally assessed in fluorescent dyetransfer experiments (e.g. calcein). In these experiments, hMSCs were loaded with a small, fluorescent, non-membrane-permeable dye (calcein, green). The glioma cell lines U87 and U373 were labeled with a different fluorescent dye (Vybrant DiI), and pre-plated before

212

C. Schichor et al. / Experimental Neurology 234 (2012) 208–219

A

B

Fig. 2. A Immunofluorescence detection of Cx43 in glioma cells and hMSC. B Western Blot of hMSCs and different glioma cell lines for expression of Cx43. Cx43 could be detected in all samples, with stronger expression in hMSCs and U87 glioma cells. Rat brain extracts were used as positive control.

addition of calcein-loaded cells. Within 2 h, calcein-loaded hMSCs had attached and formed functional contacts to glioma cells as seen by transfer of calcein (Fig. 3). Dye transfer was time dependent and detectable in co-cultivated cells after 2 h. As a control, we performed calcein transfer experiments with astrocytes and calcein-loaded hMSCs, U87 or U373. A calcein transfer could be detected from all cell lines tested to astrocytes (Fig. 4), showing their ability to form gap junctions.

Calcein transfer can successfully be inhibited by AGA and Cbx By inhibition of gap junctional coupling with AGA and Cbx, a dependency on gap junction formation for calcein transfer could be shown. Both chemicals are known to inhibit gap junctional coupling unspecifically and reversibly. For AGA, this was shown by diminished calcein transfer in a dose-dependent manner (Fig. S1). AGA as well as Cbx effectively inhibited dye transfer from calcein-loaded glioma cells to hMSCs (Fig. 5), whereas non-treated cells showed a strong exchange of the dye (Fig. 3).

Tracking of fluorescence labeled cells in co-culture shows structural syncytium formation between glioma cells and hMSCs in vitro Cell fusion events between glioma cells and hMSCs were studied using four different labeling techniques. The basic principle behind all these techniques is that large fluorescent dyes are introduced into one cell-type, whereas the other one is labeled differently. If the fluorescent molecule is too large for a transfer through gap junctional coupling, fusion events can be assumed in cells which contain two fluorescent dyes. Vybrant DiI and DiO fluorochromes are lipophilic carbocyanine dyes that bind to intracellular phospholipid bilayer membranes. They are non-translocating dyes, which are considered to be nongap junction permeable, due to their high molecular weight. After fluorescent labeling of the cells with isolated Vybrant DiO and DiI, hMSCs and glioma cells formed spontaneous two-color positive cells in co-culture (structural syncytium formation, Figs. 6d and e). Using this technique, we observed cell fusion events after 72 h of co-culture. Staining with high-molecular weight dextran tetramethylrhodamine (MW 10 kDa) is only possible by the use of scrape-loading of

C. Schichor et al. / Experimental Neurology 234 (2012) 208–219

213

Fig. 3. Functional syncytium formation of hMSCs with glioma cells: Fluorescent dye (Calcein, green) was transferred from pre-loaded hMSCs to U87 or U373 glioma cells. Transfer of calcein from one cell to another via gap junctions resulted in orange cells, all indicated with white arrows. Shown are hMSCs in green; U87 and U373 (red); nuclei (blue). No transfer of calcein to distant cells (arrow heads).

subconfluent adherent cells. By the use of this assay, early fusion events could not be detected sooner than 48 h of co-culture (Figs. 6a–c). Qtracker is a nanocrystal labeling marker, which is incorporated into vesicles of the cytoplasm after pre-treatment of cells. In our setting, we could detect syncytium formation in double stained (Qtracker 655 and 525) cells (Fig. 6f). Cell fusion events occurred in all glioma cell lines with hMSCs. Mitochondrial subcellular compartment staining of glioma cells by transfection with a baculovirus directs the expression of a red fluorescent protein in mitochondria. In this setting, we also detected double stained cells after 72 h, which could only be existent after cell fusion events (Fig. 7). Co-culture with hMSCs inhibits glioma cell proliferation Whereas hMSC proliferation was not significantly changed due to co-culturing with glioma cells (direct or indirect), glioma cell

proliferation was inhibited significantly. Both, indirect as well as direct co-culturing led to a reduced proliferation rate, but this effect was even more pronounced if cells were in close contact to each other, being able to form syncytia (Fig. 8). Co-culturing in mixed spheroids enhances migration of glioma cells and hMSCs Migration of monocultures compared to mixed cell conglomerates revealed significant differences: as expected, glioma cells showed a higher migratory capacity compared to hMSCs in the monoculture experiment. In contrast, migration out of mixed spheroids, was markedly enhanced, compared to glioma cells or hMSCs alone (Fig. 9A). Surprisingly, at the edge of the migration front of the fastest migratory cells, we did not exclusively detect non-fused glioma cells but also fused cells and even non-fused hMSCs, which migrated much faster than hMSCs in a monoculture experiment.

214

C. Schichor et al. / Experimental Neurology 234 (2012) 208–219

Fig. 4. Glioma cells (U87 and U373) as well as hMSCs were able to form functional syncytia with astrocytes. Calcein-transfer to astrocytes (red) resulted in orange cells, all indicated with white arrows. Shown are astrocytes in red; hMSC, U87 and U373 in green. No transfer of calcein to distant cells (arrow heads).

Inhibition of gap junctional coupling reduces migration capacity of cells out of multicellular spheroids in mixed cell systems Inhibiting gap junction formation with either AGA or Cbx reduced migration capacity of glioma cell-hMSC co-culture markedly. The migration behavior of hMSCs was not affected, as hMSCs alone did not migrate significantly. Glioma cells, treated with gap junction inhibitor AGA showed more than two-fold decreased migration (Fig. 9B). The enhanced migration capacity for the mixed co-cultures was also reduced more than two-fold, if treated with either gap junction inhibitor, AGA or Cbx (Fig. 9C). Discussion Gap junctional coupling has been described in most mammalian cell types, thereby coordinating their actions. Each pore is made of six structural protein subunits, known as connexins (Cx). Organized in a ring-like structure, the so-called connexon is located in the cytoplasmatic membrane (Herve et al., 2007). A functional channel is formed when a hemichannel, composed of six connexin molecules, assembles with a hemichannel from an adjacent cell. The resulting gap junction electrically couples cells by direct exchange of ions, nutrients and second messengers. An intense intercellular communication between grafted neural stem cells and recipient brain tissue via Cx43 mediated gap-junctional coupling was described by Jaderstad et al. (Jaderstad et al., 2010). The authors found that gap junction

formation was fast and led to a neural stem cell mediated protection of neurons. Also, hMSCs have been shown to be connected by adherent cell junction systems most dominantly expressing Cx43 (Herve et al., 2007; Wurmser and Gage, 2002). The direction of their differentiation was determined by transport of intracellular components from cardiomyocytes to hMSCs (Plotnikov et al., 2008). Of more than 10 different connexins characterized, the most widely expressed in malignant gliomas is Cx43 (Caltabiano et al., 2010; Zhang et al., 2003). Zhang et al. showed that glioma cells communicate with normal astrocytes via gap junctional coupling as monitored by transfer of fluorescent dyes (Zhang et al., 2003), which is in line with our experiments. Guided by microarray gene expression analysis, we identified Cx43 as the key regulated connexin in glioma-hMSC cocultures. Expression of Cx43 was confirmed on mRNA as well as on protein level. The functionality of gap junctional coupling was elucidated by transfer of fluorescent calcein and its complete inhibition via AGA or Cbx (Fig. 5). Both are known to inhibit gap junctional transfer. Despite the obvious effects, these pharmacological gap junction inhibitors are neither specific for certain connexins (Davidson and Baumgarten, 1988; Hanstein et al., 2010; Spray et al., 2002), nor selective for gap junction blocking, because Cbx might also have an effect on Ca 2 + channels (Vessey et al., 2004). Migration into the adjacent brain is a hallmark of malignant gliomas, resulting in diffuse infiltrating and recurring tumors after surgical resection. We have observed not only a massively enhanced migration of glioma cells from spheroids of co-cultured glioma cells with hMSCs compared to

C. Schichor et al. / Experimental Neurology 234 (2012) 208–219

215

Fig. 5. Gap junction inhibition in glioma cell-hMSC co-culture. Transfer of red Calcein from U373 glioma cells to DiO-stained hMSCs (green) as well as green Calcein from hMSC to DiI-stained U87 glioma cells (red). Cells were treated with 50 μM AGA (upper panel) or 200 μM Cbx (lower panel). No transfer of Calcein could be detected in cells treated with gap junction inhibitors (white arrows), whereas a strong exchange occurred in the untreated cells (Fig. 3). Images were taken after 4 h of co-culture.

monocultured spheroids, but also an increased migration of the adjacent co-cultured hMSCs alone. The amount of migration enhancement was diminished significantly, if gap junction inhibitors were added to the co-cultured spheroids, suggesting a considerable contribution of gap junctional coupling to this phenomenon. Gap junctional coupling resulted in transfer of fluorescent calcein from one cell type to another already after 1 h. After 2 h, calcein spread to almost all cells in the culture dish, depending on confluency. In contrast, we detected cell fusion events after 48–72 h. Cell fusion in general is a highly specific phenomenon (Sapir et al., 2008). Fertilization in mammalians is

perhaps the most well-known form of cell–cell fusion. In the mammalian placenta, trophoblasts fuse to form a syncytial layer (the syncytiotrophoblast). Macrophages can differentiate and fuse to form two multinucleated types of cells, osteoclasts and giant cells (Anderson, 2000). Bone marrow cells have been shown to spontaneously fuse with embryonic stem cells, thereby adopting the phenotype of the recipient. This was interpreted as dedifferentiation or transdifferentiation (Terada et al., 2002). Stem cell fusion has emerged as an unexpected and complicated mechanism in tissue regeneration strategies. Fusion of bone marrow derived cells with

216

C. Schichor et al. / Experimental Neurology 234 (2012) 208–219

A

B

D

C

E

F

Fig. 6. Structural syncytium formation, detected by different techniques (a–c) Rhodamine loading: hMSC (red), loaded with Rhodamine (arrow head) and DiO-stained glioma cells (U87, green): after 3 h, no fusions between hMSC and glioma cells can be detected. (b,c) large, multinucleated fused cells can be detected (yellow, fat arrow), DiO-stained glioma cells (green, small arrow). (d,e) Vybrant Dye labeling (DiI/DiO): detection of fusions with fluorescent DiI/DiO staining: fused, multinucleated cell (yellow, fat arrow), DiO-labeled (green) glioma cell (U373, U87 respectively) (small arrow), DiI-labeled (red) hMSC (arrowhead). (f) Qtracker labeling: Qtracker 655-labeled red hMSC (arrow head), fused cells (green and red). Fused cells: fat arrows, hMSC: arrow heads, glioma cells: small arrows.

Fig. 7. Structural syncytium formation, detected by staining with organelle light/Vybrant DiO: Mitochondria of U87 cells were labeled with a red fluorescent protein (RFP) and membranes of hMSC with a green fluorescent dye (DiO). No detection of fused cells (merged) after 48 h in co-culture (arrow head), whereas after 72 h a clear fusion of glioma cells with hMSC could be shown (white arrow).

C. Schichor et al. / Experimental Neurology 234 (2012) 208–219

217

12

10

8

6

4

2

0

Fig. 8. Co-culture with hMSCs inhibits glioma cell proliferation. Glioma cells have been seeded with hMSCs and have been counted on day 1 and day 4. After 4 days, growth of hMSCs had not been altered markedly whereas proliferation of U373 was significantly reduced in the co-culture setting. This effect was even more pronounced, if cells were in close contact to each other, being able to form syncytia. * p-value b 0.03 was considered significant.

cells at the site of transplantation like neurons, liver cells and cardiomyocytes (Alvarez-Dolado et al., 2003) has been considered to be responsible for transdifferentiation of the transplanted cells (Wurmser and Gage, 2002). Chen and Olson showed that neural stem cells in culture possess the ability to generate fused polyploidal cells in culture (Chen and Olson, 2005). A general requirement for cell fusions is that the two fusing membranes are in close contact. This is accomplished by receptor–ligand interactions as described in virus–cell fusions and in macrophage–macrophage fusions. However, proteins mediating stem cell fusion are not yet identified. Macrophages or other bone marrow-derived cells showed signs of fusion with metastasizing tumor cells in animal models as well as in myeloma patients (Pawelek and Chakraborty, 2008). Whereas fusion of normal cells in the body is tightly controlled, fusion events with tumor cells might produce up to 99% of dead or quiescent cells, but also 1% of proliferating cells with new, unwanted properties like increased proliferation rate, metastatic potential, drug resistance or resistance to apoptosis (Duelli and Lazebnik, 2003). In this study, we found rates of fusion events markedly increased in cells grown in spheroids, compared to cells grown in monolayers. On the one hand, spheroids represent inherent features of the underlying glioma much better than monolayer cultures. On the other hand, intense inter-cellular contacts in spheroids might enhance cell-fusion as well as gap-junction mediated cross-talk. In hMSCs, cell fusion was considered to be an alternative mechanism to cell reprogramming by transdifferentiation, leading to the generation of hybrid cells with donor cell origin and simultaneous expression of recipient cell markers and characteristics (Garbade et al., 2005). These authors detected spontaneous fusion of hMSCs and cardiomyocytes in vitro by Vybrant DiI and DiO staining. Application of this technique to co-cultures of hMSCs and glioma cells showed fusion events in our experimental setting. Fusion of cells has been investigated by a variety of other methods. As described by Murasawa et al., endothelial cells show signs of fusion, detected by double Quantum-dot stained cells (Murasawa et al., 2005). This assay for detection of cell fusion events was already used in coculture experiments of endothelial progenitor cells and myoblast cells (Pisati et al., 2007). Thus, we confirmed fusion events of glioma cells and MSCs. Nevertheless, further investigations are necessary in order to distinguish complete fusion of two cells and formation of

multinucleated cells at genetic level from partial fusion events, where nuclei may remain separated. The concept of partial heterocellular fusion was thought to provide a description for the mere transfer of cytoplasmic components without incorporation of the donor nucleus into the acceptor cell as investigated for the fusion between cardiomyocytes and bone marrow-derived cells (Alvarez-Dolado et al., 2003). Our results show that both, cell-fusion and gap junction mediated crosstalk, are present in hMSCs and glioma cells in vitro. These fusion and crosstalk events could have beneficial effects on cellbased hMSC-mediated tumor therapies (Hamada et al., 2005; Lawler et al., 2006), eventually accelerating delivery and improving therapeutic efficiency. On the other hand, unwanted effects should be taken into account. Kim et al. described successful isolation of MSCs contributing to malignant gliomas in an experimental murine tumor model (Kim et al., 2011). Hence, the conclusion might be drawn that hMSC recruitment from the patients' bone marrow towards malignant gliomas could eventually lead to enhanced migration and invasion via syncytium formation. It cannot be excluded that certain hMSC properties might be misused by the tumor cells, which could be beneficial for tumor cell invasion, migration or survival and as a consequence may even lead to tumor progression. Although encouraging concepts of hMSC-based therapies gain attention in the scientific community, the risks of cell transplantation must not be underestimated. Recently, Snyder et al. thoroughly discussed current drawbacks of hMSC-transplantations due to the observation of unintended hMSC cell masses in non-neoplastic, inflammatory brain lesions after transplantation (Snyder, 2011). Future cell therapy concepts, which even augment a putative hMSC-population in malignant gliomas via exogenous administration, will therefore have to take these considerations into account and make sure that tumor treatment does not lead to the exact opposite. In our opinion, this study highlights a crucial consequence that coculture experiments should always be controlled for serious influences of structural or functional syncytium formation. On the other hand, our experimental approach suggests that multinucleated cells in cancer could be generated by fusion events of cells of different origin. Subsequent studies can now be conducted, focusing on proliferative and migratory properties of isolated, multinucleated fused cells, compared with multinucleated cells, present in malignant gliomas.

218

A

C. Schichor et al. / Experimental Neurology 234 (2012) 208–219

45 40 35 30

syncytium leads to intense intercellular cross-talk. Furthermore, hMSCs form fusion-mediated structural syncytia with malignant glioma cells. These phenomena should be taken into account, if experimental data on glioma cell–stem cell interaction have to be interpreted and future cell based treatment strategies are developed. Supplementary materials related to this article can be found online at doi:10.1016/j.expneurol.2011.12.033.

25 20

Acknowledgments 15 10 5 0

This work was supported by the EU ERA net bilateral INREMOS project on Systems Biology Tools development for cell Therapy and Drug Development/SYSTHER (Con. No.: 3211-06-000539) (2006–2011) funded by the German and Slovenian Federal Ministries of Education and Research and by the Deutsche Forschungsgemeinschaft (DFG) SFB824/B2. Parts of this work are part of the doctoral thesis of Benjamin Korte.

B 45

References

40

Alvarez-Dolado, M., Pardal, R., Garcia-Verdugo, J.M., Fike, J.R., Lee, H.O., Pfeffer, K., Lois, C., Morrison, S.J., Alvarez-Buylla, A., 2003. Fusion of bone-marrow-derived cells with Purkinje neurons, cardiomyocytes and hepatocytes. Nature 425, 968–973. Anderson, J.M., 2000. Multinucleated giant cells. Curr. Opin. Hematol. 7, 40–47. Beraneck, M., Uno, A., Vassias, I., Idoux, E., De, W.C., Vidal, P.P., Vibert, N., 2009. Evidence against a role of gap junctions in vestibular compensation. Neurosci. Lett. 450, 97–101. Caltabiano, R., Torrisi, A., Condorelli, D., Albanese, V., Lanzafame, S., 2010. High levels of connexin 43 mrna in high grade astrocytomas. Study of 32 cases with in situ hybridization. Acta Histochem. 112, 529–535. Chen, E.H., Olson, E.N., 2005. Unveiling the mechanisms of cell–cell fusion. Science 308, 369–373. Choi, S.A., Lee, J.Y., Wang, K.C., Phi, J.H., Song, S.H., Song, J., Kim, S.K., 2012. Human adipose tissue-derived mesenchymal stem cells: characteristics and therapeutic potential as cellular vehicles for prodrug gene therapy against brainstem gliomas. Eur. J. Cancer 48 (1), 129–137. Corcoran, A., Del Maestro, R.F., 2003. Testing the “go or grow” hypothesis in human medulloblastoma cell lines in two and three dimensions. Neurosurgery 53, 174–184. Davidson, J.S., Baumgarten, I.M., 1988. Glycyrrhetinic acid derivatives: a novel class of inhibitors of gap-junctional intercellular communication. structure–activity relationships. J. Pharmacol. Exp. Ther. 246, 1104–1107. Djouad, F., Plence, P., Bony, C., Tropel, P., Apparailly, F., Sany, J., Noel, D., Jorgensen, C., 2003. Immunosuppressive effect of mesenchymal stem cells favors tumor growth in allogeneic animals. Blood 102, 3837–3844. Duelli, D., Lazebnik, Y., 2003. Cell fusion: a hidden enemy? Cancer Cell 3, 445–448. el-Fouly, M.H., Trosko, J.E., Chang, C.C., 1987. Scrape-loading and dye transfer. A rapid and simple technique to study gap junctional intercellular communication. Exp. Cell Res. 168 (2), 422–430. Garbade, J., Schubert, A., Rastan, A.J., Lenz, D., Walther, T., Gummert, J.F., Dhein, S., Mohr, F.W., 2005. Fusion of bone marrow-derived stem cells with cardiomyocytes in a heterologous in vitro model. Eur. J. Cardiothorac. Surg. 28, 685–691. Hamada, H., Kobune, M., Nakamura, K., Kawano, Y., Kato, K., Honmou, O., Houkin, K., Matsunaga, T., Niitsu, Y., 2005. Mesenchymal stem cells (Msc) as therapeutic cytoreagents for gene therapy. Cancer Sci. 96, 149–156. Hanstein, R., Zhao, J.B., Basak, R., Smith, D.N., Zuckerman, Y.Y., Hanani, M., Spray, D.C., Gulinello, M., 2010. Focal inflammation causes carbenoxolone-sensitive tactile hypersensitivity in mice. Open Pain J. 3, 123–133. Herve, J.C., Bourmeyster, N., Sarrouilhe, D., Duffy, H.S., 2007. Gap junctional complexes: from partners to functions. Prog. Biophys. Mol. Biol. 94, 29–65. Jaderstad, J., Jaderstad, L.M., Li, J., Chintawar, S., Salto, C., Pandolfo, M., Ourednik, V., Teng, Y.D., Sidman, R.L., Arenas, E., Snyder, E.Y., Herlenius, E., 2010. Communication via gap junctions underlies early functional and beneficial interactions between grafted neural stem cells and the host. Proc. Natl. Acad. Sci. U. S. A. 107, 5184–5189. Kamijo, M., Haraguchi, T., Tonogi, M., Yamane, G.Y., 2006. The function of connexin 43 on the differentiation of rat bone marrow cells in culture. Biomed. Res. 27, 289–295. Keating, A., 2006. Mesenchymal stromal cells. Curr. Opin. Hematol. 13, 419–425. Kim, S.M., Kang, S.G., Park, N.R., Mok, H.S., Huh, Y.M., Lee, S.J., Jeun, S.S., Hong, Y.K., Park, C.K., Lang, F.F., 2011. Presence of glioma stroma mesenchymal stem cells in a murine orthotopic glioma model. Childs Nerv. Syst. 27, 911–922. Lawler, S.E., Peruzzi, P.P., Chiocca, E.A., 2006. Genetic strategies for brain tumor therapy. Cancer Gene Ther. 13, 225–233. Liu, S., Ginestier, C., Ou, S.J., Clouthier, S.G., Patel, S.H., Monville, F., Korkaya, H., Heath, A., Dutcher, J., Kleer, C.G., Jung, Y., Dontu, G., Taichman, R., Wicha, M.S., 2011. Breast cancer stem cells are regulated by mesenchymal stem cells through cytokine networks. Cancer Res. 71, 614–624. Murasawa, S., Kawamoto, A., Horii, M., Nakamori, S., Asahara, T., 2005. Niche-dependent translineage commitment of endothelial progenitor cells, not cell fusion in general, into myocardial lineage cells. Arterioscler. Thromb. Vasc. Biol. 25, 1388–1394.

35 30 25 20 15 10 5 0

C 45 40 35 30 25 20 15 10 5 0

Fig. 9. Migration of glioma cells and mesenchymal stem cells from bone marrow (MSC) out of spheroids. Spheroids were grown from either U373-GFP, MSC or a mixed population with and without gap junction inhibitors. A. Both cell types showed significantly enhanced migration in mixed spheroid migration experiments. B. For U373, only AGA led to a marked inhibition of migration out of monoculture spheroids, but only little effect was seen for hMSCs, as these cells have only little migration potential. C. Nevertheless, a significant antimigratory effect of the inhibitors was observed in the mixed spheroid migration assay.

Conclusion Our study shows that hMSCs generate predominantly connexin 43 mediated gap junctions with glioma cells. This so-called functional

C. Schichor et al. / Experimental Neurology 234 (2012) 208–219 Otsu, K., Das, S., Houser, S.D., Quadri, S.K., Bhattacharya, S., Bhattacharya, J., 2009. Concentration-dependent inhibition of angiogenesis by mesenchymal stem cells. Blood 113, 4197–4205. Pawelek, J.M., Chakraborty, A.K., 2008. Fusion of tumour cells with bone marrowderived cells: a unifying explanation for metastasis. Nat. Rev. Cancer 8, 377–386. Pisati, F., Belicchi, M., Acerbi, F., Marchesi, C., Giussani, C., Gavina, M., Javerzat, S., Hagedorn, M., Carrabba, G., Lucini, V., Gaini, S.M., Bresolin, N., Bello, L., Bikfalvi, A., Torrente, Y., 2007. Effect of human skin-derived stem cells on vessel architecture, tumor growth, and tumor invasion in brain tumor animal models. Cancer Res. 67, 3054–3063. Plotnikov, E.Y., Khryapenkova, T.G., Vasileva, A.K., Marey, M.V., Galkina, S.I., Isaev, N.K., Sheval, E.V., Polyakov, V.Y., Sukhikh, G.T., Zorov, D.B., 2008. Cell-to-cell cross-talk between mesenchymal stem cells and cardiomyocytes in co-culture. J. Cell. Mol. Med. 12, 1622–1631. Ramasamy, R., Lam, E.W., Soeiro, I., Tisato, V., Bonnet, D., Dazzi, F., 2007. Mesenchymal stem cells inhibit proliferation and apoptosis of tumor cells: impact on in vivo tumor growth. Leukemia 21, 304–310. Ricci-Vitiani, L., Pallini, R., Larocca, L.M., Lombardi, D.G., Signore, M., Pierconti, F., Petrucci, G., Montano, N., Maira, G., De, M.R., 2008. Mesenchymal differentiation of glioblastoma stem cells. Cell Death Differ. 15, 1491–1498. Sapir, A., Avinoam, O., Podbilewicz, B., Chernomordik, L.V., 2008. Viral and developmental cell fusion mechanisms: conservation and divergence. Dev. Cell 14, 11–21. Schichor, C., Birnbaum, T., Etminan, N., Schnell, O., Grau, S., Miebach, S., Aboody, K., Padovan, C., Straube, A., Tonn, J.C., Goldbrunner, R., 2006. Vascular endothelial growth factor a contributes to glioma-induced migration of human marrow stromal cells (hmsc). Exp. Neurol. 199, 301–310.

219

Snyder, E.Y., 2011. The risk of putting something where it does not belong: mesenchymal stem cells produce masses in the brain. Exp. Neurol. 230, 75–77. Sonabend, A.M., Ulasov, I.V., Tyler, M.A., Rivera, A.A., Mathis, J.M., Lesniak, M.S., 2008. Mesenchymal stem cells effectively deliver an oncolytic adenovirus to intracranial glioma. Stem Cells 26, 831–841. Spray, D.C., Rozental, R., Srinivas, M., 2002. Prospects for rational development of pharmacological gap junction channel blockers. Curr. Drug Targets 3, 455–464. Staflin, K., Honeth, G., Kalliomaki, S., Kjellman, C., Edvardsen, K., Lindvall, M., 2004. Neural progenitor cell lines inhibit rat tumor growth in vivo. Cancer Res. 64, 5347–5354. Terada, N., Hamazaki, T., Oka, M., Hoki, M., Mastalerz, D.M., Nakano, Y., Meyer, E.M., Morel, L., Petersen, B.E., Scott, E.W., 2002. Bone marrow cells adopt the phenotype of other cells by spontaneous cell fusion. Nature 416, 542–545. Timmins, N.E., Nielsen, L.K., 2007. Generation of multicellular tumor spheroids by the hanging-drop method. Methods Mol. Med. 140, 141–151. Vessey, J.P., Lalonde, M.R., Mizan, H.A., Welch, N.C., Kelly, M.E., Barnes, S., 2004. Carbenoxolone inhibition of voltage-gated ca channels and synaptic transmission in the retina. J. Neurophysiol. 92, 1252–1256. Werbowetski, T., Bjerkvig, R., Del Maestro, R.F., 2004. Evidence for a secreted chemorepellent that directs glioma cell invasion. J. Neurobiol. 60, 71–88. Wurmser, A.E., Gage, F.H., 2002. Stem cells: cell fusion causes confusion. Nature 416, 485–487. Zhang, W., Nwagwu, C., Le, D.M., Yong, V.W., Song, H., Couldwell, W.T., 2003. Increased invasive capacity of connexin 43-overexpressing malignant glioma cells. J. Neurosurg. 99, 1039–1046.

Lihat lebih banyak...

Comentários

Copyright © 2017 DADOSPDF Inc.