Brain Advance Access originally published online on June 9, 2005
Brain 2005 128(9):2200-2211; doi:10.1093/brain/awh563
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Lessons from the bone marrow: how malignant glioma cells attract adult haematopoietic progenitor cells
1 Laboratory of Molecular Neurooncology, Department of General Neurology and Hertie Institute for Clinical Brain Research, 2 Department of Internal Medicine II (Hematology), University of Tübingen, Tübingen, Germany, 3 Department of Neurosurgery, University of Erlangen and 4 Institute of Neuropathology, University of Erlangen, Erlangen, Germany
Correspondence to: Wolfgang Wick, MD, Laboratory of Molecular Neurooncology, Department of General Neurology and Hertie Institute for Clinical Brain Research, Hoppe-Seyler-Strasse 3, D-72076 Tübingen, Germany E-mail: wolfgang.wick{at}uni-tuebingen.de
 |
Summary |
|---|
 Top Summary IntroductionMaterials and methodsResultsReferences |
|---|
Stem and progenitor cells (PCs) of various lineages have become attractive vehicles to improve therapeutic gene delivery to cancers, notably glioblastoma. Here we report that adult human and murine haematopoietic PCs display a tropism for intracerebral gliomas but not for normal brain tissue in mice. Organotypic hippocampal slice culture and spheroid confrontation assays confirm a directed PC migration towards glioma cells ex vivo and in vitro. RNA interference-mediated disruption of transforming growth factor beta (TGF-ß) synthesis by the glioma cells strongly inhibits PC migration. We delineate a CXC chemokine ligand (CXCL) 12-dependent pathway of TGF-ß-induced PC migration that is facilitated by MMP-9-mediated stem cell factor cleavage in vitro. Moreover, neutralizing antibodies to CXCL12 strongly reduce PC homing to experimental gliomas in vivo. Thus, we define here the molecular mechanism underlying the glioma tropism of the probably most easily accessible PC population suitable for cancer therapy, that is, adult haematopoietic PC.
Key Words: CXCL12; hippocampal slice culture; metalloproteinase; TGF-ß
Abbreviations: CA1 = cornu ammonis 1; CA3 = cornu ammonis 3; CD117 = cluster of differentiation 117; CD34+PC = G-CSF mobilized peripheral blood CD34+ progenitor and stem cells; CFSE = carbofluorescein diacetate succinimidyl ester; CXCL12 = CXC chemokine ligand; CXCR4 = chemokine receptor 4; DG = dentate gyrus; EC = entorhinal cortex; FCS = fetal calf serum; G-CSF = granulocyte colony stimulating factor; HBSS = Hanks' balanced salt solution; lin– = lineage-depleted; MACS = magnetic activated cell sorting; MMP = matrix metalloproteinase; nu PC = bone marrow-derived murine lin–Sca1+ haematopoietic stem and progenitor and lin– stem and progenitor cells from athymic nude mice; PKB = protein kinase B; RO28-2653 = 5-biphenyl-4-yl-5-[4-(nitro-phenyl)-piperazin-1-yl]-pyrimidine-2,4,6; SC = stem cell; sca1 = stem cell antigen; SCF = stem cell factor; SDF = stromal cell-derived factor; SEM = standard error of the mean; SFM = serum-free medium; sKitL = soluble Kit ligand; TGF = transforming growth factor; VM/Dk PC = bone marrow-derived murine lin–Sca1+ haematopoietic stem and lin– stem and progenitor cells from immunocompetent VM/Dk mice
Received March 11, 2005. Revised May 3, 2005. Accepted May 17, 2005.
|
Introduction |
|---|
|
TopSummaryIntroductionMaterials and methodsResultsReferences |
|---|
Gliomas are paradigmatic for their deep infiltration of adjacent brain, precluding definitive surgical resection and limiting the efficacy of other local therapies (DeAngelis, 2001
). The benefit derived from traditional cancer therapies is low because of defects in the apoptotic machinery of glioma cells, accounting for their resistance to radiation and chemotherapy. Similarly, gene therapy has its limits in the failure of currently available viral vectors to reach infiltrating glioma cells. An easily accessible autologous cellular vector that targets disseminated glioma cells and expresses a therapeutic transgene would, therefore, represent a major step ahead in the experimental treatment of these tumours. Neural progenitor cells (PCs) from the cortex of C57BL 6 mice have been used as vehicles to deliver interleukin 4 to C6 experimental rat gliomas in Sprague-Dawley rats and to GL261 mouse gliomas in C57BL6 mice. Animals treated with PC delivering interleukin 4 lived significantly longer than animals treated with fibroblast-derived packaging cells for in vivo transfer of interleukin 4 (Benedetti et al., 2000). Furthermore, LacZ-expressing cells derived from the immortalized murine neuronal precursor cell line C17.2, injected either into an experimental CNS-1 rat glioma, the tail vein or into non-tumorous brain tissue, diffusely transmigrated the whole glioma diameter within 3 days (Aboody et al., 2000).
Adult haematopoietic PCs are a highly migratory cell population of the bone marrow (Krause et al., 2001). Primary non-transformed adult bone marrow-derived cells with neuronal properties migrate efficiently towards distant sites of brain tumour when injected intracranially (Lee et al., 2003). The homing of haematopoietic stem cells (SCs) to the bone marrow and chemotaxis involve CXCL12 [stromal cell-derived factor (SDF)-1
] acting on chemokine receptor 4 (CXCR4) expressed on the SC, interleukin 8 and non-peptide leukotriene D4 (Aiuti et al., 1997; Peled et al., 1999; Bautz et al., 2001).
In bone marrow cells, CXCL12 upregulates the expression of MMP-9. The release of soluble Kit ligand (s-KitL)/stem cell factor (SCF) from its membrane-bound form on bone-marrow stromal cells is mediated by MMP-9. Binding to its receptor, CD117 on the haematopoietic SC contributes to their mobilization from a quiescent state in the bone marrow niche (Heissig et al., 2002).
Matrix metalloproteinases (MMPs) are also essential for the invasiveness of malignant gliomas (Kondraganti et al., 2000). Furthermore, glioma cells express SCF mRNA and protein, and the receptor c-Kit (Stanulla et al., 1995). However, the impact of SCF on the growth and motility of glioma cells is unclear. CXCL12/CXCR4 interactions mediate chemotaxis of glioma cells and protect from cell death induced by serum withdrawal via phosphorylation of protein kinase B (PKB/Akt) (Zhou et al., 2002). A small molecule CXCR4 antagonist, AMD 3100, inhibited the growth of experimental glioblastomas and medulloblastomas (Rubin et al., 2003).
The molecular mechanisms underlying the phenomenon of a possible lesion- or tumour-directed migration of either primary embryonic or adult SC, primary neuronal or glial precursor cells have remained unknown. Here we characterize the possible in vitro and in vivo use of adult peripheral CD34+ human haematopoietic PC and adult bone marrow-derived murine lin–Sca1+ haematopoietic PC as vehicles for glioma therapy. We delineate a cascade for their active glioma tropism which involves two transforming growth factor (TGF)-ß-dependent pathways, (i) CXCL12 synthesis and (ii) MMP-9-mediated cleavage of SCF, facilitating the promigratory effect of CXCL12.
|
Materials and methods |
|---|
|
TopSummaryIntroductionMaterials and methodsResultsReferences |
|---|
Isolation of haematopoietic PC from mice
Nude mice were purchased from Charles River Laboratories, Sulzfeld, Germany. VM/Dk mice were originally purchased from TSE Research Center (Berkshire, UK) and are now bred in our laboratory. Nude or VM/Dk mice that were 4 to 6 weeks-old were killed. Femurs and tibiae were flushed with phosphate-buffered saline (PBS) containing 2% fetal calf serum (FCS). Bone marrow cells (106/ml) were incubated in PBS containing 2% FCS with the lineage marker antibodies from BD Biosciences Pharmingen (San Diego, CA), rat anti-mouse CD4, CD8a, CD45R/B220, Gr-1, CD11b and TER119 at 4°C for 60 min, and washed three times with PBS containing 2% FCS. Prewashed sheep anti-rat IgG magnetic beads (Dynabeads, Dynal Biotech ASA, Oslo, Norway) were added at a 4:1 ratio of beads/cells in PBS containing 2% FCS and incubated on a rotating platform for 45 min at 4°C. Beads and attached cells were magnetically removed. The remaining lineage-depleted (lin–) cells were MACS-isolated with anti-SC antigen (Sca)1+-beads (Miltenyi Biotech, Bergisch Gladbach, Germany). Flow cytometry revealed that Lin– Sca1+ cells were between 97.5 and 98.5% positive for Sca1 (FITC-conjugated antibody) and >99% positive for c-kit (PE-conjugated antibody, BD Biosciences Pharmingen). Also, <4% of the isolated Lin–Sca1+ cells expressed CD4, CD8, CD11b, B220 or Gr-1. Isolated cells were used directly for the in vitro experiments. For the in vivo experiments, isolated cells were pretreated with murine IL-3 (20 ng/ml), IL-6 (50 ng/ml) and SCF (50 ng/ml) (R&D Systems, Minneapolis, MN) for 24 h.
Isolation of human CD34+ haematopoietic PC
After informed consent, peripheral blood mononuclear cells were obtained. Mononuclear cells were separated by Ficoll density gradient centrifugation. CD34+ cells were isolated using immunomagnetic microbeads (MACS system; Miltenyi Biotec). The purity of CD34+ cells isolated by positive selection was between 95.1 and 98%.
Cell culture and primary cells
All glioma cell lines were kindly provided by Dr N. de Tribolet (Lausanne, Switzerland). In contrast to other LN-229 sublines, LN-229 glioma cells cultured in our laboratory exhibit wild-type p53 status (Wischhusen et al., 2003) and are designated LNT-229. The generation of LNT-229 sublines depleted from TGF-ß1 and TGF-ß2 by RNA interference has been described (Friese et al., 2004). T189 and T204 are short term cultures from glioblastoma biopsy samples (Bähr et al., 2003). Staining of glioma cells with VybrantTM carbofluorescein diacetate succinimidyl ester (CFSE) (Molecular Probes, Invitrogen, Carlsbad, CA), staining of haematopoietic SC with PKH26 or DiI and staining of the DNA on frozen sections with 4',6-diamidino-2-phenylindole (DAPI) were performed according to manufacturer's protocols.
Chemotaxis assay
Lin– cells were suspended in RPMI containing 10% FCS and incubated in tissue culture flasks at 37°C for 1 h for resensitization of potentially desensitized chemokine responses (Bowman et al., 1998). Cells that had been washed three times with PBS were placed into a transwell chemotaxis chamber (24 wells, 5 µm pore size; Corning Inc., Acton, MA) (Wright et al., 2002). Chemokines or conditioned media were added to the bottom well. Murine lin– cells (105 per top well) or 105 human CD34+ or 5 x 105 Lin–Sca1+ cells per top well were added in 200 µl serum-free medium (SFM) and allowed to migrate for 4 h at 37°C. The following antibodies, chemicals and proteins were used in these assays: anti-human/mouse CXCL12; anti-human SCF, monoclonal anti-human CXCR4 (clone 12G5) (all R&D Systems); anti-human TGF-ß2, 5-biphenyl-4-yl-5-[4-(nitro-phenyl)-piperazin-1-yl]-pyrimidine-2,4,6 (RO28-2653) (15 µM) (Roche Diagnostics, Penzberg, Germany); and human recombinant CXCL12, SCF (R&D Systems), active MMP-9 (Oncogene; Merck Biosciences GmbH, Darmstadt, Germany) and TGF-ß2 (PeproTech, London, UK).
Quantification of migrated cells
The principle of the CyQuant Assay (Molecular Probes) is a proprietary green fluorescent dye, CyQuant GR dye, which exhibits fluorescence when bound to cellular nucleic acids. The CyQuant Assay has a linear detection range from 50 or fewer to at least 5 x 104 cells in 200 µl volumes using a single dye concentration. A reference standard curve for haematopoietic PC was established for converting sample fluorescence values into cell numbers. Migrated cells were carefully collected from bottom wells, centrifuged and resuspended in 200 µl freshly prepared working solution containing CyQuant cell lysis buffer and CyQuant GR dye. The samples were transferred to a microplate and fluorescence was measured at 480 nm excitation and 520 nm emission. Samples without cells as well as samples with defined cell numbers were included as controls.
Organotypic entorhino-hippocampal slice cultures
Slice cultures were prepared and maintained as described (Ullrich et al., 2001; Eyupoglu et al., 2003). Seven-day-old Wistar rats were used for explantation. After decapitation, the brains were rapidly removed and placed into ice-cold preparation medium containing Hanks' balanced salt solution (HBSS) (Invitrogen, Karlsruhe, Germany) with 10% normal horse serum (Biochrom, Berlin, Germany). After dissection of the frontal pole of the hemispheres and the cerebellum, the brains were sectioned into 350 µm thick horizontal slices using a vibratome (Leica VT 1000S, Bensheim, Germany) in preparation medium. The slices were transferred into culture plate insert membrane dishes (Becton Dickinson, Franklin Lakes, NJ; pore size 0.4 µm) and subsequently transferred into 6-well culture dishes (Becton Dickinson) containing 1.2 ml culture medium (MEM–HBSS, 2:1, 25% normal horse serum, 2% L-glutamine, 2.64 mg/ml glucose, 100 U/ml penicillin, 0.1 mg/ml streptomycin, 10 µg/ml insulin–transferrin–sodium selenite supplement and 0.8 µg/ml vitamin C) (Stoppini et al., 1991). The slices were cultured in a humidified atmosphere at 35°C and 5% CO2. The medium was changed one day after preparation and every second day thereafter.
Organotypic glioma transplantation model
Aliquots of 5 x 103 CFSE-stained glioma cells or 1.4 x 103 DiI-stained human CD34+ haematopoietic PC in 0.1 µl medium were implanted into the entorhinal cortex (layers II and III) or the dentate gyrus 1 day after slice preparation. Glioma as well as PC were evaluated daily using an inverse fluorescence microscope (Olympus microscope, IX 70) equipped with a FITC (excitation filter 450–490 nm, band filter 520–550 nm) narrow band filter, a CCD camera (F-View II; Soft Imaging System) and respective image software (analySIS).
Collagen gel spheroid invasion assay
Glioma spheroids were generated by incubating 1.5 x 103 cells for 120 h in 96-well plates precoated with 20% Noble Agar (Difco Laboratories, Detroit, MI). Spheroids were embedded into a collagen matrix containing collagen type I (Vitrogen 100, Cohesion Technologies, CA), 10% FCS and 10% NaHCO3 in a 12-well plate. The sprouting of the spheroids was monitored by time-lapse video. To monitor the invasion of PC into the glioma spheroids, 105 DiI-stained 2 x 105 nu PC or 5 x 105 VM/Dk PC were embedded into the collagen matrix with or without a glioma spheroid. Invasion of the spheroids by DiI-stained PC was monitored by fluorescence microscopy.
RT–PCR
RNA was isolated using the RNeasy Mini kit (Qiagen, Hilden, Germany). First strand synthesis was performed using Superscript II reverse transcriptase (Gibco, Gaithersburg, MD). The primer sequences were as follows: CXCL12 forward GAGAATTCATGAACGCCAAGGTCGTGG (nucleotides 235–261), reverse GATCTAGATCACATCTTGAACCTCTTG (nucleotides 522–496), yielding a 287 bp product. The PCR conditions included predenaturation at 95°C for 5 min and 35 cycles consisting each of denaturation at 95°C for 45 s, annealing at 50°C for 45 s, synthesis at 72°C for 1 min, and a post-synthesis run of 72°C for 7 min. The PCR products were separated in 2% agarose gels and visualized by ethidium bromide staining. The ß-actin cDNA fragment was amplified as an internal control for equal amplification (Friese et al., 2004). A water control was run in each amplification to control for cross-contamination between tubes.
ELISA
Glioma cell supernatants generated in SFM for 48 h were concentrated with the Centriplus centrifugal filter device YM-3 (3000 Da cut-off; Millipore, Eschborn, Germany). CXCL12 and SCF levels were measured by ELISA (R&D Systems).
Flow cytometry
The levels of CXCR4 protein at the cell surface were quantified by flow cytometry using the specific monoclonal antibodies MAB173 or isotype IgG2B (R&D Systems). The cells were incubated with MAB173 or mouse IgG2B isotype control (10 µg/ml) antibodies for 1 h at 4°C. After three washes with ice-cold PBS containing bovine serum albumin (0.2%), the cells were stained with PE-labelled goat anti-mouse IgG (Sigma) for 30 min at 4°C and washed three times. Fluorescence intensity was measured using a Becton Dickinson FACSCalibur (Heidelberg, Germany). The specific fluorescence index was calculated by dividing the mean fluorescence intensity of cells stained with the specific antibody by the mean fluorescence intensity obtained with an isotype-matched control antibody.
Animal studies
Stereotactic implantation of 5 x 103 murine SMA-560 or 5 x 104 human LNT-229 glioma cells into the right striatum of 6- to 12-week-old VM/Dk mice (SMA-560) or athymic mice (CD1 nu/nu, Charles River) was carried out for tumour for matrix. After implantation it takes 5 (VM/Dk mice) or 7 days (athymic mice) to reliably generate a tumour. To resemble the clinical situation of larger tumour bulks, we let the tumours grow for 15 (VM/Dk mice) or 22 (athymic mice) days. PKH26-stained VM/DK (SMA-560 model) or CD34+ (LNT-229 model) PC (106) were injected into the tail vein on days 15 (SMA-560) or 22 (LNT-229). Neutralizing CXCL12 antibody (50 µg) or control antibody was injected intraperitoneally within 10 min. Mice were killed 48 h after injection of the PC. Cryostat sections were DAPI-stained. The number of PKH26-labelled PC in brain or tumour tissue was assessed on serial 8 µm sections by fluorescence microscopy by two independent observers (G.T., W.W.). CD34 expression was determined using a FITC-conjugated anti-human CD34 antibody (DAKO, Glostrup, Denmark; clone QBEnd 10, F7166). Immunohistochemistry on frozen sections was performed to determine GFAP (DAKO, clone 6F2, M0761, dilution 1:50) or MAP2 (Chemicon, Temecula, CA, AB5622, dilution 1:1000) staining. FITC-conjugated IgG was used as isotype-control for CD34-FITC-staining. Negative controls were performed without primary antibody.
The experiments were performed according to the German animal protection law.
Statistical analysis
Quantitative data were obtained for chemotaxis, ELISA, flow cytometry and the number of PC within the volume of the experimental glioma as indicated. Data are expressed as mean and SEM. Statistical significance was assessed by one-way ANOVA followed by Tukey's post hoc test (Excel, Microsoft, Seattle, WA). Interrelationships among PC migration and CXCL12 levels were determined using the Pearson product–moment correlation. All experiments reported here were performed at least three times in triplicate with similar results.
|
Results |
|---|
|
TopSummaryIntroductionMaterials and methodsResultsReferences |
|---|
Adult haematopoietic PC home into orthotopic experimental gliomas
To investigate the homing of adult haematopoietic PC to gliomas in vivo, we used G-CSF-mobilized peripheral blood human CD34+ cells (CD34+ PC) in a xenograft nude mouse paradigm. PKH26-stained CD34+ PC were injected intravenously on day 22 after the implantation of human LNT-229 glioma cells into the brains of nude mice. Two days after PC injection, the mice were killed and liquid nitrogen-frozen brains were cut into 10 µm sections. Immunofluorescence microscopy revealed CD34+ PC within the tumour on day 24 (Fig. 1A, Panels 1 and 2) but no cells within the normal homolateral or contralateral brain parenchyma (Fig. 1A, Panel 4). Similar results were obtained using Lin– stem and progenitor cells of immunocompetent VM/Dk mice (VM/Dk PC) administered to SMA-560 glioma-bearing VM/Dk mice (Fig. 1A, Panels 5–8). We further questioned whether the injected CD34+ PC still expressed CD34 or whether the PC that homed to the glial tumour expressed glial or neuronal markers. Figure 1B shows a strong CD34 signal on PKH26-positive PC (Panels 1 and 2) but not the astrocytic marker, glial fibrillary acidic protein (GFAP) (Panel 3) nor microtubule associated protein (MAP)2 (Panel 4) that is expressed on adult neurons. The slight positivity for MAP2 surrounding the PC is due to LNT-229 glioma cells that express this marker.
|
PC are responsive to chemotactic cues in vitro
VM/Dk and CD34+ PC were analysed in chemotaxis assays to examine the molecular mechanisms underlying this glioma tropism. These assays revealed increased migration of VM/Dk or CD34+ PC, but not LNT-229 glioma cells (data not shown), towards CXCL12, the major cytokine for SC homing to the bone marrow niche, and towards supernatant from human LNT-229 or SMA-560 cells. Heat-denatured LNT-229 or SMA-560 glioma supernatant did not enhance migration, suggesting that migration is stimulated by a heat-labile molecule, presumably a protein. In fact, the impact of LNT-229 and SMA-560 supernatant on migration was blocked by a neutralizing antibody to CXCL12 (Fig. 2A), suggesting that gliomas may use bone marrow cues to attract PC. RT–PCR confirmed that CXCL12 mRNA expression is common in human glioma cell lines including LNT-229 (Fig. 2B). We next correlated the attraction of glioma cell supernatants for PC with CXCL12 levels. Figure 2C shows that all glioma cell lines and two primary ex vivo glioblastoma cell cultures released CXCL12 protein in the supernatant and that all supernatant attracted CD34+ PC. T98G, T189 and T204 supernatant contained the highest levels of CXCL12 and induced the greatest increase of PC migration. There was a significant correlation between migration towards the supernatant and CXCL12 levels (r = 0.69, P < 0.05). SMA-560 cells contained 1350 pg/ml CXCL12 and enhanced CD34+ PC 3-fold as did T204 cells. Similar results were obtained with bone marrow-derived murine lin–Sca1+ haematopoietic stem and progenitor cells from athymic nude mice (nu PC) (data not shown). CXCL12 interacts with CXCR4 and is its only known ligand. Flow cytometry confirmed that CD34+ PC, VM/Dk PC and nu PC express CXCR4 (Fig. 2D). Neutralizing antibodies to CXCR4 did not affect basal migration, but blocked CXCL12-induced migration of human CD34+ PC in a concentration-dependent manner (10–40 µg/ml). Moreover, these antibodies also blocked the promigratory effect of T98G supernatant (Fig. 2E).
|
We next studied the molecular mechanism of PC attraction by glioma supernatant in more detail. MMP-9 contributes to the mobilization of haematopoietic PC from the bone marrow to the circulation (Heissig et al., 2002
). TGF-ß promotes MMP-9 expression by glioma cells (Wick et al., 2001). Consequently, we determined the possible modulation of glioma-directed chemotaxis of PC by TGF-ß.
RNA interference targeting TGF-ß reduces chemoattractivity of LNT-229 glioma cells for adult haematopoietic PC in organotypic culture. We confronted LNT-229 puro or siTGF-ß1/2 glioma cells (green) and CD34+ PC (red) on 7-day-old adult rat hippocampal slices in an organotypic brain environment (Fig. 3A) (Ullrich et al., 2001). Tumour and SC migration were monitored up to 4 days after implantation using fluorescence microscopy. PC migration did not occur at random, but was directed towards the glioma cells. Moreover, PC migration was reduced when LNT-229 cells were TGF-ß-depleted (Fig. 3B).
|
TGF-ß-dependent attraction of PC involves CXCL12, MMP-9 and SCF
To translate the results from the slice cultures into a three-dimensional paradigm using nu PC, the invasion of PC into glioma cell spheroids in a collagen matrix was examined (Fig. 4A). Nu PC were used in the following experiments instead of CD34+ PC because of the high numbers of cells needed. The distribution of PC in empty gels was diffuse (Panels 1 and 5), whereas LNT-229 spheroids strongly attracted the PC (Panels 2 and 6). In contrast, spheroids formed by LNT-229 siTGF-ß1/2 showed much less attraction for nu PC (Panels 3 and 7). Spheroids formed by SMA-560 cells were similarly chemoattractive for VM/Dk PC in this assay (data not shown). Furthermore, supernatant from siTGF-ß1/2-transfected LNT-229 cells displayed significantly less chemoattraction for nu PC than control supernatant. Reduced chemotaxis in siTGF-ß1/2 cells was overcome by exposure of the LNT-229 siTGF-ß1/2 cells to exogenous TGF-ß2 (2 ng/ml) for 7 days but not for 2 days. Previously, zymography had revealed reduced MMP-9 activity in the LNT-229 siTGF-ß1/2 supernatant (Friese et al., 2004
). Here, the reduction of PC migration was rescued by exogenous MMP-9. Moreover, HPC migration significantly decreases after treatment of supernatants from LNT-229 with neutralizing antibodies against TGF-ß for 2 days (Fig. 4B). However, neither the addition of neutralizing TGF-ß antibodies nor the addition of exogenous recombinant TGF-ß to SFM lead to any change in HPC migration as compared with SFM alone. These data suggest an indirect effect of TGF-ß in PC migration. In fact, there were also lower CXCL12 protein levels in cells treated with a neutralizing TGF-ß antibody or transfected with siTGF-ß1/2, and higher CXCL12 protein levels were observed in the LNT-229 siTGF-ß1/2 supernatant after exposure to exogenous TGF-ß for 7 days but not 2 days nor by exposure to MMP-9 (Fig. 4C).
|
Similarly, SCF levels were reduced in the supernatant of siTGF-ß1/2 cells, and the loss of SCF was reversed by 7 days exposure to TGF-ß2 (Fig. 4D), whereas TGF-ß2 had no effect on SCF levels in LNT-229 puro cells (data not shown). In contrast to CXCL12, SCF levels in LNT-229 siTGF-ß1/2 supernatant were also restored by MMP-9 (Fig. 4D).
The supernatant of LNT-229 puro and siTGF-ß1/2 cells enhanced PC migration when supplemented with CXCL12 (Fig. 5A). In contrast, MMP-9 (Fig. 4B) and SCF had a specific stimulating effect on migration induced by the supernatant of TGF-ß depleted but not puro control cells, consistent with a critical loss of MMP-9 and SCF resulting from TGF-ß depletion. Neutralizing antibodies to CXCL12 or SCF, or the specific MMP inhibitor RO28-2653, reduced migration stimulated by control or siTGF-ß1/2 supernatant. The combination of CXCL12 and SCF neutralization exerted no further effect (Fig. 5A).
|
We next ordered the interrelations between TGF-ß, CXCL12, SCF and MMP-9 in that process. Importantly, neither SCF nor MMP-9 retained their activity when CXCL12 was neutralized, indicating that CXCL12 is essential for the chemotaxis towards LNT-229 cell supernatant. Exogenous CXCL12 but not MMP-9 compensated for the neutralization of SCF, indicating that SCF is dispensable when CXCL12 levels are high. The effects of RO28-2653 were overcome by CXCL12 or SCF (Fig. 5B).
These data suggest that CXCL12 is absolutely essential for glioma-directed PC migration and that this pathway is positively modulated by MMP-9 and SCF. In fact, CXCL12 alone enhanced migration towards LNT-229 siTGF-ß1/2 and control supernatant, and SCF alone enhanced migration towards LNT-229 siTGF-ß1/2 but not control supernatant in a concentration-dependent manner, and their combination displayed synergy (Fig. 5C). Altogether, these data place (i) TGF-ß upstream of CXCL-12, MMP-9 and SCF, (ii) MMP-9 upstream of SCF and (iii) CXCL12 parallel to MMP-9 and SCF. Furthermore, CXCL12 is indispensable, whereas SCF needs CXCL12 to induce migration and in fact facilitates CXCL12-induced migration.
Homing of adult haematopoietic PC into orthotopic SMA-560 gliomas in vivo critically depends on CXCL12
Finally, we investigated the role of CXCL12 for PC homing in an intracranial syngeneic glioma model in vivo. VM/Dk PC (106) were administered intravenously on day 15 after SMA-560 glioma tumour inoculation into the right striatum. Injection of PC was followed by the intraperitoneal administration of neutralizing CXCL12 or control antibodies within 10 min. PC were found within the tumour but neither in the adjacent brain tissue nor elsewhere distant from the visible tumour mass. Co-treatment with anti-CXCL12 antibodies significantly reduced PC homing to the tumour (Fig. 6).
|
Discussion
The insufficient delivery of therapeutic molecules to deeply infiltrating glioma cells is the major reason for the failure of novel local treatment strategies for malignant gliomas. There is, therefore, a great need for the evaluation of versatile cellular vectors that specifically approach tumour cells to deliver therapeutic molecules. Recent studies have demonstrated the potential of PC, transformed neuronal precursors, neuronally differentiated bone marrow cells and adult neural SC to engraft the lesioned brain (Kim et al., 1998
; Aboody et al., 2000; Priller et al., 2001; Lee et al., 2003) The molecular mechanisms underlying this apparent attraction of the lesioned brain for PC and other SC have not been elucidated.
In the present study, we explored how a readily available PC type, adult haematopoietic PC, are attracted by glioma cells in vitro and in vivo. Haematopoietic PC reside in the adult bone marrow and are easily mobilized by G-CSF (Kimura et al., 2004). They possess the unique ability to self-renew and are an unlimited source of potential cellular vectors. We first confirmed the observation made with other SC types (Aboody et al., 2000; Lee et al., 2003) that haematopoietic PC administered systemically can be traced within intracerebral gliomas, but not within normal brain (Fig. 1A). Careful examination with immunohistochemistry revealed no expression of glial or neuronal markers but sustained expression of CD34 on the PC residing for 2 days in the LNT-229 gliomas (Fig. 1B). In contrast, haematopoietic PC in the regenerating chicken embryo spinal cord displayed expression of neuronal markers between days 4 and 9 after implantation. This neuronal differentiation was paralleled by a loss in CD34 expression (Sigurjonsson et al., 2005). Hence, neuronal or glial differentiation of CD34 PC might be influenced by the different microenvironment in the regenerating chicken embryo spinal cord compared with a human xenograft glioma, or the interval between administration and analysis. Others have also failed to show glial differentiation of bone marrow-derived cells (Wehner et al., 2003). Cell culture assays indicated that the migration of PC towards glioma supernatant in vitro involves CXCL12/CXCR4 interactions (Fig. 2), recapitulating the mechanisms mediating PC homing to the bone marrow, extravasation of mature leukocytes to sites of tissue damage and differentiation-dependent thymocyte migration in the thymus during T lymphopoiesis (Aiuti et al., 1997; Kim et al., 1998; Peled et al., 1999). CXCL12/CXCR4 interactions have also been proposed to stimulate glioma cell survival and proliferation (Zhou et al., 2002; Rubin et al., 2003). Although all glioma cells examined here express CXCL12 and CXCR4, CXCL12 was not chemoattractive for these glioma cell lines (Fig. 2B and C and data not shown). Thus the chemoattractive effect of CXCL12 acting on CXCR4 (Fig. 2E) may be limited to specific target cell populations. Apart from glioma and bone marrow stromal cells, there is constitutive (Rostasy et al., 2003) and disease-enhanced expression of CXCL12 in astrocytes and neurons. For instance, human immunodeficiency virus-infected astrocytes, and astrocytes and neurons of patients with Alzheimer's disease show increased CXCL12 release (Zhang et al., 2003). Hence, our experiments do not rule out a contribution of the lesion induced by the implantation of the glioma cells to the PC tropism. At the same time, the exclusive localization of the PC within the tumour as well as the in vitro data support a major role for PC attraction by the glioma cells.
TGF-ß has become a major target of immunotherapy for malignant glioma (Friese et al., 2004; Uhl et al., 2004). There is also an evolving role for TGF-ß in the migration and invasion of malignant gliomas (Wick et al., 2001). However, the local release of TGF-ß by glioma cells may also be required to attract PC in vivo. Since TGF-ß-depleted LNT-229 siRNA cell are non-tumourigenic in nude mice (Friese et al., 2004), we adopted the organotypical slice culture paradigm to study a role for TGF-ß for PC homing to gliomas. Here, PC attraction was reduced by RNA interference targeting TGF-ß. These assays allowed us to conclude that PC truly migrated preferentially towards tumour cells and appeared to rule out a preferential utilization of identical substrate pathways for migration as an alternative explanation for the co-localization of implanted tumour cells and PC (Fig. 3). Similarly, the supernatant of LNT-229 siTGF-ß1/2 cells showed a strong reduction in PC attraction (Fig. 4). Since the promigratory activity of this supernatant was not rescued by TGF-ß and since TGF-ß itself was not chemoattractive for PC (Fig. 4B), we postulated an indirect pathway leading from TGF-ß to PC migration. In fact, we obtained evidence to suggest that the TGF-ß-dependent synthesis and release of CXCL12 and MMP-9 mediate PC migration. This is because PC migration and CXCL12 are reduced in LNT-229 siTGF-ß1/2 cells and only long-term TGF-ß supplementation sufficient to restore CXCL12 and MMP-9 (Friese et al., 2004), or exogenous CXCL12 rescue the TGF-ß depletion (Fig. 4B–D and Fig. 5A). Furthermore, neutralization of CXCL12 abolished migration towards TGF-ß-containing control supernatant, placing CXCL12 downstream of TGF-ß. In addition, MMP-9 expression and activity were reduced by RNA interference targeting TGF-ß (Friese et al., 2004), and neutralization of MMP-9 by RO28-2653 reduced migration towards TGF-ß-containing and TGF-ß-depleted supernatant (Fig. 5A). The target of MMP-9, in turn, was probably SCF because MMP-9 augmented SCF release and migration of PC in LNT-229 siTGF-ß1/2 cells (Fig. 4B and D). Furthermore, in the presence of CXCL12, SCF synergistically enhanced nu PC migration towards supernatant of TGF-ß-depleted and control cells (Fig. 5C). The MMP-9-mediated release of s-KitL and binding to its receptor, CD117, contribute to the mobilization of haematopoietic PC from the bone marrow to the circulation (Hattori et al., 2002). In addition to our finding that SCF is involved in glioma-directed haematopoietic PC migration, SCF is important for the migration of neuronal PC to areas of freeze-damaged brain (Sun et al., 2004). Figure 7 summarizes that the TGF-ß-dependent release of CXCL12 and MMP-9/SCF are not sequential but parallel pathways because MMP-9 does not enhance the release of CXCL12 (Fig. 4C) and because MMP-9 and SCF are unable to overcome CXCL12 neutralization (Fig. 5B).
|
In summary, we demonstrate the central role of TGF-ß and CXCL12 in the glioma-induced and -directed migration of adult peripheral CD34+ human and adult bone marrow-derived murine haematopoietic PC. CXCL12 is the essential mediator of PC migration, and the CXCL12 attraction is enhanced by MMP-9-induced SCF cleavage. However, the use of these PC as cellular vectors for gene therapy in glioblastoma may not be successfully combined with therapeutic strategies targeting TGF-ß-mediated immunosuppression and motility.
|
Acknowledgements |
|---|
We thank Brigitte Frank, Ulrike Obermüller and Gabriele von Kürthy for excellent technical assistance and Hans-Willi Krell from Roche Diagnostics GmbH, Penzberg, Germany for the generous supply of RO-28-2653. This work was supported by the Landesstiftung Baden-Württemberg, State of Baden-Württemberg, Germany (P-LS-AS/HSPA7-12).
|
References |
|---|
|
TopSummaryIntroductionMaterials and methodsResultsReferences |
|---|
Aboody KS, Brown A, Rainov NG, Bower KA, Liu S, Yang W, et al. Neural stem cells display extensive tropism for pathology in adult brain: evidence from intracranial gliomas. Proc Natl Acad Sci USA 2000; 97: 12846–51.
Aiuti A, Webb IJ, Bleul C, Springer T, Gutierrez-Ramos JC. The chemokine SDF-1 is a chemoattractant for human hematopoetic progenitor cells and provides a new mechanism to explain the mobilization of CD34+ progenitors to peripheral blood. J Exp Med 1997; 185: 111–20.
Bähr O, Rieger J, Duffner F, Meyermann R, Weller M, Wick W. P-glycoprotein and multidrug resistance-associated protein mediate specific patterns of multidrug resistance in malignant glioma cell lines, but not in primary glioma cells. Brain Pathol 2003; 13: 482–94.[ISI][Medline]
Bautz F, Denzlinger C, Kanz L, Möhle R. Chemotaxis and transendothelial migration of CsD34+ haematopoietic progenitor cells induced by the inflammatory mediator leukotriene D4 are mediated by the 7-transmembrane receptor CysLT-1. Blood 2001; 97: 3433–40.
Benedetti S, Pirola B, Pollo B, Magrassi L, Bruzzone MG, Rigamonti D, et al. Gene therapy of experimental brain tumours using neural progenitor cells. Nat Med 2000; 6: 447–50.[CrossRef][ISI][Medline]
Bowman EP, Campbell JJ, Druey KM, Scheschonka A, Kehrl RR, Butcher EC. Regulation of chemotactic and proadhesive responses to chemoattractant receptors by RGS (regulator of G-protein signaling) family members. J Biol Chem 1998; 273: 28040–8.
DeAngelis LM. Brain tumours. N Engl J Med 2001; 344: 114–22.
Eyupoglu IY, Savaskan NE, Brauer AU, Nitsch R, Heimrich B. Identification of neuronal cell death in a model of degeneration in the hippocampus. Brain Res Brain Res Protoc 2003; 11: 1–8.[CrossRef][Medline]
Friese MA, Wischhusen J, Wick W, Weiler M, Eisele G, Steinle A, et al. RNA interference targeting transforming growth factor-beta enhances NKG2D-mediated antiglioma immune response and inhibits glioma cell migration and invasiveness, and abrogates tumourigenicity in vivo. Cancer Res 2004; 64: 7596–603.
Hattori K, Heissig B, Wu Y, Dias S, Tejada R, Ferris B, et al. Placental growth factor reconstitutes hematopoiesis by recruiting VEGFR1+ stem cells from bone-marrow microenvironment. Nature Med 2002; 8: 841–9.[CrossRef][ISI][Medline]
Heissig B, Hattori K, Dias S, Friedrich M, Ferris B, Hackett NR, et al. Recruitment of stem and progenitor cells from the bone marrow niche requires MMP-9 mediated release of kit-ligand. Cell 2002; 109: 625–37.[CrossRef][ISI][Medline]
Kim CH, Pelus LM, White JR, Broxmeyer HE. Differential chemotactic behavior of developing T cells in response to thymic chemokines. Blood 1998; 91: 4434–43.
Kimura T, Boehmler AM, Seitz G, Kuci S, Wiesner T, Brinkmann V, et al. The sphingosine 1-phosphate receptor agonist FTY720 supports CXCR4-dependent migration and bone marrow homing of human CD34+ progenitor cells. Blood 2004; 103: 4478–86.
Kondraganti S, Mohanam S, Chintala SK, Kin Y, Jasti SL, Nirmala C, et al. Selective suppression of matrix metalloproteinase-9 in human glioblastoma cells by antisense gene transfer impairs glioblastoma cell invasion. Cancer Res 2000; 60: 6851–5.
Krause DS, Theise ND, Collector MI, Henegariu O, Hwang S, Gardner R, et al. Multi-organ, multi-lineage engraftment by a single bone marrow-derived stem cell. Cell 2001; 105: 369–77.[CrossRef][ISI][Medline]
Lee J, Elkahloun AG, Messina SA, Ferrari N, Xi D, Smith CL, et al. Cellular and genetic characterization of human adult bone marrow-derived neural stem-like cells: a potential antiglioma cellular vector. Cancer Res 2003; 63: 8877–89.
Peled A, Petit I, Kollet O, Magid M, Ponomaryov T, Byk T, et al. Dependence of human stem cell engraftment and repopulation of NOD/SCID mice on CXCR-4. Science 1999; 283: 845–8.
Priller J, Flügel A, Wehner T, Boentert M, Haas CA, Prinz MF, et al. Targeting gene-modified hematopoietic cells to the central nervous system: use of green fluorescent protein uncovers microglial engraftment. Nature Med 2001; 7: 1356–61.[CrossRef][ISI][Medline]
Rostasy K, Egles C, Chauhan A, Kneissl M, Bahrani P, Yiannoutsos C, et al. SDF-1alpha is expressed in astrocytes and neurons in the AIDS dementia complex: an in vivo and in vitro study. J Neuropathol Exp Neurol 2003; 62: 617–6.[ISI][Medline]
Rubin JB, Kung AL, Klein RS, Chan JA, Sun Y, Schmidt K, et al. A small-molecule antagonist of CXCR4 inhibits intracranial growth of primary brain tumours. Proc Natl Acad Sci USA 2003; 100: 13513–8.
Sigurjonsson OE, Perreault M-C, Egeland T, Glover JC. Adult human hematopoietic stem cells produce neurons efficiently in the regenerating chicken embryo spinal cord. Proc Natl Acad Sci USA 2005; 102: 5227–32.
Stanulla M, Welte K, Hadam MR, Pietsch T. Coexpression of stem cell factor and its receptor c-Kit in human malignant glioma cell lines. Acta Neuropathol (Berl) 1995; 89: 158–65.[Medline]
Stoppini L, Buchs PA, Muller D. A simple method for organotypic cultures of nervous tissue. J Neurosci Methods 1991; 37: 173–82.[CrossRef][ISI][Medline]
Sun L, Lee J, Fine HA. Neuronally expressed stem cell factor induces neural stem cell migration to areas of brain injury. J Exp Med 2004; 113: 1364–74.
Uhl M, Aulwurm S, Wischhusen J, Weiler M, Almirez R, Mangadu R, et al. SD-208, a novel TGF-(receptor I antagonist, inhibits growth and invasiveness and enhances immunogenicity of murine and human glioma cells in vitro and in vivo. Cancer Res 2004; 64: 7954–61.
Ullrich O, Diestel A, Eyüpoglu IY, Nitsch R. Regulation of microglial expression of integrins by poly(ADP-ribose) polymerase-1. Nat Cell Biol 2001; 3: 1035–42.[CrossRef][ISI][Medline]
Wehner T, Bontert M, Eyupoglu I, Prass K, Prinz M, Klett FF, et al. Bone marrow-derived cells expressing green fluorescent protein under the control of the glial fibrillary acidic protein promoter do not differentiate into astrocytes in vitro and in vivo. J Neurosci 2003; 23: 5004–11.
Wick W, Grimmel C, Wild-Bode C, Platten M, Arpin M, Weller M. Ezrin-dependent promotion of glioma cell clonogenicity, motility, and invasion mediated by BCL-2 and transforming growth factor-beta 2. J Neurosci 2001; 21: 3360–8.
Wischhusen J, Naumann U, Ohgaki H, Rastinejad F, Weller M. CP-31398, a novel p53-stabilizing agent, induces p53-dependent and p53-independent glioma cell death. Oncogene 2003; 22: 8233–45.[CrossRef][ISI][Medline]
Wright DE, Bowman EP, Wagers AJ, Butcher EC, Weissman IL. Hematopoietic stem cells are uniquely selective in their migratory response to chemokines. J Exp Med 2002; 195: 1145–54.
Zhang K, McQuibban GA, Silva C, Butler GS, Johnston JB, Holden J, et al. HIV-induced metalloproteinase processing of the chemokine stromal derived factor-1 causes neurodegeneration. Nat Neurosci 2003; 6: 1064–71.[CrossRef][ISI][Medline]
Zhou Y, Larsen PH, Hao C, Yong VW. CXCR4 is a major chemokine receptor on glioma cells and mediates their survival. J Biol Chem 2002; 277: 49481–7.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  G. Tabatabai, C. Herrmann, G. von Kurthy, M. Mittelbronn, S. Grau, B. Frank, R. Mohle, M. Weller, and W. Wick VEGF-dependent induction of CD62E on endothelial cells mediates glioma tropism of adult haematopoietic progenitor cells Brain, August 9, 2008; (2008) awn182v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Vuaillat, G. Androdias, N. Davoust, and S. Nataf About multiple sclerosis, natalizumab, and CD34+ hematopoietic progenitors Blood, July 1, 2008; 112(1): 208 - 209. [Full Text] [PDF]
|
|
|
|
 |
|
 N. Davoust, C. Vuaillat, G. Cavillon, C. Domenget, E. Hatterer, A. Bernard, C. Dumontel, P. Jurdic, C. Malcus, C. Confavreux, et al. Bone marrow CD34+/B220+ progenitors target the inflamed brain and display in vitro differentiation potential toward microglia FASEB J, October 1, 2006; 20(12): 2081 - 2092. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Tabatabai, B. Frank, R. Mohle, M. Weller, and W. Wick Irradiation and hypoxia promote homing of haematopoietic progenitor cells towards gliomas by TGF-{beta}-dependent HIF-1{alpha}-mediated induction of CXCL12 Brain, September 1, 2006; 129(9): 2426 - 2435. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||